Bioresource Technology 129 (2013) 219–223
Contents lists available at SciVerse ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Effect of organic loading rate on anaerobic digestion of thermally pretreated Scenedesmus sp. biomass C. González-Fernández a,⇑, B. Sialve b, N. Bernet a, J.P. Steyer a a b
INRA, UR50, Laboratoire de Biotechnologie de L’Environnement, Avenue des Etangs, Narbonne 11100, France Naskeo Environnement, Avenue des Etangs, Narbonne 11100, France
h i g h l i g h t s " Increasing the organic loading rate did not affect anaerobic digestion performance. " Scenedesmus sp. biofuel production is hindered by its cell wall. " Storage time of Scenedesmus sp. biomass affected organic matter hydrolysis. " Scenedesmus sp. thermal pretreatment was beneficial for anaerobic digestion.
a r t i c l e
i n f o
Article history: Received 20 March 2012 Received in revised form 3 October 2012 Accepted 8 October 2012 Available online 2 November 2012 Keywords: Microalgae Scenedesmus sp Pretreatment Anaerobic digestion
a b s t r a c t Biogas production is one of the means to produce a biofuel from microalgae. Biomass consisting mainly of Scenedesmus sp. was thermally pretreated and optimum pretreatment length (1 h) and temperature (90 °C) was selected. Different chemical composition among batches stored at 4 °C for different lengths of time resulted in organic matter hydrolysis percentages ranging from 3% to 7%. The lower percentages were attributed to cell wall thickening observed during storage for 45 days. The different hydrolysis percentages did not cause differences in anaerobic digestion. Pretreatment of Scenedesmus sp. at 90 °C for 1 h increased methane production 2.9 and 3.4-fold at organic loading rates (OLR) of 1 and 2.5 kg COD m 3 day 1, respectively. Regardless the OLR, inhibition caused by organic overloading or ammonia toxicity were not detected. Despite enhanced methane production, anaerobic biodegradability of this biomass remained low (32%). Therefore, this microalga is not a suitable feedstock for biogas production unless a more suitable pretreatment can be found. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Microalgae are a promising feedstock for the production of biodiesel and bioethanol; however, the generation of these biofuels only extracts part of the energy stored in the biomass. According to Lardon et al. (2009), oil cakes discarded after biodiesel production still contained 35–73% of the energy present in the biomass. Therefore, anaerobic digestion of these cakes would decrease costs by providing heat and electricity. Likewise, bioethanol production only ferments the carbohydrate fraction while lipids and proteins can still be used for other purposes. Production of biogas requires pretreatment of the biomass in order to disrupt the cell wall of the microalgae (González-Fernández et al., 2012a). In the present study, thermal pretreatment was evaluated for enhancing anaerobic biodegradation. Methane production at organic loading rates of 1 and 2.5 kg COD m 3 day 1, ⇑ Corresponding author. E-mail address: cgonfer@iq.uva.es (C. González-Fernández). 0960-8524/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2012.10.123
was applied to a semi-continuously fed reactor. In addition to organic matter degradation, nitrogen mineralization and potential inhibitions by organic overloading and free ammonia were assessed. 2. Methods 2.1. Feedstock and inocula Biomass containing mainly Scenedesmus sp. microalgae was grown at 20 °C in synthetic Z-8 nutrient broth (Staub, 1961). Continuous illumination with fluorescent lamps (Edison, 58 W) was employed. The photobioreactor consisted of an open plastic bag of 0.2 m3. Air flow of 12 L min 1 was provided for mixing. When the absorbance (680/800 nm) of the culture reached a plateau (after approximately two weeks), the biomass was allowed to settle by gravity. The inoculum for anaerobic digestion was granular sludge collected at a sugar factory in Marseille (France). When operating
