Analysis of Acid Returns

Ž .Journal of Petroleum Science and Engineering 28 2000 33–53www.elsevier.nlrlocaterjpetscieng

Analysis of acid returns improves efficiency of acid stimulation:a case history

K.C. Taylor a,), H.A. Nasr-El-Din a,1, Ragheb B. Dajani b,2

a Laboratories R&D Center, P.O. Box 62, Saudi Aramco, Dhahran 31311, Saudi Arabiab BJ SerÕice Arabia Ltd., P.O. Box 14037, Dammam 31424, Saudi Arabia

Received 24 June 1999; accepted 26 June 2000

Abstract

Techniques are demonstrated to evaluate acid stimulation treatments based on chemical analysis of the return fluids. Thetechniques can provide information about the source of formation damage, the amount and type of scale in the casing, theefficiency of the acid treatment, and the amount of iron control chemicals required for similar treatments. As an example,

Ž .these methods were used for the acid stimulation treatment of a typical seawater injection well Well A in a large carbonatereservoir in Saudi Arabia. Over a 2-day period, more than 50 field samples were collected. More than 700 chemical analyseswere eventually made on these samples. Analysis of the casing pickle fluids showed that 235 kg of iron from corrosion

Ž .products, as well as a significant amount of acid-insoluble iron carbide, Fe C, was produced. Measurement of iron II and3Ž .iron III concentrations showed that the iron corrosion products in the casing contained mostly iron oxides such as rust or

mill scale. The pickle treatment prevented the injection of this material into the formation around the wellbore. Otherwise,the main acid treatment would have flushed all of this material into the reservoir and reduced injectivity. Analysis of thematrix acid flowback samples showed that iron control chemical concentrations in the acid could be reduced with significantcost savings. The extent of acid dilution and the amount of unreacted acid were also determined. The results showed that 125kg of calcium sulfate was dissolved from the near-wellbore region during flowback. Acid stimulation of Well A increased

Ž . 3 Ž .the injectivity index II of the well from 0.12 to 0.21 m rdayrkPa 5 to 9 bblrdayrpsi , an increase of 75%. The acidstimulation was successful and the evaluation methods provided useful information to improve the success of futuretreatments. q 2000 Elsevier Science B.V. All rights reserved.

Keywords: injection well; stimulation; formation damage; iron control chemicals; pickling acid

) Corresponding author. Fax: q966-3-876-2811.E-mail addresses: [email protected]

Ž . Ž .K.C. Taylor , [email protected] H.A. Nasr-El-Din ,Ž [email protected] R.B. Dajani .

1 Fax: q966-3-876-2811.2 Fax: q966-3-858-8128.

1. Introduction

The seawater injection system for a large carbon-ate field in Saudi Arabia has more than 400 km of

3 Žpipeline and can supply 0.67 million m rday 4.2.million barrels per day of treated seawater. In this

0920-4105r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S0920-4105 00 00066-8

( )K.C. Taylor et al.rJournal of Petroleum Science and Engineering 28 2000 33–5334

system, corrosion must be maintained at a reasonablelevel, and water quality must be high enough to

Žprevent excessive losses of injectivity Brown et al.,.1979; Bayona, 1993 . Extensive sections of the

pipeline are internally coated to reduce corrosion.Water quality standards have been established andare shown in Table 1. Corrosion control in water

Ž .injection systems was reviewed by Patton 1993 .Ž .Chen and Ahmed 1998 recently gave a review of

the coating effectiveness and water quality.Even with these measures, injectivity of some

wells in the reservoir has gradually decreased andacidizing treatments only partially restore injectivityŽ .Bayona, 1993; Chen and Ahmed, 1998 . As thecumulative water injection increases, the acidizingeffectiveness decreases. To address these trends, adetailed investigation of an acidizing treatment of atypical injection well in the field was initiated.

Most water injection wells in this carbonate reser-voir are vertical with openhole completions over an

Ž .interval of 60 to 120 m 200 to 400 ft . Static bottomŽ .hole temperatures are 498C to 658C 1208F to 1508F .

Static bottom hole pressure is approximately 27.5Ž . Ž .MPa 4000 psig at 2271 m 7450 ft . Treated seawa-

Ž .ter Table 2 is injected to maintain the reservoirpressure.

To restore injectivity, damaged wells are stimu-lated using 15 wt.% HCl. Various acid diversion

Ž .techniques have been used. Ginest et al. 1993evaluated three different acid placement techniquesin the carbonate reservoir. These included conven-tional bullhead treatments, coiled tubing with nitro-gen foam diversion, and coiled tubing without divert-

Table 1Ž .Water quality requirements for seawater injection Bayona, 1993

Component Specifications

Total suspended solids -0.2 mgrlParticle number and size Maximum of 200 particles

greater than 2 mmper 0.5 ml of water

aDissolved oxygen -10 ppbSulfide -14 mgrl

bpH 7.0–7.5

a Ž .This is in agreement with Patton 1993 .bpH is adjusted to prevent the precipitation of calcium

carbonate.

Table 2Analysis of injected seawater and low sulfate water used inWell A

Component Injected seawater Water usedŽ .mgrl in Well

Ž .A mgrl

Sodium 16,870 210Calcium 700 137Magnesium 2,040 48Sulfate 4,150 270Chloride 30,040 371Bicarbonate 160 210Total dissolved solids 53,960 1,257Iron -0.01 1pH 7.5 8

3Density, grcm at 1.0418 1.0011Ž .15.58C 608F

ing agents. The goal of the work was to improve theefficiency of stimulation of the long openhole sec-tions in the formation as a result of drilling muddamage. They concluded that coiled tubing treat-ments with nitrogen foam diversion were more effec-tive than conventional bullhead jobs; however, sur-face washout does not currently allow the use ofcoiled tubing in some of the wells in this reservoir.

The gradual injectivity decline observed in thesewells may be a result of near-wellbore plugging ordeep damage by particulates over an extended periodof time. Possible sources of this damage include the

Ž .precipitation of iron III hydroxide from previousŽ .acidizing treatments Crowe, 1985 , the formation of

calcium sulfate scale from the seawater that is in-Ž .jected Allaga et al., 1992 or the formation of iron

sulfide and biomass as a result of microbial corro-Ž . Ž .sion reactions Lee et al., 1995 . Allaga et al. 1992

showed that calcium sulfate precipitation inside asandpack resulted in a gradual permeability decline.Iron sulfide has been identified as a corrosion prod-

Žuct in seawater injection systems Chen and Ahmed,.1998; Little et al., 2000 . The plugging of the forma-

tion over time with both acid soluble and acid-in-soluble particulates would be consistent with the

Ž .type of injectivity loss observed by Bayona 1993 .Water injection wells can contain large amounts

of iron scale from pipe corrosion. This corrosionoccurs from oxygen in the injection water or frommicrobial action. If acid is bullheaded into a water

( )K.C. Taylor et al.rJournal of Petroleum Science and Engineering 28 2000 33–53 35

injection well, iron scale will rapidly react with theacid. Total iron concentrations up to 100,000 mgrlhave been reported in acid returns from other water

Ž .injection wells Gougler et al., 1985 .Corrosion in the casing can be present in several

forms. If small amounts of oxygen are present in theinjected water, then iron oxides will form. In air, thisrust or mill scale contains three layers of iron oxideŽ .Coburn, 1984 . The outer layer is primarily ferricoxide, Fe O . The layer closest to the steel is ferrous2 3

oxide, FeO. The intermediate layer can be repre-sented as Fe O or FeOPFe O . In the latter, the3 4 2 3

Ž . Ž .ratio of iron III to iron II is two to one. If 15wt.% HCl reacts with sufficient Fe O to give a3 4

solution containing 45,000 mgrl total iron, then theŽ .reacted acid will contain 30,000 mgrl iron III and

Ž .15,000 mgrl iron II . Acid reaction with rust re-duces the acid concentration in this example from 15to 7 wt.%. This dissolved iron can lead to pluggingin the critical near-wellbore region.

