Crojfe 36(1) 2015 kb

Page 1

2015



Editorial

Journal Anniversary Dear foresters, dear readers, We have been working together for ten years now, under the new titles and new technical design. Forty years ago, the first issue of the Croatian scientific and professional journal entitled Mehanizacija šumarstva (Forest Engineering) was published. The journal dealt with a specific area of forestry – forest engineering. It played a major role especially in the period of the most intensive mechanization of forestry in the Republic of Croatia, in 1970s and 1980s of the last century, addressing primarily the main operations of timber harvesting: felling and processing, skidding/forwarding and long-distance transport, as well as planning, construction, and maintenance of forest roads. In the journal Mehanizacija šumarstva (Forest Engineering), new ideas from the area of forest engineering in our country, Europe and worldwide were published, the results of research of the Croatian forestry scientists and practitioners were presented, the results of development of new forest machines made by Croatian manufacturers were presented and promoted, and, in a way, the introduction of different machines, adapted to our stand and habitat conditions, as well as guidelines for conservation, sustainable and biodiversity forest management in different relief areas were encouraged, directed and publicly announced. Foresters, who worked in operational forestry, were not only our regular readers but also published their own papers in our journal. The first twenty years of publishing the journal Mehanizacija šumarstva (Forest Engineering) is probably also one of the best and most lasting examples of cooperation between forestry science and practice. In the field of forest engineering, forestry science was mostly concentrated in four chairs of the Faculty of Forestry, University of Zagreb (by merger of which the present Institute of Forest Engineering was established). At that time, forestry practice was almost entirely in the hands of the former forest management units throughout Croatia, i.e. at the end of that period in the hands of the public enterprise for the management of forests and forest land in the Republic of Croatia »Hrvatske šume« p.o. Zagreb (Croatian Forests Ltd. Zagreb). At the very end of the last century and at the beginning of the new millennium, partly due to objective and Croat. j. for. eng. 36(2015)1

valid reasons, and partly due to subjective and unacceptable reasons, problems started regarding regular publishing of the journal Mehanizacija šumarstva (Forest Engineering), which was published four times a year until that time. In 2004, the then technical editor of the journal Prof. Dubravko Horvat PhD. in collaboration with the management of the enterprise »Hrvatske šume« d.o.o. Zagreb (Croatian Forests Ltd. Zagreb), which was the publisher of the journal Mehanizacija šumarstva (Forest Engineering), together with the Faculty of Forestry, University of Zagreb, summoned the meeting with the aim of finding a solution to the problems raised. At the meeting, besides the conclusion that the delays in publishing the journal Mehanizacija šumarstva (Forest Engineering) should be solved as soon as possible, decision was made on starting two new journals: Croatian Journal of Forest Engineering (CROJFE) and Nova mehanizacija šumarstva – NMŠ (New Forest Engineering), both successors of the journal Mehanizacija šumarstva (Forest Engineering). The proposed members of the core editorial board of both journals were accepted with the decision to set up all other bodies of the journal as soon as possible. As a scientific journal, CROJFE was to be published twice a year in English, with a broad summary, subtitles and captions below tables and figures in Croatian, and Nova mehanizacija šumarstva (New Forest Engineering), as a scientific and professional journal, was to be published once a year in Croatian, with a summary, subtitles and captions below tables and figures in English. The core editorial board of both journals was mostly made of employees of the Department of Forest Techniques and Technology of the Faculty of Forestry, University of Zagreb. Operational continuity was provided by appointing Prof. Dubravko Horvat PhD. as the new technical editor, collaboration with the company »Hrvatske šume« d.o.o. Zagreb (Croatian Forests) was provided by appointing Željko Tomašić PhD. as the responsible editor, and fresh ideas, new objectives and positive energy was to be secured by appointing a young doctor of science as the editor-inchief and member of the technical board of the journal. On the basis of the vision presented by the core editorial board, short-term, middle-term and long-term program goals were defined. Very soon, other bodies

1


Editorial (1–9)

of the journal were established, and it should be pointed out that the Croatian and International editorial board were made of the leading Croatian and world scientists dealing with forest engineering in their scientific, teaching and professional activities. Although during ten years of continuous development and improvement of the journal CROJFE, there were many events that should be mentioned because they all resulted in a positive impulse and quantitative shift, in our opinion, the following events were the most important:    Two years after starting the journal under the new title, in early 2007, it was included in the bibliography database Web of Science, i.e. Science Citation Index Expanded. In accordance with the development strategy of the journal and its editorial policy, this was a significant achievement in reaching the final goal – to include the journal in Current Contents.    In mid 2008, a young editor was employed part time in the journal following a job announcement by the Ministry of Science, Education and Sport of the Republic of Croatia.    Since 2009 the journal has been subsidized by the Ministry of Science, Education and Sport of the Republic of Croatia.    In late 2010, the first two-year agreement was signed, and in late 2012 the second three-year agreement was signed on cooperation between the journal CROJFE and the international network Forestry Mechanization (FORMEC). Thus, FORMEC became the co-publisher of CROJFE. In early 2013, the Croatian Chamber of Forestry and Wood Technology Engineers became one of the publishers of the journal CROJFE. In late 2013, the automation system was implemented for the qualification of papers and the whole reviewing procedure. During the last ten years, a database of reviewers has been established and continually updated. Today it has a respectable volume and it consists of leading world scientists and experts working and operating in different fields of forest engineering. Regular and timely publishing, high level of technical equipment, big competition of papers qualified for publishing, quality and speed of reviewing procedure, high quality of published scientific papers of authors from different European and out-of-Europe countries are only some of the characteristics CROJFE can be proud of. From the starting volumes of the journal in which five to six papers were published in an issue, today we have come to the number of fifteen papers per an issue.

2

T. Pentek et al.

Due to a large number of qualified papers (about one hundred a year) and due to an exceptionally good cooperation with the FORMEC network, the core editorial board seriously considers the possibility of publishing annually three issues of the journal CROJFE. Unfortunately, today in the world, and especially in Croatia, many good or even excellent ideas cannot even be started, let alone implemented, because of the lack of finances. We were lucky because the company »Hrvatske šume« d.o.o. Zagreb (Croatian Forests Ltd. Zagreb) recognized from the very beginning our mutual project and they joined it as the publisher and the main financier. However, during our cooperation, not everything was always perfect, and the problems have particularly arisen in the last three years. In conditions of serious economic and financial crisis, frequent reassessing of purposefulness and justification of all projects is quite understandable, and this also applies to the journal CROJFE and Nova mehanizacija šumarstva (New Forest Engineering). It is, however, hard to understand the lack of support and continuous decrease of financial support to the projects of the highest quality in an international environment and according to international assessment criteria (the journal CROJFE is one of nine reference journals of IUFRO Division 3 »Forest Operations Engineering and Management«, it is listed in SCI bibliography database, it is the forestry journal in the Republic of Croatia with the highest impact factor of 0.526 for 2013, etc.). Still, we optimistically hope, together with our partners in this project (publishers and co-publishers), that high-quality projects, achieved as a result of many years of efforts and work, will someday be recognized and adequately valued. Plans, wishes and goals of the core editorial board and objectives of the core editorial board in the next ten-year period are synchronized with the Croatian and International editorial board, and supported by the members of the publishing council. They can be summarized as follows:    Develop and accept long-term (ten-year) and medium-term (five-year) journal development strategy.    Sign the new three-year cooperation agreement between the journal CROJFE and international FORMEC network (October 2015 to October 2018).    Include the journal CROJFE in Current Contents bibliography database.    Raise the impact factor to the level above 1.5 with an increase of minimum 0.10 a year, and enter and maintain the position in the first quarCroat. j. for. eng. 36(2015)1


T. Pentek et al.

tile (Q1) on the Web of Science list of forestry journals.    Increase the number of papers nominated for publishing to approximately 120 papers a year, by publishing 14 to 18 papers in each CROJFE issue.    Make more uniform the reviewing of papers and limit the reviewing time to maximum two months after the paper has been accepted.    Upgrade the automatic computer system for nominating and implementing the reviewing procedure of the papers.    Develop the web site of the journal.    Redesign the visual identity of the printed and digital issue of the journal CROJFE.    Manage more efficiently the journal CROJFE by greater efforts of the professional young editor and the whole editorial team.    Achieve financial stability of the journal as the guarantee of its viability andindependence by attracting a higher number of active co-publishers, donators and sponsors.    Extend the list of reviewers with most qualified scientists dealing with forest engineering, and interested in reviewing.    Extend the International Editorial Board with recognized scientists in the area of forest engineering who are willing to help achieving the set goals.    Position CROJFE as one of the leading journals of forest engineering in the world. To mark the anniversary – ten (forty) years of publishing the journals Croatian Journal of Forest Engineering and Nova mehanizacija šumarstva (New Forest Engineering) – international scientific conference »Forest engineering – current situation and future challenges«

Croat. j. for. eng. 36(2015)1

Editorial (1–9)

will be held in Zagreb and in Zalesina from March 18 to 20, 2015. The organizer of the three-day scientific conference is the Faculty of Forestry of the University in Zagreb, Department of Forest Engineering, and coorganizers are IUFRO, FORMEC, Croatian Chamber of Forestry and Wood Technology Engineers, Croatian Forestry Society and Academy of Forestry Science. The attendance of 250 participants from Croatia and abroad is expected. Thirteen papers (by invitation) from the area of forest engineering will be presented, and the invited lecturers are Croatian and international scientists and experts dealing with roundwood harvesting, production of forest biomass, construction of forest roads and mechanization of forest operations. The basic data on the Conference are presented below, and for more information, please visit our web site http:// www.crojfe2015.com. I would like to thank all of you who have contributed in any way to the results achieved by the journal CROJFE in its ten-year (forty-year) history, and of which we, the editors of the journal CROJFE, are immensely proud. A great thank to publishers and co-publisher of the journal, to all members of the journal bodies: Publishing Council, Editorial Board and International Editorial Board, all authors and co-authors of the papers nominated for publishing, all reviewers, and also thanks to donators, sponsors and supporters, thanks to all present and future readers. And finally, dear foresters, dear friends, we invite you to cooperate and to join actively in the work and life of one of the rare scientific journals strictly specialized in the area of forest engineering – the journal CROJFE. Tibor Pentek, Tomislav Poršinsky, Željko Tomašić, Mario Šporčić, Ivica Papa

3


Editorial (1–9)

T. Pentek et al.

International scientific conference CROJFE 2015 – basic data Međunarodno znanstveno savjetovanje CROJFE 2015. – osnovni podaci Title: »Forest Engineering – Current Situation and Future Challenges«, Zagreb, Zalesina, 18th – 20th of March 2015 Naziv: »Šumarsko inženjerstvo – sadašnje stanje i budući izazovi«, Zagreb, Zalesina, 18. – 20. ožujka 2015. Organizational Committee (in alphabetical order) Organizacijski odbor savjetovanja (abecednim redom)

Scientific Committee (in alphabetical order) Znanstveni odbor savjetovanja (abecednim redom)

Damir Delač, Croatia Danko Diminić, Croatia Dubravko Horvat, Croatia Hrvoje Nevečerel, Croatia Tibor Pentek, Croatia – President (predsjednik) Tomislav Poršinsky, Croatia – Vice President (dopredsjednik) Mario Šporčić, Croatia Marijan Šušnjar, Croatia Željko Tomašić, Croatia Silvija Zec, Croatia

Raffaele Cavalli, Italy Woodam Chung, USA Dubravko Horvat, Croatia Ola Lindroos, Sweden Tibor Pentek, Croatia Tomislav Poršinsky, Croatia Igor Potočnik, Slovenia Raffaele Spinelli, Italy Karl Stampfer, Austria Jori Uusitalo, Finland

4

Croat. j. for. eng. 36(2015)1


T. Pentek et al.

Editorial (1–9)

Honorary Committee

Conference Sectretariat

Počasni odbor savjetovanja

Tajništvo savjetovanja

Vladimir Jambreković, Dean of Faculty of Forestry, University of Zagreb (dekan Šumarskog fakulteta Sveučilišta u Zagrebu)

Andreja Đuka, e-mail: aduka@sumfak.hr, General Secretary (glavni tajnik savjetovanja)

Woodam Chung, IUFRO division 3 President (pred­sjednik IUFRO grupe 3) Karl Stampfer, FORMEC Network President (predsjednik međunarodne organizacije FORMEC) Damir Felak, President of Croatian Chamber of Forestry and Wood Technology Engineers (predsjednik Hrvatske komore inženjera šumarstva i drvne tehnologije) Oliver Vlainić, President of Croatian Forestry Society (predsjednik Hrvatskog šumarskog društva) Igor Anić, President of Croatian Academy of Forestry Sciences (predsjednik Akademije šumarskih znanosti)

Ivica Papa, e-mail: papa@sumfak.hr, Vice General secretary (zamjenik glavnog tajnika savjetovanja) Zdravko Pandur Dinko Vusić Kruno Lepoglavec Matija Landekić Marko Zorić Matija Bakarić Irena Šapić Jelena Kranjec

Zdenko Bogović, President of Croatian Union of Private Forest Owners' Associations (predsjednik Hrvatskog saveza udruga privatnih šumovlasnika)

Speakers – Izlagači referata Session 1 – Key note presentations (Sesija 1 – Uvodni referati) Moderators/Moderatori: Igor Potočnik/Raffaele Cavalli 1.   Tibor Pentek, Tomislav Poršinsky, Željko Tomašić, Mario Šporčić, Ivica Papa (Croatia) – Croatian Journal of Forest Engineering (CROJFE) – 40 Year History and Future Prospects (CROJFE – 40 godina povjesti časopisa sa pogledom u budućnost) 2.   Woodam Chung (USA) – IUFRO Division 3 »Forest operations engineering and management« – challenges in the Future (IUFRO Divizija 3 (Šumarski postupci, inženjerstvo i menadžment) – budući izazovi) 3.   Karl Stampfer (Austria) – Harvesting Systems – State-of-Art (Suvremeni sustavi pridobivanja drva) Session 2 – Harvesting Operations and Timber Transport (Sesija 2 – Pridobivanje i transport drva) Moderators/Moderatori: Woodam Chung/Antti Asikainen 1.   Raffaele Cavalli (Italy) – Harvesting Operations in Steep Terrain (Postupci pridobivanja drva na strmim terenima) 2.   Ola Lindroos (Sweden) – Timber Procurement – Swedish Experience (Lanac dobave drva – švedska iskustva) 3.   Igor Potočnik*, Tibor Pentek**, Boštjan Hribernik*, Hrvoje Nevečerel** (*Slovenia and **Croatia) – Maintenance of Forest Roads – the Need for Sustainable Forest Management (Održavanje šumskih cesta – potreba za održivim gospodarenjem šuma) 4.   Jori Uusitalo (Finland) – Key Components of Terrain Trafficability (Glavne značajke prometnosti terena) 5.   Radomír Klvač (Czech Republic) – Daily Use of Life Cycle Assessment (Analiza životnog ciklusa u svakodnevnoj upotrebi)

Croat. j. for. eng. 36(2015)1

5


Editorial (1–9)

T. Pentek et al.

Session 3 – Forest Biomass and Forest Machines (Sesija 3 – Šumska biomasa i šumski strojevi) Moderators/Moderatori: Karl Stampfer/Ola Lindroos 1.   Raffaele Spinelli (Italy) – Forest Biomass Supply Chains: European Experiences from the INFRES Project (Lanac dobave šumske biomase: europska iskustva u okviru INFRES projekta) 2.   Antti Asikainen (Finland) – From Cost Reduction to Value Creation in Biomass Supply Chains (Od smanjenja troškova do povećanja dobiti u lancu dobave šumske biomase) 3.   Gernot Erber (Austria) – Fuel Wood Drying Modelling – the European Experience (Modeli sušenja energijskog drva – europska iskustva) 4.   Jurij Marenče (Slovenia) – Harvesting Operations in Small Scale Forests (Postupci pridobivanja drva u šumama malih površina) 5.   Marijan Šušnjar, Dubravko Horvat, Zdravko Pandur, Marko Zorić (Croatia) – Development of Forest Machines – New Trends (Razvoj šumarskih vozila – noviteti) 6.   Dinko Vusić, Andreja Đuka (Croatia) – Repositioning of Forest Biomass as a Renewable Source of Energy – Implications to Harvesting Operations in Croatia (Značaj šumske biomase kao obnovljivog izvora energije – utjecaj na sustave pridobivanja drva u Hrvatskoj)

CONFERENCE PROGRAME – PROGRAM SAVJETOVANJA Day 1 – 1. dan (Zagreb) 19:00 –

Meeting of the Editors Office and International Editorial Board of CROJFE and NMŠ journals Sastanak uredništva i međunarodnog uredničkog odbora časopisa CROJFE i NMŠ

Day 2 – 2. dan (Zagreb) 7:30 – 9:00 9:00 – 11:00

Registration of participants – Registracija sudionika Opening ceremony, welcome speeches, awards ceremony, paper presentations (session 1) Otvaranje savjetovanja, pozdravni govori, dodjela nagrada, prezentacije referata (sekcija 1)

11:00 – 11:30

Coffee break – Pauza za kavu

11:30 – 13:15

Paper presentations (session 2) – Prezentacije referata (sekcija 2)

13:15 – 15:00

Lunch – Ručak

15:00 – 17:00

Paper presentations (session 3) – Prezentacije referata (sekcija 3)

17:30 – 19:00

Trip to TRFC Zalesina – Putovanje na NPŠO Zalesina

19:00 – 19:45

Accommodation of participants – Smještaj sudionika

20:15 –

Dinner – Večera

Day 3 – 3. dan (Zalesina) 8:00 – 9:30

Breakfast – Doručak

9:30 – 15:00

Field trip – Terenska ekskurzija

15:00 – 17:00

Lunch – Ručak

17:00

Departure of participants – Odlazak sudionika

6

Croat. j. for. eng. 36(2015)1


T. Pentek et al.

Editorial (1–9)

Obljetnica časopisa Cijenjene kolegice i kolege šumari, drage čitateljice i čitatelji! Prošlo je punih deset godina otkako se, pod novim naslovima i u novom tehničkom ruhu, zajedno družimo. Prije punih četrdeset godina iz tiska je pod naslovom Mehanizacija šumarstva izišao prvi broj hrvatskoga znanstveno-stručnoga časopisa koji se bavio jednim užim područjem šumarstva – šumarskim inženjerstvom. Časopis je, poglavito u razdoblju najintenzivnijega mehaniziranja šumskih radova u Republici Hrvatskoj, sedamdesetih i osamdesetih godina prošloga stoljeća, u prvom redu u području svih sastavnica pridobivanja drva: sječe i izradbe, privlačenja i daljinskoga prijevoza, ali i otvaranja šuma šumskim prometnicama, odigrao vrlo važnu i nezamjenjivu ulogu. U Mehanizaciji su šumarstva objavljivane nove spoznaje iz šumarskoga inženjerstva u nas, u Europi, ali i u svijetu, predstavljani su rezultati istraživanja domaćih šumarskih znanstvenika i šumara praktičara, prikazivani su i promovirani rezultati razvoja novih šumarskih strojeva hrvatskih proizvođača te je, na neki način, poticano, usmjeravano i javno praćeno uvođenje različitih mehaniziranih sredstava, prilagođenih našim sastojinskim i stanišnim uvjetima, ali i smjernicama potrajnoga, održivoga i bioraznolikoga gospodarenja šumama u različitim reljefnih područjima. Šumari koji su radili u operativnom šumarstvu vrlo su rado ne samo čitali napisano već i objavljivali vlastite radove u našem časopisu. Prvih je dvadesetak godina tiskanja časopisa Mehanizacija šumarstva vjerojatno i jedan od najboljih i najdugotrajnijih primjera suradnje šumarske znanosti i šumarske operative, šumarske znanosti koja je u području šumarskoga inženjerstva uglavnom bila koncentrirana u četiri katedre Šumarskoga fakulteta Sveučilišta u Zagrebu (spajanjem kojih je nastao današnji Zavod za šumarske tehnike i tehnologije), te šumarske operative koja se tada, gotovo u potpunosti, nalazila u nekadašnjim šumskim gospodarstvima diljem Hrvatske, odnosno pri kraju toga razdoblja u Javnom poduzeću za gospodarenje šumama i šumskim zemljištima u Republici Hrvatskoj »Hrvatske šume« p.o. Zagreb. Na samom kraju prošloga i početkom ovoga tisućljeća, što zbog objektivnih i razumljivih, a što zbog subjektivnih i teško prihvatljivih razloga, Croat. j. for. eng. 36(2015)1

započinju problemi oko redovitosti izlaženja Mehanizacije šumarstva koja je dotada izlazila četiri puta godišnje. Tadašnji tehnički urednik časopisa prof. dr. sc. Dubravko Horvat, u suradnji s upravom poduzeća »Hrvatske šume« d.o.o. Zagreb koje je, uz Šumarski fakultet Sveučilišta u Zagrebu, izdavač Mehanizacije šumarstva, pokreće 2004. godine održavanje sastanka radi rješavanja nastalih problema. Na sastanku je, uz zaključak o obvezi što skorijega rješavanja zaostataka u tiskanju časopisa Mehanizacije šumarstva, donesena odluka o pokretanju dvaju novih časopisa: Croatian Journal of Forest Engineering (CROJFE) i Nova mehanizacija šumarstva (NMŠ), oba sljednika časopisa Mehanizacija šumarstva. Prihvaćen je i prijedlog sastava novoga užega uredništva obaju časopisa uz suglasnost da se u najkraćem mogućem roku ustroje sva ostala tijela časopisa. CROJFE je, kao znanstveni časopis, trebao izlaziti dva puta godišnje na engleskom jeziku uz širi sažetak, podnaslove i potpise ispod tablica i slika na hrvatskom jeziku, a NMŠ je, kao znanstveno-stručni časopis, trebao biti tiskan jednom godišnje na hrvatskom jeziku uz sažetak, podnaslove i potpise ispod tablica i slika na engleskom jeziku. U uže su uredništvo obaju časopisa ušli većinom djelatnici Zavoda za šumarske tehnike i tehnologije Šumarskoga fakulteta u Zagrebu; kontinuitet je rada osiguran imenovanjem prof. dr. sc. Dubravka Horvata novim tehničkim urednikom, nastavak suradnje s »Hrvatskim šumama« d.o.o. Zagreb imenovanjem dr. sc. Željka Tomašića odgovornim urednikom, a svježe ideje, novi ciljevi i pozitivna energija trebali su doći imenovanjem mladih doktora znanosti za glavnoga urednika i u tehničko uredništvo časopisa. Na temelju vizije koju je uže uredništvo predstavilo, definirani su kratkoročni, srednjoročni i dugoročni programski ciljevi. Uskoro su ustrojena i ostala tijela časopisa, a posebno treba naglasiti sastav domaćega i međunarodnoga uredničkoga odbora u koje su ušli vodeći hrvatski i svjetski znanstvenici koji se u svom znanstvenom, nastavnom ili stručnom radu bave šumarskim inženjerstvom. Iako je tijekom desetogodišnjega stalnoga unaprjeđivanja i poboljšanja časopisa CROJFE bilo jako puno događaja koje bi valjalo navesti i koji su, svaki na svoj način, donijeli svojevrsni pozitivni im-

7


Editorial (1–9)

puls i kvantitativni pomak, izdvojili smo, po našem mišljenju, najvažnija događanja:    Dvije godine od početka izlaženja pod novim naslovom, od početka 2007. godine, časopis je uključen u bibliografsku bazu Web of Science, odnosno Science Citation Index Expanded. To je, sukladno strategiji razvoja časopisa i njegovoj izdavačkoj politici, bio međucilj na putu ka dosezanju konačnoga cilja – ulaska časopisa u bibliografsku bazu Current Contents.    Sredinom 2008. godine, na javnom natječaju koji je tada objavilo Ministarstvo znanosti, obrazo­va­ nja i športa Republike Hrvatske, na časopisu je, u trajanju od pola radnoga vremena, zaposlen mladi urednik.    Od 2009. godine pa do danas časopis se nalazi u sustavu potpore Ministarstva znanosti, obrazo­ vanja i sporta Republike Hrvatske.    Krajem 2010. godine potpisan je prvi dvogodišnji, a krajem 2012. godine drugi trogodišnji ugovor o suradnji časopisa CROJFE i međunarodne mre­že Forestry Mechanization (FORMEC). Time je FORMEC postao suizdavač CROJFE-a.    Početkom 2013. godine jedan od izdavača ča­so­ pi­sa CROJFE-a postaje i Hrvatska komora in­že­ njera šumarstva i drvne tehnologije (HKIŠDT).    Krajem 2013. godine u rad je pušten automatizi­ ra­­ni sustav kandidiranja radova i provedbe ci­je­ lo­ga recenzijskoga postupka. Tijekom protekloga desetljeća uspostavljena je i kontinuirano nadopunjavana baza recenzenata koja je danas dostigla respektabilan opseg i u kojoj se nalaze vodeći svjetski znanstvenici odnosno stručnjaci koji rade i djeluju u području različitih sastavnica šumarskoga inženjerstva. Redovito i pravodobno tiskanje, visoka razina tehničke opremljenosti, velika konkurencija kandidiranih radova za objavu, kakvoća i brzina provedbe recenzijskoga postupka, visokokvalitetni objavljeni znanstveni radovi autora iz različitih europskih i izvaneuropskih zemalja samo su neke od današnjih značajki kojima se CROJFE može pohvaliti. Od početnih volumena časopisa u kojima je po pojedinom broju tiskano pet-šest radova, danas smo došli do brojke od petnaestak radova po pojedinom broju. Zbog velikoga broja kandidiranih radova godišnje (brojka se kreće oko stotinu) te zbog izuzetno dobre suradnje s mrežom FORMEC uže uredništvo ozbiljno razmatra mogućnost tiskanja triju brojeva časopisa CROJFE godišnje. Kako to danas u svijetu, a poglavito u Hrvatskoj biva, mnoge dobre ili čak odlične ideje nisu ni započete, a kamoli ostvarene zbog nedostatka financijskih sred-

8

T. Pentek et al.

stava. Imali smo tu sreću što su »Hrvatske šume« d.o.o. Zagreb od samoga početka prepoznale naš zajednički projekt te su se u njega uključile kao izdavač i glavni financijer. No nije uvijek sve tijekom naše zajedničke suradnje bilo idealno, a problemi su posebice narasli u posljednje tri godine. Razumljivo je, u uvjetima teške gospodarske i financijske krize, često preispitivanje svrhovitosti i opravdanosti svih projekata, pa tako i časopisa CROJFE i časopisa NMŠ. No nije razumljivo nepodržavanje i stalno smanjivanje financijske podrške onima, u međunarodnom okruženju i prema međunarodnim kriterijima procjene, najkvalitetnijim projektima (časopis CROJFE je jedan od devet referentnih časopisa IUFRO divizije 3 »Forest Operations Engineering and Management«, nalazi se u SCI bibliografskoj bazi, šumarski je časopis u Republici Hrvatskoj s najvećim faktorom odjeka od oko 0,526 za 2013. godinu itd.). Ipak, s optimizmom se nadamo te zajedno s našim ostalim partnerima na ovom zajedničkom projektu (izdavači i suizdavači) gledamo unaprijed, jer kvalitetni projekti, u koje je uložen golem mnogogodišnji trud i rad, kad-tad budu prepoznati i dostojno vrednovani. Planovi, želje i ciljevi užega uredništva u idućem desetogodišnjem razdoblju usklađeni s domaćim i međunarodnim uredničkim odborom, a koje su podržali članovi izdavačkoga vijeća, mogu se sažeti u nekoliko točaka:    Izrada i prihvaćanje dugoročne (desetogodišnje) i srednjoročne (petogodišnje) strategije razvoja ča­sopisa.    Potpisivanje novoga trogodišnjega ugovora o su­ radnji časopisa CROJFE s međunarodnom mrežom FORMEC (listopad 2015. – listopad 2018. godine).    Uvrštavanje časopisa CROJFE u bibliografsku bazu Current Contents.    Impact factor podići na razinu iznad 1,5 povećanjem od najmanje 0,10 godišnje te ulazak i zadržavanje u prvom (Q1) kvartilu na Web of Science listi časopisa iz područja šumarstva.    Povećanje broja nominiranih radova za objavu na oko 120 radova godišnje, uz objavu 14 – 18 radova po pojedinom broju CROJFE-a.    Skraćivanje i ujednačivanje roka recenzije radova na najviše dva mjeseca od datuma primitka rada.    Poboljšanje funkcioniranja automatskoga računalnoga sustava za nominiranje i provedbu recenzijskoga postupka radova.    Unaprjeđenje mrežne stranice časopisa. Croat. j. for. eng. 36(2015)1


T. Pentek et al.

   Redizajniranje vizualnoga identiteta tiskanoga i digitalnoga izdanja časopisa CROJFE.    Učinkovitije upravljanje časopisom CROJFE povećanim angažmanom profesionalnoga mladoga urednika i čitavoga uredničkoga tima.    Postizanje financijske stabilnosti časopisa kao jamca njegove održivosti i neovisnosti privlačenjem većega broja aktivnih suizdavača, donatora i sponzora.    Proširenje liste recenzenata najkvalitetnijim znanstvenicima koji se bave šumarskim inženjerstvom, a zainteresirani su za recenzijski posao.    Proširenje Međunarodnoga uredničkoga odbora priznatim znanstvenicima u području šumarskoga inženjerstva koji su voljni pomoći dostizanju navedenih ciljeva.    Pozicioniranje CROJFE-a kao jednoga od vodećih časopisa šumarskoga inženjerstva u svijetu. Povodom obilježavanja ove vrijedne obljetnice – desete (četrdesete) godine tiskanja časopisâ Croatian Journal of Forest Engineering i Nova mehanizacija šumarstva – u Zagrebu i u Zalesini će se od 18. do 20. ožujka 2015. godine održati međunarodno znanstveno savjetovanje »Forest engineering – current situation and future challenges« (»Šumarsko inženjerstvo – sadašnje stanje i budući izazovi«). Organizator je trodnevnoga znanstvenoga savjetovanja Šumarski fakultet Sveučilišta u Zagrebu, Zavod za šumarske tehnike i tehnologije, a suorganizatori su IUFRO, FORMEC, Hrvatska komora inženjera šumarstva i drvne tehnologije, Hrvatsko

Croat. j. for. eng. 36(2015)1

Editorial (1–9)

šumarsko društvo i Akademija šumarskih znanosti. Očekuje se nazočnost oko 250 sudionika iz Hrvatske i iz inozemstva. Bit će održano trinaest pozivnih referata iz područja šumarskoga inženjerstva, a pozvani predavači vodeći su domaći i međunarodni znanstvenici i stručnjaci koji se bave problematikom pridobivanja obloga drva, pridobivanja šumske biomase, otvaranja šuma i mehaniziranja šumskih radova. Osnovne podatke o Savjetovanju možete pronaći u nastavku, a za više informacija molimo posjetite mrežnu stranicu http://www.crojfe2015.com. Hvala svima vama koji ste na bilo koji način pridonijeli rezultatima koje je časopis CROJFE dostigao u svojoj desetogodišnjoj (četrdesetogodišnjoj) povijesti, a na koje smo mi, urednici časopisa CROJFE, neizmjerno ponosni. Veliko hvala izdavačima i suizdavačima časopisa, svim članovima tijela časopisa: Izdavačkoga vijeća, Uredničkoga odbora i Međunarodnoga uredničkoga odbora, svim autorima i suautorima radova nominiranih za objavu, svim recenzentima radova, hvala donatorima, sponzorima i podupirateljima, hvala svim čitateljicama i čitateljima i onima koji će čitateljima tek postati. I na kraju vas, cijenjene kolegice i kolege šumari, dragi prijatelji, pozivamo na suradnju, pozivamo vas da se aktivno uključite u rad i život jednoga od rijetkih znanstvenih časopisa usko specijaliziranih za područje šumarskoga inženjerstva – časopisa CROJFE. Tibor Pentek, Tomislav Poršinsky, Željko Tomašić, Mario Šporčić, Ivica Papa

9



Original scientific paper

Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution for Open Drum Chippers Lars Eliasson, Henrik von Hofsten, Tomas Johannesson, Raffaele Spinelli, Tomas Thierfelder Abstract Chip size distribution is an important quality variable not only for buyers of forest fuels, but also for chipping contractors as it influences both fuel consumption and productivity of chippers. Studies of disc chippers and of drum chippers with closed drums have shown that increased chip target length increases chipper productivity and decreases fuel consumption per ton of chips produced. For open drum chippers, chip length is partly controlled by the mesh size in the sieve. In order to evaluate how this sieve affects productivity and fuel consumption of chippers, two open drum machines for professional chipping of forest fuels were studied. Small chippers were represented by a Kesla 645, and larger ones by an Eschlbรถck Biber 92. The Kesla 645 was studied with 25, 50, and 100 mm sieves and the Biber 92 with 35, 50, and 100 mm sieves. With the 100 mm sieve the Kesla chipper produced 14.5 oven dry ton (odt) of chips per effective hour and the Biber 30.0 odt per effective hour. Fuel consumption per odt was 3.0 l for the Kesla and 2.1 l for the Biber. A reduction of sieve mesh size decreased productivity and increased fuel consumption for both machines. Reducing the mesh size decreased the size of produced chips for the Kesla, but not for the Biber. The sieve on the Biber seems to be a safety measure against oversized pieces whereas chip size is, as on a closed drum chipper, mainly controlled by the cut length of the knives. Keywords: biomass, forest fuel, chip quality

1. Introduction In Sweden, 90% of logging residue biomass is chipped on or adjacent to the landing (Brunberg 2013) in order to reduce road transport costs. Terrain chipping, i.e. chipping of small piles on the cut, is not used anymore as it is too expensive (Eliasson 2011). Mainly truck mounted and forwarder mounted chippers are used for chipping of forest biomass that is stored in piles on landings or on the side of the road. If the material is stored some distance from the road, e.g. in a large pile on the cut, forwarder mounted chippers are the preferred choice. In central Europe, one of the dominating chipper types for both these conditions are farm tractor towed machines powered by the tractor power take off (PTO). The advantages with a towed chipper is that they are faster and easier to move between setups than forwarder mounted chippers as Croat. j. for. eng. 36(2015)1

they do not require flatbed trailers for the relocation and that it is possible to utilise the farm tractor for other purposes while there is no chipping work. On the other hand they are less mobile off-road, and are not able to transport the chips to/on the landing by themselves like a forwarder mounted chipper. Forwarder mounted chippers that are equipped with a chip bin, e.g. Erjo 9/93 and Bruks 806STC, usually transport the chips to a reloading spot where the chips either are dumped in containers or on a tarpaulin on the ground (cf. Eliasson et al. 2011, Lombardini et al. 2013). Towed chippers, and some forwarder mounted chippers, usually chip directly into containers (cf. Eliasson et al. 2011, Eliasson et al. 2013, Grรถnlund and Eliasson 2013) or trucks. In the former case the contractor often have a second tractor or a forwarder equipped with a hook loader to shunt the containers to a suitable reloading place.

11


L. Eliasson et al.

Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution ... (11–17)

An increased target length for the produced chips has proven to increase chipper productivity, as well as reduce the fuel consumption per produced oven dry tonne of chips both for disc chippers (Eliasson et al. 2012, Facello et al. 2013) and drum chippers with a closed drum (Johannesson et al. 2012, Spinelli and Magagnotti 2012). For both these chipper types, target length is mainly a function of the distance between the knife edge and the drum or disc surface. For drum chippers with an open drum, it is a bit more complex to control the target chip length, i.e. to control the chip size distribution, as it is influenced by the feeding speed, the amount of self-feed, distance between the knife edge and the imaginary drum surface (were there can be a stopping device to prevent overfeeding), and the mesh size in the bottom sieve that acts as a barrier to stop oversized chips leaving the drum casing. Contractors operating open drum chippers claim that by changing feeding speed and sieve they can produce chips according to the customers’ preferred chip size distribution. There are studies of open drum chippers that show that an increased sieve mesh size increases chipper productivity and reduces fuel consumption (Nati et al. 2010, Röser et al. 2012). The aim of the study was to infer the effects of sieve mesh size on chipper performance and fuel consumption and on chip size distribution for the produced chips. In order to do this, two open drum machines for professional chipping of forest fuels were studied when chipping tree sections. Both machines were powered by farm tractors. The small chippers were represented by a Kesla 645, and the larger ones by an Eschlböck Biber 92.

2. Material and methods The study was carried out on June 1 and 2 adjacent to Åre Östersund airport in northern Sweden (63°12’9.3“N 14°28’51.8“E). Two open drum chippers owned by the same contractor were studied, a Kesla 645 powered by a 270 kW John Deere 8345R farm tractor and a Eschlböck Biber 92 powered by a 358 kW Claas Xerion 5000 farm tractor. The contractor operates a chipper together with two farm tractors, where each tractor pulls two 42.5 m3 chip trailers. During chipping, the chips are blown directly into the trailers and when both trailers are filled, the tractors travel to the customer. The Kesla 645 chipper has 6 angled blades that are positioned in a spiral around the drum and the Eschlböck Biber 92 was used with 10 knives positioned on 4 positions around the drum. The cut length for the Bieber was 24 mm and approximately 25 mm for the

12

Kesla. The chip extraction is done in a similar way for the two chippers. In both cases there are augers beneath the open drum that feed the chips to a fan that throw the chips out through the chip tube. A square mesh sieve is placed between the drum and the auger to avoid that oversized chips leave the drum casing. Both chippers were studied with 3 different sieves; Coarse (100 mm mesh size), medium (50 mm mesh size), and fine (25 mm mesh size for the Kesla 645, and 35 mm mesh size for the Biber 92 chipper). The reason for the different mesh sizes in the fine sieves is that they were the sizes available to the contractor. During the study, newly harvested (i.e. in late May) tree sections from a first thinning were chipped and transported to the CHP plant in Östersund. The tree sections in chipped piles consisted of a random mix of pine, spruce, aspen, and birch. The average moisture content in the chipped material was 41.7%. For each chipper and sieve combination, it was intended to fill with chips three tractor trailers, each with a gross volume of approximately 42.5 m3. After filling three trailers, the sieve was shifted and the chipper fitted with a new set of sharp knives to avoid that knifewear should affect the results. The tractor trailers were taken to the measurement station at the CHP plant in Östersund, where the volume and weight of the chips was measured, and samples were taken for determination of moisture content and chip size distributions. For each trailer, a 10 l sieving sample and at least 3 smaller samples for moisture content determination were taken. For the combination of the Kesla chipper and fine sieve 3 sieving samples were taken from the same trailer. The moisture content samples were scaled when sampled and after drying at 105˚C for 24 hours. The sieving samples were sieved according to SIS-CEN/TS 15149–1:2006. The fuel consumption of the tractors that powered the chippers and their hydraulic loaders were measured by topping up the fuel tank after each filled chip trailer using an accurate fuel gauge. To compensate for differences between trailer loads, fuel consumption per produced amount of chips (odt) were used in the analyses. The time study of the chipping work was done as a comparative time study with snap back timing (Bergstrand 1987). Time recording was made with Allegro hand-held computers equipped with Skogforsk SDI software. Chipping work was split into 8 elements (Table 1). All measured times for each trailer load have been summarized per work element and divided by the oven dry mass of the load to get times in centiminutes per oven dry ton (odt). In some of the analyses the elements »Boom out«, »Grip«, »Boom in & feeding«, »Release & adjustment«, »Chipping«, »Move and Other« have been summarized in the main work Croat. j. for. eng. 36(2015)1


Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution ... (11–17)

L. Eliasson et al.

Table 1 Work elements used in the study Element

Definition

»Boom out«

Boom movement from the chipper to the piled material

»Grip«

Gripping of material

»Boom in & Feeding«

Boom movement from the pile to the machine and using the boom to assist feeding the chipper before the grapple load is released

»Release & adjustment«

Releasing the grapple load and possible adjustments of the material on the feeding table

»Chipping«

Chipping while the loader is idle

»Move«

Repositioning of the machine alongside the piled material

»Other«

Other work time – works not covered above that is needed to complete the work task

»Delays«

All that is not productive work

element efficient chipping time. Only effective times have been included in the analysis and no delays have been reported. The reason for not reporting any delays are that all delays either were caused by this study or by the establishment of a storage trial at the heating plant in Östersund. The storage trial substantially increased unloading times for the transport tractors, thus causing waiting times for the chippers. The study was designed as a factorial experiment with the factors »Chipper« in two nominal levels (Kesla 645 and Biber 92), »Sieve_size« in three ordinal levels (coarse, normal and fine), and »Size_class« in eight ordinal levels (<3.15, 3.15–8, 8–16, 16–31, 31–45, 45–63, 63–100, >100). All analyses of productivity and fuel consumption have been made using analysis of variance, and difference between means have been tested post hoc using t-tests and Tukey t-tests. Chip size distribution has been analysed using a general linear model (GLM) on logit transformed shares (S) using the factors »Chipper«, »Sieve_size«, and »Size_class«. The logit transformation was necessary since it transforms the primary range of shares S Î [0, 1] onto the interval [– ∞, ∞] assumed by the normal distribution (Olsson 2002). The test criteria were the respective interactions of »Size_class« within »Chipper« (Chipper × Size_class) and »Size_class« within »Sieve_ size« (Sieve_class × Size_class). If the effect of »Size_ class« (on Logit S) was found to be independent of the respective interactive factors »Chipper« and »Sive_ size«, no effect on chip size distribution may be assumed. The model used can be expressed as: S Logit S = Log ( )= 1-S = Chipper + Sieve size + Size class + Chipper × Size class + + Sieve size × Size class + e

Croat. j. for. eng. 36(2015)1

(1)

The two-way interaction Chipper × Sieve_size and the 3 way interaction Chipper × Sive_size × Size_class were not included in the model since they lack plausible interpretation. Restricted maximum likelihood methodology was used for the GLM analysis, and Type V sum of squares as implemented in the STATISTICA version 12 statistical software package.

3. Results Both the performance and fuel consumption per produced odt of chips were significantly dependent on the choice of sieve (Tables 2–4). With the 100 mm sieve, the Kesla 645 chipper produced 14.5 oven dry ton (odt) of chips per effective hour and the Biber 92 30.0 odt per effective hour. Table 2 Chipper performance and fuel consumption depending on sieve mesh size. Fuel consumptions followed by different letters within a machine are significantly different (p < 0.05) Chipper

Sieve

Performance

Fuel consumption

Odt/Eff.hour

Liter/TTV

Biber 92

100

30.0

2.1a

Biber 92

50

25.8

2.8b

Biber 92

35

23.0

3.2c

Kesla 645

100

14.5

3.0a

Kesla 645

50

13.1

3.4b

Kesla 645

25

6.7

7.0g

Decreasing sieve size to 50 mm decreased productivity by 10% for the Kesla and 14% for the Biber

13


L. Eliasson et al.

Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution ... (11–17)

Table 3 Anova for the fuel consumption per odt. n = 16

Fig. 1 Time consumption for efficient chipping work in minutes per odt of chips separated on sieve size and machine. Biber 92 denoted by black squares and Kesla 645 by blue rhombs. Bars denote 95% confidence intervals (Table 2) and a further decrease in sieve size caused further reductions in chipper performance. There were significant effects on the effective chipping time per odt of chips by both machine and sieve, and a significant interaction between the two was observed (Table 4, Fig. 1).

Source

DF

Type III SS

Mean Square

F Value

Pr > F

»Machine«

1

10.53

10.53

122.84

<.0001

»Sieve«

2

13.28

6.64

77.44

<.0001

»Machine * Sieve«

2

5.62

2.81

32.80

<.0001

The interaction is caused by the slow chipping work that occurred when the Kesla 645 was used with the fine sieve (Fig. 1). Only one trailer of chips was produced with the Kesla and the fine sieve, as the contractor was not keen to continue to operate the machine in this setting. For the Biber 92 chipper, the choice of sieve had significant effects on the time consumption per odt for the individual work elements »Boom in & feeding« and »Chipping« (Table 5). For the Kesla 645, observed time consumptions are higher for the fine sieve, but in the statistical comparison between the medium and coarse sieve no differences can be found. Table 4 Anova for the efficient chipping time per odt. n = 16 Source

DF

Type III SS

Mean Square

F Value

Pr > F

»Machine«

1

442 101.4

44 101.4

246.86

<.0001

»Sieve«

2

151 648.6

75 824.3

42.34

<.0001

»Machine * Sieve«

2

99 519.2

49 759.6

27.78

<.0001

Fig. 2 Chip size distribution as an effect of sieve size, upper part of the figure actual shares in per cent per class, lower part of the figure logit transformed data with 95 % confidence intervals. Coarse sieve denoted by black dots, medium sieve by triangles and fine sieve by squares

14

Croat. j. for. eng. 36(2015)1


Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution ... (11–17)

L. Eliasson et al.

Table 5 Time consumption per odt for the individual work elements separated on chipper and sieve. Time consumptions followed by different letters within a machine are significantly different (p < 0.05) Work element Sieve size

Kesla 645

Biber 92

Fine

Medium

Coarse

Fine

Medium

Coarse

»Boom out«

103.6

74.3a

68.8a

44.4a

42.4a

37.9a

»Grip«

50.2

29.0a

29.0a

13.8a

14.2a

12.2a

»Boom in & Feeding«

328.5

219.5a

182.8a

73.0a

71.2a

58.4b

»Release & adjustment«

19.2

18.4a

24.5a

3.4a

5.5a

7.3a

»Chipping«

392.2

126.9a

113.1a

125.9a

99.0b

85.3c

»Other«

0.0

0.0

0.0

0.0

0.0

0.0

»Move«

0.0

0.0

0.0

0.3

0.3

0.0

»Efficient chipping time«

893.7

468.0

418.2

260.6

232.4

201.1

Fuel consumption increased by approximately 50% for the Biber 92, and 130% for the Kesla 645 when the coarse sieve was replaced by the fine sieve (Table 2). The chip size distributions of the produced chips were quite uniform, and only the share of particles in size classes larger than 31 mm was significantly affected by sieve size (i.e. the significant Sieve_size × Size_class interaction in Table 5, Fig. 2). The coarse sieve produced significantly more chips in these size classes than the fine sieve. No significant differences between chippers, i.e. in the Chipper × Size_class interaction, could be found (Table 5). A visual inspection of the chips showed that the chips produced by the Kesla 645 with the fine sieve were not cut but rather ground, and were more like a hog fuel in structure than normal chips. This is probably an effect of the mesh size that was smaller than the cut length of the knives.

4. Discussion As the studied chippers represented two different size classes for professional chipping on landings, it was expected that there should be a productivity difference between them. The observed difference in productivity and fuel consumption, when the chippers used the fine sieve, is misleading for two different reasons: the sieves used did not have the same mesh size and the area of a 35 mm square hole is actually 96% larger than that of a 25 mm square hole, so the »fine« sieve in the Kesla caused more resistance to the chips than the »fine« sieve in the Biber. The operator was not able to adjust the feeding speed of the Kesla chipper, Croat. j. for. eng. 36(2015)1

Table 6 Anova table from the test of chip size distribution Source

DF

Type III SS

Mean Square

F Value

Pr > F

»Chipper«

1

1.58

1.58

0.53

0.4679

»Sieve«

2

126.83

63.41

21.26

<.0001

»Size class«

7

1498.99

214.14

71.80

<.0001

»Chipper * Size class«

7

14.85

2.12

0.71

0.6625

»Sieve * Size class«

14

218.83

15.63

5.24

<.0001

so that the cut length of the chips became smaller than the sieve size. This caused the chipper to almost grind the cut chips as it forced them through the sieve. To perform as intended with the fine sieve the operator should have needed to adjust the knife and counter blade settings on the Kesla. The productivity of the Kesla 645 was somewhat lower and the fuel consumption was higher compared to studies of the similarly sized Bruks 605 chipper (Johannesson et al. 2012, Grönlund and Eliasson 2013), which to a large extent may be caused by the material chipped, the tractors powering the chippers and the operators. The Biber 92/Claas Xerion 5000 is comparable in size and power to forwarder mounted Biber 84 and Bruks 806 chippers that were studied in the spring of 2013 (Eliasson et al. 2013, Lombardini et al. 2013) and both performance and fuel consumption were on par with those machines. Previous studies of chippers with a bottom sieve have shown that a larger mesh size gives increased productivity and improved fuel efficiency compared

15


L. Eliasson et al.

Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution ... (11–17)

to a smaller mesh size (Nati et al. 2010, Röser et al. 2012). This is confirmed for both chippers in the current study. Furthermore, both productivity and fuel efficiency will decrease radically if the cut length of the chipper exceeds the mesh size as for the Kesla chipper with 25 mm sieve. The use of a sieve between the drum and the auger that extracts the chips from the drum casing introduces a resistance in the chip extraction. This resistance is dependent on the total sieve area, the area of the individual holes in the sieve and the amount of chips that passes the sieve per minute. If the amount of chips per time unit is large, all chips smaller than the mesh size will not be able to leave the drum casing but will start to tumble around in the drum casing. In this process, oversized chips and some chips that are of acceptable size will be chipped further. However, the tumbling of material is energy demanding and time consuming. As an example of the chip samples produced with the Biber chipper and the coarse sieve, approximately 80% passed the 31 mm sieve in the fraction analysis and since the 35 mm square meshes in the fine sieve on the Biber chipper is substantially larger, most chips should in theory be able to pass it. Even if those last 20% of the material are needed to be chipped again, and this will take as long time as chipping the same amount of unchipped material, the total chipping time would only increase by 20% and not by 30%, which is the difference noted between the coarse and fine sieve. As expected, a decreasing sieve size decreased the share of coarser chips. However, the ability of the chippers to produce coarser chips and less fines by using a sieve with larger mesh size seems to be limited. For the Eschlböck Biber 92 the cut length is probably the factor that is most important to the chip length, while the effect of feeding speed and sieve seems to be of minor importance. In other words, it behaved almost as a closed drum chipper. For the Kesla 645, it may be possible for the operators to increase the chip size by changing feeding speed and sieve mesh size without changing the cut length, as long as the mesh size exceeds the cut length. On the other hand, to operate acceptably with the fine sieve, in this study the operators should have decreased the cut length of the Kesla chipper. A decrease in cut length decreases productivity and fuel efficiency for the chipper (Spinelli and Magagnotti 2012), but probably not as much as the »grinding« process observed in this study when the chips were forced through an undersized sieve. Regardless of the sieve used, both chippers produced chips that are considered on the fine side for the large CHP plants in Sweden. Many of these plants prefer chips with the highest possible proportions of chips

16

in the 31–45 mm size class and a low amount of chips smaller than 8 mm. On the other hand, the chips are well adapted to the demands of smaller heating plants. If the contractors are interested in increasing the chip size to adhere to demands from the larger plants, the cut length of the chippers must be increased. However, it is not possible to increase the chip size that much by simply changing the sieve and increasing the infeed speed. In the past, statistical analyses regarding the effects of different chip-size distributions have often been done separately for each chip size class (e.g. Spinelli et al 2013). These analyses usually use Anova or t-tests on transformed shares, most often using arcsin transformations. The drawback with this method is that each of the eight tests needed introduce standard type 1 and 2 errors that combine into an accumulated error when hypotheses are repeatedly tested across chip-size classes. The method used in this paper increases the power of the test and avoids the multiple testing that occurs when each size class is analysed separately. The study shows that there is a potential to increase chipper productivity by 10–20% and to reduce the fuel consumption as much by increasing the sieve mesh size from the normal 50 mm mesh size to 100 mm, if the customer can accept that 5% of the chips are longer than 100 mm.

Acknowledgement This research is a part of the INFRES project which has received funding from EU 7th framework programme under grant agreement n°311881, and it is also a part of the »Efficient Forest Fuel« programme funded by the Swedish forest sector and the Swedish energy agency.

5. References Bergstrand, K.G., 1987: Planning and analysis of time studies on forest technology. The forest operations institute of Sweden, 58 p. Brunberg, T., 2013: Skogsbränslets metoder, sortiment och kostnader 2012. http://www.skogforsk.se/sv/kunskap/ db/2013/Skogsbranslets-metoder-sortiment-och-kostnader-2012/, Skogforsk. Eliasson, L. 2011: Procurement systems for logging residues. In: Thorsén, Å., Björheden, R., Eliasson, L. Efficient forest fuel supply systems. Composite report from a four year R&D programme 2007–2010. Uppsala, Skogforsk: 24–26. Croat. j. for. eng. 36(2015)1


Effects of Sieve Size on Chipper Productivity, Fuel Consumption and Chip Size Distribution ... (11–17) Eliasson, L., Granlund, P., Johanneson, T., von Hofsten, H., Lundström, H., 2012: Flisstorlekens effekt på en stor skivhuggs bränsleförbrukning och prestation. Skogforsk, 9 p. Eliasson, L., Granlund, P., Johannesson, T., Nati, C., 2011: Prestation och bränsleförbrukning för tre flishuggar.. Skogforsk, 17 p. Eliasson, L., Lombardini, C., Lundström, H., Granlund, P., 2013: Eschlböck Biber 84 flishugg – Prestation och bränsleförbrukning. Rangering av fliscontainrar med en John Deere 1410 containerskyttel. Skogforsk, 13 p. Facello, A., Cavallo, E., Magagnotti, N., Paletto, G., Spinelli, R., 2013: The effect of chipper cut length on wood fuel processing performance. Fuel Processing Technology 116(0): 228–233. Grönlund, Ö., Eliasson, L., 2013: Knivslitage vid flisning av grot. Skogforsk, 11 p. Johannesson, T., Fogdestam, N., Granlund, P., Eliasson, L., 2012: Effects of chip-length settings on productivity and fuel consumption of a Bruks 605 drum chipper. Skogforsk, 15 p.

