SPE 74657
Effect of hydrate inhibitors on oilfield scale formation and inhibition
Amy T. Kan, Gongmin Fu, Malene A. Watson, and Mason B. Tomson, SPE, Rice University
Copyright 2002, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the SPE Oilfield Scale Symposium held in
Aberdeen, United Kingdom, 30–31 January 2002.
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
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
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
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
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.
Abstract
The effect of hydrate inhibitors on oilfield scale formation has
been studied. A self-consistent activity model is presented to
model the effect of methanol on carbonate equilibrium,
calcite, barite, gypsum, celestite, and halite solubility in a
gas/methanol/water/salt solution. The model uses Pitzer theory
to model the effect of salt and Born equation to model the
effect of methanol. Excellent agreements between the model
prediction versus both experimental and literature results are
observed. The model has been incroporated into
ScaleSoftPitzer®, a scale prediction software specifically
written for the oil and gas industries. From ScaleSoftPitzer®
predictions, scaling problems can be worsened with as little as
5% (by volume) of methanol in a typical oil and gas
production system. Ethylene glycol has a much lesser impact
on oilfield scale formation than methanol. The efficiency of
common phosphonate inhibitors in the methanol and ethylene
glycol containing brines will also be discussed.
Introduction
In the oil and gas industries, methanol and ethylene glycol are
often used to inhibit gas hydrate formation during production.
Gas hydrate is a crystalline solid consisting of gas molecule
surrounded by a cage of water molecules, which forms at
certain high pressure and low temperature regimes. Gas
hydrate formation is particularly troublesome for offshore gas
wells where the producing temperature is low due to both
adiabatic expansion of gas and seawater cooling. Once gas
hydrate forms, it can plug up the well and prevent gas
production. One economic solution to prevent hydrate
formation is to inject a large quantity of methanol or ethylene
glycol. However, methanol or ethylene glycol may cause
adverse scaling problems in the associated brine solution,
which often contains high concentrations of dissolved
minerals. There is little research on the solubility of mineral
salt in methanol/water/salt or ethylene glycol/water/salt
solutions. In the following, the solubility of mineral salts in
these mixed solvent systems are studied. Inhibition of scale
formation in the presence of methanol or ethylene glycol will
also be discussed. Lastly, the impact of methanol under
realistic oil field condition are predicted by ScaleSoftPitzer®, a
scale prediction software specifically written for the oil and
gas industries.
Self-Consistent Activity Coefficients in
Alcohol/water/salt
In principle, the free energy of transfer from a solution of one
dielectric constant to another dielectric constant (∆Gal/w =
∆Gtr) can be calculated via the Born equation and this in turn
is the medium-effect activity coefficient, γN:
∆G al / w = RT ln(γ N ) =
(1
N ⋅ z i2 ⋅ e 2 ⋅10 −10  1 1 
 −  (J / mole)
8 ⋅ π ⋅ r ⋅ ε 0  ε s ε w 
A discussion of the Born equation can be found in any of
several texts on the molecular nature of ions in mixed
alcohol/water solutions in that the water is selectively
concentrated around the ions 1, 2. The Born equation is strictly
approximate at all alcohol concentrations, but at greater than
50 to 80 %, w/w alcohol, the error can increase substantially1,
3. In this paper, we have simply used the Born equation as a
model template to suggest a functional form to our semi-
empirical approach. The reasonableness of this assumption
lies in how well it is able to represent the solution chemistry of
species of interest as a function of salt and methanol. In fact,
it has been shown to be an excellent assumption over the
entire range that we have tested, to date (essentially, up to
about 6 m salt, 85% w/w methanol and at 4 - 25ºC).
In this paper, a self-consistent set of activity coefficients
for mineral ions of interest to oil field brine versus salt and
methanol concentration is proposed. First, it is assumed that,
in each case, γoverall = γS⋅γN. For example, it is assumed that the
activity coefficient of aqueous carbon dioxide, γ CO ,aq , can be
2
written as a product of the effect of salt and of neutral
methanol, γ CO ,aq = γ SCO ,aq ⋅ γ CO
N
, aq
. γS is given by conventional
2 2 2
SPE 74657 EFFECT OF HYDRATE INHIBITONS ON OILFIELD SCALE FORMATION AND INHIBITION
method for activity coefficient, e.g., the Pitzer theory, with all
concentration units as “aqueous molality” (moles of solute per
kilogram of water). γN is a curve fitted parameter obtained
using an equation similar to the Born equation. The
concentration of all aqueous species is expressed in [] and is in
units of aqueous molality (moles/kg of water). It is important
to emphasize that all concentrations are expressed as moles
per kilogram of water, even in the presence of a substantial
quantities of methanol; this is as required for conventional
activity coefficient calculation, e.g., Pitzer theory, which uses
ions in pure water as the reference state in all conditions.
