当前位置：文档库 › Experimental study on water–rock interactions during CO2 flooding in the Tensleep Formation, Wyomin

Experimental study on water–rock interactions during CO2 flooding in the Tensleep Formation, Wyomin

Experimental study on water±rock interactions during CO 2

ˉooding in the Tensleep Formation,Wyoming,USA

Ryoji Shiraki*,1,Thomas L.Dunn 2

Institute for Energy Research,University of Wyoming,Laramie,WY 82071-4068,USA

Received 24February 1997;accepted 30December 1998

Editorial handling by R.Fuge

Abstract

The conditions for mineral alteration and formation damage during CO 2treatment of Tensleep sandstone reservoirs in northern Wyoming,USA,were examined through core-ˉooding laboratory experiments carried out under simulated reservoir conditions (808C and 166bars).Subsurface cores from the Tensleep sandstone,which were cemented by dolomite and anhydrite,and synthetic brines were used.The brines used were (Ca,Mg,Na)SO 4±NaCl solution (9.69g/l total dissolved solids)for Run 1and a 0.25mol/l NaCl solution for Run 2.The solution used in Run 1was saturated with respect to anhydrite at run conditions,which is characteristic of Tensleep Formation waters.

Three major reactions took place during ˉooding,including (1)dissolution of dolomite,(2)alteration of K-feldspar to form kaolinite,and (3)precipitation (in Run 1)or dissolution (in Run 2)of anhydrite.All sample solutions remained undersaturated with respect to carbonates.The permeability of all the cores (except one used in Run 2)decreased during the experiments despite the dissolution of authigenic cement.Kaolinite crystal growth occurring in pore throats likely reduced the permeability.

Application of the experimental results to reservoirs in the Tensleep Formation indicates that an injection solution will obtain saturation with respect to dolomite (and anhydrite)in the immediate vicinity of the injection well.The injection of NaCl-type water,which can be obtained from other formations,causes a greater increase in porosity than the injection of Tensleep Formation waters because of the dissolution of both dolomite and anhydrite cements.#1999Elsevier Science Ltd.All rights reserved.

1.Introduction

Carbon dioxide ˉooding,in which CO 2gas and

brine are injected into reservoirs through wells to pro-duce more oil,has been an important technique for enhanced oil recovery since the 1980s.Many studies have demonstrated experimentally that CO 2ˉooding has certain advantages;speci?cally,the dissolution of carbonate cement into carbonate water increases the permeability of sandstone and carbonate reservoir rocks (Ross et al.,1981,1982;Omole and Osoba,1983;Sayegh et al.,1990).However,carbonate scale formation and corrosion,which also accompany CO 2ˉooding,are costly nuisances.Patterson (1979)reported both scale formation and downhole corrosion

Applied Geochemistry 15(2000)265±279

Present address:Department of Chemistry and Department of Land,Air and Water Resources,University of California,Davis,CA 95616,USA.2

Present address:INTEVEP,PDVSA,Los Teques,Venezuela.

*Corresponding author..

E-mail address:rshiraki@https://www.wendangku.net/doc/ce6762158.html, (R.Shiraki).

after CO 2injection in the Kelly±Synder oil ?eld,Texas,USA,although he did not specify the type of scale minerals.Shuler et al.(1989)and Bowker and Shuler (1991)reported carbonate scale formation due to CO 2injection in the Weber sandstone,Rangely oil ?eld,Colorado,USA.The e ect of CO 2injection on hydrocarbons and minerals is another problem that requires consideration.For instance,Monger and Fu (1987),Monger and Trujillo (1991)and Wolcott et al.(1989),assessed asphaltene precipitation as a by-pro-duct of CO 2-oil interaction.

None of those studies,however,concentrated on the impact that the brine composition used for ˉooding and the water±rock interactions occurring during ˉooding have on reservoir characteristics and scale for-mation.The authors have conducted two CO 2ˉooding experiments using cores and synthesized brines from the Tensleep Formation in the Bighorn Basin,northern Wyoming,USA.Temperature and pressure conditions were 808C and 166bars,typical reservoir conditions in the Bighorn Basin (e.g.,Big Sand Draw oil ?eld,Smith,1993).The present paper describes the exper-imental results,the nature of water±rock interactions during CO 2ˉooding,and the e ect of the chemistry of the injected solution on reservoir characteristics.2.CO 2ˉooding

2.1.Outline of CO 2ˉooding

Fig.1is a schematic diagram showing how well-to-well CO 2ˉooding is used as a technique for enhanced oil recovery.As seen in Fig.1,CO 2gas is injected alternately with water into the injection wells.This process is repeated several times until no further enhancement of oil production is observed (e.g.,Beeson and Orthlo ,1959;Holm and Josendal,1974).Pressure is high in the vicinity of the injection well.The portion of the reservoir between the injection and production wells has a pressure about half that of the

area around the injection well,and near the production well,the pressure drops rapidly (Thomas et al.,1992).In reservoirs where the pressure is relatively low,CO 2ˉooding is accomplished using an immiscible cyc-lic CO 2injection.In this case,only CO 2gas is injected into the reservoir and the well is shut-in for a couple of weeks to allow CO 2gas to soak in the reservoir.Production begins again from the same well after CO 2soaking (e.g.,Palmer et al.,1986;Monger et al.,1991).2.2.Previous experimental studies

Ross et al.(1981)carried out CO 2core ˉooding ex-periments at 20and 808C using calcitic and dolomitic sandstones and oolitic limestone (the chemical compo-sition was not described).They observed increases in permeability in each case,particularly in oolitic lime-stone with a smaller pore volume of carbonated water (<50).

Omole and Osoba (1983)carried out 3sets of CO 2ˉooding experiments at 278C,in which CO 2gas was injected into dolomite cores saturated with a 0.1N KCl solution.They examined the e ects of the amount of CO 2injected into the cores,the injection pressure of CO 2,and the pressure gradient across the core on the change in permeability before and after treatments.They observed a maximum 64%increase in per-meability after the injection of 3.28pore volumes of CO 2,5to 22%increase in permeability in response to an increase of CO 2injection pressure from 72to 172bars,and up to 22%increase in permeability for press-ure gradient of up to 72bars.They assumed that any increase or decrease in permeability was due to dissol-ution and precipitation of minerals at pore throats,and concluded that CO 2ˉooding into carbonate reser-voirs would result in the dissolution of carbonate min-erals near injection wells and precipitation of them as the pressure drops.