L. Eliasson et al.

Lombardini, C., Granlund, P., Eliasson, L., 2013: Bruks 806 STC – Prestation och bränsleförbrukning. Skogforsk, 7 p. Nati, C., Spinelli, R., Fabbri, P., 2010: Wood chips size distribution in relation to blade wear and screen use. Biomass and Bioenergy 34(5): 583–587. Olsson, U., 2002: Generalized linear models. An applied approach. Studentlitteratur, Lund, Sweden. Röser, D., Mola-Yudego, B., Prinz, R., Emer, B., Sikanen, L., 2012: Chipping operations and efficiency in different operational environments. Silva Fennica 46(2): 275–286. Spinelli, R., Magagnotti, N., 2012: The effect of raw material, cut length, and disc chip discharge on the performance of a industrial chipper. Forest Products Journal 62(7–8): 584–589. Spinelli, R., Cavallo, E., Eliasson, L., Facello, A., 2013: Comparing the efficiency of drum and disc chippers. Silva Fennica 47(2): article id 930.

Author’s address: Assoc. Prof. Lars Eliasson, PhD. e-mail: Lars.Eliasson@Skogforsk.se Henrik von Hofsten, Researcher e-mail: Henrik.vonhofsten@Skogforsk.se Tomas Johannesson, Researcher e-mail: Tomas.Johannesson@Skogforsk.se Skogforsk Uppsala science park 75183 Uppsala SWEDEN Assoc. Prof. Tomas Thierfelder, PhD. e-mail: tomas.thierfelder@slu.se Swedish University of Agricultural Sciences Department of Energy and Technology Box 7032 75007 Uppsala SWEDEN

Received: April 28, 2014 Accepted: September 23, 2014 Croat. j. for. eng. 36(2015)1

Raffaele Spinelli, PhD. e-mail: spinelli@ivalsa.cnr.it CNR IVALSA Via Madonna del Piano 10 50019 Sesto Fiorentino (FI) ITALY

17



Original scientific paper

Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside Landing with a Low-Speed Double-Shaft Grinder and a Star Screen Juha Laitila, Yrjö Nuutinen Abstract Impurities in harvested stumps are a quality problem because high levels of mineral contaminants decrease the effective heating value of the stump wood, and can also affect ash melting behaviour during combustion, leading to sintering and drift problems. The aim of this case study was to clarify the productivity and screening efficiency of the Kompetech Crambo 6000 low-speed double-shaft grinder equipped with a Komptech star screen, in the integrated grinding and screening of Norway spruce and Scots pine stumps for fuel at a roadside landing, when using two different sieve sizes (250 x 320 mm and 180 x 180 mm screen baskets). Furthermore, we studied the fuel consumption of the Crambo 6000 grinder, ash content and particle size distribution of ground stump wood, and ash content and particle size distribution of the screening reject. In addition, the heating value of the produced hog fuel and screening reject were analysed. During the time of the studies, both the grinder and star screen were operating well and there were no delays due to machine breakdowns. The mobile Crambo 6000 grinder was also capable of operating well in constricted roadside landings. The quality of the produced hog fuel was high, due to low ash content (0.4–2.3%), and this highlights the significance of screening to guarantee sufficient quality when processing stump fuel. The ash content of the screening reject was 32.4–74.7%, and the effective heating value was 5.2–13.4 MJ/kg. The effective heating value of the produced hog fuel was 17.9–19.9 MJ/kg. The average grinding productivity, when using the 250 x 320 mm screen basket, was 162 loose m3 per effective hour, and the fuel consumption of the grinder was 0.44 litres per loose m3. With a narrower screen, the average grinding productivity was 101 loose m3 per effective hour, and the fuel consumption of the grinder was 0.75 litres per loose m3. Keywords: grinding, screening, stump wood, quality, procurement system, buffer-storage, hog fuel

1. Introduction The harvesting of stumps for energy generation in Finland has increased rapidly during the past ten years. The previous time when stumps were lifted was in the 1970s and in the early 1980s, when they were procured as raw material for either sulphate chemical pulp or energy generation (e.g. Hakkila 1976, Kuitto 1984, Hakkila 2004). However, high harvesting costs made this unprofitable, and the use of stumps came to an end. Stump harvesting was reintroduced ten years Croat. j. for. eng. 36(2015)1

ago, when UPM Forest commenced the harvesting of Norway spruce (Picea abies (L.) H. Karst.) stumps, along with logging residues and small-diameter trees from thinnings, for delivery to the combined heat and power (CHP) plant of Jämsänkoski paper mill in Central Finland (Markkila 2005, Backlund 2007). Although it was initially met with scepticism, the use of stump wood in energy generation then spread to other parts of the country, and in 2012, some 1.1 million m3 of comminuted stump wood were consumed by heating

19


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

plants and power plants (Ylitalo 2013), with the total techno-economical harvesting potential estimated to be 5.0 million m3 per year (Kärhä et al. 2010b). Coniferous rootstock is considered to be a promising energy source because it contains higher concentrations of the energy-rich components – lignin and extractives, than stem wood (Hakkila 1975, 1976). Norway spruce is the most interesting species for stump harvesting because it is easier to harvest and clean than Scots pine (Pinus sylvestris L.) (Nylinder 1977, Hakkila 2004). The root system of Norway spruce spreads out close to the ground surface; it has no taproot and the lateral roots are thicker and longer than those of Scots pine (Hakkila 1976). Impurities in the harvested stumps are a quality problem, as usually the stumps and roots include some rocks and soil (Spinelli et al. 2005, Laitila et al. 2010, Laurila and Lauhanen 2010). The effective heating value of wood biomass is the main parameter defining its quality as fuel. Ash content is one of the major factors decreasing the calorific value of the stump wood, and high levels of mineral contaminants can also affect ash melting behaviour during combustion, leading to sintering and drift problems (Anerud and Jirjis 2011, Anerud 2012). The stumps are uprooted and split using a tracked excavator equipped with a stump extraction head (e.g. Backlund 2007, Karlsson 2007, Laitila et al. 2008, Hedman 2008, Lazdins et al. 2009, Jouhiaho et al. 2010, Lindroos et al. 2010, Laitila 2010, Anerud and Jirjis 2011, Erkkilä et al. 2011). The splitting of stump wood into pieces accelerates its drying and increases the productivity of the comminution work. Furthermore, the risk of impurities is higher when the stump is not split properly. Excavators of about 20 tons are used in stump harvesting (Laitila 2010). After seasoning at the stand, stumps are forwarded to roadside landings by forwarders (e.g. Backlund 2007, Karlsson 2007, Laitila et al. 2008, Lazdins 2009, Laitila 2010). Harvested stump and root wood dries fast during spring and summer, and moisture content (wet basis) can decrease from 53% to 31% even in one month (Laurila and Lauhanen 2010). If Norway spruce stump wood has dried well once, water absorption is very weak and the moisture content increases only slightly in the late autumn (Laurila and Lauhanen 2010). Dry matter losses of 4% due to decay, during 1 year of storage, have been reported by Nylinder and Thörnqvist (1981), and during 3 months of storage, dry matter losses of stump wood have been in the range of 1.5–3.4% of dry weight (Anerud and Jirjis 2011). Stumps are comminuted with grinders, as the blunt tools are less sensitive to the wearing effect of contaminants such as mineral soil and stones (Rinne 2010,

20

Spinelli et al. 2012, Eriksson et al. 2013). However, grinders offer a rather coarse product, unsuitable for use especially in some smaller plants (Strelher 2000, Rinne 2010, Eriksson et al. 2013). In contrast, chippers with sharp tools are exclusively applied to clean wood and offer a more consistent product (Rinne 2010, Spinelli et al. 2011, Spinelli et al. 2012, Eriksson et al. 2013). In Finland, the majority of stumps are ground either at the plant or at terminals, whereas the majority of smalldiameter trees and logging residues are chipped at roadside landings (Strandström 2013). Until now, comminution of stumps has been done with heavy, often stationary grinders. In smaller plants, due to small comminution volumes, the construction of a stationary grinder is not economically feasible (Laitila et al. 2010). In addition, the transportation of stumps calls for a biomass truck with solid side panels and bottom, and economical transport distances are short, owing to the small potential payloads (Ranta and Rinne 2006, Laitila et al. 2010). The relative bulk density of an intact rootstock pile is of the order of 0.1, whereas that of chopped stumps is 2 to 4 times greater (Hakkila 1976). As stump parts are hard and can damage the sides of the truck, the truck must be made of strong material, which increases the kerb weight of the truck-trailer unit. Recently, effective mobile grinders suitable for the comminution or pre-comminution of stumps at roadside landings have been introduced (von Hofsten and Granlund 2010, Kärhä et al. 2011, Laitila et al. 2013). The truck- or semitrailer-mounted grinder is used in a similar manner to mobile chippers in the chipping of logging residues and small-diameter trees. The grinder moves from landing to landing, with the comminuted material transported to the end-user by truck (Asikainen 2010). According to an expert survey (Laitila et al. 2010), two major problems that can limit the growth of stump wood procurement for fuel are: 1) stump particles remain too large when applying current harvesting technologies, resulting in the truck load containing too much air, so that, in terms of weight, the trucks seldom reach maximum payload; 2) stump wood deliveries contain contaminants, which reduce the calorific value, can cause problems in combustion, and might thus limit the share of stump wood in the fuel mix of power and heating plants. There is also a need to develop logistic models for procurement and storage, because supply and demand for fuels is often diachronic, and screening of stump wood is more effective when material is unfrozen (Laitila et al. 2010). During the cold season of the year, the comminuting machinery and transportation equipment are in intensive use, while during the summer months, the problem is a lack of Croat. j. for. eng. 36(2015)1


Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

work (Laitila et al. 2010). In order to guarantee a reliable supply of fuels from roadside landings during the cold season, there is an obvious need to store comminuted stump wood in buffer stacks at terminals and plants for at least a few weeks before combustion (Laitila et al. 2010). One way to increase the payload and reduce the contaminant content of fuel chips is integrated comminuting and screening of stumps or logging residues at roadside landings (Anerud 2012, von Hofsten et al. 2012, Fogdestam et al. 2012, Eriksson et al. 2013, Dukes et al. 2013). This approach would reduce the amount of fine material contaminants at the source, provide a possibility to increase payloads and lower transportation costs, and at the same time increase the quality of the produced fuel (Anerud 2012, von Hofsten et al. 2012, Eriksson et al. 2013, Dukes et al. 2013). In a study, Laitila et al. (2013) found that the heating values of the finest stump wood hog fuel particles were significantly lower compared to larger particles, due to high contaminant and bark content. The recently introduced semitrailer-mounted Komptech Crambo 6000 low-speed double-shaft shredder equipped with a Komptech star screen is a novel mobile grinder unit, which is capable of operating both at terminals and roadside landings and of producing hog fuel with lower fine and contaminant content than usual (http://www.komptech.com). Currently, there is one Crambo 6000 grinder equipped with a Komptech star screen operating in Finland. Many questions are raised that must be addressed quickly in order to overcome possible bottlenecks. If comminution of stumps at roadside landings is becoming more common, there is also an urgent need to get information about storing comminuted stump wood in buffer stacks of plants and terminals.

1.1 Aim of the study The aim of this case study was to clarify the productivity and screening efficiency of the Crambo 6000 grinder equipped with a star screen in the integrated grinding and screening of Norway spruce and Scots pine stumps for fuel at a roadside landing, when using two different sieve sizes (250 x 320 mm and 180 x 180 mm screen baskets). Furthermore, we studied the fuel consumption of the Crambo 6000 grinder, ash content and particle size distribution of ground stump wood, and ash content and particle size distribution of the screening reject. The heating value of the produced hog fuel and screening reject were also analysed. In addition, we measured the self-warming of ground stump wood during 64 days of storage at the buffer stack in the terminal, from 6th August 2013 to 10th October 2013. Croat. j. for. eng. 36(2015)1

J. Laitila and Y. Nuutinen

2. Material and methods 2.1 The time study of integrated grinding and screening of stump wood for fuel The time study data consisted of 20 semitrailer loads of comminuted stump wood originating from five different stands located in Juva (61°54’N, 27°47’E), Eastern Finland (Table 1), and the studies were carried out during daylight hours from 5th to 8th August 2013. The observation unit was a semitrailer with a 90 m3 load volume, and each load was always completely filled by conveyer belt and levelled with a shovel. The ground materials were Scots pine and Norway spruce stumps, which had been uprooted and split by an excavator-based stump harvester. Small stumps (diameter < 30 cm) had been split into two pieces, while larger stumps were split into three or four pieces. The average stump diameter was measured to be 40 cm. The storage time of stumps at the roadside landing before comminution was in the range of 10–20 months (Table 1). In the trials, the Crambo 6000 grinder was equipped with 250 x 320 mm and 180 x 180 mm screen baskets, and 10 semitrailer loads were comminuted using both sieve sizes (Table 1). The use of 180 x 180 mm and 250 x 320 mm screen baskets gave a granularity that requires secondary grinding either at the terminal or at the plant before combustion. From the roadside landing comminuted stump wood, 6 loads were transported to the CHP plant, and 14 loads to the terminal. The CHP plant was located in Mikkeli (61°41’N, 27°17’E) and the terminal in Pieksämäki (62°15’N, 27°12’E), Eastern Finland. The payloads of the semitrailer loads that were transported directly to the CHP plant were measured with a certified weight scale at the plant, and both the filled and empty weights of the semitrailers were recorded. Unfortunately, at the Pieksämäki terminal or nearby, a certified weight scale was not available in trim. Therefore, the experimental setup was based on using the semitrailer with a 90 m3 load volume as the base unit for productivity measurements and comparison of different treatments. The fuel consumptions of the Crambo 6000 grinder and towing vehicle were measured at a local fuel station after changing screen baskets, at the end of every second grinding day. The mobile grinder and towing vehicle were parked in exactly the same place both times, and the fuel tanks were refilled to full. The accuracy of the fuel pump was 0.1 litres. The mobile industrial grinder used for the experiment was a tandem-axle semitrailer-mounted Crambo 6000 grinder equipped with a star screen (Fig. 1 and Fig. 2). The grinder was driven independently

21


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

Table 1 Properties of the ground material per semitrailer load Sequence number

Ground material &

Size of screen basket,

Moisture content

Volume, m3 &

Basic density

of loads

storage time in months

mm

of stump wood, %

payload, kg

of stump wood, kg/m3

1

Norway spruce & 11

250 x 320

17.9

90 m3 loose

461

2

Norway spruce & 11

250 x 320

20.9

90 m3 loose

466

3

Scots pine & 20

250 x 320

23.9

3

90 m loose

417

4

Scots pine & 20

250 x 320

31.5

90 m3 loose

435

5

Scots pine & 20

33.7

3

90 m loose

440

6

Scots pine & 20

250 x 320

29.8

3

90 m loose

452

7

Scots pine & 20

250 x 320

20.9

90 m3 loose

446

8

Norway spruce & 10

250 x 320

23.9

3

90 m loose

440

9

Norway spruce & 10

250 x 320

22.2

90 m3 loose

420

10

Norway spruce & 10

250 x 320

21.2

3

90 m loose

431

11

Norway spruce & 12

180 x 180

32.7

90 m3 loose & 13 900 kg

430

12

Norway spruce & 12

180 x 180

32.8

90 m3 loose & 14 800 kg

409

13

Norway spruce & 12

180 x 180

35.4

90 m loose & 14 100 kg

419

14

Norway spruce & 12

180 x 180

36.7

90 m3 loose & 15 000 kg

447

15

Norway spruce & 12

180 x 180

27.8

90 m loose & 14 300 kg

418

16

Norway spruce & 12

180 x 180

23.1

90 m3 loose & 13 050 kg

440

17

Norway spruce & 11

180 x 180

26.5

90 m3 loose

435

18

Norway spruce & 11

180 x 180

27.2

3

90 m loose

432

19

Norway spruce & 11

180 x 180

23.1

90 m3 loose

442

20

Norway spruce & 11

25.1

3

392

250 x 320

180 x 180

and it was powered by a 429 kW CAT C18 six-cylinder diesel engine. The grinder year model was 2011. The towing vehicle was a three-axle Volvo FM 12 (year model 2004) and the tractor was equipped with a heavy-duty Kesla 2012T cab timber loader, used to bring the wood to the vertical-flow in-feed hopper (Fig. 1 and Fig. 2). The timber loader was equipped with a five-spike grapple developed for handling stumps and logging residues (Fig. 2). The star screen deck underneath the screen baskets separated off the fine fractions, which were removed back to the landing area via a side conveyer belt (Fig. 1 and Fig. 2), whereas the coarse fraction was considered to be fuel. The width of the forest roads and ditches beside the ground stump wood piles were 4 m and 1 m, respectively. The area of the in-feed hopper measured 2000 mm in width and 2820 mm in length, and the two shred-

22

3

3

90 m loose

ding drums were located at the bottom of the in-feed hopper. The lengths of the shredding drums were 2820 mm, the drum diameter was 610 mm, and the maximum rotation speed of the drums was 0.68 s/r (41 rpm). The sickle teeth seized the material and pressed it, in a splitting action, against the cutting edge and screen baskets located underneath. The material did not exit the shredding area until the particle size matched the hole size of the screen basket, enabling the quantity of ground material of the desired particle size to be maximized. Ground stump wood was discharged into the walking floor semitrailer via a conveyer belt, and the semitrailers were located in a consecutive line (Fig. 2) or crossways to the Crambo 6000 grinder. In time studies, the semitrailers were located almost invariably (16 loads out of 20) in a consecutive line to the grinder. The skilful operator had extensive working experience. He had fifteen Croat. j. for. eng. 36(2015)1


Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

J. Laitila and Y. Nuutinen

Harstela 1991). The accuracy of the Rufco-900 field computer was 0.6 s (Nuutinen et al. 2008). The grinder working time was divided into effective working time (E0h) and delay time (Harstela 1991), which is a common method employed in Nordic work studies. Auxiliary times (e.g. planning of work and preparations) were included in the work phases in which they were observed. Effective working time was divided into the following main work phases, in order of priority:

Fig. 1 The tandem-axle semitrailer-mounted Crambo 6000 lowspeed double-shaft grinder equipped with a star screen and heavyduty Kesla 2012T cab timber loader

Þ Loading: The work cycle began when the grapple started to move towards the stump stack and ended when a stump bunch had been lifted and placed in the in-feed hopper of the grinder. The number of grapple loads for each semitrailer load was counted, in order to calculate the average size of the grapple load in feeding. Þ Grinding (loading is idled): Began when the stump bunch had been lifted and placed in the in-feed hopper of the grinder, the in-feed hopper was full, and the shredding drums were processing wood into pieces. The work phase ended when the grapple started to move towards the stump stack or the back doors of the semitrailer had to close. Þ Moving of the semitrailer and lifting of the conveyer belt: Began when the load of the semitrailer was almost full of comminuted wood and the back doors had to close. The work phase ended when the back doors were closed, the conveyer belt was lifted higher, and the grapple started to move towards the stump pile or the shredding drums started to process the wood into pieces.

Fig. 2 Integrated grinding and screening of stumps for fuel at the roadside landing (ground stump wood was discharged into a walking floor semitrailer via a conveyer belt, and screening reject was removed back to the landing area via a side conveyer belt) years of working experience in truck-transporting industrial roundwood, and almost three years of working experience in grinding stumps with a Crambo 6000 grinder. The total weight of the mobile grinder-truck-trailer unit was 40 tons, and that of the grinder unit was 22 tons. The time study was carried out manually using a Rufco-900 field computer (Nuutinen et al. 2008). The working time at the worksite was recorded by applying a continuous timing method, by which a clock runs continuously and the times for different elements are separated from each other by numeric codes (e.g. Croat. j. for. eng. 36(2015)1

Þ Arrangements: Repositioning of stumps at the roadside pile in order to improve loading work, or shaking off stones or other noticeable impurities. Delays or preparation time: Time not related to productive grinding work, but for which the reason for the interruption was recorded. The main reasons for delay times shorter than 15 minutes were preparing the grinder for grinding work, moving machines and vehicles at the roadside landing, cleaning the dropped material and grinding residues away from the road, organisational delays (e.g. telephone calls), and personal breaks. The data analysis was conducted for productive time only (E0h), in order to avoid the confounding effect of delay time, which is typically erratic (e.g. Spinelli and Visser 2009, Eliasson et al. 2012, Holzleitner et al. 2013). The productive time (E0h) included the work phases of loading, grinding, and arrangements.

23


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

2.2 Sampling and laboratory analyses of ground stump wood and screening reject Stump wood samples were taken directly from the arriving semitrailers as part of the normal delivery process in the yard of the terminal or plant, after unloading comminuted wood to the ground (Uusvaara 1978, Uusvaara and Verkasalo 1987). Samples were taken to define the moisture content, basic density, dry weight of the hog fuel, particle size distribution, ash content, and effective heating value of comminuted stump wood, and samples were analysed in the laboratory of the Finnish Forest Research Institute according to the following standards: SFS-EN 14780, SFS-EN 14774-1, SFS-EN 14774-2, SFS-EN 14774-3, SCAN-CM 43:95, SFS-EN 15149-1, SFS-EN 14775, and SFS-EN 14918. The solid content of the weighed stump wood loads (%) was based on the relation of the recorded dry masses (kg), dry green densities (kg/m3), and frame volumes of each load (e.g. Kanninen et al. 1979, Uusvaara and Verkasalo 1987). Four samples were taken for each semitrailer load, and wood samples were stored in plastic bags, which were carefully closed and marked. Moisture samples were packed in double bags in order to minimise the risk of bag outbreak or evaporation. The dimensions of the plastic bags were 35 x 35 cm (volume 8 litres), and the raw material, date, and time were written on the label. In addition, plastic bags were wrapped in a plastic sack, and each semitrailer load was packed in a corrugated paperboard box of its own. The samples were extracted from several locations from the load using a small shovel, so that the results would be representative of the load. The amount, volume, and properties of the screening reject were analysed from semitrailer loads 1, 4, 9, and 15 (Table 1). A tarpaulin was placed underneath the star screen conveyer belt, in order to recover the screening reject of the integrated grinding and screening process (Fig. 3). The volume of the screening reject was mea­sur­ ed by shovelling it into certified 90-litre boxes. From each load, a sample was taken to define the moisture content, weight, particle size distribution, ash content, and heating value of the screening reject according to standards SFS-EN 14780, SFS-EN 14774-1, SFS-EN 14774-2, SFS-EN 14774-3, SFS-EN 15149-1, SFS-EN 14775, and SFS-EN 14918. The samples were taken im­ me­­diately after grinding, with a shovel, in a 10-litre pla­ stic bucket, and the location, date, material, and time were written on the cover. The bucket and cover were sealed with adhesive tape, in order to prevent the screening reject from drying. The sample was extracted from four points in the middle of the screening reject mound, so that the results would be representative of the load.

24

Fig. 3 A mound of screening reject from the first semitrailer load, on a tarpaulin, after integrated screening and grinding of stumps at a roadside landing

2.3 Buffer storage of ground stump wood at a terminal stack The temperature of the ground material at the terminal stack was monitored with four a-Nap 100 temperature loggers, from 6th to 10th August 2013. The temperature loggers were placed in the middle of the stack at 1.0 m and 2.0 m depths, in sealed plastic tubes (diameter 32 mm), and recorded the temperature once an hour. The height of the terminal stack was 4.2 m, width 15 m, and length 70 m. The temperature loggers were fully protected against dust and dirt, and they were capable of measuring temperatures from -30°C up to +85°C. The programming of the temperature loggers and the reading of data after a monitoring period of 64 days 17 hours were done with a special Windows application.

3. Results 3.1 Grinding productivity and fuel consumption The average grinding productivity, when using the 250 x 320 mm screen basket, was 162 loose m3 per effective hour, and the fuel consumption of the grinder was 0.44 litres per loose m3 (Fig. 4). Grinding productivity varied in the range of 126–192 loose m3 per effective hour, and the grapple load size in the feeding was 0.9–1.1 loose m3. The average grapple load in the feeding was 1.0 loose m3. When using the 180 x 180 mm screen basket, the grinding productivity and the grapple load size in the feeding were lower and the fuel consumption per loose m3 was higher compared to the 250 x 320 mm screen basket (Fig. 4). The average grinding productivity was 101 loose m3 per effective hour, Croat. j. for. eng. 36(2015)1


Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

J. Laitila and Y. Nuutinen

Fig. 4 Grinding productivity according to load size in the feeding, for the two screen baskets considered

Fig. 5 Grinding productivity according to moisture content of stump wood (%), for the two screen baskets considered

and the fuel consumption of the grinder was 0.75 litres per loose m3. Grinding productivity varied in the range of 76–124 loose m3 per effective hour, and the grapple load size in the feeding was 0.7–0.9 loose m3. The average grapple load in the feeding was 0.8 loose m3.

whereas the basic density of stump wood (kg/m3) had no observed impact on grinding productivity when the basic density of the wood material was in the range of 392–466 kg/m3. A denser 180 x 180 mm screen basket lowered the grinding productivity, and due to that, the in-feed loading was idled for, on average, 38% of the productive grinding time (min. 31% and max. 47%).

A higher moisture percentage of the wood material improved the grinding productivity (Fig. 5),

Fig. 6 The volume of screening reject per studied semitrailer load (a screen basket of 250 x 320 mm was used for loads 1, 4, and 9, and 180 x 180 mm for load 15) Croat. j. for. eng. 36(2015)1

25


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

Fig. 7 Ash percentage of screening reject and hog fuel Table 2 Properties of the screening reject Sequence number of the load

Total volume of the screening reject, loose m3

Total weight of the screening reject, kg

Moisture percentage of the screening reject, %t

Ash percentage of the screening reject, %t

Effective heating value of the screening reject, MJ/kg (dry mass)

1

4.5

1374

14.6

47.7

10.8

4

5.7

3094

18.4

74.7

5.2

9

7.5

3554

12.0

70.0

6.1

15

11.2

3310

22.3

32.4

13.4

When using the 250 x 320 mm screen basket, the infeed loading was idled for, on average, 22% of the productive grinding time (min. 9% and max. 34%) during the time study.

3.2 Efficiency and quality of screening The volume of screening reject varied in the range of 4.5–11.2 loose m3 per studied semitrailer load (Fig. 6, Table 2), and thus, without screening, the payload of clean hog fuel would have been 78.8–85.5 loose m3 (Fig. 6). Obviously, in practice, the impact of screening on the payloads is not so great, due to the heterogeneous blend of hog fuel and contaminants, and the smaller particle size of screening reject compared to hog fuel. The total weight of the screening reject per semitrailer load was in the range of 1374–3553 kg, and the moisture of the screening reject was 12–22% (Table 2). The ash content of the screening reject was 32.4–74.7%,

26

and the effective heating value was 5.2–13.4 MJ/kg (Table 2). The effective heating value of the hog fuel was 17.9–19.9 MJ/kg (Table 3). The highest ash content and the lowest effective heating value and loose volume of the screening reject were found when using the 250 x 320 mm screen basket (Fig. 6 and Fig. 7, and Table 2). The average ash content of the hog fuel was 1% (SD 0.36%) when using the 250 x 320 mm screen basket, and 1.5% (SD 0.61%) when using the 180 x 180 mm screen basket (Fig. 7, Table 3). The estimated ash content of the harvested stump wood was 3–6% before grinding and screening. The effective heating values of hog fuel for studied semitrailer loads 1, 4, 9, and 15 were, before screening, 14.8 MJ/kg, 11.9 MJ/kg, 13.6 MJ/kg, and 12.7 MJ/kg (wet basis), respectively, and screening improved heating values to 15.2 MJ/kg, 12.5 MJ/kg, 14.4 MJ/kg, and 13.1 MJ/kg (wet basis), Croat. j. for. eng. 36(2015)1


Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

J. Laitila and Y. Nuutinen

Table 3 Properties of the ground material per semitrailer load Sequence number of the load

Ash percentage of the stump wood, %

Effective heating value of stump wood, MJ/kg (dry mass)

Payload of the semitrailer, m3 (solid)

Solid content, %

1

0.85

19.0

2

0.36

18.9

3

1.14

17.9

4

1.37

19.3

5

1.64

19.0

6

0.96

19.9

7

0.62

18.7

8

1.13

19.6

9

1.14

19.2

10

1.02

19.2

11

1.88

18.9

21.8

24

12

2.34

18.4

24.3

27

13

1.02

19.1

21.7

24

14

2.21

18.9

21.2

24

15

0.75

19.1

24.7

27

16

0.82

19.1

22.8

25

17

1.11

19.0

18

1.18

19.5

19

1.38

19.9

20

2.23

19.5

respectively. The average payload of the semitrailer load was 22.8 m3 (SD 1.45 m3) and the average solid content was 25% (SD 1.6%) (Table 3). After grinding and screening, the majority of the hog fuel was in the particle size classes 63–100 mm and > 100 mm, and screening reject was in the particle size class < 3.15 mm (Fig. 8). When using the 180 x 180 mm screen basket, the relative share of bigger particles in the screening reject was higher compared to screening reject when using the 250 x 320 mm screen basket (Fig. 8).

3.3 Stack temperatures of the ground stump wood at the terminal buffer-storage During the monitoring period of 64 days and 17 hours, the stack temperature of the ground stump wood at the terminal buffer storage fluctuated with the weather and season (Fig. 9), and rapid temperature increases caused by self-warming inside the stack Croat. j. for. eng. 36(2015)1

were not observed. The highest temperatures observed inside the stack were 20°C at a 1 metre depth and 21°C at a 2 metre depth. At its lowest, the temperature was 0.5°C at a 1 metre depth, and 1°C at a 2 metre depth (Fig. 9).

4. Discussion and conclusions Previous research on chippers and grinders comminuting woody biomass has highlighted the substantial increases in production rates that result from increasing the size of holes in the screens used (e.g. Kärhä et al. 2010a, Kärhä et al. 2011, Röser et al. 2012, Jylhä 2013), which supports the observations of this study. In addition, the fuel consumption has been noted to be higher when using narrower screens (Kärhä et al. 2010a, Kärhä et al. 2011, Jylhä 2013). In the study of Metsäteho (Kärhä et al. 2011), conifer stumps

27


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

Fig. 8 Particle size distribution of screening reject and hog fuel when using the 250 x 320 mm and 180 x 180 mm screen baskets

Fig. 9 Temperatures of ground stump wood (hog fuel) at a terminal stack at 1 and 2 meter depth were pre-ground with a Crambo 5000 low-speed double shaft grinder using 500 x 320 mm and 120 x 90 mm screen baskets, and the grinding productivities were 171 loose m3/E0h with coarse screen, and 55 loose m3/E0h with a narrower screen. The fuel consumption of the grinder was 0.33 litres and 0.61 litres per loose m3 (Kärhä et al. 2011). In this study, the average ash content of hog fuel after screening was 1% when using the 250 x 320 mm

28

screen basket, and 1.5% when using the 180 x 180 mm screen basket. However, the volumes of screening reject and wood losses were higher when using the narrower screen. The difference in the amount of screening reject was expected and visually so noticeable that one measurement was considered to be enough. The benefit of a narrower screen is that the productivity in the secondary grinding at the terminal or at the plant is higher compared to coarser hog fuel (Kärhä et al. 2011). Croat. j. for. eng. 36(2015)1


Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

The weakness of the study reported herein was that the experimental setup was based on using a semitrailer with a 90 m3 load volume as the base unit for productivity measurements and comparison of different treatments, because it was not possible to measure all loads with a certified weight scale. Therefore, the influence of the screen size on the solid content of hog fuel payloads was not examined. It is clear that dry mass is a more accurate unit compared to loose volume of hog fuel, but in recent studies in Finland, it has been noted that the variation in the solid content values of ground stump wood is unexpectedly narrow (Kärhä et al. 2011, Laitila et al. 2013). In the study of Metsäteho (Kärhä et al. 2011), when using 500 x 320 mm and 120 x 90 mm screen baskets, the measured solid content of hog fuel payloads were 28% and 27%, respectively. In the study of Laitila et al. (2013), conifer stumps were ground with a CBI 5800 fast-running grinder using a 51/76 x 152 mm sieve, and the measured solid content of hog fuel payloads was on average 28% (SD 1.9%). In both studies, conifer stumps were ground at the roadside landings, loads were loaded with belt conveyers, and the technical properties of the ground stumps were similar to those reported in Table 1. However, for defining more accurate solid content of hog fuel loads and factors affecting that, more extensive follow-up and field studies should be conducted (cf. Uusvaara and Verkasalo 1987). Compared to whole trees or logging residue chips, the solid content of hog fuel payloads is about 10% lower (Uusvaara and Verkasalo 1987). In addition, the impact of contaminant content on the payload weights should be taken into account more carefully, because the weight of sand and rocks is significantly higher compared to stump wood. During the time studies, both the grinder and the star screen were operating well, and there were no delays due to machine breakdowns. The mobile Crambo 6000 grinder was also capable of operating well in constricted roadside landings. The quality of the produced hog fuel was high, due to the low ash content, and this highlights the significance of screening to guarantee sufficient quality when processing stump fuel. The considerable variations in contamination levels result in widely varying concentrations of ash. For example, Anerud and Jirjis (2011) have reported ash content ranging between 2% and 7% for freshly ground stumps, and in the study by Laitila et al. (2013), the ash content of seasoned stumps was 13%. In the study by Korpinen et al. (2007), the ash content of the hog fuel samples varied from 1% to 24% and, for most samples, the ash content was below 10%. In Sweden (Anerud 2012, von Hofsten et al. 2012, Fogdestam et al. 2012), studies were made on coarse Croat. j. for. eng. 36(2015)1

J. Laitila and Y. Nuutinen

grinding of stumps combined with sieving the ground stump wood. The contractor ground the stumps using a Doppstad DW 3060 low-speed grinder and the ground material was sieved using a Doppstad SM 620 drum sieve. The mesh size of the drum was 20 mm. An excavator and a truck-mounted grapple loader were used to load stumps into the grinder from the stump piles during the test. In the first test, the grinder and sieve combination produced, on average, 17.7 dry tons of acceptable hog fuel per effective grinding hour, and the fuel consumption for the grinder was 2.8 litres per dry ton. During the sieving process, 22.3% of the ground material was rejected. This material had an ash content of 34%, while the accepted hog fuel had an ash content of 1.1% (Anerud 2012, von Hofsten et al. 2012). In the second test, the stumps were highly contaminated with soil and humus particles (ash content of 22%). Although, on average, 31% of the dry weight was rejected in the sieving process, the accepted material had an ash content of 7.6%, while the rejected material had an ash content of 53.9%. When grinding these stumps, the grinder produced 25.8 dry tons per effective hour, of which 18.7% was acceptable hog fuel after sieving. The average fuel consumption per acceptable dry matter ton was 1.75 litres for the grinder and 0.45 litres for the drum sieve (Anerud 2012, von Hofsten et al. 2012, Fogdestam et al. 2012). In actual operations, the effect of delays or translocations reduces the productivity of the machinery and the whole supply chain in roadside landing operations. Therefore, readers must consider that the figures in this study refer to effective grinding time (E0h) and were calculated for loading, grinding, and arrangements time only, excluding all delays and all other working time. In order to get representative data on delays in an operation, a long study period is needed, because delays or translocations can represent a significant proportion of chipper or grinder scheduled working time, and may account for up to 50% of the total site working time (Spinelli and Visser 2009, Eliasson et al. 2012, Holzleitner et al. 2013). Controlling the complex supply chain of chips from the forest to the customer is a complex task, and comminuting machines and truck-trailer units for transport must be scheduled with minimum operational delay to be profitable (e.g. Spinelli and Hartsough 2001, Stampfer and Kanzian 2006, Kanzian et al. 2009, Asikainen 2010, Holzleitner et al. 2013). Self-warming inside the coarse ground stump wood stack was not observed during the two-month buffer-storage period, which is a benefit, because it is well known from previous studies that when commi-

29


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

nuted biomass is stored, microbial activity will most likely take over. The first sign of this activity is heat generation (Kubler 1987, Nurmi 1990, Nurmi 1999, Jirjis 2005). The main reasons for the observed unresponsiveness of heat generation might be that the produced hog fuel was coarse, fines had been sieved away, wood material was dry, and buffer storage times were quite short. Jirjis (2005) pointed out that particle size has a strong impact on the storage properties of fuel chips, as it affects decay rate and durability: fines of less than 3 mm in length represent a health hazard because they reduce air circulation during storage, supporting bacteria proliferation with an increased risk of combustion. It is also known that chips made of fresh wood generate more heat and suffer greater dry material losses than if they are made of seasoned material (Björklund 1982, Kubler 1987). The geometry of the procurement area, the main and forest road network density, the availability of forest fuels, and the end-use facility location relative to the procurement area affect transportation distances (Ranta 2002, Ranta 2005, Ranta and Rinne 2006, Laitila et al. 2010, Tahvanainen and Anttila 2011, Anttila et al. 2013). In future, the transportation distances, especially in coastal areas, will increase along with the growth of forest fuel consumption in Finland, and it is a question of transporting raw materials from surplus areas to deficit areas using transportation modes suitable for long-distance transportation (Ranta 2002, Ranta and Rinne 2006, Laitila et al. 2010, Tahvanainen and Anttila 2011, Anttila et al. 2013). In addition, centralised buffer storage is characteristic of these modes, which may be vital, especially for large-scale power plants (Ranta and Rinne 2006, Kanzian et al. 2009, Laitila et al. 2010). Therefore, methods with better transport economy, fuel quality, and storage properties, such as screened hog fuel, will obviously gain in competitiveness in the future. The produced parameters reported in this study are valuable information when developing novel supply systems for stump wood procurement.

5. References Anerud, E., Jirjis, R., 2011: Fuel quality of Norway spruce stumps – influence of harvesting technique and storage method. Scandinavian Journal of Forest Research 26(3): 257–266. Anerud, E., 2012: Stumps as fuel – the influence of handling method on fuel quality. Doctoral thesis. Swedish University of Agricultural Sciences. 60 p. Anttila, P., Nivala, M., Laitila, J., Korhonen, K.T., 2013: Metsähakkeen alueellinen korjuupotentiaali ja käyttö (Spatial

30

analysis of the regional balance of potential and demand of forest chips for energy in Finland). Working Papers of the Finnish Forest Research Institute 267, 24 p. Asikainen, A., 2010: Simulation of stump crushing and truck transport of chips. Scandinavian Journal of Forest Research 25(3): 245–250. Backlund, C., 2007: Stump wood fuel in large scale industrial use. Bioenergy 2007, 3rd International Bioenergy Conference and Exhibition, 3rd–6th September 2007, Jyväskylä Paviljonki. Finbio Publications 36: 375–377. Björklund, L., 1982: Lagring av bränsleflis I fraktionen 25–30 mm (Storage of fuelwood chips in fraction 25–30 mm. Sveriges Lantsbruksuniversitet. Institutionen för Virkeslära. Uppsatser 115, 26 p. Dukes, C.C., Shawn, A.B., Greene, W.D., 2013: In-wood grinding and screening of forest residues for biomass feedstock applications. Biomass and Bioenergy (54): 18–26. Eliasson, L., Granlund, P., von Hofsten, H. Björheden, R., 2012: Studie av en lastbilsmonterad kross – CBI 5800 (Study of a truck-mounted CBI 5800 grinder). Arbetsrapport Från Skogforsk nr. 775 – 2012. 16 p. Eriksson, G., Bergström, D. Nordfjell, T., 2013: The state of the art in woody biomass comminution and sorting in Northern Europe. International Journal of Forest Engineering 24(3): 194–215. Erkkilä, A., Hillebrand, K., Raitila, J., Virkkunen, M., Heikkinen, A., Tiihonen, I., Kaipainen, H., 2011: Kokopuu ja mäntykantojen korjuuketjujen sekä varastoinnin kehittäminen (Developing the supply chains and storing of whole trees and Scots pine stumps). Tutkimusraportti VTT-R-10151-10. 52 p. Fogdestam, N., Granlund, P., Eliasson, L., 2012: Grovkrossning och sållning av stubbar på terminal (Coarse grinding of stumps and sieving of the produced hog fuel). Arbetsrapport Från Skogforsk nr. 768 – 2012. 9 p. Hakkila, P., 1975: Kanto- ja juuripuunkuoriprosentti, puuaineen tiheys ja asetoniuutteitten määrä. (Bark percentage, basic density and amount of acetone extractives in stump and root wood). Folia Forestalia 224. 14 p. Hakkila, P., 1976: Kantopuu metsäteollisuuden raaka-aineena (Stumpwood as industrial raw material). Folia Forestalia 292, 39 p. Hakkila, P., 2004: Developing Technology for Large-Scale Production of Forest Chips. Wood Energy Technology Programme 1999–2003. Technology Programme Report 6/2004. National Technology Agency. 98 p. Harstela, P., 1991: Work studies in forestry. University of Joensuu. Silva Carelica 18, 41 p. Hedman, L., 2008: Produktivitet vid stubbskörd (Productivity at stump harvesting). Department of Forest Resource Management, Swedish University of Agricultural Sciences, Report 219. 40 p. Croat. j. for. eng. 36(2015)1


Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32)

J. Laitila and Y. Nuutinen

Holzleitner F., Kanzian C., Höller N., 2013: Monitoring the chipping and transportation of wood fuels with a fleet management system. Silva Fennica 47(1): article id 899, 11 p.

Laitila, J., Ranta, T., Asikainen, A., 2008: Productivity of stump harvesting for fuel. International Journal of Forest Engineering 19(2): 37–46.

Jirjis, R., 2005: Effects of particle size and pile height on storage and fuel quality of comminuted Salix viminalis. Biomass and Bioenergy 28(2): 193–201.

Laitila, J., 2010: Kantojen korjuun tuottavuus (Productivity of the stump harvesting). Working Papers of the Finnish Forest Research Institute 150, 29 p.

Jouhiaho, A., Rieppo, K., Mutikainen, A., 2010: Kantoharan ja kantoharvesterin tuottavuus ja kustannukset (Productivity and costs of stump rake and stump harvester). TTS tutkimuksen tiedote. Luonnonvara-ala: Metsä 2/2010 (738) 8 p.

Laitila, J., Leinonen, A., Flyktman, M., Virkkunen, M., Asikainen, A., 2010: Metsähakkeen hankinta- ja toimituslogistiikan haasteet ja kehittämistarpeet (Challenges and development needs of forest chips procurement and delivery logistics). VTT Tiedotteita 2564. 143 p.

Jylhä, P., 2013: Autohakkurin seula-aukon koon vaikutus kokopuun haketuksen tuottavuuteen ja polttoaineen kulutukseen (The effect of screen mesh size on the productivity and fuel consumption of whole-tree chipping with a truckmounted chipper). Working Papers of the Finnish Forest Research Institute 272, 19 p. Kanninen, K., Uusvaara, O., Valonen, P., 1979: Kokopuuraaka-aineen mittaus ja ominaisuudet (Measuring and properties of whole tree raw material), 53 p. Kanzian, C., Holzleitner, F., Stampfer, K., Ashton, S., 2009: Regional energy wood logistics – optimizing local fuel supply. Silva Fennica 43(1): 113–128. Karlsson, J., 2007: Produktivitet vid stubblyftning (Productivity at stump lifting). Department of Forest Resource Management, Swedish University of Agricultural Sciences, Report 168. 52 p. Kärhä, K., Mutikainen, A., Hautala, A., 2010a: Vermeer HG6000 terminaalihaketuksessa ja -murskauksessa (Grinding and chipping at the terminal with Vermeer HG6000). Metsätehon tuloskalvosarja 15/2010. 37 p. Kärhä, K., Elo, J., Lahtinen, P., Räsänen, T., Keskinen, S., Saijonmaa, P., Heiskanen, H., Strandström, M., Pajuoja, H., 2010b: Kiinteiden puupolttoaineiden saatavuus ja käyttö Suomessa 2020 (Availability and the use of solid wood fuels in Finland in 2020). Työ- ja elinkeinoministeriön julkaisuja. Energia ja ilmasto. 66/2010. 68 p. Kärhä, K., Hautala, A., Mutikainen, A., 2011: Crambo 5000 kantojen tienvarsimurskauksessa (Crambo 5000 and crushing of stumps at roadside landings). Metsätehon tuloskalvosarja 4/2011. 48 p. Korpinen, O.-J., Ranta, T., Jäppinen, E., Hämäläinen, E., Laitila, J., 2007: Forest fuel supply chain based on terminals and stumps. In: Savolainen, M. (ed.). Bioenergy 2007. 3rd International Bioenergy Conference and Exhibition from 3rd to 6th of September 2007, Jyväskylä, Finland. Proceedings. FINBIOn julkaisusarja – FINBIO Publications 36: 399–404. Kubler, H., 1987: Heat generation process as cause of spontaneous ignition in forest products. Forest Products Abstracts 10(11): 299–322. Kuitto, P-J., 1984: Kantopuun korjuu kivennäismailla (Harvesting of stumps from mineral soils). Metsätehon Tiedotus 385. 16 p. Croat. j. for. eng. 36(2015)1

Laitila, J., Rytkönen, E., Nuutinen, Y., 2013: Kantojen, latvusmassan ja harvennuspuun murskaus tienvarsivarastolla kuorma-autoalustaisella CBI 5800 murskaimella (Grinding of stumps, logging residues and whole trees at roadside landing for fuel with a truck-mounted CBI 5800 grinder). Working Papers of the Finnish Forest Research Institute 260. 29 p. Laurila, J., Lauhanen, R., 2010. Moisture content of Norway spruce stump wood at clear cutting areas and roadside storage sites. Silva Fennica 44(3): 427–434. Lazdins, A., Von Hofsten, H., Dagnija, L., Lazdans, V., 2009: Productivity and costs of stump harvesting for bioenergy production in Latvian conditions. Jelgava, 28–29 May 2009, 194–201. Lindroos, O., Henningsson, M., Athanassiadis, D., Nordfjell, T., 2010: Forces required to vertically uproot tree stumps. Silva Fennica 44(4): 681–694. Markkila, M., 2005: Kannot energianlähteenä (Stumps as a source of forest energy). Kehittyvä puuhuolto 2005 – seminaari metsäammattilaisille, 16.–17.2.2005. Paviljonki, Jyväskylä. Presentation material. Nurmi, J., 1990: Polttohakkeen varastointi suurissa aumoissa (Longterm storage of fuel chips in large piles). Folia Forestalia 767. 18 p. Nurmi, J., 1999: The storage of logging residue for fuel. Biomass and Bioenergy 17(1): 41–47. Nuutinen, Y., Väätäinen, K., Heinonen, J., Asikainen, A., Röser, D., 2008: The accuracy of manually recorded time study data for harvester operation shown via simulator screen. Silva Fennica 42(1): 63–72. Nylinder, M., 1977: Upptagning av stubb- och rotved (Harvesting of stump and root wood). Skogsarbeten Redogörelse 5/77. 19 p. Nylinder, M., Thörnqvist, T., 1981: Lagring av stubbved I fingerad miljö (Storage of stumpwood in a simulated environment). Institutionen för skogsproduktion. Rapport 121. Sveriges landbruksuniversitet, Garpenberg. Ranta, T., 2002: Logging residues from regeneration fellings for biofuel production – a GIS-based availability and cost supply analysis. Lappeenranta University of Technology. Finland. Acta Universitatis Lappeenrantaensis 128. 180 p.

31


J. Laitila and Y. Nuutinen Efficiency of Integrated Grinding and Screening of Stump Wood for Fuel at Roadside ... (19–32) Ranta, T., 2005: Logging residues from regeneration fellings for biofuel production – a GIS-based availability analysis in Finland. Biomass and Bioenergy 28: 171–182.

Stampfer, K., Kanzian, C., 2006: Current state and development possibilities of wood chip supply chains in Austria. Croatian Journal of Forest Engineering 2006(2): 135–145.

Ranta, T., Rinne, S., 2006: The profitability of transporting uncomminuted raw materials in Finland. Biomass and Bioenergy 30(3): 231–237.

Strandström, M., 2013: Metsähakkeen tuotantoketjut Suomessa vuonna 2012 (Production chains of forest chips in Finland in 2012). Metsätehon tuloskalvosarja 4/2013. 24 p.

Rinne, S., 2010: Energiapuun haketuksen ja murskauksen kustannukset (The cost of wood fuel chipping and crushing). Lappeenranta University of Technology. Master’s thesis. 102 p.

Strelher, A., 2000: Technologies of wood combustion. Ecological Engineering (16): 25–40.

Röser, D., Mola-Yudego, B., Prinz, R., Emer, B., Sikanen, L., 2012: Chipping operations and efficiency in different operational environments. Silva Fennica 46(2): 275–286. Spinelli, R., Hartsough, B., 2001: A survey of Italian chipping operations. Biomass and Bioenergy 21(6): 433–444. Spinelli, R., Nati, C., Magagnotti, N., 2005: Harvesting and transport of root biomass from fast growing poplar plantations. Silva Fennica 39(4): 539–548. Spinelli, R., Visser, R., 2009: Analyzing and estimating delays in wood chipping operations. Biomass and Bioenergy 33(3): 429–433. Spinelli, R., Ivorra, L., Magagnotti, N., Picchi, G., 2011: Performance of a mobile mechanical screen to improve the commercial quality of wood chips for energy. Bioresource Technology (102): 7366–7370. Spinelli, R., Cavallo, E., Facello, A., Magagnotti, N., Nati, C., Paletto, G., 2012: Performance and energy efficiency of alternative comminution principles: Chipping versus grinding. Scandinavian Journal of Forest Research 27(4): 393–400.

Tahvanainen, T., Anttila, P., 2011: Supply chain cost analysis of long-distance transportation of energy wood in Finland. Biomass and Bioenergy 35(8): 3360–3375. Uusvaara, O., 1978: Teollisuushakkeen ja purun painomittaus (Estimation of industrial chip and sawdust weight). Folia Forestalia 341, 18 p. Uusvaara, O., Verkasalo, E., 1987: Metsähakkeen tiiviys ja muita teknisiä ominaisuuksia (Solid content and other technical properties of forest chips). Folia Forestalia 683, 53 p. von Hofsten, H., Granlund, P., 2010: Effectivare transport om stubbarna grovkrossas på avlägg (Haulage gains from crushing stumps to coarse chips at landing). Resultat från Skogforsk. Nr. 2/2010. 4 p. von Hofsten, H., Anerud, E., Fogdestam, N., Granlund, P., Eliasson, L., 2012: An alternative supply system for stump biomass – coarse grinding combined with sieving of the produced hog fuel (a manuscript and 5th paper of Erik Aneruds doctoral thesis) Ylitalo, E., 2013: Puun energiakäyttö 2012 (Wood consumption in energy generation in the year 2012). Metsätilastotiedote 15/2013. 7 p.

Authors’ address:

Received: November 20, 2013 Accepted: February 28, 2014

32

Juha Laitila, PhD.* e-mail: juha.laitila@luke.fi Nuutinen Yrjö, PhD. e-mail: yrjo.nuutinen@luke.fi Natural Resources Institute Finland (Luke) Yliopistokatu 6, FI–80101 Joensuu FINLAND * Corresponding author Croat. j. for. eng. 36(2015)1


Original scientific paper

Tree Damage in Mechanized Uneven-aged Selection Cuttings Matti Sirén, Juha Hyvönen, Heikki Surakka Abstract The amount of selection cuttings in uneven-aged forest stands is supposed to increase in Finland with the new Forest Act. In uneven-aged management, it is estimated that the cutting could be repeated every 15–20 years with the removal of around 100 m3/ha, depending on the site type and stand growth. This interval and volume highly depend on survival of lower canopy trees in cuttings. The number of these trees (2.5–15 m in height) is typically limited. Felling larger trees from above means a high damage risk for smaller trees and also restricts the harvesting outside the heavy frost period due to top damage risk. Damage to trees taller than 2.5 meters was studied in three selection cutting stands. Mechanized harvesting (harvester – forwarder) was carried out in late winter with no frost, which is the optimal time for selection cuttings. On the average 21.5% of the remaining trees were damaged. The percentage of damage to smaller (2.5–10 m) trees was highest, 28.4%. Stem damage and breakage were the most common types of injury. A logistic mixed model was used to model the probability of tree injury (uninjured/injured). Distance from the nearest removed tree, harvested basal area within 25 m of the tree and diameter of the tree were the explanatory variables taken into the model. The model discrimination ability by the ROC curve was 72.2%. With a classification cutpoint of 0.5 for the model fitted injury probabilities, the rate of correct classification was 79.1%. There is a need to develop optimal working practices for mechanized selection cuttings. Information on the stand structure, practical operator tutoring and knowledge of the goals of the forest owner are needed for successful harvesting implementation. Keywords: uneven-aged forest management, selection cutting, mechanized harvesting, tree damage, spatial analysis.