C. González-Fernández et al. / Bioresource Technology 129 (2013) 219–223
the semi-continuous fed reactor, anaerobic biomass was collected from a reactor digesting Chlorella vulgaris (Ras et al., 2010). 2.2. Thermal pretreatment The microalgae suspension was adjusted to 14 g tCOD L 1 and 0.6 L were poured into a 1-L reactor. In order to fully use the energy supplied by thermal treatment, pressure build-up (1.2 bar) was allowed by sealing the reactor. The pretreatment temperatures of 70, 80 and 90 °C were achieved within 15 min by pumping appropriately heated water into the reactor jacket. The elapsed time was measured from the moment that the biomass started to be heated. Biomass was continuously stirred during the thermal treatment to avoid temperature gradients. Thermal pretreatment lasted for a total of 3 h. Samples were taken every 15 min for the first hour and every 30 min for the following 4 h. Biomass solubilization was calculated as: [(sCOD sCODo)/(tCODo sCODo)] 100 where sCOD refers to soluble chemical oxygen demand, tCOD refers to total oxygen demand and the subscript ‘‘o’’ stands for ‘‘before pretreatment’’. 2.3. Biomethane production Tests were run in a room heated to 35 °C in 0.5-L glass bottles over a two-months period. Batch tests were carried out in duplicates for each sample and blanks. For the determination of endogenous methane production, blanks containing only anaerobic sludge were run. Microalgae biomass and anaerobic sludge were mixed to attain a ratio tCOD substrate/VS inoculum of 0.5. Quantities were calculated to obtain a final liquid volume of 0.3 L. Calcium carbonate was added as buffering agent to a final concentration of 50 g L 1. The pH ranged from 7–8 in all the reactors before the test started. Oxygen was removed by flushing the headspaces with nitrogen for 10 min. The volume of biogas produced was calculated by measuring the pressure of the bottle headspace. The substrates tested were fresh microalgae harvested from the photobioreactor and allowed to settle in a room at 4 °C. In this manner, biomass was stored in the cold room for 15, 45 and 180 days. 2.4. Semi-continuous digestion A continuous stirred-tank reactor (CSTR) with a working volume of 1 L with magnetic agitation was used for anaerobic digestion. In order to prevent any photosynthetic activity, the digester was covered with aluminium foil. The reactor temperature was maintained at 35 °C by circulating water through a water jacket. The pH was monitored by an in situ probe. Biogas was measured by water displacement and composition was analyzed periodically. The hydraulic retention time (HRT) was 15 days. The continuous stirred-tank reactor was started up using fresh microalgae biomass as substrate. After the reactor reached steady state at organic loading rate (OLR) of 1 kg tCOD m 3 reactor day 1, the reactor was fed with the thermally pretreated biomass. Steady state was characterized by a stable gas production and COD concentration (3 cycles of HRT). Likewise, two OLR were evaluated when the reactor was fed with thermally pretreated biomass diluted with distilled water to 1 and 2.5 kg COD m 3 reactor day 1, respectively. 2.5. Analyses
