SPE 87449

Measurement of total Alkalinity and Carboxylic Acid and Their Relation

to Scaling and Corrosion
Mason B. Tomson, SPE, Amy T. Kan, SPE, Gongmin Fu, Lili Cong

Copyright 2004, Society of Petroleum Engineers Inc.

intricate inter-relationship of total alkalinity, carboxylic acids,
This paper was prepared for presentation at the 6th International Symposium on Oilfield Scale and pH on scale and corrosion will be discussed by using
held in Aberdeen, UK, 26-27 May 2004.
case studies.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to Introduction
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at Based on charge neutrality relation, Eq. 1 has to be satified for
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
all brine solution chemistry1.
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 ([Na +
) ( )
] + 2[Ca 2 + ] + L − [Cl- ] + 2[SO24 - ]L = (1
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
[HCO ] + 2[CO ] + [Ac ] + [HS ] + 2 S
-
3
2-
3
- -
[ ] + [OH ] − [H ]
2− - +

The left hand side of Eq. 1 represents the net difference of the
Abstract summation of all strong base cations in brine and the
Alkalinity is needed in many water treatment calculations, summation of all strong acid anions in brine. The right hand
scale, corrosion, precipitation, oxidation, etc., yet the concept side of Eq. 1 represents the summation of all weak acid anion
is often misunderstood. In natural waters, alkalinity often is in brine and water. The [S2-] term is always small and will be
not equal to bicarbonate concentration since natural waters neglected hereinafter. Alkalinity is defined as the strong-acid
contain base contributing anions that can significantly affect neutralizing capacity of a solution2,3. From Eq. 1, the
alkalinity. However, alkalinity is commonly assumed to be alkalinaity can be defined as the sumation of either the terms
equal to the bicarbonate concentration in many scale and on the left or the terms on the right. The definition on the left
corrosion prediction algorithms. When other anions, e.g. would be fine if the analyses of the metals and anions were
carboxylates, are present, bicarbonate concentration in sufficiently accurate. Typically, the sum of the weak acids on
production tubing is not a conservative quantity; it varies with the right is only 0.001 to 0.010 M and the water terms, ([OH-]-
CO2 partial pressure, temperature and carboxylate [H+]), are small. The concentrations of strong base cations
concentrations in a complicated manner, up and down a well. and strong acid anions are about 1.000 to 6.000 M. Therefore,
Reliable methods to accurately measure true alkalinity are the difference would have to be accurate to about ±0.1 to 1.0
scarce, especially when multiple weak acids are present and %, which would require the individual numbers to be even
the effects of TDS on electrode and color end point are more accurate which is extremely unlikely for routine
significant. Oil field brines contain aliphatic carboxylic acids sampling and analyses. In fact, some other commercial
of one to six carbons, e.g. acetate, up to 5000 mg/L. The computer codes in the business have been known to calculate
highest concentrations of carboxylates tend to be in waters the solution pH from just such an analysis; in those cases the
from reservoirs at temperatures of 80 to 100 ºC. In this paper, calculated pH was routinely either about 10 to 11 or 3 to 4,
a new analytical procedure and computation routine to depending upon whether the random error in the difference
determine alkalinity and carboxylic acids simultaneously will was positive or negative, respectively.
be discussed. The procedure was recently debugged and This charge balance-type definition of alkalinity helps
simplified by the Rice University Brine Chemistry Consortium clarify one of the most important features of why the alkalinity
(Rice BCC). The new titration method is based upon concept is so important, that is, that the addition or removal or
simultaneous analysis of the titration curve determined at CO2 or H2S gas, up and down a well, has absolutely no effect
fixed PCO2 and emphasizes the titration shape (profile) on the total alkalinity. Therefore, the alkalinity is a
instead of the endpoint inflection as is done presently. “conservative” property of the solution. The pH does increase
A wide range of natural and synthetic waters has been as CO2 and H2S gas are released from solution as the brine
tested. Excellent agreement was observed between the true flows up a well. As the pH changes the relative amounts of
and calculated carboxylic acid concentration with a correlation the various weak acids on the right hand side of the equation
coefficient squared of 0.9986. Once the total alkalinity and will change, but the total alkalinity does not change at all.
acetic acid concentrations are determined, the theoretically Similarly, adding acetic acid will lower the pH, but will not
correct bicarbonate concentration and/or pH at any given alter the alkalinity, because it does not add any strong cations
operation temperature, pressure can be calculated. The nor strong anions to the left hand side of the above equation.
2 SPE 87449