1. Introduction Current forest management in Finland is based on even-aged management. The use of alternative forest management methods including selection cuttings has been marginal concentrating to urban forests, landscape protection areas, valuable habitats, riparian and other buffer zones (e.g. Hyvän metsänhoidon suositukset 2006). The amount of selection cuttings in uneven–aged forests has been only some thousands of hectares per year (Selvitysraportti metsälain 6 §: n 2003). The structure of private forest ownership has changed and today 36% of forest owners live outside the municipality, where their forest property is located (Hänninen and Karppinen 2010). Many forest owners are not highly dependent of forest income anymore and emphasize multiple values in management deciCroat. j. for. eng. 36(2015)1

sions. Even more than a half of the forest owners are satisfied with the current forest management practices, one of six forest owners feels unsatisfied especially with clear cuttings, lack of management alternatives, soil preparation and damage caused by heavy machinery. The attitude towards uneven-aged forest management is positive. Near half of forest owners see a potential for uneven-aged management, and 27% of forest owners are ready to try it at least in a part of their forest property (Kumela and Hänninen 2011). Since 2010 there has been a renewal process of the Finnish Forest Act. The new Forest Act (http://www. finlex.fi/fi/laki/smur/1996/19961093) was set in the beginning of 2014, leaving more freedom and choices in forest management. The demand for alternative forest management is likely to increase, and forest own-

33


M. Sirén et al.

Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

ers as well as professionals need more information on effects and implementation of alternative management practices. Compared to even-aged management, our knowledge on alternative methods is vague. This lack of knowledge covers stand development and production, economy, ecology as well as harvesting. Kuuluvainen et al. (2012) present a thorough review of even-aged and uneven-aged forest management in Boreal Fennoscandia. They conclude that, although the number of relevant studies has increased in recent years, the ecological and economic performance of alternative management methods still remains insufficiently examined. The uneven-aged forest management consists of a range of methods, in which the forest cover is only partially removed. In this definition of uneven-aged management, the tree age distribution does not necessarily conform to the reverse J-shaped form and an uneven-aged forest can also consist of spatially segregated groups of tree age classes created by the group selection method. This broader group of uneven-aged management is also referred to as continuous cover forestry (Pommering and Murphy 2004). Pukkala et al. (2010) optimized the structure and management of uneven-aged stands in Finland. The post thinning diameter distribution of stands was optimized with 20-year cutting cycle when aiming at maximum economic profitability. Spruce stand optimizations were done for fertile and medium sites. The optimal post-thinning distributions had a truncated reverse J shape. The harvesting removal on fertile site was 145 m3/ha and 95 m3/ha on medium site. Optimization of economic result meant removing all logsized trees at 20-year intervals. Increasing discounting rate and decreasing site productivity improved the relative performance of uneven-aged management compared with even-aged management. Near 100% of harvesting carried out by the forest industry in Finland is done with the mechanized cutto-length method. In selection cuttings, harvesting carried out by the forest owner would be a proper alternative, making also the cuttings with small removals possible. However, the ability of forest owners living in towns to carry out cuttings themselves is limited. If selection cuttings are carried out on a larger scale, mechanized cutting is the main alternative. The efficient use of machinery needs sufficient removals, at least 70–100 m3/ha. In selection cuttings, the removal mainly consists of larger trees. Felling and processing of these trees means a high risk of damage to smaller trees and saplings. In uneven-aged stands, future development, as well as harvesting conditions, depend on the structure

34

of the stand, harvesting intensity, forest regeneration, but also on the amount of damage in harvesting. In Finland, Norway spruce (Picea abies (L.) Karst) stands have the highest potential for uneven-aged management on a commercial scale (Valkonen and Maquire 2005, Lähde et al. 2002), but they also have high risk of pathogen infections following harvesting damage (Hakkila and Laiho 1967, Isomäki and Kallio 1974). There is also a high risk of Heterobasidon root rot in all tree size classes in uneven-aged stands (Piri and Valkonen 2013). Knowledge on harvesting damage in selection cutting in Scandinavia is limited, and it has mainly focused on small saplings. Granhus and Fjeld (2001) found that the injury probability of saplings depends both on stand and operational characteristics, the most important factor being the interaction between these two variables. Sapling height and spatial distribution of saplings relative to the strip roads and larger trees of the residual stand represented stand characteristics, whereas operational characteristics were described by the operational method and harvesting intensity. Surakka et al. (2011) studied injuries on 0.5–2.5 m saplings. Depending on the stand, the percentage of injured saplings varied between 17.6–61.0%. The distance of the sapling to the nearest strip road, sapling height, harvested basal area within a distance of 25 m from the sapling and sapling distance to the nearest remaining tree explained the probability of injury. Saplings near the strip road and taller saplings were more prone to damage than saplings located further away from the strip road and small saplings. Earlier Fjeld and Granhus (1998) compared the effect of two operating systems (motor-manual cutting followed by cable skidding and one-grip harvester followed by forwarding) and three harvest intensities on the injury rate in multi-storied Norway spruce stands. The average injury rate was higher in mechanized than in motor-manual harvesting. The largest differences were at high harvest intensities in densely stocked stands. The average injury rate was 13% for small trees (diameter under 10 cm) and 7.5% for larger trees. The long term future of an uneven-aged stand depends on ingrowth, survival and height growth of small trees. The growth of small trees is very low, and with average growth rates, it takes about 60 years for a spruce germinant to achieve 1.3 m in height (Eerikäinen et al. 2014). Thus shorter term harvesting possibilities are highly based on survival of smaller and medium-sized trees in cuttings. If a significant percentage of these trees is damaged in harvesting, repeated cuttings every 15–20 years are not possible. Croat. j. for. eng. 36(2015)1


Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

2. Aim of the study The aim of this study was to evaluate the amount, type and quality of damage to trees taller than 2.5 m in mechanized selection cutting (cut-to-length method, harvester – forwarder) of uneven-aged Norway spruce stands and to construct a model for the description of damage.

3. Materials and methods

Table 1 Stand characteristics before and after cuttin Stand

Before harvesting, trees of commercially valuable tree species (i.e. Norway spruce, Scots pine (Pinus sylvestris L.), silver birch (Betula pendula Roth), downy birch (B. pubescens Ehrh) and aspen (Populus tremula L.) with a height of > 2.5 m were mapped (x and y coordinates) and measured for diameter at breast height (d). Tree heights were measured from a sample of 75–125 trees per stand. The sample trees were used to estimate the tree heights for the rest of the trees and the stem volumes for all the trees. The tree heights were estimated with the models by Näslund (1937), and the stem volumes with the models by Laasasenaho (1982) and Kärki et al. (1999). Trees to be felled were selected in three phases (Surakka et al. 2011): 1. trees located on the strip roads, 2. trees of weak health or poor technical quality, 3. Single tree selection from the remaining stand using a computerized tree selection procedure, with a classical negative exponential distribution (de Liocourt 1898) as the structural framework. The target basal area after harvesting was set at 20 m2/ha, and the trees to be removed were marked before cutting. Stand parameters before and after harvesting are summarized in Table 1 and diameter distributions are presented in Fig.1. Croat. j. for. eng. 36(2015)1

A

B

C

Volume, m /ha

289

296

295

Basal area, m2/ha

27.9

29.2

31.1

Stems/ha, h>2.5 m

766

1021

1262

Volume, m3/ha

185

200

173

Basal area, m /ha

18.3

20.3

18.7

Stems/ha, h>2.5 m

625

891

947

Volume, m3/ha

104

96

122

Basal area, m /ha

9.6

8.9

12.4

Stems/ha, h>2.5 m

141

130

315

3

Before cutting

Remaining stand

3.1 Study stands and measurements The harvesting experiments were carried out in three Norway spruce dominated stands located in Sounenjoki, Northern Savonia region. The sites represented the submesic Myrtillus type and the mesic Oxalis – Myrtillus in terms of Cajander (1909). The stands (A, B and C, 1.08, 0.85 and 0.42 hectares, respectively) were earlier harvested motor-manually in 1987. After that, the stands A and B were harvested again in 1999, A mechanically with a one-grip harvester and B motor-manually. All these harvesting operations aimed to an uneven-aged structure. The whole strip road network was already in place after these previous entries, with a few additions. The average distance between strip roads was 25, 27 and 23 m in stands A, B and C, respectively.

M. Sirén et al.

Removal

2

2

Cutting was carried out with a one-grip harvester Ponsse Ergo HS16 in late March and forwarding in early April 2007 with a Ponsse Buffalo forwarder. The temperature varied from –1 to +13 °C, snow depth from 0 to 20 cm, and visibility was good during the cutting. Two skilled harvester operators carried out the cuttings. The harvester and forwarder drivers were instructed to prevent damage to both trees and saplings. The harvester operators were instructed to fell the marked trees away from the strip roads. After cutting, the damage caused by the harvester was marked and inventoried to keep it separate from the damage caused by forwarding. After harvesting, the total number of remaining trees higher than 2.5 m was 1808. Injuries were assessed for each remaining tree. Trees were classified into a) uninjured, b) injured, will survive or c) fatally injured. Fatally injured consists of perished small trees, felled trees and trees broken near the ground. The damage type of a tree could be one of the following (for injury b) types 1–6 and for injury c) types 2 or 7 were possible): 1) stem damage including root collar damage, 2) stem breakage, 3) root damage, 4) crown damage, 5) tilt, 6) several types of damage or 7) disappeared. A stem or root damage was recorded, if the damage area was at least 1 cm2. Damage size was measured at two dimensions, lengthwise and crosswise. Distance from root collar to the beginning of the damage was measured for stem and root damage. Stem and root damages were further divided into bark damage (bark removed) and wood damage (wood smashed), depending on how deep the damage was. For stem breakage, the height of the breaking point was measured. Crown damage was recorded if the green crown loss was noticeable (more than 10% of

35


M. Sirén et al.

Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

Fig. 1 Diameter distributions of study stands green crown volume). A tilt was recorded if a tree tilted at least 10 degrees from its original (vertical) position. Trees that were not found after harvesting were entered as »disappeared«. Usually they were smaller trees often under piles of slash or logs. For the modeling purposes, we simplified the injury classification into two groups: uninjured and injured.

3.2 Modelling Variables affecting the probability of a tree being injured yij (a tree j sampled from a stand i) were explored using a logistic mixed model:

36

yij = 1, if a tree was injured (living or dead), probability = pij yij = 0, if a tree was uninjured, probability = 1–pij yij ~ Binomial (1,)  pij   = b0 + b1X1i + b2 X2ij + b3 X1i × X2ij + ... + ui (1) ln   1 − pij   

(

)

where pij is the modelled injury probability, ln is the natural logarithm function, β0, β1, β2, … are fixed effects parameters to be estimated, X1i, X2ij, … are stand-specific (i) or tree-specific (ij) explanatory variables (conCroat. j. for. eng. 36(2015)1


Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

M. Sirén et al.

Table 2 Statistics of tested possible explanatory variables for the injury model Variable

n

Minimum

Median

Mean

Maximum

Tree species

1808

0.00

1.00

0.82

1.00

Diameter

1808

1.00

10.80

14.28

90.00

Height

1808

2.54

11.60

12.73

35.50

Basal

1808

0.00

6.21

6.08

12.73

Stems

1808

0.00

96.77

105.78

325.95

Distance

1808

0.10

4.14

6.05

33.96

Distance_2

1808

0.65

5.93

7.51

34.95

Tree species – tree species group (1 = coniferous tree, 0 = broadleaved tree) Diameter – diameter of tree at breast height, cm Height – height of tree, m Basal – harvested basal area at distance of 25 m from the tree, m2/ha Stems – harvested number of trees at distance of 25 m from the tree, stems/ha Distance – distance of tree to the nearest removed tree, m Distance_2 – distance of tree to the centre of the nearest strip road, m

tinuous or dummy), »×« denotes an interaction effect, and ui is a random stand effect, ui ~ Normal (0, σ2u). Adding the stand as a random, categorical variable to the model takes into account the possible correlation of the tree observations within the stands. In this data, all the explanatory variables were tree-specific and continuous except the categorical variable tree species group (coniferous or broadleaved). To observe better the effect of spatial variation on a tree injury, several tree-specific explanatory variables were performed (Table 2) from the mapped tree data to describe the remaining stand around the tree. Various transformations of the continuous explanatory variables or interactions of the explanatory variables were also checked in candidate models. The model fit was assessed by the discrimination and the correct classification of the data, using the fixed part of the model to predict injury probability pij. From the model (1) we get:

pij =

(

injured (value 1 is closer), otherwise uninjured (value 0 is closer). The cutpoint 0.5 does not usually give the best correct classification result, but it tries to be an objective cutpoint. After the model classification the rate of correct classification could be counted by crosstabulating observed and predicted tree injuries. The analyses were carried out by the SAS statistical software, version 9.3. (SAS Institute Inc. 2014). The GLIMMIX procedure was used for the model estimation, the LOGISTIC procedure for the ROC curve calculation and the FREQ procedure for the cross-tabulation.

) ) (2) + b ( X × X ) + ...) (

exp b0 + b1X1i + b2 X2ij + b3 X1i × X2ij + ...

(

1 + exp b0 + b1X1i + b2 X2ij

3

1i

2ij

where exp is the exponential function. The area under the ROC curve is a measure of the discrimination ability of a statistical model for a binary response variable: it is the probability (or the percentage) that, for a randomly selected pair of an injured and uninjured tree, the model pij is greater for the injured one. The probability 0.5 was used as a cutpoint value for the model classification of an observed tree: if the model pij ≥ 0.5, the tree was classified Croat. j. for. eng. 36(2015)1

Fig. 2 Percentages of uninjured and injured trees by stands

37


M. Sirén et al.

Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

Fig. 3 Percentages of uninjured and injured trees by height classes

4. Results

Fig. 4 Percentages of injured trees in damage types by height classes

4.1 Amount, type and severity of injury Fig. 2 presents the percentage of trees of the stands A–C divided into three classes: uninjured, injured, will survive fatally injured. The percentages of trees in different height classes are shown in Fig. 3, and percentages of different damage types in Fig. 4. Stem damage and stem breakage were the most common damage types, together affecting near 70% of the damaged trees. Stem breakage or disappearing occurred mostly for lower canopy trees, and these trees formed the injury class »fatally injured«. The percentage of root damage was under 10% for all damaged trees, but nearly 50% for damaged trees taller than 20 m. Near 15% of the damaged trees had several types of

damage. Most damage, 88.4% of all damage, was caused in cutting. Forwarding caused 11.6% of damage, and near all root damage was caused in forwarding. The distribution percentiles of size distributions and locations of stem and root collar damage are presented in Table 3 and distribution percentiles of stem breakage heights in Table 4. The amount of damage was also calculated according to the classification system of the Forest Act, where only trees with d1.3 ≥ 7 cm are included and for superficial stem damage the minimum wound size is 12 cm2 under d1.3 or 30 cm2 in the whole tree (Fig. 5).

Table 3 Distributions percentiles (minimum, lower quartile, median, upper quartile and maximum) of stem damage area (cm2) and the distance to lowest damage point from root collar (dm, in parenthesis) by damage type and height of tree (m) Damage

Height of

type

tree, m

Bark

Wood

38

Lower

Minimum

2.5–10

84

1 (0)

8 (2)

15 (5)

45 (8)

380 (35)

10–20

62

2 (0)

8 (5)

20 (10)

48 (18)

1600 (78)

20–35.5

19

5 (0)

21 (4)

100 (9)

600 (27)

4200 (70)

2.5–10

22

15 (0)

36 (2)

120 (5)

210 (7)

300 (33)

10–20

10

6 (0)

60 (2)

155 (5)

320 (16)

560 (42)

20–35.5

3

20 (1)

20 (1)

75 (16)

400 (70)

400 (70)

quartile

Median

Upper

n

quartile

Maximum

Croat. j. for. eng. 36(2015)1


Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

M. Sirén et al.

Table 4 Distribution percentiles (minimum, lower quartile, median, upper quartile and maximum) of stem breakage height from root collar (dm) by height of tree (m) Height of tree, m

n

Minimum

Lower quartile

Median

Upper quartile

Maximum

2.5–10

100

0

0

0

20

95

10–20

30

0

4

53

90

200

20–35.5

1

200

200

200

200

200

degrees of freedom). The area under the ROC curve for the model was 0.722 (72.2%), which is considered acceptable discrimination ability by Hosmer and Lemeshow (2000). With a classification cutpoint of 0.5 for the model fitted injury probabilities pij using the formula (2), the rate of correct classification was 79.1% (uninjured 99.2%, injured 5.4%). The influence of the explanatory variables to the injury probability of a tree in the model is presented in Fig. 6.

5. Discussion

Fig. 5 Percentages of uninjured and injured trees by stands with injury classification according to the Forest Act

4.2 Model for spatial variations in the probability of injury In the modelling, there were 1420 uninjured and 388 injured trees. The following model was used to explain the injury probability of the tree (3):  pij   = −0.685 − 0.195 × DISTANCE − ln   1 − pij    − 0.040 × DIAMETER + 0.117 × BASAL

(3)

Where: DISTANCE – distance of tree to the nearest removed tree, m DIAMETER – diameter of tree at breast height, cm BASAL – harvested basal area at distance of 25 m from the tree, m2/ha All the explanatory variables were clearly significant in the model: p < 0.0001 (F-test with 1 and 1802 Croat. j. for. eng. 36(2015)1

The study material used in this study was collected in 2007 and was earlier used in modelling damage to 0.5–2.5 m saplings (Surakka et al. 2011). Almost the entire strip road network of the study stands was established in earlier cuttings, and thus there was very little removal from strip roads. Trees to be removed were marked before cutting, which is not typical in practical harvesting operation, and the harvester operators were instructed to fell the marked trees away from the strip roads. As the felled trees were mainly large, the possibility to lift them under felling to the direction of the strip roads would have been limited. However, utilizing the strip road openings in felling could have decreased the amount of damage to some extent. Miettinen (2005) simulated the effect of harvester working technique on damage risk for saplings in selection cutting on three permanent study plots (Eerikäinen et al. 2007), where the locations of larger trees and saplings had been mapped. A working method where all trees were felled into the stand away from the strip roads was compared with a method where trees nearer than five meters from the strip road were directed towards the strip road, also lifting them on the strip road when possible. When felling trees into the stand without utilization of the strip roads, 45% of saplings were exposed to damage. When strip roads were used, the percentage of damaged saplings was somewhat lower, at 38%. After harvesting, injuries were assessed for each remaining tree. This kind of total inventory, where all remaining trees on the area of 2.35 hectares were as-

39


M. Sirén et al.

Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

Fig. 6 Influence of harvested basal area at distance of 25 m from the tree (A = 5 and B = 10 m2/ha), diameter of tree at breast height (cm) and distance of tree to the nearest removed tree (m) on injury probability of tree sessed, is a laborious operation. However, when areas with suitable uneven-aged structure for the study were hard to find, total measurement of quite limited areas allowed collecting a proper material for modelling. Typically the damage inventories are made on sample plots. When quality of harvesting is followed by Finnish Forest Centre, circular sample plots of 100 m2 are used. When studying harvesting quality and tree damage in even-aged stands, Sirén (1998) used 240 m2 rectangular sample plots divided into eight 30 m2 measuring zones. In the study of selection cutting damage by Fjeld and Granhus (1998), the post-harvest inventory was based on 4 m wide and 24 m long rectangular sample plots stretching from strip road centre to strip road centre. Trees were classified into uninjured, injured, will survive and fatally injured. For small lower canopy trees the damage was often fatal, and the class fatally injured consists of small trees with severe damage. In most damage inventory methods (Eriksson 1981, Björheden and Fröding 1986, Sirén 1998), tree damage is monitored only for trees with commercial value. In this study, 78.5% of the trees had no damage, while 15.7 trees were damaged but will survive and 5.8% were fatally damaged. Near 90% of the damage was caused by the cutting operation. The trees with very severe damage were mainly 2.5–10 m tall lower canopy trees, which had stem breakage or had disappeared. In the study of Fjeld and Granhus (1998), the percentage of damaged trees was 8.7–13.7%. In their study the injury rate increased with harvesting intensity, and was higher in mechanized than in motor-manual cutting. The injury rates were highest near strip roads, where the injury rates with both methods exceeded 20%.

40

The probability of damage was explained by distance to the nearest removed tree, harvested basal area within 25 m from the tree and diameter of tree. The distance to the nearest removed tree best predicted the damage probability. The lower canopy trees have higher injury probability than larger trees. These explaining variables are logical. With increasing amount of work per area unit, the risk for damage increases. When larger trees are felled and processed, smaller nearby trees are at high risk for damage. The model for injury probability was able to correctly classify 79.1% of the trees as injured or not injured. In the model for sapling damage (Surakka et al. 2011), the rate of correct classification was 73.0%, and with model presented by Granhus and Fjeld (2001) 70.5%. In Forestry recommendations (Äijälä et al. 2014), the amount of tree damage is one element of silvicultural harvesting quality. Finnish Forest Centre measures annually the quality of more than 200 thinning stands, and the percentage of damaged trees in 2013 was 3.6 (Korjuujälkitarkastukset 2013), which is quite typical in long run. If the percentage is over 15%, it exceeds the limit of the Forest Act. Although there is a large variation between stands, the limit of the Forest Act is seldom exceeded in even-aged stands. In this study, the percentage of damaged trees after the classification of the Forest Act was 13.8%. As the amount of damage in good conditions carried out with skilled operators was near to exceed the limit, we can find that in selection cuttings it can be very difficult to reach low damage levels typical for even-aged stands. Also, new results (Hämäläinen 2014) on the amount of tree damage in uneven-aged stands show quite high Croat. j. for. eng. 36(2015)1


Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

damage numbers and high damage risk especially in lower canopy trees. High percentages of damaged trees with mechanized cut-to-length method have also been presented outside Scandinavia in even-aged stands. In Germany 12.6% of remaining trees were damaged (Sauter 1995). In North-America Bettinger and Kellogg (1993) reported very high damage numbers. In their study near 40% of remaining trees were damaged. Košir (2008) modelled the amount of tree damage with motor-manual and cut-to-length methods in Slovenia. With motor-manual method, the share of injured trees in thinnings was 17–19% and with cut-to-length method 13–15%. During a 160-year rotation period 10 thinnings take place, and the total number of damaged trees continuously grows and reaches 90% at the end of rotation. The modelled percentage of damaged trees is very high, but a high share of damaged trees (64–70%) has also been recorded in old stands (Košir 1998). There are many interesting study topics in selection cutting operations and working methods. As stand structures and goals of forest owners vary substantially, the harvester operator needs information on stand structure, slopes and soil bearing capacity. There are new interesting ways to utilize multi source information and operator tutoring in harvesting (Räsänen et al. 2014, Väätäinen et al. 2013). Selection cuttings are a challenging but interesting scene to test these tools.

6. References Bettinger, P., Kellogg, L.D., 1993: Residual stand damage from cut-to-length thinning of second growth timber in the Cascade Range of western Oregon. Forest Products Journal 43(11– 12): 59–64. Björheden, R., Fröding, A., 1986: Ny rutin för praktisk gallringsuppföljning. (A new routine for checking the biological quality of thinning in practice) Sveriges lantbruksuniversitet, Institutionen för skogsteknik. Uppsatser och Resultat 48. 14 p. Cajander, A., 1909: Über Waldtypen. (Forest site types) Acta Forestalia Fennica 1(1): 175. De Liocourt, F.,1898: De L´aménagement des sapinieres. (Spruce stand management) Bull. Soc. Franche-Comté et Belfort. Eerikäinen, K., Miina, J., Valkonen, S., 2007: Models for the regeneration establishment and the development of established seedlings in uneven-aged, Norway spruce dominated forest stands of southern Finland. Forest Ecology and Management 242(2–3): 444–461. Eerikäinen, K., Valkonen, S., Saksa, T., 2014: Ingrowth, survival and height growth of small trees in uneven-aged Picea abies stands in southern Finland. Forest Ecosystems 2014(1:5), 10 p. Eriksson, L., 1981: Stickvägar och körskador i gallringsbestånd. Resultat från Riksskogstaxeringens inventering åren Croat. j. for. eng. 36(2015)1

M. Sirén et al.

1978–1979 (Strip roads and damages caused by machines when thinning stands. Results from the Swedish National Forest Survey for 1978 and 1979). The Swedish University of Agricultural Sciences. Department of Operational Efficiency. Report No 137, 44 p. Fjeld, D., Granhus, A., 1998: Injuries after selection harvesting in multi-storied spruce stands – the influence of operating systems and harvest intensity. Journal of Forest Engineering 9(2): 33–40. Granhus, A., Fjeld, D., 2001: Spatial distribution of injuries to Norway spruce advance growth after selection harvesting. Canadian Journal of Forest Research 31: 903–913. Hakkila, P., Laiho, O., 1967: On the decay caused by axe marks in Norway spruce. Communicationes Instituti Forestalis Fenniae 64(3): 1–34. Hosmer, D.W., Lemeshow, S., 2000: Applied logistic regression, 2nd edition. Wiley series in probability and statistics, New York, 392 p. Hyvän metsänhoidon suositukset, 2006. (Forestry recommendations) Metsätalouden kehittämiskeskus Tapio, 100 p. Hämäläinen, J., 2014: Poimintahakkuun nykykäytännöt: työohjeistus, ajanmenekki ja korjuujälki (Practices in selection cuttings: operator tutoring, time consumption and harvesting damage) Pro gradu – work. University of Helsinki, Department of Forest Sciences, 46 p. Hänninen, H., Karppinen, H., 2010: Metsänomistusrakenteen muutos ja puuntarjonta. (Change in forest ownership and wood supply) In: Hänninen, R., Sevola, Y., (eds.). Finnish Forest Sector Economic Outlook 2010–2011: 52–55. Isomäki, A., Kallio, T., 1974: Consequences of injury caused by timber harvesting machines on the growth and decay of spruce (Picea abiea (L.) Karst). Acta Forestalia Fennica Vol. 136: 1–25. Korjuujälkitarkastukset 2013. (Harvesting quality survey 2013) Finnish Forest Centre bulletin. Košir, B., 1998: Damage to mountain spruce stands due to harvesting. Conference proceedings »Gorski gozd«, Ljubljana. Biotechnical Faculty, Department of Forestry and Renewable Forest Resources: 95–107. Košir, B., 2008: Modelling stand damages and comparison of two harvesting methods. Croat. j. for.eng. 29(1): 5–13. Kumela, H., Hänninen, H., 2011: Metsänomistajien näkemykset metsänkäsittelymenetelmien monipuolistamisesta. (Forest owners´opinions on alternative forest management practices) Working Papers of the Finnish Forest Research Institute 203, 76 p. Kuuluvainen, T., Tahvonen, O., Aakala, T., 2012: Even-aged and uneven-aged forest management in Boreal Fennoscandia: a review. AMBIO 41(7): 720–737. Royal Swedish Academy of Sciences. Kärki, T., Eerikäinen, K., Heinonen, J., Korhonen, K. T., 1999: Harmaalepän (Alnus incana) tilavuustaulukot. (Volume functions for grey alder) Metsätieteen aikakauskirja, 1/1999: 39–49. Laasasenaho, J., 1982: Taper curve and volume functions for pine, spruce and birch. Communicationes Instituti Forestalis Fenniae 109, 74 p. Lähde, E., Laiho, O., Norokorpi, Y., Saksa, T., 2002: Development of Norway spruce dominated stands after single-tree

41


M. Sirén et al.

Tree Damage in Mechanized Uneven-aged Selection Cuttings (33–42)

selection and low thinning. Canadian Journal of Forest Research 32: 1577–1584. Miettinen, A., 2005: Paikkatietoanalyysien soveltaminen eriikäisrakenteisten metsien hakkuiden tutkimuksessa. (Spatial analysis in uneven-aged forestry research) Päättötyö, paikkatiedonhallinnan erikoistumisopinnot, Hämeen ammattikorkeakoulu, 10 p. Näslund, M., 1937: Skogsföranstaltens gallringsförsök i tallskog. (Thinning trials in pine stands) Meddelanden från Statens Skogsförsökanstalt 29, 169 p. Piri, T., Valkonen, S., 2013: Incidence and spread of Heterobasidion root rot in uneven-aged Norway spruce stands. Canadian Journal of Forest Research 43(9): 872–877. Pommering, A., Murphy, S.T., 2004: A review of the history, de­finitions and methods of continuous cover forestry with spe­cial attention to afforestation and restocking. Forestry 77(1): 27–44. Pukkala, T., Lähde, E., Laiho, O., 2010: Optimizing the structure and management of uneven-sized stands of Finland: Forestry 83(2): 129–141. Räsänen, T., Hämäläinen, J., Lamminen, S., Lindeman, H., Salmi, M., Väätäinen, K., 2013: Uudet informaatiolähteet puunhankinnan tukena. (Support of new information sources for wood supply) Metsätehon raportti 279, 28 p. SAS Institute Inc. 2014. SAS/STAT 9.3 User’s Guide. (http:// support.sas.com/documentation/onlinedoc/stat/index. html#stat93). Sauter, U., 1995: Competing long and short wood cutting systems using harvesters for thinning conifer stands. In: Kellogg,

L., Milota, G. (eds.): The way ahead with harvesting and transportation technology. Proceedings of Iufro P3.04 meeting. Iufro XX World Congress. Tampere, Finland, 70–76. Selvitysraportti metsälain 6 §: Mukaisesta hakkuusta erityiskohteilla Maa-ja metsätalousministeriölle 8.12.2003. (Report on cuttings on specific areas after Forest Act 6 §) Forestry Development Center Tapio and WWF Finnish Fund. 20 p. Sirén, M., 1998: One-grip harvester operation, it´s silvicultural result and possibilities to predict tree damage. Metsäntutkimuslaitoksen tiedonantoja, 694, 179 p. Surakka, H., Sirén, M., Heikkinen, J., Valkonen, S., 2011: Damage to saplings in mechanized selection cutting in unevenaged Norway spruce stands. Scandinavian Journal of Forest Research 26(3): 232–244. The new Forest Act. (http://www.finlex.fi/fi/laki/ smur/1996/19961093). Valkonen, S., Maguire, D.A., 2005: Relationship between seedbed properties and the emergence of spruce germinants in recently cut Norway spruce selection stands in Southern Finland. Forest Ecology and Management 210(1–3): 255–266. Väätäinen, K., Lamminen, S., Ala-Ilomäki, J., Sirén, M., Asikainen, A., 2013: Kuljettajaa opastavat järjestelmät koneellisessa puunkorjuussa (Operator tutoring systems in mechanized harvesting) Working Papers of the Finnish Forest Research Institute 279, 24 p. Äijälä, O., Koistinen, A., Sved, J., Vanhatalo, K., Väisänen, P. (eds.) 2014: Metsänhoidon suositukset (Forestry recommendations) Metsätalouden kehittämiskeskus Tapion julkaisuja. 264 p.

Authors’ address:

Received: May 5, 2014 Accepted: September 2, 2014

42

Matti Sirén, PhD.* e-mail: Matti.Siren@luke.fi Natural Resources Institute Finland Vantaa, PB 18 01301 Vantaa FINLAND Juha Hyvönen, BSc. e-mail: Juha.Hyvonen@luke.fi Natural Resources Institute Finland Rovaniemi, PB 16 96301 Rovaniemi FINLAND Heikki Surakka, MSc. e-mail: Heikki.Surakka@ramboll.fi Ramboll Finland Oy, PB 25 02601 Espoo FINLAND * Corresponding author Croat. j. for. eng. 36(2015)1


Original scientific paper

Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting Benjamin Hatton, Guillaume Pot, Belhassen-Chedli Bouzgarrou, Vincent Gagnol, Grigore Gogu Abstract Delimbing is the process used to cut the branches off the trunk during tree processing by forest harvester. This process can be described as chipless cutting of green wood at a feed speed of 3 to 7 m s-1. This work aims to identify the parameters influencing the efficiency of the delimbing process. To this end, the main parameters are defined, and different experimental tests are presented. The first experiment was conducted using a dynamometric pendulum that can reach cutting speeds of 10 m s-1. A Digital Image Correlation method was used in order to compute the deformation field in the branch. The deformation fields observed are consistent with previous studies in the literature. The second experimental device was a slow speed test bench. It uses a hydraulic actuator to translate the knife through the branch while measuring force and displacement. Tests were conducted, varying the diameter of branches, to analyze its effect on the cutting force. Proportionality between branch area and cutting forces were verified, and empirical coefficients were obtained for both speeds. Keywords: cutting force, broad-leaved tree, dynamometric pendulum, digital image correlation

1. Introduction This paper presents a research work from the ECOMEF project (Eco-design of a mechanized equipment for hardwood harvesting), which aims to develop a harvester head more specifically designed to fell and process broad-leaved trees, in order to face the lack of motor manual workers and low rate of mechanization in hardwood logging operations (Cacot et al. 2006). This work focuses particularly on the modeling and simulation of the single-grip harvesters used in the cut-to-length method (CTL), like the Kesla 25 RH (Kesla Oyj 2013) harvester head (Fig. 1). The CTL method implies felling and processing of trees. The processing operation is composed of delimbing trunks and bucking them into logs (Ćuprić and Bajić 2009). In order to identify the models to be developed, a topdown functional analysis has been used. Four functions have been identified, associated with different phases of the CTL process: modeling the trunk gripping (function A1), feeding (function A2), delimbing Croat. j. for. eng. 36(2015)1

(function A3) and felling and bucking (function A4). These models presented in Fig. 2, which illustrates the central place of the delimbing force, are included in a closed loop between feeding and delimbing modeling. Delimbing force is also a key parameter in designing

Fig. 1 A Kesla 25 RH single-grip harvester head

43


B. Hatton et al.

Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

a harvester head and particularly in dimensioning the feeding equipment. The specificity of the ECOMEF project is the focus on broad-leaved trees. Due to crooked trunks, bigger branches with sharper angles and hardwood (Cacot 2009), these species are indeed harder to feed and delimb than conifers, and thus require specific tools to improve the energy efficiency and productivity of the operation. For these reasons, hardwood mechanization in France only reaches 4% of the total harvest, whereas it is at approximately 55% for coniferous forests. Previously, the European ForstINNO project emerged on similar observations, and allowed the development of the CTL-40HW harvester head (CTL Technology GmbH 2008). The absence of lower knife

and the geometry of the feed rollers gripping mechanism greatly increased the compactness of the head. The hydraulic control of the fixed knife also improved the feeding of crooked trunks, but disabled momentarily the delimbing function of this knife. However, the productivity announced, compared to existing solutions, (Suchomel et al. 2012, Mederski et al. 2011) suggest that room for improvement is still possible for more innovative heads in the field of hardwood harvesting. This will probably involve a better consideration of morphological characteristics of broad-leaved trees in the design of harvester heads for hardwoods. This work focuses on the experimental study of the delimbing process. Cutting the branches off the trunk is usually achieved by feeding the tree at 3 to 7 m s-1

Fig. 2 Top-down function analysis of CTL modeling

44

Croat. j. for. eng. 36(2015)1


Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

B. Hatton et al.

The following parameters are then defined: α = clearance angle β = sharpness angle γ = rake angle γr = real rake angle ε = implantation angle (angle of the grain direction with the cutting speed direction) e = blade thickness δ = oblique cutting angle Vc = cutting speed Ø = diameter of the branch

Fig. 3 Cutting angles and parameters using feed rollers, and creating an impact of the branches against delimbing knives. Unlike many cutting studies based on the type and section of a chip (Frantz 1958) for several machining operations (Woodson 1979), it is not possible in this case to define a chip, since the »chip« is the whole branch. To identify the cutting situation process involved, the definition proposed by McKenzie (1961) can be used. In this notation, two angles are used to define the cutting process. The first one is formed by the cutting edge with the grain direction, the second one by the cutting speed direction with the grain direction. Delimbing can thus be considered as a 90°–90° orthogonal cutting process, sometimes called cross-cutting. In the same way, chipping can follow the same definition. That is why the work of Abdallah (2011) can be adapted to define the cutting situation, as presented in Fig. 3a. However, δ angle is added to the proposed description to allow considering the case of oblique cutting (Fig. 3b). Croat. j. for. eng. 36(2015)1

At the best of our knowledge, only a few, mainly experimental, studies of the delimbing process have been conducted. Johnston and St-Laurent (1974) studied the delimbing of softwood and analyzed the influence of blade thickness, bevel ratio, wood species, cutting direction, bending of the branch, position of the branch on the trunk and temperature on the cutting force. Mattson and Sturos (1996) conducted a similar study of the delimbing of sugar maple and focused on the influence of branch diameter, cutting speed, blade thickness, oblique angle and sharpness angle on cutting force. Lately, Mikleš (2013) carried out experimental tests in order to optimize the knife geometry and minimize energy consumption while delimbing softwoods, and optimal values of clearance angle, sharpness angle and blade thickness were proposed. In these studies, several experimental devices were used. Some of them were developed specifically for the purpose of the tests, whereas others rely on well-known principles and architectures. Chardin (1958) considered the dynamometric pendulum as a low-cost and simple bench, able to accurately reproduce the machining processes. Eyma et al. (2005) also used such a device in order to simulate the routing process and observed good correlation of the observed phenomena between the router and the pendulum. More recently, Pfeiffer (2012) carried out tests on a dynamometric pendulum to reproduce the chipping process. The relationship between cutting force and branch diameter has been discussed in the past. Whereas a quasi-linear regression between these two variables was first considered (Dunfield 1971), later works focused on a quadratic regression, i.e. a linear regression between the branch area and the cutting force (Pszkit and Wiesik 1985, Mikleš 2013). In a similar way, Mattson and Sturos (1996) studied the influence of several parameters on the average cutting force, i.e. the cutting force divided by the branch diameter, and still noticed a positive influence of the diameter despite the normalization. In order to simulate the CTL process and to optimize the delimbing, a model has to be proposed, al-

45


B. Hatton et al.

Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

lowing to predict the cutting force needed to cut a branch for a given configuration (geometry of the system, mechanical and dynamical parameters known). The work presented here is an introduction to this study mainly based on experimental cutting tests, the first step being to understand the delimbing process in order to be able to model it.

2. Materials and methods All the tests were carried out on fresh wood of sessile oaks (Quercus petraea (Mattuschka) Liebl.) between July and September. Pole-size trunks were used instead of branches implanted on stems. This choice was made for setting and sample procurement convenience, and was motivated by the study of Mattson and Sturos (1996), who compared the cutting force obtained for actual branches and saplings and concluded that similar, though not identical, values of average cutting force were obtained for both types. Because of the complex fiber implantation at the base of the branches and the eventual resulting growth stresses, working on pole-size stems also allows an easier understanding of the observed phenomena by limiting the sources of unexplained differences (knots, different moisture content, etc.). Samples were taken from each stem. Both fresh and oven dry weights were systematically measured in order to determine moisture content of the wood. During the second campaign realized with the slow speed test bench, samples were better calibrated in order to know their fresh volume. This allows to calculate the specific gravity as the oven dry mass divided by the fresh volume of the sample (Williamson and Wiemann 2010). Even if this configuration is not really representative of the hardwood morphology, the implantation angle ε was chosen as close as possible to 90° in every test except for the deformation field analysis, i.e. the branch was almost perpendicular to the knife. For both test benches, the same cutting knife was used (Fig. 4). The knife was made of Hardox 400 wear resistant steel. The blade profile was inspired from existing cutting knives used on the Kesla 25 RH harvester head. The sharpness angle ß was 29°, and the rake angle 61°. The clear face was composed of three consecutive parts, the first and the last being parallel to the cutting plane, whilw the second one was oriented with a negative clearance angle of -12°, in order to keep the cutting edge away from the trunk and to avoid unexpected penetration of the knife into the trunk. For machining and fastening convenience, the decision was taken to work with a flat knife instead of a curved one as those used on real harvester heads.

46

Fig. 4 Profile of the knife (units: mm, deg)

2.1 Dynamometric pendulum To reproduce the cutting conditions met on a real harvester as accurately as possible, a dynamometric pendulum has been designed and developed as a first experimental device to study delimbing (Fig. 5). As explained by Eyma et al. (2005), this kind of bench is well known in the study of wood cutting, because of its simplicity and its ability to deal with interesting cutting speeds and impact phenomena. To be able to cut branches with diameters up to 100 mm, even at relatively low cutting speed, the pendulum had a 2.5 m and 68 kg arm, which can be equipped with up to 100 kg extra weight. The pendulum was designed as a foldable trailer in order to move it as close as possible to the fresh wood. Then, the trailer was anchored using chains and/or concrete blocks. An electric winch was used to reach the desired release position. An adjust-

Fig. 5 The dynamometric pendulum Croat. j. for. eng. 36(2015)1


Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

able clamping plate allowed the positioning of the branch and the setting of the implantation angle ε. By orientating the knife support relative to the arm, it was possible to modify the couple of angles γ and α (rake and clearance angles). The sharpness angle β was changed by machining new knives. An angle sensor measured the angular position of the arm. Due to its position on the rotation axis, this measurement was perturbed by the vibrations of the arm, and gave only information about the amplitude of oscillations. The precise position of the arm during the cut was computed using the autotracker feature of the OSP Tracker software (Brown and Christian 2011). The video used for autotracking was shot with an SMOTION high speed camera set up at 2000 frames per second. A force transducer could be used to measure the force applied on the cutting edge (Fig. 6). This sensor returned a voltage function of the hydraulic pressure (up to 600 bar) in a cylinder chamber placed just behind the knife. An appropriate calibration allowed the conversion of pressure to force. However, in the series of tests presented below, the force was computed using the equations of dynamics and the acceleration of the arm, obtained by smoothing the tracked position using a Savitzky-Golay filter (Savitzky and Golay 1964), that directly returns the position and its first (speed) and second (acceleration) derivatives, smoothed by fitting second degree polynomials on successive periods of 4.5 ms. A second study was performed to film the cutting of the branch with the same 8-bits camera and frame rate. Digital Image Correlation (DIC) using 7D software (Vacher et al. 1999) was used to measure full field displacements while delimbing. The DIC method consists of matching, before and after displacement, a set of Zones of Interest (ZOI) which make up a Region of

Fig. 6 The force transducer behind the knife Croat. j. for. eng. 36(2015)1

B. Hatton et al.

Interest (ROI) on the observed surface of the object. In the present study, ROI compares with about 400 (width) × 500 (height) pixels, depending on branches widths. ZOI were 12 × 12 pixels. This size of ZOI gives the best compromise between spatial resolution and resolution of the DIC method. Since the measured surface must be planar for DIC, branches were sawed longitudinally to obtain two half branches. Then, a small part was removed by a new longitudinal cut to obtain a tangential flat face, which helps maintaining the branch on the clamping plate (Fig. 7). It was the flat inside face of the branches that was observed with the camera, avoiding any out of plane effect. Black paint was sprayed on the sample surface before testing to obtain a speckle pattern and improve contrast. An image, taken before the knife begins to cut the branch, was referred to as the reference image, and the following images were correlated with this reference image using the DIC method. Displacements in longitudinal and radial direction were then computed using 7D software. These displacements can be noisy. Therefore, before their differentiation to obtain strains, displacements were smoothed by a moving average over adjacent ZOI. Then, planar deformations were calculated.

2.2 Slow speed test bench The second device was a linear test bench (Fig. 8). The knife was guided on two shafts using linear bear-

Fig. 7 Setting for Digital Image Correlation study

47


B. Hatton et al.

Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

Fig. 8 The slow speed test bench ings. The translation of the blade through the branch was actuated by a hydraulic ram without controlling displacement, force or pressure. The branch support allowed the modification of the implantation angle ε, and the translating support was designed to allow the modification of both oblique cutting angle δ and couple of angles (γ, α). A 0–125 kN range force sensor and an inductive displacement transducer were used to measure cutting force and penetration of the knife through the branch. Using this bench, the cutting speed was reduced (up to 20 mm s–1). This slow speed and the stiffness of the test bench ensured minimum vibrations and precise measures.

3. Results This section presents some experimental results on deformation fields and cutting force measured at various cutting speeds.

3.1 Deformation field The evolution of strains during the whole cutting process was measured by DIC. These measurements

enabled a quantitative study of the cutting process. Fig. 9 presents the planar deformations obtained with DIC on a branch with an 81° implantation angle and plotted on the corresponding picture. The chosen image is representative of the behavior mostly observed during the delimbing process. DIC did not give results in zones very close to the knife because of very high deformation of wood: pattern recognition was thus impossible for the software. However, it shows, interestingly, that deformation up to 2% occurred in zones distant from the knife. Indeed, in the direction of cutting (x direction Fig. 9a), a large zone of compression between the knife and the clamping plate appeared. It is due to the force of the knife in the x direction. There was also a localized strip of tensile deformations in x-direction above the knife, which follow the direction of the grain. This phenomenon is actually due to the growth of a crack parallel to the grain: the crack starts from the knife and propagates along the grain. These cracks appeared also below the knife, but their opening is less important. It can be seen in Fig. 9c that there were also strips of high shear strains. In the direction perpendicular to cutting (y direction Fig. 9b), strains are lower than in x direction. However, two zones can be distinguished: the zone in front of the cutting edge, which is principally in tension (about +0.5%, with higher values close to the tip), and the zone behind the cutting edge, which is mostly in compression (about –0.5%). Fig. 10 depicts what happens at the end of the cut. Another stage was reached, during which wood fiber failure below and ahead of the cutting plane can be observed. During this phenomenon, DIC was unable to compute matching of ZOIs because of huge displacements.

3.2 Cutting force Both test benches were used to measure cutting force and displacement of the knife through the branch during the cut. Fig. 11 and 12 present the force as a func-

Fig. 9 Deformation in direction x (a), y (b) and shear (c)

48

Croat. j. for. eng. 36(2015)1


Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

Fig. 10 Wood fiber failure during the second stage tion of dimensionless displacement for some tests of the two series that have been carried out, respectively, using the slow speed test bench and the dynamometric pendulum. The dimensionless displacement or unit displacement was calculated by dividing the displacement by the branch diameter in the direction of cutting, and was defined in order to be able to compare test results. The maximum force was reached for a displacement between 50% and 65% of the branch diameter (Fig. 11). This range was larger for the series of tests carried out with the pendulum (Fig. 12). Quickly after (or before)

B. Hatton et al.

the maximum force, a more erratic behavior was observed, corresponding to the second stage shown in Fig. 10. During this stage, the fracture frequently occurred very suddenly for the whole remaining part of the branch, which explains the fluctuation of the cutting force. The observation of the high-speed camera picture (Fig. 10), taken during the cuts with the pendulum, clearly showed that the fracture of the whole remaining part of the branch occurs almost instantaneously (at least really faster than the knife travel through the branch). Furthermore, this fracture takes place below the cutting plane and leads to the appearance of a stump on the branch. This phenomenon is not visible in Fig. 12 because of the smoothing realized on the raw data to compute the force for the dynamometric pendulum. Fig. 13 illustrates the proportionality obtained between maximum cutting force and branch area, mentioned in the literature about chipless cutting (Pszkit and Wiesik 1985, Mikleš 2013). Linear regressions for both data sets were computed. Maximum force Fmax as a function of branch area A can then be expressed for pendulum and slow speed as follows: Dynamometric pendulum:

Fmax = 520 × A + 1367

(1)

Slow speed test bench:

Fmax = 638 × A + 581

(2)

Both linear regressions provided good determination coefficient (R² = 0.85 for pendulum and 0.95 for

Fig. 11 Force as a function of dimensionless displacement (slow speed) Croat. j. for. eng. 36(2015)1

49


B. Hatton et al.

Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

Fig. 12 Force as a function of dimensionless displacement (pendulum) slow speed). The area considered is that of an ellipse, both the radii along and perpendicular to the cutting direction being measured and used as semi-major and semi-minor axes. In order to understand the difference between the gradients of the curves for both cutting speeds, it is important to notice that Moisture Content (MC) was not exactly similar for each series of tests. Table 1 presents the mean and standard deviation of MC for both speeds.

Table 1 Mean and standard deviation of Moisture Content for both cutting speeds Average MC

s (MC)

Slow speed

46.91

4.23

Dynamometric pendulum

63.05

5.91

Fig. 13 Maximum force as a function of branch area and associated correlations

50

Croat. j. for. eng. 36(2015)1


Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

3.3 Cutting force To study the influence of the speed, the scatter plot of the maximum cutting force as a function of the cutting speed is plotted in Fig. 14, and Pearson’s correlation coefficient has been calculated:

(

)

r = cov ( speed , Fmax ) / s ( speed ) × ( Fmax ) = −0.27 (3) were cov(speed, Fmax) is the covariance between speed and maximum cutting force.

B. Hatton et al.

if tensile deformations parallel to the grain are high enough to induce wood failure in tension. Finally, thanks to these full field measurements, it can be assumed that the theories developed by McKenzie (1961) in the frame of wood machining in 90°–90° situation can be used for the study of delimbing process. The trend of different cutting force curves was similar for both cutting speeds, even if there was more disparity between the curves obtained with the pendulum. This can be easily explained by the dynamic nature of the test, which causes naturally more vibrations and noises than the tests carried out at slow speed, which can be considered as quasi-static experiments. Considering the value of the Pearson correlation coefficient, it is difficult to conclude on the influence of the cutting speed on maximum cutting force. Furthermore, due to the lack of data of maximum force versus cutting speed, particularly in the range 2.5–7 m s–1, it is not possible to obtain concluding results on this influence of speed on the maximum cutting force. In traditional wood cutting, this effect has been discussed in the literature as presented in McKenzie (1961), but because of the shock produced during delimbing and the high moisture content of green wood, the cutting speed may affect the cutting force more than in other types of cutting. However, for a lower speed range (from 0.64 to 1.15 m s-1) than that considered in this study, the influence of the cutting speed has already been considered as non-relevant by Mattson and Sturos (1996).

Fig. 14 Maximum cutting force as influenced by cutting speed

5. Conclusions 4. Discussion Using DIC, the observation of strips of both shear and tensile deformations perpendicular to the grain in the zone where a crack is propagating is consistent with the interpretations of the 90°–90° cutting process: McKenzie (1961) assumed that the cracks along the grain in zones below and above the cutting edge are due to shear failure along the grain. As assumed in the case of wood machining (McKenzie 1961), the tension perpendicular to the cutting direction in front of the knife can be the cause of failure in the cutting plane, given the fact that wood failure by shear in this plane is generally considered impossible due to the anisotropy of wood and its great strength in this direction. However, this explanation needs further elaboration, in particular by more local DIC measurements just in front of the knife to verify Croat. j. for. eng. 36(2015)1

Using DIC, the strains involved in the delimbing process have been analyzed and the main failure modes in the study were compared with the literature. The present study is not considered sufficient to determine if the cutting speed is relevant or not as a parameter influencing the maximum delimbing force. Comparing the strains during the cut at both low (20 cm s-1) and high (5–10 m s-1) cutting speeds could be an interesting step towards answering this question and ensuring that further results (i.e. optimization of the blade geometry) obtained on the slow speed test bench will be transposable on the studied system. This first stage of understanding has been followed by a series of tests carried out on the two test benches (slow speed test bench and dynamometric pendulum). Experimental data were obtained, particularly force and displacement of the knife through the branch during the cut. The proportionality between maximum

51


B. Hatton et al.

Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

cutting force and branch area has been confirmed. This correlation will be the first step towards the modeling approach of the delimbing process. Indeed, first perspective of the work is the comparison of the area of the intersection of the knife and the branch to the force during the cut, to see if only maximum values are proportional or if a similar trend of that observed can be obtained throughout the penetration of the knife. Based on the collected data, no clear-cut conclusions could be drawn about the relevance of the cutting speed as a parameter influencing the maximum cutting force. Further investigations on this topic, but also on other parameters such as MC or SG, are needed to explain the different gradients observed between the two curves of maximum cutting force as influenced by the branch area. The influence of tree species should also be further investigated.

Acknowledgements This research work is part of ECOMEF project funded by Conseil Regional Auvergne and by the program Regional competitiveness and employment 2007–2013 (European Regional Development Fund – Auvergne region). These organizations are acknowledged for their financial support.

6. References Abdallah, R., Auchet, S., Méausoone, P.J., 2011: Experimental study about the effects of disc chipper settings on the distribution of wood chip size. Biomass and Bioenergy 35(2): 843–852. Brown, D., Christian, W., 2011: Simulating what you see. In: Proceeding book of the joint international conference MPTL 16 – HSCI 2011, 10–17. Cacot, E., 2009: Mécanisation de l’exploitation en feuillus: évolutions et questions (Mechanization of the hardwood exploitation: evolutions and questions). Forêt Wallonne 102: 34–44. Cacot, E., Bigot, M., Cuchet, E., 2006: Developing full-mechanized harvesting systems for broadleaved trees: a challenge to face the reduction of the manual workforce and to sustain the supply of hardwood industries. In: 29th Council on Forest Engineering (COFE) Conference Proceedings: »Working globally – Sharing forest engineering challenges and technologies around the world«. Chardin, A., 1958: Utilisation du pendule dynamométrique dans les recherches sur le sciage des bois (Usage of the dynamometric pendulum in wood sawing research). Bois et Forêts des Tropiques 58: 49–61.

