Oil-well Casing Corrosiont R H GOODNIGHT*AND J P BARRETT*
ABSTRACT Information 1s presented to outl~nethe major causes of both ~nternaland external corroslon and the poss~ble damage that could occur as a result of thrs act1011 The use of a volat~lelnhlb~torto control ~nternalcorroslon 1s proposed In those cases where ~t 1s ~mpract~cal to use a packer to seal the tub~ng-cas~ng aiinulus The use of ~nternallycoated caslng may become effect~vewith the development of niore su~table coat~ng materials and methods of appl~cation.Control of external corros.on on well caslng caused by bacterial a c t ~ v ~ t yelectrolyt~c . effects, or a c ~ dwater attack, by cathod~cproteet~on1s proposcd Protect~on from external corroslon on new wells by the use of coat~ngsplus cathod~c~ ~ r o t e - t ~1s on suggested.
INTRODUCTION There have been widely separated views on the matter of corrosion prevention 3s l~ractlced by tlle pipeline and procluction companies Pipeline pzoi11e generally coat their lines, wrap them, and even give them extra mechanical protection in rocky country Cathocl~c-protect1011systems are installeel and frequent inspections are made to clctei.n~ineif corrosion is being controlled All of this is on lines which can be dug u p and repaired a t a relatively low cost On the other hand, production people have installeel thousands of nliles of bare vertical pipel ~ n e s No protection llas been supplied against external corrosion, and in illany cases, no measures are taken to prevent internal chen~ical attacli Casing has failed rapldly in illany areas a s a result of either internal 01 external corrosion Casing strings in a nunlber of fields have been perforated by internal attack in a s little a s six years, whereas in other fields, fallures have resulted fro111 external attack in a like p?rlod. Typically, a number of wells in the same fielcl are affected Repairs of caslng failures are extremely costly and tests for detesmlning corrosion of casing a r e not well-clevelopecl The 011 ~ndustry'sinvestlnent in oil-well casing is well over $5 billion The yearly increase in this \
*Pan Amerlcan Petroleum Carp. Tulsa, Okla ?Presented at the spring meetlng of the Southweztern D.str,et, D ~ v l s ~ oof n P r u d ~ . c t l o n . Fort Worth. Texas, March 1956
investment is over $85 million The cost of corroslon-mitigation measures .becomes relatively cheap when one considers t h a t uncontrollecl casing corrosion results not only in the loss of the casing but also in tlle interruption of procluction, and map-result in loss of the well, or even llernlanent damage to the reservoir.
INTERNAL CORROSION Although internal corroslon of the casing surfaces can be caused by carbon dioxide, l~ydrogen sulfide, and organic acids known to be responsible for tubing corrosion, production techniques i~lodifythe opportunity for attack In general, the nlain cause of lnternal attack is hydrogen sulfide The general mechanism f o r this attack may be expressecl a s follows H,S + Fe + FeS\ + 2H ~Glro~en Sulfide + Iron (yields in the Iron Sulfides presence of and water) Hyclrogen As this react1011is primarily one of acid attack, carbon dioxide, when present, accelerates the reaction by increasing the total acid present. The iron sulfide forilled sets up a galvanic cell in which tlle steel pipe becomes the anode This reaction is generally assumed to be responsible for the deep irregular pitting observed in sulfide COrroSlo11 Fig 1shows a sectlon of casing renloved from a west Texas sour well after only six years' exposure It is significant t h a t although the Inside of the casing was severely corroded, tlle external surfaces of tlle tubing were virtually unattacked This specinlen was in the upper part of the string above the fluid level. The mechanisnz of attack by hydrogen sulfide is a s follows. In the absence of a paclter, the casing-tubing annulus is exposed to the reservoir gases, which in this case contain hydrogen sulfide. The gas is saturated with water a t reservoir conditioils As tlle gas diffuses up the hole, t h e water condenses on the surface of the casing a t
344 - -
R. H. GOODNIGHT AND J. P. BARRETT
of lnternal Surface of Casing by Hydrogen Sulfide
a r e a s which a r e cooled below the dew point. Hydrogen sulfide then dissolves in t h e water droplet and corrosion is initiated. Because there is little or no flow through t h e annulus, the iron sulfide scale is not r e m ~ e dand galvanic action caused by this scale accelerates t h e corrosion.