Sayegh et al.(1990)carried out a CO 2ˉooding ex-periment at 458C and 138bars using sandstone cores from the Upper Cretaceous Cardium Formation at the Pembina oil ?eld,Alberta,Canada.They observed that the permeability of the cores decreased rapidly at an early stage during the runs (10to 60%of the orig-inal values),and then gradually increased,although the permeability did not regain the original values during runs of 25to 50h.They also observed the dis-solution of carbonate cement (calcite and siderite)under a microscope.They concluded that the decrease in permeability was due to the migration of ?nes (mainly illite)to pore throats and the increase due to the dissolution of carbonates.

Bowker and Shuler (1991)ˉooded cores from the Weber sandstone,Rangely oil ?eld,Colorado,USA with carbonated water,but the core permeability did not change substantially throughout the runs.

The

Fig.1.Schematic representation of well-to-well CO 2ˉooding for enhanced oil recovery.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

266

cores contained about 10vol.%carbonate cements (ferroan calcite and dolomite)and about 10vol.%carbonate cements (ferroan calcite and dolomite)and about 5vol.%of illite and mixed-layer illite/smectite.An increase in Ca,Mg and Fe in e uent solutions was observed,suggesting dissolution of carbonates.They concluded that no change in permeability was due to o setting factors.The increase in permeability by car-bonate dissolution was o set by a reduction caused by migration of clay minerals into pore throats.3.Experimental 3.1.Materials

Drill cores from 981to 1208m depth from the Tensleep Formation in the Oregon Basin oil ?eld,northern Wyoming,USA were drilled parallel to lami-

nation to yield the 6Tensleep core samples used in the experiments.The 6core samples were each 3.8cm in diameter and about 7.6cm in length,and were con-siderably cemented by dolomite and anhydrite.Crude oil present in the samples was removed by organic sol-vent before the experiments commenced.

The injection solution used in Run 1was prepared by dissolving CaSO 4,MgSO 4,Na 2SO 4,and NaCl,into distilled water to approximate a composition of Tensleep Formation water (Shiraki and Dunn,1995;Dunn et al.,1994;Dunn et al.,1995).The solution was equilibrated with respect to anhydrite at exper-imental P±T conditions (808C and 166bars total press-ure).A solution of 0.25mol/l NaCl solution was prepared for Run 2.3.2.Apparatus and runs

Fig.2shows a schematic diagram of the

experimen-

Fig.2.Schematic diagram of the core-ˉooding apparatus used.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279267

T a b l e 1C h e m i c a l c o m p o s i t i o n o f i n j e c t i o n a n d s a m p l e s o l u t i o n s

D t a T i m e b p H c N a K C a M g F e M n A l S r B a S i C l S O 4A l k a l .T I C d

R u n S a m p l e

(m i n )(h )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )(m m o l /l )

I n i t .S o l .±±3.1190.10.02215.717.50.1464.2812.031.40.080.08490.132.5±

2.341±01380

3.2

4.1587.82.4318.618.70.21329.523.447.90.892.5689.729.78.550.6581±0248014.73.6382.40.71414.618.70.55717.537.828.50.491.708

5.028.75.801.661±0348022.73.6585.80.44014.519.20.48310.645.627.40.441.1289.529.54.821.311±0440030.63.7085.70.36115.019.50.40610.638.925.10.550.99388.329.95.341.261±0540037.33.738

6.40.28415.119.50.3378.9222.22

7.40.420.8448

8.530.24.921.101±0640043.93.7687.90.23814.91

9.50.3388.9223.026.20.470.76588.029.95.341.101±0740050.63.9086.70.20515.319.80.3468.9222.224.00.480.69888.530.45.770.8221±0821056.43.7785.90.18214.819.70.3699.1031.935.40.520.64189.530.25.431.101±0936061.23.8083.50.16415.419.70.35810.920.427.40.390.54589.031.25.561.011±1030066.73.8285.60.15316.019.80.2707.2814.826.20.360.50988.129.35.881.011±1130071.73.8482.50.13815.820.00.2695.4612.225.10.400.50289.028.96.171.011±1231077.43.8787.00.13616.020.10.2495.4614.825.10.390.48890.929.16.380.9751±1335082.93.9185.50.12516.120.00.2545.4611.934.20.360.47389.299.76.390.8791±1435088.73.9283.20.11816.420.00.2565.4611.526.20.390.46690.429.76.550.8811±1535094.63.9388.60.12016.420.00.2675.6414.127.40.350.48189.029.66.610.8791±16400101.53.9582.70.11016.620.30.2385.6411.124.00.400.48888.830.06.960.8781±17400108.23.9187.90.09716.319.90.3265.6413.725.10.37054188.529.66.880.9461±18400114.93.9074.70.08715.920.00.3407.4614.122.80.370.46680.029.16.700.9481±19400121.53.9084.70.09515.819.90.4037.4617.425.10.370.49889.328.86.610.9431±20400129.03.8082.20.10515.920.00.3927.4614.539.90.350.53089.429.26.881.231±21400135.73.8783.20.09015.320.10.3979.2820.021.70.340.50288.928.66.611.001±22400142.43.8786.80.08415.520.10.3969.2825.921.70.350.48888.528.76.540.9871±23400151.73.8585.70.09215.620.20.4019.2819.326.20.360.50690.628.96.341.011±24400

158.4

3.9186.30.08716.520.20.4219.2823.73

4.20.350.48490.930.8

5.67

0.784

I n i t .S o l .±±2.922460.0160.030.010.07111.21.670.190.270.0622500.49±2.292±014003.34.852444.0423.18.390.06316.913.774.32.041.0125414.016.40.2302±023209.34.312381.5019.48.110.06612.47.6445.42.040.82725510.616.70.7962±0324015.34.032410.66717.58.110.10312.06.7533.12.040.55524810.39.610.8802±0424019.24.232360.49617.78.120.17712.47.8928.92.400.4672569.4116.50.9352±0532024.04.642380.40519.37.940.27810.44.0035.01.970.74025511.015.80.3562±0632029.34.282340.31016.37.610.47012.48.4526.02.480.4602548.3316.50.8502±0740035.34.262350.25715.47.910.47514.05.5621.32.990.3242387.2916.60.8962±0840042.94.222360.23515.98.540.45714.25.5220.22.990.3202557.7517.00.9962±0948050.24.272490.20414.58.090.38117.33.1917.83.280.3112416.6817.50.9252±1048058.24.262390.18014.08.210.39115.74.8616.13.280.2812506.4117.20.9332±11480