Ž .Two reactions can occur with iron III that canlead to precipitation of iron compounds. If hydrogen

Ž .sulfide is present, then iron III will react to formŽelemental sulfur Crowe, 1985; Taylor and Nasr-El-

.Din, 1999 . In the absence of hydrogen sulfide, ironŽ .III hydroxide will precipitate as the pH increases

Žabove 1 Taylor et al., 1999a, Taylor and Nasr-El-. Ž .Din, 1999 . Iron II will not precipitate as a hydrox-

ide until the pH increases to a value of about 7Ž .Crowe, 1985 .

Iron control chemicals are not able to maintainŽ .iron III concentrations above 10,000 mgrl in spent

acidizing fluids. For instance, to keep 10,000 mgrlŽ . Ž .iron III in solution at 528C 1258F requires 22 l of

acetic acid and 12 kg of citric acid per 1000 m3 ofŽ .15% HCl Dill and Smolarchuk, 1988 . It is not

recommended to exceed this level of acetic acid andcitric acid, since precipitation of calcium citrate mayoccur. Calciumriron complexes with citric acidformed at high citric acid concentrations in partially

Ž .spent acids Taylor et al., 1999a .Several chemical studies of acid flowback sam-

Ž .ples have been reported. Gdanski and Peavy 1986 ,Ž . Ž .Almond et al. 1990 , Shuchart 1995 and Nasr-El-

Ž .Din et al. 1996 examined acid return samples inŽ .sandstone reservoirs. Nasr-El-Din et al. 1999 ,

Ž .Dahlan and Nasr-El-Din 2000 and Mohamed et al.Ž .1999 examined acid returns in carbonate reservoirs.

No comprehensive chemical studies of acid flow-back samples from water injection wells have beenpublished, to the best of our knowledge. The objec-

Ž .tives of this study are to: 1 demonstrate techniquesto evaluate acid stimulation treatments based on

Ž .chemical analysis of the return fluids, 2 use thesetechniques in Well A to identify the source of forma-tion damage, the amount and type of scale in the

Ž .casing, and the efficiency of the acid treatment, 3evaluate the amount and type of iron control chemi-

Ž .cals that were used, and 4 make recommendationsto improve future similar treatments.

1.1. Well description

Well A is a seawater injection well in a largecarbonate reservoir in Eastern Saudi Arabia. Theinjectivity of this well has been decreasing sinceinjection began in 1993. The injectivity index de-

3 Ž .creased from 0.53 m rdayrkPa 22 bblrdayrpsiŽ . 3 ŽNov.r93 to a value of 0.12 m rdayrkPa 5 bblr

.dayrpsi before the treatment. The damage to thiswell is significant and has been occurring over anumber of years.

The well has an openhole completion in the inter-Ž .val from 2215 to 2324 m 7268 to 7626 ft . The wellŽ .was acidized using 3.8 cm 1.5 in. coiled tubing in

1992 to improve well injectivity. At that time, 15 m3

Ž .95 barrels of 15 wt.% HCl containing a surfactantŽ . Ž .0.2 vol.% , a corrosion inhibitor 0.4 vol.% and a

Ž .friction reducer 0.2 vol.% were injected into theformation. No iron control chemicals were used andthe well was not flowed back after the treatment.Instead, the reacted acid was displaced into theformation using seawater.

ŽA caliper log indicated a washout approximatelyŽ .0.6 m 25 in. in the zone from 2215 to 2217 m

Ž .6751 to 6757 ft . A spinner test indicated that mostof the injected water was going into a narrow zone

Žthat extended from 2228 to 2256 m 6791 to 6876.ft .

Ž .The well has a 17.8 cm 7 in. liner from theŽ .openhole section to 742 m 2262 ft . All of the

casing in the well is J-55 low carbon steel. The3 Ž .wellbore volume is nearly 61 m 383 barrels and

Ž .the static bottom hole temperature is 588C 1378F . Atypical geochemical analysis of the seawater injectedinto Well A is given in Table 2.


1.2. Description of the treatment

Mixing and water tanks were inspected, and sam-ples were taken of all chemicals on site before thetreatment began. The treatment chemicals collectedat Well A were:

Ø Breakerriron controlŽ . Ž .Ø Erythorbic acid, reduces iron III to iron II

Ž .Ø Citric acid, prevents precipitation of iron III byforming a soluble complex in spent acid

Ø Cross-linking agentØ Surfactant, to improve the contact between the

live and spent acids with reservoir rockØ Polymer, to increase the viscosity of the in situ

gelled acidØ Mixed samples of the pickle acid containing all

additivesØ A mixed sample of the 10 wt.% in situ gelled acid

containing all additivesØ A mixed sample of the 15 wt.% acid containing

all additivesØ Low salinity water for mixing of acids, pre-flush

and post-flush

The corrosion inhibitor was already present in theacid when it arrived on site.

On the first day of the treatment, the well wasflowed back using reservoir pressure to flush out thewellbore. Samples were collected and the flowratewas determined by measuring the time required to

Ž 3.fill a container of known volume 0.16 m . The rate3 Ž .was found to be 22.3 m rh 140 bblrh .

On the second day, the casing was pickled beforethe matrix acid stimulation treatment. Compositionsof the fluids that were used in the treatment aregiven in Table 3. Treatment stages, volumes, pump-ing rates and surface pressures are listed in Table 4.

The pickle treatment was performed to removerust and scale from the casing. This part of thetreatment was designed so that penetration of theacid into the formation would be minimal. The acidŽ 3. 38 m was displaced down the casing with 50 mŽ .315 barrels of a low salinity low sulfate water untilthe acid was close to the openhole section. The wellwas then flowed back to flush the spent acid fromthe casing. Chemical analysis of the low sulfatewater used in the treatment is given in Table 2.

To enhance acid placement efficiency, matrixstimulation of the well was carried out in 15 stagesŽ .Table 4 . The first stage used low sulfate water as aspacer between the reservoir water and the acid. Thisreduces mixing of the spent acid with seawater.Seawater contains high concentrations of sulfate,while spent acid contains high levels of calcium.Mixing of the two fluids leads to precipitation ofcalcium sulfate that can damage the formation in thecritical near-wellbore area.