Normally, the external surface of the tubing will show the effects of hydrogen sulfide corsosion. However, as in the foregoing case, there a r e exceptions. T h e lack of corrosion on t h e tubing is believed to be a result of its higher temperature a s compared to t h a t of the casing. The flow of production through t h e tubing can raise t h e temperature of t h e tubing surfaces above the dew point of the gas in the annulus. If this happens, no condensation of water will occur; and, as a result, t h e external tubing surfaces will be f r e e from attack. Internal corrosion of casing can also occur below the fluid level. Here the attack should generally be t h e same type and at t h e same r a t e as t h a t experienced by t h e tubing. Preventive Measures Several methods of protection against internal casing corrosion a r e used. If a packer can be installed in t h e well and t h e annulus filled with a sweet oil or inhibitor sdution, essentially 100-percent protection can be obtained. This method of protection h a s been successfully used i n a g r e a t number of wells f o r over 10 years. It should be noted t h a t this method of control will provide corrosion protection to only t h a t portion of t h e casing above t h e packer. Corrosion will continue below t h e packer a t i t s normal rate.
Specimens after Six Weeks Exposure in Simulated Sour Vapor Space Conditions Specimens on the left were inhibited
Treatment of rods ancl tubing with inhibitors is difficult when there is a packer in the annulus Probably the most widely used method of controlling internal casing corrosion IS by circulating inllibitors In the well To be effective in prot e c t ~ n gall the internal surfaces esp3secl in an oil-well annulus, the inhibitor must b2 volatile under well conditions A illaterial of t!lis type that found early use was formaldehycle Anhydrous ammoma is also being usecl by some operators Pan Aillerican Petroleum Corp has developed an inhibitor which is satisfactory for this service The inllibitor has a volatile component carried in a hyclrocarbon base Laboratory data show the material 1s very effective in a simulated casing vapor zone condition where water, hydrogen sulfide, and carbon dioxide are present The effectiveness of this inllibitor is evident as noted in Fig 2 The illaterial has been found to operate successfully in gases containing up to 15-percent carbon dioxide and 12 percent hydrogen sulfide. As a result of laboratory tests, this inhibitor has been used in over 1,000 wells for more than one year Althougll fully conclusive results wlll not be available for several years, laboratory studies lead one to expect t h a t the inhibitor will be highly effective in preventing internal vapor space sulfide corrosion The illaterial is hanclled much the same a s the noilvolatile coi~lpounds wllich are widely used througl~outthe industry Fig 3 is a schematic drawing showing the inVALVE - V E N T
Fig 3-Typical Wellhead Connections for Chemical Lubricator
jection eclui~ment usecl for treating flowing wells Treatment wit!^ this compound is generally a t the rate of 11b2gal injected into the well every two weeks Simple colonmetric field tests have been devised to test the gas for the presence of the inhibitor It has been found that if any inhibitor can be detected in the gas by this test, protection will be obtained No reaction proclucts are formed w111ch could interfere with the production of the well, and no solids, which would be troublesome in disposal operations, are present in the procluced water. Good protective coatings applied to the inside of the casing should also be effective in preventing this attack This method would be satisfactory only if the casing and tubing could be landed in the well without damage to the coating A t the present tiille this method has not, to the authors' knowledge, been extensively used With the inlprovenlents t h a t are being made, both in coating illaterials and application methods, the use of internally coated casing may becoille econonlical and practical EXTERNAL CORROSION Causes Esternal casing corrosion is now recognized a s being caused by: 1 Bacterial action 2. Electrolytic effects. 3. Clleill~calaction by acidic subsurface waters. Bacterial action can cause external corrosion by chemical attack througll the l~roductsof bacterial metabolism o r decomposition, or through depolarizatioil of the metal surfaces by tllelr consumptioil of hydrogen Sulfate-reducing organisills have receiver1 the most attention in the corrosion field the last few years. Sulfate (SO,) which occurs 111 most water and soils, mainly a s the salts of calcium, sodium, and potassium, is reduced by these organisills to hydrogen sulfide (H.,S) , and thereby causes corrosion by cheinical hyclrogen sulfide (H,,S) attack and by depolarization This reaction- is shown by the follo~vingchem~calecluations ]References a r e a t the end of the paper
R H GOODNIGHT AND J P BARRETT
Anodzc solutson o f w o n 8H"O water (dissociates to) 8H+ Fe
8H+ Hydrogen Ion