This last concept, that adding or removing neutral CO2 or H2S equilibrium constants7 , K = 10-5.94, K = 10-2.04
1, H 2 CO 3 CO
from solution has no effect on the alkalinity, is made use of in 2

our laboratory determination alkalinity and carboxylic acids. atm/M, K HAc = 10-4.48.
We sparge the original brine with CO2 gas to remove all the
sulfide in solution as H2Sg, which has no effect on alkalinity. K1, H CO ⋅ K CO ⋅ PCO 2 , gas K HAc
Alkalinity ≅ 2 3 2
+ THAc ⋅
Once the charge balance-based definition of alkalinity is [H ] + +
[H ] + K HAc
understood, the equations relevant to each physical situation, − 7.98
e.g., free to exchange gas or not, etc., can be substituted from 10 ⋅ PCO 2 , gas (atm) 10 - 4.48
Alkalinity ≅ + THAc ⋅
the previous definitions and rearrangements4,5: [H ]+ +
[H ] + 10 - 4.48
Alkalinity = [ HCO 3- ] + 2[CO 32 - ] + [ Ac - ] + [ HS- ] + [OH - ] − [H + ] (7
Since the Alkalinity and the THAc are both constants and the
(2
PCO2, gas changes due to the flow of the well, the pH varies with
If a well has no acetic acid nor hydrogen sulfide, as is often the PCO2, gas. When the PCO2, gas is low, toward the surface of
the case: the well -or the laboratory- and depending upon the percentage
CO2 in the gas, the pH will be higher or [H+] will be lower and
Alkalinity = [ HCO 3- ] + 2[CO 32 - ] + [OH - ] − [H + ] (3
the second term on the right hand side will become THAc⋅1.00,
The pH of most oil field wells is in the range of from 4 to 7/8 i.e., all the acetic acid will be ionized and [HCO 3- ] will be less
pH. In this case, generally the majority of the weak acid than it otherwise should be. Then, as the gas is compressed
charge balance is given by the [ HCO 3- ] term alone. Therefore, down the well the CO2 in gas phase will be forced back into
the alkalinity equation can be simplified for many wells the brine which will lower the pH making [H+] → K HAc at pH
to simply: = 4.48. At this pH only ½ or the THAc will be ionized and the
Alkalinity ≅ [ HCO 3- ] (4 actual amount of [HCO 3- ] will increase which will increase the
SI(calcite) quadratically (see Eq.13 below for calcite SI).
When there is also an amount of carboxylic acids present6, For example, with Alkalinity = 610 mg/l = 0.01 equiv./l
but no sulfide, for most wells the alkalinity becomes and THAc = 0.009 M with 5 % CO2 (STP) conditions at the
approximately the following:
surface the pHSTP ≈ 6.00. At this point, [HCO 3- ] ≈0.00112 M
Alkalinity ≅ [ HCO 3- ] + [ Ac - ] (5 and [ Ac - ] ≈ 0.00888 M, assuming the same values for the
Numerous approximate methods to determine alkalinity have equilibrium constants. In the well at one hundred times this
been based upon this equation. Typically, something along pressure, PCO2, gas = 5 atm: [HCO 3- ] ≈0.00370 M and [ Ac - ] ≈
the lines of adding a known excess amount of strong acid 0.00630 M at 4.85 pH. This shows that the [HCO 3- ] has
which converts the [ HCO 3- ] to carbon dioxide that is removed increased by 330 %. This will change the SI by log(3.3)2 =
by boiling or sparging with nitrogen, Then the excess strong 1.04 SI units vs. what would have been calculated without
acid is normally titrated with strong base to a known end correcting for the acetic acid effect.
point. Sometimes the carbon dioxide is trapped and measured. With more work on oil wells that often contain significant
The Rice U research group presented two methods along these quantities of either or both hydrogen sulfide or carboxylic
lines several years ago. If the carboxylic acids are also acids, we included both into our computer models,e.g.,
measured by GC or ion chromatography, the determination is ScaleSoftPitzerTM. Even though the calculations are
even better. Most of these methods are based around the goal considerably more tedious, they can easily be handled
of accurately measuring the [ HCO 3- ] concentration in the brine rigorously in ScaleSoftPitzerTM with ease once the problem is
sample. As will be illustrated more precisely below, even if properly stated.
the [ HCO 3- ] is measured accurately in the laboratory, this is If a gas phase is present and allowed to exchange with
the solution:
generally not the conceptually correct objective for predicting
pH and scale formation in a flowing well. Often this is Alkalinity = [HCO3- ] + 2[CO32 - ] + [Ac- ] + [HS- ] + [OH - ] − [H + ] (8
reasonably close. If the PCO2, gas is measured or controlled, the
interrelationship between [ HCO 3- ] and [ Ac - ] for this As stated previously, the original brine solution is sparged
with constant PCO2, gas to remove any H2S as a gas and to
approximate case of only carbonate and carboxylic acids establish equilibrium. Then, as a titration is performed with
present in significant amounts: strong acid, HCl, there are only two unknowns, Alkalinity and
Alkalinity ≅ [HCO3- ] + [Ac- ] THAc.
(6
K1, H 2 CO 3 ⋅ K CO 2 ⋅ PCO 2 , gas
K HAc
Alkalinity ≅ + THAc ⋅
[H + ] [H + ] + K HAc
This last approximate equation can be used to illustrate the
effect of carboxylic acids on the concentration of [ HCO 3- ] by
substituting in specific values for all the constants using 200 F,
1,000 psi, and 1 M TDS ≈ 60,000 mg/L; the conditional
SPE 87449 3