In Table 1 are listed the equilibrium parameters that have
been evaluated in this paper. Numerical values at 25 °C and 1
atm are used for illustration. The value of γ CO ,g can be
2
calculated using an equation of state for carbon dioxide, but at
room temperature and pressure it can be taken to be one
(1.00). The activity of water in methanol water solution is
determined by the ratio of vapor pressure of water over the
methanol/water mixture and over water. The term a * refers to
H+
the activity of the hydrogen ion in the mixed solvent, which is
corrected for medium effect in the mixed solvent via the ∆pHj
and δ-parameter2, 4, i.e.,
pH = − log(a *H + ) = pH meter reading + ∆pH j − δ
The magnitude of ∆pHj is between 0.07 to 0.23 for a salinity of
0.1 to 3 m NaCl solution, respectively. δ is less than 0.1 pH
unit when methanol concentration is less than 40% by weight.
Effect of hydrate inhibitors on mineral solubilities
A large number of halite, barite, calcite, celestite, and gypsum
dissolution and carbonate equilibrium experiments have been
conducted in the laboratory. In Table 2 is listed the range of
conditions that was used for these experiments. Due to the
limitation in space, detailed experimental procedure and
analytical results will be published in a separate paper 5.
Typical observations can be found in Figures 1 and 2.
Methanol significantly reduces the solubility of all mineral
salts tested in this study, i.e., halite, calcite, barite, celestite,
and gypsum. For example, the effects of methanol on three
sulfate mineral solubilities are illustrated in Figure 1. The
solubility of the three sulfate minerals are reduced by more
than one order of magnitude with about 0.35 mole fraction of
methanol, which is equivalent to approximately 50% (wt) in
methanol concentration. In addition to methanol, we have
limited data on the effect of ethylene glycol on barite and
halite solubility. Interestingly, there is a significant difference
in the effect of ethylene glycol and methanol on barite and
halite solubilities. Ethylene glycol reduces both barite and
halite solubilities to a lesser extent than methanol does(See
Figure 2).
The measured equilibrium concentrations from many
experiments (see Table 2), similar to those presented in
Figures 1 and 2, are compared to the equilibrium constants of
Table 1 after correcting for the temperature and salt effects
from the Pitzer theory and well established temperature
dependence of the equilibrium constants 6. The deviation
2
between experimental observation and the thermodynamic
equilibrium constants are attributed to γN. For the carbonate
system, γ CO
N can be derived from CO2 solubility experiments
2,g
and Henry's law. After γ CO
N is determined, γ N can be
2,g HCO −
3
determined from titration of bicarbonate and the first
ionization constant of carbonic acid. Likewise, γ N and γ N 2− 2+
CO Ca
3
can be determined from carbonate titration and calcite
dissolution experiments. Ideally, γ N can be determined from
SO 24 −
gypsum dissolution experiments with γ N determined from 2+
Ca
calcite dissolution experiments. However, there is insufficient
gypsum dissolution data available at this time (See discussion
below). Therefore, the mean activity coefficients
( )
N±
γ MA ≡ γ N ⋅ γ N are reported for barite, celestite, gypsum
M+ A−
and halite in this paper.
It can be shown that the dielectric term of Eq.
1, (1 ε s − 1 ε w ) , is a second order function of methanol
concentration (mole fraction, xMeOH). For methanol, eq 2 can be
derived:
Log ( γ N ) = ∆G al / w /(2.303 ⋅ R ⋅ T )
Z2
(2
= 3.628 ⋅10 ⋅ i (0.0115 ⋅ x MeOH + 0.0068 ⋅ x 2MeOH )
4
r ⋅T
It is proposed that the functional form of Eq. 2 can be used
to correlate all γ N s to methanol concentrations. The linear
term in Eq. 2 represents the primary Born effect of dielectric
constant, εs, on the free energy of the ions relative to water.
Experimentally, the first term represents the majority of the
observed data up to about 0.3 mole fraction. At higher
methanol concentration, it has been demonstrated that the
experimental data often deviates from the predictions of Born
equation1, 2, 7.
When we correlate γN values with methanol concentrations
using the functional form of Eq. 2, we obtained the
correlations listed in Table 3. In Table 3 also lists the standard
deviation of the curve fitted parameters and correlation
coefficient (see Table 2 for the number of data points used in
each correlation). Literature data on halite solubility in
ethanol/water solution was used to yield γ NNaCl ±
for ethanol8.
Generally, excellent correlations were observed between the
experimental data and model equations, except for γ N (for
HCO3−
methanol) and γ N±
(for ethylene glycol). This is because both
NaCl
γ N (for methanol) and γ NaCl (for ethylene glycol) are not
N±
HCO3−
significantly differed from one (1.00). These correlations have
been independently verified with calcite and barite dissolution
experiments where excess gas phase CO2, and solution phase
Ca and sulfate are present in a closed reaction vessel. These
analytical data will be presented elsewhere.
Interestingly, the first term of Eq. 2 for the monovalent and
divalent ions equal to 1.4 and 5.6, respectively, by assuming
an ion radius of 1 A at 25 ºC. For the correlations in Table 3,
SPE 74657 EFFECT OF HYDRATE INHIBITONS ON OILFIELD SCALE FORMATION AND INHIBITION 3