soluble phase with a gas chromatograph (GC-8000 Fisons instrument) equipped with a flame ionization detector. Ammonium (NHþ 4 ) was analyzed in the soluble phase by ion chromatography system (DIONEX 100) using conductivity detection. Biogas composition was determined by gas chromatography (Shimadzu GC-8A) equipped with a thermal conductivity detector. 3. Results and discussion 3.1. Pretreatment: length and temperature selection Fig. 1 shows the sCOD released upon pretreatments at 70, 80 and 90 °C. sCOD increased with time. The same pattern in terms of organic matter hydrolysis was achieved with treatments at 70 and 80 °C. Maximum COD hydrolysis yield at these temperatures was approximately 6% of initial tCOD. Thermal treatment at 90 °C achieved this value within the first 15 min of heating. Increasing the length of the pretreatment to up to 4 h led to a hydrolysis yield of 8%. A temperature of 90 °C is sufficient to break hydrogen bonds linking cellulose and hemicellulose in the cell wall of the microalgae (Laureano-Perez et al., 2005), and to cause damage to the cell wall (González-Fernández et al., 2012b). Therefore, treatment at 90 °C for 1 h was selected for biomass pretreatments. 3.2. Effect of thermal pretreatment on solubilization of organic matter The characterization of fresh and thermally pretreated biomass (90° for 1 h in a closed system) is shown in Table 1. Batches collected from the photobioreactor exhibited more or less the same content of TS with VS accounting for around 95% of the TS. In contrast, the tCOD/TKN and sCOD/tCOD ratios were slightly different among batches. Despite these differences, tCOD hydrolysis ranged 6–7%. Surprisingly, batches 4 and 5 used for feeding the reactor at 2.5 kg COD m 3 day 1 exhibited a lower hydrolysis percentage. More specifically, organic matter hydrolysis decreased to 3–4%. The difference among batches employed for feeding the reactor at 1 and at 2.5 kg COD m 3 day 1 was storage time. 3.3. Effect of storage length effect on biomass digestion During OLR of 1 and 2.5 kg COD m 3 day 1, the collected biomass was used after approximately 15 and 45 days of storage, respectively. Conditions such as low-N culture medium (van Donk et al., 1997) or low irradiance (van Donk and Hessen, 1995) have been described to cause cell wall thickening. In order to confirm whether or not cell wall thickened during storage occurred (and hence decreasing organic matter hydrolysis), biomethane production assays were carried out. Three different biomasses stored for 15, 45 and 180 days for were tested. After 63 days of 1400 1200 sCOD (mg L-1)
220
1000 800 600
70 C
400
80 C
90 C
200
Total solids (TS), volatile solids (VS) and total Kjeldahl nitrogen (TKN) were measured according to Standard Methods (Eaton et al., 2005). Chemical oxygen demand (COD) was analyzed by a colorimetric method using Hach 0–1500 mg L 1 vials. Soluble fractions (sCOD) were obtained after centrifugation (mini-spin Eppendorf, 15 min, 12,100g). Volatile fatty acids (VFA) were measured in the
0 0
50
100
150
200
250
300
Elapsed Time (min) Fig. 1. sCOD concentrations over time of microalgae biomass treated at 70, 80 and 90 °C.
221
C. González-Fernández et al. / Bioresource Technology 129 (2013) 219–223
Table 1 Chemical characterization of different batches employed for the feeding of the CSTR for the two tested OLR, 1 and 2.5 kg COD m 3 day 1. F and T stand for fresh and treated batches. (Average ± standard deviation). OLR of 1 kg COD m 3 day 1 Batch F1 TS (g L 1) VS (g L 1) tCOD (mg L 1) sCOD sol (mg L 1) TKN (mg L 1) 1 NHþ 4 (mg L )
23.8 21.9 41950 153 2394.7 27.4
Batch T1 ±0.5 ±0.4 ±707 ±6.4 ±17.3 ±2.7
23.8 21.5 421300 2908 2395.7 318.4
Batch F2 ±0.4 ±0.2 ±300 ±5.7 ±33.6 ±4.2
23.2 21.9 37125 147 4495.6 83.6
Batch T2 ±0.2 ±0.4 ±35.4 ±15.6 ±7.3 ±12.4
23.0 21.8 37320 2834 4502 105.2
Batch F3 ±0.2 ±0.1 ±267 ±59.4 ±5.4 ±2.3
