See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/288160605

Corrosion by HCl in dry steam wells controlled using potassium carbonate

without destroying superheat

Article  in  Transactions - Geothermal Resources Council · January 2010

CITATIONS READS

3 487

2 authors, including:

Oleh Weres
Chemtreat
61 PUBLICATIONS   964 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Kesterson View project

All content following this page was uploaded by Oleh Weres on 08 June 2017.

The user has requested enhancement of the downloaded file.

 Corrosion by HCl in dry steam wells controlled using potassium carbonate
without destroying superheat

Oleh Weres
PowerChem Technology, 2555 Business Parkway, Minden, NV 89423
oleh@powerchemtech.com

Cheri Kendrick
Bottle Rock Power LLC, P.O. Box 326, Cobb, CA 95426
ckendrick@bottlerockpower.com

Keywords: Hydrogen chloride, HCl, corrosion, dry steam, superheated, caustic, potassium
carbonate, Bottle Rock

This paper will be presented at the Geothermal Resources Council 2010 Annual Meeting,
October 24-27, 2010 in Sacramento, CA and will be published in the Transactions of that
meeting.

Abstract

An increasing amount of the dry geothermal steam produced in Sonoma and Lake Counties,
California contains sufficient hydrogen chloride (HCl) to severely corrode well casing and
surface piping. Sodium hydroxide is injected into the surface piping to neutralize HCl, but
requires that enough water be injected with it to desuperheat the steam and provide sufficient
liquid to keep the reaction products in solution. Sodium hydroxide and water cannot be injected
downhole for this purpose because localized cooling of the casing might induce condensation of
acidic and highly corrosive salt solutions. Also, the concentrated caustic employed is itself
corrosive and hazardous.

Potassium carbonate is much less hazardous than caustic, but equally effective in neutralizing
HCl. Excellent corrosion control has been demonstrated in field trials using potassium
carbonate.1 Because potassium carbonate and the potassium salts produced are highly soluble
up to high levels of superheat, only a small amount of carrier water needs to be added, and the
steam remains superheated with no risk of inducing localized corrosion.

Introduction

Geothermal steam contains a variety of impurities. Carbon dioxide and hydrogen sulfide usually
are present, and ammonia, hydrogen chloride (HCl) and/or boric acid may also be present. The
concentration of hydrogen chloride sometimes is high enough to corrode well casings, well
heads, steam pipelines and other steam gathering equipment or to damage turbines powered by
the steam. This problem is most particularly encountered in relation to dry steam produced
directly from the ground; for example, in the large geothermal field in Sonoma County and Lake

1
This method of controlling corrosion is patent pending.
Weres and Kendrick

County, California.

The boiling point of pure water (also called saturation temperature Tsat) is a unique function of
pressure Tsat(P). When temperature T = Tsat(P) at a given pressure P, pure water vapor or steam
can coexist at equilibrium with pure water. Likewise, steam and water can coexist at a given
temperature T when P = Psat(T). (The saturation pressure Psat(T) is the same mathematical
relation as Tsat(P) with the variables T and P reversed.) If, however, T > Tsat(P) the steam must
be dry, and a small amount of liquid water will evaporate if brought into contact with the steam.
In this condition steam is said to be superheated, and the degree of superheat is defined as

ΔTsup = T – Tsat(P)

Likewise, steam is superheated when P < Psat(T). Thus, steam which starts out saturated when
first separated from liquid water (that is, T = Tsat(P) and P = Psat(T)) can become superheated
either of two ways:

(a) By addition of heat to the steam thereby increasing the temperature at constant P, or
(b) By decrease of the pressure at constant T.

If there is a dry zone around the wellbore where the steam enters the well, both processes will
contribute to creating superheat; that is, additional heat will be added to the steam as it flows
through the dry rock near to the wellbore, increasing temperature, and the pressure will decrease
as it flows. In this case, the produced steam may become superheated by many degrees.

If there is liquid water present near to the wellbore, the steam entering the wellbore will be
saturated or nearly so. However, the inevitable pressure drop as the steam flows up the well and
through the surface pipes typically will produce a few degrees of superheat.