the first methanol terms are equals to 1.4 for γ N + , 3.4 for application. The above methanol activity coefficient ( γ N and
Na
N±
N±
γ N 2+ , 3.5 for γ N 2 − , 4.3 for γ CaSO ⋅2 H O
, 5.0 for γ N±
SrSO
, and 6.9 γ MA ) have been incorporated into the program. Details of the
Ca CO 3 4 2 4

N±
program have been discussed in an earlier paper by the
for γ which are certainly quite reasonable comparing to
BaSO 4 authors10. The calcite simulation is done for a brine containing
the prediction by Born theory. It should be emphasized that 4,972 mg/L Ca, 660 mg/L bicarbonate alkalinity, 69,801 mg/L
N± N±
the validity of equations for γ SrSO N±
γ BaSO and γ CaSO ⋅2 H O
total dissolved solid, equilibrated with 1% CO2 in the gas
4 4 4 2

should be restricted to the range of the experimental data, phase, at 25 ºC and 2,940 psig pressure. The barite scaling
which is approximately 0-0.36 mole fraction methanol (or 0 - tendency was simulated for a brine containing 57,211 mg/L
~50% by weight). The apparent gypsum solubility increases at TDS, 15 mg/L Ba, and 11 mg/L sulfate at 25 ºC and 2,940
methanol concentrations > 50% (data not shown). This may psig pressure. In both simulations, the brine composition is
imply a strong CaSO4 solution complex formation in the adjusted such that the SI is zero when xMeOH = 0. The
presence of high methanol concentrations. Preliminary data simulation shows that considerable barite or calcite scale will
also indicates the possibility of gypsum dehydration to form in the presence of as little as 5-20% methanol (by
semihydrate or anhydrite that would have a different solubility volume). The effect of methanol on halite scale formation can
limit. These hypotheses will be tested in the future. be evaluated from Figure 5. In Figure 5 is plotted the amount
of halite that will precipitate from brines versus TDS (assumed
Scale Prediction and implication. to be equal to NaCl concentrations) where different methanol
The solution supersaturation index can be calculated according concentrations (Vol%) are compared. In the absence of
to Eq. 3-7. methanol, NaCl is soluble in water up to 6.1 m or
approximately 318,000 mg/L TDS. However, NaCl will
 [ ][3 ]
 Ca 2 + ⋅ HCO− 2 ⋅ γ S 2 + ⋅ γ N 2 + ⋅ γ S − 2 ⋅ γ N − 2 ⋅ K 
2  (3
precipitate out of a brine containing 250,000 mg/L TDS when
= log 
Calcite Ca Ca HCO 3 HCO 3
SI  30% methanol is added to the system. At 50% methanol, the
 PCO 2 ⋅ γ CO 2 , g ⋅ K H ⋅ K1 ⋅ K Calcite
sp  system may experience halite problem if the brine contains >
210,000 mg/L TDS, etc.
 [ ][ ]
 Ba 2 + ⋅ SO 24 − ⋅ γ S 2 + ⋅ γ S 2 − ⋅ γ BaSO
N± 2
4 

(4)
= log 
Barite Ba SO 4
SI Barite  Effect of methanol and ethylene glycol on barite
 K sp  scale nucleation and inhibition
It has been observed that barite nucleation rate is significantly

 [ ][ ]
 Ca 2+ ⋅ SO 24− ⋅ γ S 2 + ⋅ γ S 2− ⋅ γ CaSO
N± 2
⋅ a SH 2O ⋅ a HN2O 
2 2 affected by the addition of methanol. We have systematically
measured the barite nucleation time at both a varied methanol
SI Gypsum = log 
Ca SO 4

4

Gypsum
 K sp  concentrations (0 - 40% by wt.) and a varied barium and
sulfate (at equal molal) concentrations from 0.5 -1.1
(5)
mmoles/Kg H2O. The experiments were done by first mixing a
[ ][ ]
barium containing solution and a sulfate containing solution.
 Sr 2 + ⋅ SO 2 − ⋅ γ S 2 + ⋅ γ S 2 − ⋅ γ N ± 2 
 4 SrSO 4  (6) The onset of nucleation was monitored by a turbidity meter
SI Celestite = log 
Sr SO 4

 K Celestite
sp  over time using a data logger 11. The induction time is
operationally defined as the time when the rapid increase in
turbidity is observed. In Figure 6a is plotted the turbidity
SI Halite = log 
[ ][ ]
 Na + ⋅ Cl− ⋅ γ S + ⋅ γ S − ⋅ γ NNaCl
± 2
 (7) reading versus logarithmic reaction time in seconds for seven