In the absence of hydrogen sulfide gas, the above equation from temperature and ionic strength. The values of fHCO3 and
simplifies to: fAc are used to correct for inability to know the "true" activity
Alkalinity = [HCO3- ] + 2[CO32 - ] + [Ac- ] + [OH - ] − [H + ] coefficients, variations in electrode response, for errors in the
equilibrium, and true PCO2, etc. The program, "Rice BCC
K1, H 2 CO 3 ⋅ K CO 2 ⋅ PCO 2 , gas K1, H 2 CO 3 K 2, HCO - ⋅ ⋅K CO 2 ⋅ PCO 2 , gas
= + 2⋅ 3 Alk-TAc at [Link]" computes the sum of squared errors of
+
[H ] [H + ]2 the calculated and observed HCl (M) vs. pH and minimizes
K HAc K the error by changing the Alk, TAc, fAc and fHCO3 values
+ THAc ⋅ +
+ w+ − [H + ] using Newton-Raphson procedure in Excel Solver©.
[H ] + K HAc [H ]
Correlation coefficient of the fit and standard deviation of the
(9
curve fitted parameters are calculated according to a procedure
This last equation is the “working” equation used in our
modified from Billo, "Excel for Chemists"8.
alkalinity determination.
RESULTS AND DISCUSSION
MATERIALS AND METHOD
Testing With Synthetic Brines. In figure 2 is shown a
Material and apparatus. The apparatus (Figure 1) consists
typical brine titration result entered to the "Rice BCC Alk-
of a 1% CO2 in N2 mixed gas cylinder, a sparging apparatus
TAc at PCO2©.xls" program. We have tested this procedure
and a titration vessel, pH electrode and pH meter, magnetic
with thirteen synthetic brines containing 610 mg/L total
stirrer and a titrator. A 1 N HCl standard solution was used
alkalinity and between 0 to 600 mg/L of carboxylic acid.
in titrations.
Between 0.18 to 5.22 M NaCl was added to these brines as
background electrolytes. After the titration by this method, the
Method. Briefly, the procedure involves carefully titrating a
total alkalinity and carboxylic acid concentrations were
brine solution, at fixed PCO2, to a pH of 3.0. The procedure
calculated with "Rice BCC Alk-TAc at PCO2©.xls" program.
uses the brine sample as its own reference. The procedure is
The curve fitted titration curves yield a calculated total
not very sensitive to temperature and PCO2. Typically, brine
alkalinity of 598±6 mg/L. In Figure 3 is plotted the true
should be titrated at near room temperature. The PCO2 for the
carboxylic acid concentration versus the carboxylic acid
titration should probably be the larger of 0.01 atm or greater
concentration calculated from "Rice BCC Alk-TAc at
than the PCO2 in the well. Approximately 100 ml brine sample
PCO2©.xls". Excellent agreement was also observed between
(~77 °F) is poured into the titration vessel. Then, the sample is
the true and calculated carboxylic acid concentration with a
sparged vigorously with CO2 gas at a typical flow of ~1 l/min.
correlation coefficient squared of 0.9986.
The sample equilibrates with CO2 in approximately 30
minutes. After the sample is saturated with CO2, titrate the
Testing With Oilfield Brines. We also have tested this
sample solution with HCl in small increment and allow the pH
procedure with six representative field brines (See Table 1) of
to re-establish equilibrium after each acid addition. Ideally,
wide variation in composition. In Table 1 is listed the brine
about 10 equal pH increments of titration data should be taken
data reported by the operators and the alkalinity and
between the initial pH and the end point of about 3 pH.
carboxylic acid concentrations measured by this method. The
The titration curve is fitted to a theoretical equation, shown
TDS concentrations of these brines varied from 9,972 to
below, using a non-linear least square curve-fitting routine.
343,346 mg/L, the calcium concentrations varied from 128 to
Equation 6 is derived from Eq 4 and is used in the Rice BCC
13,340 mg/L and the alkalinity varied from 0 to 3000 mg/L.
Alk-TAc at [Link]© Excel program to calculate HCl (M)
The brine pH varies from 4.51 to 7.67 pH. Notice that the
needed and compare with the experimental values at each
reported alkalinity varied by at least a factor of 2-3 and the
titration point.
reported pH is often quite different from the pH that we
10− pH measured in our laboratory. This causes the comparison of
HCl(M, calc.) = Alk + − total alkalinity from our method and the operator results nearly
γH+
impossible. In the following, we will briefly discuss what are
⎧ ⎫ the alkalinity and carboxylic acid concentrations determined
⎪ ⎪
⎪ fHCO3 ⋅ K1, H 2 CO 3 ⋅ K CO 2 ⋅ PCO 2 ,gas ⋅ γ g ,CO 2 + ⎪ by our method and what are the implication to scale and
(10
⎪ ⎪ corrosion tendency.
10− pH ⋅ γ HCO −
⎪ 3 ⎪ In the last three rows of Table 1 are listed the initial pH,
⎪⎪ TAc ⎪⎪ alkalinity and carboxylic acid concentrations determined by
⎨ + ⎬
⎪ ⎛⎜ 10 ⋅ γ Ac −
− pH
⎞ ⎪ this procedure. Note that initial pH for brines D and F are
+ 1 ⎟ considerably different from operator supplied data, indicating
⎪ ⎜ fAc ⋅ K ⋅ γ ⎟ ⎪
⎪⎝ HAc HAc, aq ⎠ ⎪ there may be other problems, which will be discussed below.
⎪ Kw ⎪ Various concentrations of carboxylic acid (0-338 mg/L) were
⎪ − pH ⎪ added to brine A and brine B for testing the procedure. The
⎪⎩10 ⋅ γ OH − ⎪⎭
carboxylic acid concentrations measured by this method are
where Alk is total alkalinity (equiv/L), TAc is equivalent sum generally in excellent agreement to the carboxylic acid
of carboxylic acids, PCO2, gas is the CO2 partial pressure (atm), concentrations measured independently by GC. Similarly,
K's are equilibrium constants, and γ's are the activity excellent agreement between the calculated and known
coefficients. The values of the activity coefficients and carboxylic acid concentrations can be seen in Figure 3, which
equilibrium constants are calculated in Excel Visual Basic includes additional data from Brines A and B with added
4 SPE 87449