52

Ćuprić, L., Bajić, V., 2009: Cut-to-length machines. In: Bošnjak, S., Zrnić, N. (eds) Proceedings of the XIX International Conference on »Material Handling, Constructions and Logistics«, University of Belgrade, Faculty of Mechanical Engineering, Serbia, 109–118. CTL Technology GmbH, 2008: Harvester technologie CTL – 40 HW. URL http://www.ctl-technology.de/images/prospekt/CTL-40HW.pdf, [Online; accessed 6-January-2014]. Dunfield, J., 1971: Annotated Bibliography on Delimbing of Trees. Information report, Canadian Forestry Service, Department of the Environment. Eyma, F., Méausoone, P.J., Larricq, P., Marchal, R., 2005: Utilization of a dynamometric pendulum to estimate cutting forces involved during routing. Comparison with actual calculated values. Annals of Forest Science 62(5): 441–447. Frantz, N., 1958: An analysis of the wood-cutting process. Tech. rep., Engineering Research Institute Ann Arbor. Johnston, J.S., St-Laurent, A., 1974: Force and energy required for removal of single branches. Tech. rep., Department of Forestry, Forest Products Laboratory. Kesla Oyj., 2013: Kesla harvester heads. URL http://www. kesla.fi/en/c/docu-ment_library/get_file?uuid=49b7f0e23b7d-407c-8044-bef802b6e3b2&groupId=10304, [Online; accessed 9-September-2013]. Mattson, J.A., Sturos, J.B., 1996: Reducing the forces required to delimb hardwoods. Tech. rep., U.S. Department of Agriculture, Forest Service, North Central Forest Experiment Station. McKenzie, W.M., 1961: Fundamental analysis of the woodcutting process. PhD thesis, Dept. of Wood Technology, School of Natural Resources, The University of Michigan. Mederski, P.S., Bembenek, M., Erler, J., Giefing, D.F., 2011: Effects of innovative thinning operation in a birch stand. Acta Scientiarum Polonorum. Silvarum Colendarum Ratio et Industria Lignaria 10(4): 29–38. Mikleš, J., 2013: Research into the geometry of the delimbing head of cutting knives. Research in Agricultural Engineering 59(1): 29–34. Pfeiffer, R., Denaud, L., Collet, R., Fromentin, G., 2012: Analysis of chips production by slabber. In: IUFRO Conference Division 5 – Forest products Proceedings. Pszkit, J., Wiesik, J., 1985: Studies of the resistances at debranching pine trees with passive knives. Sylwan 129(8): 17– 24. Savitzky, A., Golay, M.J., 1964: Smoothing and differentiation of data by simplified least squares procedures. Analytical chemistry 36(8): 1627–1639. Croat. j. for. eng. 36(2015)1


Experimental Determination of Delimbing Forces and Deformations in Hardwood Harvesting (43–53)

B. Hatton et al.

Suchomel, C., Spinelli, R., Magagnotti, N., 2012: Productivity of processing hardwood from coppice forests. Croatian Journal of Forest Engineering 33(1): 39–47.

Williamson, G.B., Wiemann, M.C., 2010: Measuring wood specific gravity... correctly. American Journal of Botany 97(3): 519–524.

Vacher, P., Dumoulin, S., Morestin, F., Mguil-Touchal, S., 1999: Bidimensional strain measurement using digital images. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science 213(8): 811–817.

Woodson, G., 1979: Tool forces and chip types in orthogonal cutting of southern hardwoods. Tech. rep., U.S. Department of Agriculture, Forest Service, Southern Forest Experiment Station.

Authors’ address:

Received: January 13, 2014 Accepted: May 6, 2014 Croat. j. for. eng. 36(2015)1

Benjamin Hatton*, dipl. ing. email: benjamin.hatton@ifma.fr Assoc. Prof. Belhassen-Chedli Bouzgarrou, PhD. email: belhassen-chedli.bouzgarrou@ifma.fr Assoc. Prof. Vincent Gagnol, PhD. email: vincent.gagnol@ifma.fr Prof. Grigore Gogu, PhD. email: grigore.gogu@ifma.fr Institut Pascal (UMR 6602 CNRS / UBP / IFMA), Campus de Clermont Ferrand les Cézeaux – CS 20265, 63175 Aubière cedex Assoc. Prof. Guillaume Pot, PhD. email: guillaume.pot@ensam.eu LaBoMaP, Arts et Métiers – ParisTech, rue Porte de Paris, 71250 Cluny FRANCE * Corresponding author

53



Original scientific paper

Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology Mika Yoshida, Hideo Sakai Abstract With the increase of importance of bioenergy, the whole tree harvesting system was reconsidered. As a simple way of a whole tree harvesting system, this study analyzed two winching procedures in terms of productivity and influence from terrain conditions. One of the winching procedures was observed on flat terrain, where the felled trees were scattered and then, one-by-one, attached to the winch rope. The other was observed on a steep terrain, where the felled trees were concentrated in a line and then attached, at one time, to the winch rope. A significant difference caused by slope conditions appeared only in the reeling velocity. When increasing the harvested volume in a cycle, the productivity increased. The difference in productivity between two sites was slight with the same harvested volume. By the one-by-one procedure, the full volume harvested in a cycle was only skidded from the nearest felling points, while the bundled procedure skidded the full volume from various points including the farthest point. The one-by-one procedure will save the winch rope tension and reduce engine acceleration. Fuel consumption, therefore, will be decreased by applying the one-by-one procedure instead of the bundled procedure. Additionally, reducing the weight of felled trees by natural drying in a forest is effective for reducing fuel consumption not only in transportation but also in winching operation, and for enhancing the calorific value of bioenergy. Keywords: fuel consumption, thinning, whole tree harvesting, winching, winch rope tension

1. Introduction FIT (Feed in Tariff) law came into effect on July 1st, 2012, in Japan and the price of electric energy from unused wood materials was set at 32 JPY/kWh (0.3Â USD/kWh, 1 USD = 101.6 yen, 2014/5/09), which was decided on the basis of the high logging cost and intended to encourage harvesting of unused wood materials. In this context, the whole tree harvesting method using cable harvesting systems had been reconsidered. A previous work that studied the relationship between material prices and production cost of line thinning emphasized the need to fully utilize felled trees in order to reduce the production cost (Yoshino et al. 2010). The whole tree harvesting method is suitable to fully utilization of trees because the method keeps the recovery rate at an operation site (Latia et al. 2010). Moreover, logging residues from whole tree harvesting can be a by-product of conventional timber harvesting, which covers the extraction cost of these residues (Stampfer and Kanzian 2006). EconomCroat. j. for. eng. 36(2015)1

ical profit from fuel utilization of unused wood materials, such as tree tops and slashes, is expected. On-ground winch harvesting is one of the basic harvesting systems. It can be applied to uphill and downhill harvesting on flat and steep slopes. In 2010, wheel tractors with a winch were introduced on both flat and steep slope areas in Japan, while until then the winch equipment had mostly been installed on excavators. Fei et al. (2008) pointed out that the high production cost and energy/fuel cost associated with mechanical harvesting of forest biomass were the major factors that could impede the use of forest biomass for energy. Multifunction tractors have various kinds of attachments and can reduce the harvesting cost by being used efficiently in many ways (Johanson 1997). Therefore, the winch system on a wheel tractor can be expected to realize lower harvesting cost especially in young forest stands. Moreover, such multi-functional characteristic seems to be appropriate to small scale operational sites because they have insufficient operational areas to keep working on a single operation.

55


M. Yoshida and H. Sakai Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology (55–61)

It is important to know the appropriate time standards to follow the proper harvesting procedures and to plan rationally the operations (Sowa and Szewczyk 2013). Besides the time standards, for above purposes it is also important to study the effect of slope conditions on harvesting productivity, as well as the winch harvesting procedure itself. In this study, two winching procedures were analyzed during the actual thinning operations performed by a tractor with a winch on flat and steep slopes. Winch harvesting productivity and operator’s walking speed on flat and steep terrains were analyzed based on results of time studies, and two types of winching procedure, from the aspect of load and fuel reduction, were evaluated.

2. Materials and methods 2.1 Materials Two typical different thinning sites, sites A and B, were investigated. Site A was located in a forest in Tsurui Village, Hokkaido Island, whose terrain was moderately flat. The surface of skidding corridors was covered with bamboo grass and logging residues (Fig. 1). On-ground winch harvesting was practiced by the tractor, Fendt Werner Wario 714 (96 kW), with Schlang & Reichrt's remote controlled double winches. One of the winch drums was equipped with a wire rope. Its diameter was 14 mm and the weight was 0.704 kg/m. The other, which was used during this investigation, was equipped with a textile rope. Its diameter was 14 mm and the weight was 0.1 kg/m. The maximum rope

length was 90 m for both drums. The tractive force of both winches was 8.2 tonnes. Trees were previously thinned and delimbed by chainsaw. Tree species was larch (Larix laptorepis) and it was easy to delimb because of fewer branches. The average volume of harvested trees was 0.28 m3. An operator harvested logs from the forest to spur roadside by the remote controlled winch in a tree length condition with the textile rope. The operator had two and a half years’ experience, and was mainly engaged in tractor operations for two years. Site B was located in a forest on a steep terrain in Kami City, Shikoku Island, where the average degrees of slope was 35 degrees. The forest surface was black soil. The tree species was Japanese cedar (Cryptomeria japonica) and the average volume of harvested trees was 0.50 m3. On-ground uphill winch harvesting was practiced by the tractor, John Deere 6930 (114 kW), equipped with Schlang & Reichrt’s remote controlled double winches – same as Site A. One of the two winch ropes was wire rope with 10 mm in diameter and the other was textile rope with 10 mm in diameter. The maximum rope length was 90 m. An experienced operator harvested logs from the forest to forest roadside in a whole tree condition with the wire rope. At site A, the felled trees were scattered because of selection thinning, whereas at site B, the felled trees were concentrated in a line. In this way, the felling situations and the harvesting procedure was different between sites A and B. For convenience, the harvesting procedure at sites A and B was named as »Procedure A« and »Procedure B«, respectively, not depending on terrain conditions.

2.2 Observation methods

Fig. 1 Surface conditions of Site A

56

Time of winch harvesting was measured by stopwatch to calculate productivities. Work elements of winch harvesting were classified into 6 elements: releasing, pulling the winch rope to logs; attaching, attaching felled trees to the winch rope; reeling, reeling the winch rope; detaching, detaching winch rope at roadside from logs; stop, all operational stops of winch due to search of logs and ways and others; and delay, all delays due to mechanical, personal and organizational causes. Operations were also recorded by a video camera. At site A, the position of the worker and tractor was recorded by GPS logger with one second interval to measure skidding distances because the felled trees were scattered and it was difficult to measure the distance during the operation. Skidding distances were calculated from the data of latitude, longitude and altitude referred as straight line. At site B, it was possible to observe the operation by eyesight because the felled trees were concentrated along a line and the understory vegetation was sparse. Skidding distances were Croat. j. for. eng. 36(2015)1


Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology (55–61) M. Yoshida and H. Sakai

measured by a laser range finder. Other components needed for the analysis were obtained by interview. The velocities of releasing and reeling operations were calculated from operation time and skidding distance. Average times of attaching and detaching operations and stop were also calculated. The data obtained from these observations were used in the following analysis.

2.3 Productivity calculation method Cycle time and productivity of winch harvesting, Cy (sec) and P (m3/h), were derived theoretically from equations (1) and (2), respectively, according to the maximum skidding distance based on the equations by Sakai and Kamiizaka (1985).

1 1 C y = D ⋅  +  + i ⋅ Ta + Td + Ts  v1 v2 

(1)

where D is the maximum winch skidding distance (m); v1 is the velocity of reeling (m/sec); v2 is the velocity of releasing (m/sec); i is the number of attaching operations in a cycle (i times/cycle); Ta is the time for attaching operation (sec); Td is the time for detaching operation (sec); and Ts is the time for stop (sec).

P=

3600 ⋅ j ⋅ V Cy

i , j =≤ 3

(2)

(3)

where j is the number of harvested trees in a cycle (j trees/cycle); and V is the average tree volume (V m3/tree). Productivities of both procedures on actual sites were calculated and also simulated to define when the average volume of trees V was set to 0.2 m3, which represented a young forest. The combinations of i and j were expressed as (i, j), and based on the observation it was assumed that a maximum of 3 trees could be harvested in a cycle. Thus, the possible combinations of (i, j) were (3,3), (2,2) and (1,1) for Procedure A, and (3,3), (2,3), (1,3), (1,2), (2,2) and (1,1) for Procedure B. The average of reeling and releasing velocities, and the average time of attaching, detaching and stop were used in this calculation.

3. Results The examples of GPS data are shown in Fig. 2. The average of Horizontal Dilution of Precision (HDOP) was 0.922, so that the GPS data was considered enough accurate for the analysis. The situation of Procedures A and B could be schematically illustrated as shown in Fig. 3 from this GPS data. In Procedure A, logs scattered around a skidding corridor were, one-by-one, attached to the winch rope during reeling operations. On the contrary, a couple of logs could be bundled by a sling rope and attached at one time in Procedure B. When harvesting in these typical situations, with 9 logs located at locations 1 to 3, the number of har-

Fig. 2 Two examples of GPS data of harvesting operation on Site A Croat. j. for. eng. 36(2015)1

57


M. Yoshida and H. Sakai Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology (55–61)

Fig. 3 Harvesting situation on flat terrain (site A) and steep slope (site B) vested trees and cycles were the same in Procedures A and B, as discussed later. At site A, among the total of 18 observed cycles, 17 cycles of winch harvesting operation lasting 1.3 hours were used to analyze the velocity because there was an error of GPS. The average winch skidding distance was 28 m. Average releasing and reeling velocities that were calculated from the time, including walking with logs, were 0.53 m/sec (n = 17, SD = 0.20, max = 1.05, min = 0.20) and 0.43 m/sec (n = 17, SD = 0.17, max = 0.78, min = 0.04), respectively. The average time for attaching operation and detaching operations were 18 sec (n = 53, SD = 13, max = 67, min = 2), and 22 sec (n = 18, SD = 14, max = 51, min = 6), respectively. Forty nine trees were hauled. The average number of attaching operation in a cycle was 2.94 times, and 2.72 trees were harvested in a cycle. The reason why the number of harvested trees was less than that of attaching operations was that attaching operations failed sometimes. The actual harvesting productivity of Procedure A at site A was 13.5 m3/h for the average tree volume of 0.28 m3. At site B, the total time of the observed winch harvesting operation was 42 minutes. Skidding distances of all 9 cycles were measured and the average distance was 18 m. In releasing, data of 8 cycles were used to calculate the average releasing velocity because there was an extra operation and it was difficult to separate it from releasing operation in a cycle. Data of 7 cycles were used to calculate the average reeling velocity be-

58

cause of the same reason as above. The average releasing and reeling velocities, calculated including walking time, were 0.67 m/sec (n = 8, SD = 0.34, max = 1.16, min = 0.22) and 0.26 m/sec (n = 7, SD = 0.14, max = 0.48, min = 0.09), respectively. The average time for attaching operation and detaching operation was 18 sec (n = 11, SD = 12, max = 49, min = 6), and 16 sec (n = 9, SD = 10, max = 31, min = 5), respectively. Eighteen trees were hauled. The average number of attaching operation in a cycle and harvested trees in a cycle was 1.22 times and 2 trees, respectively. The actual harvesting productivity of Procedure B at site B was 17.4 m3/h for the average tree volume of 0.5 m3. These data were summarized in Table 1 including the average times for stop. Differences in velocities and time consumptions between sites A and B were checked by t-test, and there was a significant difference in reeling velocity (p < 0.05), while there were no differences in releasing velocity and time consumptions of attaching, detaching, and stop (p > 0.05) although the experience of operators was much different. The productivity P on each site was simulated as shown in Fig. 4. The results showed that the operation was the most productive at both sites A and B when j was 3. The productivity decreased rapidly within the short skidding distance. As a matter of course, the productivity of Procedure B decreased with the increase of the number of attaching i, although the reduction was slight. Comparing the Procedures A (3, 3), B (3,3) Croat. j. for. eng. 36(2015)1


Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology (55–61) M. Yoshida and H. Sakai

Fig. 4 Harvesting productivity (for the number of attaching operation in a cycle – i; and the number of harvested trees in a cycle – j) cedure A (3, 3) and Procedure B (1, 3) rather than between Procedure A (3, 3) and Procedure B (3, 3). Multiple attaching operations are inevitable for Procedure A, so that such productivity reduction caused by multiple attaching operations should be taken into account when comparing Procedure A to Procedure B with a single attaching operation. Table 1 Averages of winch harvesting efficiencies Site

Site A

Site B

Releasing velocity, v1

m/sec

0.53

0.67

Reeling velocity, v2

m/sec

0.43

0.26

Time for attaching, Ta

sec

18

18

Time for detaching, Td

sec

22

16

Time for stopping, Ts

sec

68

57

Fig. 5 Comparison of productivity among Procedures A and B (operations in a cycle – i; the number of harvested trees in a cycle – j was 3)

4. Discussion

and B (1, 3), as shown in Fig. 5, it was recognized that the procedures with the maximum harvested volume could achieve high productivity in a short skidding distance even in a young stand with the average tree volume of 0.2 m3. The difference between the productivities was larger between the productivities of Pro-

Only in the reeling velocity, there was a significant difference between sites A and B in spite of similar engine horse powers. On the other hand, in a similar study of Imatomi (1997) that experimented releasing and reeling velocities on paved roads with four different terrain conditions, both velocities were affected by

Croat. j. for. eng. 36(2015)1

4.1 Difference between velocities

59


M. Yoshida and H. Sakai Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology (55–61)

terrain conditions. This difference seemed to be caused by the difference of skidding corridor conditions of two studies. The skidding corridors were covered with bamboo grass and logging residues at site A, and with slippery black soil at site B. The actual releasing velocity including operator’s walking might be affected not only by slopes but also by the condition of understory vegetation and surface soil.

4.2 Mechanical work of winch harvesting The reeling acceleration a (ms-2), and winching power p (kg) of a winch drum can be defined by the following equations (4) and (5) (Sakai and Kamiizaka 1985), 2⋅p ⋅R⋅r a= (4) h where R is effective diameter of winch drum (m); r is acceleration of the engine rotation (s-2); and η is the speed reduction ratio, and

p = W ⋅ ( m ⋅ cos q + sin q ) + d ⋅ w

(5)

where W is the weight of harvested logs (kg); μ is the coefficient of friction between soil and logs; θ is the degree of slope (where downhill skidding uses minus); d is the skidding distance (m); and w is the weight of rope (kgm–1). The mechanical work of winch harvesting F (N) was calculated by the following equation (6).

F = a ⋅ p

(6)

From equations (4) to (6), it is understandable that the mechanical work of winch harvesting becomes heavier as the load increasing. Additionally, the effective diameter of winch drum decreases as the extension of skidding distance, so that it is necessary to maintain the reeling speed by accelerating the engine rotation. Fuel consumption, therefore, increases as attaching logs and extending skidding distance during a cycle. When harvesting in the sample situation of Fig. 3, the mechanical work of Procedure A will be lighter than that of Procedure B because the full load occurs when harvested from only location 1 with Procedure A (3,3) while that occurs from further points as locations 2 and 3 with Procedure B (1,3). The procedure A is preferable in operation site B to reduce fuel consumption with keeping higher productivity especially when the maximum skidding distance exceeds about 40 m as shown in Fig. 5.

4.3 Winch harvesting productivity The results of productivity simulation show that the harvested volume in a cycle had the highest effect on productivity in both procedures. Therefore, to increase on-ground winch harvesting productivity, it is effective to increase the harvested volume in a cycle

60

even if the number of attaching operations in a cycle increases. Hence, it was concluded that when applying Procedure A, productivity could be improved by gradually increasing the number of attaching trees in a cycle within selection thinning conditions. The productivities of actual operations were higher than those of simulations because the volume of trees was larger than that in the simulations. It indicates that in order to improve productivity, the volume of trees should be increased. This fact is also supported by the result showing that the productivity in late thinning sites was higher than the productivity in early thinning sites when the maximum skidding distance was fixed at 50 m (Sowa and Szewczyk 2013). The upturn of winch harvesting productivities was rapid within 30 m of the maximum skidding distance so that such short skidding distance is appropriate for efficient winch harvesting. For harvesting at longer distances, the use of a mini tower yarder attachment for tractors was recommended (Spinelli et al. 2010). Its productivity did not vary much with longer skidding distances compared to on-ground winch harvesting. Harvesting productivity will be improved by using properly different attachments according to the maximum harvesting distance from the forest road.

5. Conclusion The number of harvested logs is the most effective factor for productivity of winch harvesting. The productivity was also affected by the soil and ground conditions in actual operations as well as geographical terrain conditions. However, the difference in productivity between sites A and B was slight within the skidding distance of 30 m when the number of attaching operation was the same. The important point is that high productivity can be achieved by winch harvesting even in a young forest within a short skidding distance and by increasing the harvested volume in a cycle. The number of tests was insufficient and more investigations should be made for accurate productivity estimation. However, we proposed a new method of winch harvesting related to felling operations. As the mechanical work will become heavier with the increase of the harvested volume in a cycle and as the extension of skidding distance, Procedure A with characteristic is one-by-one attaching operations will reduce fuel consumption of winch harvesting. Procedure A will, therefore, benefit the economy and ecology of winch harvesting especially at longer skidding distances from the aspect of saving fuel consumption, although the skidding distance is usually short in actual logging operations. Croat. j. for. eng. 36(2015)1


Winch Harvesting on Flat and Steep Terrain Areas and Improvement of its Methodology (55–61) M. Yoshida and H. Sakai

It is said that reducing log weight by natural drying is important for transporting bioenergy because the moisture in a wood material makes poor use of transportation energy and reduces the calorific value of raw material for bioenergy (Talbot and Suadicani 2006). Loosing the weight of logs also can bring economic and ecological benefits to winch harvesting. Harvesting planning including drying whole trees in the forest after felling should be considered.

Acknowledgment We are deeply grateful to Mr. Matsui, the president of Tsurui-Village Forest Owner’s Association; Mr. Monma, the manager; Mr. Inamori, the operator of the same Association; Mr. Nojima, the president of KamiCity Forest Owner’s Association; Mr. Morimoto and Mr. Ueda, the chief of operator team, Mr. Takemasa and Mr. Ishikawa, the members of operator team, for winch harvesting experiment. This study was partly aided by JSPS Grant-in-Aid for Scientific Research (C) No.2458214.

6. Literature Pan, F., Han, H., Johnson, L.R., Elliot,W. J., 2008: Net energy output from harvesting small–diameter trees using a mechanized system. For. Prod. J. 58(1): 25–30.

Imatomi, Y., 1997: An ergonomic study on the geometrical design and density of tractor skidding-roads. Bull. For. For. Prod. Res. Inst. 373:1–71. Johansson, J., 1997: Small tree harvesting with a farm tractor and crane attached to the front. J. For. Eng. 8(1):21–33. Latia, J., Heikkilä, J., Anttila, P., 2010: Harvesting alternatives, accumulation and procurement cost of small-diameter thinning wood for fuel in central Finland. Silva Fenn. 44(3): 465–480. Sakai, H., Kamiizaka, M., 1985: Prehauling systems for logs from thinnings. J. Jpn. For. Soc. 67(3): 82–91. Sowa, J.M., Szewczyk, G., 2013: Time consumption of skidding in mature stands performed by winches powered by farm tractor. Croat. J. For. Eng. 34(2): 255–264. Spinelli, R., Magagnotti, N., Lombardini, C., 2010: Performance, Capability and Costs of Small-Scale Cable Yarding Technology. Small Scale For. 9(1): 123–135. Stampfer, K., Kanzian, C., 2006: Current state and development possibilities of wood chip supply chains in Austria. Croat. J. For. Eng. 27(2): 135–145. Talbot, B., Suadicani, K., 2006: Road transport of forest chips: containers vs bulk trailers. Forestry studyes Metsanduslikud Uurimused 45: 11–22. Yoshino, S., Yabe, K., Sato, T., 2010: The possibility of wood biomass usage by line thinning, the case of the profit sharing forest of Tokyo University of Agriculture. J. Agric. Sci. Tokyo Nogyo Daigaku 55(1): 31–37.

Authors’ address:

Received: May 13,2014 Accepted: November 13,2014 Croat. j. for. eng. 36(2015)1

Mika Yoshida, MSc.* e-mail: yoshida@fr.a.u-tokyo.ac.jp Prof. Hideo Sakai, PhD. e-mail: sakaih@fr.a.u-tokyo.ac.jp The University of Tokyo Graduate School of Agricultural and Life Sciences Department of Forest Sciences Bunkyo-ku Yayoi 1-1-1 113-8657 Tokyo JAPAN * Corresponding author

61



Original scientific paper

Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran Fatemeh Berenji Tehrani, Baris Majnounian, Ehsan Abdi, Ghavammodin Zahedi Amiri Abstract Forest roads facilitate various activities such as forest management, tending of forest, timber logging, and fire and pest control, but the fact remains that roads can interrupt the natural function of forest ecosystem. They divert water flow and increase the amount of sediment. They can also alter plant species composition. Furthermore, the network structure of roads divides the land to small patches, which ends in habitat fragmentation. In this study, which was carried out in the Hyrcanian Forest in the north of Iran, effects of a forest road on plant species diversity (including trees, saplings and herbs) was investigated on both cut and fill slopes. At 10 points along the road, toward the fill slope and cut slope, ten 100 meter transects perpendicular to the road were established. Within each transect, ten 10Ă—10 m plots were sampled to record the tree and sapling species and ten 2 Ă— 2 m plots to record the herbal species. Results showed that road segment had no significant effects on plant species diversity. Since the studied road is constructed using environmentally sound techniques and criteria, unnecessary cut and fill operations were avoided. The other factor is the width of the road, which is as narrow as possible, so the habitat fragmentation is not so considerable. The traffic on this road is also limited; therefore soil pollution does not affect plant composition. These items together with the ability of the Hyrcanian ecosystem to repair itself can mitigate negative effects of this road. Keywords: Caspian Forest, road effects, plant species diversity, habitat fragmentation, environmentally sound techniques, Iran

1. Introduction Forest roads provide access to a variety of activities including forest management, tending of forest, timber logging, and fire and pest control. Roads are also associated with economic growth and national wealth (Wikie et al. 2000), but they have various direct and indirect effects on their adjacent environment. In fact, since the roads are external factors imposed on nature, they can have some profound ecological effects. Roads are accompanied by ecological disturbances and landscape degradation, and they introduce broad changes in ecosystem structure and function. They may alter the natural composition of the forest ecosystem, such as plant species composition and diversity (Findlay and Houlahan 1997, Buckley et al. 2003, Godefroid and Koedom 2004, Marchand and Houle 2005, Avon et al. 2010). Habitat destruction, introduction of exotic species and changes to the physical and chemical environment (Trombulak and Frissel 2000, Formann 2000) can Croat. j. for. eng. 36(2015)1

be cited as some ecological disorders. Moreover, the network structure of roads divides a large forest land into smaller patches, which results in some degree of habitat fragmentation. Roads are a barrier for natural drainage and, by cutting the natural flow, they divert water and disturb the natural aquatic balance in the region (Burroughs et al. 1972, Megahan 1972, Rummer 1997, Wemple 1998). Typically the amount of sediment will be increased (Reid and Dunne 1984, Luce and Cundy 1994, Rummer et al. 1997, Connolly et al. 1999, Luce and Black 1999, Elliott 2001, Fransen et al. 2001, Kahklen 2001, Appelboom et al. 2002, Jha et al. 2006, Lopez et al. 2008, Baihua et al. 2009).

2. Overview of studies Many different investigations have been done concerning road ecology and the effects of roads on plant species. Some of them are summarized below.

63


F. B. Tehrani et al.

Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

According to the results by Avon et al. (2010) in young and adult oak stands in French lowland forest with a long history of management and road construction, plant composition strongly differed between road verge and forest interior habitats. The main road effect extends less than 5 m into the forest stand. Non forest species were not observed from the forest interior, while many bryophytes and vascular plants kept away from the road. According to the results obtained by BernhardtRomermann et al. (2006) in the Munich-area, Southern Germany, motorways have an impact on the vegetation composition in the neighbourhood of roads. Depending on wind direction, influences of motorways reaches up to 230 m on the downwind side and up to 80 m on the upwind side. A study of Godefroid and Koedom (2004), carried out in the Sonian forest, south of Brussels, Belgium, showed that forest paths have significant effect on the surrounding plant assemblages, and they increase the number of ruderal species, disturbance indicators, nitrogen-demanding indicators and indicators of basic conditions. Buckley et al. (2003) evaluated impacts of haul roads and skid trails on understory vegetation. The results showed that understory plant richness was significantly greater in haul roads than in skid trails and forest, as a result of significantly greater percentages of introduced species (13%) and wetland species. Skid trails had a greater percentage of wetland species (9%) than in forest, but differences in richness between skid trails and forest were not statistically significant. Results obtained by Findlay and Houlahan (1997) in Southeastern Ontario, Canada wetlands showed a strong positive relationship between wetland area and species richness. The species richness was negatively correlated with the density of paved roads on lands up to 2 km from the wetland.

3. The problem area In this study, the effects of a forest road on plant species diversity, which includes trees, saplings and herbs, was investigated so that some environmentally friendly principles for forest road construction and maintenance can be achieved in order to alleviate their negative effects and conserve the natural cycle of the forest ecosystem. This study specifically addresses the following questions:  Does the forest road affect the plant species diversity?

64

 Are there any significant differences between plant species diversity on cut slope and fill slope by the effects of road?  What are the environmentally friendly principles to alleviate the roads negative effects?

4. Area of study The study was conducted in the educational and experimental forest of the University of Tehran (Kheyrud Forest). It is located on the northern slopes of the Albourz Mountains, approximately 7 km east of the port of Noshahr (36°34’–36°37’N, 51°32’E). The area of the forest is about 8000 ha, and the altitude ranges from 0 m to 2200 m from sea level. This broad range of altitude is the basis for different plant communities. The plant communities are as listed below (Marvie Mohajer 2005):  Querco-Buxetum, Pterocaryo-Alnetum

0–100 m

 Querco-Carpinetum, Parrotio-Carpinetum

100–700 m

 Fagetum hyrcanum Rusco-Fagetum 700–1800 m Vaccinio-Fagetum  Quercetum macranthera, Carpinetum oriental

1800–2200 m

Kheyrud forest has seven districts. The current research was done in the lowest district, which is called Patom. It covers 900 ha and its altitude ranges from 0 m to 934 m above sea level. The Patom district has 18 parcels. Five parcels are protected and 13 parcels are productive and under logging since 1969. The plant community in this region is mostly Querco-carpinetum, that is to say Quercus castaneifolia and Carpinus betulus are characteristic species, and the other tree species are mostly Diospyros lotus, Fagus orientalis, Acer velutinum, Alnus glutinosa. (Tab. 1, 2, 3). The road in Kheyrud forest was constructed in the period 1971–1994 with the density of 21 m/ha, and in the study area (Patom district) the forest road route was constructed in 1972–1973. The width of the road is 5–6 m, which includes roadbed with the width of 3.5–4 m, two one-meter-ditches, and two banks each half a meter wide. Material for the road is from the mine and the river bed: mineral sand and gravel of the region. However, now riverbed exploitation for road building is banned. Some parts of the road, which has clay soil, were first fixed by the lime stabilization method. Croat. j. for. eng. 36(2015)1


Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

F. B. Tehrani et al.

Table 1 Herbal species No.

Scientific name

Points along the road

1

Oplismenus undolatifolius

10–9–8–7–6–5–4–3–2–1

invasive

2

Euphorbia amygdaloides

10–9–8–7–6–5–4–3–2–1

invasive

3

Carex sylvatica

10–9–8–7–6–5–4–3–2–1

endemic

4

Pteridium aquilinum

10–9–8–7–6–5–4–3–2–1

invasive

5

Epimedium pinnatum

10–9–8–7–6–5–4–3–2–1

endemic

6

Rubus spp.

10–9–8–7–6–5–4–3–2–1

semi-endemic

7

Urtica dioica

10–9–6

endemic

8

Hedera pastoshawii

10–9–8–7–6–5–3–2–1

endemic

9

Phyllitis scolopendrium hill

10–9–5–3–2–1

endemic

10

Pteridium aquilinum

10–9–8–7–5–4–3–2–1

invasive

11

Mentha longifolia

10–7

endemic

12

Ruscus hyrcanus

10–9–8–7–6–5–4–3–2–1

semi-endemic

13

Hypericum androsaemum

9–8–7–6–4–3–2–1

endemic

14

Cyclamen coum

5–3–2–1

endemic

15

Convolvulus sp

4

Indefinite*

16

Smilax excelsa

7–5–4–1

endemic

17

Equisetum ramosissimum

4

endemic

18

Potentilla reptans

8–7–1

endemic

19

Malva sp.

1

Indefinite*

20

Circaea lutetiana

7–6–2–1

endemic

* Since the species is not identified, it is not possible to recognize according to genus

Table 2 Shrub species No.

Scientific name

Points along the road

1

Crataegus spp.

10–9–2–1

endemic

2

Crataegus spp.

10–9–8–7–6–5–4–3–2–1

endemic

3

Prunus divaricata Ledeb

9–1

endemic

4

Prunus divaricata Ledeb

9–5–4–3–2–1

endemic

5

Mespilus germanica

9–5–3–2–1

endemic

6

Mespilus germanica

9–8–7–6–5–4–3–2–1

endemic

7

Ilex spinigiera

10–9–8–7–6–5–4–3–2–1

endemic

5. Materials and methods In this study, at 10 points along the road, toward fill and cut slope, ten 100 meter transect were sampled from the top of the embankment and established perpendicular to the road. Ten 10 × 10 m plots to record Croat. j. for. eng. 36(2015)1

the trees and shrubs species and ten 2 × 2 m plots to record the herbal species were set up along each transect (Fig. 1). The species diversity of cut and fill slope for each macro and micro plot was analysed by the Shannon

65


F. B. Tehrani et al.

Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

Table 3 Tree species No. 1 2

3 4 5 6

7 8 9 10 11 12 13

Scientific name Frangula alnus Diospyros lotus Carpinus betulus Fagus orientalis Acer velutinum Acer cappadocicum Tilia platyphyllus Alnus glutinosa Parrotia persica Ulmus glabra Ficus carica Quercus castaneifolia Fraxinus excelsior

Points along the road 4 10–9–8– 7– 6– 5– 4– 3– 2–1 10–9–8– 7– 6– 5– 4– 3– 2– 1 10–9–8–7–6–5–4–3–2–1 10–9–8–7–6–5–4–3–2–1 10–9–8–7–6 10–9–7–5 10–7–6–4 10–6–5–4–1 10–9–8 10–8–7 6–5–4–2 4

endemic endemic endemic endemic semi-endemic endemic endemic endemic endemic endemic endemic semi-endemic semi-endemic

Fig. 1 The position of plots scaled to road index1 in PAST software2 (Hammer and Harper 2006, Hammer et al. 2001, Harper 1999). Then, normality of data was tested by Kolmogorov- Smirnov of SPSS

1

2

t he Shannon index (H) is a diversity index that is commonly used to characterize species diversity in a community. Shannon’s index accounts for both abundance and evenness of the species present. The proportion of species i relative to the total number of species (pi) is calculated, and then multiplied by the natural logarithm of this proportion (lnpi) a software for scientific data analysis, with functions for data manipulation, plotting, univariate and multivariate statistics, ecological analysis, time series and spatial analysis, morphometrics and stratigraphy.

66

software. The significance of the effect was tested by ANOVA, and then the comparison of means was tested by Duncan.

6. Results and Discussion In this study, as above said, two questions should have been addressed. Does the forest road affect the plant species diversity? What are the environmentally friendly principles to alleviate the roads negative effects? According to the results, the diversity of trees, saplings and herbs are not influenced by road (distance Croat. j. for. eng. 36(2015)1


Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

F. B. Tehrani et al.

from road) on both cut (Fig. 2, 4, 6) and fill slope (Fig. 3, 5, 7). The only significant difference for tree species diversity was recorded at the first plot on the cut slope, set up between the embankment and 10 meters (Tab. 4). The number of tree and herb species is not influenced by road (Fig. 8, 9, 12, 13), but the number of

sapling species on the cut slope shows significant differences (Fig. 10). To answer our first research question, these results showed that the studied road had no significant effect on the number of species and diversity, which is not in accordance with the results of Avon (2010), Marchand and Houle (2005) and Houlahan (1997), who con-

Fig. 2 Variation in tree species diversity from road to interior forest on cut slope

Fig. 4 Variation in sapling species diversity from road to interior of forest on cut slope

Fig. 3 Variation in tree species diversity from road to interior of forest on fill slope

Fig. 5 Variation in sapling species diversity from road to interior of forest on fill slope

Croat. j. for. eng. 36(2015)1

67


F. B. Tehrani et al.

Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

Fig. 6 Variation in herb species diversity from road to interior of forest on cut slope

Fig. 8 Variation in the number of tree species from road to interior of forest on cut slope

Fig. 7 Variation in herb species diversity from road to interior of forest on fill slope

Fig. 9 Variation in the number of tree species from road to interior of forest on fill slope

cluded that the plant species diversity changes with the distance from the road. The only significant differences in the present study are for tree species diversity on cut slope at the first plot (between the embankment and 10 m from the road) and the number of saplings on the cut slope. This result can be due to the following reasons: firstly, the Kheyrud forest road has not only been planned, designed and constructed by road construction standards, but the environmentally

friendly principles have also been considered, especially in design. These environmentally friendly principles include the minimum dimensions of travel way, right of way, sub grade, curves and radius and avoiding straight lines and huge curves. That is to say that nature has not been sacrificed by the road, and that this road which is synchronized and accorded with the topography and nature. Also, local materials were used for construction: mineral sand and gravel of the

68

Croat. j. for. eng. 36(2015)1


Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

F. B. Tehrani et al.

Fig. 10 Variation in the number of sapling species from road to interior of forest on cut slope

Fig. 12 Variation in the number of herbal species from road to interior of forest on cut slope

Fig. 11 Variation in the number of sapling species from road to interior of forest on fill slope

Fig. 13 Variation in the number of herbal species from road to interior of forest on fill slope

region from the mine in the forest and the river bed, which is an important factor in order to not disturb the local soil composition by introducing new materials. Secondly, the width of the road is as narrow as possible, which is a critical factor in maintaining the ecological balance. In the other words, the new patches made by the road are almost as large as their former size. As a result, the connection between two patches is still retained, and this connection can be a guarantee

to prevent habitat fragmentation. This is proved by the study of Hilty et al. (2006). This is completely in accordance with the results provided by the studies of Laurance (2000) and Fahrig (2009), who showed that more species in the newly made habitat provided more chance for conserving the native species. The significant difference in the number of saplings, which are fewer on the first plot from the road, might be due to the grazing by cattle, which is more considerable on

Croat. j. for. eng. 36(2015)1

69


F. B. Tehrani et al.

Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71)

Table 4 The result of Duncan’s test Significance Cut slope

Fill slope

Trees

0.045

0.079

Saplings

0.091

0.068

Herbs

0.157

0.171

the road verge. On the other hand, the snow and rainfall are more intense on the road verge than in the interior forest, which can cause more damage to the saplings near the road. Thirdly, the limited traffic of this road can be mentioned as another beneficial item for reducing the adverse effects of the road. This last observation needs more investigation, but according to the study of Godefroid and Koedam (2004), pollution from traffic can affect plant composition. Fourthly, the ability of the Hyrcanian ecosystem to repair itself should not be ignored. In this ecosystem, the growing season is about 300 days per year, so it would be long enough to repair all the unnatural disturbs. Therefore, our second research question was answered by all the items mentioned, which can mitigate the negative effects of this road, when compared with previous studies, in which the roads are wider and less synchronized with nature (Coffin 2007).

7. Conclusion In the experimental area, plant species diversity does not show any significant changes related to distance from the road. According to the structure of the studied road, it can be concluded that to limit the effects of the road on adjacent forest, environmentally sound techniques and criteria should be considered as well as road construction standards. In other words, nature should dictate the route to conserve the natural habitat by avoiding unnecessary cut and fill operations. As the size of retained patches is an important factor influencing species survival, the width of the forest road should be as narrow as possible to mitigate the disconnection between two patches. All in all, considering the items analysed in this paper can prevent major hazardous effects, so the Hyrcanian forest ecosystem can proceed along its natural path of development.

Acknowledgements The author wishes to thank all reviewers for their worthwhile comments, especially Prof. Raffaele Ca-

70

valli and Dr. Gary Blank. In particular, the author appreciates Professor Baris Majnounian, Dr. Ehsan Abdi and Dr Ghavamodin Zahedi Amiri for their helpful advice.

8. Literature Appelboom, T. W., Chescheir, G. M., Skaggs, R. W., Hesterberg, D. L., 2002: Management practices for sediment reduction from forest roads in the coastal plains. Transactions of the Asae 45: 337–344. Avon, C., Bergès, L., Dumas, Y., Dupouey, J. L., 2010: Does the effect of forest roads extend a few meters or more into the adjacent forest? A study on understory plant diversity in managed oak stands, Forest Ecology and Management 259: 1546– 1555. Baihua, F., Newham, L. T. H., Field, J., 2009: Integration of a road erosion model, WARSEM, with a catchment sediment delivery model, Catch MODS. International Congress on Modeling and Simulation (MODSIM 2009), ed. Anderssen, R. S., Braddock, R. D., Newham, L. T. H., Modeling and Simulation Society of Australia and New Zealand Inc., Australia, 4085–4091. Bernhardt-Romermann, M., Kirchner, M., Kudernatsch, T., Jakobi, G., Fischer, A., 2006: Changed vegetation composition in coniferous forests near to motorways in Southern Germany: the effects of traffic-born pollution. Environ. Poll. 143: 572–581. Buckley, D. S., Crow, T. R., Nauertz, E. A., Schulz, K. E., 2003: Influence of skid trails and haul roads on understory plant richness and composition in managed forest landscapes in Upper Michigan. USA, Forest Ecology and management 175: 590–520. Burroughs, E. R., Marsden, M. A., Haupt, H. F., 1972: Volume of snowmelt intercepted by logging roads, J. Irrig. Drain. Div. Am. Soc. Civ.Eng. 98: 1–12. Coffin, A. W., 2007: From road kill to road ecology: A review of the ecological effects of roads. Journal of Transport Geography 15: 396–406. Connolly, R. D., Costantini, A., Loch, R. J., Garthe, R., 1999: Sediment generation from forest roads: bed and eroded sediment size distributions, and runoff management strategies. Australian Journal of Soil Research 37(5): 947–964. Elliot, W. J., Hall, D. E., Graves, S. R., 1999: Predicting sedimentation from forest roads. J. For. 97: 23–29. Fahrig, L., Rytwinski, T., 2009: Effects of Roads on Animal Abundance: an Empirical Review and Synthesis. Ecology and Society 14(1): 21 p. Findlay, C. S., Houlahan, J., 1997: Anthropogenic Correlates of Species Richness in Southeastern Ontario Wetlands. Conservation Biology (published online 2003) 11: 1000–1009. Forman, R. T. T., 2000: Estimate of the area affected ecologically by the road system in the United States. Conserv. Biol. 14: 31–35. Croat. j. for. eng. 36(2015)1


Impacts of Forest Road on Plant Species Diversity in a Hyrcanian Forest, Iran (63–71) Forman, R. T. T., Deblinger, R. D., 2000: The ecological roadeffect zone of a Massachusetts (USA) suburban highway. Conserv. Biol. 14: 36–46. Fransen, P. J. B., Philips, C. J., Fahey, B. D., 2001: Forest road erosion in New Zealand: overview. Earth Surface Processes and Landforms 26: 165–174. Godefroid, S., Koedam, N., 2004: The impact of forest paths upon adjacent vegetation: effects of the path surfacing material on the species composition an soil compaction. Biological Conservation 119: 405–419. Hammer, Ø., Harper, D. A. T., 2006: Paleontological Data Analysis. Blackwell Publishing, 351 p. Hammer, Ø., Harper, D. A. T., Ryan, P. D., 2001: PAST: Paleontological Statistics Software Package for Education and Data Analysis. Palaeontologia Electronica 4(1): 9 p. Harper, D. A. T. (ed.), 1999: Numerical Palaeobiology. Computer-Based Modelling and Analysis of Fossils and their Distributions. John Wiley & Sons, 468 p. Hilty, J. A., Lidicker, Jr. W. Z., Merenlender, A. M., 2006: Corridor Ecology, the Science and Practice of Linking Landscapes for Biodiversity Conversation, Island Press, 323 p. Jha, S., Western, A., May, D., Turner, J., 2006: A Monte Carlo analysis of sediment load from unsealed forest road crossings. Land and Water Management, 2390–2395. Kahklen, K., 2001: A method for measuring sediment production from forest roads. USDA Forest Service. Pacific Northwest Research Station, Research Note, PNW-RN-529, 17 p. Laurance, W. F. 2000: Do edge effects occur over large spatial scales? Trends in Ecology and Evolution 15: 134–135. López, A. J., Zavala, L. M., Bellinfante, N., 2008: Impact of different parts of unpaved forest roads on runoff and sediment yield in a Mediterranean area. Science of the Total Environment 937–944.

F. B. Tehrani et al.

Luce, C. H., Black, T. A., 1999: Sediment production from forest roads in western Oregon. Water Resources Research 35(8): 2561–2570. Luce, C. H., Cundy, T. W., 1994: Parameter identification for a runoff model for forest roads. Water Resources Research 30(4): 1057–1069. Marchand, P., Houle, G., 2005: Spatial patterns of plant species richness along a forest edge: What are their determinants? Forest Ecology and Management 223: 113–124. Marvie Mohajer, M. R., 2005: Silviculture. University of Tehran Press, 387 p. Megahan, W. F., 1972: Subsurface flow interception by a logging road in mountains of central Idaho. In Proceedings, National Symposium on Watersheds in Transition, American Water Resources Association: Fort Collins, CO, 350–356. Noss, R. F., 1987: Corridors in real landscapes: A reply to Simberloff and Cox. Conservation Biology 1:159–164. Reid, L. M., Dunne, T., 1984: Sediment Production from Forest Road Surfaces. Water Resources Research 20: 753–1761. Rummer, R. B., Stokes, B., Graeme, L., 1997: Sedimentation associated with forest road surfacing in a bottomland hardwood ecosystem, 195–200. Trombulak, S. C., Frissell, C. A., 2000: Review of ecological effects of roads on terrestrial and aquatic communities. Conserv. Biol. 14: 18–30. Wemple, B. C., 1998: Investigations of runoff production and sedimentation on forest roads. PhD. dissertation, Department of forest science, Oregon State University, Corvallis, 168 p. Wilkie, D., Shaw, E., Rotberg, F., Morelli, G., Auzel, P., 2000: Roads, Development, and Conservation in the Congo Basin. Conservation Biology 14(6): 1614–1622.

Authors’ address:

Received: July 28, 2013 Accepted: July 8, 2014 Croat. j. for. eng. 36(2015)1

Fatemeh Berenji Tehrani, MSc.* e-mail: fb.tehrani@yahoo.com University of Tehran, Department of Forestry Unit 23-no. 28-Sheydayi st. Yakhchal st. Shariati st. Tehran IRAN Prof. Baris Majnounian, PhD. e-mail: bmajnoni@ut.ac.ir Assist. Prof. Ehsan Abdi, PhD. e-mail: abdie@ut.ac.ir Assoc. Prof. Ghavammodin Zahedi Amiri, PhD. e-mail: ghavamza@ut.ac.ir University of Tehran, Department of Forestry Karaj-31585 IRAN * Corresponding author

71



Original scientific paper

Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance Steven A. Weaver, Zennure Ucar, Pete Bettinger, Krista Merry, Krisha Faw, Chris J. Cieszewski Abstract The ability to measure trees remotely or at a distance may be of value to forest inventory processes. Within three forest types (young coniferous, old coniferous, and deciduous), we compared laser caliper measurements collected at distances up to 12 m from each tree, to direct contact caliper measurements. Bitterlich sector-fork measurements and diameter tape measurements were also collected for reference purposes. We used non-parametric tests to evaluate three of our four hypotheses that suggest there are no significant differences between direct and remote diameter measurements, between caliper measurements and sector-fork measurements, and between diameter measurement errors across forest types. In general, most of the differences in diameters were small (≤ 0.8 cm) and were observed within the 0–6 m measurement distance from each tree. These results suggest that forest characteristics and measurement distance may play a role in remote diameter measurement accuracy. We also performed a correlation analysis between light conditions and remote measurements. The correlation analysis suggested light conditions were not significantly correlated to diameter measurement accuracy. Keywords: dendrometer, precision forestry, Bitterlich sector-fork, HaglÜf Gator Eyes, laser caliper

1. Introduction Examinations and tests of analog and digital tools for measuring tree diameters (dendrometers) have been reported in the literature for nearly 100 years. The main concerns associated with forest sampling procedures when using these instruments relate to accuracy, efficiency, economy, and rationality (Rhody 1975). Sophisticated instruments have been devised to measure trees from a distance or remotely (e.g. Henning and Radtke 2006) and to measure trees with special characteristics, such as fluted basal swells (e.g. Parresol and Hotvedt 1990). A range of results have been presented in comparing measurements of diameter directly obtained by using calipers or tapes. In some cases, practically no importance has been associated with the choice of instrument (Behre 1926). In other cases, the differences between two types of measurements have been very small (Krauch 1924), while others have found the differences to be statistically significant (Binot et al. 1995). Some have even suggested Croat. j. for. eng. 36(2015)1

that the most accurate method is one that involves direct measurements of inside bark diameter (Chacko 1961). Although most dendrometers can provide estimates of outside bark diameters that are adequate for a number of field inventory applications, when minor differences between measured tree diameters have been found among dendrometers, these differences can translate into significant variations in tree volume estimates (Parker and Matney 1999). In addition to precision dendrometers that are strapped or affixed to a tree (e.g. Yoda et al. 2000, Drew and Downes 2009), panoramic (Rhody 1975) and wide angle photography (Clark et al. 2000b) have also been tested for their ability to assist in diameter measurements. Optical sensor systems that use lasers have also been developed to count and determine the sizes of trees (Fairweather 1994, Delwiche and Vorhees 2003). A machine vision system has been recently tested that, through the detection of illuminated line segments, can count stems and determine diameters (Zhang and

73


S. A. Weaver et al.

Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

Grift 2012). Further, tree diameters have been correlated with measurements obtained through the use of Lidar (Popescu 2007). Skovsgaard et al. (1998) found that remote measurements tended to overestimate tree diameters by 2 to 5%, with increasing deviations as measurement distance from a tree increased. On the other hand, Nicoletti et al. (2012) found that the two optical dendrometers tested tended to result in an underestimation of stem biomass. Williams et al. (1999) also noted that the variability of measurements increases with the distance from a tree. While the sophistication of remote methods is increasing, results generated by some of these methods can be affected by the inability of a sensor to locate blocked tree stems or measurement errors arising from stem and bark irregularities (Bell and Groman 1971). For practical purposes, dendrometers need to be inexpensive, precise, and easy to use (Kalliovirta et al. 2005). Some efficient and reliable instruments may be expensive, complex (e.g. Parker 1997), or too heavy (e.g. Liu et al. 1995) for regular field work. Laser dendrometers might be suitable for use in practical forestry applications, yet the accuracy of the devices needs to be tested under typical operating conditions. In our case, we are interested in the ability of the laser calipers to accurately provide estimates of tree diameters from distances up to 12 m, the approximate radius of a circular inventory plot (0.04 ha) commonly used in the southern United States. The accuracy of some types of laser dendrometers may be associated with distance from a tree, measurement time, and tree diameter. We tested three dendrometers, a diameter tape, the Haglöf Gator Eyes system mounted on an 18-inch Mantax Black caliper (when collecting diameters at a distance, remotely, hereafter called the laser caliper), and the Bitterlich sektorkluppe (hereafter called the sector-fork). A diameter tape measures the girth of a tree and estimates the quadratic mean diameter of a tree measured from all possible directions. A caliper measures the distance between parallel tangents of closed convex regions to arrive at an estimate of a diameter in a selected direction, and a sector-fork uses principles of perspective geometry to arrive at an estimate of a diameter also from a selected direction (Clark et al. 2000a). In contrast to the Laser-relascope used by Kalliovirta et al. (2005), there is no relationship between the position of the dendrometer (when in use) and a person’s eye with the laser caliper; therefore theoretically, the laser caliper should be more user-friendly than other laser dendrometer devices. As with previous evaluations (e.g. Skovsgaard et al. 1998), our study is concerned with detecting pos-

74

sible bias when using remote and direct (contact) instruments for measuring outside bark tree diameters. The objectives of this research were to determine the relative consistency in measurements obtained using different techniques, remotely and directly, and whether there were significant differences between these. We attempt to examine several hypotheses: H1: There is no significant difference between direct and remote laser caliper measurements of tree diameters. H2: There is no significant difference between caliper (direct and remote) measurements and sector-fork measurements of tree diameters. H3: Light conditions have no significant effect on tree diameter measurements. H4: There is no significant difference between tree diameter measurement errors for data collected in different forest types.