Na Cl

 K Halite
sp 
nucleation kinetic experiments where the only variable is the
methanol concentrations. In this example, the barium solution
It should be noted that Eq. 3 is independent of pH9. The model contains 1 m NaCl, 0.09 m Ca, 1.50 mm Ba and 5 mm PIPES
parameters for γN and γ MA
N±
can be depicted in the nomogram of buffer at pH 6.4. The sulfate solution contains 1 m NaCl, 0.09
m Ca, 1.44 mm sulfate, and 5 mm PIPES buffer at pH 6.4.
Figure 3. Figure 3 can be used to estimate activity coefficients
N±
Methanol was added to both the cationic and anionic solution
(γN and γ MA ) and dielectric constant values in methanol
to a fixed concentration. An equal volume of these two
solutions at 25 °C, as well as to correlate between vol%, wt%, solutions were then mixed and the turbidity of the solution
N±
and mole fraction. Combining γ N and γ MA with common was monitored. As shown in Figure 6a, barite induction time
activity coefficient calculations, e.g., Pitzer theory, the was affected by as little as 5% (by wt) of methanol and the
saturation index of calcite, barite, celestite, gypsum, and halite induction time is shortened more at higher methanol
at a given methanol concentration can be predicted. concentrations.
In Figure 4 is plotted the predicted change in mg/L of scale 0
It has been observed that barite nucleation time ( t ind , sec)
formation and saturation index of calcite and barite versus the
is related to the supersaturating state of the solution (SI),
methanol concentration (% vol) under realistic field condition.
The calculations were done with ScaleSoftPitzer®, which is temperature (ºK) and methanol concentration by Eq. 8:
written in Microsoft Excel® program specifically for oil field
SPE 74657 EFFECT OF HYDRATE INHIBITONS ON OILFIELD SCALE FORMATION AND INHIBITION
log10 ( t 0ind , sec)
 [1087.2 − 0.297 ⋅ T ]
2
[1087.2 − 0.297 ⋅ T ] 
= − 2.235 + 2
− 0.123
 T SI Brine T 3SI Brine
− 13.68 ⋅ x MeOH
In Eq. 8, the first term represents the effect of ions and
temperature on barite nucleation11 and the second term is an
empirical function to account for the effect of methanol on
nucleation rate. Very good agreement (r = 0.97) between the
observed and calculated induction time (log10( t 0ind )) are
observed (Figure 6b).
The inhibitory effect of a phosphonate scale inhibitor
(BHPMP, Bis-hexamethylene triamine-penta(methylene
phosphonic) acid) on baite nucleation has been studied at 0-
40% methanol or ethylene glycol and 0.7 to 1.1 mm barium
and sulfate concentrations. BHPMP is one of the most
effective barite inhibitors12. In Figure 7 is plotted a typical set
of nucleation studies where barium and sulfate concentrations
are 1.1 mm (SI=2.0). The solutions also contain 1 m NaCl,
0.09 m Ca, 5 mm PIPES buffer (pH 6.4) and various
concentrations of BHPMP and methanol or ethylene glycol.
The six plots in Figure 7 are grouped by different methanol or
ethylene glycol concentrations. In each plot of Figure 7,
multiple nucleation data are plotted. These nucleation
experiments were done at identical solution compositions,
except for BHPMP concentrations. Increasing concentrations
of BHPMP were added to successive experiments until a
BHPMP concentration capable of inhibiting barite
precipitation up to ~24 hours (logt=4.94) was observed. As
shown in Figure 7a, only 0.33 mg/L BHPMP was needed to
inhibit the barite nucleation for 24 hours when no methanol
was present. However, more BHPMP was needed at higher
methanol concentrations. For example, 0.65, 1.90 2.52 mg/L
BHPMP were needed to inhibit barite nucleation in 10-30%
methanol. Even though only 2.52 mg/L BHPMP was needed
to inhibit barite precipitation in 30% methanol, low turbidity
was observed at 3-5.3 mg/L BHPMP for the same solution.
The reason for the deteriorating inhibition effect at higher
BHPMP concentrations is probably caused by the precipitation
of Ca-BHPMP at higher methanol concentrations. The
solubility of BHPMP in a similar sulfate free solution is
measured to be ~2.0 mg/L in 30% methanol concentration and
~1.2 mg/L in 40% methanol concentration. If the phosphonate
inhibitor solubility in methanol solution is limited, there
should be a limiting methanol concentration that no
phosphonate inhibitor will work. As shown in Figure 7e, no
inhibitory effect was observed for BHPMP up to 5.8 mg/L
concentration at 40% methanol. Interestingly, only 0.38 mg/L
BHPMP is needed to inhibit barite precipitation from a similar
solution containing 40% ethylene glycol (Figure 7f). These
results are consistent with the solubility data, indicating a
significant advantage in using ethylene glycol to control
hydrate formation.
4
Conclusion
A new activity model (ScaleSoftPitzer®) is proposed to model
3
2  the methanol effect on mineral salt scaling tendency in oil and
 gas production system. The model uses Pitzer theory to model
the effect of salt and Born equation to model the effect of
(8) methanol. The model predicts potentially significant barite and
calcite scaling problem with as little as 5% to 20% methanol
in a well. Halite scale problem can occur at >210,000 mg/L
TDS when 50% (vol.) methanol is present in the production
system. Barite nucleation rate is accelerated in as little as 5%
methanol. A semiempirical equation has been developed to
predict the kinetics of barite scale formation as a function of
methanol concentration. More scale inhibitors are needed to
inhibit barite scale when methanol is added to the production
system. Barite scale control may become impossible when
substantial amount of methanol are used for hydrate control.
Ethylene glycol has less adverse effects than methanol in both
scale formation and inhibition.
Acknowledgement
The financial support of Rice University Brine Chemistry
Consortium of companies: Aramco; B.J. Services; Baker-
Petrolyte; British Petroleum, Champion Technologies, Inc.;
Chevron Petroleum Technologies, Inc.; Conoco, Inc.; Ondeo
Nalco, Shell, Solutia, and Texaco, Inc. to this research is
greatly appreciated.
Nomenclature
A = an anion
*
aH +
= the hydrogen ion activity in the mixed solvent
aH O = the activity coefficient of water
2
a HS 2O =the activity coefficient of water due to the salt
effect
a HN2O = the activity coefficient of water due to the alcohol
effect
aM and aA = the activities of a cation and anion
aq = the aqueous phase species
e = the elementary charge (=1.6x10-19 coulomb)
g = the gas phase species
I = the ionic strength (moles/Kg H2O)
K spMA = the solubility product of a mineral salt
KH = the Henry's law constant (mole/Kg H2O-atm)
K1 and K2 = the first and second dissociation constants of
carbonic acid
M = a cation
[] = concentrations in moles/Kg H2O
N =the Avogadro constant (= 6.02x1023 mol-1)
pHmeter reading = the observed pH-meter reading in the mixed
solvent when the pH electrode has been
calibrated using normal aqueous buffers.
∆pHj = the correction term to represent the changes in
electrode response due to the presence of salt.
∆pHj can be estimated from the following eq.
∆pH j = 0.129 ⋅ I 0.5 r =0.99
SPE 74657 EFFECT OF HYDRATE INHIBITONS ON OILFIELD SCALE FORMATION AND INHIBITION 5