carboxylic acid, with a correlation coefficient squared A total Fe concentration of ~50 mg/L was measured by
of 0.9946. acidifing the brine to dissolve the Fe precipitate. This
Interestingly, our alkalinity data agrees to the operator's corresponds to 0.90 mmol/L of total iron and at two
Brine A result quite well. Good agreement is reported for equivalents of H+ per mole of iron this corresponds to 1.80
Brine A, as this brine does not contain carboxylic acids and meq/L of H+ produced by the oxidation of iron and the
the solution composition is moderate. For brine C, the operator subsequent precipitation as Fe(III)Ox. This would destroy 1.80
did not measure carboxylic acid concentration. However, we milliequivalents per liter of total alkalinity, i.e., the total
observed nearly 42% of total alkalinity is due to the presence alkalinity would be equal to 110 mg/L higher than measured
of carboxylic acid. If one assumes the operator's brine by our method. By adding 110 mg/L of alkalinity that was
analysis, the calcite SI would be quite high (0.26 to 0.94 destroyed by Fe oxidation to Brine D, the original brine would
calcite SI between bottom hole to well head at 1,150 mg/L have a total alkalinity of 490 mg/L. Assuming a downhole SI
alkalinity). Up to 244 mg/L of calcite would be precipitated of 0, this brine should contain ~533 mg/L total alkalinity,
from this well as predicted by ScaleSoftPitzerTM using the which is quite reasonable and a self-consistent interpretation
operator's data. However, the well may not be scaling and of the field data.
possibly even slightly corrosive (-0.34 to 0.34 calcite SI
between bottom hole to well head), when we used the What If The Predominant Carboxylic Acid In Brine Is
alkalinity and carboxylic acid data from this study. That is, Formic Acid? In the above discussion, we assumed that all
one might have over treated for scale and under treated carboxylic acids are similar to acetic acid in acid/base
for corrosion. titration. This is a valid assumption for most carboxylic acid,
Brine D, E, and F have very high TDS. The solution pH is except when formic acid is the predominant carboxylic acid in
considerably different from the operator's reported result, the brine. The pKa for formic acid is approximately one log
indicating there are other issues that need to be resolved. unit lower than the other carboxylic acids9. There are two
Interestingly, the carboxylic acid concentrations determined aspects to be considered if formic acid is the predominant
by this method are in very good agreement to that reported by carboxylic acid in the brine of concern. First, what will be the
the operator, but not the case for alkalinity. In our study, we Alk and TAc values that Rice BCC Alk-TAc at [Link]
observed that nearly all alkalinity is due to the presence of calculates when a solution contains a large fraction of formic
carboxylic acid. In other words, the bicarbonate concentration acid instead of acetic acid? Second, how much error will it
is negligibly small. This is quite reasonable due to the very result in Calcite SI and pH predictions?
high calcium concentrations and low pH. At this point, we do In the Rice BCC Alk-TAc at [Link] program, four
not know why the pH of brine D and F are so much lower than adjustable parameters, Alk, TAc, fHCO3, and fAc, are used to
the operator's reported values. However, we noticed that brine fit the titration curve (see Eq. 10). Therefore, the program is
D contained significant amount of iron precipitate. We suspect self-correcting for uncertainty in activity coefficients and
that the brine may be compromised due to oxidation after the equilibrium constants, etc. with the two adjustable parameters:
brine was sampled. fAc and fHCO3. If a solution contains formic acid instead of
acetic acid, the fAc value ~10 would be used by Excel solver
The Effect Of Iron (II) Oxidation And Precipitation On to fit the data and compensate for the error in equilibrium
Alkalinity, HCO3-, And pH. Iron oxidation would produce constants. To prove this, we first generated a theoretical
two equivalents of protons per iron due to Fe(II) to Fe(III) titration data with 0.01 M alkalinity, 0.001 M formic acid in 1
oxidation and subsequent precipitation1 as illustrated in the M ionic strength at 25 ºC. We then used the theoretically
following equations. generated data as input data in Rice BCC Alk-TAc at
[Link] program to calculate the alkalinity and carboxylic
Fe2 + + 1 4 O 2 + H + → Fe3+ + 1 / 2H 2O Net Redox reaction
acid concentrations. The calculated Alk and TAc are 636
3+ −
Fe + 3OH → Fe(OH)3, solid Precipitation mg/L and 85 mg/l. Note that these concentrations are
3H 2O → 3H + 3OH+ −
Ionization of three waters equivalent to 0.010426 M alkalinity and 0.001427 M of acetic
acid and a net bicarbonate concentration of 0.00901 M, i.e.,
Fe2 + + 1 4O 2 + 2.5H 2O → Fe(OH)3, solid + 2H + Net Reaction the true bicarbonate concentration.
(11 Next, we used ScaleSoftPitzer to determine how the
The net effect, Eq. 11, of Fe2+ oxidation and Fe(OH)3,sold equilibrium constants affect calcite SI predictions by assuming
precipitation is to replace each Fe2+ ion in solution by two the equilibrium constant of acetic acid versus that of formic
hydrogen ion equivalents, 2H+, and this neutralizes exactly acid. We used a hypothetical brine composition of 70,900
two equivalents of bicarbonate alkalinity, 2HCO3- + 2H+ → mg/L TDS, 1000 mg/L Ca, 636 mg/L alkalinity and 85 mg/L
2H2CO3. For each 0.001 M Fe2+ that oxidizes and precipitates carboxylic acid and the operation condition of 340 °F BHT,
the alkalinity will drop by 122 mg/L as HCO3-. The amount of 120 °F WHT, 7000 psi BHP, 500 psi WHP, and 3% CO2. The
Fe(II) that has oxidized and precipitated can generally be SI calculations were compared by assuming the ionization
determined by filtering the solid sediment in a sample bottle, constant of formic acid and acetic acid. As shown in Tables 2,
strong acid digestion, and analyze TFe by ICP, etc. If the both calcite SI and brine pH were not significantly affected if
solid is dark colored and no precautions to exclude O2 were we mistaking formic acid as acetic acid. The error of assuming
taken the TFe in the solid can probably be assumed to be all carboxylic acids to be acetic acid are believed to be
Fe(III) from oxidation and precipitation. negligibly small in most realistic field conditions.
SPE 87449 5