66.24.272430.16213.88.160.40916.04.1115.03.640.2752516.24

17.2

0.921

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

268

2±1240074.04.262410.15513.38.310.36916.22.6714.03.790.2742456.0617.10.9312±1348081.64.252340.13713.08.260.38416.03.6313.53.790.2532486.0916.80.9482±1448089.54.312430.13012.17.410.41916.43.6012.94.080.1922485.7915.90.7962±1548098.24.272380.12612.17.850.45716.42.6311.54.220.1932445.5216.40.8932±16480106.24.262430.12411.87.80.44715.12.7811.05.590.1822444.9016.40.9042±17480114.24.262440.11612.08.030.45415.53.2210.65.100.1792474.8416.20.9032±18480123.24.272450.11511.87.770.44216.02.4510.35.030.1942514.8416.70.9012±19480131.24.302410.10911.37.570.46119.52.5611.05.540.2022434.4416.30.8222±20480139.24.322470.10411.27.550.48822.62.979.245.970.2042454.4416.60.8112±21480148.24.252400.09911.37.830.46024.62.228.675.750.2082514.5516.50.9382±22480156.24.232480.09811.37.850.49626.02.789.025.970.2192494.4916.30.9562±23480164.24.282440.09311.48.070.49021.82.748.226.410.2162484.3616.90.895

D u r a t i o n o f s a m p l e c o l l e c t i o n .b

T i m e o f t h e m i d d l e p o i n t o f d u r a t i o n o f s a m p l e c o l l e c t i o n s i n c e t h e r u n h a s s t a r t e d .c C a l c u l a t e d i n s i t u p H .d C a l c u l a t e d .

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

269

Fig.3.Chemical composition of e uent solution vs.reaction time.A,C and E show the results of Run 1and B,D and F show those of Run 2.The Y axes are common for A and B,C and D,and E and F,respectively.Solid vertical lines in B,D and F show the time when the interruption of the injection took place.Data after the interruption are plotted by neglecting the duration of the interruption.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

270

tal apparatus used in both runs.For each run,3cores wrapped with a Teˉon sheet were placed into rubber tubing and both sides of the tubing were tightly attached to stainless steel core-ends with tubing clamps.This core assemblage was placed in the core holder(high pressure vessel),and water was injected into the free space in the vessel.Nitrogen gas was introduced into the tank connected to the pressure vessel to obtain an overburden pressure about100bars higher than the internal pressure.This overburden pressure ensures a tight?t between the rubber tubing and the outside cores,thus preventing solution from passing through the space between the rubber tubing and the outside of the cores.

The injection solution and CO2gas were injected into cylindrical reservoirs(500ml volume)from the bottom side of the apparatus.After the temperature of the system reached808C,the CO2pressure was set to precisely166bars by adjusting the position of pistons. The constancy of the CO2pressure suggests that equili-brium between the CO2gas and the injection solution was obtained.A Ruska pump was used to send min-eral oil into the upper part of the reservoirs,continu-ously displacing the piston and injection solution.The injection rate of the solution was set to7.5ml/h for both runs.

During each run,the e uent solution was collected continuously into cylinders of the autosampler after passing through the back-pressure regulator(BPR);the cylinders shifted position at about80min intervals. CO2degassed out of sample solutions was collected in a gas collector and the volume measured.Once a day, 5or6e uent solutions were mixed together in a plas-tic bottle to obtain enough solution to permit chemical analyses.During the sample collection and mixing,the injection of solution was interrupted for about1h. The collected solutions were not?ltered.

The?rst experiment was run for161.7h and a total of24mixed-samples were collected(denoted1±01 through1±24).Table1shows the amount of time required to collect each sample and the midpoint of each sample interval(i.e.,the reaction time).In the sec-ond experiment,the injection of the solution was inter-rupted for44.9h,beginning at a reaction time of21.3 h,due to a malfunction in the autosampler.However, this experiment was resumed normally and ran for 168.2h,excluding the44.9h interruption,with a total of23mixed-samples collected.

3.3.Analyses

The pH and alkalinity of the e uent for both runs were measured daily using a combination pH electrode and HCl titration,respectively.The residual samples were acidi?ed using1:1HNO3in order to keep all ions soluble for later analyses.Contents of Na,K,Ca,Mg,Fe,Mn,Al,Sr,Ba,and Si were measured by ICP and atomic absorption spectrometry,and Cl and SO4were measured by ion chromatography.The accuracy of the analyses is estimated to be better than5%.

The porosity and air permeability of the cores were measured before and after each run using the Core Laboratories Instruments CMS±300,an automated core measurement system.Following measurement of the porosity and permeability,the cores were examined by scanning electron(SEM)and optical microscopy. Rock fragments obtained during the preparation of the sample cores were also examined using a scanning elec-tron microscope to determine the original mineral mor-phology.

3.4.Speciation of solutions

SOLMINEQ.88(Kharaka et al.,1988)was used to calculate the speciation of each solution.In these cal-culations,the amount of CO2degassed from the sol-ution after sampling was taken into account to calculate in situ pH https://www.wendangku.net/doc/ce6762158.html,ing the results of specia-tion,the saturation index,S.I.,of minerals given by:

S X I X log

IAP

was calculated,where IAP is the ion activity product and K sp is the solubility product.The value of S.I. should be zero at equilibrium;positive and negative values indicate supersaturation and undersaturation, respectively.All values for K sp,with the exception of that for dolomite,were taken from the original data-base of SOLMINEQ.88.The value for dolomite was calculated by adopting an generally accepted value,log K sp=à17.0at258C and1bar(Hsu,1963;Langmuir, 1971;Busenberg and Plummer,1982),and by using values for the temperature and pressure dependence of K sp used in SOLMINEQ.88.

4.Results

4.1.Solution chemistry

The analytical results of the collected sample sol-utions for both runs are summarized in Table1.Fig.3 shows the change of solution chemistry with reaction time for both runs.The vertical lines at t=21.3h in Fig.3B,D,F show the time when the interruption of injection occurred in Run2.