r = the ion radius (A) 9.Oddo, J.E. and Tomson, M.B.:"Why Scale Forms in the Oil Field
SI = the saturation indices of mineral salt and Methods to Predict It," SPE production & facilities,
SIBrine =the barite SI of the brine in the absence of methanol February, 1994 (1994), 47-54.
T = temperature (ºK) 10.Kan, A.T., Fu, G., and Tomson, M.B. "OTC13236 Mineral scale
control in subsea completion.". in 2001 Offshore Technology
tind = the barite induction time (sec) Conference. (2001). Houston, TX: SPE.
xMeOH = mole fraction concentration of methanol in 11.Xaio, J.A., Kan, A.T., and Tomson, M.B.:"Prediction of BaSO4
methanol-water mixture
precipitation in the presence and absence of a polymeric inhibitor:
xEG = mole fraction concentration of ethylene glycol in Phosphino-polycarboxylic acid," Langmuir, 17 (2001), 4668-
ethylene glycol-water mixture 4673.
xEtOH = mole fraction concentration of ethanol in 12.He, S.L., Kan, A.T., and Tomson, M.B.:"Mathematical Inhibitor
ethanol-water mixture Model for Barium Sulfate Scale Control," Langmuir, 12 (1996),
zi = the ion charge 1901-1905.
δ = the correction term used to represent the changes
in electrode response due to methanol. δ can be
estimated from the following eq.
δ = xMeOH
2
⋅ {− 0.897 ⋅ exp( 3.262 ⋅ xMeOH
2
) − 1.309 ⋅ exp( −4.008 ⋅ xMeOH
3
)}
r = 1.00
ε0 = the vacuum permittivity (= 8.854x10-12J-1·C2·m-1)
εs = the dimensionless dielectric constant of the mixed
solvent
εw = the dimensionless dielectric constant of water
γS⋅ = the activity coefficient due to the salt effect
γN = the activity coefficient due to alcohol effect
γ MS and γ AS = the acitivty coefficient of a cation or anion due
to the salt effect
γM N
and γ A the acitivty coefficient of a cation or anion due
N
=
to the cosolvent effect
γN± =the mean activity coefficient due to cosolvent
effect

References
1.Robinson, R.A. and Stokes, R.H.,"Electrolyte Solutions: The
Measurement and Interpretation of Conductance, Chemical
Potential and Diffusion in Solutions of Simple Electrolytes". 2nd
ed. Butterworth & Co., London, 571. London (1970).
2.Bates, R.G.,"determination of pH - Theory and Practice". Second
Edition ed. A Wiley-Interscience Publication. Canada (1973).
3.Sen, B., Roy, R.N., Gibbons, J.J., Johnson, D.A., and Adcock, L.H.,
in "Thermodynamic behavior of electrolytes in mixed solvents-
II", (W.F. Furter, Editor). p. 215-248. ACS, Washington, DC,
(1978).
4.Brezinski, D.P.:"Kinetic, static and stirring errors of liquid junction
reference electrodes," The analyst, 108 (1983), 425-442.
5.Kan, A.T., Fu, G., and Tomson, M.B.:"Effect of methanol on
carbonate equilibrium and calcite solubility in a
gas/methanol/water/salt mixed solution," (To be submitted to
Langmuir), (2002).
6.Langmuir, D.,"Aqueous Environmental Geochemistry". Prentice
Hall. Upper Saddle River, NJ (1997).
7.Sen, J. and Gibbons, J.J.:"Ionic Equilibria in Mixed Solvents:
Formation of Calcium Lactate," Journal of Chemical and
Engineering Data, 22 (1977), 309-314.
8.Stephen, H. and Stephen, T.,"Solubilities of inorganic and organic
compounds, Vol 1 Binary Systems, Part 2". The MacMillan
Company. New York (1963).
Table 1. Equilibrium parameters tested in methanol/water/salt solution; values are at 25 ºC and 1
atm with ions in pure water as a reference state.
molal a CO , aq [CO 2, aq ]⋅ γ CO , aq [CO 2, aq ]⋅ γ CO ⋅ γ CO
S N