The low sulfate water pre-flush was followed byalternating slugs of 10 wt.% in situ gelled HCl andregular 15 wt.% HCl. The in situ gelled acid uses apolymer and a cross-linking agent to form a highviscosity acid over a narrow pH range. The purposeof the in situ gelled acid was to provide acid diver-sion and increase the flow of acid into the lower

3 Žpermeability zones. Approximately 0.96 m rm 77

Table 3Acid formulation for Well A

Fluid Chemical additives

Ž .15 wt.% HCl acid for pickling 0.3 vol.% corrosion inhibitoraPre-flush 0.2 vol.% surfactant in low sulfate water

b b15 wt.% HCl 0.3 vol.% corrosion inhibitor, 1.1 vol.% acetic acid , 0.56 wt.% citric acid , 0.2 vol.% surfactantIn situ gelled 10 wt.% HCl 0.3 vol.% corrosion inhibitor, 2.0 vol.% gelling agent, 0.045 wt.% breakerriron control,

0.2 vol.% buffer, 0.45 wt.% cross-linkeraOverflush Low sulfate water

aComposition of low sulfate water is given in Table 2.b Ž .This amount of acetic acid and citric acid can keep approximately 5000 mgrl of iron III in solution in spent acid at reservoir

temperature.


Table 4Treatment stages, volumes, rates and surface pressures for Well A

3 3Ž . Ž . Ž .Stage of treatment Fluid Volume m Pumping rate m rmin Surface pressure MPa

Well pickle 15 wt.% HCl 8.0 1.1 7.6Pickle post-flush Low sulfate water 50 2.4 12.8Pre-flush Low sulfate water 60 3.2 4.11 15 wt.% HCl 2.2 1.3 15.22 10 wt.% gelled acid 1.9 1.3 14.53 15 wt.% HCl 2.7 1.4 14.54 10 wt.% gelled acid 2.2 1.6 14.55 15 wt.% HCl 4.5 1.7 14.86 10 wt.% gelled acid 4.0 1.9 15.27 15 wt.% HCl 5.1 1.9 15.28 10 wt.% gelled acid 4.5 1.9 15.29 15 wt.% HCl 5.7 2.1 15.210 10 wt.% gelled acid 5.1 1.9 15.211 15 wt.% HCl 4.5 2.1 15.512 10 wt.% gelled acid 5.7 1.9 15.513 15 wt.% HCl 10.6 2.1 15.514 Overflush 77 1.3–3.3 16.5

3Ž .Total 15 wt.% acid m 373Ž .Total in situ gelled 10 wt.% acid m 23

.galrfoot of acid was injected into the openholesection. The treatment was followed by injection of

3 Ž .approximately 77 m 483 barrels of low sulfatewater. The well was shut in for 70 min for reactionof the acid with the formation before it was flowedback.

Over a 2-day period, more than 50 field sampleswere collected. More than 700 chemical analyseswere eventually made on these samples.

2. Experimental studies

2.1. Field measurements

Total dissolved iron and ferrous iron were mea-sured at the wellsite using the procedure of Taylor et

Ž .al. 1999b . Ferrous iron was measured in the fieldŽ . Ž .because iron II is rapidly oxidized to iron III in

the presence of air. Acid content of the injected acidswas measured by titration of a known acid volumewith 1 N sodium hydroxide solution to a phenolph-

Ž .thalein endpoint pH 9 . Accurate volumes weremeasured using calibrated autopipettes. Density wasmeasured with a portable densitometer. A portable

pH meter was used at the well site and was cali-Ž .brated with buffers pH 4 and pH 7 before sample

measurement.

2.2. Laboratory measurements

Calcium, magnesium, iron, sodium, aluminum andzirconium concentrations were measured by induc-tively coupled argon plasma emission spectroscopyŽ .ICP . Acid content of flowback samples was mea-sured by titration of a known acid volume with 0.1 Nsodium hydroxide solution to an endpoint of pH 4.2using an autotitrator. Chloride ion was measuredwith an autotitrator using 0.1 N silver nitrate solu-tion. Sulfate ion was measured by turbidity with 0.1N barium chloride solution. Density measurementswere made with a digital density meter at 15.58C.The pH was measured with an AgrAgCl singlejunction pH electrode. Total dissolved solids weredetermined by addition of the concentration of thechemical species measured. Analysis of filtered sus-pended solids was done using X-ray diffractionŽ .XRD and by the energy dispersive X-ray spectrom-eter attachment of a scanning electron microscopeŽ .SEM .


3. Results and discussion

3.1. Chemicals used during the treatment

ŽWater of low salinity and sulfate content Table.2 was used in the treatment for the pre-flush, for

mixing the acids and for the post-flush. The chemicalcomposition of the injected fluids is shown in Table5. Two different samples of the pickling acid weretaken after it was mixed to demonstrate the repro-ducibility of the measurement methods. The acidconcentration was very close to the target value of15 wt.%, and the measurement made at the well sitewas in good agreement with the value measured inthe lab.

The 10 wt.% in situ gelled acid contained 180mgrl aluminum and 325 mgrl zirconium from thecross-linking additive. The sulfate concentration, at924 mgrl, was much higher than the value of 270mgrl found in the mixing water. This additionalsulfate is either from the acid additives or from aninterference in the method.

All four of the injected acid samples had low ironconcentrations with values from 44 to 84 mgrl. Thisshows that some iron contamination occurred duringtransport and mixing of hydrochloric acid.

( )3.2. Well flowback before acid treatment

This stage of the treatment was included to flushout any suspended material in the wellbore. Flowrate was measured at the beginning and at the end of

3 Žthe flowback, and was 22.3"1.5 m rh 140"10

.barrelsrh . This measurement was extremely usefulin later calculations of the casing pickle fluid returns.If the flowrate of the well is known, then materialbalance calculations can be conducted using the con-centrations of produced chemical species. Based onthe well flowrate, 2.75 h should be required to flushout the wellbore volume of 383 bbls.

Fig. 1 shows the concentrations of sulfate, magne-sium, calcium and bicarbonate in the produced flu-ids. The sulfate concentration shows a large increaseafter 1.4 h of production. The concentration of sul-fate in injected seawater is approximately 4150 mgrland is shown in the figure. An increase in calciumconcentration also occurs 1.5 h after flowback. The

Žvalue increases from approximately 650 mgrl the.value expected in seawater to approximately 1500

mgrl. In addition, an increase in density of theŽ .flowback fluids Fig. 2 and the TDS occurs at the

same time as the increase in the calcium and thesulfate concentrations. No change is observed in themagnesium or bicarbonate concentrations. At the

Žsame time as the density increase, the pH measured. Ž .at the wellsite gradually decreases Fig. 2 . If the

observed increases in sulfate concentration was dueonly to calcium sulfate, then the calcium concentra-tion would increase by approximately 875 mgrl.This value is very close to the increase of 850 mgrlthat was observed. Therefore, it is likely that theincrease in the calcium and sulfate concentrations isdue to the dissolution of calcium sulfate. The con-centration of calcium sulfate increased at 1.5 h after

3 Žflowback started, corresponding to 33.4 m 210.barrels of brine production. This is lower than the

Table 5Chemical analysis of acid injection fluids

Variable Pickle acid a1 Pickle acid a2 10 wt.% in situ gelled acid 15 wt.% acid

Ž .Aluminum mgrl -10 -10 180 -10Ž .Calcium mgrl 116 110 214 112

Ž .Magnesium mgrl 30 30 39 30Ž .Sodium mgrl 146 146 418 145Ž .Chloride mgrl 148,544 148,125 85,599 138,928

Ž .Sulfate mgrl 535 481 924 438Ž .Zirconium mgrl -10 -10 325 -10Ž .Total iron mgrl 44 44 59 84

3Ž .Density at 15.58C grcm 1.0722 1.0723 1.0561 1.0716Wtrwt.% HCl at wellsite 14.21 14.75 9.67 13.50Wtrwt.% HCl at lab 14.21 14.20 9.67 13.50


Ž .Fig. 1. Flowback analysis before acid treatment : sulfate, calcium, magnesium and bicarbonate.