In Run1,similar changes occur in pH,Ca,and al-kalinity(Fig.3A).The solution pH of Sample1±01 jumped from an initial value of3.11for the injection solution to4.15.However,the pH dropped to3.63in Sample1±02and then gradually increased to a steady

R.Shiraki,T.L.Dunn/Applied Geochemistry15(2000)265±279271

value of about 3.85at a reaction time of about 90h.The Ca content of Sample 1±01was 18%more than that of the injection solution;however,it returned to nearly the value of that of the injection solution during the rest of the run.The Mg content increased with reaction time and the last sample,1±24,contained 1.15times as much Mg as the injection solution.The SO 4content was about 10%lower than that of the injec-tion solution throughout the run,and the Ba content also remained low (<0.9m mol/l)throughout the run (Fig.2C).The SiO 2and K contents in Sample 1±01were 2.56and 2.43mmol/l,respectively;however,they decreased with reaction time to steady values (0.49and 0.095,respectively)within 80h (Fig.3E).Similar changes to those of SiO 2and K occurred in the Al content,but it was the third sample (1±03)that had

the highest value;the Al content increased during the last stage of the run (t >100h).

In Run 2,the contents of Ca,K,Al,Sr,SiO 2,and SO 4for Sample 2±01increased considerably compared to those for the injection solution (Fig.3B,D,F).All these contents then decreased with reaction time,although the contents of Ca,Sr,and SO 4of Sample 2±05,which was taken just after the interruption of injec-tion,increased by 10to 21%compared with those of Sample 2±04.The pH of Sample 2±01increased from 2.92for the injection solution to 4.85and then decreased to 4.03in Samples 2±02and 2±03.The pH of Samples 2±04and 2±05increased to 4.23and 4.64,respectively,and then remained constant until the end of the run (values in the range 4.22and 4.31).The al-kalinity and the content of Mg were very constant throughout the run (alkalinity 15.9±17.5mmol/l;Mg 7.41±8.39mmol/l),although Sample 2±03had a signi?-cantly low alkalinity (9.61mmol/l)compared to the other samples.Unlike the contents of the other el-ements,that of Ba increased consistently with reaction time to reach 6.41mol/l by the end of the run (Sample 2±23).

Total inorganic carbon,TIC,which was calculated from the measured solution pH,alkalinity,and the CO 2recovered from the degassed solutions using SOLMINEQ.88was much less than that of the orig-inal injection solution (Table 1).This was likely due to the loss of CO 2by di usion through the Teˉon sheet and rubber tubing when the solution passed through the core assemblage (H.Haines,personal communi-cation).For t >100h,the TIC of the solution was 34to 43%of the original value of the injection solution for Run 1,and 35to 42%for Run 2.4.2.Mineral saturation

Fig.4shows the change in the S.I.of minerals with reaction time.In Run 1,the sample solutions were saturated with respect to anhydrite and barite through-out the run.The S.I.of dolomite for Sample 1±01was à2.79,whereas the value dropped to à4.14and à4.24for Samples 1±02and 1±03,respectively.The S.I.then increased gradually with reaction time,remaining at about à3.5for the period when the reaction time was >100h.The S.I.of aragonite varied in a similar man-ner to that of dolomite,but the values were larger.In Run 2,changes in the S.I.values of carbonates during early stages of the run were complex.The S.I.of dolomite for Samples 2±01through 2±03decreased from à1.13to à3.28,and then began to increase in Samples 2±04and 2±05up to à1.63.The value of Sample 2±06dropped to à2.38,but then S.I.remained constant until the end of the run.The range of values for steady states were à2.37to à2.58.The S.I.of anhydrite for Sample 2±01was à0.43and

decreased

Fig.4.Saturation indices of minerals vs.reaction time (top:Run 1,bottom Run 2).Values of the log of the solubility pro-duct at run conditions are:anhydrite à4.88;celestite à6.62;barite à9.45;dolomite à18.30;aragonite à8.72;strontianite à9.48;witherite à8.95;and K-feldspar à1.48.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

272

gradually toà1.18with reaction time;the sample taken just after the injection interruption showed a slight increase.The S.I.of celestite showed a similar trend to that of anhydrite,but with smaller values. The solution was saturated with respect to barite throughout the run,as in Run1.

4.3.Rock phase

Table2lists the pore volume,porosity,and per-meability of cores before and after each run.The per-meability of the cores used in Run1decreased by16 to21%during the run,but no signi?cant change in pore volume and porosity was observed.The upstream and middle cores used in Run2showed respective increases in pore volume of0.7and0.2cc,resulting in respective increases in porosity of9and2%.The downstream core showed no signi?cant change in pore volume and porosity.The permeability of the top core decreased by31%and the bottom core by44%,but the middle core showed a27%increase in per-meability.

Fig.5shows scanning electron photomicrographs of the cores used in both runs as well as the original unˉooded Tensleep Formation cores.Fig.5A and5B show K-feldspar grains before and after the injection. The pre-injection grains have a smooth surface,but afterˉooding they exhibit a skeletal structure.Fig.5C and5D show dolomite grains before and after the injection;Fig.5D shows step-like structures,which are not seen in the pre-injection sample(5C).Some round corners are also observed in the post-injection sample. Fig.5E and F show anhydrite present in cores used in Runs1and2.Fig.5F shows the dissolution texture on the surface of anhydrite,whereas such texture is not observed in Fig.5E.Fig.5G and H show platy hexa-gonal kaolinite crystals,which were not observed in the pre-injection samples.5.Discussion

5.1.Water±rock interaction

The fact that changes in the contents of SiO2,K, and Al with reaction time were similar in both runs suggests that the contents of these ions were controlled by the same reaction.The dominant feldspar species of the Tensleep Formation is K-feldspar(Todd,1964; Mankiewicz and Steidmann,1979).Using SEM,it was observed that K-feldspar was altered and kaolinite crystals were formed during the experiment(Fig.5). Fig.6shows that all sample solutions for both runs fall in the kaolinite?eld in the activity diagrams of the system Na2O±K2O±Al2O3±SiO2±HCl±H2O at808C and166bars,which were constructed using the ther-modynamic data of Helgeson et al.(1978,1981).These results strongly suggest that the contents of SiO2,K, and Al are mainly controlled by the hydrolysis reaction of K-feldspar to form kaolinite,which is given by:

2KAlSi3O8 2H 9H2O

4Al2Si2O5 OH 4 2K 4H4SiO4 2

The fact that the contents of these constituents increase drastically for the?rst sample of each run but then decrease consistently(Fig.3)suggests that the rate of the K-feldspar-to-kaolinite reaction is greatest at an early stage of the run,and then decreases with reaction time to reach a steady state.The higher pH of the?rst sample of both runs and the subsequent drop in pH may also be due to this reaction.