The functions Tsat(P) and Psat(T) describe the equilibrium of pure water with pure water vapor or
steam. If dissolved salts are present in the water, the vapor pressure of the solution will be less
than Psat(T) because the ions effectively dilute the water in the solution, and the intense electric
fields of the ions tend to tie down water molecules and keep them in the liquid phase. In other
words, a salt solution will be at equilibrium with water vapor at some value of P < Psat(T), and,
therefore, a salt solution can coexist with superheated steam. A salt solution that coexists with
superheated is also said to be superheated.

The degree of superheat of steam at equilibrium with a salt solution will increase with the
concentration of salt in the solution. If a small amount of salt solution is brought into contact
with superheated steam, water will evaporate from the solution until the concentration of salt in
the solution is sufficient to match the superheat of the steam. (Or the solution may absorb water
from the steam until the superheat is equal, if the solution is initially more concentrated.)
However, if the superheat is increased beyond a certain point, the concentration of the salt
solution will exceed the solubility of the salt, and the solution will completely dry out with solid
salt remaining. If the solution contains more than one salt and at least one of the salts is
sufficiently soluble, the less soluble salts will precipitate as water evaporates until the remaining
solution is at equilibrium with the superheated steam.

2
Weres and Kendrick

The maximum superheat that a solution of a given salt can support is determined by the
solubility of the salt or, in the case of ammonium chloride, by the volatility of the salt. The more
soluble the salt, the greater the degree superheat of the steam in equilibrium with a saturated
solution of that salt. Some very highly soluble salts can form highly concentrated solutions in
contact with strongly superheated steam, wherein the weight fraction of water is less than 50%.
A salt is said to be capable of supporting a certain degree of superheat if contact of a solution of
that salt with a large volume of steam characterized by that degree of superheat will not cause the
solution to dry out.

Contact of superheated geothermal steam with a metallic surface sometimes produces a highly
concentrated liquid phase which corrodes the metal. For example, if HCl and ammonia are both
present in the steam, they will react to form ammonium chloride and, because ammonium
chloride is very highly water soluble and deliquescent, a concentrated solution of ammonium
chloride is produced:

HCl(g) + NH3(g) → NH4Cl(aq)

Sufficient HCl and ammonia sometimes are present in the Sonoma/Lake County steam field to
produce a solution of ammonium chloride that can be superheated by as much as 11°C before the
ammonium chloride evaporates together with the water and the solution dries out. This solution
is acidic and aggressively corrodes well casings, steam pipes and surface equipment.

When steel is attacked by HCl in the absence of oxygen, the initial reaction product is another
highly soluble salt, ferrous chloride:

2 HCl(g) + Fe0 → H2 + FeCl2(aq)

At about 171°C – a typical temperature for geothermal steam – a concentrated solution of

ferrous chloride can exist in contact with steam superheated by as much as 24°C. If the degree
of superheat is such that can be supported by ferrous chloride, a solution of ferrous chloride can
form, and some of the HCl present in the steam will dissolve in this solution, making it highly
corrosive. This means that once corrosion starts, a highly corrosive liquid phase can form which
causes corrosion to accelerate; that is, positive feedback.

Possible formation of these corrosive liquids and corrosion of the casing is governed by the
degree of surface superheat; that is, the degree of superheat determined by the temperature and
pressure at the inner surface of the casing, where corrosion actually takes place. The most
corrosive conditions are encountered in the shallower part of the well where the steam flowing
up the well is hotter than the surrounding rock. Therefore, some heat is lost from the well by
conduction, and the casing is cooler than the steam. The conductive heat flux through the casing
can be calculated from the temperature gradient in the well and mass flow, and the temperature
difference between steam and casing can be estimated based on the temperature gradient;
typically, the inner surface of the casing is about 5°F = 3°C cooler2 than the steam, whereby the

2
These values refer to the difference between two temperatures rather than to a

3
Weres and Kendrick

degree of superheat at the metal surface (“surface superheat”) is smaller by an equal amount than
the superheat of the steam flowing through the pipe (“bulk superheat”).

While a corrosive solution of ferrous chloride can exist up to 24°C superheat and corrosion can
continue once initiated, corrosion is more likely to commence at a lower degree of superheat
which is more conducive to formation of the corrosive liquid. Hirtz et al. (1991) suggest that
corrosion is unlikely when bulk superheat exceeds 30°F = 17°C (corresponding to surface
superheat of approximately 25°F = 14°C), but very likely at bulk superheat less than 20°F =
11°C (corresponding to surface superheat 15°F = 8°C). A more conservative “rule of thumb”
quoted in the industry is that 40°F = 22°C bulk superheat (corresponding to 35°F = 19°C surface
superheat) is “safe”.