Iron Ion + Hydrogen (in solution)
H,S 4H,O ~ ~ d l ~ o+~ iViter e n Sulfide
(in the presence of water yields)
6 (OH) Hydroxyl Ions
8(OH)Hydroxyl Ions 4Fe++ 8H
(in water yields)
(ylelds by bacterial activity)
Co?-ros.lonp,~ocklLcts: Fe++ H.,S Iron ~ y c l r o g e nSulfide 3Fe++ Iron
Cathocllc deyoba?.lxn.tlon: H,,SO, + 8H Sourze of Sulfate + Hydrogen
+ + + +
(in the presence of water yields)
FeS Iron Sulfide
Sulfuric acid is used in the foregoing equation only a s a n example of a source of sulfate Several instances of casing corrosion by bacterial a c t ~ o nhave been observed Fig 4 sllows a section of casing showing this attack and removed after six years' exposure Pit depths equal to two thlrds the wall thickness of the pipe were found Electrolytic effects a r e broadly defined a s either galvanic (two clifferent electrocles) or electrolytic (same electrode in different environments) in which electric currents flow f o r perceptible distances tllrougll inetal Corrosion occurs a t the anode where the current leaves the metal and enters the electrolyte, which in this case would be the earth Electrolytic effects, resulting from surfacesubsurface galvanic cells a s well a s completely subsurface cells, have been reported Tlle effect of a galvanic couple, formed by' the surface lines coupled to tlle casing, is well-known ' Lately, some evidence has been found which suggests t h a t the minor corrosion caused by a short-term exposure of casing to this surface-subsurface
3Fe(OH), Iron Hydroxide
Summanzed, the reaction is considered to be: H,SO, 2H,O 5 - 3Fe(OH), 4Fe Iron Iron + Source of Sulfate + Water (by bacterial Hydroxicle activity yields)
FeS Iron Sulficle
galvanic cell nlay be sufficient to cause serious corrosion problen~seven after the surface line is insulated from the well The anodic areas which develop before insulation of the surface line may couple with catllodic areas on the casing, resulting in the formation of subsurface cells wl~ichthen become the controlling corrosion mechanism Differences in the geocl~enlistryof subsurface formations can cause differences in electrical potential between these formations. When these formations are connected by the casing string, current will flow causing corrosion where it leaves the pipe. These subsurface electrolytic cells can often be determined by subsurface surveys I n come cases, a n indication of their presence may be obtained from the self-potential log " Chemical attack of the outside of casing can be caused by the dissolved acid gases (hydrogen e sulfide and carbon dioxide) in s u b ~ u r f a c waters. These acid gases can react by clirect conlbination wit11 the metal causing chen~icalcorrosion. Examples of this attack are the cas:ng failures
Corrosion Outside of Casing
opposite the Dakota water sand in central Kansas and possibly the casing failures in t h e E a s t Texas Field. I t should be pointed out t h a t in most cases two or more corrosion mechanisms a r e in operation which modify t h e effects of the foregoing discussed causes of corrosion.
Preventive Measures With t h e exception of cathodic protection, little can be done to control external casing corrosion on old wells. Fortunately, however, cathodic protection used in conjunction with adequate insulation of surface lines from the well will, if properly applied, control the previously mentioned types of external attack. Cathodic protection is the reduction or prevention of corrosion of a metal surface by changing i t s electrical characteristics through t h e use of
sacrificial anodes or impressed currents t o make the metal cathodic. This prevents or reduces corrosion by restricting current discharge on the metal surface. This action also causes a deposition of hydrogen on the cathode surface. By chemical action, this hydrogen can cause the formation of various alkaline chemical compounds. If the cathodic protection currents a r e of sufficient magnitude to reverse t h e flow of current, electrolytic effects and direct chemical attack can be controlled. The deposition of t h e alkaline film promotes a n environment which the bacteria cannot tolera t e ; and control of corrosion by their activity is achieved. Although bacteria can only be determined by visual inspection, the use of the casing potential profile tool h a s been shown to be helpful in determining if electrolytic effects a r e causing corrosion and t h e effect of cathodic protection on the pipe:' This method of survey consists of contacting the internal surface of the casing with two spaced probes a t selected depths. The voltage drop across the probes is read on surface instruments. These data a r e then plotted a s shown in Fig. 5. Portions of the curve to t h e left of the zero microvolt line indicate current flowing down the pipe, and a negative slope indicates current leaving t h e pipe. A positive slope of the curve indicates current entering t h e pipe, and t h a t portion of the curve to the right of the z2ro line indicates current flow up t h e casing.' I n some fields, design of a cathodic-protection system presents some difficult problems. F o r exPOTENTIAL BETWEEN ELECTRODES IN MICROVOLTS NEGATIVE BASE OF SURFACE CASING ABOVE HERE READINGS