According to Busenberg and Plummer(1982),the dissolution of dolomite is controlled by the following three reactions:

CaMg CO3 2 s 2H 4Ca2 Mg2 2HCOà3 3

Table2

Physical properties of cores

Pore volume(cc)Porosity(%)Air permeability(md)

Run Core

before(b)after(a)a±b before(b)after(a)a/b before(b)after(a)a/b

1upstream12.612.60.014.814.60.9976.064.20.84 middle12.012.10.114.214.3 1.0164.653.70.83 downstream12.512.50.014.614.9 1.0261.848.60.79 total in pore volume37.137.20.1

2upstream7.88.50.79.210.0 1.09 6.62 4.540.69 middle12.112.30.214.414.7 1.0299.7126.9 1.27 downstream8.58.50.011.911.9 1.0027.215.10.56 total in pore volume28.429.30.9

R.Shiraki,T.L.Dunn/Applied Geochemistry15(2000)265±279273

CaMg CO 3 2 s 2H 2CO o 34Ca 2

Mg 2 4HCO à3

CaMg CO 3 2 s 2H 2O 4Ca 2 Mg 2 2HCO à3 2OH

They observed that the dissolution rate was pro-portional to the activity of H +below a pH of 6,

whereas independent of the activity of H +above a pH

of 8in the absence of CO 2.In the presence of CO 2(0to 0.96atm),the dissolution rate was proportional to the square root of P CO 2(i.e.,equivalent to a H 2CO o 3)below pH 5.Busenberg and Plummer (1982)also observed that the backward reaction becomes domi-nant in proportion to the activity of HCO à3.

As the pH range of the present solutions was 4.03to 4.85,Eq.(3)may be the dominant mechanism of dolo-mite dissolution for CO 2ˉooding.The range of P CO 2of the present study was,however,much higher than that of Busenberg and Plummer (1982),so that the contribution of Eq.(4)and the reduction of dissol-ution rate due to backward reaction are not clear.

Another reaction that may have had a great impact on the chemistry of the e uent solution is the dissol-ution or precipitation of anhydrite,which controlled the concentration of Ca and SO 4.The injection sol-ution of Run 1was originally saturated with respect to anhydrite,so that the mineral precipitated to consume the excess Ca released by dolomite dissolution.The precipitation kept the Ca content constant and led to a decrease in the SO 4content in solution.All solutions were saturated with respect to anhydrite throughout Run 1(Fig.4).In contrast,in Run 2,anhydrite dis-solved rapidly during the early stages of the run,but the dissolution rate decreased with reaction time in the same manner as K-feldspar dissolution.The Sr content also decreased in this run (Fig.3D),suggesting that Sr was also released by the dissolution of anhydrite.As shown in the following section,calculating the di er-ence in the contents of Ca and SO 4(i.e.,C Ca àC SO 4;C is the molar concentration)allows us to determine how much dolomite was dissolved during the runs.This di erence may also be indicative of the amount of Ca released by dolomite dissolution;this is in good agree-ment with the Mg increase,which supports the idea that dolomite and anhydrite provide the source of Ca in the run.

All solutions were saturated with respect to barite despite the di erent contents of SO 4and Ba in both runs,suggesting that the content of Ba was controlled by its solubility (Fig.3C,D).

5.1.1.E ect of the interruption of injection on solution chemistry in Run 2

The interruption of the injection of carbonated water in Run 2caused the solution remaining in the pores (initially 28.4ml,Table 2)to be stagnant.This allowed some minerals,with which the solution was undersaturated,to dissolve until equilibrium was reached with the solution in the cores.Increases in the contents of Ca,SO 4,and Sr in Sample 2±05suggest that anhydrite dissolution occurred during the inter-ruption.

The calculations indicate that the interruption

Fig.5.Scanning electron photomicrographs of pre-and post-CO 2ˉooding cores.A.K-feldspar before ˉooding;B.K-feld-spar after ˉooding (Run 2);C.dolomite before ˉooding;D.dolomite after ˉooding (Run 2);E.anhydrite after ˉooding (Run 1);F.anhydrite after ˉooding (Run 2);G,H.kaolinite formed during ˉooding (Run 2).Each scale bar denotes 5m m.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

274

injection caused the TIC to decrease in the solution,likely due to the loss of the C supplied by the injection solution and the loss of CO 2due to di usion through the Teˉon sheet and rubber tubing.This caused an increase in of a CO 2à3

the solution,which likely resulted

in the dolomite saturation without any additional dolomite dissolution.The solution trapped in the pore spaces was diluted by 41%with newly injected solution in Sample 2±05(total 40ml).Speciation by SOLMINEQ.88based on the chemical composition of Sample 2±05,in which the e ects of dilution were con-sidered,indicated that the TIC of the solution decreased to 67.9mmol/l during the interruption,thereby allowing the saturation with respect to dolo-mite at a pH of 5.45.

5.2.Dissolution/precipitation rate of minerals

Fig.7shows the total amount of mineral dissolved or precipitated (i.e.,dolomite,anhydrite,and K-feld-spar)as a function of the total volume of solution injected.This amount is the sum of the product of the content of each mineral's constituent (Mg for dolo-mite,SO 4for anhydrite,and K for K-feldspar)and the volume of solution injected for each sample.The amount of dolomite dissolved in Run 2was calculated from the di erence of C Ca àC SO 4,as well as from Mg solely;these amounts are in good agreement (errors of only 10%;5%for both C Ca and C SO 4).Note that the linear trend suggests a constant dissolution (positive slope)and precipitation (negative slope)rate;the

stee-

Fig.6.Activity diagrams of the systems K 2O±Al 2O 3±SiO 2±H 2O (left)and Na 2O±K 2O±Al 2O 3±SiO 2±HCl±H 2O (right)at 808C and 166bars.Thermodynamic data were taken from Helgeson et al.(1978,1981).

Table 3

Volumes of minerals dissolved or precipitated and their rates based on solution chemistry a Species used Run 1

Run 2

Minerals

for calculation volume (cm 3)rate (mmol/l)volume (cm 3)rate (mmol/l)Dolomite (Mg)0.17 2.3720.030.627.9420.02Anhydrite (SO 4)à0.16à2.9320050.34 4.2720.07?

K-feldspar (K)0.040.10220.002??0.050.10020.002??Kaolinite

(0.5K)

à0.02n.d.