Ž 3 Ž ..volume of the wellbore 60.9 m 383 barrels ,indicating that the calcium sulfate may be present asa scale on the lower part of the casing. Calciumsulfate may form a scale due to the temperatureincrease that occurs when the seawater flows down

the wellbore to the formation. The solubility ofcalcium sulfate in seawater gradually decreases as

Ž .the temperature is increased Silcock, 1979 . In de-salination plants, untreated seawater will usually pro-duce a calcium sulfate scale when heated above 778C

Ž .Fig. 2. Flowback analysis before acid treatment : pH and density.


Ž . Ž .1708F Cowan and Weintritt, 1976 . The reservoirŽ .temperature is 938C 2008F , but the bottomhole

Ž .temperature at well A is now only 588C 1378F dueto long-term water injection. If water is producedfrom the well, then the seawater in the formationflowing back to the wellbore will be cooled. If itflows through an area of precipitated calcium sulfateor anhydrite, then it will dissolve some of it. It isthen possible for the seawater flowing out of theformation to contain more calcium and sulfate thanthe concentration seen in the injected seawater.

The amount of calcium sulfate dissolved by theflowback fluid was calculated as 125 kg. The calcu-

3 Ž .lation used a flow rate of 22.3 m rh 140 bblrhand an injection water sulfate concentration of 4150mgrl.

Chloride and sodium concentrations in the flow-back samples did not change significantly during theflowback, and were very close to those expected in

Ž .seawater Table 2 . During the last hour of the wellflowback stage, the average chloride and sodiumconcentrations were 30,873 and 16,972 mgrl, re-spectively. The weight ratio of chloride to sodium inthe injected brine is 1.82 to 1, based on these values.

Total iron and ferrous iron concentrations in theproduced fluids were measured and were less than10 mgrl. No suspended solids were present in the

flowback samples. In general, the samples were clear.Ž .Sulfate-reducing bacteria SRB are known to be

Žpresent in the seawater injection system Bayona,.1993 . The samples collected from 3 h until the end

of the flowback were initially clear and had a slightodor of hydrogen sulfide. The iron concentration inthese samples was 2 mgrl or less. Within 2 days,they had become cloudy with a light yellow color. Itis likely that this material was elemental sulfurformed by the oxidation of hydrogen sulfide by

Ž .oxygen, see for example Little et al. 2000 .

3.3. Casing pickle fluid returns

The pickling acid contained only a corrosion in-hibitor as an additive, and the acid concentration wasnearly 15 wt.%. The chemical composition of thepickling acid is shown in Table 5.

Fig. 3 shows the acid concentration and the totaliron concentration in the flowback samples. Themaximum acid concentration in the samples was 9.7wt.%, while the maximum total iron concentrationwas 23,600 mgrl. The dissolution of rust from thecasing can account for most of the decrease noted inthe acid concentration, as will be described in thenext paragraph. This means that the acid was not

Ž .Fig. 3. Flowback analysis pickle treatment : hydrochloric acid and total iron.


significantly diluted during the pickle treatment.Some dispersion is observed in Fig. 3 as the acidconcentration profile is slightly broadened. The flowwas turbulent as the acid was pumped into the casingand during the flowback.

The composition of the dissolved iron that flowedback following the pickle treatment is shown in Fig.3. The total iron content was measured by twodifferent methods. ICP spectroscopy was used in the

Ž .lab analyses. Total iron and iron II concentrationswere also measured by the field method of Taylor et

Ž .al. 1999b after the samples were brought back tothe lab. The results obtained by the field method arein excellent agreement with the ICP method. Maxi-

Ž .mum concentration of iron II was 10,000 mgrl.Measurements in the field with and without samplefiltration gave results that agreed within 5%. Themeasured maximum iron concentrations give a ratio

Ž . Ž .of iron II to iron III of 0.67 to 1, compared to aŽ .theoretical ratio of 0.5 to 1 in Fe O FeOPFe O .3 4 2 3

Ž . Ž .The ratio of iron II to iron III produced during theentire flowback of the pickle fluids was 0.9 to 1.This was calculated using the iron concentrations inthe produced fluids and the flowback rate of thewell. The source of this iron could be gradual surfacecorrosion before the pickle treatment, or from rustthat was already present in the casing. In addition to

Ž .millscale or rust, a source of iron II may be pre-sent, most likely as FeS. The latter is usually presentas a byproduct in systems contaminated with SRBŽ .Nasr-El-Din et al., 1996; Little et al., 2000 .

Several research groups have reported the mea-Ž . Ž .surement of ratios of iron II to iron III during

various acidizing operations. In field samples, SmithŽ . Ž .et al. 1969 observed that the ratio of iron II to

Ž .iron III in spent acidizing fluids varied from 5:1 to10:1. In contrast, Well A showed an average ratio of

Ž . Ž .iron II to iron III of 0.9 to 1. The relative amountŽ .of iron III in this well was much higher than

Ž .previously reported results. More details on iron IIŽ . Ž .to iron III ratios are given by Loewen et al. 1990

Ž .and Taylor et al. 1999b .From Fig. 3, the mass of iron dissolved from

corrosion products during the pickle treatment can becalculated, since the flow rate of the well duringflowback is known. A total of 235 kg of iron wasproduced in the pickle flowback samples. Less than0.5 kg of iron was present in the injected pickle acid.

Scale deposits on the casing were dissolved by theacid, since direct reaction of steel and hydrochloric

Ž . Ž .acid produces only iron II Kaesche, 1985 . If ironsulfide was the dissolved corrosion product, then

Ž .only iron II would be produced. The high amountŽ .of iron III present clearly indicates that an oxygen-

containing corrosion product was present as the ma-jor scale component. This corrosion product couldhave been present as mill scale when the casing wasoriginally put in place. If oxygen was present in theinjection water, rust could also have formed overlong periods of time.