Furthermore, mistaking formic acid with acetic acid will acid addition during work over or due to microbial activities
always produce a conservative estimate of calcite SI. that results in bicarbonate consumption. Once the total
alkalinity and acetic acid concentrations are determined by the
The Effect Of Sulfide Oxidation On Alkalinity, HCO3-, procedure discussed above, the theoretically corrected
And pH. As long as the pH is less than about 7.0, the bicarbonate concentration and/or pH at any given operation
predominant sulfide species in solution will be H2Saq and the temperature, pressure can be calculated from the scale
oxidation will proceed as follows: prediction programs such as Rice University
H 2Saq + 2O 2 → SO 24 − + 2H + (12 ScaleSoftPitzerTM program.
From this there would be a loss of two equivalents of
Conclusions.
alkalinity for each mole of hydrogen sulfide oxidized. In fact,
Alkalinity and pH are needed in many water treatment
the oxidation of sulfide species in solution is extremely
calculations, scale, corrosion, precipitation, oxidation, etc., yet
complex but generally in the absence of bacteria as catalysts
the concept is often misunderstood and reliable methods to
and the absence of ferric hydroxide solid phase the rate will be
accurately measure alkalinity are scarce, especially when
rather slow. Bacterial oxidation can be inhibited by adding
multiple weak acids are present. A new titration method that
any of numerous bactericides/inhibitors, such as sodium azide,
is based upon simultaneous analysis of the titration curve
ethylene oxide, mercuric nitrate, quaternary ammonium salts,
determined at fixed conditions, instead of complete reliance on
formaldehyde, cresol, copper (II) hydroxide, to mention a few.
the end points, has been developed and tested for a wide range
of natural and synthetic waters. The simultaneous
Relation To Scale And Corrosion Prediction. Good
alkalinity/carboxylic acids method proposed avoids the
scale and corrosion prediction relies upon precise knowledge
tedious hydrogen sulfide problem by sparging the sample with
of water chemistry10,11. For example, the true value of HCO3
fixed PCO2, gas and this will remove any H2Saq without any
concentration in water is needed for use with calcite Saturation
effect on the true alkalinity nor the carboxylic acids. Then, the
Index (Eq. 13) calculations, in the presence of a
effect of hydrogen sulfide on the brine chemistry is rigorously
known PCO2, gas:
calculated from the reported PH2S and Henry’s law constants or
from measured H2Saq (mg/L or M) and Henry’s law –the
⎧⎪ a Ca 2 + ⋅ a CO 32 − ⎪⎫ reported PH2S is recommended as probably the most reliable
SI Calcite = Log10 ⎨ ⎬ method with field samples and handling. Presently, these
⎪⎩ K sp , Calcite (T, P) ⎪⎭ (13 procedures are being tested in a systematic round-robin blind
⎧⎪ 2+ − 2
[Ca ] ⋅ [HCO ] ⋅ γ Ca 2 + γ HCO −
3
⎫⎪ test with personnel from several of the Rice BCC companies.
= Log10 ⎨ 3
⎬ Results of these blind tests will be published with the
⎪⎩ K sp , Calcite K1, H 2 CO 3 K CO 2 PCO 2 , gas γ gas (T, P) ⎪⎭
CO 2
participants later.
As mentioned in Introduction, there are some fundamental
theoretical issues related to the true bicarbonate concentration Acknowledgment
in the brine versus pressure, and to the presence of carboxylic The financial support of the Rice University Brine Chemistry
acids. The value of (HCO 3− ) changes with pressure changing Consortium of companies: Aramco, B.J.-Unichem, Baker-
Petrolite, Champion Technologies, Inc., Chevron-Texaco,
down a well. The change with pressure is strongly dependent Inc., ConocoPhillips, Inc., Marathon Oil, Ondeo Nalco, Shell,
upon the specific PCO2,gas, the relative volume of water, gas, Center for Biological and Environmental Nanotechnology,
and oil produced, and the total carboxylic acids present. Nanoscale Science and Engineering Initiative of the National
Figures 5-7 are illustrations of such complicated relationshiops Science Foundation and U.S. EPA Hazardous Substance
between bicarbonate and acetic acid at different operation Research Center/South & Southwest Region is
conditions. In Figure 5 is plotted the influence of water greatly appreciated.
production rate and pressure on CO2 concentration of the gas
phase. Assumed was a CO2 concentration of 3% at STP Nomenclature:
condition, the downhole gas phase CO2 concentration could [CO 32- ] = Carbonate ion concentration, m
varied drastically with the water production rate and pressure.
In Figure 6 and 7 are plotted the (HCO3-) and brine pH as a [ HCO 3- ] = Bicarbonate ion concentration, m
function of PCO2, gas and acetic acid concentration in the brine. [HS ]-
= Hydrogen sulfide ion concentration, m
As shown in Figure 6, the bicarbonate concentration is nearly [S2− ] = Sulfide ion concentration, m
identical between 0.1 to 1000 psi PCO2, gas in the absence of
acetic acid. However, the bicarbonate concentration changed [OH - ] = Hydroxide ion concentration, m
considerably in the presence of acetic acid. A useful way to [H + ] = Hydrogen ion concentration, m
address this problem is by using the concept of alkalinity. The [Ac - ] = Acetate ion concentration, m
advantage of this concept derives from the fact that although Alk =total alkalinity (equiv/L)
the actual concentration of the weak-acid anions and pH TAc = equivalent sum of carboxylic acids (equiv/L)
change with pressure and PCO2, gas, the alkalinity is constant –as PCO2, gas = the CO2 partial pressure (atm)
long as no net acids or bases are added, e.g., by precipitation K CO 2 = Henry's law constant for CO2 partition in water
or dissolution of CaCO3 during production, by oxidation (most
oxidation reactions produce net acid) during storage, by direct (mol·Kg-1·atm-1)
6 SPE 87449