Total in volume change

0.03

0.99

Positive and negative numbers represent dissolution and precipitation,respectively.Volume of kaolinite precipitated was calcu-lated based on the stoichiometry of Eq.(2).Dissolution and precipitation rates were calculated by data regression of linear parts shown in Fig.7;that is,values with ?and ??were calculated for the range of the injected solution of >600ml and >500ml,re-spectively,whereas others were done for the full range.n.d.:not determined.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279275

per the slope,the greater the dissolution/precipitation rate.The dissolution/precipitation rate of each mineral,which is an average rate that depends upon the resi-dence time of solution in the cores,was calculated from the slope of the plot in Fig.7for regions with constant slope.The results of regression are listed in Table 3.The volume of minerals dissolved or precipi-tated was calculated by multiplying the amount of mineral's constituent by the molar volume of each mineral (64.3for dolomite,45.9for anhydrite,and 109cm 3/mol for K-feldspar)(Robie et al.,1978;Helgeson et al.,1981).

Busenberg and Plummer (1982)suggested that the mechanism of dolomite dissolution may shift from sur-face-reaction control to di usion-control between 508and 1008C.Talman and Gunter (1992)supported their observation through experiments of dolomite dissol-ution at higher temperatures (100to 2008C).Therefore,it is likely that dolomite dissolution in the

present experiments was controlled partially by di u-sion (i.e.,mixed control)under experimental conditions (808C).The rounded dissolution texture seen in Fig.5may support this point.However,the di erence in the hydrodynamic conditions between the two runs was too small to account for the di erence in their dissol-ution rates of dolomite.It seems likely that the di er-ence in surface area of the mineral exposed to carbonated solution caused the di erence in the dissol-ution rate.In addition,microscopic observation indi-cated that the cores used in Run 2contained much more dolomite cement than those used in Run 1.Thus,the di erence in the dissolution rate of dolomite between the runs was probably caused by the di er-ence in the surface area that was exposed to the sol-ution in cores.

As seen in Fig.7,K-feldspar dissolved rapidly during the early stages of the runs;the initial rate in Run 2was greater than that in Run 1.The dissolution rate,however,gradually decreased and became con-stant.The calculated dissolution rate for volumes of solution >500ml are very similar in both runs (Table 3).This pattern follows the parabolic rate law described in early studies on feldspar dissolution,in which the dissolution rate of a mineral is proportional to the square of time (Wollast,1967;Helgeson,1971;Busenberg and Clemency,1976).However,a sub-sequent study by Holdren and Berner (1979)showed that feldspar etched by HF does not follow the para-bolic rate law;they attributed its kinetic pattern to the dissolution of very ?ne grains and strained particles formed during grinding,which had high surface energy during the early stages of experiment.The cores used in the present study were not treated with acid;thus,the higher dissolution rate during the early stages of the runs is probably due to the dissolution of smaller grains and parts having higher lattice energy due to crystal defects.SEM observation of altered K-feldspars (Fig.5)supports this interpretation.5.3.Porosity/permeability change

As seen in Table 2and Table 3,the measured pore volume increase is in good agreement with the total volume of minerals dissolved and precipitated,as cal-culated from the solution chemistry.No net change in pore volume occurred in Run 1;the decrease caused by the precipitation of anhydrite and kaolinite likely compensated for the increase caused by the dissolution of dolomite and K-feldspar.In Run 2,the upstream core showed the largest increase,followed by the middle core,suggesting that more minerals dissolved at the injection side than at the e uent side.

The decrease in the permeability of the 3cores used in Run 1and the downstream core of Run 2is prob-ably due to changes in the geometry of pore

spaces,

Fig.7.Total amount of minerals dissolved vs.volume of sol-ution injected.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

276

because no change in pore volume was observed.The upstream core used in Run 2showed a 31%decrease in permeability despite 9%increase in porosity.Such changes in pore geometry may have resulted from the crystallization of kaolinite in pore spaces (Fig.5G,H).This,as well as the rough surfaces on dolomite grains eroded by solution (Fig.5D),would have hindered ˉuid ˉow through the core.We are uncertain what led to the increase in the permeability of the middle core used in Run 2.

Previous experimental studies have shown that the permeability of carbonate rocks increases during CO 2ˉooding,whereas that of sandstones decreases or remains unchanged (see preceding section).The decrease in core permeability in the present study (except for one core used in Run 2)is consistent with the decreases observed in other CO 2ˉooding exper-iments (Sayegh et al.,1990;Bowker and Shuler,1991).These authors accounted for the permeability decrease by demonstrating that authigenic clays had migrated into pore throats.However,optical microscopy and SEM studies of the present cores showed that few clay minerals were present originally.Thus,the mechanism of permeability decreaseDthe formation of kaolinite crystalsDmay be characteristic of CO 2ˉooding for formations containing K-feldspar or other soluble alu-minosilicates grains.

6.Application to reservoirs in the Tensleep Formation Based on the experimental results obtained,the authors carried out a simulation on the change in min-eral saturation of the injected solution with migration.In this simulation,the dissolution rate of the minerals shown in Table 3was corrected for the unit distance by dividing the length of the core assemblage used in

the runs assuming a constant axial concentration gra-dient within cores (noted as R 1in mol/l/m).The con-stituents of dolomite,anhydrite,and K-feldspar were added to the injection solution as the ?rst step of the simulation.The amount of each constituent dissolved/precipitated (mol/l)can be calculated from R 1?D x (D x is the migration distance in m ).Then,the satur-ation state of each mineral,O 1(=IAP/K sp ),was calcu-lated based on the speciation of the solution using SOLMINEQ.88(Kharaka et al.,1988).Using the sat-uration state obtained,the dissolution rate (R 2)for the solution was calculated using the relation given by:R n R 1 1àO n à1

Where n denotes the number of core considered.Again,the mineral constituents were added into the solution using the value of R 2for the next distance of migration,and these procedures were repeated.Throughout the simulation,the following assumptions were adopted:T=808C (no temperature gradient in the reservoir),pressure gradient of 1bar/m,TIC=0.4TIC initial ,C SiO 2=0.5mmol/l (H quartz saturation),C Al =2.85m mol/l.A constant linear ˉow rate of 6.65?10à4l/cm 2/h is also assumed.The pressure gra-dient was estimated from Thomas et al.(1992)and other assumptions were based on our experimental results.