The returning acid samples were black in colorand contained suspended solids. These solids were

Ž .isolated by filtration 0.45 mm , dried, weighed, andanalyzed. Fig. 4 shows the concentration of solids byfiltration together with the acid concentration in theproduced fluids. The filtered material was a fineblack powder and was collected from samples withacid concentrations up to nearly 10 wt.%. XRDanalyses of selected samples are shown in Table 6. Itis important to emphasize that the XRD results showcrystalline material only. Based on these tests, manyof the samples contained iron carbide as a majorcomponent. However, the sample at 2.2 h into theflowback contained major amounts of akaganeiteŽ Ž ..ß-FeO OH . This mineral should be acid soluble. Itmay have occurred in the sample by oxidation of aniron compound after the sample was isolated. At 2.9h into the flowback, acid is no longer produced andthe solids contain mostly silica, gypsum, and clays.

The approximate compositions of the suspendedsolids are shown in Table 7. These results were madeby energy dispersive X-ray spectrometry. Thismethod gives an approximate elemental composition

Ž .for atomic weights greater than 16 oxygen . Themethod does not distinguish between crystalline andamorphous materials. In all cases, iron was the majorcomponent. Sulfur was a significant component insome cases, meaning that iron sulfide could be acomponent of the suspended solids.

Ž .Iron carbide Fe C is known as cementite or3Žcohenite and is a component of carbon steel Dugs-

. Ž .tad et al., 2000 . Cron et al. 1971 studied thereaction of various acids and bases with low carbonsteels. The mode of acid attack was found to be afunction of the applied potential, pH of the environ-ment, and the anion present. Either the ferrite or the


Ž .Fig. 4. Flowback analysis pickle treatment : suspended solids and HCl concentration.

carbide phases could be reactive or non-reactive,depending on the electrochemical conditions. In WellA, it is likely that the iron in the steel casing wascorroded and the acid dissolved the corrosion prod-ucts, but left small particles of iron carbide behind.

The results shown in Fig. 4 also support this hypoth-esis. If the solids were 100% Fe C, then it can be3

calculated that 1500 mgrl of the solids would con-tain nearly 100 mgrl of carbon. If a flowbacksample contains 25,000 mgrl total iron, then the

Table 6XRD analysis of solids from pickle flowback

Major: )20 wt.%.Minor: 1 to 20 wt.%.Trace: 0 to 1 wt.%.Shaded area shows samples that contained HCl when collected.


Table 7Ž .Energy dispersive X-ray spectrometry of solids from pickle flowback concentration, wt.%

Shaded area shows samples that contained HCl when collected.

percentage of carbon in the iron is 0.4 wt.%. This isremarkably close to the amount of carbon expectedin low carbon steel, typically 0.4 to 0.5 wt.%. In theconditions of the casing pickle treatment, it appearsthat the iron carbide was not soluble in the acid. Itmay have been coated with a non-soluble material orwith biomass.

The chloride concentration in the flowback sam-ples reached a maximum of 132,000 mgrl, com-

pared to a value of 148,000 mgrl in the injectedŽ .pickle acid sample Table 5 . This result shows that

dispersion led to a small amount of dilution of theinjected pickle acid. The final chloride concentrationin the pickle flowback samples is the value expectedfor seawater.

Values of sulfate, calcium and magnesium in theflowback samples are shown in Fig. 5. The low

Žvalues of calcium during the acid flowback 2.0 to

Ž .Fig. 5. Flowback analysis pickle treatment : sulfate, calcium, and magnesium.


.2.4 h show that the acid slug did not contact theŽ .formation. The sulfate concentration 2.0 to 2.4 h in

the pickle acid increases from 590 to 880 mgrl,higher than the injected value of 510 mgrl. The

Ž .solubility of calcium sulfate gypsum increases inaqueous solutions at low pH or high ionic strengthŽ .Carlberg and Matthews, 1973 . This indicates thatsome sulfate scale is being removed from the casing.Assuming an initial concentration of 510 mgrl sul-

3 Žfate, and using the well flow rate of 22.3 m rh 140.bblrh , it was calculated that 3.7 kg of gypsum

Ž .CaSO P2H 0 was dissolved from the casing. After4 2

2.6 h, the calcium and magnesium levels approachedthe values of the formation water that were seen inFig. 1. These results show that the flowback timewas sufficient to completely lift the pickle acid fromthe wellbore.

In Fig. 6, aluminum concentration and pH areshown for the flowback fluids. The concentration ofzirconium was also measured and found to be less

Žthan the detection limit of the ICP instrument 1.mgrl . These measurements were made because both

aluminum and zirconium are present in the cross-lin-ker used in the acid diverting stage. It was necessaryto establish a baseline for aluminum and zirconiumconcentrations to evaluate results from the matrix

acid treatment. The presence of aluminum in thepickle flowback samples probably results from acid

Žinteractions with clay minerals Simon and Ander-.son, 1990 . The source of these clays is the upstream

sand filters. The clays are deposited on the casingand leached aluminum ion on contact with the acid.

3.4. Matrix acidization fluid returns

Well A was flowed back for 5 h after the matrixacidization treatment. The pH values of the producedfluids are shown in Fig. 7. Values measured at thewellsite and in the lab were in good agreement. Thehigh pH values measured at the start of the flowbackare due to sodium carbonate used at the very end ofthe treatment to flush out the surface lines. This highpH fluid did not reach the formation. From Fig. 7, aminimum pH value was reached at 2.7 h after thestart of the flowback. The pH was above a value offive for two consecutive samples after 5 h. Theflowback was terminated at this stage, as planned.

The hydrochloric acid concentration of the matrixacid flowback samples is shown in Fig. 7. A maxi-mum concentration of 1.7 wt.% HCl was reached at2.88 h after the start of the flowback. The acidconcentration gradually tapered off to zero after 5 h.

Ž .Fig. 6. Flowback analysis pickle treatment : aluminum and pH.


Ž .Fig. 7. Flowback analysis matrix treatment : acid concentration and pH.

3 ŽIf the flow rate is estimated to be 22.3 m rh 140.barrelsrh , the area under the curve can be inte-

3 Ž .grated to show that 1.9 m 12 barrels of 15 wt.%HCl did not react with the formation and was flowed

back from the well. It is likely that this unreactedacid was in a cavity or washout and was not in directcontact with the formation. The unreacted acid isonly 1% of the total volume of acid that was in-

Ž .Fig. 8. Flowback analysis matrix treatment : total iron.


jected. Increasing the soaking time from 70 min isnot likely to have a large effect on the amount ofunreacted acid.

Iron concentrations in the flowback samples wereŽ .measured, Fig. 8. Iron II was measured at the

wellsite and is compared to total iron measured byICP in the lab. The maximum iron concentration of170 mgrl occurred at 2.5 h and was almost entirely

Ž . Ž .in the iron II form. The values of iron II mea-sured in the field are higher than some of the totaliron measurements made in the lab due to experi-

Ž .mental errors sample oxidation . Using the wellflowrate and the data in Fig. 8, a total of 6 kg of ironwas produced in this stage. No solids were producedwith the samples containing iron. These iron concen-

Ž .trations are relatively low, and since iron II is thepredominant species, very little iron control chemi-cals are required. For future treatments in similarwater injection wells, when a pickle treatment isused, citric acid can be used as the iron controlchemical at a concentration of 0.22 wt.%. This quan-

Ž .tity of citric acid is sufficient for iron III concentra-tions up to 500 mgrl at temperatures up to 658CŽ . Ž .1508F Taylor et al., 1999a .