2. Methods Repeated measurements are necessary for obtaining statistical stability and for assessing accuracy and precision (Bruce 1975). For this study, one hundred trees were randomly selected within each of three forest types; an older deciduous (Quercus spp., Carya spp., Ostrya virginiana, and others) forest (60–70 years old), an older coniferous (Pinus echinata, Pinus taeda) forest (60–70 years old), and a young coniferous (Pinus taeda) forest (18 years old). These three forest types had different characteristics (Table 1) and diameter distributions (Fig. 1), and thus were included in this study to assess differences with light conditions (as a function of tree density and canopy closure) and bark characteristics (as a function of tree species, as suggested by Liu et al. 2011). At the location of the study area (the University of Georgia Whitehall Forest, in northeast Georgia, USA), these are also the only main forest types present. Data were collected in the afternoon for Table 1 Characteristics of the forested test areas Forest type

Approximate age, years

Basal area, m2 ha–1

Stem count, trees ha–1

Canopy closure, %

Young coniferous

18

35.4

1,495.3

93

Old coniferous

65

22.9

303.4

85

Deciduous

65

26.2

421.7

94

Croat. j. for. eng. 36(2015)1


Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

S. A. Weaver et al.

12 days (4 days per forest type) during October and November 2012. Light conditions ranged between 101–60,000 lux with an average of 4,906. We based our sample size, where the sample units were trees to measure in each of the three forest types, as a compromise between time availability and estimated precision of the population mean. Of primary interest to us was the difference between the direct caliper measurement and the other measurements made with the caliper at a distance. The computation of the deviation in diameter values, DEVidj, or the deviation between the direct measurement and the measurement made for tree i at distance d in forest type j. DEVidj = DBHi0j – DBHidj

(1)

DBHi0j represents the direct caliper measurement for tree i in forest type j. DBHidj, which could either be smaller or larger than DBHi0j, represents the caliper measurement for tree i in forest type j, collected at distance d. After collecting 100 samples, the standard deviation of these values was computed to determine whether the sample size was appropriate. The standard deviation for each forest type (j) and each distance (d) was thus computed using the following formula:   n  n  ∑ DEVidj  ∑ DEV 2 −  i=1 idj  i=1 n     sdj =  n−1      

(

)

(

)

2

           

0.5

∀d , j

(2)

We assessed the sample size required for each distance and forest type (ndj), assuming a desired 95% confidence interval, using the following sample size formula, 2

 1.96 sdj  ndj =   E   

(3)

where sdj represents the standard deviation for deviations in values found at distance d in forest type j. The value E represents an assumed objective for estimating the population mean deviation in values between direct measurements and measurements collected at a distance (i.e. to within a certain number of units, represented by E). When we assumed an objective of estimating the population mean deviation to be within Croat. j. for. eng. 36(2015)1

Fig. 1 Diameter distributions of the young coniferous, older coniferous and deciduous test areas 0.15 cm, we found that 100 samples was sufficient. This assumption (0.15 cm) was at worst, about 33% of a single standard deviation representing the difference in direct and remote measurements. In only one case (the deciduous forest at the 12 m distance) was the suggested sample size larger than 100 trees (102 trees). This sample size (tree count) was also consistent with other recent work in the southern United States (Parker and Matney 1999, Liu et al. 2011). In all cases, the selected trees were measured along their stem to collect the diameter at breast height (DBH) outside bark at 1.37 m above ground. We used masking tape to mark the location just below where DBH would be measured so that measurements would all be made at the same place on each tree at the edge of the actual bark. Each tree was visited one time during the study period to collect all seven diameter measurements. Three measurements involved directly touching each tree (diameter tape, sector-fork, and Mantax Black caliper), and the other four involved single remote measurements of DBH with the laser caliper along a consistent line of sight from the tree at 3 m, 6 m, 9 m, and 12 m (Fig. 2). These distances from each tree were marked on the ground with wire stake flags. The sector-fork and direct (0 m) caliper measurements were also made along this same line of sight. Measurements were made in order of diameter tape, sector-fork, 0 m, 3 m, 6 m, 9 m, and 12 m caliper measurements. This was to ensure efficient use of field

75


S. A. Weaver et al.

Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

Fig. 2 Remote measurements conducted with the laser caliper dendrometer time and consistent measurement collection. For this study, the direct caliper measurement was assumed to be the best or the »true« diameter. We allowed up to 30 seconds for each individual remote caliper measurement. The diameter tape measurements were collected for reference purposes, as this is a common method used in the southern United States. Measurements made using diameter tapes have been shown to be different than those collected using calipers (McArdle 1928), and technically should not be directly compared to single caliper measurements or sector-fork measurements given the irregular shape of most tree boles (Brickell 1970, Moran and Williams 2002). However, for illustrative purposes, we make those comparisons in this study.

Fig. 3 Light conditions being measured using the Mastech LX1330B light meter

76

While a variety of electronic dendrometers and scanning systems are available, due to availability, time, and cost limitations, only the Mantax Black caliper was chosen for testing. For the same reasons, all of the measurements were collected by one individual after several hundred practice measurements with both the sector-fork and the laser caliper, and after practice on fixed width, non-natural targets. This process helped avoid differences between individuals, although they could be small (Elzinga et al. 2005). The only environmental variable that was collected with the sampling of each tree was the incident light luminous emittance (lux) using a Mastech LX1330B light meter (Fig. 3). This lux data were collected to determine whether light conditions are correlated with remote diameter measurement accuracy. Errors in successive measurements of tree diameters can occur with some instruments, and may be due to the following (McCarthy 1924, Robertson 1928):  misjudging points of successive measurements;  failing to place the instrument in its proper plane;  measuring within close proximity to tree deformations;  failing to account for differences in the tension of bark on trees;  misreading instrument divisions;  failing to notice weathering and scaling of tree bark;  failing to know that instruments can be out of adjustment. To limit potential errors such as these, we developed a set of standard methods for data collection. These methods included measuring the diameter of a tree all seven times with each visit, using the same person to collect all of the measurements, and conducting six of the seven measurements from the same perspective with respect to the tree; the exception involved the use of the diameter tape. The caliper tongs were also closed after each measurement to avoid biasing subsequent measurements. While direct caliper measurements can be subject to error described by Abbé’s Principle, remote caliper measurements will not (Clark 2003). This principle states that measurement errors with calipers will increase as the object being measured moves away from the caliper bar, causing the caliper’s tongs to bend outward, which introduces error. To minimize this problem when using the caliper to make direct measurements of tree diameters, the bole of the tree was placed as close as possible to the caliper bar, which reduced the bending force on the tongs. When larger trees were Croat. j. for. eng. 36(2015)1


Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

measured, this type of error could be introduced when pressure from the tree bole was placed further out on the tongs. We did not employ a correction factor in these instances, and given that the caliper is relatively new, we assumed that the forces acting on the tongs, perhaps requiring them to bend outward rather than to slide naturally along the caliper bar, would be minimized. Ideally, the set of laser caliper measurement deviations for a specific distance d (direct measurement – distance d measurement) in a forest type j should be normally distributed around zero (no deviation). However, the ability to place the laser lights exactly on the edge of each tree at exactly the same time was difficult, perhaps due to a combination of general light conditions, bark conditions (wet, dry, fragmented, etc.), and shadows within the crevasses of the bark. While we tested the correlation between accuracy and general light conditions, the other potential factors were not tested. We used BestFit software (Palisade Corporation 1996) to test whether each set of laser caliper measurement deviations for a specific distance d was normally distributed. In 10 of the 12 cases, sets of deviations were not statistically significant with respect to their ability to represent a normal distribution, according to Chi– squared, Anderson-Darling, or Kolmogorov-Smirnov tests. Therefore, a non-parametric method, Wilcoxon’s matched-pairs signed-ranks test, was used to determine whether pairs of sample sets arose from the same population having the same location. Another non– parametric method, the Mann-Whitney test, was used to determine whether unpaired data of sample sets from different forest types had significantly different median values. When applying the Wilcoxon’s matched-pairs signed-ranks test, if the rank sums of the paired samples are approximately the same, we would expect that they are not significantly different (Sokal and Rohlf 1995). Although we initially assumed they are different, we applied this test to assess the difference between diameter tape measurements and other direct measurements. In applying this test for an analysis of Hypothesis 1, the test statistic was the tree diameter, and we compared the remotely obtained caliper measurements (3–12 m) to the direct caliper measurement (0 m) within each forest type. In applying this test for an analysis of Hypothesis 2, the test statistic was again the tree diameter, we compared the sector-fork measurements to all caliper measurements (direct, 3–12 m) within each forest type. In assessing Hypothesis 3, Pearson’s product-moment correlation coefficient was computed to estimate the linear correlation or association between illuminance (lux measurements at DBH) and the deviations computed for reCroat. j. for. eng. 36(2015)1

S. A. Weaver et al.

motely measured tree diameters using the caliper (direct measurement – remote measurement). Both the actual deviation (positive or negative value) and the absolute value of the deviation were assessed in this correlation analysis. The aim was to determine whether the Pearson product-moment correlation coefficient was significantly different from zero at the p = 0.05 level. In other words, if the associated p-value for each pair of data was less than 0.05, then the Pearson product-moment correlation coefficient was considered significantly different from zero. For Hypothesis 4, we focused on measurements collected at a specific distance from each tree (e.g. 3 m), and attempted to determine whether the unpaired sample data (the diameter deviations) from the three different forest types were significantly different. Here, the Mann-Whitney test was employed to determine whether the median value of the deviation in diameters was significantly different among the three forest types.

3. Results The average diameters measured within each forest type and the associated measurement process are shown in Table 2. In general, tree diameters estimated using the diameter tape were significantly greater (p < 0.05) than measurements of diameters estimated using other methods. However, the other methods only considered one viewing perspective of a tree, thus do not fully account for irregularities in the shape of tree boles. The general pattern of results within a forest type is similar, yet the use of the sector-fork consistently produced a lower mean diameter when compared to the other measurements. Table 2 also provides a measure of variation (standard deviation) among the sets of diameters, reflecting the fact that there is more diversity among tree sizes in the deciduous forest than in the two coniferous forests. Interestingly, 12 m remote caliper measurements were consistently slightly smaller with regard to the standard deviation than diameter measurements collected with the other processes. Since the diameter distribution of trees within each forest type is different, another way to view the results is to compare the deviation in diameters with respect to the 0 m caliper measurement (Fig. 4). In general, most of the deviations were 0.8 cm or less for individual trees. There are two interesting results here; first, the remote measurements 9 m and greater within the young coniferous stand tended to overestimate tree diameters, and second, the sector-fork measurements across all forest types tended to slightly underestimate tree diameters. The measurement deviations

77


S. A. Weaver et al.

Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

Table 2 Mean tree diameter and standard deviation of tree diameters by forest and measurement type Sample measurement

Young coniferous Old coniferous

Deciduous

Mean, cm

SD*, cm

Mean, cm

SD*, cm

Mean, cm

SD*, cm

Diameter tape

18.36

4.87

34.64

8.42

30.11

14.04

Sector–fork

17.88

4.68

33.74

8.62

29.50

13.69

Caliper – 0 m

18.03

4.79

34.35

8.52

29.79

13.98

Caliper – 3 m

18.07

4.72

34.34

8.52

29.56

13.77

Caliper – 6 m

18.11

4.68

34.35

8.45

29.69

13.75

Caliper – 9 m

18.24

4.63

34.34

8.41

29.67

13.63

Caliper – 12 m

18.31

4.56

34.26

8.38

29.68

13.56

* Standard deviation

also suggest that the use of the sector-fork tended to result in a noticeably larger amount of variation across forest types. In general, the variation in caliper measurement deviations (as compared to the 0 m caliper measurement) tended to increase slightly the farther one moved away from the tree. When examining the differences between the direct caliper measurement and the remote caliper measurements within the deciduous stand, we reject the null hypotheses (p < 0.05) that samples obtained at 3 m, 6 m, and 9 m from each tree are the same as the direct measurement. However, the 12 m remote measurements (p > 0.05) were not significantly different from the direct caliper measurement. Therefore, in assessing Hypothesis 1, we found mixed results from measurements collected in the deciduous stand. When examining the differences between direct and remote caliper measurements within the older coniferous stand, there are no statistically significant (p > 0.05) differences between the direct and remote measurements. Therefore, we could not reject the null hypothesis that the samples arose from the same population. The same can be said about the direct and 3 m remote measurements obtained from the young coniferous stand. However, measurements obtained from 6–12 m were statistically significantly different than the direct measurement (p < 0.05); therefore, we reject the null hypothesis in these cases. In comparing the caliper measurements to the sector-fork measurements, we found no statistically significant differences (p > 0.05) in the deciduous stand. For the older coniferous stand, we found statistically significant differences between the sector-fork measurements and the direct caliper and 3 m caliper mea-

78

Fig. 4 Box-and-whisker plot of the deviation in tree diameters when compared to the 0 m (direct) caliper measurements Croat. j. for. eng. 36(2015)1


Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

surements (p < 0.05); all other comparisons of diameters collected remotely in the older coniferous stand with the calipers were not significantly different than the sector-fork measurements. According to the results obtained from the application of the Wilcoxon two-sample test, the sector-fork data collected within the young coniferous stand were considered statistically significantly different (p < 0.05) than the data collected at all distances with the calipers. The correlation analysis between illuminance (lux) and the deviation in remote caliper measurements from direct caliper measurements indicated very weak relationships in many instances (Table 3). In this analysis the deviation could be either positive or negative, and therefore it is assumed that light characteristics may force an overestimate or underestimate of the tree diameter when measured remotely. However, based on the p-values of this analysis, illuminance was not significantly correlated with the deviation in diameter measurements between the direct measurement and the remote measurements. We also assessed the correlation between illuminance and the absolute value of the difference between remote caliper measurements and direct caliper measurements, assuming that the direction of the deviation (either an overestimate or underestimate of the tree diameter) was not necessarily forced by illuminance, but that changes in illuminance simply caused a deviation one way or the other (Table 4). As with the prior analysis, it did not appear that illuminance was significantly correlated with the absolute value of the difference between remote caliper measurements and direct caliper measurements based on the p-values (p > 0.05) produced. Table 3 Pearson’s product-moment correlation (rxy) between illuminance (lux) and the deviation in remote caliper measurements from direct caliper measurements Sample distance

Young coniferous

Old coniferous

Deciduous

rxy

p-value

rxy

p-value

rxy

p-value

3m

0.126

0.213

0.024

0.810

0.042

0.678

6m

0.074

0.467

–0.070

0.486

–0.062

0.540

9m

0.092

0.364

–0.021

0.833

–0.103

0.310

12 m

0.075

0.456

–0.008

0.939

–0.117

0.247

In assessing differences between forest types, using the Mann-Whitney non-parametric test and the deviations between direct and remote measurements as the test statistic, at 3 m we found that there were Croat. j. for. eng. 36(2015)1

S. A. Weaver et al.

Table 4 Pearson’s product-moment correlation (rxy) between illuminance (lux) and the absolute value of the deviation in remote caliper measurements from direct caliper measurements Sample distance

Young coniferous

Old coniferous

Deciduous

rxy

p-value

rxy

p-value

rxy

p-value

3m

0.048

0.636

0.045

0.659

–0.013

0.897

6m

–0.100

0.322

0.101

0.319

–0.027

0.792

9m

–0.076

0.455

–0.077

0.444

–0.057

0.573

12 m

–0.108

0.286

–0.113

0.264

–0.004

0.971

significant differences between the deciduous stand and both coniferous stands (p < 0.05), yet there was no significant difference between the young and old coniferous stands. When using the absolute value of the deviation as the test statistic, no significant differences were observed. In comparing the 6 m remote measurements, the only significant differences (p < 0.05) were observed between the deciduous and young coniferous stands. When the absolute value of the deviations were used as the test statistic, significant differences (p < 0.05) were only observed between the deciduous and old coniferous stands, interestingly. There were two significant differences in the 9 m measurements among forest types: between the young coniferous and old coniferous stands, and between the young coniferous and deciduous stands. However, when the absolute value of the deviations was used as the test statistic, no significant differences (p < 0.05) were observed. Similarly, these same results were observed with the 12 m measurements. In sum, when comparing measurements collected from the same distance away from a tree, yet within different forest types, when the absolute value of the measurement deviations were compared, in only one case was there a significant difference. And when using the original (positive and negative values) measurement deviations, the results were mixed, but when comparing the longer distances there seemed to be differences between the measurement of small trees (young coniferous stand) and the measurement of larger trees (deciduous and old coniferous stand).

4. Discussion The ability to remotely measure the diameter of trees has practical value for field technicians in that travel time to individual trees at sample locations can be reduced. Further, upper-stem diameters necessary to understand the extent of merchantability within a

79


S. A. Weaver et al.

Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

tree can be estimated more reliably, as these otherwise generally are ocularly estimated. Perhaps the efficiency of data collection processes can be increased, however, the efficiency of using laser calipers to measure tree diameters remotely was not assessed in this research. We found significant differences in diameters measured using a diameter tape and using the calipers. We recognize that it is commonly accepted that diameter tape measurements will more likely lead to different results than caliper or sector-fork measurements, due to variations in tree bole and bark shape (McArdle 1928). Two or more sector-fork or caliper measurements acquired from different perspectives of the tree bole can alleviate some of these concerns. However, in this work we assumed that only one direction (or perspective) of a tree bole would be used in conjunction with the laser calipers. This assumption arises from the notion that a field technician should be able to stand in the middle of a circular measurement plot and use the laser calipers to remotely measure all of the trees in the plot without having to move away from the plot center. We further only measured tree diameters with the sector-fork from one perspective in order to be consistent with, and comparable to, the laser caliper measurements. These limitations in measurement standards do not detract from the practical value of collecting remote measurements, and associated decisions were made to accommodate the study design. In our work, we did find that the use of the sectorfork resulted in greater variation among the deviations from the direct (0 m) caliper measurement conducted at the same point on a tree and viewed from the same perspective. In fact, on average, the sectorfork diameter measurements were slightly smaller than the caliper measurements. We attribute a great deal of this problem to the scale of each device. Cummins (1937) found that differences in scale between instruments can contribute to differences in diameter measurements. The calipers have a graduated scale in 0.25 cm (0.1 inch) increments, yet the sectorfork scale has a graduated scale in 1 cm increments, and diameters were estimated to the nearest 0.5 cm. The scale on the sector-fork is also non-linear, and larger diameter measurements seemed to be more difficult to refine, while the caliper scale is linear and consistent. One issue that could have potentially introduced error into the measurement of tree diameters with the laser calipers was the ability of the person performing the measurements to consistently measure a tree diameter at the same height and same angle (horizontal or vertical) to the tree bole. The calipers, while not

80

overly heavy (in terms of weight), needed to be held steady for 10–20 seconds each time a diameter was measured. If fatigue sets in after numerous repeated measurements, this practice can become a burden on the field technician and possibly affect the quality of results. Further, any uncertainty on behalf of the field technician regarding where the tree diameter should be measured can affect the person’s ability to position the laser points correctly on the edge of a tree bole. The extra time required to ensure the correct position of the laser points on a tree bole could affect the increase in efficiency expected when using a remote instrument, and perhaps lead to greater error. Therefore, one drawback to our analysis was the time limit we placed on measuring diameters when the calipers were used remotely. While effort was made to apply similar amounts of time at each stage in the measurement collection process, there may have been an association between measurement time and measurement accuracy of which we are unaware. Another issue that may have introduced error during the measurement process was distraction on behalf of the operator of the equipment. One particular distraction was glare caused by the Sun. At times, depending on the arrangement of the field technician, the tree, and the Sun, the laser points were difficult to see on the edges of tree boles. Although the field technician practiced using each device for several weeks prior to the onset of the study, not all environmental factors could be replicated during the practice period. This potentially introduced error into the analysis. Further, we did not design the study to control for stem density or understory vegetation composition or density. Each set of 100 samples was contained within one of three stands, represented by one of the three forest types, and conditions within each stand (density, understory) were very similar. We recognize the fact that stem density and understory vegetation composition can play a role in the ability of a person to accurately measure tree bole diameters with a laser caliper, and the slight variations in these that were evident at the study site could have potentially introduced error into the analysis. One issue we discovered through a review of the literature was that over the course of a study period (and even over the course of a day) tree diameters may change slightly due to cambial growth, water balance, or due to the angle from which the remote measurements were made (Haasis 1934, Pesonen et al. 2004, Devine and Harrington 2011). Paired comparisons in our analysis were made with measurements that were collected within about five minutes of each other during each visit to a tree; therefore, this Croat. j. for. eng. 36(2015)1


Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

issue should have been minimized through the study design. We also understood that there may be some aspects of tree and understory vegetation, bark color, stem density, and forest type in general that could cause error and affect the ability to distinguish bark edges with a high level of accuracy. For example, slight variations in tree or bark condition could act to misguide a field technician into collecting remote measurements that do not necessarily represent the true edge of a tree bole. Tree lean and the shape of a tree’s cross-sectional area may have also contributed to the variations in measurements between instruments (Grosenbaugh 1963). The differences between forest types with respect to these types of issues appear to be most pronounced at distances of 6 m or less to the target tree, after which there are no significant differences in remote measurements. Thus the viewing perspective (i.e. being too close to the measured object) may be a concern. In summary, even while there were significant differences in diameters measured using a diameter tape and using the calipers, due to variations in tree bole and bark shape, these differences were, on average, 0.55 cm or less in each of the three forest types related to this study. In most cases of the distances from the subject tree, this represents a 2% or less deviation from the diameter estimated using a diameter tape. The differences among average forest caliper measurements of tree diameters, from 3 to 12 m distances, are also less than 0.3 cm. Therefore, the usefulness of the laser caliper system for measuring tree diameters within the forest conditions represented by this study seems good for distances up to at least 12 m. However, trees within measurement plots that have easily accessible boles from the center of the plots may be more efficiently measured using traditional methods (diameter tape, sector fork). In addition, it would seem necessary to measure the distance from a plot center to a borderline tree (situated on a plot edge). If this is necessary, these trees may also be more efficiently measured using traditional methods rather than remote methods, since the technician will likely be in physical contact with the tree when measuring

Croat. j. for. eng. 36(2015)1

S. A. Weaver et al.

its distance from the plot center. This of course assumes that remote instruments (laser rangefinders) are not employed to measure distances.

5. Conclusions Tree diameters are one of the main components of forest volume estimation processes. Assuming the same level of sampling intensity with and without remote measurements of tree diameters, if remote measurements can accurately represent forest conditions, management costs can possibly be reduced. While laser caliper measurements were only collected with respect to one viewing perspective of a tree, they were consistently smaller on average than diameter measurements collected with a diameter tape. While it seemed that most of the significant differences in remote measurements were observed within the first 6 m of trees, these differences were rather small (0.8 cm or less for individual trees). The direction of the difference (over or under the direct caliper measurement) was different for each forest type, which if consistently observed, might suggest the use of a small correction value for each type of forest measured. However, reasonably accurate remote measurements may be attractive to field personnel for the time saved not having to travel to and physically touch each tree. While significant differences were found, the small differences found in this study may not have a significant impact on in field practices when tree diameters are grouped into one–inch diameter classes, as they often are in the southern United States. The laser calipers are able to provide accurate diameter readings at a distance within 12 m, and measurements that are traditionally collected remotely (e.g. upper stem diameters, or lengths of the merchantable portion of a stem) can perhaps be estimated or measured more accurately.

Funding This work was supported by the Warnell School of Forestry and Natural Resources at the University of Georgia.

81


S. A. Weaver et al.

Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83)

6. References Behre, C.E., 1926: Comparison of diameter tape and caliper measurements in second-growth spruce. Journal of Forestry 24(2): 178–182. Bell, J.F., Groman, W.A., 1971: A field test of the accuracy of the Barr and Stroud Type FP-12 optical dendrometer. Forestry Chronicle 47(2): 69–74. Binot, J.-M., Pothier, D., Lebel, J., 1995: Comparison of relative accuracy and time requirement between the caliper, the diameter tape and an electronic measuring fork. Forestry Chronicle 71(2): 197–200.

Grosenbaugh, L.R., 1963: Optical dendrometers for out-ofreach diameters: A conspectus and some new theory. Forest Science Monograph 4: 47 p. Haasis, F.W., 1934: Diametral changes in tree trunks. Carnegie Institution of Washington, Washington, D.C. Publication No. 450: 103 p. Henning, J.G., Radtke, P.J., 2006: Detailed stem measurements of standing trees from ground-based scanning lidar. Forest Science, 52(1): 67–80. Kalliovirta, J., Laasasenaho, J., Kangas, A., 2005: Evaluation of the laser-relascope. Forest Ecology and Management, 204(2): 181–194.

Brickell, J.E., 1970: More on diameter tape and calipers. Journal of Forestry 68(3):169–170.

Krauch, H., 1924: Comparison of tape and caliper measurements. Journal of Forestry 22(5): 537–538.

Bruce, D., 1975: Evaluating accuracy of tree measurements made with optical instruments. Forest Science 21(4): 421– 426.

Liu, C.J., Huang, X., Eichenberger, J.K., 1995: Preliminary test results of a prototype of Criterion. Southern Journal of Applied Forestry 19(2): 65–71.

Chacko, V.J., 1961: A study of the shape of cross section of stems and the accuracy of calliper measurement. Indian Forester 87(12): 758–762.

Liu, S., Bitterlich, W., Cieszewski, C.J., Zasada, M.J., 2011: Comparing the use of three dendrometers for measuring diameters at breast height. Southern Journal of Applied Forestry 35(3): 136–141.

Clark, N.A., Wynne, R.H., Schmoldt, D.L., 2000a: A review of past research on dendrometers. Forest Science 46(4): 570– 576. Clark, N.A., Wynne, R.H., Schmoldt, D.L., Winn, M., 2000b: An assessment of the utility of a non-metric digital camera for measuring standing trees. Computers and Electronics in Agriculture 28(2): 151–169. Clark, R., 2003: Understanding errors in hand-held measuring instruments. Modern Machine Shop. (Source: http:// www.mmsonline.com/articles/understanding–errors–in– hand–held–measuring–instruments, accessed on 25 May, 2012). Cummins, W.H., 1937: Tree-fork and steel tape for close measurement of small diameters. Journal of Forestry 35(7): 654– 660.

McArdle, R.E., 1928: Relative accuracy of calipers and diameter tape in measuring Douglas fir trees. Journal of Forestry 26(3): 338–342. McCarthy, E.F., 1924: Comment on tapes and calipers. Journal of Forestry 22(4): 539. Moran, L.A., Williams, R.A., 2002: Comparison of three dendrometers in measuring diameter at breast height. Northern Journal of Applied Forestry 19(1): 28–33. Nicoletti, M.F., Batista, J.L.F., Carvalho, S.P.C., Castro, T.N., 2012: Accuracy of two optical dendrometers for non-destructive determination of woody biomass. Pesquisa Florestal Brasileira 32(70): 139–149. Palisade Corporation, 1996: BestFit for Windows, version 2.0d. Palisade Corporation, Newfield, NY.

Delwiche, M., Vorhees, J., 2003: Optoelectric system for counting and sizing field-grown deciduous trees. Transactions of the ASAE 46(3): 877–882.

Parker, R.C., 1997: Nondestructive sampling applications of the Tele-Relaskop in forest inventory. Southern Journal of Applied Forestry 21(2): 75–83.

Devine, W.D., Harrington, C.A., 2011: Factors affecting diurnal stem contraction in young Douglas fir. Agricultural and Forest Meteorology 151(3): 414–419.

Parker, R.C., Matney, T.G., 1999: Comparison of optical dendrometers for prediction of standing tree volume. Southern Journal of Applied Forestry 23(2): 100–107.

Drew, D.M., Downes, G.M., 2009: The use of precision dendrometers in research on daily stem size and wood property variation: A review. Dendrochronologia 27(2): 159–172.

Parresol, B.R., Hotvedt, J.E., 1990: Diameter measurement in bald cypress. Forest Ecology and Management 33/34(1–4): 509–515.

Elzinga, C., Shearer, R.C., Elzinga, G., 2005: Observer variation in tree diameter measurements. Western Journal of Applied Forestry 20(2): 134–137.

Pesonen, E., Mielikäinen, K., Mäkinen, H., 2004: A new girth band for measuring stem diameter changes. Forestry 77(5): 431–439.

Fairweather, S.E., 1994: Field tests of the Criterion 400 for hardwood tree diameter measurements. Northern Journal of Applied Forestry 11(1): 29–31.

Popescu, S., 2007: Estimating biomass of individual pine trees using airborne lidar. Biomass and Bioenergy 31(9): 646–655.

82

Croat. j. for. eng. 36(2015)1


Assessing the Accuracy of Tree Diameter Measurements Collected at a Distance (73–83) Rhody, B., 1975: A new approach to terrestrial and photographic forest sampling: The use of a panoramic lens. Photogrammetria 30(2): 75–85. Robertson, W.M., 1928: Review of the case of diameter tape vs calipers. Journal of Forestry 26(3): 343–346. Skovsgaard, J.P., Johannsen, V.K., Vanclay, J.K., 1998: Accuracy and precision of two laser dendrometers. Forestry 71(2): 131–139. Sokal, R.R., Rohlf, F.J., 1995: Biometry, third ed. W.H. Freeman and Company, New York. 887 p.

S. A. Weaver et al.

Williams, M.S., Cormier, K.L., Briggs, R.G., Martinez, D.L., 1999: Evaluation of the Barr & Stroud FP15 and Criterion 400 laser dendrometers for measuring upper stem diameters and heights. Forest Science 45(1): 53–61. Yoda, K., Suzuki, M., Suzuki, H., 2000. Development and evaluation of a new type of opto-electronic dendrometer. IAWA Journal 21(4): 425–434. Zhang, L., Grift, T.E., 2012. A monocular vision-based diameter sensor for Miscanthus giganteus. Biosystems Engineering 111(3): 298–304.

Authors’ addresses:

Received: May 7, 2013 Accepted: June 12, 2014 Croat. j. for. eng. 36(2015)1

Steven A. Weaver, MSc.* e-mail: sw3av3r@uga.edu Zennure Ucar, MSc. e-mail: zennucar@uga.edu Prof. Pete Bettinger, PhD. e-mail: pbettinger@warnell.uga.edu Krista Merry, MSc. e-mail: kmerry@warnell.uga.edu Krisha Faw, BSc. e-mail: kfaw@uga.edu Prof. Chris J. Cieszewski, PhD. e-mail: biomat@warnell.uga.edu University of Georgia Warnell School of Forestry and Natural Resources 180 E Green Street USA – GA 30602 Athens USA * Corresponding author

83



Preliminary note

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW Electric Disc Chipper Fulvio Di Fulvio, Gunnar Eriksson, Dan Bergstrรถm Abstract The development of efficient woody biomass comminuting processes and systems is of great importance for establishing bio-refineries. Using hybrid systems, which store excess energy from a diesel engine during periods of low loading for use during peak loading times, may yield higher energy efficiency compared to direct diesel-powered comminuting systems. In order to design hybrid chippers, a series of data are required on the load variations, in order to estimate the amount of energy that needs to be stored, and the peak power required. As a consequence, a detailed knowledge of the effects of wood properties on the direct power consumption during chipping is relevant. Therefore, the objectives of this work were to study the effects of wood properties (size and density) of pine, spruce and birch trees from early thinnings in the north of Sweden on the specific power and energy demand and time consumption of a 30 kW electric chipper while producing chips of two sizes. The study has generated models that replicate the processes, which can be used when designing efficient hybrid systems. The butt area had a significant effect on the power requirements when chipping and, along with chip length, had a significant effect on the energy requirements. Butt area and chip length also had a significant effect on the chipping productivity. There were small effects caused by the OD densities and by different species. These findings agree with previous studies and can be used for designing future hybrid chippers. Keywords: chipper, energy consumption, productivity, efficiency, hybrid systems

1. Introduction Forest biomass is increasingly utilized worldwide for different markets, such as construction, pulp and paper, chemicals and fuels. Before being used for industrial chemical and thermo-chemical processes, the wood must be comminuted. This is mainly carried out using mechanical processes such as chipping or crushing. For most processes, the quality of chips is defined by their homogeneity i.e. their particle size distribution. Thus, chipping is preferable to crushing since it produces particles with more predictable properties. For many processes, such as combustion of solids, a wide range of fuel particle sizes may significantly reduce the efficiency of the conversion process. Thus, the development of efficient woody biomass comminuting processes and systems is of great importance in order to achieve these targets. Croat. j. for. eng. 36(2015)1

In general, chippers operating at industrial sites are run with electric engines while machines operating at terminals, or close to forests, are run using diesel engines. Using electricity is preferred because of technical (easier to build and maintain), economic (cheaper and higher energy efficiency) and environmental (e.g. powered by electricity from renewable sources) benefits, among others. However, such systems require connection to a power grid which is only possible at industrial sites and some terminals. A third option is to use hybrid systems, which store excess energy from the diesel engine during periods of low loading for use during peak loading times, either in batteries, capacitors, flywheels or hydraulic storage devices (Sun et al. 2010). This means that the diesel engine is running under more efficient operating conditions and may also make it possible to recover some kinetic and po-

85


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

tential energy, for instance from braking and from crane movements. Such systems may give higher energy efficiency compared to a direct diesel-powered system. For forest machines, potential savings from hybridization of 15–30% have been predicted (Swedish Forest Agency 2012). When designing a hybrid system for chipping, it is essential to understand the variations in the power required over time. Work to develop such systems is currently being undertaken in many different markets (Larsson 2012). In order to design hybrid chippers, an understanding of the operational conditions is necessary. Data on load variations, to estimate the amount of energy that needs to be stored, and on the power that is required during peaks, are required. Direct measurements of the torque and rotation speed (Nurmi 1986, Liss 1987, Van Belle 2006, Spinelli et al. 2012) make it possible to focus on the chipping process and to eliminate the influence of other factors such as operator skill, the log loading system and engine efficiency. The exclusion of these factors was necessary in order to increase the accuracy in this study, but factors such as the loading system may increase the chipper idling time, usually accounting for approximately 10% of the productive work time (c.f. Röser et al. 2012), and could generate energy for charging an electrical accumulator. The measurement of the operational fuel consumption as a measure of energy consumption for chipping is commonly carried out (e.g. Aman 2011, Yoshioka 2006). Magagnotti and Spinelli (2011) found that the energy used for chipping represents only about 3% of the energy return (solid fuel wood) for an industrial diesel drum chipper. However, diesel consumption makes up 36% of the operational costs for roadside chipping (Marchi et al. 2011) and the use of it significantly contributes to greenhouse gas emissions (Van Belle 2006). There are many factors influencing the efficiency of chipping processes. The influence of chipper geometry (parameters such as knife sharpness, spouting angle, side angle and the number of knives) on disc chipping has been studied by Hartler (1986), Uhmeier (1995), Hellström (2008, 2009) and Abdallah (2011). Among the wood parameters, density, temperature (especially if the wood is frozen) and moisture content are known to influence the process (Kivimaa and Murto 1949, Papworth and Erickson 1966, Liss 1991). Kivimaa and Murto (1949) showed that an increase in chip length from 2.5 to 50 mm resulted in a reduction of energy requirement per produced energy by about 88%. The production of larger chips can decrease the required energy per tonne, per oven dry-tonne (OD t) and per cubic metre (m3) significantly (cf. Nurmi 1986).

86

However, Spinelli et al. (2012) showed that a novel knife design for disc chippers can significantly improve the quality of chips for small boilers, without reducing the machine energy efficiency. Wood piece size has a significant effect on productivity and energy demand, as efficiency was found to be proportional to the size of wood used (Liss 1987; 1991, Van Belle 2006, Ghaffariyan et al. 2013). Different tree species have different wood properties, e.g. depending on the growth rate of trees and species, the resultant wood has a different density/hardness, which also affects the chipping efficiency (Liss 1991). Small diameter trees from early thinnings represent a great resource of biomass for refinery purposes in Finland and Sweden, where, respectively, 2.5 and 1.3 million m3 of forest chips are produced annually from small trees, that is around 33% and 20% of the respective total potential resource (Routa et al. 2012). The main tree species in the region are Scots pine (Pinus sylvestris L.), Norway spruce (Picea abies (L.) Karst) and birch (Betula spp.), all of which have different wood properties. Knowledge about the influence of factors such as wood properties on direct power consumption during chipping is of relevance to the development of efficient comminuting systems. Knowledge of the specific electricity consumption is of importance for designing hybrid systems i.e. describing the variation and peaks in workload that the electricity generator (diesel engine) must handle. The objectives of this work were to study the effects of wood properties (size and density) of pine, spruce and birch on the specific power energy demands and time consumption of a 30 kW electric chipper when producing chips in two sizes, in order to create models which can be used when designing efficient hybrid systems based on similar chippers.

2. Material and methods Delimbed stemwood logs from pine, spruce and birch were sampled from a thinning site in the coastal area of Västerbotten (64°06’ N, 20°37’ E and 60 m.a.s.l.), northern Sweden. The sampled stand had a density of 3200 trees/ha, an average tree size diameter of 8.1 cm at breast height (dbh) and an average height of 8.3 m. In total, about 30 m3 of solid round-wood were harvested by thinning from below and transported to the experimental site (Biofuel Technology Centre in Umeå) a few days after cutting. The biomass was stored in three piles (by species) on open asphalted ground (uncovered) for 3 weeks before trials. The chipping trials were carried out over five days in October 2012. A toCroat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

F. Di Fulvio et al.

Table 1 Properties of trees per species and diameter size class (1–5) used in experiments using average and standard deviation (Sd) values. Diameter at butt end = Dbutt, diameter at half length = Dmiddle, diameter at top = Dtop, moisture content, wet basis = MC Class, cm

1: <9

2: 9 <11

Birch (n = 53)

Properties

Sd

Average

Sd

Average

Sd

Dbutt, cm

7.91

0.77

7.73

1.01

8.05

0.70

Dmiddle, cm

6.26

0.71

6.39

1.20

6.88

0.97

Dtop, cm

4.33

0.78

4.85

1.10

4.95

0.95

Length, m

3.55

0.38

3.67

0.50

3.24

0.46

Mass, fresh, kg

3.55

0.38

12.7

5.09

11.5

3.17

MC, %

45.2

1.1

61.7

1.4

62.9

2.9

Density, OD kg m-3

499

46

407

26

383

44

Dbutt, cm

10.20

0.66

10.3

0.63

9.88

0.55

Dmiddle, cm

7.96

0.63

8.54

0.85

8.19

0.51

Dtop, cm

5.94

0.99

6.57

1.05

5.97

0.51

Length, m

3.75

0.44

3.61

0.31

3.72

0.68

Mass, fresh, kg

16.80

2.86

20.9

4.81

19.3

4.77

44.1

0.1

62.0

0.2

65.0

2.9

Density, OD kg m

481

33

378

28

368

20

Dbutt, cm

11.60

0.27

11.5

0.27

11.5

0.29

Dmiddle, cm

9.14

0.92

9.56

0.64

9.38

0.82

Dtop, cm

7.40

0.99

7.59

1.08

7.18

1.12

Length, m

3.90

0.45

3.77

0.53

3.75

0.65

Mass, fresh, kg

22.70

2.36

27.6

3.84

24.6

4.94

43.9

1.3

62.6

0.1

60.5

4.8

Density, OD kg m

482

24

378

25

403

40

Dbutt, cm

12.50

0.20

12.4

0.3

12.5

0.3

Dmiddle, cm

9.90

0.90

10.3

0.5

10.3

0.6

Dtop, cm

7.15

1.33

7.74

1.70

8.40

1.32

Length, m

3.88

0.49

4.04

0.44

3.74

0.42

Mass, fresh, kg

26.10

4.38

32.7

2.81

30.3

4.45

43.6

0.6

62.6

1.3

62.3

0.5

Density, OD kg m

483

70

382

20

351

37

Dbutt, cm

13.60

0.27

13.3

0.28

13.5

0.38

Dmiddle, cm

11.30

0.71

10.9

1.69

11.4

1.02

Dtop, cm

9.50

1.03

9.54

1.04

9.63

1.52

Length, m

3.63

0.56

3.60

0.39

3.40

0.65

Mass, fresh, kg

34.20

5.58

35.9

4.9

34.5

5.1

44.1

0.3

60.9

2.5

60.8

1.5

506

35

380

58

372

33

-3

MC, % -3

4: 12 <13

MC, % -3

5: 13  14

MC, % -3

Density, OD kg m

Croat. j. for. eng. 36(2015)1

Spruce (n = 61)

Average

MC, %

3: 11 <12

Pine (n = 62)

87


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

tal of 185 logs (63 pine, 61 spruce, 52 birch) were randomly sampled from the piles. Their lengths fell in the range 2.6–5.1 m, the diameters of their butt ends (dbutt) ranged from 5.6 to 14.4 cm and their mass (fresh weight) ranged from 3.5 to 43.5 kg. The total mass was 4189 kg. The diameter was measured for each log at the butt end (dbutt) at half length (dmiddle) and at the top (dtop). The length and mass of each tree log were also measured. The logs were sorted by species and dbutt into five classes: 1) Ø < 9 cm, 2) 9 ≤ Ø < 11 cm, 3) 11 ≤ Ø < 12 cm, 4) 12 ≤ Ø < 13 cm and 5) 13 ≤ Ø ≤ 14 cm (Table 1). Each combination of species (n = 3) and diameter class (n = 5) (15 treatment combinations) was repeated three times. Each treatment combination was carried out using two different disc chipper knife settings (nominal lengths of 8 mm (»short«) and 12 mm (»long«)). The chipper used was an Edsbyhuggen 250H (Edsbyhuggen AB, Sweden, year of manufacture 2011) with a 30 kW electric motor (Busck T1C 200L-4), which had an efficiency of 91.2–93.5% at full load (Swedish Energy Agency 2010) (Fig. 1). It had a hydraulic system (15 l) that drove a vertical pair of feedrollers with an in-feed opening of 250×250 mm. The speed of the feed rollers was constant over the whole experiment. The chipper was equipped with a steel disc 825 mm in diameter and 38 mm thick, with 4 knives, giving a total mass of 205 kg. The disc formed an angle of 45° with the feed direction and rotated at 540 rpm. The 4 knives were adjustable to produce chips from 5 to 12 mm in target length. The chips produced were blown for 2.0 m into an expulsion tube by means of a fan and then collected in 1.5 m3 plastic bags.

The first batch of logs was chipped to a length of 12 mm and the second to 8 mm. The electricity supply to the chipper was connected to a Fluke Power Log data logger during the experiment, giving instantaneous measurements of the electricity used by the engine. These data were logged at a sampling frequency of 2 Hz (1 observation every 0.5 seconds). The instrument was calibrated before starting the experiments. The chipping output was defined as the scaled mass of each log and its solid volume, which was calculated using the butt and top diameters with the length in the formula for the volume of a truncated cone. The dry density of each log was determined as the ratio between its dry mass and volume. Individual logs were manually fed into the chipper, butt end first. To determine the moisture content (MC, wet basis), 30, five litre buckets of chips were collected (150 litres in total); the chips were collected from the output stream. Each bucket was filled by merging together sub-samples from three trees per diameter class (for each species and treatment combination). It was assumed that all trees in each diameter class had approximately the same MC. The MC was determined in accordance with CEN/TS 14774-2. The same ovendried samples used for determining MC were used for the determination of particle size distribution in accordance with CEN/TS 15149-1. The sieves used were suitable for length classes 31.5 mm, 16 mm, 8 mm and 3.15 mm. The time taken to chip each individual log (»chipping time«) was determined from the power measurements (the time was also manually recorded with a stopwatch for comparison). Chipping took place from

Fig. 1 Edsby chipper (left) and preparation of logs for experiments (right)

88

Croat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

F. Di Fulvio et al.

Fig. 2 Example of graph showing the chipping time and power demand for a pine stem (Dbutt=10.4 cm, length= 3.23 m). The approximate cross-sectional area (grey curve) in contact with the knives is shown on the y-axis on the right (assumes constant feeding rate and linear variation in diameter from butt to top) and the power demand for chipping on the y-axis on the left the moment that 1) the power had reached a threshold of at least 6.0–9.2 kW above the idle power of the chipper (in the absence of logs) and 2) the power consumption increased at a rate of 1 kW s-1. The end of each repetition occurred when the power consumption

had dropped below 5.5 kW at a rate of 0.6–0.7 kW s-1 (Fig. 2). The absolute maximum power (kW) demand reached over the chipping time for each run represented the »maximum power demand«. The »energy

Table 2 Average results of the chipping experiments (standard deviation in parentheses). Required maximum power = P, energy demand = E, productivity = Pr Treatment Chip length, 8 mm

Chip length, 12 mm

Birch (n = 26)

Pine (n = 31)

Spruce (n = 30)

Birch (n = 26)

Pine (n = 32)

Spruce (n = 31)

25.0 (9.08)

24.1 (8.43)

21.7 (6.78)

22.1 (9.6)

24.9 (7.70)

25.0 (9.08)

3.91 (0.56)

3.33 (0.56)

3.27 (0.51)

3.37 (0.48)

2.77 (0.59)

3.91 (0.56)

2.85(0.20)

2.37 (0.23)

2.32 (0.31)

2.35 (0.24)

1.97(0.14)

2.85(0.20)

6.98 (1.0)

8.77 (1.38)

8.45 (1.29)

6.02 (0.89)

7.27 (1.42)

6.98 (1.0)

Enet*, kWh OD t

5.09 (0.37)

6.23 (0.64)

5.98 (0.81)

4.201 (0.43)

5.21 (0.38)

5.09 (0.37)

E, kWh m-3

3.54 (0.64)

3.38 (0.66)

3.19 (0.43)

2.84 (0.51)

2.78 (0.63)

3.54 (0.64)

Enet*, kWh m

2.57 (0.29)

2.39 (0.22)

2.26 (0.20)

1.98 (0.27)

1.99 (0.27)

2.57 (0.29)

Pr, OD t h-1

2.33 (0.88)

1.86 (0.72)

1.74 (0.64)

2.28 (0.99)

2.36 (0.86)

2.33 (0.88)

Pr, m h

4.42 (1.71)

4.89 (1.94)

4.59 (1.60)

4.87 (2.09)

6.24 (2.41)

4.42 (1.71)

Feeding speed, m s-1

0.19 (0.01)

0.19 (0.01)

0.19(0.01)

0.23 (0.02)

0.24 (0.02)

0.23(0.02)

Pmax, kW -1

E, kWh t

Enet*, kWh t-1 -1

E, kWh OD t

-1

-3

3

-1

*The power used when the machine was idling has been deducted from the total power

Croat. j. for. eng. 36(2015)1

89


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

Fig. 3 Required maximum power (Pmax) for chipping as a function of a log’s butt-end area; a) birch, b) pine, c) spruce

90

Croat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

demand« (kWh ODt-1; kWh m-3) was obtained by integration of the total power demand over the chipping time for each run divided by the mass/volume of each log. The mechanical energy required for the direct interaction between the knives and the wood, henceforth referred to as »net energy demand«, was calculated by integrating the difference between the total power over chipping time and the power used running the chipper in the absence of logs (estimated as the average power over at least 4 s immediately before or after the chipping). The »chipping productivity« (ODt h-1; m3 h-1) was obtained as the ratio between the OD mass/ solid volume of each log (output) and its chipping time. Statistical analyses of data were carried out using the software Minitab 16 (Minitab Ltd). General Regression Models (GLM) were used to detect significant factors and differences (p < 0.05) between chipping treatments and to model the results in the experiment. For the analysis of required maximum power (P = kW), energy demand (E = kWh ODt-1; kWh m-3) and productivity (Pr = ODt h-1; m3 h-1), the following hypothetic linear models were used:

P , E = m + ai(L) + b1 ( A) + b2 ( r) + ei

(1)

ln(Pr) = m + ai(L) + b1 (ln( A)) + b2 (ln( r)) + ei (2)

Where µ = overall mean, α = factor fixed effect, β1, β2 = constants and ε = random error. »L« is the length class (»short« and »long«) of the chips, »A« is the butt area (cm2) and »ρ« is the wood density (OD kg m–3). These factors were selected based on the work of Liss (1987, 1991) and the specific experimental factors. A correlation test was used to exclude collinearity among factors. The original data were transformed to the natural logarithmic scale for the productivity models in order to obtain linear relationships. Only the significant factors (p ≤ 0.05) were included into the final models. The dataset contained 52, 63 and 61 observations, respectively, for birch, pine and spruce; 10% of the original observations for each of the species were randomly extracted and reserved as witness samples for model validation by using paired t-tests with a 5% significance level.

3. Results The observed time for chipping each log varied between 11.5 and 24.9 s. There were obvious differences in required power (P) and energy demand (E) and in productivities (Pr) between the different tested chip sizes (Table 2). Croat. j. for. eng. 36(2015)1

F. Di Fulvio et al.

The power required for chipping consisted of two parts: one part was proportional to the mass flow through the chipper (for a given chip size), with the other being constant, regardless of chip production (including friction and powering of the hydraulic pump for the feed-rolls). The maximum power required for chipping (Pmax) was roughly proportional to the butt cross-sectional area of the stems, which explained most of the variability; it was almost independent of the chip length, which accounted for only slight variation (1%) (Fig. 3, Table 3). However, longer chips required significantly higher peak power than the shorter ones in the case of spruce (Eq. 5). In the case of birch, the maximum power absorbed was also directly correlated to the density, which accounted for a small part of the variability (2%) (Table 3, Eq. 3). The higher mass flow through the chipper for thicker logs meant that the chipping energy per m3 increased linearly with the reciprocal cross-sectional area (Fig. 4, Eqs. 9–11); the relationship was statistically significant in all cases (Table 3). This means that the chipping energy per m3 decreased with increasing stem diameter. When the energy required for running the chipper whilst not chipping was subtracted (Enet), the dependence on stem diameter became less evident. The chip size also had a significant effect for all species (Table 3), since longer chips reduced the energy requirements (Eqs. 6–11), due to the minor refinement. The chipping energy per m3 was about 20% less for a nominal chip size of 12 mm, compared to 8 mm. At the same time, an increase of OD density significantly increased the energy demand per m3, while the density was inversely proportional to the energy per OD t, since density and output (OD t) are directly related. The wood density effect was generally less than the stem size and the chipping length, becoming significant only in some of the cases (Eqs. 8–9), (Table 3). The significant relationships shown in Table 3 are described by the following equations:

Pmax birch = −14.6240 + 0.2651 ⋅ A + 0.0272 ⋅ r; kW (3) R2 (adj) = 89.18%, F = 190.499, p < 0.001

Pmax pine = 3.7002 + 0.2040 ⋅ A; kW

(4)

R2 (adj) = 83.93%, F = 293.425, p < 0.001

Pmax spruce = 1.6388 + 1.7673 ⋅ L + 0.2089 ⋅ A; kW

(5)

R2 (adj) = 88.47%, F = 208.157, p < 0.001

91


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

Fig. 4 Chipping energy per m3 as a function of the reciprocal of butt area; a) birch, b) pine, c) spruce

92

Croat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

F. Di Fulvio et al.

Table 3 Analysis of variance table for maximum power (Pmax) and energy demand (E) as a function of significant factors and variables included in the models Birch

Pmax, kW

E, kWh OD t-1

E, kWh m-3

Pine Pmax, kW

E, kWh OD t-1

E, kWh m-3

Spruce

Pmax, kW

E, kWh OD t-1

E, kWh m-3

Variables

DF

Adj SS

Adj MS

F

p-value

Butt area, cm2

1

3497.220

3497.220

366.98

<0.001

Density, OD t m-3

1

60.210

60.210

6.32

0.016

Residual error

44

419.310

9.530

Chip length, mm

1

15.408

15.408

65.52

<0.001

Butt area, cm2

1

29.451

29.451

125.23

<0.001

Residual error

44

10.348

0.235

Chip length, mm

1

3.141

3.141

47.73

<0.001

Butt area, cm2

1

6.984

6.984

106.12

<0.001

Density, OD t m

1

2.358

2.358

35.83

<0.001

Residual error

43

2.830

0.066

Variables

DF

Adj SS

Adj MS

F

p-value

Butt area, cm2

1

3069.290

3069.290

293.43

<0.001

Residual error

55

575.310

10.460

Chip length, mm

1

30.654

30.654

50.12

<0.001

Butt area, cm2

1

84.703

84.703

138.49

<0.001

Residual error

54

33.028

0.612

Chip length, mm

1

4.596

4.596

51.54

<0.001

Butt area, cm2

1

19.890

19.890

223.02

<0.001

Residual error

54

4.816

0.089

DF

Adj SS

Adj MS

F

p-value

Chip length, mm

1

42.810

42.810

6.66

0.013

Butt area, cm2

1

2591.600

2591.600

402.90

<0.001

Residual error

52

334.490

6.430

Chip length, mm

1

22.585

22.585

46.28

<0.001

Butt area, cm2

1

62.204

62.204

127.46

<0.001

Density, OD t m-3

1

30.520

30.520

62.53

<0.001

Residual error

51

24.891

0.488

Chip length, mm

1

3.021

3.021

51.91

<0.001

Butt area, cm2

1

9.695

9.695

166.57

<0.001

Residual error

52

3.027

0.058

-3

Variables

Croat. j. for. eng. 36(2015)1

93


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

(6)

Etot birch = 5.2067 − 1.1704 ⋅ L + 149.088 ⋅ A −1 ; kWh ODt –1

R2 (adj) = 77.48%, F = 80.127, p < 0.001 Etot pine = 6.5568 − 1.4683 ⋅ L + 186.598 ⋅ A −1 ; kWh ODt –1

(7)

Prbirch = e −3.5721+0.2166⋅L ⋅ A1.1030 ; m 3 h –1 (8)

kWh ODt –1

R2 (adj) = 77.99%, F = 64.788, p < 0.001

(16)

R2 (adj) = 92.26%, F = 334.658, p < 0.001 (9)

R (adj) = 82.39%, F = 72.731, p < 0.001 2

Prspruce = e −3.3025+0.1921⋅L ⋅ A1.0596 ; m 3 h –1

(17)

R2 (adj) = 92.27%, F = 323.116, p < 0.001

Etot pine = 2.3135 − 0.5697 ⋅ L + 90.4222 ⋅ A −1 ; kWh m –3 (10)

R2 (adj) = 83.45%, F = 142.135, p < 0.001 Etot spruce = 2.1027 − 0.4704 ⋅ L + 91.3848 ⋅ A −1 ; kWh m –3 (11)

R2 (adj) = 81.18%, F = 117.475, p < 0.001 where L = 0 for the chip length »8 mm« and 1 for the chip length »12 mm«. The productivities (m3 h-1), based on effective chipping work time (excluding all waiting times), were strongly correlated to the stem diameter (Fig. 5, Table 4).The productivities were roughly proportional to the cross-sectional area of the logs fed into the chipper, while the feeding speed difference was small (0.04 m/s) almost constant with a little reduction in feeding rate seen for thicker logs (Fig. 5). As the chip length was increased from 8 to 12 mm, the productivity increased, and the difference between the two sizes was 22% at a cross-sectional area of 80 cm2 (Fig. 5). The OD density became significant, in the case of spruce (Eq. 14), and it was directly correlated with the productivity in terms of OD t, due to the fact that the mass fed to the machine increased with log density. Based on the significant relationships found in the analyses (Table 4), the following two models were created:

94

(15)

R2 (adj) = 92.68%, F = 292.080, p < 0.001

Prpine = e −3.5059 +0.2170×L ⋅ A1.1032 ; m 3 h –1

Etot birch = −0.3146 − 0.5782 ⋅ L + 72.7758 ⋅ A −1 + 0.0059 ⋅ r;

R2 (adj) = 92.91%, F = 201.835, p < 0.001

Prspruce = e −11.5447 +0.2011⋅L ⋅ A1.0713 ⋅ r1.2157 ; ODt h –1 (14) R2 (adj) = 92.03%, F = 208.781, p < 0.001

Etot spruce = 13.8361 − 1.3218 ⋅ L + 236.837 ⋅ A −1 − 0.02154 ⋅ r;

Prbirch = e −4.2031+0.1521⋅L ⋅ A1.0866 ; ODt h –1

(13)

R2 (adj) = 91.03%, F = 285.275, p < 0.001

R2 (adj) = 77.61%, F = 98.057, p < 0.001

kWh m –3

Prpine = e −4.1708 +0.2150⋅L ⋅ A1.0372 ; ODt h –1

(12)

where L = 0 for the chip length »8 mm« and 1 for the chip length »12 mm«. The t-tests for validation showed, in all cases, pvalues > 0.05. The smallest p-value was 0.124 in the case of Eq. 12 for birch productivity (m3 h-1). When the 12 mm chipper knife setting was used (long), birch logs with smaller diameters tended to result in finer chips than those with larger diameters (Fig. 6), with a similar trend for spruce but not for pine. When the 8 mm chipper knife setting was used (short) on pine and spruce, there was an opposite trend with smaller diameter stems resulting in coarser chips than the larger diameters. There were very few particles (< 0.5%) that did not pass through the 31.5 mm sieve, which means that the chip quality should be adequate for combustion in small-scale generators, even with the chipper set for 12 mm. In general, if chipper settings (i.e. short and long) are considered together, a significantly higher mass fraction of the birch chips was in the size interval l8–16 mm, compared to spruce (p = 0.037). As a consequence, compared to the two softwoods, significantly (p = 0.015) fewer birch chips fell into the size interval 3.15–8 mm. Thus, birch chips tend to be slightly larger than softwoods in general; however the combination of species and chipping length masked those differences to a certain extent (Table 5).