R E N T FLOW IS DOWN PlPE WHEN E N T I A L S ARE NEGATIVE
CURRENT FLOW UP PlPE WHEN P O T E N T I A L S ARE P O S I T I V E
N:;u;p:;sT;M T;lH o R ;;E; I S TOWARD
R H GOODNIGHT AND J P BARRETT
ample, in a typical problem field, the high resistance of the soil and the estimated 1 to 5 amperes of current required for protection, may nlalte the economics of protection questionable. Where rectifiers are required for protection, their 11ig11 driving potential may result In more corrosion on foreign structures by stray electncal currents, i e interference, than the protection gained by tlle equipi~lenton which they were installed. Preliminary work has shown that by decreasing the distance between the ground bed and the well head, the depth to which tlle protective current is projected is also decreasd This technique makes possible t h e use of lower current for protection by controlling the area to which current is projected Also, in certain cases of rectifier installations, less interference will occur a s a result of the lower current requirenlents In a field under consideration, where 80 percent of the external casing corrosion has occurred in the upper 700 to 800 f t , this principle could be of great importance. Fig. 6, curve A, shows the native-state potential profile of a well in this fleld Curve B in thls figure shows t h a t protection to the top 70 f t has been obtained by the application of 0.6 ampere of protective current from a ground bed 20 f t from the \\re11 Curve C shows the potential profile obtained when 1.5 ampheres were used in the conventional system (grouncl bed located a minimum of 75 f t from the well). The use of the casing potential profile tool requires the pulling of the tubing This can become an expensive operation and linlits its use. The surface potential current requirement test4 has been developed to eliminate the need for the casing potential profile tool in deterinining t h e amount of current required for the cathodic protection of casing This surface illethod is rela-
Curve A-Natlve-state condlt~on Curve 8-Ground bed 20 f t from well, 0 6 amp Impressed Curve C-Ground bed 75 f t from well, 1 5 amp Impressed All curves run wtth 25-ft electrode spaclng, polar~tyasslgned to bottom electrode.
Fig 6-Effects of Ground-bed Spacing as Demonstrated by Casing Potential Profiles
tively inexpensive to run, requiring approximately two hours of engineering tiille; and the test may be run wllile tlle well is producing if the well head is insulated froin the flow line This surface nlethocl would allow specific design data to be obtained for each individual well a t a mininlum cost. I11 the surface potential current requiren~ei~t test, the extent of polarization of the casing is measured using a reference electrocle and a temporary current source Current is applied in small increments in increasing amounts; and the potential of tlle well casing to grouncl is measurecl. These data are plotted a i d occasionally result in a sharp-breaking curve The point a t whlch the curve breaks iildicates tlle amount of current requirecl for polar~zationof the casing. Usually, the data give a snlooth curve in which case a n arbitrary method, using tangents to the curve to cletermine the break point, is recommendecl Lines AE and CD a r e dra\vn a s tangents to the curve a s shown 111 F i g 7 The angle AOC is bisected by line EF. A line, GH, is drawn touching the curve ancl parallel to line EF Where line GH intersects line AB, the break point of the curve on the abscissa is indicated Experience has shown that, when this amount of current is applied to the casing, protection is obtained.
Fig. 7-Current-requirement Measurements Ground Bed and Reference Electrode 75 Ft from Well
It has been clemonstrated by pipeline operations that the best method of protecting buried structures against external corrosion would be tlle use of coatings and cathodic protection. The coatings reduce the ainount of current required
The surface potential current requirement test can also be used to deterinine tlle ainount of current required for protecting the casing string t o a given depth Flg.7 shows the current requlred for protection when the reference electrode and ground bed are 75 f t from tlle well By inoving the ground bed and electrode to within 20 f t , it can be seen 111 Fig. 8 t h a t the amount of current required for protection IS only 0 6 ampere. When these amounts of current were applied t o the well, the subsurface potential profiles, previously shown in Fig. 6, curves B and C, were obtainecl By proper spacing of the ground bed, the depth of current penetration can be coi~trollecl. This results in lower current requirements where corros~oilis confinecl to the upper portion of the casing It is possible that overall greater use of cathodic protection for oil-well casing mill result from this technique.