The possible formation of these corrosive liquids in relation to surface superheat is

schematically illustrated on the left side of Figure 1.

Treatment with caustic

The HCl in geothermal steam may be neutralized by injecting caustic (that is, a concentrated
solution of sodium hydroxide, NaOH) into a well head or steam pipe. The NaOH reacts with the
HCl in the steam to produce sodium chloride, which is much less corrosive. This method has a
variety of drawbacks and limitations which are described at length in the paper by Hirtz et al.
(1991). Treatment with caustic is illustrated on the right side of Figure 1.

Geothermal steam usually contains a substantial amount of carbon dioxide. When sodium
hydroxide solution is contacted with steam to neutralize HCl, it also reacts with carbon dioxide
to produce sodium carbonate:

2 NaOH + CO2 → Na2CO3 + H2O

Sodium carbonate is less soluble than sodium hydroxide and the vapor pressure of a saturated
sodium carbonate solution is nearly that of pure water. Table 1 shows the solubility of sodium
carbonate, and the degree of superheat of steam in contact with a saturated solution of sodium
carbonate at various temperatures.

Thus, adding NaOH to steam that is superheated by more than a few degrees can cause rapid
accumulation of solid sodium carbonate mixed with sodium chloride, and crevice corrosion can
occur underneath the solid deposits. In particular, NaOH may react with hydrogen sulfide in the
steam forming sodium bisulfide NaSH which would then accumulate as a highly concentrated
and corrosive solution underneath the solids at the surface of the metal.

In order to avoid precipitation of sodium carbonate, treatment with caustic requires adding
enough water with the caustic to essentially eliminate superheat, and to produce a significant

temperature; therefore, the proper conversion formula applied is ΔT(°F) = 1.8 ΔT(°C). Values
referred to as “superheat” or “degree of superheat” are also in the nature of temperature
differences, and the same conversion formula applies.

4
Weres and Kendrick

amount of liquid at equilibrium with the steam. Boric acid in the steam dissolves in the liquid in
sufficient concentration to make the liquid corrosive, wherefore additional NaOH must be added
to neutralize the boric acid, substantially increasing the amount of caustic consumed. The
solubility of sodium borate produced by reaction with boric acid is limited, and more water must
be added to keep it in solution. Typically, an amount of water equivalent to 3 or 4 weight % of
the steam must be added to desuperheat the steam and provide a sufficient volume of liquid to
keep the sodium chloride and sodium borate in solution. The resulting liquid phase must be
removed from the steam before it enters the turbine. With superheat gone, the turbine operates
with decreased thermodynamic efficiency, and significantly less power is generated.

Finally, caustic corrodes many materials, is extremely hazardous to personnel, and freezes well
above the freezing point of pure water, commonly requiring storage in heated tanks.

The substantial amount of water that must be added to the steam together with NaOH effectively
precludes application of NaOH downhole to protect the casing from corrosion. Sufficient water
must be added to desuperheat the steam with some liquid remaining. If this water is added to the
steam inside the casing, portions of the inside surface of the casing will be wetted, with
temperature nearly as low as the saturation temperature of pure water at the ambient pressure.
Because the steel casing is an excellent conductor of heat, adjacent areas of the casing will also
be cooled to the point that a highly corrosive solution of ammonium chloride condenses. In this
way, injection of NaOH solution can actually induce severe localized corrosion. Fear of such
corrosion has precluded sustained downhole application of NaOH in dry steam wells.

Treatment with potassium carbonate

Hydrogen chloride in superheated steam can safely and more conveniently be neutralized using
potassium carbonate. Potassium hydroxide, potassium bicarbonate or potassium sesquicarbonate
could also be used, as a chemically equivalent amount of any one of these compounds would
produce the same result as potassium carbonate when contacted with superheated steam.
Potassium carbonate is preferred, because it is much safer to handle than potassium hydroxide,
and is less bulky and less expensive than an equivalent amount of potassium bicarbonate or
potassium sesquicarbonate.

Saturated solutions of potassium carbonate can coexist with steam much more strongly
superheated than is the case with sodium carbonate (compare Table 2 to Table 1), and, unlike
sodium carbonate, the maximum degree of superheat increases with increasing temperature.