fHCO3 = Correction factor for bicarbonate terms

fAc = Correction factor for acetic acid terms
K1, H 2 CO 3 = First ionization constant of carbonic acid
K 2, HCO − = Second ionization constant of carbonic acid
3

K HAc = Ionization constant of acetic acid.

K sp , Calcite = Solubility product of calcite
γ H+ = Hydrogen ion activity coefficient
γ HCO − = Bicarbonate ion activity coefficient
3

γ g ,CO 2 = CO2 gas fugacity coefficient

γ HAc ,aq = Acetic acid activity coefficient
γ OH − = Hydroxide ion activity coefficient
γ AC− = Acetate ion activity coefficient
a Ca 2 + = Activity of Calcium ion
a CO 2 − = Activity of Carbonate ion
3

SI Calcite = Saturation index for calcite

REFERENCES
(1)Stumm, W.; Morgan, J. J. Aquatic Chemistry Chemical
Equilibriua and Rates in Natural Water; 2nd edition ed.; Wiley-
Interscience: New York, NY, 1996.
(2)Butler, J. N. Carbon Dioxide Equilibria and Their Applications;
Addison-Wesley Publ.: Reading, Mass., 1982.
(3)Eaton, A. D.; Clesceri, L. S.; Greenberg, A. E. Standard Methods
for the Examination of Water and Wastewater; 19th ed.; APHA,
AWWA and WPCF: Washington, D.C., 1998.
(4)Oddo, J. E.; Tomson, M. B. "Simplfied Calculation of CaCO3
Saturation at High Temperatures and Pressures in Brine
Solutions"; J. Pet. Tech. 1982, 34, pp. 1583-1590.
(5)Kharaka, Y. K.; Gunter, W. D.; Aggarwal, P. K. "Solmineq 88: A
computer program for geochemical modeling of water-rock
interactions," U. S. Geological Survey, 1988.
(6)Collins, A. G. Geochemistry of Oilfield Waters; Elsevier
Scientific: Amsterdam, 1975.
(7)He, S. L.; Kan, A. T.; Tomson, M. B. "A new interactive software
for scale prediction, control and management"; In SPE Annual
Technical Conference and exhibition;SPE:
San Antonio, TX, 1997
(8)Billo, J. Excel for Chemists; 2nd Ed. ed.; John Wiley and Sons:
New York, 2001.
(9)Smith, R. E.; Martell, A. E. Critical Stability Constants; Plenum
Press: New York, NY, 1975.
(10) Oddo, J. E.; Tomson, M. B. "Why Scale Forms in the Oil Field
and Methods to Predict It"; SPE Prodcution and Facilities 1994,
1994, 47-54.
(11) Kaasa, B. "Alkalinity in oil field waters, what alkalinity is and
how it is measured." In SPE International Symposium on
Oilfield Chemistry;SPE: Houston, TX, 1997
SPE 87449 7

Table 1. Composition of six field brines reported by the operators and the total alkalinity and carboxylic acid
composition determined by this method.