Measurements Ground Bed and Reference Electrode 20 Ft from Well
R. H. GOODNIGHT AND J. P. BARRETT
To investigate the problems of coating casing, a 1,000-ft string of 10%-in. spiral-weld slipjoint pipe was coated by spray techniques with a n epoxy-coal t a r coating. The coated pipe was stored outdoors for 8 months prior to running. No special precaution other than the use of excelsior pads was talren to protect the pipe during
Fig. 9-Coating Casing Collar and weld are being cleaned by wire brushing.
for protection and result in better corrosion control. In view of this pipeline experience, the best method for protection against external corrosion for new well casing would be coatings and cathodic protection. Various attempts have been made to run externally coated casing. However, the suc-ess of such work has been doubtful. This has been due largely to the lack of a suitable coating material. Recently these have been made available epoxy-coal t a r copolymer coatings which appear to p3ssess the required properties for an external coating for casing. These materials have good abrasion and chemical resistance, excellent adhesion to metal, and, being catalytically set, lend themselves to rapid field patching.
Fig. 1 1-Coating
Collar and welded area have been coated and the completed joint is ready to be lowered into the hole.
Fig. 10-Coafing Casing Workman is applying coating-glass-fiber mixture to collar and welded area.
transportation from the coating yard to the well. The pipe was racked a t the wellsite on a steel rack with no pads used for protection. In the cou.rse of unloading and racking, one joint of pipe was dropped on a n asphalt road sufficiently hard to crack one collar and knock it out of round without damaging the coating on the remainder of tlze joint.
T1w pipe was weldecl using three beads on each joint. Fig 9 shows tlle amount of burn-back on the pin end of the joint This burlung occurred only where the temperature of the pipe exceeded 400 IF. The collars, which were left bare during the ~ n ~ t i coating al operation, were coated after welding by wire bruslllng and applying a brush coat of a glass-fiber filled mixture of the coating (Fig 9 and 10) Fig. 11 sllows a coinplete patch. Tlir action of the slips caused only a slight amount of damage and did not result in the removal of any large anlount of coating Altllougll the coated pipe was exposed to severe mechanical abuse during storage, transportation, and surface handling during running, less than 5 percent of the coating was damaged as determined by visual inspection Based on laboratory burial ancl water-adsorption tests and the condition of the coating on the casing as it was lowered into the hole, it appears t h a t these materials will be satisfactory for use a s coatings for the external surfaces of caslng It sl~ouldbe noted that the use of coatings alone could possibly lead to increased corrosion of the casing string a t areas where the coatings are damaged ancl bare metal is exposed For this reason, tlle use of catllodic protection in conjunction with such coatings is considerecl to be almost manclatory, ancl together, they should give virtually 100-Percent protection against external attack.
CONCLUSIONS W11en one considers t h a t corrosion of the casing could result in expensive workover jobs, loss of the well, or even permanent clanlage to the
reservoir, it IS concluded t h a t the most important corrosion problem facing production operations today is t h a t of casing corrosion It has been shown t h a t casing failures can be the result of rapid ancl severe internal corrosion caused by acid-gas attack or external corrosion froill bacterial action, electrolytic effects, or cllemical action by acicllc waters Internal corrosion can be controlled by the proper application of cl~emical~nhibitors,and by packers and oil-filled annuli Altllougll internal coatings for casing have not been successful in preventing this corrosion, future work along this llne may be fruitful. Fortunately, the several types of external corrosion can, under certain conditions, be prevented by a single preventive measure-cathodic protection On new wells. it is belleved that coatings on the outside of casing seem to offer advantages when used in conjunction with catl~odicprotection.
ACKNOWLEDGMENT The authors wish to express al~preciationto Pan American Petroleunl Corporation for the release of ancl pernlission to present thls paper REFERENCES IStarh-y, R L ancl Wlght, K M Anaerobic Corrosion of Iron and Steel, G o w o s l o , ~ ,3, May (1947) Wa~iimlnon,R E ancl Ewmg, S P Protect1011 of Well Caslng Agalnst External Corroslon, I,l,'o~Id011, 132 [21 154, Feb 1 (1051) "De Witte, Leendert, and Raclcl, Fred J Corroslon of 011-Well Caslng by E a i t h Currents, J P c t ~Tech , 7 [dl, Aprll (1955) -'Barrett, J P and Goulcl, E D Cathodlc Protectloll of 011-Well Casllies. NACE. Chlcaro. March 7-11. 1955