4. Discussion The results from this study mostly agree with findings in the literature on wood chipping processes. The Croat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

F. Di Fulvio et al.

Fig. 5 Chipping productivity as a function of butt area; a) birch, b) pine, c) spruce Croat. j. for. eng. 36(2015)1

95


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

Table 4 Analyses of variance table for the productivity (Pr) as a function of significant factors and variables included in the models Birch

Ln (Pr), OD t h-1

Ln (Pr), m3 h-1

Pine

Ln (Pr), OD t h-1

Ln (Pr), m3 h-1

Spruce

Ln (Pr), OD t h-1

Ln (Pr), m3 h-1

Variables

DF

Adj SS

Adj MS

F

p-value

Chip length, mm

1

0.232

0.232

14.89

0.001

Ln (Butt area, cm2)

1

8.009

8.009

514.83

<0.001

Residual error

44

0.685

0.016

Chip length, mm

1

0.521

0.521

36.88

<0.001

Ln (Butt area, cm2)

1

8.253

8.253

583.91

<0.001

Residual error

44

0.622

0.014

Variables

DF

Adj SS

Adj MS

F

p-value

Chip length, mm

1

0.592

0.592

26.96

<0.001

Ln (Butt area, cm2)

1

11.745

11.745

534.56

<0.001

Residual error

54

1.186

0.022

Chip length, mm

1

0.607

0.607

28.89

<0.001

Ln (Butt area, cm2)

1

13.241

13.241

630.25

<0.001

Residual error

54

1.135

0.021

Variables

DF

Adj SS

Adj MS

F

p-value

Chip length, mm

1

0.527

0.527

37.88

<0.001

Ln (Butt area, cm2)

1

8.109

8.109

582.65

<0.001

Ln (Density, OD t m-3)

1

0.711

0.711

51.12

<0.001

Residual error

51

0.710

0.014

Chip length, mm

1

0.505

0.505

35.86

<0.001

Ln (Butt area, cm2)

1

8.287

8.287

588.57

<0.001

Residual error

52

0.732

0.014

Table 5 Influence of species and chipper knife settings on the particle size distribution as percentage of total sieved mass Birch

Pine

Spruce

Particle size interval, mm

Long (n = 5)

Short (n = 5)

Long (n = 5)

Short (n = 5)

Long (n = 5)

Short (n = 5)

<3.15

4.8 bc*

4.5 c

4.9 bc

6.2 a

5.7 ab

6.1 a

3.15–8

23.4 c

42.7 ab

32.1 bc

47.3 a

33.9 bc

45.7 a

8–16

65.9 a

48.4 bc

58.9 ab

44.8 c

55.5 abc

46.2 c

16–31.5

5.5 a

4.3 ab

4.0 abc

1.6 c

4.8 a

2.0 bc

>31.5

0.4 a

0.1 a

0.1 a

0.0 a

0.1 a

0.0 a

*The values in rows that do not share any letters are statistically different (p 0.05) according to Tukey’s test

96

Croat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

F. Di Fulvio et al.

Fig. 6 Size distribution of the chips from birch (top), pine (middle) and spruce (bottom) for the two chippers settings (long/short chips) and tree butt-end diameter class (1–5) Croat. j. for. eng. 36(2015)1

97


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

maximum power required for chipping is not strongly affected by the chip length, but is mostly related to the diameter of a log’s butt end. One explanation for this is that the butt end of a log causes the maximum »absorption« of power during the chipping process; the peak chipping power was 24–125% higher than the average absorbed when chipping the rest of the log. If the power required to run the chipper empty was removed from the total (i.e. when the chipper was idling), the net power for chipping was proportional to the butt area. This is in agreement with the results of Nurmi (1986) and Liss (1987), while Van Belle (2006) found an exponential relationship with the diameters of wood parts when chipping logging residues. The productivity was significantly lower when producing shorter chips and more energy was used. This is due to the increased number of required cuts per wood piece, which slows down the process. These findings agree with Liss (1987, 1991), Nurmi (1986), Facello et al. (2013) and Spinelli and Magagnotti (2012). The energy consumption for the production of small chips (8 mm; 1.3–4.5 kWh m-3) was in line with results found by Liss (1987), who studied tree species, similar stem sizes and similar feeding inclination angle (50°). However, Liss’ (1987) measurements were made directly on the shaft connecting the tractor and the chipper, which means that it did not include additional energy losses in the process. In the present study, these losses amounted to 8.1% of the total energy requirements. The energy requirements for chipping in this study corresponded, on average, to 0.085% of the produced fuel energy content (lower heating value). In most other studies, losses in the diesel engine are included in the measured values, which change the results considerably. For example, Liss (1987) estimated that the amount of energy produced by the diesel fuel was a factor of 3.3 larger than the resulting power on the tractor PTO shaft. He also considered that the fuel consumption under operational conditions would be doubled when the waiting time is included. In this study, the productivity (ODt h–1) increased with butt end area by up to a power of 1.04–1.09, whereas Van Belle (2006) found it to be 1.05. It seems that the productivity increase is almost proportional to the volume feeding rate and that the specific energy requirements can be reduced with more efficient feeding. Thus, an efficient use of the feed opening and available chipping time are both important. For very small trees and logging residues, feeding of many stems at a time should increase the efficiency. One way to reduce the waiting time could also be to have a larger feeding board. Nurmi (1986) found that, for a given cross-sectional area, bunches required less pow-

98

er when compared to single stems. This fact indicates that there are advantages to having the biomass compressed/bundled to diameter sizes that match the feeding inlet of the chipper in comparison to loose material (cf. Nurmi 1986). As birch wood had a lower MC than the softwoods in this study, only the combined influence of OD density and MC could be estimated. However, for the studied three species, the experiment was representative for chipping fresh wood. Although wood logs were all sampled from one stand, different levels of variability in wood densities were observed for the three species due to their different growth characteristics. This study made it possible to observe some of the effects due to this variability, but a larger sampling could have made it easier to investigate the effect of densities to a larger extent. Tree species seems to have some effect on the energy demand, due to the higher density of hardwood (birch), which corresponded to a 3–4% higher energy demand per m3 than for softwood (i.e. pine and spruce) at a butt diameter of 10 cm (a butt-end area of 80 cm2). The feed to the chipper used in the present study had a disc side angle of about 45° and should be rather energy efficient compared to feeding the logs in perpendicular (90°) to the plane of rotation of the disc. Papworth and Erickson (1966) found that an increasing side angle decreased the specific energy consumption. In addition to reducing the waiting time for the chipper, the energy efficiency of the process can be improved, if it is possible to reduce the energy demands of the hydraulic system and the fan when no chipping takes place. Compared to a tractor-driven »Farmi« chipper of similar size studied by Van Belle (2006), the chipper used in the present study required about 3.5 kW when no feeding took place, below the 4.6–4.9 kW measured by Spinelli et al. (2013), but above the mechanically required minimum of 1.4 kW as measured by Van Belle (2006). Thus, it seems that the Edsby chipper could be improved. For example, Liss (1987) suggested using a conveyor instead of a fan for removing the chips and suggested that the waiting periods during crane movements could be used for emptying the chips as a way to reduce the power requirement. The energy used to start-up the chipper was about 0.050 kWh, which is comparable to the chipping energy required for a single log. From the mass and diameter of the disc, and from the rotational speed of 540 rpm, it can be roughly estimated that accelerating and breaking the disc requires about 0.016 kWh. A disc with a larger inertial momentum would require less power; on the other hand, this would increase the energy needed for start-up. These observations give an Croat. j. for. eng. 36(2015)1


Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

idea of how much energy is wasted when starting or turning off a similar machine. In addition, it was observed that the idling power decreased when the chipper had warmed up. This was not expected for an electric engine and was probably due to the warming of the hydraulic oil, which affected the results (on average 1.3% of the maximum power demand). During the experiments, some wood pieces about 10 cm long, from previous chipped logs, remained in the chipper in each run (i.e. wood pieces stayed between the feed rollers and the disc). This could also have generated some errors in the recording of absolute power; however, the error was almost constant between repetitions. Nevertheless, this could have had some influence on the quality of the chips. For this reason, we avoided sampling chips immediately at the starting of a new run. Different bouncing disturbances were observed in case of different species and chip lengths, thus a variable threshold was used to identify the starting of chipping process, and this could have also introduced some marginal uncertainty in the analyses. With regard to the prospects of using hybrid power systems for chipping and grinding equipment, several basic decisions have to be made. The system may consist of only a chipper/grinder, or be a more complex system including other components such as a crane used for loading. Depending on the actual system, the engine load pattern will be different. The amount of energy to be stored and the required power will have to be estimated. When optimizing such a system, the cost of additional components such as an electric motor and a battery must be compared to the efficiency gains.

5. Conclusions This study provides models which could be used for estimating the power and energy requirements, along with the chipping productivity, of a 30 kW electric disc chipper for chipping small diameter pine, spruce and birch trees from early thinnings in the north of Sweden. The following factors have a significant effect on the power requirements and productivity when chipping small logs from thinnings: chip length, butt area and density. The maximum power required for chipping was roughly proportional to the butt cross-sectional area of the stems, which explained most of the variability; it was almost independent of the chip length. The energy needed per m3 decreased by 4% for each centimetre increase in stem butt diameter. The Croat. j. for. eng. 36(2015)1

F. Di Fulvio et al.

chip size also had a significant effect, and the chipping energy per m3 was about 20% less for a nominal chip size of 12 mm, compared to 8 mm. Increases in OD density increased the energy demand per m3. The productivities were roughly proportional to the cross-sectional area of the logs fed into the chipper. As the chip length increased from 8 to 12 mm, the productivity increased, and the difference between the two sizes was 20% at a cross-sectional area of 80 cm2. The chipping of hardwood (birch) corresponded to a slightly higher (3–4%) energy demand per m3 than for softwood (i.e. pine and spruce). These findings are in line with previous studies and the results could be used for designing small hybrid chippers.

Acknowledgements The research program Botnia–Atlantica – the project Forest Refine – is gratefully acknowledged for financial support. We especially acknowledge Otto Läspä and Raul Fernandez Lacruz for their assistance in the chipping test. The forest company Sveaskog is acknowledged for providing biomass for the experiments.

6. References Abdallah, R., 2011: Experimental study about the effects of disc chipper settings on the distribution of wood chip size. Biomass and Bioenergy 35: 843–852. Aman, A.L., Baker, S.A., Greene, W.D., 2011: Productivity and product quality measures for chippers and grinders on operational southern US timber harvests. International Journal of Forest Engineering 22(2): 7–14. Facello, A., Cavallo, E., Magagnotti, N., Paletto, G., Spinelli, R., 2013: The effect of chipper cut length on wood fuel processing performance. Fuel Processing technology 116: 228–233. Hartler, N., 1986: Chipper design and operation for optimum chip quality. Tappi Journal 69(10): 62–66. Ghaffariyan, M., Spinelli, R., Brown, M., 2013: A model to predict productivity of different chipping operations. Southern Forests: a Journal of Forest Science 75(3): 129–136. Hellström, L.M., Gradin, P.A.,Carlberg, T., 2008: A method for experimental investigation of the wood chipping process. Nordic Pulp and Paper Research Journal 23(3): 339–342. Hellström, L.M., Isaksson, P., Gradin, P.A., Eriksson, K., 2009: An analytical and numerical study of some aspects of the wood chipping process. Nordic Pulp and Paper Research Journal 24(2): 225–230. Kivimaa, E.,Murto, J., 1949: Investigations on factors affecting chipping of pulp wood. VTT, the State Institute for Technical Research, Helsinki, Finland, Publication 9.

99


F. Di Fulvio et al.

Effects of Wood Properties and Chipping Length on the Operational Efficiency of a 30 kW ... (85–100)

Larsson, G., 2012; Hybridization of off-road mobile machinery (Hybridiseringavmobilamaskiner, in Swedish). Swedish University of Agricultural Sciences, Department of Energy and Technology. Liss, J. E., 1987: Power requirement and energy consumption in fuel-chip production using a tractor-mounted chipper. Swedish University of Agricultural Sciences, Department of Operational Efficiency, Report No 173, Garpenberg, Sweden. Liss, J. E., 1991: Bränsleflisensfraktionsfördelning – Enstudieavnågramaskin- ochvedparametrarsinverkanpåfliskvaliteten. Fractional distribution on fuel chips – Machine and wood parameters influence on chip quality. UppsatserochResultat nr 208. Department of Operational Efficiency, Swedish University of Agricultural Sciences, Garpenberg. Nurmi, J., 1986: Chunking and chipping with conescrew chipper. Folia Forestalia 659(12): 22 p. Papworth, R.L., Erickson, J.R., 1966: Power requirements for producing wood chips. Forest Products Journal 16(10): 31–36. Röser, D., Mola-Yudego, B., Prinz, R., Emer, B., Sikanen,L., 2012: Chipping operations and efficiency in different operational environments. Silva Fennica 46(2): 275–286. Routa, J., Asikainen, A., Björheden, R., Laitila, J., Röser,D., 2012: Forest energy procurement: state of the art in Finland and Sweden. WIREs Energy Environ 2012. doi: 10.1002/ wene.24. Spinelli, R., Magagnotti, N., Paletto, G., Preti, C., 2011: Determining the impact of some wood characteristics on the performance of a mobile chipper. Silva Fennica 45(1): 85–95.

Spinelli, R., Cavallo, E., Facello, A., 2012: A new comminution device for high-quality chip production. Fuel Processing Technology 99: 69–74. Spinelli, R., Magagnotti, N., 2012: The Effect of Raw Material, Cut Length, and Chip Discharge on the Performance of an Industrial Chipper. Forest Products Journal 62(7–8): 584–589. Spinelli, R., Cavallo, E., Eliasson, L., Facello, A., 2013: Comparing the efficiency of drum and disc chipper. Silva Fennica 2013(2), article id. 930, 11 p. Sun, H., Yang, L., Jing, J., 2010: Hydraulic/electric synergy system (HESS) design for heavy hybrid vehicles. Energy 35(12): 5328–5335. Swedish Energy Agency, 2010: Motor data, IE2 and IE3 motors (Motordata, IE2- och IE3-motorer (50 Hz, 400 V), in Swedish), available at www.energimyndigheten.se, downloaded on 15/03/2013. Swedish Transport Administration, 2012: Climatic impact of working machines and possible reductions. Data from the Swedish Forest Agency, 2012. Uhmaier, A., 1995: Some fundamental aspects on wood chipping. Tappi Journal 78(10): 79–86. Van Belle, J.-F., 2006: A model to estimate fossil CO2 emissions during the harvesting of forest residues for energy – with an application on the case of chipping. Biomass and Bioenergy 30(12): 1067–1075. Yoshioka, T., Aruga, K., Nitami, T., Sakai, H., Kobayashi, H., 2006: A case study on the costs and the fuel consumption of harvesting, transporting, and chipping chains for logging residues in Japan. Biomass and Bioenergy 30(4): 342–348.

Authors’ address:

Received: December 20, 2013 Accepted: July 28, 2014

100

Fulvio Di Fulvio, PhD.* e-mail: fulvio.di.fulvio@slu.se Gunnar Eriksson, PhD. e-mail: gunnar.eriksson@slu.se Dan Bergström, PhD. e-mail: dan.bergstrom@slu.se SLU, Swedish University of Agricultural Sciences Department of Forest Biomaterials and Technology Skogsmarksgränd, SE-901 83 Umeå SWEDEN * Corresponding author Croat. j. for. eng. 36(2015)1


Preliminary note

Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops Vincenzo Civitarese, Angelo Del Giudice, Alessandro Suardi, Enrico Santangelo, Luigi Pari Abstract The particle size distribution of wood chips, along with the moisture content, are some of the main parameters for defining the quality of most wood fuels. A new experimental rotor, powered by the self-propelled forage harvester Claas Jaguar was developed by the Consiglio per la ricerca in agricoltura e l'analisi dell'economia agraria (CRA), Agricultural Engineering Research Unit (CRA-ING). The rotor allowed for improved dimensional features of wood chips. The comminution achieved with the CRA–ING drum increased the percentage of 16–45 mm wood chips fraction from 63.69% to 73.29%, and progressively reduced the fraction of chips less than 16 mm from 35.20 to 25.35%. Consequently, the bulk density of the chips decreased by 8.57% in comparison with products obtained by standard devices. The dimensional increments achieved by the rotor and the percentage reduction of the smallest fractions represent two valuable elements affecting the behaviour of the wood chips during storage and handling. Keywords: rotor, poplar, short rotation coppice, harvesting, comminution, wood chips

1. Introduction As a source of lignocellulosic biomass, energy crops are used as fuel in power stations, combined heat and power (CHP) plants, biogas stations, large heating plants and small combustion units (Abdallah et al. 2011). Among lignocellulosic perennial crops, poplar is considered as the best suited for the Italian environmental conditions, and plantation programmes in Italy are largely based on this species (Spinelli et al. 2009). Several studies have reported high productivity levels of the energy crop grown in Italy (Paris et al. 2011, Bergante et al. 2010, Makeschin 1999). In general, techniques for harvesting poplar can be broadly grouped into the single pass cut and chip, and the whole stem systems (Mattison and Mitchell 1995). Since the latter method is not commonly used at present, biomass is usually stored in comminuted form. Although harvesting coincides with period of high demand for wood fuel, short time storage is unavoidable. The storage phase plays a pivotal role on the energy balance, and the economic efficiency is linked to several factors, especially the handling cost and the susceptibility to decay. It is therefore necessary to idenCroat. j. for. eng. 36(2015)1

tify methods and conditions for the storage of comminuted wood, which could minimise loss of biomass or decline in quality. Particles size distribution of wood chips is one of the most important parameters characterizing the biomass quality (Hartmann et al. 2006, Paulrud and Nilson 2004, Suadicani and Gamborg 1999), since it influences the storage behaviour (Jirjis 1995, Jirjis 2005, Barontini et al. 2013), the handling properties (Nati et al. 2010, Spinelli et al. 2012) and the combustion efficiency (Wu et al.2011). High proportion of small particles or fines may be detrimental during storage of chips since it results in compaction of chip piles and subsequent reduction in air movement. It may contribute to the maintenance of a high moisture content and, consequently, a delay in the dissipation of heat, which is a key factor in cooling chip piles (Kubler 1982, Afzal et al. 2010). Under such conditions, the risk of spontaneous fuel ignition increases (Kubler 1987) and the loss of dry matter may rise up to a rate of 3% per month (Mattison and Mitchel 1995). Another important aspect to consider is the heterogeneous particle size distribution that might

101


V. Civitarese et al.

Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

create problems during combustion of the fuel (Jirjis 1995). Shifting particle size distribution towards higher dimensional classes and/or improving the homogeneity of the wood chips would yield a high grade fuel. Comminution relies on some mechanical properties of the wood such as the cleavage strength, the shear strength parallel to the grain, and the compressive strength parallel to the grain (Abdallah et al. 2011, Mc Lauchlan and Lapointe 1979, Twaddle 1997). Such properties vary according to the anatomical plane because the cell elements are inter-connected by forces of various type and intensity, mainly covalent bonds along the longitudinal axis and secondary bonds on the transverse plane (Goli et al. 2004). The components of the compressive forces react with the splitting force of the wood and with the shear stress parallel to the arrangement of fibres leading to chip formation. The thickness of the chips can be increased by reducing the cutting angle (Buchanan and Duchinicki 1963, Monico and Soule 1979), while the increasing of the cutting speed results in high percentage of small chips and fine sized particles (Edelma and Stuart 1992, Hartler 1986, Hernandez and Jacques 1997). Currently, harvesting of short rotation coppice (SRC) in Italy is performed using large size foragers equipped with dedicated SRC headers such as the Claas (Spinelli et al. 2009), Krone or John Deere (Spinelli et al. 2011), although the Claas Jaguar is by far the most popular (Spinelli et al. 2008). Rotors mounted on such machines were designed for harvesting grass crops. Jirjis et al. (2008) and Pari et al. (2008) noted that the reduced size of product obtained from such rotors could negatively affect the storability. With this background, CRA–ING started a program aimed at developing a new drum chipper for application in self propelled forage harvesters in order to shift the particles size distribution towards the 16– 45 mm dimensional class. A new rotor was designed after a preliminary experience, that showed the feasibility of increasing the longitudinal section of the chips (Pari et al. 2009, Pari et al. 2010). The objective of this study was to evaluate the quality of products obtained from the CRA–ING rotor compared with that of a standard design in terms of bulk density, particle size distribution and size of individual chips.

side, equally distributed on a drum weighting 195 kg and having diameter and length of 630 mm and 750 mm, respectively. The inclination of the bladeholders is 136° while the cutting angle of the knives reaches 32.5°. When operating on wood species, the device usually uses 12 out of the 24 knives (6 per side) but the other bladeholders interact passively during the chipping. The CRA–ING rotor differs from the standard in terms of weight, number of bladeholders and knives, bladeholders inclination and cutting angle of the knives. The new device weighs 256 kg and is equipped with 10 bladeholders and 10 knives. The bladeholders are inclined at 129° resulting in a cutting angle of the knives of 22°, as shown in Table 1. Table 1 Technical characteristics of the tested rotors Characteristics

Units

Rotor Standard

CRA–ING

195

256

Weight

kg

Diameter

mm

630

Length

mm

750

Bladeholder Length Inclination Knives Angle of cut

n

24

mm

10 340

degrees

136°

129°

n

12–24

10

degrees

34°

22°

The study was carried in the year 2011 in the Treviso province (Italy). A Claas Jaguar 890 was used to harvest a poplar plantation in an area measuring 4000 m2. The spacing between plants were 3 x 0.5 m, and the roots and stems were aged four and two years (R4S2), respectively. Two treatments were compared, the standard and the CRA–ING rotor both over an area of 2000 m2 at the same forward speed of 1.25 m s–1. During the study, the forage harvester worked first with the standard rotor and, after a week, with the rotor designed by CRA–ING (Fig. 1).

2.2 Qualitative assessment of wood chip

2. Materials and methods 2.1 The rotors The standard rotor of Claas Jaguar 800 series mounts 24 bladeholders each 340 mm long, 12 per

102

In accordance with Mitchel et al. (1997), the following traits were recorded over the entire area: plant density, percentage of lacking plants, diameter and height of the main and secondary stems, number and size of the branches (Table 2). Croat. j. for. eng. 36(2015)1


Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

V. Civitarese et al.

Table 2 Site description and morphological characterization (means ±standard deviation) of the poplar stand at the second cycle (R4S2: root 4 years; stem 2 years) Site

Ca` Tron Treviso

Cutting cycle

R4S2

Surface, ha

0.41

Elevation (m a.s.l.)

0

–1

Plant density, p ha

Theoretical

6666

Effective

5866

Shoots/coppice, n.

5.1±2.18

Main stem

Height, m

8.58±0.49

Diameter, mm

68.18±9.60

Secondary stems

Height, m

5.35±1.97

Diameter, mm

35.84±15.16

Branches, n.

Fig. 1 Experimental drum mounted on self propelled chipper Claas Jaguar 890 At the collection point, wood chips produced by the two rotors were stored in two separate piles. Moisture content was determined by collecting six samples of approximately 500 g from each heap. The samples were immediately sealed in suitable non breathable bags and transported to the laboratory, where they were left to dry in an oven with forced ventilation at a temperature of 103 ± 2°C according to EN 14774-2. The bulk density was measured using a steel cylinder of known internal volume following EN 15103. A cylinder of 0.026 m3 was filled with wood chips and then weighed using a dynamometer. For each treatment, the weights of 8 samples were measured and recorded. The ratio between the net weight of samples in the cylinder and its internal volume represented the bulk density, expressed in kg m-3. For the particle size characterization, an amount of 10 kg (corresponding approximately to 24 l) of wood chips for each rotor was collected. After dying in air, ten sub-samples of 1 kg were used for the sieving in order to avoid the overloading of the mechanical sieve shaker (Analysette 18, Fritsch) and favouring the optimal separation of the wood chips. As required by the Croat. j. for. eng. 36(2015)1

52.80±21.09

Diameter, mm

6.60±2.64

Length, m

0.83±0.33

European Standard, this is to ensure that the filling height on the upper sieve shall never exceed 5 cm (CEN/TS 15149-1). Four sieves (normalized in accordance with ISO 3310-1) were used in order to separate the five following chip length classes: 100–63 mm, 63–45 mm, 45–16 mm, 16–3.15 mm and < 3.15 mm (CEN/TS 15149-1).

Table 3 Means (± standard deviation) of the main dimensional parameters measured for the chips falling in the 16–45 mm class obtained from the two rotors (for each rotor n. = 50) Parameters

Rotor

Increment, %

Claas

CRA–ING

Length, mm

38.57±5.85

43.39±8.07

12.5

Width, mm

22.67±6.25

28.62±6.94

26.2

Thickness, mm

13.05±3.41

16.76±5.81

28.4

Weight, g

3.28±1.69

5.47±2.56

66.8

Volume, ml

8.18±4.49

11.07±7.06

35.3

103


V. Civitarese et al.

Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

Fig. 2 Position of measurements carried out on each anatomical direction of chips: A – length, B – width, C – thickness

2.3 Statistical analysis

The study was completed through a careful analysis of the sizes of chips belonging to the most representative size fractions (16–45 mm). A sample of 50 chips was drawn from the 16–45 mm size fraction of each rotor and their weight, volume, dimensions (length, width and thickness) of each chip were measured using a precision balance (d:0.01), a xilometer and an electronic caliper, respectively. Owing to the uneven shapes of the chip profiles, their dimensions were measured and reported as means of three measurements made at three different positions on each anatomical direction (Fig. 2).

All data were analyzed with the PAST, Statistica and MSTATC software, in order to check the statistical significance of the differences between treatments. Homoscedasticity and normality were checked before testing. The data collected were statistically analyzed using the Student’s t-test for the bulk density and moisture content evaluation. The particle size distribution was analyzed using two way ANOVA, where the rotor and the size class were the factors analyzed. For the analy-

Table 4 Results from the 50–50 Manova (for each rotor n. = 50) Source

DFa

exVarSS

nPC

nBu

exVarPC

exVarBU

p-Value

Rotor

1

0.134360

2

3

0.835

1.000

0.000000

Error

98

0.865640

DFa – Degrees of Freedom; exVarSS – explained variances based on sums of squares; nPC – number of principal components used for testing; nBu – number of principal components used as buffer components; exVarPC – variance explained by nPC components; exVarBU – variance explained by (nPC+nBU) components; pValue – the result from 50–50 MANOVA testing

Table 5 Rank of variables analyzed by rotation simulation test (for each rotor n. = 50) RankNra

varName

pRaw

pAdjFDR

p99999

1

Weight

0.000020

0.000020

0.000020

2

Width

0.000018

0.000065

0.000100

3

Thickness

0.000183

0.000293

0.000520

4

Length

0.000906

0.001242

0.002070

5

Volume

0.021323

0.021294

0.020620

Rank Nra – rank of the variables analyzed; pRaw – ordinary univariate p-values; pAdjFDR – adjusted p-values according to false discovery rates; p99999 – adjusted p-values according to the familywise error rate

104

Croat. j. for. eng. 36(2015)1


Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

Fig. 3 Chip size distribution (% ±S.E) for the wood chips collected using Claas and CRA rotors. Before the ANOVA analysis (Two Way Anova: interaction factors A – B <0.001), the data were transformed as square root of the arcsine. Different letters indicate a significant difference at the level of p 0.05 after HSD Tukey’s test sis, the data of the frequency (%) of chip classes were transformed as square root of the arcsine. After ANOVA, the Tukey’s post-hoc test (significance level α = 0.05) was applied. The 50–50 MANOVA was used for the dimensional analysis of the chips. The method is a modified variant of classical MANOVA that integrates the Principal Component Analysis (PCA) in its algorithm. In this way, the dimensionality of the data is reduced by using principal component decompositions and the final tests are performed by ordinary MANOVA. The data were standardized before MANOVA. Ranks of the variables observed (volume, length, width, weight and thickness) were analyzed by using the rotation test, an application of the 50–50 MANOVA. The rotation test adjusts the single response p-values keeping the type I error controlled according to the familywise error rate criterion. In this way the adjusted p-values can now be compared to the same threshold level of significance (α).

3. Results A first clue of the rotor effect on the modification of chip size was appreciable by observing the values of bulk density. These varied from 288.12 ± 4.08 kg m–3 obCroat. j. for. eng. 36(2015)1

V. Civitarese et al.

tained from the standard device to 263.42 ± 13.91 kg m–3 by the CRA–ING rotor. Moisture contents were 55.27% ± 0.63 and 54.60% ± 0.60 for the standard and CRA rotor, respectively. The results of t-tests showed a statistically significant difference between the bulk densities (p-value 0.0012) but not for the moisture contents (p-value 0.055). From a practical standpoint, the lower bulk density could lead to a 8.57% reduction of material delivered per trip, that is 1.97 t less for a truck of 80 m3 capacity. Moreover, concerning the storage, a heap of 50 t would require a volume around 190 m3, 16 m3 more than the corresponding pile built with wood chips comminuted by the standard rotor and mainly attributable to a higher proportion of macroporosity. Concerning the particle size distribution (Fig. 3), it is apparent that the presence of oversized (45–63 mm) and undersized (< 3.15 mm) materials were extremely limited in both cases. Fractions ranging from 45 to 8 mm were the most represented for both rotors, accounting for about 97%, which assured compliance with the quality specifications described in the CEN/TS 14961 standard for wood chip commercialization. Accordingly, for the P45 class, the gross fraction identified as the particles quantity with dimensions exceeding 63 mm, has a limit of 1%, while the finer parts (< 1 mm) have a limit of 5%. Interestingly, the comminution achieved with the CRA–ING drum increased the percentage of 16–45 mm fraction from 63.69 to 73.29% and progressively reduced the proportion of classes lower than 16 mm (from 35.20 to 25.35%). Such differences assumed a statistical significance for the fractions 16–45 mm and 3.15–8 mm. These results suggest that the increased percentage of chips produced by CRA–ING rotor in the 16–45 mm class was due to a general increase in the size of the individual chips at the expense of the lower size fractions. The additional dimensional analysis carried out on individual wood chips revealed more details on the positive effects deriving from the use of the new rotor (Table 3), emphasizing the positive percentage increment for all the dimensional factors, especially for weight and volume. The results of 50–50 MANOVA test showed a statistically significant difference among the size of the chips produced by the two rotors (Table 4). The rotation test included in the 50–50 MANOVA test (Table 5) helped to identify the rank of the variables analyzed (rank Nr column), according to adjusted p-values controlled by familywise error rate (p99999). Among these, weight and width were found to be the most important variables. Such analysis is extremely helpful to identify the main factors that cause an increase in the dimensions of chips.

105


V. Civitarese et al.

Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

4. Discussion The main devices producing wood chips are drum and disc chipper. The size of wood chips is one of the parameters used to measure the quality of the chips produced, bought or sold and that must comply with quality standards. Therefore, comminution required for massive mobilization of feedstock is a crucial step along the supply chain. Experimental works on parameters affecting the wood chips shape are rather scarce and, as stated by Abdallah et al. (2011, 2014) and Krajnc and Dolsac (2014), there is inadequate information on mechanism of chip formation. Spinelli and coworkers (Nati et al. 2014, Spinelli et al. 2013, Nati et al. 2010, Spinelli et al. 2005) have conducted a series of studies focused on the comparison among two main chipping devices (drum and disk). Also, Abdallah et al. (2011, 2014) performed a deep analysis about the effect of various factors affecting disc chipper, whilst Krajnc and Dolsac (2013, 2014) studied the parameters influencing the wood chip production using a drum chipper. Therefore, to our knowledge, the present work gives a significant contribution to this issue. As reported, CRA–ING brought some changes to the standard rotor in order to produce larger wood chips aimed at improving some physical properties such as air permeability during the storage. The reader must keep in mind that the study, starting from a basic research on constructional parameters, was aimed at answering an applicative issue such as improving storage conditions; hence, wood chip quality. The new rotor was heavier than the standard, but with less bladeholders (10 vs 24) and knives (10 vs 12), a reduced inclination of both the bladeholders (129° vs 136°) and cutting angle of the knives (22° vs 32.5°). A first concrete result, suggesting the effectiveness of the modifications, was the reduction of the bulk density of the chips, due to the increase of their size. From a practical standpoint, such a reduction would lead to an increase in transport costs; on the other side, it determines an improvement of wood grade during the storage phase due to the improved air the circulation through the chip pile. Our results also suggest that modifications of the drum elements that are involved in the cutting action (number and inclination of the bladeholders, number and cutting angle of the knives) may lead to the achievement of the desired objective. The basic reasoning that guided such choices was that increasing the dimension of the raw material inlet to the cutting knives would lead to an increase in the percentage of larger chips. These results suggest that the positive effect of the CRA–ING rotor affected two aspects: i) a shift in par-

106

ticle size distribution towards the 16–45 mm fraction; ii) just for this class, a repositioning of the wood chip size toward the higher class limit, whereas the result may differ for the remaining fractions. When a feeding system is used, reducing the number of knives increases the feed per tooth and consequently the chip length, but also results in proportionate increase in the other dimensions (Abdallah et al. 2014). Such observation was confirmed by the results given in Table 3, which suggests that removal of two knives contributed (beside the other factors) to the increase in the dimension of wood chips. The positive effect of a smaller knife angle observed in this study is consistent with the observations by Krajnc and Dolsac (2014) that obtained a better form and size structure of the single particles using an inclination of 29° rather than 34°. However, the results could not be adequately comparable since the authors reported the average grain size and a subjective estimation of form constancy, without a complete particle size distribution. Our results and the findings of Krajnc and Dolsac (2014) are in contrast with the conclusion drawn by Abdallah et al. (2014) for the chipping process on a disc chipper. Among the factors analyzed, Abdallah et al. (2014) observed that an increase in the cutting angle would lead to a higher proportion of large size classes. However, the dimensional classes ranged from 1 to > 10 mm, generally less than the size distribution examined in the present work, without any reference to recognized standards. More importantly, the data are referred to a disc chipper, which is conceptually different from the drum chipper studied. Nevertheless, such work provides new insights for future studies on how additional factors such as cutting speed, feed per tooth, and the sharpness angle could influence the size of wood chip.

5. Conclusions The production of wood chip matching the quality requirements of heating plants is a tricky process influenced by several variables related to mechanical aspects (chipper type, rotor configuration, blade wear, screen type) as well as feedstock type (species, age, comminuted branches or stems). The outcome of the study appears worthy and susceptible to prompt further in depth studies. By varying the number of bladeholders and knives, bladeholders inclination and cutting angle of the knives, it is possible to substantially modify the particle size distribution and improve the fuel grade. When compared to the standard ones, comminution accomplished by the CRA–ING rotor led to a feedstock with a higher percentage of wood Croat. j. for. eng. 36(2015)1


Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

V. Civitarese et al.

chips in the 16–45 mm fraction (+ 15%) at the expense of the classes finer than 16 mm (– 28%), thus lowering the bulk density (– 8.57%). In addition, the accurate dimensional analysis of the single chips disclosed a secondary effect of the CRA–ING rotor concerning a distribution closer to the higher class limit of the wood chips within the most representative fraction.

Goli, G., Marchal, R., Uzielli, L., 2004: Superfici e loro formazioneXylon 3: 68–73.

The dimensional increments obtained and the percentage reduction of the smallest fractions represent two valuable elements having a positive influence on both the behaviour of the wood chips during storage and the fuel handling in processing plants. Starting from the present and previous studies on the factors affecting the wood chip quality, future activities could also addresse storage, in order to check the storage quality benefit offered by different particle size distributions.

Hernandez, R., Jacques, B., 1997: Effect of the rotation speed on the size distribution of black spruce pulp chips produced by chipper-canter. Forest Products Journal 47(4): 43–49.

6. References

Krajnc, M., Dolsak, B., 2013: Computer and experimental simulation of biomass production using drum chipper. International Journal of Simulation Modelling 12(1): 39–49.

Abdallah, R., Auchet, S., Méausoone, P.J., 2011: Experimental study about the effects of disc chipper settings on the distribution of wood chip size. Biomass and Bioenergy 35(2): 843–852. Abdallah, R., Auchet, S., Méausoone, P.J., 2014: A dynamic measurement of a disc chipper cutting forces. Biomass and Bioenergy 64: 269–275.

Hartler, N., 1986: Achievement and significance of optimal chip quality. Tappi Journal 79(2): 259–264. Hartmann, H., Böhm, T., Jensen, P.D., Temmerman, M., Rabier, F., Golser, M., 2006: Methods for size classification of wood chips. Biomass and Bioenergy 30(11): 944–53.

Jirjis, R., 1995: Storage and drying of wood fuel. Biomass and Bioenergy 9(1–5): 181–190. Jirjis, R., 2005: Effects of particle size and pile height on storage and fuel quality of comminuted Salix viminalis. Biomass and Bioenergy 28(2): 193–201. Jirjis, R., Pari, L., Sissot, F., 2008: Storage of poplar wood chips in northern Italy. In: Proceedings of the World Bioenergy 2008 Congress. Jonkoping, Svezia 2008, 85–89 p.

Krajnc, M., Dolšak, B., 2014: The influence of drum chipper configuration on the quality of wood chips. Biomass and Bioenergy 64: 133–139. Kubler, H., 1982: Air convection in self-heating piles of wood chips. Tappi Journal 65(8):63–79.

Afzal, M.T., Bedane, A.H., Sokhansanj, S., Mahmood, W., 2010: Storage of comminuted and uncomminuted forest biomass and its effect on fuel quality. Bioresources 5(1): 55–69.

Kubler, H., 1987: Heat generation processes as cause of spontaneous ignition in forest products. Forest Products Abstracts 10(11): 299–322.

Barontini, M., Scarfone, A., Santangelo, E., Gallucci, F., Spinelli, R., Pari, L., 2013: The CRA ING experience on storage of poplar wood chips. In: Proceedings of the 21st European Biomass Conference and Exhibition, Copenhagen, Denmark 2013, 170–172 p.

Makeschin, F., 1999: Short Rotation Forestry in Central and northern Europe – Introduction and Conclusions. Forest Ecology and Management 121(1–2): 1–7.

Bergante, S., Facciotto, G., Minotta, G., 2010: Identification of the main site factors and management intensity affecting the establishment of Short-Rotation-Coppices (SRC) in Northern Italy through Stepwise regression analysis. Central European Journal of Biology 5(4): 522–530. Buchanan, J.G., Duchinicki, T.S., 1963: Some experiments in low-speed chipping. Pulp & Paper Magazine of Canada 5: 235–245. CEN/TS 14961:2005 Standards: Solid biofuels – fuel specifications and classes. CEN/TS 15149-1:2010 Standards: Solid biofuels-determination of particle size distribution part 1: oscillating screen method. Edelma, J.S., Stuart, W.B., 1992: Effect of disk speed on sawmill residue wood chip quality. In: Proceeding of TAPPI pulping conference, Boston, USA, 375–380 p. EN 14774-2, 2009: Solid biofuels–determination of moisture content-oven dry method – part 2 simplified method. EN 15103, 2010: Solid biofuels – determination of bulk density. Croat. j. for. eng. 36(2015)1

Mattison, J.E., Mitchell C.P., 1995: IEA bioenergy agreement task IX harvesting and supply of woody biomass for energy 1992 – 1994. Biomass and Bioenergy 9(1–5): 117–125. McLauchlan, T.A., Lapointe J.A., 1979: Production of chips by disc chippers. In: Chips quality monograph. Pulp and paper technology series, No 5, Joint Textbook Committee of the Paper Industry, CPPA/TAPPI, (Hatton J.V. ed.), Vancouver, 15–32 p. Mitchell, C.P., Kofman, P.D., Angus–Hankin, C.M., 1997: Guidelines for conducting harvesting trials in short rotation forestry. Aberdeen University, Department of Forestry (eds), Aberdeen 1–50 p. Monico, J.A., Soule, H.M., 1979: A machine to harvest and chip brush stands. Winter Meeting. American Society of Agricultural Engineers, New Orleans, USA 1979, 16 p. Nati, C., Eliasson, L., Spinelli, R., 2014: Effect of chipper type, biomass type and blade wear on productivity, fuel consumption and product quality. Croatian Journal of Forest Engineering 35(1): 1–7. Nati, C., Spinelli, R., Fabbri, P., 2010: Wood chips size distribution in relation to blade wear and screen use. Biomass and Bioenergy 34(5): 583–587.

107


V. Civitarese et al.

Study on the Effect of a New Rotor Designed for Chipping Short Rotation Woody Crops (101–108)

Pari, L., Civitarese, V., Del Giudice, A., 2010: Quality of wooden chips produced by CLAAS Jaguar equipped with experimental CRA–ING rotor. In: Proceedings of the 18th European Biomass Conference and Exhibition. From research to industry and markets, Lyon, France, 1717–1720 p. Pari, L., Civitarese, V., Gallucci, F., 2009: Development of a chipping apparatus prototype mounted on a Claas Jaguar 890. In: Proceedings of the 17th European Biomass Conference & Exibition. From research to industry and markets, Hamburg, Germany, 247–251 p. Pari, L., Fedrizzi, M., Ciriello, A., 2008: SRF poplar chips stocking methods comparison to lessen fuel depletion from production to utilization. In: Proceedings of the 16th European Biomass Conference & Exbition, Valencia, Spain, 74–78 p. Paris, P., Mareschi, L., Sabatti, M., Pisanelli, A., Ecosse, A., Nardin, F., Scarascia–Mugnozza, G., 2011: Comparing hybrid Populus clones for SRF across northern Italy after two biennial rotations: survival, growth and yield. Biomass and Bioenergy 35(4): 1524–1532. Paulrud, S., Nilsson, C., 2004: The effects of particle characteristics on emissions from burning wood fuel powder. Fuel 83(7): 813–821. Suadicane, K., Gamborg, C., 1999: Fuel quality of whole-tree chips from freshly felled and summer dried Norway spruce on a poor sandy soil and a rich loamy soil. Biomass and Bioenergy 17(3):199–208. Spinelli, R., Cavallo, E., Eliasson, L., Facello, A., 2013: Comparing the efficiency of drum and disc chippers. Silva Fennica 47(2): article id 930.

Spinelli, R., Hartsough, B.R., Magagnotti, N., 2005: Testing mobile chippers for chip size distribution. International Journal of Forest Engineering 16(2): 29–35. Spinelli, R., Nati, C., Magagnotti, N., 2008: Harvesting short rotation poplar plantations for biomass production. Croatian Journal of Forest Engineering 29(2): 129–139. Spinelli, R., Nati, C., Magagnotti, N., 2009: Using modified foragers to harvest short-rotation poplar plantations. Biomass and Bioenergy 33(5): 817–821. Spinelli, R., Magagnotti, N., Picchi, G., Lombardini, C., Nati, C., 2011: Upsized harvesting technology for coping with the new trends in short-rotation coppice. Applied Engineering in Agriculture 27(4): 551–557. Spinelli, R., Nati, C., Pari, L., Mescalchin, E., Magagnotti, N., 2012: Production and quality of biomass fuels from mechanized collection and processing of Vineyard pruning residues. Applied Energy 89(1): 374–379. Tharakan, P., Volk, T., Abrahamson, L., White, E., 2003: Energy feedstock characteristics of willow and hybrid poplar clones at harvest age. Biomass and Bioenergy 25(6): 571–580. Twaddle, A., 1997: The influence of species, chip length, and ring orientation on chip thickness. Tappi Journal 80(6): 123– 131. Wu, M.R., Schott, D.L., Lodewijks, G., 2011: Physical properties of solid biomass. Biomass and Bioenergy 35(5): 2093– 2195.

Authors’ address:

Received: April 30, 2014 Accepted: September 7, 2014

108

Researcher Vincenzo Civitarese, PhD.* e-mail: vincenzo.civitarese@entecra.it Angelo Del Giudice, PhD. e-mail: angelo.delgiudice@entecra.it Alessandro Suardi, PhD. e-mail: alessandro.suardi@entecra.it Enrico Santangelo, PhD. e-mail: enrico.santangelo@entecra.it Senion researcher Luigi Pari, PhD. e-mail: luigi.pari@entecra.it Consiglio per la Ricerca in Agricoltura e l’analisi dell’economia agraria Unità di ricerca per l‘ingegneria agraria Via della Pascolare 16, 00015 Monterotondo, Rome ITALY * Corresponding author Croat. j. for. eng. 36(2015)1


Preliminary note

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain from Wooden Biomass Srđan Vasković, Velid Halilović, Petar Gvero, Vlado Medaković, Jusuf Musić Abstract Production of solid fuels from wooden biomass is defined with appropriate energy chain of supply. Production procedure of solid fuels from wooden biomass, starting with technology for gathering wood residues and residues from logging up by the system of fuel production (system for milling, crushing, chopping, drying and pressing of wood residues), represents the energy chain of supply of solid fuel from biomass. Every single energy chain of supply and production of certain form of solid fuel from wooden biomass can be uniquely defined with three general criteria. These criteria are: energy efficiency of production, economy of production and environmental criteria. Efficiency of production is the relation of overall energy consumption per 1 kWh of heating value of produced fuel. When we talk about the economical aspect of production of solid fuels we take into account all production costs per 1 kWh of heating value of biofuel produced. Forest biomass is scattered and the need for its collection and transport require certain consumption of fossil fuel. Consumption of fossil fuel is needed to run mechanization to collect, transport and prepare biomass. Consumption of fossil fuels causes the emission of GHG. Ecological criteria for the estimation of production process of bio energy can be defined as emission of GHG per 1 kWh of heating value of produced fuel. Besides general criteria to estimate the quality of production of energy from biomass, there are specific criteria. Specific criteria regarding several characteristics are tightly related to applied technologies, potentials and barriers during the use of biomass. This paper analyzes only specific investment in selected chain of energy supply. The paper mathematically describes four characteristic cases of solid biofuel production from wooden biomass. These cases are: production of wooden chips from forest biomass with mobile chipper, production of wooden chips from wooden residues transported from sawmill to processing terminal, production of wooden briquette from mill residues transported into briquette factory, production of wooden pellet from mill residues transported into pellet factory. For overall ranking of energy chain of wooden biomass supply and selection of optimum variant, multicriteria optimization and VIKOR method is used. Keywords: forest biomass, chips, briquette, pellet, energy chain, multi-criteria optimization, VIKOR method

1. Introduction Biomass represents a solar energy stored in the photosynthesis process in the form of chemical compounds that form the structure of plants. During the time of biomass combustion, the oxygen from the atmosphere joins with the carbon in the biomass, and CO2 and water are obtained again as the combustion products. This is the main reason why the energy obtained in this way is considered as CO2 neutral. The use of biomass as a fuel provides significant opportunities for the decrease of harmful influences of GHG to the environment (WEA 2004, Karp and Shield 2008, Demirbas 2009). With the increase of crude oil price, various countries and institutions do the research trying to find the best ways of energy production from renewable sources (Inyang 2005). However, certain Croat. j. for. eng. 36(2015)1

109


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

amounts of CO2 emissions occur directly from fossil fuels used at cutting, collecting, processing and transportation of biomass. The energy production from biomass has been defined by adequate energy chain. The notion of energy chain is little known in the available literature. However, the notion of supply chain has been more defined in some other research areas, which are not of technical nature. Nowadays, in the scientific and professional literature of numerous papers, mixed opinions can be found on defining, management, basic elements, coverage and characteristics of a supply chain (Cooper et al. 1997). The energy chain concept has been defined as the trajectory of energy transformations from the fuel source to the end users (Hamamatsu et al. 2004). An energy chain is the way in which the energy is used from an adequate fuel, starting from the fuel collection technology to the system for energy or fuel production (systems for transformation of energy from one form to the other). The production of solid fuels from wood biomass is defined for sure with an adequate energy chain. In different conditions in which the fuels are produced from wood biomass, different production costs of these fuels also occur. However, it is not enough to talk only about the production cost of particular fuels from wood biomass. It is also necessary to perceive some other aspects of fuel production process such as: energy efficiency of the process, consumption of fossil fuels, CO2 emission and investment cost in energy chain. The research in the field of the structures of energy chains for production of fuels and energy from biomass is of a relatively recent date. Generally, the increased interest for such a research has occurred with global problems such as: global warming process, increase of fuel prices due to the decrease in reserves of fossil fuels, tendency to decrease the dependence on fossil fuels supply with the use of local biomass resources, environmental pollution, etc.