Well Name/Species Brine A Brine B Brine C Brine D Brine E Brine F
(mg/L)
Data supplied by operators
Na & K 7,057 63,821 2,733 99,680 115,881 50,875
Mg 292 488 15 2,401 2,380 1,502
Ca 2,880 1225 128 12,000 13,340 9,300
Sr 0 37 73 0 571 42
Ba 708 0.1 0.1 0 7 160
Fe 0 49 2.6 30 0.1 17
Cl 17,000 97,107 2,008 173,330 210,045 95,565
sulfate 12 5,140 3461 83 8 15
Alkalinitya 317 2,013 1150-3000 0-115 0-531 146-256
Carboxylic acidc <10 N.A.b N.A.b 496 840 288
TDS 33,345 169,843 9,972 249,346 343,249 158,072
pH 6.65 6.97 6.90-8.90 4.51-5.40 6.8-6.82 6.8-7.88
The following are initial pH, alkalinity, and carboxylic acid measured by this method
PHd 7.00 7.20 7.67 4.45 5.84 4.30
Alkalinity 304 575 986 380 955 273
Carboxylic acid 0 96 411 506 909 255
a
Reported by operator or titrated by conventional method (HACH Co).
b
N.A. = not available.
c
Analyzed by GC or other methods by operators.
d
pH measured at PCO2, gas = 0.01 atm.

Table 2. ESTIMATION OF THE IMPACT OF MISTAKEN FORMIC ACID AS ACETIC ACID ON

CALCITE SI PREDICTION.

T P Assume as Acetic acida Assume as Formic acidb

(°F) (psi) pH CalciteSI pH CalciteSI
120 500 5.61 -0.29 5.61 -0.30
144 1,222 5.39 -0.42 5.39 -0.43
169 1,944 5.34 -0.39 5.33 -0.41
193 2,667 5.33 -0.33 5.33 -0.35
218 3,389 5.35 -0.26 5.34 -0.27
242 4,111 5.38 -0.17 5.37 -0.19
267 4,833 5.42 -0.08 5.41 -0.09
291 5,556 5.47 0.02 5.46 0.00
316 6,278 5.52 0.13 5.51 0.11
340 7,000 5.58 0.24 5.57 0.21
8 SPE 87449

10
4
2 3
9 pH 6.55
7
1 5 6

8
6
Sparging Titration
apparatus apparatus

List of apparatus:
1. CO2/N2 Gas Cylinder
2. Gas regulator
3. Regulating valve
4. 1/8" OD Teflon tubing
5. Glass bottle with a GL-45 2 valve cap
6. Sparging filter
7. 8 oz jar, with a 1a1/2” stirring bar
8. Magnetic stirrer
9. pH electrode
10. pH meter
Figure 1. Schematic diagram of the titration apparatus.
SPE 87449 9

Rice University Brine Chemistry Consortium

Total Alkalinity, Carboxylic Acid Simultaneous Titration at Constant PCO2©
Sample ID Temp (F) 77.0 <Enter T, TDS, & CO2% values used
Date 9/20/2000 TDS (mg/L) 9,972 in titration.
Operator BJ CO2 (%) 1.00 Std. Err.
Total Alk Click here to run.
Well Name Brine A 986 5
(mg/L)
Carboxylic <Check box if TAc is known and
Location 411 6
acid (mg/L) fixed. Enter value in cell D8.
Field r2 0.9997

Titration Data and Plot

HCl(M) exp pH exp
0.0000 7.671 8.0
0.0010 7.620
7.0
0.0020 7.563
6.0
0.0030 7.501
pH