K H = 10 −1.47 = = 2
= 2 2, g 2

atm f CO , g PCO , g ⋅ γ CO , gw
PCO ⋅ γ CO , g 2 2 2,g 2

− ( pH meter + ∆pH j − δ ) − − ( pH meter + ∆pH j − δ )

a *
H+
⋅ a HCO − 10 reading
⋅ [HCO ] ⋅ γ HCO −
3 10 reading
⋅ [HCO 3− ] ⋅ γ SHCO − ⋅ γ HCO
N
−
K1 = 10− 6.35 = 3
= 3
= 3 3

a CO 2 , aq [CO 2 ,aq ] ⋅ γ CO 2 ,aq [CO 2 ,aq ] ⋅ γ SCO 2 , aq ⋅ γ CO

N
2 , aq

a *H + ⋅ a CO 2− 10
(
− pH meter reading + ∆pH j − δ )
[ ]
⋅ CO32 − ⋅ γ CO 2− 10
(
− pH meter reading + ∆pH j − δ )
[ ]
⋅ CO 32 − ⋅ γ SCO 2− ⋅ γ CO
N
2−
K 2 = 10−10.33 = = =
[HCO ]⋅ γ [HCO ]⋅ γ ⋅ γ
3 3 3 3
− − S N
a HCO − 3 HCO 3− 3 HCO 3− HCO 3−

= [Ca ]⋅ γ ⋅ [CO ]⋅ γ = [Ca ]⋅ γ ⋅ [CO ]⋅ γ

3
calcite
K sp = 10 −8.36 =a ⋅aCa 2+
CO 32 −
2+
Ca 2 +
2−
3 CO 32 −
2+ S
Ca 2 +
⋅γ N
Ca 2 + 3
2− S
CO 32 −
⋅ γ CO
N
2−

= [Ba ]⋅ γ ⋅ [SO ]⋅ γ = [Ba ]⋅ γ ⋅ [SO ]⋅ γ

3

K spBarite = 10 −9.69 =a ⋅aBa 2 + SO 24 −

2+
Ba 2 +
2−
4 SO 24 −
2+ S
Ba 2 +
⋅γ N
Ba 2 +
2−
4
S
SO 24 −
⋅ γ SO
N
2−
4

= [Ba ]⋅ γ ⋅ [SO ]⋅ γ ⋅ (γ )
2+ S 2− S N± 2
Ba 2 + 4 SO 24 − BaSO 4

Celestite
K sp = 10 −6.47 =a ⋅a = [Sr ]⋅ γ ⋅ [SO ]⋅ γ
Sr 2 + SO 24 −
= [Sr ]⋅ γ
2+
Sr 2 +
2−
4 SO 24 −
2+ S
Sr 2 +
[ ]
⋅ γ SrN 2 + ⋅ SO 24 − ⋅ γ SSO 2 − ⋅ γ SO
N
2−
4 4

= [Sr ]⋅ γ ⋅ [SO ]⋅ γ ⋅ (γ )
2+ S 2− S N± 2
Sr 2 + 4 SO 24 − SrSO 4

Gypsum
K sp = 10 −4.63 =a ⋅a ⋅a
Ca 2 +
= [Ca ]⋅ γ
SO 24 −
⋅ [SO ]⋅ γ
2
H 2O ⋅a 2+
Ca 2 +
2−
4 SO 24 −
2
H 2O

= [Ca ]⋅ γ ⋅ [SO ]⋅ γ ⋅ (γ ) ⋅ (a ⋅ a )
2+ S 2− S N± 2 s N 2
Ca 2 + 4 SO 24 − CaSO 4 H2O H 2O

K spHalite = 101.59 = a ⋅ a = [Na ]⋅ γ ⋅ [Cl ]⋅ γ = [Na ]⋅ γ ⋅ γN

Na + Cl −
+
Na +
−
Cl −
+ S
Na +
S
Na +
[ ]
⋅ Cl − ⋅ γ SCl − ⋅ γ Cl
N
−

= [Na ]⋅ γ ⋅ [Cl ]⋅ γ ⋅ (γ )
+ S
Na +
− S
Cl −
N± 2
NaCl

Table 2. Range of equilibrium experimental conditions studied in this paper.

Exp. type Phase Cosolvent Cosolvent Ionic T (°C) No. of
Concentratio strength, I Exp.
n Range (moles/Kg
(wt%) H2O)
Dissolutio Halite Methanol 0 - 90 2.9 - 6.1 5 - 25 25
n
Dissolutio Halite Ethylene 0 - 50 3.8 - 6.1 25 5
n Glycol
Dissolutio Barite Methanol 0 - 46 0-3 4 - 25 21
n
Dissolutio Barite Ethylene 0 - 70 1-3 4 - 25 15
n Glycol
Dissolutio Celestite Methanol 0 - 55 1-3 4 - 25 16
n
Dissolutio Gypsum Methanol 0 - 90 0-3 25 18
n
Dissolutio Calcite Methanol 0 - 85 0-3 4 - 25 48
n
Dissolutio CO2 (gas) Methanol 0 - 80 0 - 2.9 4 - 25 30
n
Titration Bicarbonat Methanol 0 - 84 0-3 4 - 25 36
e
Titration Carbonate Methanol 0 - 85 0-3 4 - 25 69
SPE 74657 EFFECT OF HYDRATE INHIBITONS ON OILFIELD SCALE FORMATION AND INHIBITION
SPE 74657 EFFECT OF HYDRATE INHIBITONS ON OILFIELD SCALE FORMATION AND INHIBITION 8