20.5 19.1 34225 283.5 2681.1 137.3
Batch T3 ±0.3 ±0.2 ±672 ±2.1 ±82.7 ±1.0
20.0 18.6 37250 3548 2375.7 590.2
±0.1 ±0.1 ±1555 ±17 ±38.5 ±7.8
OLR of 2.5 kg COD m 3 day 1 Batch F4 TS (g L 1) VS (g L 1) tCOD (mg L 1) sCOD sol (mg L 1) TKN (mg L 1) 1 NHþ 4 (mg L )
21.2 20.8 47975 908.5 2466.4 74
Batch T4 ±0.3 ±0.2 ±35.4 ±0.7 ±11.2 ±3.7
21.7 20.7 48150 2678 2436.9 124.5
Batch F5 ±0.2 ±0.2 ±353.6 ±53.7 ±37.8 ±6.4
23.9 23.0 38550 443 3382.5 80.2
Table 2 Chemical characterization of Scenedesmus sp. biomass stored for different periods. (Average ± standard deviation.) Storage time
TS (g L 1) VS (g L 1) tCOD (mg L 1) sCOD (mg L 1) TKN (mg L 1) 1 NHþ 4 (mg L )
15 days
45 days
180 days
8.2 ± 0.2 7.5 ± 0.1 13140 ± 226 74 ± 14 1237 ± 32 2±1
28.4 ± 0.3 25.9 ± 0.2 47428 ± 35 254 ± 37 3034 ± 90 76 ± 3
23.8 ± 0.2 21.9 ± 0.3 37300 ± 495 147 ± 16 2395 ± 18 50 ± 1
digestion, similar anaerobic biodegradability (32–34%) was achieved with the biomass stored for 45 and 180 days. Methane production was 25% higher when the biomass was stored for only 15 days. The methane content of the biogas averaged 69% regardless of storage time, but the methane production rate profile was different. For biomass stored for 45 and 180 days, the methane production rate started at around 3–4 mL CH4 day 1 and decreased with time. The opposite tendency was observed for methane production rate of biomass stored for 15 days. The methane production rate increased from 2 to 4.4 mL CH4 day 1 during the first 7 days of digestion and decreased afterwards, but remained above 2.5 mL CH4 day 1 until day 16. At digestion day 16, the methane yields were 79, 62 and 49 mL CH4 g tCOD initial 1 for biomass stored for 15, 45 and 180 days, respectively. These differences are corroborating the hypothesis of the development of a stronger cell wall during storage. While the sCOD/tCOD ratio ranged from 0.4% to 0.5% for all the tested biomasses tested, the tCOD/TKN ratio was around 10 for biomass stored for 15 days and 15 for biomass stored for 45 and 180 days (Table 2). Likewise, the NHþ 4 /TKN ratio of biomass stored for 15 days was quite low when compared to that of biomass stored for longer times, likely due to protein degradation during storage. Furthermore, the protein content (assumed to be 6.5 TKN according to Wilkie and Mulbry, 2002) was almost double in biomass stored for 15 days than in the other two biomass samples tested. Thus, bio-chemical reactions are conferring a higher resistance to the cell wall during long storage time.
3.4. Semi-continuous anaerobic digestion 3.4.1. Organic matter fate Anaerobic digestion of untreated Scenedesmus sp. biomass under semi-continuous conditions resulted in low methane
Batch T5 ±0.2 ±0.2 ±282.8 ±2.8 ±194.4 ±7.4
24.0 23.1 38780 2326 3382.5 186.6
±0.3 ±0.2 ±175.8 ±19.8 ±194.4 ±1.3
production. When feeding at an OLR of 1 kg tCOD m 3 day 1, the methane yield was around 33 mL CH4 g tCOD initial 1 while the methane content in the biogas ranged 63–68%. Therefore, 90% total tCOD fed to the reactor was still present in the effluent after anaerobic digestion. This fact proved once again the high resistance of Scenedesmus sp. to anaerobic bacterial attack. Previous research on anaerobic digestion of this biomass also showed low methane production even at higher HRTs (González-Fernández et al., 2012b). The presence of sporopollenin and algaenan polymers might have contributed to the low digestibility (González-Fernández et al., 2012a). When the thermally pretreated microalgae were digested in the semi-continuous system at higher OLR, methane production was increased from 96.9 mL CH4 g tCOD initial 1 registered at 1 kg tCOD m 3 day 1 to 111.4 