0.0040 7.427 5.0

0.0050 7.336 4.0

0.0060 7.222 3.0
0.0070 7.065 0.0000 0.0050 0.0100 0.0150 0.0200
0.0080 6.820 HCl added (M, Exp)
0.0090 6.305
"Exp" "Calc"
0.0096 5.795
0.0100 5.503 Instructions for use of Rice BCC total alkalinity, carboxylic acid simultaneous
0.0110 5.060 titration at constant PCO2:
0.0120 4.764 Experimental:
0.0130 4.502
1. Calibrate a pH electrode (Orion Ross Combination Electrode) vs standard buffers.
2. Equilibrate a 100 ml sample at fixed PCO2 by sparging brine in a covered beaker
0.0140 4.241
with ~1% CO2 gas cylinder for approximately 30 minutes and record the pH.
0.0150 3.914
3. Titrate the sample with 1 N HCl and record the pH versus HCl added. Allow
0.0160 3.528
about 5-7 minutes per titration point for solution to re-equilibrate with CO2 gas.
0.0170 3.201
4. Continue the titration with HCl to a final pH of ~3.0.
Data analysis:
1. Enter the temperature (Cell D4), the approximate solution TDS (Cell D5), and
CO2 gas concentration (Cell D6) of the sparging gas at which the titration was
performed.
2. Delete the existing titration data (in Columns A, B and Rows 13 and below) from
the worksheet (Select the old titration data and press the Delete key on the key
board). One may also use the "Clear Content" command on the Edit menu. NOTE:
Do not use the "Clear All" command on the Edit menu.
3. Enter the titration data in Columns A and B and Rows 13 and below.
4. Click the Click Here to Run button to calculate the total alkalinity and carboxylic
acid concentration of the sample.
5. If you have measured the value of TAc by GC, etc., analysis, the value can be
entered Cell D8 and check the Check Box next to it. If the Check box is checked, the
program will calculate the total alkalinity using the input carboxylic acid value.

Figure 2. Rice BCC Alk-TAc at [Link] program data input/output sheet.

10 SPE 87449

700
R2 = 0.9986

Carboxylic Acid (mg/L)

600

Calculated Conc.
500
400
300
200
100
0
0 200 400 600 800
Known Conc.
Carboxylic Acid (mg/L)

Figure 3. Plot of the calculated versus known carboxylic acid concentrations in thirteen synthetic brines. The calculated carboxylic acid
concentrations was determined by fitting the titration results by this method and calculated with the Rice BCC Alk-TAc at [Link] program.

1000
R2 = 0.9946
Carboxylic acid (mg/L)

800
Calc. Conc.

600

400

200

0
0 200 400 600 800 1000
Known Conc. (by GC)
Carboxylic acid (mg/L)

Figure 4. Plot of the calculated versus known carboxylic acid concentrations in six field brines. Additional data were from the analyses of
Brine A and Brine B with addition of carboxylic acid (0-338 mg/L) to the brines. The calculated carboxylic acid concentration was determined
by fitting the titration result of this method and calculated with the Rice BCC Alk-TAc at [Link] program. The known carboxylic acid
concentrations in the field brine were measured by GC.
3.50
Pressure
3.00 15
20
2.50
Percentage CO2

50
2.00 100
200
1.50 500
1.00 1000
2000
0.50 5000
0.00 10000
15000
0

0
10

20

50

10 0
00
1

10

20

50

00

00

00

,0
1,

2,

5,

Barrels of water per day

Figure 5. Plot of the changes in gas phase CO2 concentration as a function of pressure and water production rate, where the CO2
concentration in the gas phase is assumed to be 3% at STP and the gas production rate of 1MMSCF. The equilibrium constants used are at
212 °F and 15 psia.
SPE 87449 11

TAc (M)
0.012

0.01 0
0.001

HCO3 conc., M
0.008 0.002
0.003
0.006 0.004
0.005
0.004
0.006
0.007
0.002
0.008
0 0.009

1000
100
200
500
10
20
50
0.1
0.2
0.5
1
2
5
0.01
0.011
PCO2, gas, psia

Figure 6. Plot of the variation of (HCO3 ) versus PCO2, gas and (TAc) concentrations, all at a final Alk = 0.01 eq/L ≈ 610 mg/l as HCO3 . The
- -

equilibrium constants used are at 212 °F and 15 psia and 1 M I. Note that with any specific well, the change in pressure will typically be a
factor of 100x, or less.

TAc (M)
9.00 0
0.001
8.00 0.002
0.003
7.00 0.004
0.005
pH

6.00 0.006
0.007
5.00 0.008
0.009
4.00 0.01
0.011
3.00
1
2
5
10
20
50
0.1
0.2
0.5

100
200
500
1000

PCO2, gas, psia

Figure 7. Plot of brine pH as a function of PCO2,gas and TAc concentrationsall at a final Alk = 0.01 eq/L ≈ 610 mg/l as HCO3 . The equilibrium
-

constants used are at 212 °F and 15 psia and 1 M I. Note that with any specific well, the change in pressure will typically be a
factor of 100x, or less.