Table 3. The correlation of activity due to methanol effect (γγ and γ ±) with methanol concentrations (mole
N N±

fraction), ionic strength and temperature.

log( γ CO
N
2 , aq
) = ( −6.029 + 1444.9 / T (o K )) ⋅ x MeOH − 1.170 ⋅ x 2MeOH r = 1.00
Standard deviations for the three parameters are 0.612, 175.4, and 0.135, respectively.

log(γ HCO
N
− ) = ( −3.338 + 955.9 / T ( K )) ⋅ x MeOH + 0.565 ⋅ x MeOH
o 2 r = 0.85
3

Standard deviations for the three parameters are 0.929, 260.8, and 0.211, respectively.

log(γ CO
N
2 − ) = ( −1.957 + 1580.9 / T ( K )) ⋅ x MeOH − 1.601 ⋅ x MeOH
o 2 r = 0.99
3

Standard deviations for the three parameters are 0.520, 150.9, and 0.097, respectively

log( γ Ca
N
2 + ) = (19.343 − 4714.2 / T ( K ) − 0.333 ⋅ I ) ⋅ x MeOH − 2.519 ⋅ x MeOH
o 2
r = 0.97
Standard deviations for the three parameters are 0.861, 247.2, 0.0316 and 0.189, respectively

N±
log( γ CaSO4 ⋅2 H 2 O
) = (4.706 − 0.423 ⋅ I) ⋅ x MeOH - 4.153 ⋅ x 2MeOH r = 0.97
Standard deviations for the three parameters are 0.314, 0.092, and 1.205, respectively

N±
log( γ SrSO 4
) = (4.963 − 0.206 ⋅ I) ⋅ x MeOH − 4.017 ⋅ x 2MeOH r = 1.00
Standard deviations for the three parameters are 0.221, 0.0484, and 0.613, respectively

N±
log( γ BaSO 4
) = (3.022 + 1167.6 / T( o K)) ⋅ x MeOH − 10.89 ⋅ x 2MeOH r = 0.99
Standard deviations for the three parameters are 2.507, 723.2, and 1.046, respectively

N±
log( γ BaSO4
) = 2.590 ⋅ x EG - 5.316 ⋅ x 2EG r = 0.52
Standard deviations for the three parameters are 0.623 and 1.684, respectively

±
log(γ NNaCl ) = 1.385 ⋅ x MeOH - 1.104 ⋅ x 2MeOH r = 0.96
Standard deviations for the three parameters are 0.070 and 0.110, respectively

±
log(γ NNaCl ) = 1.366 ⋅ x EtOH r = 0.99
Standard deviations for the parameter is 0.018.

±
log( γ NNaCl ) = −0.429 ⋅ x EG r= 0.90
Standard deviations for the parameter is 0.118.
 1.00E-01

1.00E-02 Gypsum

Ba, Sr, or Ca Conc.

(moles/Kg H 2 0) 1.00E-03 Celestite

Barite
1.00E-04

1.00E-05

1.00E-06
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Methanol Conc. (mole fraction)

Figure 1. Plot of Ba, Sr, and Ca concentrations versus methanol concentrations measured in the dissolution experiments
where barite, celestite and gypsum solids were dissolved in methanol-NaCl (1 m)-H2O solutions at 25 ºC.

7
0.12
BaSO4 conc. (mm)

6
NaCl Conc. (m)

1 m NaCl EG
0.10 5
0.08 EG 4 MeOH
0.06 3
0.04 2
0.02 MeOH 1
0.00 0
0 0.1 0.2 0.3 0.4 0.5 0.0 0.1 0.2 0.3 0.4 0.5

Cosolvent Conc. (mole fraction) Cosolvent Conc. (mole fraction)

Figure 2. Comparison of barite and halite solubilities in methanol-NaCl-Water and Ethylene glycol-NaCl-water solutions at
25 C. In the figures are plotted (a) the aqueous phase Ba and (b) the aqueous phase Cl concentrations measured in barite
and halite dissolution experiments versus cosolvent concentrations. The barite dissolution experiments were done in
cosolvent - NaCl (1m) - water solution and the halite dissolution experiments were done in cosolvent - water solutions.
 Calcite Halite Sulfate
100 5 16
γ CO
N
2− SrSO4
3

4
10 γ Ca
N
2+ 12
γ N
N±
γN 1
HCO −3 γ NaCl 3
NaCl
N±
γ MSO
BaSO4
4 8

0.1 2
CaSO4·2H2O
γ N
CO 2 , aq 4
1
0.01
0 80
100 80 100 80

(Vol% or Wt%)
Methanol Conc.
80 70
(Vol% or Wt%)

(Vol% or Wt%)
Methanol Conc.