mL CH4 g tCOD initial 1 at 2.5 kg tCOD m 3 day 1. Moreover, tCOD and sCOD removals were 30–44% and 83–87%, respectively (Fig. 2). When compared to untreated biomass, the methane yield increased 2.9- and 3.4-fold at OLR of 1 and 2.5 kg tCOD m 3 day 1, respectively. The higher methane yield exhibited during the highest OLR denotes an underestimated activity of anaerobic microorganism at the lowest OLR. The different percentage of hydrolysis attained during thermal pretreatment for the different batches (6–7% and 3–4% for the OLR of 1 and 2.5 kg tCOD m 3 day 1, respectively) did not have a detrimental effect on anaerobic digestion. No direct link between organic matter solubilization during thermal pretreatment and methane production can be withdrawn. Nevertheless, weak hydrogen bonds broken are likely responsible for the enhancement in methane production. VFAs were also measured in order to evaluate organic overloading of the process. VFA levels were maintained at 10–15 mg COD (VFA) L 1 throughout the experiment. The generally stable VFA and pH (7.9–8.1) levels indicate that organic overload did not occur. 3.4.2. Nitrogen fate Anaerobic digestion causes not only COD removal but also organic nitrogen mineralization. Scenedesmus biomass presents a high content of proteins (González-Fernández et al., 2010) which can hamper anaerobic digestion by releasing a high amount of ammonium. High ammonium released during organic nitrogen break down together with high pH can mediate methane production inhibition by high free ammonia concentration. Effluent ammonium concentrations averaged 550 mg L 1 for OLR of 1 kg tCOD m 3 day 1 and 700 mg L 1 for OLR of 2.5 kg tCOD m 3 day 1 (Fig. 2).
222
C. González-Fernández et al. / Bioresource Technology 129 (2013) 219–223
OLR 1 Kg COD m-3d-1
OLR 2.5 Kg COD m-3d-1
40000
3000
35000
2500 2000
25000
1500
20000 15000
1000
sCOD (mg L-1)
tCOD (mg L-1)
30000
10000 500
5000
0
0 0
20
40
60
80
100
120
Digestion Time (days) OLR 1 Kg COD m-3d-1
3500
OLR 2.5 Kg COD m-3d-1
TKN/NH4+ (mgL-1)
3000 2500 2000 1500 1000 500 0 0
20
40
60 80 Digestion Time (days)
100
120
Fig. 2. Inlet (closed symbols) and outlet (open symbols) tCOD (circles), sCOD (triangles), TKN (circles) and NHþ 4 (triangles) concentrations obtained for the CSTR at OLRs of 1 and 2.5 kg COD m 3 day 1.
Table 3 TS and VS% removal, methane yield, gas quality, % of N mineralized and free ammonia concentration obtained for the CSTR digesting thermally pretreated Scenedesmus sp. biomass at OLRs of 1 and 2.5 kg COD m 3 day 1.
TS% reduction VS% reduction Specific CH4 yield (mL CH4 g COD initial 1) CH4 % in biogas % N mineralization Free NH3 (mg L 1)
OLR 1 kg COD m 3 day 1
OLR 2.5 kg COD m 3 day 1
26.2 30.2 96.9
34.3 36.1 111.4
75 43.5 48.2
72 34.1 68.9
The free ammonia concentrations were 48.2 and 68.9 mg L 1 for 1 and 2.5 kg tCOD m 3 day 1, respectively (Table 3). It is therefore unlikely that ammonia inhibition occurred since an ammonia concentration of 1.1 g L 1 previously measured during digestion of microalgae biomass did not inhibit methane production (González-Fernández et al., 2011). A second important parameter with regard to the nitrogen fate is nitrogen mineralization. This value gives an idea of how much ammonium could be recycled to the culture medium for microalgae growth. With thermally pretreated biomass, nitrogen mineralization was 43.5% and 34.1% for 1 and 2.5 kg tCOD m 3 day 1, respectively (Table 3). In comparison, 40–48% nitrogen mineralization was reported after 40 days of digestion of microalgae biomass (González-Fernández et al., 2011). This higher value could be attributed to the longer retention time of the substrate in the reactor.