The density of the produced fluids was measuredas a function of flowback time. At 2.5 h, the densityrapidly increases then stabilizes near 1.041 grcm3,

indicating that the spent acid begins to flow back atthat point. However, the density then increased to amaximum of 1.065 grcm3 at 5 h, in contrast to the

Ž 3.density of the seawater 1.044 grcm . At 5 h, theŽ .pH value increased to more than 5 Fig. 7 . It is clear

that the flowback time was not sufficient, and thatthe pH value is not effective for determining thelength of the flowback time. Measurement of thedensity of the produced fluids should be used todetermine the optimum flowback time.

Chloride and sodium concentrations in the flow-back samples are shown in Fig. 9. The mass averageinjected chloride concentration was 118,300 mgrl,based on both the concentrations and relative vol-umes of the 10 wt.% in situ gelled acid and the 15wt.% regular acid. The acid that initially flowed back

Žwas diluted with both the low sulfate water chloride. Žconcentration of 370 mgrl and seawater chloride

.concentration of 30,000 mgrl . This dilution factorcan be calculated, because the sodium concentrationshould result only from the injected seawater. Theweight ratio of chloride to sodium in the seawaterwas reported as 1.82 to 1, as mentioned previously in

Žthe section on the initial well flowback before acid.treatment .

Fig. 10 shows the results of the calculations thatcan be made using the total chloride concentration,

Ž .Fig. 9. Flowback analysis matrix treatment : chloride and sodium.


Ž .Fig. 10. Flowback analysis matrix treatment : measured and calculated chloride values.

the sodium concentration, and the measured acidconcentration. The chloride from the live acid de-creases from approximately 17,000 to 1800 mgrl.This is calculated from the measured hydrochloricacid concentration. At the same time, the chloridecontributed from the seawater increased from 9000to 26,000 mgrl. The remaining chloride was con-tributed from the acid that reacted with the forma-tion, which decreases slightly from 6700 to 5600mgrl. By using the well flowback rate and theconcentration values of chloride that reacted with theformation, it can be calculated that less than 7% ofthe injected acid was produced. The same techniquecould be used on calcium and magnesium to calcu-late the amount of dolomite that reacted with the

Ž .acid Dahlan and Nasr-El-Din, 2000 . For Well A,however, insufficient spent acid was produced forthese calculations. This occurred because pH wasused to determine the amount of time the well shouldflow back after the treatment. The flowback timeshould be increased and the optimum time deter-mined using a portable density meter.

The acid flowing back from the well was signifi-Ž .cantly diluted Fig. 11 . At 2.70 h when acid first

flowed back from the well, it had been diluted toŽ20% of the injected concentration a dilution factor

.of five by mixing with injected seawater and theoverflush. By 4.1 h, the produced acid was at 6%, adilution factor of 17. When the fluids flowed back,they were a mixture of spent and live acids. Fig. 11also shows how much of the acid produced from theformation did not react with it. Initially, 70% of theacid was unreacted and that amount decreased to25% by 4.1 h. At 4.1 h, the acid was extremely

Ž .diluted by a factor of 17 and 25% of this acid hadnot reacted with the formation. This agrees with welllogs that show a large washout zone or cavity ispresent in the well below the casing shoe. Thiswashout zone, at the top of the openhole section,occurs in a zone that contains anhydrite. The largevolume of this washout can act as a mixing containerfor acid and spent acid flowing back from the forma-tion. If unreacted acid remains in the washout zoneat the end of injection, then it will gradually mixwith spent acid flowing up the casing as the well isflowed back.

As the spent acid flowed back from the well, itwas produced with seawater, the post-flush and thepre-flush. The proportion of low sulfate water de-creased and the proportion of seawater increased

Ž .with time Fig. 11 . At 2.7 h after flowback, whenthe acid and spent acid production began, seawater


Fig. 11. Matrix flowback composition.

already made up 30% of the total produced fluids.By 4.1 h, nearly 90% of the total fluids producedwas seawater.

Finally, Fig. 11 shows the percentage of chloridefrom spent acid in the flowback samples. Thesevalues decrease from 25% to approximately 15%,and show that a relatively constant amount of spent

acid was produced, although the amount of unreactedacid was decreasing rapidly. This may mean that thespent acid was produced from a different zone orzones than the live acid.

Fig. 12 shows the concentrations of sulfate, mag-nesium and calcium in the fluid returns. The calciumconcentration is close to the expected value from the

Ž .Fig. 12. Flowback analysis matrix treatment : sulfate, magnesium and calcium.


amount of spent acid calculated in Fig. 11, for theperiod 2.7 to 4 h. After that, the calcium concentra-tion increases, indicating that a larger amount ofspent acid is being produced. However, a calciumconcentration of approximately 80,000 mgrl resultswhen 15 wt.% HCl completely reacts with limestone.This shows that the spent acid flowing back from thewell was highly diluted. Magnesium concentration isnear 2200 mgrl, as compared to a value of 2000mgrl in the seawater. Some magnesium is probablypresent in the formation in the form of dolomite. Alarge amount of sulfate is seen in the flowbacksamples. The peak values of 5000 to 6000 mgrl aresimilar to those observed in the flowback samplesŽ .before acid treatment shown in Fig. 1. Sulfatevalues fluctuate at 4 h and later, because precipita-

Ž .tion of gypsum CaSO P2H O occurred due to the4 2

high calcium levels in these samples. Gypsum wasidentified by XRD analysis as the major componentŽ .Table 6 . These results show that calcium sulfatehas a limited solubility in the flowback samples.Once the solubility limit is exceeded, precipitation ofgypsum occurs.

ŽThe solubility of gypsum in seawater with a 3.wt.% chloride content at 258C is 0.458 wt.%

Ž .Silcock, 1979 . This corresponds to a calcium con-

centration of 1060 mgrl and a sulfate concentrationof 2540 mgrl. The concentrations of calcium andsulfate in the spent acid in Fig. 12 are both signifi-cantly higher than these values. The calcium concen-tration is approximately 7000 mgrl and the sulfateconcentration is approximately 4000 mgrl. Eitherthe samples are not at equilibrium with respect togypsum, or additives in the acid stimulation fluidsincrease the solubility of gypsum in seawater. It isknown that partially hydrolyzed polyacrylamides andpolyacrylic acids are used as calcium scale inhibitorsŽ .Przybylinski, 1989 . The 10 wt.% in situ gelled acidcontained an acrylamide copolymer and may inhibitgypsum precipitation.