2. Overview of research of supply chains based on biomass In the developing countries, about 22% of the used energy is obtained from biomass, however that is a traditional way of use with a very low energy efficiency and increased emission of pollutants. Many scenarios predict a significant increase of share of the energy from biomass in the future (IEA 2010). For that reason, it is necessary to work constantly on the process of introducing the new technologies for energy production from biomass with an improved energy efficiency. The research dealing with the composition of energy chains of supply and general use of biomass as a fuel is relatively new. The optimization of supply chains by biomass is mostly performed in accordance with the transportation distances and moisture of the biomass to be transported. The description of modelling a regional supply structure of wood biomass as a fuel, depending on the transportation costs, was given by Gronalt and Rauch (2007). The model of linear biomass supply chain, which includes the transportation, storage and preparation of biomass, was discussed in Silke Van Dyken et al. (2009). The main focus of the paper is finding the linear dependence between the moisture content in biomass and energy content of biomass and economic indicators. The case of passive drying of biomass in the storage process was specifically discussed. The planning and logistics in the use of wood biomass for energy production were discussed by Frombo et al. (2009). The decision making variables in this approach are the plant capacity and collected biomass from the adequate area, while the target function is minimization of total costs in the process of wood biomass utilization. The productivity and cost of mechanized cutting and collecting of wood for energy purposes were discussed by Roser et al. (2011). The paper presents an analysis and overview of costs that occur at different production combinations of wood chips production (in plants, near the road, at the terminal), according to transportation distance. In the analysis of transportation costs of energy wood supply chains at greater distances in Finland, the ways of transportation by means of trucks for wood chips, transportation of baled cutting residues by means of a truck or train, transportation of scattered residues, etc. were analysed (Tahvanainen and Anttila 2011). Also, there are several studies about wooden residues as follow: Feasibility study for commercial use of wood residues in central Bosnia as a project for regional economic development 2006, Wood Energy Technologies, Partnership Programmes – TCDC/TCC –TCP/YUG/3201 (D), Belgrade, March 2011, etc.

3. Problem definition and approach to solution Utilization of wood biomass for the production of fuels depends on many factors. Primarily, biomass is scattered and must be collected and processed to become a fuel. For cutting, collecting and transportation of biomass

110

Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

S. Vasković et al.

to the processing plant, it is necessary to engage the following mechanization: chainsaws, tractors, forwarders, trucks, technology for milling and crushing, draying, densification, etc. It, logically, results in corresponding energy consumption. Each operation of collecting, processing and transportation of biomass requires some energy consumption at related costs. The consumed fuel in the biomass supply process is usually of fossil origin. That consequently carries the harmful CO2 emissions, which occur in collecting and processing of biomass. Every machine for cutting, chopping and transportation requires certain investment costs, as well as a biofuel production plant. For that reason, to be able to select an optimum variant of a fuel production energy chain, it is necessary to apply a multi-criteria optimization method. In this paper, VIKOR optimization method is applied. Before selecting the optimum variant of fuel production from wood biomass, it is necessary to define the optimization criteria. The following criteria have been defined:  consumed energy per 1 kWh of the lower heating value of produced biofuel, kWh/kWh (K1);  energy chain production cost per 1 kWh of the lower heating value of produced biofuel, EUR/kWh (K2);  CO2 emission in the total chain due to the fossil fuels consumption for 1 kWh of the lower heating value of produced biofuel, kg/kWh (K3);  specific investment cost per totally installed power of all machines and plants in the energy chain, EUR/kW (K4). As we know, lower calorific value of a fuel portion is defined as the amount of heat evolved when a unit of weight (or volume in the case of gaseous fuels) of the fuel is completely burnt and water vapor leaves with the combustion products without being condensed. This is the main reason why the above three criteria were defined with lower heat value of biofuels. It is the useful heat energy obtained by combustion from chemical energy stored in biomass. We can say that the first three criteria are of general character, while the fourth criterion is specific, and they are all together used to evaluate the acceptability of biomass based energy supply chains. The above mentioned criteria are different for differently defined initial conditions of a problem. The solution of the problem of selecting the optimum variant of fuel production from wood biomass comes down to the development of:  a mathematical model for the calculation of optimization criteria;  a mathematical method for the selection of optimum variant of fuel production from wood biomass with the application of VIKOR method. The significance of energy chain analysis from the aspect of invested energy is very important. In the literature, we can find the so-called EROEI factor (Energy Returned on Energy Invested), which presents the quotient of utilizable energy from a certain fuel (or from a way of energy production) and the energy consumed to convert the fuel or energy to a useful form (Hall 2011). The diagram (Fig. 1) shows that among the discussed fuels and ways of energy production, there is no biomass. The answer to the question why the EROEI factor for biomass is not defined in the diagram is very simple. The energy production from biomass depends on many variable parameters and that is why a clear value of this parameter cannot be defined. However, EROEI factor interval related to the production of fuels and energy from biomass can be obtained by mathematical modelling of biomass supply chains and numerical calculations for different initial conditions. Such an approach to the analysis of energy production from biomass has not been sufficiently researched and requires further research. To be able to evaluate the total energy consumption in an energy chain of fuel production at all, it is necessary to convert all consumed energies into the primary energy form (heating value). In this way, all the consumed energy in the energy chain of fuel production Fig. 1 EROEI – USA, Ratio of Energy Returned on Energy Invested from biomass can be summed. It practically means that for different kinds of renewable energy (Murphy and Hall 2010) Croat. j. for. eng. 36(2015)1

111


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

all the energy conversions that occur in machines and plants in an energy chain are defined by corresponding efficiency factors. One of the main indicators of energy contents within a biomass fuel is the heating value. The heating value of biomass depends on its chemical composition, as well as on the moisture content.

4. Specific characteristics of wood biomass as a fuel The most important characteristic of biomass related to combustion and its other thermochemical processes is the moisture content, with the increase of which the heating value of biomass decreases. The value of the lower heating value of wet wood can be calculated by the following equation (Hartmann et al. 2000): ehvw =

ehv0 × (100 − w) − (2.44 × w) (1) 100

Where: ehvw lower heating value of wood in relation to moisture content, MJ/kg; ehv0 heating value of dry wood, MJ/kg; 2.44 energy needed for water evaporation at 25oC, MJ/kg; w moisture content in total mass expressed in percentage. The volumetric mass or density of wood r0 is defined as the relation between the dry mass of wood (kg) and the volume it occupies. The value varies widely, depending on the type of wood but is mostly in the range between 320 and 720 kg/m3. The heating value per a volume unit can be calculated by taking into account the lower heating value ehvw and density of wood: ehvvw = ehvw × rw

(2)

For the moisture content per dry wood basis higher than 30%, the density of wet wood is:  u   1 + 100  10 4 =r × rw = r0 ×  0  av  (100 − w ) × (100 + av ) 1+   100 

(3)

For the moisture content per dry wood basis lower than 30%, the density of wet wood is:  u   1 + 100  3000   rw = r0 × = r0 × 3000 − 30 w + av × w   av u  × 1+   100 100   

(4)

Where: ehvvw heating value per volume unit, MJ/m3; u moisture content per dry basis u = r0

100 × w , %; 100 − w

(5)

density of dry wood, kg/m3;

rw density of wood with moisture content, kg/m3; av

percentage of swelling, % (Hellrigl 2006).

The volume occupied by wood fuels depends on the shape, size and organization of particular pieces of wood. Fulfilment factor of the volume also depends on that. It should be mentioned that a significant factor in biofuel supply and at the selection of their transportation is the shape in which biofuels are transported. For that reason, the SVF

112

Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

S. Vasković et al.

Table 1 Criteria K1, K2, K3, K4 for all four variants of fuel production from wood biomass OPTIMIZATION CRITERIA: K1, K2, K3, K4 (kWh/kWh, EUR/kWh, kg/kWh, EUR/KW) j=1

j=2

j=3

j=4

CHAIN CH1

CHAIN CH2

CHAIN CH3

CHAIN CH4

mobile chipper

plant for production of wood chips

plant for briquetting

plant for pelleting

i=1

E11, C11, G11, P11 chainsaw

E12, C12, G12, P12 chainsaw

E13, C13, G13, P13 chainsaw

E14, C14, G14, P14 chainsaw

i=2

E21, C21, G21, P21 tractor provided with winch for forest operations

E22, C22, G22, P22 tractor provided with winch for forest operations

E23, C23, G23, P23 tractor provided with winch for forest operations

E24, C24, G24, P24 tractor provided with winch for forest operations

i=3

E31, C31, G31, P31 forwarder

E32, C32, G32, P32 lifter for loading of timber

E33, C33, G33, P33 lifter for loading of timber

E34, C34, G34, P34 lifter for loading of timber

i=4

E41, C41, G41, P41 medium power chipper

E42, C42, G42, P42 truck for transportation of timber

E43, C43, G43, P43 truck for transportation of timber

E44, C44, G44, P44 truck for transportation of timber

i=5

E51, C51, G51, P51 truck for transportation of wood chips to the end user

E52, C52, G52, P52 crane for unloading of timber

E53, C53, G53, P53 crane for unloading of timber

E54, C54, G54, P54 crane for unloading of timber

i=6

0

E62, C62, G62, P62 sawmill

E63, C63, G63, P63 sawmill

E64, C64, G64, P64 sawmill

i=7

0

E72, C72, G72, P72 crane for loading of wood residues on truck

E73, C73, G73, P73 crane for loading of wood residues on truck

E74, C74, G74, P74 crane for loading of wood residues on truck

i=8

0

E82, C82, G82, P82 truck for transportation of wood residues to the chips production plant

E83, C83, G83, P83 truck for transportation of wood residues to the briquettes production plant

E84, C84, G84, P84 truck for transportation of wood residues to the pellets production plant

i=9

0

E92, C92, G92, P92 crane for unloading of wood residues

E93, C93, G93, P93 crane for unloading of wood residues

E94, C94, G94, P94 crane for unloading of wood residues

i = 10

0

E102, C102, G102, P102 chips production plant

E103, C103, G103, P103 plant for rough and fine chipping

E104, C104, G104, P104 plant for rough and fine chipping

i = 11

0

E112, C112, G112, P112 loading of wood chips

E113, C113, G113, P113 drying plant

E114, C114, G114, P114 drying plant

i = 12

0

E122, C122, G122, P122 truck for transportation of wood chips to the end user

E123, C123, G123, P123 plant for briquetting

E124, C124, G124, P124 plant for pelleting

i = 13

0

0

E133, C133, G133, P133 loading of briquettes on trucks

E134, C134, G134, P134 loading of pellet on trucks

i = 14

0

0

n

f11 = ∑ Ei1 i=1 n

f21 = ∑ Ci1 i=1

F

n

f31 = ∑ Gi1 i=1 n

E143, C143, G143, P143 truck for E144, C144, G144, P144 truck for transportation of briquettes to the end transportation of pellet to the end user user

n

f12 = ∑ Ei2 i=1 n

f22 = ∑ Ci2 i=1 n

f32 = ∑ Gi2 i=1 n

n

f13 = ∑ Ei3 i=1 n

f23 = ∑ Ci3 i=1 n

f33 = ∑ Gi3 i=1 n

n

f14 = ∑ Ei4 i=1 n

f24 = ∑ Ci4 i=1 n

f34 = ∑ Gi4 i=1 n

f41 = ∑ Pi1

f42 = ∑ Pi2

f43 = ∑ Pi3

f44 = ∑ Pi4

n=14

n=14

n=14

n=14

i=1

Croat. j. for. eng. 36(2015)1

i=1

i=1

i=1

113


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

– solid volume content has been defined. The SVF represents the relation between the solid volume the wood would occupy without air holes at its current density and the total volume which it occupies in the form of various types of complex wood assortments and fuels. The factor is lower than one (Pottie M.A. and Guimier D.Y., 1986). In order to calculate the amount of energy (heating value) of wood fuel in a volume in which it is, for example, transported, we will use the previously mentioned equations (1, 2, 3, 4, 5), but also the values of density r0 and the SVF, so as to get the following relations for the moisture content higher than 30% and lower than 30%;

H w≥30 =

H w<30 =

ehv0 (100 − w) − (2.44 w) 10 4 × r0 × × V × SVF 100 (100 − w ) × (100 + av ) 3.6

ehv0 (100 − w) − (2.44 w) 10 4 × r0 × × V × SVF 100 (100 − w ) × (100 + av ) 3.6

(6)

(7)

Where: V volume occupied by a material, for example the volume of a truck during transportation;

H w≥30 , H w<30 ,  kWh  total energy value of a fuel in a given shape and the volume it occupies as a function of moisture percentage w for the moisture content higher or lower than 30%.

5. Mathematical model for the calculation of criteria for energy chains optimization The energy chains modelling should be based on modularity. It practically means that it is necessary to do a mathematical modelling of every energy chain element as an independent entity, which will for itself present a mathematical model as an elementary part of the energy chain. In this paper, the mathematical modelling is performed for: machines for cutting and collecting of biomass, means of transportation, plants for chopping and preparation of biomass, plants for drying, plants for pressing and pelleting of biomass. The approach to the modelling of all the elements is based on the analysis from the aspect of consumed energy in every element of the chain. The calculation of other criteria comes down to the calculation of production costs, CO2 emission and investment cost per the installed power of all energy consumers in the chain. Table 1 gives an overview of all the elements of an energy chain for biofuel production defined by the criteria (K1, K2, K3, K4). The calculation of all the criteria functions of the matrix F is obtained by summation of all the elements in a table column for each of the chain variants (CH1, CH2, CH3, CH4). Where: Eij ratio of invested energy per an obtained kWh of heating value of the processed wood for the ith element of chain and jth energy chain of production, kWh/kWh; Cij production cost for obtaining a kWh of heating value of processed wood for the ith element of chain and jth energy chain of production, kWh/kWh, EUR/kWh; Gij CO2 emission per a produced kWh of heating value of processed wood for the ith element of chain and jth energy chain of production, kg/kWh; Pij investment cost per installed power for the ith element of chain and jth energy chain of production, EUR/KW. Mathematical formulations of the functions Eij, Cij, Gij, Pij are given in the Enclosure 7.

6. Mathematical model for the selection of optimum variant of energy chain for the production of solid fuels from wood biomass The VIKOR method (Multi-criteria compromise ranking) has been developed for the determination of a multicriteria optimal solution. The VIKOR method has been developed on such a methodological basis that a decision maker is suggested the alternative (or solution), which (Opricović 1998) presents a compromise between:

114

Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

S. Vasković et al.

 wishes and opportunities;  different interests of the decision-making participants. The VIKOR method has been developed for a multi-criteria optimization of complex systems. It is focused on ranking and selection of the best solution from the given set of alternatives, with conflicting criteria. The VIKOR method requires the values of all the criteria functions for all the alternatives in the form of a matrix to be familiar. Because of that, at the beginning of the optimization process, we set a general form of the problem (evaluation matrix) for our case. The matrix of criterion functions for all four variants of the production of solid fuels from wood biomass is of 4 x 4 dimensions, (4 alternatives of biofuel production from wood biomass and 4 criteria), obtained from Table 1. a1 f1  f11  f f F = 2  21 f3  f31  f4  f41

a2 f12 f22 f32 f42

a3 f13 f23 f33 f43

a4 f14   f24  f34   f44 

(8)

Where: {a1, a2, a3, a4} is a finite set of possible alternatives to which the four energy chains of production correspond (CH1, CH2, CH3, CH4), m = 4; {f1, f2, f3, f4} is a finite set of criterion functions for four adopted criteria on the basis of which the chains of fuel production from wood biomass are analysed (K1, K2, K3, K4), n = 4; {f11(.),f12(.),...,fij(.),...,f44} is the set of all the criterion functions values. An ideal solution is determined on the basis of the criterion function values from the equation:

fi = ext fij , i = 1, 2, ..., n. j

(9)

The operator ext denotes a maximum if the function fi describes a benefit or profit, and a minimum if fi describes damages or costs. This is the best way to define an ideal solution. The criterion functions within the matrix F are commonly not expressed in the same units of measurement (i.e. the belonging criterion space is heterogeneous). For that reason, in order to perform the multi-criteria optimization, it is necessary to convert all the criterion functions to dimensionless functions whose values will be in the interval [0, 1]. Firstly, the best values of criterion functions are determined. In our case, these are the minimum values of all the criterion functions (minimization of: invested energy per the obtained one, production cost, CO2 emission and investment cost per an installed kilowatt in the production chain):

min f1 (f1j ) = f1* , min f2 (f2j ) = f2* , min f3 (f3j ) = f3* , min f4 (f4j ) = f4*

(10)

In the same way, the worst values of the criterion functions can be determined, which are obtained by maximization of the criterion functions, i.e.

max f1 (f1j ) = f1− , max f2 (f2j ) = f2− , max f3 (f3j ) = f3− , max f4 (f4j ) = f4−

(11)

Then all the elements of the matrix f are converted in the value domain [0, 1]. Thit is achieved by the following equation: nij =

f * − fij f i* − f i −

, and a matrix is formed, in the form D = ( −1) ⋅ nij  = dij  , for i = 1,... n and j = 1,…m

(12)

Considering that there is a negative difference fi* − fi− in the expression for nij it is necessary to multiply all the elements of nij with – 1 to satisfy the condition that the values of elements of the matrix D are in the interval [0, 1]. The negative difference occurs due to the nature of the problem in which a lower value of criterion functions is obtained by maximization, while a higher value is obtained by minimization. Croat. j. for. eng. 36(2015)1

115


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

The criterion weights for our problem, for the named four criteria are mutually equal. The reason for that is very simple, because we strive for the minimal: energy consumption, production cost, CO2 emissions and investment cost per totally installed power in the energy chain. Consequently, the criterion weights are: w1 = w2 = w3 = w4 = w j =

1 4

(13)

After that, the values of the elements of matrices Sj i Rj are calculated. Considering the equality of the criterion weights, they are obtained as:

1 4  1 4 1 4 1 4 1 4 × ∑ dij =  × ∑ di1 × ∑ di2 × ∑ di3 × ∑ di4  , 4 i=1 4 i=1 4 i=1 4 i=1  i=1  4 i=1  d max di2 max di3 max di4 max  = w j × max  dij  =  i1  i 4 4 4   4 4

Sj=1...4 = w j × ∑ dij = R j=1...4

(14)

In this way, the problem is reduced from a multi-criteria space to a two-criterion problem. The values of minimal and maximal element are determined from the matrices Sj and Rj.

S* = min Sj , S − = max Sj , R* = min Rj , R − = max Rj j

j

j

(15)

j

The decision strategy weight will be taken as u = 0,5 . This is valid for the criterion number m ≤ 4 (Opricović 1998). On the basis of that, it is possible to calculate the value of the matrix Qj pursuant to the equation:

( S − S ) + ( 1 − u) R − R Q = u⋅ j

j

*

j

S − − S*

*

R − − R*

(16)

A certain value of Qj corresponds to every chain, as shown in the following matrix: CH1 CH 2 Qj = Q1

Q2

CH 3 Q3

CH 4 Q4 

(17)

The optimum variant of production is defined by the minimal value Q * = min Qj . The ranking of alternatives j is formed from the lowest value of Qj to the highest value of Qj, that is from the best to the worst variant. In our case, the alternatives are the mentioned chains of solid fuel production from wood biomass.

7. Enclosure of mathematical functions Eij, Cij, Gij, Pij from the energy chain elements For the production of biofuel from wood biomass, it is necessary to engage: different types of mechanization, plants for converting biomass to useful fuel, human and other resources. Due to the fact that the energy chains for biofuel production are analysed from the energy aspect in this paper, in the text that follows mathematical descriptions will be given for particular elements of an energy chain pursuant to the previously adopted concept for the calculation of functions (Eij, Cij, Gij, Pij).

7.1 Biomass collection machines in a supply chain Biomass collection machines are the first element in a chin from which the entire biomass supply process starts. Different operations in wood biomass collection require different machines, whose selection for use practically depends on the application conditions. In the structure of the analysed energy chains discussed in this paper, the following machines are used: chainsaw, tractor, truck, hydraulic crane, mobile chipper and forwarder. For all the production machines whose fuel consumption is expressed in litres per hour (l/h), and the labour productivity in the volume unit per hour (m3/h), the following relations apply:

116

Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

n1

Eij =

rFij × Fcijq × t ijq 1000

q=1

S. Vasković et al.

× Hvijq

(18)

(19)

 eFij  n1  rFij × Hvijq × 6  × ∑ Fc ijq × t ijq 10  q=1  Gij = n 1 ehv (100 − wijq ) − (2.44 wijq ) 10 4 0ijq × × × × × r Pr t SVF ( ) ∑ 0ijq ijq ijq ijq 100 100 − wijq × 100 + avijq q=1

(20)

n1

ehv0ijq (100 − wijq ) − (2.44 wijq ) 100

q=1

× r0ijq ×

10 4

(100 − w ) × (100 + a ) ijq

n1

Cij =

∑ Fc q=1

n1

ehv0ijq (100 − wijq ) − (2.44 wijq ) 100

q=1

× r0ijq ×

ijq

vijq

× Prijq × tijq × ( SVF )ijq

× t ijq × c ijq 10 4

(100 − w ) × (100 + a ) ijq

(

vijq

) (

× Prijq × tijq × ( SVF )ijq

)

n1

∑I q=1

Mijq

Pij = n 1 (21) ∑ PMijq q=1

Where: q = 1...n1 the number of machines included in the work; Fcijq

specific fuel consumption of the observed working machine, l/h;

Prijq

productivity of the working machine, m3/h;

tijq

working time of machine, h;

Hvijq lower heating value of fuel (gasoline or oil, depending on the fuel type the machine uses), MJ/kg; wijq ≥ 30% wood moisture, %; r0ijq

wood density, kg/m3;

avijq

percentage of wood swelling, %;

(SVF ) c ijq rFij

ijq

fulfilment factor of volume (0,...1); price of a litre of fuel (gasoline or oil); density of fuel at atmospheric conditions, kg/m3;

eFij coefficient of CO2 emission for different fuels in a kilogram of CO2 per a gigajoule of the fuel heating value, kg CO2/GJ; I Mijq

cost price of a new working machine, €;

PMijq

maximum power of working machine in kilowatts, at which j = 1, 2...4.

It must be mentioned that the above equations are valid only for the working machines whose productivity is expressed in working hours. Table 1 is related to the table elements which are in the cells marked with grey colour. Thit is, for the first energy chain: Eij, Cij, Gij, Pij (i = 1...4, j = 1). The second energy chain is defined with Eij, Cij, Gij, Pij (i = 1...3, 5, 7, 9, 11, j = 2). The third energy chain is the one with: Eij, Cij, Gij, Pij (i = 1...3, 5, 7, 9, 13, j = 3). The fourth energy chain: Eij, Cij, Gij, Pij (i = 1...3, 5, 7, 9, 13, j = 4). Also, for the operation of a hydraulic crane for loading into trucks, a minimal averCroat. j. for. eng. 36(2015)1

117


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

age fuel consumption of a truck will be used (work in idle state). A truck, as an element of biomass supply chain used for transportation of either wood chips or timber, will differ slightly in terms of calculating the values Eij, Cij, Gij, while for the calculation of Pij value a previously defined equation will be used completely. The reason for that lies in the calculation of the truck fuel consumption per passed kilometres for an average defined load. n1

rFij × Ftcijq × lij

× Hvijq 1000 Eij = n 1 ehv (100 − wijq ) − (2.44 wijq ) 0ijq 1 000 × M × ∑ tijq 100 q=1 q=1

n1

Cij =

∑ Ftc q=1

ijq

(22)

× lij × c ijq

(23)

 eFij  n1  rFij × Hvijq × 6  × ∑ Ftcijq × lij × c ijq 10  q=1 Gij = n  1 ehv (100 − wijq ) − (2.44 wijq ) 0ijq × M tijq × 1 000 ∑ 100 q=1

(24)

n1

ehv0ijq (100 − wijq ) − (2.44 wijq ) 100

q=1

× M tijq × 1 000

Where: Ftcijq specific fuel consumption of trucks, l/km; lij

transportation distance, km;

M tijq maximum truck load, t. It must be emphasized that the load of a truck for wood chips is different from the load of a truck for timber transportation. If the analysis of a truck, as an element for biomass transportation in an energy chain, is observed through the maximum volume it can transport, then the equations for the calculation of the heating value which they transport are given in the expressions (6, 7). The fuel consumption of machines, which take part in the wood biomass supply chain, is mostly expressed in litres per hour. Also, the productivity of work of particular machines is given in the volume of biomass processed, attracted, collected or loaded by the machine in a time interval. To obtain some proper units of fuel consumption and productivity of different machines for wood biomass collecting, it is necessary to perform different measurements and explorations in the exploitation conditions (Krajnc 2011).

7.2 Primary mechanical wood processing Mechanical wood processing implies the type of processing at which, in the first place, the shape and dimensions of wood are changed by using mechanical means (saws, knives, etc.). The residues generated in sawmills present a significant amount of wood biomass for the production of solid biofuels. Besides the main product at sawmills such as planks, lumber, different forms of semi products, wood residues generated from processing is less important. The energy in primary wood processing is collectively consumed per the volume unit of a final product. Thus, the mathematical functions (Eij, Cij, Gij, Pij) for a sawmill are as follows:

Eij = r ×

1 hcel

 n1  ×  ∑ Fpijq × tijq × Ecijq   q=1 

 1  n1 ehv0ijq (100 − wijq ) − (2.44 wijq ) 10 4 × ∑ × r0ijq × × Fpijq × tijq × ( SVF )ijq   3.6  q=1 100 100 − wijq × 100 + avijq  

(

118

) (

(25)

)

Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123) n1

Cij = r ×

∑ Fp

ijq

q=1

× tijq × Ecijq × Ceijq

 1  n1 ehv0ijq (100 − wijq ) − (2.44 wijq ) 10 4 × ∑ × r0ijq × × Fpijq × tijq × ( SVF )ijq   3.6  q=1 100 100 − wijq × 100 + avijq  

(

n1

Gij = r ×

S. Vasković et al.

eFij ×

∑ Fp q=1

ijq

× tijq × Ecijq hcel

) (

×

3.6 10 3

) (

(26)

(27)

)

 1 n1  ehv0ijq (100 − wijq ) − (2.44 wijq ) 10 4 ×∑ × r0ijq × × Fpijq × tijq × ( SVF )ijq   3.6 q=1  100 100 − wijq × 100 + avijq  

(

)

n1

Pij =

∑I q=1

Mijq

n1

∑ PMijq

(28)

q=1

If we look at Table 1, the above defined functions correspond to the table elements marked with Eij, Cij, Gij, Pij (i = 6, j = 2, 3, 4) in the cells. Where: Q = 1...n1 number of sawmills; productivity (sawmill capacity), m3/h; Fpijq 3 Ecijq specific consumption of electricity per a processed cubic metre kWh/m3 (20–30 kWh/m soft and hard wood) (Danon et al. 2003); tijq working time of machine, h;

hcel factor of efficiency of electricity production from a thermal power station (coal as a fuel, assumption); R factor of wood residues in primary processing, in the interval from 0.25 to 0.35 (soft and hard wood without bark) (Danon et al. 2003); wijq ≥ 30% wood moisture, %; r0ijq

wood density, kg/m3;

avijq

percentage of wood swelling, %;

(SVF ) Ceijq

ijq

= 1 fulfilment factor of volume (timber); price of a kWh of electricity;

eFij coefficient of CO2 emission for coal in kilograms of CO2 per a gigajoule of the fuel heating value, kg CO2/GJ; I Mijq

cost price of a machine for wood cutting and primary processing, €;

PMijq

maximum power of cutting machine in kilowatts kW.

It must be emphasized that it has been assumed that the sawmill consumes the electricity produced in a thermal power station. The factor hcel = 0,36 takes into account all the energy losses from the thermal power station to the motor, which drives the system for wood cutting. The factor of loss includes the losses in boiler, turbine, generator and electric supply network (Honorio et al. 2003). It must be emphasized that all energy losses are reduced to the primary form (heating value). In such a way, the opportunity of a simple summation of heating values equivalent to certain forms of energy consumption is obtained, regardless of whether thermal energy or electricity is in question. The equation 28 has been previously defined. In this case, the power of motor PMijq which drives the cutting system is taken for the sawmill, while the price of a plant for primary processing is taken as the cost price I Mijq . Croat. j. for. eng. 36(2015)1

119


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

7.3 Plant for production of wood chips, drying, briquetting and pelleting For the wood chips, briquettes or pellets production process, it is first necessary to chop the initial wood residues to a certain granulation, and then to dry them. If wood chips are produced, then the production line ends at machines for rough chopping of wood to certain granulation. When producing wood briquettes or pellets, after the rough chopping, the obtained wood chips are dried in rotary dryers, and then additionally fine-chopped, to be briquetted or pelleted later. The mathematical functions (Eij, Cij, Gij, Pij) by which the production of wood chips, briquettes an pellets within a plant have been described, are related to electricity consumption due to the mechanical work of wood residue chopping. In this case, the following relations apply:

Eij =

 1  n1  ∑ Pcijq × ht × tijq  hcel  q=1 

 1  n1 ehv0ijq (100 − wijq ) − (2.44 wijq ) × Fpcijq × tijq × 1000   ∑  3.6  q=1 100  n1

Cij =

∑ Pc q=1

ijq

(29)

× ht × tijq × Ceijq

 1  n1 ehv0ijq (100 − wijq ) − (2.44 wijq ) ×∑ × Fpcijq × tijq × 1000   3.6  q=1 100 

(30)

eFij × 3.6

 n1  ×  ∑ Pcijq × ht × tijq  hcel × 10  q=1  Gij = n1  ehv (100 w ) (2.44 w ) − −  1 0ijq ijq ijq × ∑ × Fpcijq × tijq × 1000   3.6 q=1  100  3

(31)

n1

Pij =

∑I q=1

Mijq

n1

∑P q=1

(32)

Mijq

For the elements of Table 1, which are in the cells denoted with Eij, Cij, Gij, Pij to which an ordered set of counters corresponds (i = 10, j = 2, 3, 4), (i = 12, j = 3, 4), the above mathematical formulations apply. It should be emphasized again that the electricity for driving a plant has been produced from a thermal power station. Of course, this does not have to be the case. If, it is assumed, for example, that the drive of a plant has used the electricity obtained from a hydroelectric power station, then the CO2 emission factor is equal to zero for the plant. In the equations (29, 30, 31, 32), the following values are introduced: q = 1...n1 number of sawmills; Pcijq electrical power of the plant, kW; ht simultaneity factor of the operation of all electric motors in the plant (0.7–0.95), which depends on whether the plant has an installed electric power compensation system or not; tijq working time of machine in hours, h; Fpcijq output productivity of the plant, t/h; Ceijq price of 1 kWh of electricity; hcel = 0,36 takes into account all the energy losses from the thermal power station to the electricity user in a factory; eFij coefficient of CO2 emission for coal in kilograms of CO2 per a gigajoule of the fuel heating value, kg CO2/GJ. In the case of wood chip production plants, and pelleting and briquetting plants, the total installed electric power PMijq in the plant is taken into account, while the price of the plant installation is taken as the cost price I Mijq .

120

Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

S. Vasković et al.

The output moisture value of wet wood chips can vary significantly depending on the input moisture of wood to be chopped, and is usually in the interval from 30 to 50%. Pellets and briquettes have a prescribed moisture value, between 9 and 12%. It can be concluded that the difference in production plants of wood chips, pellets ad briquettes is seen only in the installed power of a plant, productivity, and in the electric power compensation factor. The mathematical functions (Eij, Cij, Gij, Pij), which describe the drying of chopped material before briquetting or pelleting, are as follows:  n1  w − wijq   1  ∑  1ijq  × Mijq × tijq × L e   3.6 × hb × hd  q=1  100 − wijq    Eij = n1 ehv   (100 − wijq ) − (2.44 wijq ) 100 − w1ijq 1 0ijq ∑ × Mijq × tijq ×    3.6 q=1 100  100 − wijq 

   

 1  n1  w1ijq − wijq   ∑  × Mijq × tijq × L e  × Chijq  3.6 × hd  q=1  100 − wijq    Cij = n1 ehv   100 − w1ijq (100 − wijq ) − (2.44 wijq ) 1 0ijq ∑ × Mijq × tijq ×   3.6  q=1 100  100 − wijq 

   

 n1  w1ijq − wijq    ∑  Mijq × tijq × L e   hb × hd × 10  q=1  100 − wijq    Gij = n1 ehv   − − − w1ijq w w (100 ) (2.44 ) 100 1 0ijq ijq ijq ∑ × Mijq × tijq ×   3.6  q=1 100  100 − wijq 

(33)

(34)

   

(35)

eFij

3

n1

Pij =

∑I q=1 n1

Mijq

∑ PMijq

(36)

q=1

For the elements in Table 1, which are in the cells denoted with Eij, Cij, Gij, Pij to which an ordered set of counters corresponds (i = 11, j = 3, 4), (i = 12, j = 3, 4), the above mathematical formulations for the dryer apply. In pelleting and briquetting plants, multi pass rotary dryers are used. Due to the complexity of the mathematical model of the rotary dryer for sawdust drying, this paper presents a simplified approach to the calculation of needed thermal energy for chopped wood residue drying. It is assumed that for the evaporation of every kilogram of water from a wet material, it is necessary to invest the amount of heat equal to the latent heat of evaporation increased by the coefficient of losses in the dryer. Also, the reduction of heat energy consumed for drying to primary energy has been done via the coefficients of losses in the boiler which supplies the dryer. The following parameters are used in the equations (33, 34, 35, 36): Q = 1...n1 is the number of dryers; Mijq input capacity of raw material, m3/h; tijq

dryer operation time, h; w1ijq moisture of material at the entrance of the dryer (0…1); w1ijq moisture of material at the exit of the dryer (0…1); Le = 2.27 latent heat of evaporation for water (Perrot, 1998), MJ/kg;

hb ≈ 0.9 boiler efficiency (Honorio et al. 2003); hd rotary dryer efficiency, usually within the limits from 0.4 to 0.6 (Devki Energy Consultancy, 2006); Chijq price of 1 kWh of thermal energy, €/kWh; eFij coefficient of CO2 emission for different fuels in kilograms of CO2 per a gigajoule of the fuel heating value, kg CO2/GJ. Croat. j. for. eng. 36(2015)1

121


S. Vasković et al.

Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123)

If wood biomass is used in the boiler for producing drying heat, then the CO2 emission is equal to zero. In this paper, the data from a real pellet production plant »Enterprise for making of wood packaging and production of eco-briquettes – pellets EU PAL factory Pale« has been used for defining particular mathematical functions and logistic concept.

8. Conclusion and further research directions The significance of energy production from biomass has been increasingly expressed in recent times. Basically, the most significant part in the process of energy production from biomass is the supply chain. If we succeed in performing the minimization of production costs, significant savings occur, especially in terms of energy. Due to the fact that there are various opportunities for the composition of energy chains of supply with wood biomass solid fuels, it is necessary to try to make a unique mathematical approach to this problem. With the mathematical model, it is possible to unify different types and a great number of parameters. This is exactly the main issue when dealing with wood biomass supply chain. This paper contains a synthesis of a few approaches to solving the problem of wood biomass solid fuels supply, such as:  approach to logical composition of energy chains based on wood biomass,  approach of multi-criteria optimization for the selection of optimum variant of an energy chain,  approach of mathematical modelling of energy chain elements based on wood biomass,  approach of reduction of all the types of consumed energies to the primary value of energy (heating value). The above mentioned approaches 2 and 4 are especially important. For the selection of the optimum variant of an energy chain of supply, VIKOR method has been selected, as it was completely adjusted to the given problem. For the adopted criteria (K1, K2, K3, K4), an equal degree of importance has been taken at the selection of the optimum variant of an energy chain for biomass fuel supply. The approach of reduction to primary energy form has enabled a mathematical estimation of the total energy consumed for the production of particular types of wood fuels. With this approach, the consumed thermal energy and electricity could be summed. For model testing and specific numerical calculations for the selection of the optimum variant of an energy chain of supply, it is necessary to provide good input data, and based on them to verify the model and check its practical applicability. At the same time, it presents a new research direction. Also, efforts will be made to develop a significantly greater number of mathematical functions of energy chain elements, which have not been described in this paper. The developed mathematical apparatus and model have a very practical application in the selection of the optimal variant of solid biofuels production from wood biomass for the defined conditions. As this paper contains many equations and mathematical descriptions, we did not show the numerical calculation and verification model. However, the next paper in this journal will present all results and discussion with more interesting conclusions.

9. References Cooper, M.C., Lambert, D.M., Pagh, J.D., 1997: Supply Chain Management: More Than a New Name for Logistics. The International Journal of Logistics Management 8(1): 1–14. Demirbas M.F., Balat M., Balat H., 2009: Potential contribution of biomass to the sustainable energy development. Energy Conversion and Management 50(7): 1746–1760. Danon, G., Bajić, V., Isajev, V., Bajić, S., Orešćanin, S., Rončević, S., 2003: Residues from wood biomass pocessing and posibilites for energy crops and forest. Chapter two from study »Energy potential and charateristics biomass residues and technologies for application in Serbia«, Faculty of forestry, Belgrade, Public company, Srbijašume, Belgrade, Institute for poplars, Novi Sad, 25–26 p. Devki Energy Consultancy Pvt. Ltd., 2006: Best practice manual »Drayers«. Ivory Terrace, R.C. Dutt Road, Vadodara – 390007, India, 405 p.

122

Frombo, F., Minicardi, R., Robba, M., Rosso, F., Scaile, R., 2009: Planing woody biomass Logistic for energy production: A strategic decision model. Biomass and Bioenergy 33(3): 372–383. Gronalt, M., Rauch, P., 2007: Designing a regional forest fuel supply network. Biomass and Bioenergy 31(6): 393–402. Hamamatsu, T., Saikawa, M., Hashimoto, K., 2004: »Energy Chain«, A New Concept in Evaluating Future Energy Conservation and Greenhouse Abatement Alternatives and Effectiveness. Proceedings 19th World Energy Congress, Sydney. Hartmann, H., Böhm, T.E., Maier L., 2000: Naturbalassene biogene Festbrennstoffe-umweltrelwante Eigenschaften und Einflussmoglichkeiteiten. Umwelt und Entwicukling. Bayerisches Statsministerium für Landsentiwicukling und Umweltfragen, 154 p. Hall, C., 2011: »Introduction to Special Issue on New Studies in EROI (Energy Return on Investment)«. Sustainability 3(10): 1773–1777. Croat. j. for. eng. 36(2015)1


Multi-Criteria Optimization Concept for the Selection of Optimal Solid Fuels Supply Chain ... (109–123) Hellrigl, B., 2006: Elementi di xiloenergetica (Elements of wood energy). Associazione Italiana Energie Rinnovabili. Legnaro (PD), AIEL, 320 p. Honorio, L., Bartaire, J., Bauerschmidt, R., Ohman, T., Tihany, Z., Zeinhofer, H., Scowcroft, J., Vasco de Janerio, Kruger, H., Meier, H., Offermann, D., Lnagnickel, U., 2003: Efficiency in electricity generation, Report drafted by Eurelectic »Preservation of resources«, working groups »Upstream«, Subgroup in collaboration with VGB. Inyang, H.I. (Ed), 2005: Bridging the gaps for global sustainable development. International Conference on Energy, Environment and Disasters, Charlotte, NC. (IEA, 2010). Karp, A., Shield, I., 2008: Bioenergy from plants and the sustainable yield challenge. New Phytoloist 179 (1): 15–32. Krajnc, N., 2011: Wood Energy Technologies, Partnership Programmes – TCDC/TCC –TCP/YUG/3201 (D), Belgrade. Murphy, D.J., Hall, C.A.S., 2010: Year in review EROI or energy return on (energy) invested. Annals of the New York Academy of Sciences 1185: 102–118. Opričović, S., 1998: Višekriterijumska optimizacija sistema u građevinarstvu (Multi-criteria system optimization in civil

S. Vasković et al.

engineering). Faculty of Civil Engineering, ISBN 86-80049-824, Belgrade. Pottie, M.A., Guimier, D.Y., 1986: Harvesting and transport of logging residuals and residues. Forest Engineering Research Institute of Canada, IEA Cooperative Project n. CPC6. FERIC special report, SR33. Perrot, P., 1998: A to Z of Thermodynamics. Oxford University Press. ISBN 0-19-856552-6. Roser, D., Sikanen, L., Asikanen, A., Parikka, H., Vaatainen K., 2011: Productivity and cost of mechanized and energy wood harvesting in North Schotland. Biomass and Bioenergy 35(11): 4570–4580. Van Dyken, Silke., Bakken, B.H., Skjelberd, H.I., 2009: Linear mixed-integer models for biomass supply chains with transport, storage and processing. Energy 35(3):1338–1350. Tahvanainen, T., Anttila P., 2011: Supply chain cost analysis of long-distance transportation of energy wood in Finland. Biomass and Bioenergy 35(8): 3360–3375. WEA (World Energy Assessment) 2004: Energy and the challenge of sustainability (2004 updated version). UNDP, UNDESA and the World Energy Council United Nations Development Programme, New York.

Authors’ address:

Received: June 12, 2014 Accepted: August 20, 2014 Croat. j. for. eng. 36(2015)1

Srđan Vasković, MSc. e-mail: srdjan_vaskovic@yahoo.com Assist. Prof. Vlado Medaković, PhD. e-mail: vlado.medakovic@gmail.com Faculty of Mechanical Engineering, East Sarajevo Department of Production Engineering Vuka Karadžića 30 71 123 Istočno Novo Sarajevo BOSNIA AND HERZEGOVINA Assist. Prof. Velid Halilović, PhD. e-mail: velidha@yahoo.com Assist. Prof. Jusuf Musić, PhD. e-mail: jusufmusic@yahoo.com Faculty of Forestry, Sarajevo Department of Forest Utilization Project Design and Construction in Forestry and Landscape Architecture Zagrebačka 20 71 000 Sarajevo BOSNIA AND HERZEGOVINA Assoc. Prof. Petar Gvero, PhD. e-mail: gvero.petar@gmail.com Faculty of Mechanical Engineering, Banja Luka Department of Thermal Engineering Bulevar vojvode Stepe Stepanovica 71 78 000 Banja Luka REPUBLIC OF SRPSKA BOSNIA AND HERZEGOVINA

123



Preliminary note

The Effect of Wider Logging Trails on Rut Formations in the Harvesting of Peatland Forests Jori Uusitalo, Marika Salomäki, Jari Ala-Ilomäki Abstract Peatlands are very problematic from the bearing capacity point of view. Therefore, logging activities on peatlands in Finland are mainly carried out during the coldest weeks in winter time. More intensive utilisation of peatland forests requires logging activities to be increasingly carried out during unfrozen conditions. Multiple passages of a harvester and a loaded forwarder used for the transportation of timber cause deep ruts on the forest floor. Wider logging trails have been presented as an interesting approach to increasing the number of forwarder passages along a single logging trail. It might be advantageous not to follow the same ruts on each pass but to choose a new parallel route, so that new ruts are formed alongside the previous ones. The study aimed at investigating whether it is beneficial to use wide trails in reducing rutting in forests growing on drained peatland. Field studies were conducted on a drained peatland in Alkkia experimental forest, located in Karvia in Western Finland. In the driving test, a forwarder was driven in the same forest site on logging trails with widths of 4.5 m, 6 m and 10 m. The results indicate that wider logging trails provide the forwarder driver with opportunities to reduce rutting in peatland forests. Keywords: tree harvesting, trafficability, rut formation, soil bearing capacity

1. Introduction The low bearing capacity of peatlands forms a severe obstacle for the prevailing harvesting machinery. Multiple passes of a harvester and a loaded forwarder used for the transportation of timber may cause deep ruts on the forest floor. Excessive machine sinkage has a direct bearing on the cost of machine operations and may lead to excessive site disturbance and soil damage (Tiernan et al. 2004, Zeleke et al. 2007, Ala-Ilomäki et al. 2011). Currently, logging activities on peatlands are mainly carried out during the coldest periods in the winter. There are, however, local and annual differences in terms of the extent to which peatlands freeze. Warm autumn months and the isolative effects of snow cover and dry upper peat layers hinder the freezing process and make many areas impossible to access, even in the winter. The pronounced climatic change at high latitudes, predicted to occur as a result of global warmCroat. j. for. eng. 36(2015)1

ing, is expected to prevent winter logging on peatland in these areas. More intensive utilisation of peatland forests requires logging activities to be increasingly carried out in unfrozen conditions. Peatlands consists of a 10 to 20 cm thick top layer that includes roots of trees and shrubs, followed by layer of decomposed peat. From the bearing capacity point of view, the top layer, with its considerable tensile strength provided by the roots of trees and shrubs, is essential, with the supporting function of the decomposed peat being of secondary importance. If the bearing capacity of peat soil is weak, the first forest machine passes may break the vital root matrix, drastically reducing trafficability. The mobility of forwarders in the summer logging of peatland forests has been studied rather extensively of late (O’Mahony et al. 2000, Tiernan et al. 2004, Zeleke et al. 2007, Ala-Ilomäki et al. 2011, Uusitalo and Ala-Ilomäki 2013). These studies show that the mobil-

125


J. Uusitalo et al.

The Effect of Wider Logging Trails on Rut Formations in the Harvesting of Peatland Forests (125–130)

ity of forwarders during the summer is critical. In many peatlands, the bearing capacity of peat soil is so weak that only a few forest machine passes along one logging trail will break the vital root matrix and may prevent forwarding operations (Nugent et al. 2003). Recent studies indicate that the most important characteristics affecting the bearing capacity of pine bogs are tree volume, the strength and moisture content of the uppermost layers of moss (Ala-Ilomäki et al. 2011, Uusitalo and Ala-Ilomäki 2013). In addition to that, logging residual is known to prevent excessive rutting although the role of logging residual has mostly been researched in mineral soils (McDonald and Seixas 1997, Eliasson and Wästerlund 2007, Gerasimov and Katarov 2010, Labelle and Jaeger 2012). Wider logging trails have been presented as an interesting approach to increasing the number of forwarder passages along a single logging trail. Recent simulation studies have proved that doubling the number of forest passes along the most critical logging trail sections would in many cases markedly improve the possibility of successfully completing summer logging operations on peat soils (Uusitalo et al. 2010, Haavisto and Uusitalo 2010). It has also been shown that increasing the width of a logging trail from 4 m to 6 m does not drastically decrease the profitability of peatland forestry in the long term (Salomäki et al. 2012). Wider logging trails enable the machine operator to choose a new route so that new ruts are formed

alongside the previous ones. Ultimately, increased trail width could double the number of forwarder passages on each logging trail. This paper presents the results of a test whereby a forwarder was driven in the same forest site on logging trails with widths of 4.5 m, 6 m and 10 m. The study aimed at investigating whether it is beneficial to use wide trails in reducing rutting in forests growing on drained peatland.

2. Materials and methods Field studies were conducted on a drained peatland in Alkkia experimental forest, located in Karvia in Western Finland. The study stand was drained for the first time in 1969 and the ditch network was cleaned in 1988. For the study, three straight logging trails, each roughly 350 m long, were marked prior to harvesting. The logging trails were laid out in parallel with 20 m spacing. The study trails were marked with coloured strips prior to harvest. All three study trails started with the 100 m long sections of 10 m wide trails, continued with 100 m long sections of 6 m wide trails and ended with 100 m long sections of 4.5 m wide trails (Fig. 1). There were 20 m long transition sections between the consecutive 100 m long sections. The borders of the study trails on the 6 m or 10 m wide sections were also marked with coloured strips prior to harvesting. The 4.5 m wide section was not marked, as it was considered the normal working method. Each 100 m long section included one study plot system-

Fig. 1 Schematic layout of the study trails and sample plots within the study stands

126

Croat. j. for. eng. 36(2015)1


The Effect of Wider Logging Trails on Rut Formations in the Harvesting of Peatland Forests (125–130)

J. Uusitalo et al.

Table 1 Technical data for the machines used in the study Equipment

Ponsse Fox harvester

Ponsse Buffalo forwarder

8

8

710/45-26.5

800/40-26.5

Clark Terra TL85 (front) wheel chains (rear)

Clark Terra TL85 (rear) wheel chains (front)

Net mass without tracks or chains, kg

18,200

18,900

Nominal Ground Pressure (NGP), kPa

41 (front)

46 (front)

(machines equipped with tracks or chains)

43 (rear)

43 (rear), without load

8

8

Number of wheels Tyres Tracks/chains

Number of wheels

atically laid out along the test trails. The study plots, 20 m in length and 20 m in width, were placed parallel to the study trails. Along the centre line of the sample plot, five sample lines, perpendicular to the centre line, were marked on the ground at 4 m intervals, the first sample line starting 2 m and the last sample line 18 m from the beginning of the plot. Harvesting and test drives were carried out in September 2011. The site was harvested with an 8-wheeled Ponsse Fox harvester that had a boom reach of 10 m and was equipped with tracks on the front bogie and chains on the rear bogie. The technical details of the harvester are provided in Table 1. The harvester was operated by an experienced operator. Normal thinning procedures were followed. All the trees removed from the logging trail and the trees adjacent to the trail were processed above the trail, resulting in logging residue accumulation on the logging trail. Trees located more than roughly 3 meters from border of the trail were processed outside the logging trail. Along the 6 m wide trail sections, the amount of removed trees in areas outside the logging trails was reduced, so that the volume of the remaining trees per hectare would equal that along the 4.5 m wide trail sections. No thinning outside the logging trails was accomplished along the sections with 10 m wide trails. Along the 6 m wide and 10 m wide trail sections, all trees were processed above the trail. It was estimated that brash mat along the 6 m wide trail section was distributed quite evenly along the whole 6 m wide area and along 10 m wide section quite evenly along 7 m wide area. These values were later used in making theoretical calculations on the mass of logging residuals per hectare. While harvesting, the automatic measurement and bucking optimization system stored the profiles of the removed stems in the STM format. After harvesting, the depth of the ruts caused by the harvester in each sample line was measured with Croat. j. for. eng. 36(2015)1

a horizontal laser levelling device and a laser levelling rod equipped with a laser beam detector. The device was first placed at a random location along the trail to obtain a reference height, which was then carved on the surface of a nearby tree. The reference level of the ground outside the wheel ruts was first measured to the left and the right of the sample line. The laser levelling rod used for measuring the height was pushed lightly against the ground to compact any loose surface layer vegetation and the relative level to the reference mark was calculated by reading the level of the laser beam detector attached to the surface of the laser levelling rod. The depth of ruts on both sides of the centre line was measured by placing the laser levelling rod at the bottom of the rut, reading the relative height of the bottom and calculating the depth of the ruts by comparing these values to the closest reference level of the ground. Rut depth after harvesting (Rutharv) is the mean of the rut depths of both tracks caused by the harvester. Test drives were carried out with an 8-wheeled Ponsse Buffalo forwarder (Table 1). Since the soil was very soft due to a long period of rain, the tests were carried out with no load. All three test trails were first driven once along the ruts caused by the harvester. The rut depth was then measured exactly as it was after the harvester pass. Rut depth after one forwarder pass (Rutforw_one) is therefore the mean of the rut depths of both tracks caused by one harvester pass and one forwarder pass. Next, the forwarder drove along the test trails three more times, except on the third test trail, which was driven only twice more. These results were treated equally regardless of the number of passes. While driving these two to three additional times, the forwarder operator was allowed to utilise the whole width of the trail and choose new parallel routes, so that new ruts were formed alongside the previous ones along the 6 m-wide and

127


J. Uusitalo et al.

The Effect of Wider Logging Trails on Rut Formations in the Harvesting of Peatland Forests (125–130)

Table 2 Means of the ruts measured in the study

Study trail

Nominal width

Rutwidth_tot

Rutharv

Rutforw_one

Rutforw_sev

Rutincr_forw (Rutforw_sev – Rutharv)

Vplot

Vharvested

Dry mass of logging residues

Mean m

m3

cm

kg/ha

1

4.5

341

8.4

11.4

14.1

5.7

188

73

15 700

2

4.5

337

9.3

17.2

22.4

13.1

192

53

12 300

3

4.5

361

16.0

17.2

21.3

5.3

177

54

11 500

All

4.5

346

11.2

15.3

19.3

8.0

185

60

13 200

1

6

436

14.5

19.1

18.4

3.9

172

81

12 800

2

6

386

9.2

10.8

13.2

4.0

202

92

14 900

3

6

448

14.4

15.0

18.5

4.1

141

62

10 600

All

6

423

12.7

15.0

16.7

4.0

172

78

12 800

1

10

433

18.0

19.1

21.2

3.2

202

90

12 300*

2

10

625

14.3

22.4

18.5

4.2

173

86

10 700*

3

10

625

18.0

19.5

23.2

5.2

171

84

11 700*

All

10

561

16.7

20.3

21.0

4.2

182

87

11 600*

*It was estimated that along 10 m wide trails the logging residues distributed quite evenly across a 7 m wide area. Rutwidth_tot = Total width of ruts; Rutharv = Rut depth caused by harvester; Rutforw_one = Rut depth after one forwarder pass; Rutforw_sev = Rut depth of several forwarder passes

10 m wide sections. Along the 6 m wide trail sections, both the depth of the first (original) ruts on both sides of the trail (Rutfirst_left, Rutfirst_right) and the depth of the second (new) ruts (Rutsecond_left, Rutsecond_right) on both sides of the trail were measured. In two of three sample plots along the 10 m wide sections, even third ruts were formed (Rutthird_left, Rutthird_right). The rut depth of several passes (Rutforw_sev) is the maximum of the means of the first, second and third ruts:

Rutforw_sev = Max (Mean (Rutfirst_left, Rutfirst_right), Mean (Rutsecond_left, Rutsecond_right), Mean (Rutthird_left, Rutthird_right))

(1)

Along the 4.5 m wide trails, no second ruts were formed, which means that Rutforw_sev is equal to the mean of the first ruts (Mean (Rutfirst1, Rutfirst2). The increase in rut depths after harvesting (Rutincr_forw) is calculated by subtracting Rutforw_sev from Rutharv. The total width of all ruts was also measured along all sample lines. The total width of ruts (Rutwidth_tot) is the distance from the left border of the rut on the far left to the right border of the rut on the far right.