4. Conclusion Hydrolysis of biomass consisting mainly of Scenedesmus sp. was affected by prolonged storage. Thermal pretreatment at 90 °C was selected to improve methane production of Scenedesmus sp. biomass. Feeding the reactor with thermally pretreated biomass at OLR of 1 kg tCOD m 3 day 1 improved methane production by 3-fold. Raising OLR to 2.5 kg tCOD m 3 day 1 did not cause organic overloading or ammonia toxicity. Despite the pretreatment, Scenedesmus sp. was refractory to anaerobic degradation. For biofuel purposes, this investigation suggests that Scenedesmus is not a suitable microalga unless another type of cost-effective pretreatment can be found. Acknowledgement This research was financially supported by the French National Research Agency for the Symbiose project (ANR-08-BIO-E11). References Lardon, L., Hélias, A., Sialve, B., Steyer, J.P., Bernard, O., 2009. Life-cycle assessment of biodiesel production from microalgae. Environ. Sci. Technol. 43, 6475–6481. González-Fernández, C., Sialve, B., Bernet, N., Steyer, J.P., 2012a. Impact of microalgae characteristics on their conversion to biofuel. Part II: Focus on biomethane production. Biofuels, Bioprod. Biorefin. 6, 205–218. Staub, R., 1961. Ernährungsphysiologisch-autökologische Untersuchungen an Oscillatoria rubescens. Z. Hydrologia 82, 198 (D.C.-Schweiz). Ras, M., Lardon, L., Sialve, B., Bernet, N., Steyer, J.P., 2010. Experimental study on a coupled process of production and anaerobic digestion of Chlorella vulgaris. Bioresour. Technol. 102, 200–206. Eaton, A.D., Clesceri, L.S., Greenberg, A.E., 2005. Standard Methods for the Examination of Water and Wastewater, 21st ed. American Public Health
C. González-Fernández et al. / Bioresource Technology 129 (2013) 219–223 Association/American Water Works Association/Water Environment Federation, Washington DC, USA. Laureano-Perez, L., Teymouri, F., Alizadeh, H., Dale, B.E., 2005. Understanding factors that limit enzymatic hydrolysis of biomass. Appl. Biochem. Biotechnol. 124, 1081–1099. González-Fernández, C., Sialve, B., Bernet, N., Steyer, J.P., 2012b. Thermal pretreatment to improve methane production of Scenedesmus biomass. Biomass Bioenenergy 40, 105–111. van Donk, E., Lurling, M., Hessen, D.O., Lokhorst, G.M., 1997. Altered cell wall morphology in nutrient-deficient phytoplankton and its impact on grazers. Limnol. Oceanogr. 42, 357–364.
223
van Donk, E., Hessen, D.O., 1995. Reduced digestibility of UV-B stressed and nutrient-limited algae by Daphnia magna. Hydrobiologia 307, 147–151. Wilkie, A.C., Mulbry, W.W., 2002. Recovery of dairy manure nutrients by benthic freshwater algae. Bioresour. Technol. 84, 81–91. González-Fernández, C., Molinuevo-Salces, B., García-González, M.C., 2010. Open and enclosed photobioreactors comparison in terms of organic matter utilization, biomass chemical profile and photosynthetic efficiency. Ecol. Eng. 36, 1497–1952. González-Fernández, C., Molinuevo-Salces, B., García-González, M.C., 2011. Evaluation of anaerobic codigestion of microalgal biomass and swine manure via response surface methodology. Appl. Energy 88, 3448–3453.