Zirconium and aluminum concentrations in theflowback samples are shown in Fig. 13. The concen-trations of these metals were measured because theyare present in the in situ gelled acid that was injectedinto the reservoir. Both zirconium and aluminumwere produced during the flowback time 2.7 to 4 h.After 4 h, aluminum was present in the samples butzirconium concentrations were below the detectionlimit of 1 mgrl. From 2.7 to 4 h, a significantamount of unspent acid flowed back from the wellŽ .Fig. 11 , accounting for the produced zirconium.After 4 h, the produced fluids contained only seawa-

Ž .Fig. 13. Flowback analysis matrix treatment : zirconium and aluminum.


ter and spent acid. These results indicate that zirco-nium may precipitate or adsorb in the formation.Work is underway to determine the rate of propaga-tion of zirconium and aluminum in the formation.

3.5. Treatment eÕaluation

Before the acid stimulation treatment, the injectiv-Ž . 3 Žity index II of the well was 0.12 m rdayrkPa 5

. 3bblrdayrpsi . This value increased to 0.21 m rŽ .dayrkPa 9 bblrdayrpsi after the treatment, an

increase of 75%. Flowmeter results could not beobtained because of the large washout present di-rectly below the casing shoe. However, it is likelythat a significant percentage of the flow into theformation is into the region identified in the 1993

Ž Ž ..injection profile 2228 to 2256 m 7310 to 7400 ft .The injectivity of the well may have declined as a

result of particulates carried into the well by the 5.63 Ž .million m 35 million barrels of seawater that have

been injected to June, 1999. These particulates con-tain significant amounts of acid-insoluble material.This includes the iron carbide identified during thepickle treatment and in previous sampling at theseawater treatment plant. Other acid insoluble partic-ulates include fine sand particles and clay minerals.

3.6. The importance of pickling the casing

Return samples from the pickle treatment con-Ž .tained up to 15,000 mgrl of iron III and up to

1500 mgrl of suspended solids. The maximum prac-Ž .tical amount of iron III that can be complexed is

Ž .10,000 mgrl Taylor et al., 1999a . This requires 123 3 Žkgrm of citric acid and 22 lrm of acetic acid Dill

.and Smolarchuk, 1988 . At higher concentrations ofcitric acid, precipitation of calcium citrate may oc-cur. When the casing is pickled, a citric acid concen-tration of 2.4 kgrm3 in the main acid treatment is

Ž .sufficient to prevent precipitation of iron III hy-droxide in the near-wellbore region. The pickle treat-ment also prevents the injection of acid-insolublecasing scale into the formation. The cost of thepickle treatment itself was very small compared tothe chemical cost of the treatment. For a 37.9 m3

Ž .10,000 gal matrix acid stimulation, using a pickletreatment and reducing the amount of citric acid and

acetic acid reduced the treatment cost by more than15%.

3.7. Acid diÕersion

In the treatment design, alternate stages of regular15% HCl acid and 10% in situ gelled HCl werebullheaded. The objective of this staged process wasto provide effective acid diversion during the stimu-lation treatment. In situ gelled acid can be used tostimulate carbonate reservoirs that have significantpermeability contrast. Basically, these acids consistof an acid-soluble polymer, a pH buffer, a cross-lin-ker and a breaker. The polymer in this system formsa gel within a narrow pH range. As a result of gelformation, the viscosity of the acid increases in situand acid diversion can be achieved. Ultimately, thegelled acid will form wormholes evenly distributedover the entire target zone. This gel will improve

Žacid placement and control acid fluid loss Yeager.and Shuchart, 1997; McGee et al., 1997 .

When hydrochloric acid is injected into the forma-tion, it has a pH of nearly zero. The pH of the acidincreases as the acid reacts with the carbonate rock.At a pH value of approximately 2 to 4, the polymerreacts with the cross-linker and forms a very viscousgel. This reaction is reversible, but is dominant atthis pH range. The viscosity of the acid can reach

Ž .1000 cP Yeager and Shuchart, 1997 and is able todivert unreacted acid further into the well bore. Asthe acid continues its reaction, the pH will risefurther. At pH values greater than 4 to 5, the viscos-ity of the gel decreases as the polymer and cross-lin-ker dissociate. A breaker is used in some types ofgelled acid to ensure a complete reversal of the

Ž .cross-linking process Yeager and Shuchart, 1997 .The reduced viscosity of the spent acid is designedto improve its removal from the formation.

In the treatment design, maximum pumping ratewas to be achieved while holding pressure constant

Ž Žbelow wellhead maximum pressure 17.2 MPa 2500..psi and formation fracture pressure. The significant

parameters to be evaluated are pumping rate andvolume of the injected fluids.

During the pre-flush stage, the pressure steadilyŽ . Ž .climbed to 15.9 MPa 2300 psi Fig. 14 . During

this time, the average injection rate of 3.3 m3rminŽ . Ž .20.7 bpm was maintained Fig. 15 . Upon switch-


Fig. 14. Surface pressure and bottom hole pressure as a function of time.

ing to the acid, the rate was reduced to 1.4 m3rminŽ .8.8 bpm to facilitate switching between the lowvolume stages. The rate was then increased to 2.0m3rmin 12.8 bpm, but was limited by the surfacepressure.

As the first stage of acid reached the formation,the injection rate increased to maintain a constantsurface pressure. The rate increases from 2.0 to 3.3

3 Ž .m rmin 12.8 to 21 bpm while the treating pressureremained constant. This indicates that as the acid

Fig. 15. Injection rate and cumulative injection as a function of time.


reached the formation, injectivity increased. As in-jection of the acid continued, the injection rate de-creased until the post-flush reached the formation.This shows that diversion was occurring during acidinjection. After 83 min, the injection rate increasedas the post-flush was injected into the formation.

4. Summary

Techniques were demonstrated to evaluate acidstimulation treatments based on chemical analysis ofthe return fluids. These techniques were used in WellA to identify the source of formation damage, theamount and type of scale on the casing, and theefficiency of the acid treatment.

Ž1. During flowback of the well before the pickle.treatment , 125 kg of calcium sulfate dissolved

from the near-wellbore area.2. The pickle treatment was highly effective, dis-

solving an estimated 235 kg of iron from tubularcorrosion products.

3. The pickle fluid returns contained black sus-pended solids that were mostly iron carbide, Fe C.3

Ž4. For the pickle treatment, the acid volume 0.29. Ž .galrft, 0.36 lrm and the flowback time 4 h

were sufficient.5. The iron control chemical concentration can be

reduced for similar low temperature water injec-tion wells that are pickled before acid stimulation.

ŽCitric acid at 0.22 wt.% 20 lbr1000 gal of 15.wt.% HCl is sufficient to keep 500 mgrl iron

Ž .III in solution.6. When it flowed back from the well, the injected

acid was diluted between 5 and 17 times based oncalculations using the chloride concentration as atracer.

7. The pH value of the flowback fluid is not anadequate criterion to determine the flowback time.The density of the aqueous phase should be usedinstead.

Acknowledgements

The authors thank the Saudi Arabian Oil Com-Ž .pany Saudi Aramco for permission to publish this

work. The Chemistry Unit and Advanced Instru-ments Unit of Saudi Aramco provided valuable assis-tance with this work. The authors thank M. Al-Ramadhan, Dr. H.R. Rosser and G. Polkowski foruseful discussions.