70

Methanol Conc.
80 Vol.% Vol.%
Wt.% Wt.% 40 70
60 60
60 60
εs εs
εs εs
50 40 50 Vol.%
40 60
εs εs
20
Wt.%
20 40 20 40

0 30 0 30 0 50
0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 0 0.1 0.2 0.3
Methanol Conc. (mole fraction) Methanol Conc. (mole fraction) Methanol Conc. (mole fraction)

N±
Figure 3. Nomograms to determine γ N, γ MA and solution dielectric constants for calcite, barite and halite from either methanol vol%, wt%, or mole fraction concentrations,
N±
where γN and γ MA values are calculated from equations listed in Table 3. For illustration, arrows are drawn to show the users how to find all parameters at 50% (vol)
±
methanol concentrations. All data is for 25 ºC. For example, 50 vol % methanol = 44 wt % = 0.308 mole fraction; γ N 2− = 7.5, γ N 2+ = 5.6, γ N − =1.0, γ N = 0.3, γ NNaCl = 2.1,
CO Ca HCO 3
CO 2 ,aq 3

N± N± N±
γ SrSO 4
= 12.1, γ BaSO
4
= 12.7, γ CaSO
4 ⋅2 H 2 O
= 8.4, and εs = 59.
 Calcite
600 2.5

(mg/L Calcite)
500 2

Amount ppt
a.
400
1.5
300 SI ppt
1
200 SI
100 0.5
0 0
0 20 40 60 80 100
Methanol Conc. (Vol%)

Barite
30 2.5
25
(mg/L Barite)

b. 2
Amount ppt

20
1.5
15 SI ppt
1
10 SI
5 0.5
0 0
0 10 20 30 40 50 60
Methanol Conc. (Vol %)

Figure 4. Plots of (a) amount of calcite scale precipitated (mg/L) and calcite SI and (b) amount of barite scale precipitated
(mg/L) and barite SI versus the methanol concentrations where the condition for the simulation is discussed in text.

75% 60% 50%

40% 30%
150 εal/w
NaCl ppt (g/L)

100

50

0
180 200 220 240 260 280
TDS (g/L)

Figure 5. Plot of the amount of NaCl precipitated (mg/L) versus the brine TDS when various volume % of methanol is
present in the brine. The condition for the simulation is discussed in the text.
a.

MeOH
20
0%
15
5%
10 10%
14%
5
18%
0 23%
1 10 100 1000 10000
40%
Reaction Time (sec)

b.
Ba and
5
SO4 Conc.
Pred. log (Induction time) (sec)

4
0.5 mm
3 0.72 mm
1.1 mm
2
1:1 line
1

0
0 1 2 3 4 5
Observed log (Induction time) (sec)

Figure 6. Plots of (a) the turbidity reading of a solution after mixing an equal volume of a cationic solution (1 m NaCl, 0.09
m Ca, and 1.50 mm Ba) and an anionic solution (1 m NaCl, 0.09 m Ca, and 1.44 mm SO4) at 25 ºC versus time (sec). The
final Ba and SO4 concentrations are 0.75 and 0.72 mm, which correspond to SIbarite = 1.6 in the absence of methanol; and (b)
the predicted logarithmic induction time (sec) by Eq. 8 versus observed logarithmic induction time (sec). The data were
from nucleation kinetics experiments at three different Ba and SO4 concentrations and 0 - 40% methanol concentrations.
a.
50 0% MeOH BHPMP
40
30 0.22 mg/L
20 0.27 mg/L
10 0.33 mg/L
0
0 1 2 3 4 5
b. 50
10% MeOH
e.
40 0 mg/L
30 0.43 mg/L 40% EG
Turbidity (NTU)

BHPMP

Trubidity (NTU)
20 0.52 mg/L 50
40 0 mg/L
10 0.65 mg/L
30 0.19 mg/L
0
20
0 1 2 3 4 5 0.29 mg/L
10
c. 20% MeOH 0 0.27 mg/L
50 0.38 mg/L
0 mg/L 0 1 2 3 4 5
40
0.63 mg/L
30 Log t (sec)
1.28 mg/L
20
1.59 mg/L
10
1.71 mg/L
0
1.90 mg/L
0 1 2 3 4 5
d.
50 30% MeOH
0 mg/L
40 1.56 mg/L
30 2.04 mg/L
20 2.52 mg/L
10 3.02 mg/L
0 5.32 mg/L
e. 0 1 2 3 4 5

50
40% MeOH
0 mg/L
40 1.58 mg/L
30 2.62 mg/L
20 3.11 mg/L
10 3.63 mg/L
0 4.78 mg/L
0 2 4 5.83 mg/L

Log t (sec)
Figure 7. Plots of the turbidity reading versus time following the mixing of a barium containing solution (1 m NaCl, 0.09 m
Ca, 1.1 m Ba, pH 6.4) and a sulfate containing solution (1 m NaCl, 0.09 m Ca, 1.1 m SO4, pH 6.4) in the presence of 0 - 40%
(wt) methanol or ethylene glycol and various concentrations of BHPMP at 25 C. The mixed solution is supersaturated with
respect to BaSO4. In the absence of methanol, the SI for barite = 2.0.