Fig.8shows the results of the simulation.The ?gure shows that the injected solution becomes saturated with respect to dolomite at a migration distance of 1.5m and 0.5m,respectively.The solution reaches satur-ation with respect to anhydrite at a migration distance of 3m in the simulation in which data obtained in Run 2were used,whereas it remains saturated with respect to the mineral throughout the simulation in which data obtained in Run 1were used.These values are extremely small compared to the distance

between

Fig.8.Results of the simulations on changes in the mineral saturation indices of the injected solution with migration.Left and right ?gures show simulation results using data obtained in Runs 1and 2,respectively.The Y axis is common for both ?gures.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279277

the injection and production wells in oil ?elds which are typically on a 40-acre spacing.Thus,in CO 2ˉood-ing,the dissolution of cement minerals likely takes place only in the immediate vicinity of the injection as suggested qualitatively by Sayegh et al.(1990).

In this simulation,the change in the surface area of minerals with reaction time has been neglected.As the amounts of mineral is limited in formations,the sur-face area of the dissolving minerals at a constant mi-gration distance would decrease,and eventually the amount of minerals dissolving would decrease.Therefore,the rate of change in the saturation state will be smaller than that predicted in the simulation.Another limitation of the simulation is that the rate of dissolution of minerals used depends upon the resi-dence time of the solution (i.e.,migration rate).Therefore,the results of the simulation would be a ected by the rate of the injection.

The other implication of the present results to CO 2ˉooding in the Tensleep Formation is the e ect of the chemistry of the injection solution on the increase in porosity.The increase in porosity observed in Run 2was greater than in Run 1(Table 2).In Run 2,in which a NaCl solution was used,the dissolution of both dolomite and anhydrite contributed to the increase in porosity.However,in Run 1,the increase in porosity by dolomite dissolution was compensated by the precipitation of anhydrite because the injection solution was saturated with respect to anhydrite at run conditions.

Tensleep Formation is cemented by dolomite and anhydrite (Todd,1964;Mankiewicz and Steidmann,1979)and formation waters are saturated with anhy-drite at reservoir conditions,whereas they are supersa-turated with respect to dolomite (Shiraki and Dunn,1995;Dunn et al.,1994;Dunn et al.,1995).(The satur-ation state of carbonates may be overestimated,because a fair amount of carbon dioxide outgases from water samples from the water steam when they are col-lected at wellheads in the ?eld.)Unlike carbonates,the saturation state of anhydrite is not a ected signi?-cantly by increasing P CO 2.Therefore,anhydrite would precipitate as Ca is released by the dissolution of dolo-mite if a Tensleep Formation water is used as the injection water,as observed in Run 1.Porosity increase due to CO 2ˉooding would be smaller than when a solution undersaturated with respect to anhy-drite is injected.Thus,more e ective injection sol-utions in terms of porosity increase during CO 2ˉooding can be obtained from other formations which do not contain sulfate cement minerals.7.Conclusions

The present study experimentally examined the

nature of water±rock interactions during CO 2ˉooding in dolomite-and anhydrite-cemented reservoir sand-stones under typical reservoir conditions in Wyoming,USA.

The major reactions observed were the dissolution of dolomite,the precipitation or dissolution of anhy-drite,and the alteration of K-feldspar to form kaoli-nite.All solutions were saturated with respect to barite throughout both runs.Core porosity increased when both dolomite and anhydrite dissolved into the sol-ution,but no signi?cant change in porosity was observed when anhydrite precipitation accompanied dolomite dissolution.The permeability of the cores (except the middle core used in Run 2)decreased through runs even when the porosity increased.Thermodynamic calculations and SEM examination showed that kaolinite was a stable mineral phase under the experimental conditions,suggesting that the decrease in permeability was due to the growth of kao-linite crystals in pore spaces,which restricted ˉuid ˉow.

Simulations using experimental data showed that the injected solution reached saturation with respect to dolomite and anhydrite in the immediate vicinity of the injection well.The porosity increase due to the injection of a solution that was undersaturated with respect to anhydrite was greater than that due to a sol-ution saturated with respect to anhydrite from the Tensleep Formation.Acknowledgements

We thank T.G.Monger-McClure of Marathon Oil Company for permission to use their core-ˉooding ap-paratus and H.Haines,G.Kennedy,and M.Jessee (MOC)for their technical support during the exper-iments.Porosity and permeability of cores were measured at MOC,and S.Boese of the University of Wyoming analyzed sample solutions with ICP and IC.S.Stookey (UW)helped us with examination of core mineralogy.Editorial improvements were provided by K.Kirkaldie,Institute for Energy Research.We also thank W.Gunter and B.Hitchon for their helpful reviews.This research was supported by the U.S.Department of Energy (DE-AC22-93BC14897).References

Beeson,D.M.,Orthlo ,G.D.,1959.A laboratory investi-gation of the water-driven carbon dioxide process for oil recovery.Amer.Inst.Min.Eng.Trans.216,388±391.

Bowker,K.A.,Shuler,P.J.,1991.Carbon dioxide injection and resultant alteration of the Weber sandstone,Rangely ?eld,Colorado.Amer.Assoc.Petrol.Geol.Bull.75,1489±1499.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

278

Busenberg,E.,Clemency,C.V.,1976.The dissolution kinetics of feldspars at 258C and 1atm CO 2partial pressure.Geochim.Cosmochim.Acta 40,41±49.

Busenberg,E.,Plummer,L.N.,1982.The kinetics of dissol-ution of dolomite in CO 2±H 2O systems at 1.5to 658C and 0to 1atm P CO 2.Amer.J.Sci.282,45±78.

Dunn,T.L.,Iverson,W.P.,Crabaugh,M.,Shiraki,R.,1994.Anisotropy and spatial variation of relative permeability and lithologic character of Tensleep sandstone reservoirs in the Bighorn and Wind River basins,Wyoming.Department of Energy Contract No,DE-AC22-93BC14897:First Annual Progress Report.

Dunn,T.L.,Iverson,W.P.,Crabaugh,M.,Shiraki,R.,1995.Anisotropy and spatial variation of relative permeability and lithologic character of Tensleep sandstone reservoirs in the Bighorn and Wind River basins,Wyoming.Department of Energy Contract No,DE-AC22-93BC14897:Second Annual Progress Report.

Helgeson,H.C.,1971.Kinetics of mass transfer among sili-cates and aqueous solutions.Geochim.Cosmochim.Acta 35,421±469.