128

After the test drives, all the remaining trees on the study plots were measured for diameter at breast height (DBH). The volume of trees on the plot (Vplot) is the sum of all the harvested (Vharvested) and remaining trees. The DBH of the remaining trees was converted to cubic metres by using taper equations by Laasasenaho (1982). The DBH, height and volume of each harvested tree was derived by using the harvester stem profile data (STM format) that gives the diameter of the stem profile in 10 cm intervals. Cubing of trees was derived by summing up 10 cm long sections. Estimates of dry mass of logging residuals were calculated using biomass equations by Repola (2009).

3. Results Rut depths are significant. The mean of rut depths after one harvester pass (Rutharv) varies between 8.4 cm and 18.0 cm and the mean of rut depths after one harvester pass and one forwarder pass (Rutforw_one) between 10.8 and 22.4 cm (Table 2). Increasing the logging trail width from 4.5 m to 6 m increased the total width of the Croat. j. for. eng. 36(2015)1


The Effect of Wider Logging Trails on Rut Formations in the Harvesting of Peatland Forests (125–130)

ruts by 77 cm (from 346 cm to 423 cm), and increasing the trail width from 4.5 to 10 m by 215 cm (from 346 cm to 561 cm). Wider logging trails resulted in a decreased rut depth. The increase in rut depths after harvesting (Rutincr_forw) is clearly lower on 6 m wide (4.0 cm) and 10 m wide (4.2 cm) test trails than on 4.5 m wide trails (8.0 cm). Characteristics of the test trail sections were similar in terms of growing stock (Vplot) and mass of logging debris, but the means of the ruts on different test trail sections indicate that the strength of the soil increased from the start to the end of the trails. The first 100 m long sections, with 10 m wide trails, clearly had the deepest ruts after harvesting, the 100 m long sections with 6 m wide trails were the second deepest, and the third 100 m long sections with 4.5 m wide trails had the shallowest ruts (Table 2).

4. Discussion The test arrangements were affected by very wet conditions. The autumn of the test period was very rainy, making the peaty soil very humid and soft. Therefore, the forwarder drove the tests with no load. Overall, the mean rut depth after harvesting was considerably deeper than in earlier tests carried out in Finland in August. Uusitalo and Ala-Ilomäki (2013) measured mean rut depths at 7.0 to 13.5, the mean of all study plots (with a brash mat) being 10.1 cm in circumstances with smaller growing stock. In the study by Uusitalo et al. (2012), rut depths after harvesting varied from 2.0 to 17.1 cm, the mean of all study plots being 9.0 cm. Pre-planning of the test trail sections failed to some extent. The soil properties clearly changed along the 350 m long test trails. It would have been wise to change the order of the 4.5 m, 6 m and 10 m wide sections within the three trails. No visible differences in trial conditions were found prior to the tests, since the properties of the soil were not adequately measured. However, the sections were similar in terms of growing stock, which is the variable most often used for describing trafficability in peatland forests. The study material comprises only one forest and the operations were carried out with only one harvester and one forwarder. Pine bogs in Finland vary to some extent in terms of growing stock and thickness of the peat layer. The study stand was quite typical in terms of growing stock. Thickness of the peat layer was not measures accurately but few random measurements indicated that the thickness of the peat layer was more than four meters. Moreover, the machines were rather poorly equipped for peatland operations. Equipping the machines with special soft soil tracks on both front and rear bogies would probably have decreased rutting. Croat. j. for. eng. 36(2015)1

J. Uusitalo et al.

Despite the many shortcomings of the test, the results of the study can be considered promising. The widening of the trail from 4.5 m to 6 m was clearly beneficial from the point of view of the maximum rut depth. Allowing the forwarder driver to choose a new parallel route, so that new ruts were formed alongside the previous ones, resulted in 5 cm lower maximum rut depths on average after three to four forwarder passes. However, this study does not reveal the extent to which wider logging trails increase the maximum number of passages allowed on one trail. The increase in trial width from 4.5 m to 10 m also markedly decreased the maximum rut depths, yet the increase from 6 m to 10 m did not result in a further decrease in rut depth. Trails wider than 6 m bring additional variables that have an effect on the results into play. Increasing brash accumulation on the trail also raises the question of how the trees should be processed. In this trial, no extra instructions were given to the harvester operator as to how to accumulate the brash mat. If the brash mat on the 10 m wide trail was forced to accumulate on a narrow belt, it would most probably have reduced rutting. Widening the trail to 10 m will, however, reduce the economic output of the forest in the long run (Salomäki et al. 2012). The test driver gave very positive feedback on driving along wider trails. He felt that wider trails gave him more opportunity to plan his driving, allowing him to fully benefit from the bearing capacity of the trail and thus reduce rutting. The test was carried out in very difficult conditions, with inadequately equipped machines and only in one forest. Therefore, the test should be repeated with larger study material. It is also important that the key soil properties are measured.

5. References Ala-Ilomäki, J., Högnäs, T., Lamminen, S., Sirén, M., 2011: Equipping a conventional wheeled forwarder for peatland operations. International Journal of Forest Engineering 22(1): 7–13. Eliasson, L., Wästerlund, I., 2007: Effects of slash reinforcement of strip roads on rutting and soil compaction on a moist fine-grained soil. Forest Ecology and Management 252(1-3): 118–123. Gerasimov, Y., Katarov, V., 2010: Effect of bogie track and slash reinforcement on sinkage and soil compaction in soft soils. Croatian Journal of Forest Engineering 31(1): 35–45. Haavisto, M., Uusitalo, J., 2010: Turvemaan ensiharvennusten maastovaurioita minimoiva reitinoptimointi LOGTRACK-simulattorilla. The Finnish Forest Research Institute. Unpublished manuscript 19 p.

129


J. Uusitalo et al.

The Effect of Wider Logging Trails on Rut Formations in the Harvesting of Peatland Forests (125–130)

Laasasenaho, J., 1982: Taper curve and volume functions for pine, spruce and birch. Communicationes Instituti Forestalis Fenniae 108, 74 p. Labelle, E.R., Jaeger, D., 2012: Quantifying the use of brush mats in reducing forwarder peak loads and surface contact pressures. Croatian Journal of Forest Engineering 33(2): 249–274. McDonald, T., Seixas F., 1997: Effect of slash of forwarder soil compaction. Journal of Forest Engineering 8(2): 15–26. Nugent, C., Kanali, C., Owende, P.M.O., Nieuwenhuis, M., Ward, S., 2003: Characteristic site disturbance due to harvesting and extraction machinery traffic on sensitive forest sites with peat soils. Forest Ecology and Management 180(1-3): 85–98. O’Mahony, M.J., Ueberschaer, A., Owende, P.M.O., Ward, S.M., 2000: Bearing capacity of forest access roads built of peat soils. Journal of Terramechanics 37(3): 127–138. Repola, J., 2009: Biomass equations for Scots pine and Norway spruce in Finland. Silva Fennica 43(4): 625–647. Salomäki, M., Niemistö, P., Uusitalo, J., 2012: Ensiharvennuksen toteutusvaihtoehdot ja niiden vaikutukset männikön tuotokseen ja kasvatuksen kannattavuuteen ojitetuilla tur-

vemailla – simulointitutkimus. Metsätieteen aikakauskirja 3/2012, 163–178. Tiernan, D., Zeleke, G., Owende P.M.O., Kanali, C.L., Lyons, J., Ward, S.M., 2004: Effects of working conditions on forwarder productivity in cut-to-length timber harvesting on sensitive sites in Ireland. Biosystems Engineering 87(2): 167–177. Uusitalo, J., Ala-Ilomäki, J., 2013: The significance of aboveground biomass, moisture content and mechanical properties of peat layer on the bearing capacity of ditched pine bogs. Silva Fennica 47(3): 1–18. Uusitalo, J., Haavisto, M., Niemistö, L., Kataja, J., 2010: Assessing the effect of harvesting method on soil disturbances with a spatial harvesting simulator. Proceedings FORMEC 2010 Conference, July 11–14, 2010, Padova, Italy, 5 p. Uusitalo, J., Kaakkurivaara, T., Haavisto, M., 2012: Utilizing airborne laser scanning technology in predicting bearing capacity of peatland forest. Croatian Journal of Forest Engineering 33(2): 329–337. Zeleke, G., Owende, P.M.O., Kanali, C.L., Ward, S.M., 2007: Predicting the pressure-sinkage characteristics of two forest sites in Ireland using in situ soil mechanical properties. Biosystems Engineering 97(2): 267–281.

Author’s address: Prof. Jori Uusitalo, PhD.* e-mail: jori.uusitalo@luke.fi Natural Resources Institute Finland (Luke) Kaironiementie 15 39700 Parkano FINLAND Marika Salomäki MSc. e-mail: marika.salomaki@pp.inet.fi Könnönkyläntie 148 61940 Hyyppä Kauhajoki FINLAND

Received: August 28, 2013 Accepted: September 12, 2014

130

Jari Ala-Ilomäki PhD. e-mail: jari.ala-ilomaki@luke.fi Natural Resources Institute Finland (Luke) PL 18 01301 Vantaa FINLAND * Corresponding author Croat. j. for. eng. 36(2015)1


Preliminary note

Electronic Control of Motor Axles of Forestry Trailers Marco Manzone, Paolo Balsari Abstract Timber transport in the forest is a very complex operation, and expensive from an energy point of view. Furthermore, in some cases, this operation can be very difficult and dangerous in unfavorable conditions of the road surface, which mainly occur in winter (frozen ground) and in spring (muddy ground). The goal of this study was to develop an innovative transmission hydraulic control system for trailer motor axle that would allow correlating the forward speed of the trailer with that of the tractor, which is not necessarily always the same. In detail, the innovative motor axle of the trailer is driven by a hydraulic motor through a pump controlled electronically. A specific software is able to correlate the forward speed of the trailer with the speed of the tractor in any operating situation, thanks to the information given by a potentiometer screwed behind to the trailer component coupled to the tractor. The innovative system developed to control the trailer motor axle provides the possibility to use the trailer with any type of tractor, not requiring long and complex adjustment of tractor and trailer, since it is completely independent from the type of the tractor used. Keywords: Timber transport, forestry trailer, motor axle, electronic control

1. Introduction Italian forestry is characterized by steep terrain and high ownership fragmentation (Mason et al. 1999). All these factors tend to slow down the introduction of mechanized harvesting, determining the current prevalence of requirements of labor-intensive operations (Magagnotti et al. 2012). In order to try to solve this problem, versatile low-investment machinery should be developed that could offer a suitable balance between capital and labor inputs (Spinelli et al. 2013). Among all forestry operations, timber transport is a very complex operation, and also expensive from an energy point of view (Antoniade et al. 2012, Johansson et al. 2006, Lindholm and Berg 2005, Angus-Hankin 1995). Furthermore, in some cases, this operation could be very difficult and dangerous in unfavorable conditions of the road surface, which mainly occur in winter (frozen ground) and in spring (muddy ground). For these reasons, the use of trailers with motor axle is increasing in large and small forests. The motor axle drive system can be mechanical or hydraulic. In the first case, due to the constant transmission ratio, it Croat. j. for. eng. 36(2015)1

is possible to use only the tractor with the tire configuration for which the trailer was designed. Furthermore, to allow this type of transmission, trailer manufacturer companies are forced to use a transmission ratio that can reduce the trailer forward speed by 3–5% compared to the coupled tractor. It is also necessary to allow the convoy to make the curves without having the trailer, which has a smaller radius of curvature, pushing the tractor. At present, the hydraulic transmission mounted on forestry trailers, generally, does not allow to manage the speed of the hydraulic motor and the system is used in forest only for short distances and only when it is necessary to increase the tires grip. Unlike the mechanical traction, this solution has the advantage to not determine potential breaks of the mechanical components of the transmission, as before breakage of mechanical parts a viscous joint slippage is present but, for the same reason, it cannot be used downhill where it is absolutely necessary to reach the maximum tires grip. The goal of this study was to develop an innovative transmission hydraulic control system for trailer mo-

131


M. Manzone and P. Balsari

tor axle that would allow correlating the forward speed of the trailer with that of the tractor, which is not necessarily always the same.

2. Materials and methods In developing this innovation, efforts have been made to enhance the positive aspects and to reduce those negatives of the two trailer motor axle driving systems (mechanical and hydraulic). To be specific, in the innovative system developed, the motor axle of the trailer is driven by a hydraulic motor through an electronically controlled pump. A specific software is able to correlate the forward speed of the trailer with the speed of the tractor in any operating situation, thanks to the information given by a potentiometer screwed behind to the trailer component coupling to the tractor. In order to identify more useful components, a market survey was carried out to examine the technical characteristics of the solutions already developed by other Research Centers and manufacturers of agricultural equipment and components. This investigation showed that all the technical solutions developed could only solve a part of the problem, and required sophisticated components not suitable for forestry conditions. Otherwise, components and solutions used for the innovative system were studied or at least protected by steel guards. The trailer used for the application of the innovative system is a commercial forestry trailer (Offine

Electronic Control of Motor Axles of Forestry Trailers (131–136)

Terpa® TM60) with a single motor axle, full load mass of 6.0 t and length and width of 4.0 m and 1.8 m, respectively. The trailer axle had a transmission ratio of 1/30. A pump (Sauer-Danfoss® M46-20777) was fixed on the rudder with variable torque (37–106 Nm) and under the frame a hydrostatic motor (Sauer-Danfoss® M4443028) with a displacement of 250 cm3 rev-1. A mechanical gearbox was inserted between the hydraulic motor and the axle that could disengage the system when necessary. The transmission between the gearbox and the axle was realized with a drive shaft. An oil tank (80 liters of capacity) was mounted behind the axle and under the trailer load floor. The electronics central unit was placed at the side of the rudder in order to make an easy access and to protect it from possible damage during the use. With the aim to modulate the »movement« of the towing eye of the trailer and cushion the impact between the wheels and a possible obstacle present on the soil, the towing eye has been prepared by interposing between the tightening nut and the bushing welded to the rudder of the disc springs made in Carbon steel (Fig. 1). The size and thickness of the disc springs can be varied depending on the total weight of the trailer and the sensitivity assigned to the system. This latter can also be adjusted using a different number and arrangement of the disc springs (in series or in parallel). Furthermore, the internal diameter of the spring disc must

Fig. 1 Scheme of components of the innovative system

132

Croat. j. for. eng. 36(2015)1


Electronic Control of Motor Axles of Forestry Trailers (131–136)

Fig. 2 Scheme of engage/disengage system of the gearbox be larger than that of the tow pin so that the latter can slide into the disc springs. The engaging and disengaging of the system is obtained automatically by setting the P.T.O. rotation of the tractor. In fact, activating the P.T.O. of the tractor, a hydraulic cylinder, powered by the hydraulic pump of the trailer, activates the mechanical clutch of the gearbox. The disengaging of the system takes place by a coil spring that, when the P.T.O. rotation is stopped, takes the mechanical clutch of the gearbox back in the initial position. When the system is activated, a warning green light is switched on (Fig. 2). In order to assess the functionality of the developed system, field tests aimed at determining the synchronization of the trailer forward speed with the speed of the tractor have been made. The speed synchronization was determined through the data acquired by a potentiometer mounted behind the towing eye. The towing eye is used in order to provide the information to the electronic control unit necessary to modulate the flow and pressure of oil input to the hyCroat. j. for. eng. 36(2015)1

M. Manzone and P. Balsari

drostatic motor, switch the position of the towing eye with respect to »0 point« (neutral point) with current pulses of different intensity (the further one moves away from the »0 point«, the greater is the current intensity). In detail, when the values coming from the potentiometer are positive, it means that the tractor pulls the trailer, while when the value are negative the trailer pushes the tractor. The trials were carried out using a 4WD tractor (Fiatagri® 88–94) with the nominal power of 67 kW and mass of 3500 kg and a trailer with empty mass of 1180 kg and full mass of 3 tones. The load of the trailer was made of concrete blocks. The tests were carried out on different itineraries traced on natural soil (ground) with the presence of curves (L and R) and in different slope condition. Two routes were chosen for the tests: Itinerary 1 had the length of about 300 m in a flat area with turf with two curves of 180° and a radius of curvature of 25 m; itinerary 2 had the length of about 120 meters in an area with an average slope of 30% and in bare soil. The convoys (tractor + trailer) have been operating with three different forward speeds (2–3–4 km h-1), while in itinerary 2, two directions were used (uphill and downhill) with constant forward speed (3 km h-1). The tensile force exerted by the tractor to pull the trailer has also been measured in each itinerary. This measurement was performed using a digital dynamometer (SAUTER® FH50k) with the capacity of 50 kN and resolution of 10.0 N.

3. Results In all tests carried out, the electronic system has guaranteed an efficient control of the mechanical coupling, synchronizing the speed of the trailer with the speed of the tractor. This has been proved by the capacity to maintain the towing eye of the trailer in »neutral« position. Operating on flat soil and with the developed control system not activated, the towing eye tended to move forward increasing the trailer forward speed, while with the innovative system activated, this phenomenon was very limited and the towing eye position also remained stable when operating with different tractor forward speeds. The peaks highlighted in Fig. 3 are mainly due to the unevenness present in the itinerary. The values obtained in the second itinerary show that thanks to the use of the innovative system, it is possible to reduce to minimum the thrust on the coupling pin of the tractor at different slopes. Fig. 4 and 5

133


M. Manzone and P. Balsari

Electronic Control of Motor Axles of Forestry Trailers (131–136)

Fig. 3 Towing eye position with disengaged/engaged innovative system with different forward speeds

Fig. 4 Towing eye position with disengaged/engaged innovative system in uphill travel

show a neutral position of the towing eye independently from the travel direction (uphill and downhill). When the developed system was not active, the tensile force required to tow the trailer was proportional to the mass of the trailer and to the slope, and when the system was active, the tensile forces registered in the trails were similar (between 144 and 186 daN) for all the operating conditions (Table 1).

134

4. Discussion Graphs showed that the electronic control unit can properly modulate the flow of oil to the hydraulic motor while maintaining the ÂťzeroÂŤ position of the towing eye, independently from the slope soil. Furthermore, the innovative system also shows a good performance In circuitous itineraries; in fact, the Croat. j. for. eng. 36(2015)1


Electronic Control of Motor Axles of Forestry Trailers (131–136)

M. Manzone and P. Balsari

Fig. 5 Towing eye position with disengaged/engaged innovative system in downhill travel Table 1 Tensile force with deactivated/activated innovative system Slope

Trailer mass

Exerted force

%

kg

daN

1180

98

3000

256

1180

426

3000

1289

1180

144

3000

156

1180

152

3000

186

0 Not activated system 30

0 Activated system 30

trailer, while making a radius of curvature different from that of the tractor, does not generate dangerous pressures like those in tractions where the transmission ratio is the same (mechanical transmission). Tests carried out have shown that, independently of whether the trailer is loaded or unloaded, the forces acting on the coupling device (trailer–tractor) are limited. The force required to activate the developed system is about 150 daN, and this value that can be ensured by any forestry tractor, including tractors of lower power (Table 1). The innovative system developed to control the trailer motor axle provides the possibility to use the Croat. j. for. eng. 36(2015)1

trailer with any type of tractor, not requiring long and complex adjustment of tractor and trailer, since it is completely independent from the type of the tractor used. Consequently, all the forestry tractors could be used, which reduced the maintenance costs thanks to more uniform yearly hours of use of the tractors. The developed device could also be conveniently used in small forestry companies that cannot purchase a forestry trailer with motor axle due to limited capital. They could share this investment with other forestry companies having the possibility to use the same trailer with different types of tractors. The new developed system, working with the same tensile forces in different operating conditions, also results in increasing the general safety level, because it protects the tractor from dangerous solicitations generated by the trailer while driving on forestry roads, which happens regularly when the current commercial solutions are used.

5. Conclusions This innovation system could be considered a viable alternative to the trailer traction systems present today on the market, because it can reduce the farm investment and improve the versatility of the forestry trailer. Furthermore, thanks to its control system, it also increases the general safety level by protecting the tractor from dangerous solicitations generated by the trailer while driving on bad roads.

135


M. Manzone and P. Balsari

Electronic Control of Motor Axles of Forestry Trailers (131–136)

This developed system is, at present, protected by an industrial patent (n° TO2012A000958) deposited by the University of Turin.

Johasson, J., Liss, J.E., Gullberg, T., Björheden, R., 2006: Transport and handling of forest energy bundles-advantages and problems. Biomass and Bioenergy 30(4): 334–341.

6. References

Magagnotti, N., Spinelli, R., Güldner, O., Erler, J., 2012: Site impact after motor-manual and mechanised thinning in Mediterranean pine plantations. Biosystems Engineering 113(2): 140–147.

Angus-Hankin, C., Stokes, B., Twaddle, A., 1995: The transportation of fuelwood from forest to facility. Biomass and Bioenergy 9(1–5): 191–203. Antoniade, C., Şlincu, C., Stan, C., Ciobanu, V., Ştefan, V., 2012: Maximum loading heights for heavy vehicles used in timber transportation. Bulletin of the Transilvania University of Braşov, Series II: Forestry, Wood Industry, Agricultural Food Engineering 54(5): 7–12. Lindholm, E.-L., Berg, S., 2005: Energy requirement and environmental impact in timber transport. Scandinavian Journal of Forest Research 20(2): 184–191.

Mason, B., Kerr, G., Simpson, J., 1999: What is continuous cover forestry? Forestry Commission, Information Note 29, Edinburgh, 1–8. Rozt, C.A., 1987: A standard model for repair costs of agricultural machinery. Applied Engineering in Agriculture 3(1): 3–9. Spinelli, R., Magagnotti, N., Facchinetti, D., 2013: Logging companies in the European mountains: an example from the Italian Alps. International Journal of Forest Engineering 24(2): 109–120.

Authors’ addresses:

Received: July 3, 2013 Accepted: May 29, 2014

136

Marco Manzone PhD.* e-mail: marco.manzone@unito.it Prof. Paolo Balsari, PhD. e-mail: paolo.balsari@unito.it Università degli studi di Torino – DISAFA Via Leonardo da Vinci, 44 Grugliasco (TO) ITALY * Corresponding author Croat. j. for. eng. 36(2015)1


Preliminary note

Estimating Time Consumption and Productivity of Roundwood Skidding in Group Shelterwood System – a Case Study in a Broadleaved Mixed Stand Located in Reduced Accessibility Conditions Stelian Alexandru Borz, Gheorghe Ignea, Bogdan Popa, Gheorghe Spârchez, Eugen Iordache Abstract In Romanian forestry, skidders represent the most used equipment for wood extraction, while the group shelterwood system is one of the most used silvicultural forest management strategies. Production rates are important indicators when trying to assess the efficiency of a process, since they can be used in different practical applications, starting with operational planning and ending with energetic analyses or LCA studies. The reduced accessibility of forest stands is one of the main problems of Romanian forestry, and therefore a lot of time and energy is usually spent in harvesting operations. In the present days, it has become very important to know the effects of these operational conditions. In order to evaluate the effects of very long skidding distances on the time inputs and productivity, a time study was conducted for a Romanian skidder that operated in a mixed hardwood stand located in Central Romania, where group shelterwood cuttings were applied. We found out that the time input of a winching work cycle was most affected by winching distance and the number of logs, and developed time consumption models for the main groups of operations. For the mean conditions (winching distance of 8.7 m, mean skidding distance of 1706.3 m, a load volume of 4.89 m3 and 6.48 logs per turn) the net and gross production rates were of 4.41 m3/h and 3.12 m3/h, respectively. The results of this study may be helpful in operational costing or harvesting planning when dealing with reduced accesibility conditions. Keywords: skidding, reduced accessibility, group shelterwood system, time consumption, production rates

1. Introduction In Romania, forests cover almost 27% of the national territory, and about 37% of them are located in the hills zone. Such forests are usually composed of mixed broadleaved stands with an important participation of oaks. Also, high forest regime is mostly used for managing such forests, which are generally harvested at the age of 100–120 years. Therefore, the use of chainsaws seems to be the only option in harvesting such forests, especially when associated with large Croat. j. for. eng. 36(2015)1

trees (Jourgholami et al. 2013). On the other hand, it is quite commonly accepted that the increment of mechanization level in harvesting operations will result in a greater economic efficiency (Oprea 2008). It depends in a great measure on a series of factors such as forest and intervention type, particular terrain conditions, level of social acceptance, and legislative aspects. Harvesting systems, such as those associated with chainsaws and skidders, are quite common in Europe when dealing with extractions that yield important wood

137


S. A. Borz et al.

Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

quantities, e.g. selective (Sabo and Poršinsky 2005, Ghaffaryian et al. 2013, Vusić et al. 2013) and conversion cuttings (Picchio et al. 2009), where the owning and operating costs are better amortized (Oprea 2008). It is also possible to use such equipment in thinning operations (Zečić 2005, Gallis and Spyroglou 2012, Vusić et al. 2013), or in other particular conditions such as salvage logging (Borz et al. 2013) or on sensitive sites when they may be used along with animal traction (Magagnotti and Spinelli 2011). On the other hand, in the Romanian forestry conditions, animal traction is used intensively in thinning operations (Borz and Ciobanu 2013), especially in the first ones, while skidders or farm tractors are more frequently used when dealing with main cuttings. Also, in Romania, the harvesting system associated with chainsaws and skidders is the most frequently used (Sbera 2007) especially when dealing with increased volumes of logs, even if other modern equipment has been introduced and used on small scale in harvesting operations (Borz et al. 2011). Furthermore, knowing the efficiency of the equipment used in harvesting operations may help in a better production organization, while the assessment of their environmental impact is very helpful in formulating strategies and policies (Magagnotti and Spinelli 2012a). The efficiency assessment of a production system in harvesting operations is carried out by using work measurement techniques, which are applied in order to develop models or to compare between two or more alternative treatments (Magagnotti and Spinelli 2012a). Modeling studies are usually done in order to obtain empirical models either in case of testing new equipment or in case of equipment working in new, unstudied conditions (Visser and Spinelli 2011). In timber harvesting operations, an adequate development of the forest transportation infrastructure may offer the premises for an increased productivity (Oprea 2008) and a reduced impact, by shortening the time of equipment deployment in the forest and reducing the energy expenditure in harvesting operations. In fact, the whole strategy related to forest road network development at optimal parameters is related to the efficiency of upstream processes such as timber harvesting (Bereziuc et al. 2006) and in particular to timber extraction operations (Ghaffaryian and Shobani 2008, Naghdi and Limaei 2009). Unfortunately, one of the main particularities of harvesting operations in Romania is that they are frequently developed in reduced accessibility conditions due to an insufficient development of the forest transportation network (Oprea 2008), which is characterized by a density index of only 6.5 m/ha (Olteanu 2008), and which

138

could be one of the limiting factors when considering the use of new short or medium distance harvesting equipment such as processor tower yarders (Borz et al. 2011). Furthermore, if compared with forests located in mountainous areas, Romanian hill forests are characterized by quite different work conditions due to preponderant presence of hardwoods such as oaks, beech and hornbeam as well as other mixed hardwood forests (Şofletea and Curtu 2008) located on clayeyloamy soils and managed by applying, mainly, group shelterwood silvicultural systems. Since the harvesting strategy plays a key role in forestry, a lot of attention has been given to efficiency assessment of different harvesting equipments. In case of skidding operations, the studies done so far revealed that the time consumption per turn may be explained by certain independent variables. Furthermore, skidding distance has been found to be one of the most relevant independent variables when explaining the time consumption (Borz et al. 2013, Borz et al. 2014a, Gallis and Spyroglou 2012, Ghaffaryian et al. 2013, Özturk 2010a, Özturk 2010b, Sabo and Poršinsky 2005, Vusić et al. 2013). However, most of the studies have been conducted for short extraction distances, usually up to 400 m (Gallis and Spyroglou 2012, Ghaffaryian et al. 2013, Özturk 2010a, Özturk 2010b, Sabo and Poršinsky 2005, Vusić et al. 2013) reporting different productivities depending on other factors such as load volume, while the effect of very long extraction distances on time consumption and productivity has never been thoroughly studied. In this context, the aim of this paper was to estimate the time consumption and skidding productivity for a felling area located in reduced accessibility conditions, where group shelterwood cuttings were applied. In order to fulfill the aim of the study, the objectives were set to: (i) modeling of time consumption for skidding, (ii) estimating the productivity of the entire system and (iii) provide an overview of general time structure in such operational conditions.

2. Material and methods The study was carried out in compartment 77C located in Laslea-Floreşti forest administrated by the State Forest District of Sighişoara. This forest is located in the central part of Romania, right at the border between Mureş and Sibiu counties (Fig. 1), and is characterized by altitudes ranging from 400 to 500 m, presence of natural forests containing native deciduous species such as beech, sessile oak and hornbeam. Felling area was located on a plateau having the general slope between 0 and 10% and it was opened up by two Croat. j. for. eng. 36(2015)1


Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

S. A. Borz et al.

Fig. 1 Study location skid roads developed near a peak and a stream, respectively (Fig. 1). In such conditions the Romanian forest management is done by applying the high forest regime and group shelterwood silvicultural system. Following the tree inventory in the felling area, a volume of 605 m3 was proposed for extraction (beech – 119 m3, sessile oak – 234 m3 and hornbeam – 252 m3). The forest was 120 years old. As usual in Romania (Oprea 2008), harvesting operations were performed by a team composed of three workers, of which one was responsible for tree felling and processing, while the other two dealt with skidding operations. Also, following tree felling and processing, in such operations the resulting logs having lengths up to 20 m (usually imposed by forest administrator) are extracted by cable skidders to the roadside landings, where supplementary operations are required in order to obtain final assortments and stack them into piles (Oprea 2008). Skidding operations were performed by a TAF 690 OP wheeled winch skidder (Fig. 2), equipped with a double–drum TA-2AM Romanian made winch, mechanically powered and steered electro-pneumatically, Croat. j. for. eng. 36(2015)1

Fig. 2 TAF 690 OP winch skidder in the studied area having a nominal pulling force of up to 11 tons. This skidder is equipped with a Perkins 1104C-44T made engine with the nominal power of 67 kW. One of the supplementary equipment is the hydraulically powered adjustable front blade, which is used for landing operations (stacking) and for skid roads cleaning. A team composed of 4 operators collected data during the field study. Time and quantity inputs/outputs,

139


S. A. Borz et al.

Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

Table 1 Description of work (time) elements and process variables Factor

Description

Measurement/calculation

Winching Maneuvring and parking, Tmp

(Optional) Begins when the skidder has just ended its empty travel (arrived at winching Camera + office processing area) and ends when all the positioning maneuvers end of media files

Cable releasing, Tcr

Begins when the worker has just grabbed the cable and ends when the worker arrives Camera + office processing at the log to be winched of media files

Log hooking, Tlh

Begins when the worker performs the first action in order to hook the log and ends when it signals the skidder operator that the log can be winched

Camera + office processing of media files

Mechanical winching, Tmw

Begins when the winch starts pulling and ends when the log arrived in the rear part of the skidder

Camera + office processing of media files

Log unhooking, Tlu

(Optional) Begins when the worker initiates the first action in order to detach the log and ends when the cable is free again

Camera + office processing of media files

Species, S

Visually assessed

Winching distance, WD

Slope distance measured between the rear part of the skidder and the end at which the log was attached

Tape

Diameter at the small end, Dthin

Caliper

Diameter at the large end, Dthick –

Caliper

Log length, LL

Distance between log ends, measured on the log

Tape

Log volume, LV

Volume of the log calculated in function of Dthin, Dthick and LL using classical relations

MS Excel database + calculus relations

Number of logs, NL

Number of logs winched in order to make a load for on–road skidding

Visually assessed

On-road skidding Empty travel, Tet

Begins when the skidder has ended its actions on landing (or first start of a day) – initiStopwatch ates movement, and ends when the skidder arrives at the winching area

Load attachment, Tla

Begins when worker starts to hook all the logs in the rear part of the skidder and ends when the load is suspended in the rear part of the skidder

Stopwatch

Loaded travel, Tlt

Begins when the skidder initiates its movement from winching area and ends when the skidder stops on the landing for detachment

Stopwatch

Load detachment, Tld

Begins with the first action of landing the load and ends when both cables are released Stopwatch

Empty travel distance, ETD

Slope distance between the point in which the load was attached and the point in which the load was detached on landing

GPS receiver + Digital Terrain Model + GIS software

Loaded travel distance, LTD

Slope distance between the point in which the load was detached on landing and the point in which the skidder stopped for winching or maneuvering and parking

GPS receiver + Digital Terrain Model + GIS software

Load volume, V

Sum of volumes of logs within a load

MS Excel calculation

Landing operations Maneuvering and stacking, Tlo

Begins when first movement maneuver is initiated and ends when all maneuvering and Stopwatch stacking tasks are accomplished. Maneuvers up to 25 m in distance.

as well as process variables, were measured differently according to the studied group of operations, by adapting a time study to the general concepts ­described by Björheden et al. (1995), respectively Magagnotti and Spinelli (2012a). In case of winching operations, time was recorded using a video camera while distance was

140

measured using a measuring tape. In case of on-road skidding, time was measured using a professional stopwatch, while distance was measured using a GPS receiver. All the operational variables were recorded in a field book, and, depending by the data acquisition method, some of them were further processed during Croat. j. for. eng. 36(2015)1


Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

the office phase of the study (Table 1) in order to obtain the needed data. For each winched log, we measured the diameters at the ends (Dthin and Dthick, respectively), length (LL) as well as the winching distance (WD). The volume of each log (LV), as quantitative input/output, was computed based on its diameters and length. Log volumes were used for determining the volume of each load (V). Species (S) of each log, as well as the number of winched logs (NL) within a winching work cycle, were visually assessed/counted and recorded in the field book. A particularity of on–road skidding operations was that the skidder operator used one skid road for the empty travel and the other for loaded travel, while loaded travel was performed downhill (Fig. 1). A description of work and time elements, as well as of recorded process variables, is presented in Table 1. Skidding operations were studied in a period of one week. A total number of 285 logs were winched and a total number of 44 work cycles were studied for winching, on–road skidding and landing operations. During the field study, we collected separately the whole operational duration, as well as the duration of mechanical (MD), operational (OD) and personal delays (PD). Empirical models concerning the time consumption were developed using Statistica 8.0 software package. These models were developed in order to estimate the time consumed in a winching work cycle (WCT), empty travel (ETT) and loaded travel (LTT). The net production rates for winching and on–trail skidding operations were calculated as the ratio between the realized production and the delay–free time consumption. Overall net and gross production rates (including landing operations) were calculated the same way, by using the total time (including delays) in case of gross production rate.

3. Results and discussion During the field study, a number of 285 logs were skidded, having a total volume of 215 m3 (Table 2), which represented about 36% of the volume to be harvested within the felling area. This meant that, on average, a number of 6.48 logs were skidded per turn, in the conditions of a mean volume of 4.89 m3 per load. In the above described conditions and for a mean winching distance of about 9 m, the time consumption for a delay-free winching work cycle was about 12 minutes (Table 2). Also, by excluding the winching time, in the same conditions regarding the volume and the number of logs per load, and for a mean skidding distance of about 1 706 m, the time consumption for a delay-free on-road skidding cycle time was about 50 minutes (Table 2). This means that it took about 61.4 minutes for a full skidding turn, excluding landing Croat. j. for. eng. 36(2015)1

S. A. Borz et al.

operations, which is comparable to the results obtained when using a farm tractors equipped for skidding in case of winching distances of 17 m and skidding distances up to 500 m (Spinelli and Magagnotti 2012). Given the reduced slope and the character of felling specific to the local conditions, which eased the cable releasing, within a winching work cycle the most time consuming category (Fig. 3) was that referring to mechanical winching. However, this would be characteristic to the local, studied conditions since other studies revealed that slope may affect the time consumption in winching operations as well as the distribution of time consumed by certain work elements within a work cycle (Borz et al. 2014b, Magagnotti and Spinelli 2012b). Skidder maneuvering and parking took less time than other elements as a result of the same local easy work conditions, while log hooking and unhooking accounted for almost the same share (Table 2, Fig. 3). While the empty travel took, on average, about 18 minutes in conditions of a mean distance of about 1764 m, in case of loaded travel about 23 minutes were necessary in conditions of a mean distance of about 1649 m (Table 2), because the loaded travel imposed a different driving behavior when compared with the empty travel. Furthermore, the field conditions (wet clayey soil) posed, frequently, adherence problems. Therefore, as shown by field results, the driving speed was 5.73 km/h in case of empty travel and 4.31 km/h in case of loaded travel, and these results are comparable with those reported by Spinelli et al. (2012) in case of a 40 kW Hittner Ecotrac V55 mini–skidder made in Croatia, and somehow in the range of those reported by Spinelli and Magagnotti (2012) in case of a number of farm tractors equipped for skidding operations. Also, Sabo and Poršinsky (2005) reported similar figures (5.33 and 3.99 km/h, respectively in case of unloaded and loaded travels), when analyzing the time consumption of a Timberjack 240C skidder used for extracting wood in selective harvests. Furthermore, they found out that travelling speed was also different when traveling on strip roads and landings, respectively. Within an on-road work cycle (including load attachment and detachent), loaded travel work element accounted for almost half of the consumed time (Table 2, Fig. 4). Also, an important amount of time was required by load attachment. This was related with several attachments/detachments performed on groups of logs until the load was finally attached and suspended in the rear part of the skidder. When introducing significantly simpler procedures, load detachment was the work element that required the smallest amount of time. It took about five minutes, on average, to perform landing operations that consisted of log manipulation and stacking (Table 2).

141


S. A. Borz et al.

Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

Table 2 Statistics of inputs/outputs and operational variables of skidding operations in the field study Factor Number of logs per load – NL 3

Volume per load – V, m

Sum

Mean

Minimum

Maximum

Standard Deviation

285

6.48

2

13

2.95

214.97*

4.89

3.04

6.63

0.71

Time consumption Maneuvering and parking – Tmp, s

1873.11

42.57

0.00

134.65

33.45

Cable releasing – Tcr, s

5005.48

113.76

24.39

334.71

69.12

Log hooking – Tlh, s

7737.03

175.84

48.12

442.27

97.70

10,724.59

243.74

44.31

641.16

139.55

5239.01

119.07

20.19

340.88

76.18

Empty travel – Tet, s

48,803.00

1109.16

852.00

1756.00

165.77

Load attachment – Tla, s

17,233.32

391.67

72.79

967.29

235.32

Loaded travel – Tlt, s

60,580.00

1376.82

1088.00

1711.00

143.75

Load detachment – Tld, s

4829.00

109.75

45.00

228.00

42.16

Landing operations – Tlo, s

13,534.00

307.59

62.00

1068.00

182.40

Winching cycle time – WCT, s

30,579.22

694.98

204.04

1519.49

338.27

On-road skidding cycle time – RSCT, s

131,445.32

2987.39

2425.00

3578.11

337.61

Overall skidding cycle time – OSCT, s

175,558.54

3989.97

2920.83

5228.78

625.12

Winching distance – WD, m

8.66

2.6

16.5

3.94

Empty travel distance – ETD, m

1764.07

1604.00

1971.00

109.50

Loaded travel distance – LTD, m

1648.55

1448.00

1879.00

96.60

Skidding distance – SD, m

1706.31

1597.00

1897.00

71.65

Mechanical winching – Tmw, s Log unhooking – Tlu, s

Process variables

*Total skidded volume

Although not effectively measured in the field, a characteristic of landing operations was that the maneuvers were deployed on distances up to 25 m, while stacking included only a few log movements using the front blade of the skidder. Within an entire skidding work cycle, the loaded travel work element also accounted for the greatest share (Fig. 5). Surprisingly, it took more time to attach the load than to perform landing operations, even if the latter involved several maneuvers and movements. This was related with a

relatively large number of logs within a load, as well as with the procedure of attaching the load. Most of the delays occurred due to operational and personal reasons – including meals taking about an hour each day (Fig. 6) – and represented about 29.3% of the total studied time. This fact reflects an improved operational time management, if compared with studies done for the same kind of skidding equipment (Borz et al. 2013, Borz et al. 2014a). The regression analysis conducted for a winching cycle time has revealed that

Table 3 General statistics for the empirical model developed for winching operations Model

R2

F

Sig.

WCT, s = 18.02 × WD + 94.78 × NL – 75.02

0.84

107.00

p<0.000

142

Independent variable

p

WD

0.003

NL

<0.000

Croat. j. for. eng. 36(2015)1


Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

S. A. Borz et al.

Fig. 3 Structure of elemental time consumptions (%) within a delayfree winching work cycle (Tamp: Time consumption for manoeuvering and parking, Tcr: Time consumption for cable releasing, Tlh: Time consumption for log hooking, Tmw: Time consumption for mechanical winching, Tlu: Time consumption for log unhooking)

Fig. 5 Structure of elemental time consumptions (%) within a delayfree on-road skidding work cycle (Tamp: Time consumption for maneuvering and parking, Tcr: Time consumption for cable releasing, Tlh: Time consumption for log hooking, Tmw: Time consumption for mechanical winching, Tlu: Time consumption for log unhooking, Tla: Time consumption for load attachment, Tet: Time consumption for empty travel, Tlt: Time consumption for loaded travel, Tld: Time consumption for load detachment, Tlo: Time consumption for landing operations)

Fig. 4 Structure of elemental time consumptions (%) within a delayfree on-road skidding work cycle (Tla: Time consumption for load attachment, Tet: Time consumption for empty travel, Tlt: Time consumption for loaded travel, Tld: Time consumption for load detachment)

Fig. 6 Delay time distribution by categories (MD: Mechanical delays, OD: Operational delays, PD: Personal delays)

Croat. j. for. eng. 36(2015)1

143


S. A. Borz et al.

Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

Table 4 Calculation of net and gross production rates for skidding operations Operation / Group of operations

Delay free time, s

Total time, s

Production, m3

Net production rate, m3/h

Winching

30,579.22

214.97

25.31

On-road skidding

131,445.32

214.97

5.89

Winching + Skidding (excluding landing operations)

162,024.54

Overall

175,558.74

248,202.14

214.97

4.78

214.97

4.41

Gross production rate*, m3/h

3.12

*Includes all the delays

the winching distance as well as the number of winched logs significantly affected the time consumption. These independent variables were significant at the chosen confidence level p < 0.05 as shown in Table 3. Given the particularities of the study (empty and loaded travels on different skid roads), as well as the fact that, for these work elements, only the distance was recorded in the field, we have developed time consumption models for each element by considering only this independent variable. The resulted models are presented in Equations 1 and 2.

Tet ,s = 0.428 ⋅ ETD ,m + 354.04

(1)

Tlt ,s = 0.314 ⋅ ETD ,m + 858.64

(2)

In case of winching operations, the study yielded a net production rate of 25.31 m3/h, which seems to be fairly acceptable for the operational conditions presented in Table 2, if compared with other results reported for such equipment (Borz et al. 2013, Borz et al. 2014a). However, the net production rates for on-road skidding of only 5.89 m3/h (Table 4), as well as the net production rate of only 4.71 m3/h for winching and skidding (excluding landing operations) were strongly affected by the increased skidding distances. Of course, this is also correlated with a more reduced load per turn, if compared with results reported in different conditions for the same equipment (Borz et al. 2013, Borz et al. 2014a). Furthermore, in the case of gross production rate, the situation was even worse, since our study yielded only 3.12 m3/h. If compared with the results by other studies, our results regarding the net production rate are similar with those reported by Spinelli et al. (2012) when using the Hittner Ecotrack mini-skidder made in Croatian in conditions of a mean skidding distance of only 130 m. This fact also emphasizes the effect of long skidding distance in our study. Also, even if the mean skidding distance was a little bit unclear, Ghaffaryian and Shobani (2008) reported production rates in the range of 5.93 to 8.33 m3/h, when

144

analyzing a predecessor of the TAF skidder, which was considered in our study. On the other hand, Spinelli and Magagnotti (2012) obtained similar results when studying a 96 kW farm tractor in conditions of a mean skidding distance of 1119 m. In their study, the mean load size was smaller if compared with the load presented in this study, but they found out that empty and loaded travel speeds were of 8.1 and 7.3 km/h, respectively, which were significantly higher than those determined in this study. The results presented above have several implications. Although our intention was not to calculate the harvesting operation costs, it is quite obvious that the reduced accessibility conditions have negative repercussions over both, the harvesting company and the workers’ paychecks, since the latter are paid based on the realized production. On the other hand, in Romania the stands located up to 2 km from a permanent transportation infrastructure are considered accessible. However, when using skidders, and, especially in steep terrain, bladed roads are frequently required (Oprea 2008). When taking into consideration that the external skid roads are required for harvesting a 5 hectare felling area (this study case), it is obvious that due to erosion processes the environmental impact will also be greater. This is of great importance in hilly forests of Romania, usually located on clayey soils, and otherwise very sensitive to the combined action of skidding and moisture (Oprea 2008). Also, the time consumption models developed in this study may have their limits and should be used carefully by considering only the conditions similar to those observed in this study (Table 2).

4. Conclusion Thinking globally, timber skidding is a highly variable operation, with many process variables affecting the relations between inputs and outputs. The aim of this paper was not to identify new predictors or to model new unknown relations, but to emphasize and quanCroat. j. for. eng. 36(2015)1


Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

tify how and to what extent, the reduced accessibility conditions (in terms of extraction distance) may affect the time consumption and operational productivity in skidding operations. As shown by the results, our study yielded very low production rates, which can be particulary related to the increased skidding distance. Also, the conditions imposed by a cleaner extraction (such as limiting the log lengths during skidding) affects the loading capacity of the skidder, which in combination with very long extraction distances leads to reduced productivities, increased number of turns in order to extract the same quantity of wood and increased energy expenditure per unit of product.

Acknowledgements We hereby acknowledge the structural funds project PRO–DD (POS–CCE, O.2.2.1., ID 123, SMIS 2637, ctr. No. 11/2009) for providing the infrastructure used in this work and the South East Europe Transnational Cooperation Programme, an EU program for financially supporting this research within the project »FOROPA – Sustainable Networks for the Energetic Use of Lignocellulosic Biomass in South East Europe«. Also, the authors would like to thank the State Forest District of Sighişoara for making possible this research and to Eng. Alina Neagu, Eng. Elena Catişov, Eng. Vişan Laurenţiu and Eng. Andrei Apăfăian for their active collaboration during field and office phases of the study.

5. References Bereziuc, R., Alexandru, V., Ciobanu, V., Ignea, G., Abrudan, I., Derczeni, R., 2006: Ghid pentru proiectarea, construirea şi întreţinerea drumurilor forestiere (Handbook on designing, building and maintaining forest roads). Editura Universităţii Transilvania din Braşov. 296 p. Björheden, R., Apel, K., Shiba, H., Thompson, M. A. 1995: IUFRO Forest work study nomenclature, Swedish University of Agricultural Science, Dept. of Operational Efficiency, Grapenberg, 16 p. Borz, S.A., Ignea G., Popa, B., 2014a: Assessing timber skidding efficiency in a group shelterwood system applied to a fir-beech stand. African Journal of Agricultural Research 9(1): 160–167. Borz, S.A., Ignea, G., Popa, B., 2014b: Modeling and comparing timber winching performance in windthrow and uniform selective cuttings for two Romanian skidders. Journal of Forest Research 19(6): 473–482. Borz, S.A., Ignea G., Popa, B., Dinulică, F., 2013: Time consumption and productivity of skidding Silver fir (Abies alba Mill.) round wood in reduced accessibility conditions: a case Croat. j. for. eng. 36(2015)1

S. A. Borz et al.

study in windthrow salvage logging from Romanian Carpathians. Annals of Forest Research 56(2): 363–375. Borz, S.A., Ciobanu, V. D., 2013: Efficiency of motor–manual and horse logging in small scale firewood production. African Journal of Agricultural Research 8(24): 3126–3135. Borz, S.A., Bîrda, M., Ignea, G., Oprea, I., 2011: Technological aspects regarding timber exploitation using Mounty 4100 cable yarder. Bulletin of the Transilvania University of Braşov 4(53): 1–6. Gallis, C., Spyroglou, G., 2012: Productivity linear regression models of tree–length harvesting system in natural coastal Aleppo Pine (Pinus halepensis L.) forests in the Chalkidiki area of Greece. Croatian Journal of Forest Engineering 33(1): 115–123. Ghaffariyan, M.R., Naghdi, R., Ghajar, I., Nikooy, M., 2013: Time Prediction Models and Cost Evaluation of Cut-ToLength (CTL) Harvesting Method in a Mountainous Forest. Small Scale Forestry 12(2): 181–192. Ghaffariyan, M.R., Shobani, H., 2008: Optimum road spacing of ground based skidding operations in Nowshahr, Iran. Caspian Journal of Environmental Sciences 6(2): 105–112. Jourgholami, M., Majnouinan, B., Zargham, N., 2013: Performance, capability and costs of motor–manual tree felling in Hyrcanian hardwood forest. Croatian Journal of Forest Engineering 34(2): 283–293. Magagnotti, N., Spinelli, R., Spinelli, R. (Eds.), 2012a: Good Practice Guideline for Biomass Production Studies – COST Action FP0902. CNR IVALSA, Florence, Italy. ISBN978-88901660-4-4. 41 p. Available at: www.forestenergy.org. Magagnotti, N., Spinelli, R., 2012b: Replacing Steel Cable with Synthetic Rope to Reduce Operator Workload During Log Winching Operations. Small Scale Forestry 11(2): 223–236. Magagnotti, N., Spinelli, R., 2011: Integrating animal and mechanical operations in protected areas. Croatian Journal of Forest Engineering 32(2): 489–499. Naghdi, R., Limaei, S.M., 2009: Optimal forest road density based on skidding and road construction costs in Iranian Caspian forests. Caspian Journal of Environmental Sciences 7(2): 79–86. Olteanu, N., 2008: Drumuri forestiere. Volumul I: Proiectarea drumurilor foresteire. Reţele de drumuri forestiere (Forest roads. Volume I: Designing forest roads. Forest roads networks) Editura Universităţii Transilvania din Braşov. 339 p. Oprea, I., 2008: Tehnologia exploatării lemnului (Timber harvesting technology). Transilvania University Press Braşov. 273 p. Özturk, T., 2010a: Productivity of New Holland farm tractor at beech stands on mountainous areas in Black Sea region. Forestry Ideas 16(1): 52–57. Özturk, T., 2010b: Productivity of MB Track 900 tractor at beech stands on mountainous areas in Black Sea region. African Journal of Agricultural Research 5(1): 28–33. Picchio, R., Maesano, M., Savelli, S., Marchi, E., 2009: Productivity and energy balance in conversion of a Quercus cer-

145


S. A. Borz et al.

Estimating Time Consumption and Productivity of Roundwood Skidding in Group ... (137–146)

ris L. coppice stand into high forest in Central Italy. Croatian Journal of Forest Engineering 30(1): 15–26. Sabo, A., Poršinsky, T., 2005: Skidding of fir roundwood by Timberjack 240C from selective forests of Gorski Kotar. Croatian Journal of Forest Engineering 26(1): 13–27. Sbera, I., 2007: Resursele de lemn şi potenţialul pieţei din România (Wood resources and market potential in Romania). Meridiane Forestiere 2: 3–7.

Spinelli, R., Magagnotti, N., 2012: Wood Extraction with Farm Tractor and Sulky: Estimating Productivity, Cost and Energy Consumption. Small Scale Forestry 11(1): 73–85. Visser, R., Spinelli, R., 2012: Determining the shape of the productivity function for mechanized felling and felling– processing. Journal of Forest Research 17(5): 397–402.

Şofletea, N., Curtu, L., 2008: Dendrologie (Dendrology). Editura Pentru Viaţă. Braşov. 418 p.

Vusić, D., Šušnjar, M., Marchi, E., Spina, R., Zečić, Z., Picchio, R., 2013: Skidding operations in thinning and shelterwood cut of mixed stands – work productivity, energy inputs and emissions. Ecological Engineering 61: 216–223.

Spinelli, R., Magagnotti, N., Relaño, R.L., 2012: An alternative skidding technology to the current used crawler tractors in Alpine logging operations. Journal of Cleaner Production 31: 73–79.

Zečić, Z., Krpan, A.P.B., Vukušić, S., 2005: Productivity of C Holder 870F tractor with double drum winch Ingland 4002 in thinning beech stands. Croatian Journal of Forest Engineering 26(2): 49–57.

Authors’ address: Assist. Prof. Stelian Alexandru Borz, PhD.* e-mail: stelian.borz@unitbv.ro Prof. Gheorghe Ignea, PhD. e-mail: igneagh@unitbv.ro Assist. Prof. Bogdan Popa, PhD. e-mail: popa03@gmail.com Assist. Prof. Eugen Iordache, PhD. e-mail: i.eugen@unitbv.ro Transilvania University of Braşov Faculty of Forestry Department of Forest Engineering Şirul Beethoven No. 1 500 123 Braşov ROMANIA

Received: April 30, 2014 Accepted: July 3, 2014

146

Prof. Gheorghe Spârchez, PhD. e-mail: sparchez@unitbv.ro Transilvania University of Braşov Faculty of Forestry Department of Silviculture Şirul Beethoven No. 1 500 123 Braşov ROMANIA * Corresponding author Croat. j. for. eng. 36(2015)1













ISSN 1845-5719

9 771845 571000


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.