References

Allaga, D.A., Wu, G., Sharma, M.M., Lake, L.W., 1992. Bariumand calcium sulfate precipitation and migration inside sand-

Ž .packs. SPE Formation Evaluation, March . pp. 79–86.Almond, S.W., Brady, J.L., Underdown, D.R., 1990. Return fluid

analysis from the Sadlerochit formation, Prudhoe Bay, AK:FIeld study — Part 1. SPE 18223, 63rd Ann. Tech. Conf.,Houston, TX, USA.

Bayona, H.J., 1993. A review of well injectivity performance inSaudi Arabia’s Ghawar field seawater injection program. SPE25531, SPE Middle East Oil Tech, Conf., Manama, Bahrain.

Brown, J.S., Dubrevil, L.R., Schneider, R.D., 1979. Seawaterproject in Saudi Arabia — early experience of plant operation,water quality and effect on injection well performance. SPE7763, Middle East Oil Tech. Conf., Manama, Bahrain.

Carlberg, B.L., Matthews, R.R., 1973. Solubility of calcium sul-fate in brine, SPE 4353, SPE Oilfield Chemistry Symp.,Denver, CO, USA.

Chen, E.Y., Ahmed, T., 1998. Why internally coated piping isused for the world’s largest seawater injection system. SPE49211, SPE Annual Tech. Conf., New Orleans, LA, USA.

Ž .Coburn, S.K. Ed. , Corrosion Source Book. American Society forMetals, Metals Park, OH, pp. 12–13.

Cowan, J.C., Weintritt, D.J., 1976. Water-Formed Scale Deposits.Gulf Publishing, Houston, TX, p. 68.

Cron, C.J., Payer, J.H., Staehle, R.W., 1971. Dissolution behaviorof Fe-Fe structure as a function of pH, potential, and anion3

— an electron microscopic study. Corrosion 27, 1.Crowe, C.W., 1985. Evaluation of agents for preventing precipita-

tion of ferric hydroxide from spent treating acid. J. Pet.Ž .Technol. 37 4 , 691.

Dahlan, M.N., Nasr-El-Din, H.A., 2000. A new technique toevaluate matrix acid treatments in carbonate reservoirs, SPE58714, International Symposium on Formation Damage Con-trol, Lafayette, LA.

Dill, W., Smolarchuk, P., 1988. Iron control in fracturing andŽ .acidizing operations. J. Can. Pet. Technol. 27 3 , 75–78.

Dugstad, A., Hemmer, H., Seiersten, M., 2000. Effect of steelmicrostructure upon corrosion rate and protective iron carbon-ate film formation. Paper 00024, Proceedings Corrosion 2000,NACE, Houston, TX.

Gdanski, R.D., Peavy, M.A., 1986. Well return analysis causesre-evaluation of HCl theories, SPE 14825, 7th SPE Symp. onFormation Damage Control, Lafayette, LA, USA.

Ginest, N.H., Phillips, J.E., Al-Gamber, A.W.A., Wright, D.W.,1993. Field evaluation of acid stimulation diverter materialsand placement methods in Arab-D injection wells with open-


hole completions. SPE 25412, SPE Middle East Tech. Conf.,Manama, Bahrain.

Gougler, P.D. Jr., Hendrick, J.E., Coulter, A.W., 1985. Fieldinvestigation identifies source and magnitude of iron prob-lems, SPE 13812, SPE Production Operations Symp., Okla-homa City, OK, USA.

Kaesche, H., 1985. Metallic Corrosion. National Association ofCorrosion Engineers, Houston, TX, p. 7.

Lee, W., Lewandowski, Z., Nielsen, P.H., Hamilton, W.A., 1995.Role of sulfate-reducing bacteria in corrosion of mild steel: areview. Biofouling 8, 165–194.

Little, B.J., Ray, R.I., Pope, R.K., 2000. The relationship betweencorrosion and the biological sulfur cycle. Paper 00394, Pro-ceedings Corrosion 2000, NACE, Houston, TX.

Loewen, K., Chan, K.S., Fraser, M., Leuty, B., 1990. A wellstimulation acid tube clean methodology. CIMrSPE 90-47,Petroleum Society of CIMrSPE International Tech. Meeting,June 10 to 13, Calgary, Canada, Preprints, vol. 1.

McGee, J., Buijse, M.A., Pongratz, R., 1997. Method for effectivefluid diversion when performing a matrix acid stimulation incarbonate formations, SPE 37736, Middle East Oil Show,Bahrain, 15–18 March.

Mohamed, S.K., Nasr-El-Din, H.A., Al-Furaidan, Y.A., 1999.Acid stimulation of power water injectors and saltwater dis-posal wells in a carbonate reservoir in Saudi Arabia: labora-tory testing and field results. SPE 56533, SPE Annual Techni-cal Meeting, Houston, TX. pp. 3–6.

Nasr-El-Din, H.A., Rosser, H.R., Hopkins, J.A., 1996. Stimulationof injection water supply wells in central Arabia. SPE 36181,7th ADIPEC, Abu Dhabi, UAE. pp. 13–16.

Nasr-El-Din, H.A., Al-Anazi, H.A., Mohamed, S.K., 1999. Stimu-lation of water disposal wells using acid-in-diesel emulsion:

case histories, SPE 50739, SPE International Symposium onOilfield Chemistry, Houston, TX, USA.

Patton, C.C., 1993. Corrosion control in water injection systems.Ž .Mater. Perform. 32 8 , 46–49.

Przybylinski, J.L., 1989. Adsorption and desorption characteristicsof mineral scale inhibitors as related to the design of squeezetreatments, SPE 18486, SPE International Symposium on Oil-field Chemistry, Houston, TX, USA.

Shuchart, C.E., 1995. HF acidizing returns analyses provide un-derstanding of HF reactions, SPE 30099, SPE European For-mation Damage Control, The Hague, The Netherlands.

Ž .Silcock, H.L. Ed. , Solubilities of Inorganic and Organic Com-pounds vol. 3 Pergamon, New York, USA, pp. 754–755.

Simon, D.E., Anderson, M.S., 1990. Stability of clay minerals inacid, SPE 19422, SPE Formation Damage Control Sympo-sium, Lafayette, LA, USA. .

Smith, C.F., Crowe, C.W., Nolan, T.J. III, 1969. Secondarydeposition of iron compounds following acidizing treatments.J. Pet. Technol., 1121, September.

Taylor, K.C., Nasr-El-Din, H.A., 1999. A systematic study of ironcontrol chemicals part 2, SPE 50772, SPE International Sym-posium on Oilfield Chemistry, Houston, TX, USA.

Taylor, K.C., Nasr-El-Din, H.A., Al-Alawi, M.J., 1999a. System-atic study of iron control chemicals used during well stimula-

Ž .tion. SPE J. 4 1 , 19–24.Taylor, K.C., Nasr-El-Din, H.A., Al-Alawi, M.J., 1999b. Field

test measures amount and type of iron in spent acids, SPE50780, SPE International Symp. on Oilfield Chemistry, Hous-ton, TX, USA.

Yeager, V., Shuchart, C., 1997. In situ gels improve formationacidizing. Oil Gas J., 70, Jan. 20.

Documents

Analysis of Acid Returns