Helgeson,H.C.,Delany,J.M.,Nesbitt,H.W.,Bird,D.K.,1978.Summary and critique of the thermodynamic proper-ties of rock-forming minerals.Amer.J.Sci.278A,1±229.Helgeson,H.C.,Kirkham, D.H.,Flowers,G.C.,1981.Theoretical prediction of the thermodynamic behavior of aqueous electrolytes at high pressures and temperatures.VI.Calculation of activity coe cients,osmotic coe cients,and apparent molal and standard and relative partial molal properties to 6008C and 5kb.Amer.J.Sci.281,1249±1516.

Holdren Jr.,G.R.,Berner,R.A.,1979.Mechanism of feldspar weatheringDI.Experimental studies.Geochim.Cosmochim.Acta 43,1161±1171.

Holm,L.W.,Josendal,V.A.,1974.Mechanism of oil displace-ment by carbon dioxide.J.Petrol.Tech.26,1427±1438.Hsu,K.J.,1963.Solubility of dolomite and composition of Florida ground waters.J.Hydrol.1,288±310.

Kharaka,Y.K.,Gunter,W.D.,Aggarwal,P.K.,Perkins,E.H.,De Braal,J.D.,1988.SOLMINEQ.88:a computer program code for geochemical modeling of water±rock in-teractions.In:US Geol.Surv.Water Res.Investigations Rept,pp.88±4227.

Langmuir,D.,1971.The geochemistry of some carbonate ground waters in central Pennsylvania.Geochim.Cosmochim.Acta 35,1023±1045.

Mankiewicz,D.,Steidmann,J.R.,1979.Depositional environ-ments and diagenesis of the Tensleep sandstone,eastern Big Horn basin,Wyoming.Soc.Econ.Paleontol.Mineral.Spec.Issue 26,319±336.

Monger,T.G.,Fu,J.C.,1987.The nature of CO 2-induced or-ganic deposition.Proc.1987SPE Annual Tech.Conf.,147±159(SPE 16713).

Monger,T.G.,Ramos,J.C.,Thomas,J.,1991.Light oil recovery from cyclic CO 2injection:inˉuence of low press-ures,impure CO 2,and reservoir gas.SPE Reservoir Eng.6,25±32.

Monger,T.G.,Trujillo,D.E.,https://www.wendangku.net/doc/ce6762158.html,anic deposition during CO 2and rich-gas ˉooding.SPE Reservoir Eng.6,17±24.Omole,O.,Osoba,J.S.,1983.Carbon dioxide±dolomite rock interaction during CO 2ˉooding process.Paper 83-34-17

Petrol.Soc.Canadian Inst.Min.Metal.1±13,Ban ,Alberta.

Palmer,F.S.,Landry,R.W.,Bou-Mikael,S.,1986.Design and implementation of immiscible carbon dioxide displace-ment projects (CO 2Hu -Pu )in south Louisiana.Proc.61st SPE Annual Tech.Conf .1±10(SPE 15497).

Patterson,K.W.,1979.Fighting downhole corrosion and scale in ˉood CO 2at SACROC.Petrol.Eng.Intl.50,36±44.

Robie,R.A.,Hemingway, B.S.,Fisher,J.R.,1978.Thermodynamic properties of minerals and related sub-stances at 298.15K and 1bar (105Pascals)pressure and at higher temperatures.U.S.Geol.Surv.Bull.1452,456.Ross,G.D.,Todd,A.C.,Tweedie,J.A.,1981.The e ect of simulated CO 2ˉooding on the permeability of reservoir rocks.In:Proc.3rd European Symp.Enhanced Oil Recovery,pp.351±366.

Ross,G.D.,Todd,A.C.,Tweedie,J.A.,Will,A.G.S.,1982.The dissolution e ects of CO 2-brine systems on the per-meability of U.K.and North Sea calcareous sandstones.Proc.3rd Joint SPE/DOE Symp.Enhanced Oil Recovery 149±162.(SPE/DOE 10685).

Sayegh,S.G.,Krause,F.F.,Girard,M.,DeBree,C.,1990.Rock/ˉuid interactions of carbonated brines in a sand-stone reservoir:Pembina Cardium,Alberta,Canada.SPE Formation Eval.5,399±405.

Shiraki,R.,Dunn,T.L.,1995.Origins of Tensleep Formation water chemistry and prediction of scale formation during CO 2treatments,Bighorn Basin,Wyoming.Amer.Assoc.Petrol.Geol.Abst.Prog 88A.

Shuler,P.J.,Freitas,E.A.,Bowker,K.A.,1989.Selection and application of barium sulfate scale inhibitors for a carbon dioxide ˉood,Rangely Weber sand unit,Rangely,Colorado.Proc.SPE Joint Rocky Mountain Regional/Low Permeability Reservoirs Symp.451±464(SPE 18973).Smith,L.K.,1993.Aspects of oil?eld water chemistry:Part A.Induced diagenetic alteration and Part B.The e ect of organic alkalinity on cation rations.Ph.D.thesis,Univ.of Wyoming.

Talman,S.J.,Gunter,W.D.,1992.Rates of dolomite dissol-ution in CO 2and HCl bearing solutions from 100to 2008C.In:Proc.7th Intl.Symp.Water±Rock Interaction,pp.119±122.

Thomas,C.E.,Mahoney,C.F.,Winter,G.W.,1992.Water-injection pressure maintenance and waterˉood processes.In:Bradley,H.B.(Ed.),Petroleum Engineering Handbook.Soc.Petrol.Eng.Chapt.44,pp.1±52.

Todd,T.W.,1964.Petrology of Pennsylvanian rocks,Bighorn basin,Wyoming.Bull.Amer.Assoc.Petrol.Geol.48,1063±1090.

Wolcott,J.M.,Monger,T.G.,Sassen,R.,Chinn,E.W.,1989.The e ects of CO 2ˉooding on reservoir mineral proper-ties.Proc.1989SPE Intl.Symp.Oil?eld Chem.101±110(SPE 18467).

Wolcott,J.M.,Groves,F.R.Jr,Trujillo,D.E.,Lee,H.G.,1991.Investigation of crude-oil/mineral interactions:fac-tors inˉuencing wettability alteration.Proc.1991SPE Intl.Symp.Oil?eld Chem.411±420(SPE 21042).

Wollast,R.,1967.Kinetics of the alteration of K-feldspar in bu ered solutions at low temperature.Geochim.Cosmochim.Acta 31,635±648.

R.Shiraki,T.L.Dunn /Applied Geochemistry 15(2000)265±279

279