The degree of superheat of a solution of potassium carbonate at 171°C as a function of weight

fraction water is illustrated in Figure 2 and compared with ferrous chloride.3 Potassium
carbonate supports a much higher degree of superheat (about 45°C) than ferrous chloride (about
24°C). Potassium carbonate remains soluble and be effective throughout the entire range of

3
The data presented in Figures 2, 6 and 7 were generated by heating a concentrated salt
solution of the indicated composition in a pressure vessel and gradually removing water vapor
from the vessel while recording T and P. The apparatus and experimental method employed
were updated versions of those described by Weres and Tsao (1986) and Weres (1995).

5
Weres and Kendrick

superheat at which a solution of ammonium chloride or ferrous chloride might otherwise form,
thus precluding the formation of a corrosive liquid phase.

Potassium carbonate is most conveniently transported, stored and applied as a solution

containing about 40 to 47 weight % of K2CO3. The lowest freezing point of the solution is about
–37°C = –34°F at 40 weight % (these are actual temperatures rather than temperature
differences), and 47 % solution is commonly available in commerce.

The mole ratio of potassium carbonate injected to HCl in the steam should be at least 0.5 to 1,
which is the proportion required to fully neutralize the HCl. An on-line corrosion meter and/or
corrosion coupons can be used to monitor corrosion rates, and to adjust the amount of potassium
carbonate injected to provide adequate control with some margin of safety.

The degree of superheat at which a solid starts to precipitate from a salt solution near to 171°C is
shown for several salts and salt mixtures in Table 3, together with a description of the physical
state of the mixture at superheat slightly above that at which solid starts to precipitate.

Steam that is sufficiently superheated to dry out a solution of ferrous chloride (that is, by more
than about 24°C, the exact value depending on temperature) need not be treated, because no
corrosive liquid phase can form if ferrous chloride dries out. In principle, steam containing HCl
may be corrosive at superheat up to 24°C because a corrosive solution of ferrous chloride may
exist up to that degree of superheat (Figure 2). Thus, ideally, the potassium carbonate should be
injected downhole at a depth where the surface superheat slightly exceeds 24°C. However, as
noted above, the “safe” zone actually extends to somewhat lower values of superheat, whereby
potassium carbonate can actually be injected at shallower depth and lower superheat without
much risk of corrosion.

Preventing deposition of solids

Reaction of HCl in the steam with potassium carbonate produces potassium chloride:

K2CO3 + 2HCl → 2 KCl + CO2 + H2O

Potassium chloride cannot support as much superheat as potassium carbonate or ferrous chloride;
for example, at 171°C, a solution of KCl will dry out if superheat exceeds 16°C (Figure 3). This
means that treating steam superheated by more than 16°C with potassium carbonate solution
risks precipitating solid KCl in the well bore. On the other hand, injecting the potassium
carbonate at a depth where surface superheat is less than 24°C risks formation of a corrosive
ferrous chloride solution. The actual depth chosen to inject the potassium carbonate will
represent a compromise between these concerns.

Several ways are being investigated to counter possible precipitation of KCl, allowing potassium
carbonate to be injected at greater depth thus reliably protecting more of the casing.

1. Together with potassium carbonate, inject a second potassium salt which is able to
support a greater degree of superheat than KCl alone, and will therefore combine with

6
Weres and Kendrick

KCl to form a more concentrated solution which will remain fully liquid at superheat in
the range 16-24°C, or which will at least produce a fluid slurry instead of drying out or
depositing solid incrustations upon metal surfaces. Potassium acetate would serve this
purpose, as would potassium nitrate or potassium methanesulfonate. Figure 3 illustrates
how adding an amount of potassium acetate equivalent to the amount of KCl produced
delays the initial precipitation of KCl to about 21.5°C of superheat, and allows the
resulting slurry of KCl crystals in concentrated salt solution to remain liquid up to about
32°C of superheat. Thus, applying potassium acetate together with potassium carbonate
would allow the well casing to be reliably protected to a greater depth as illustrated on the
right side of Figure 4.

2. If the superheat at the point of injection is large enough to dry out a solution of KCl,
additional water can be added to decrease the degree of superheat of the steam enough to
keep the KCl in solution. For example, adding 0.67% water to steam at 176°C and 5.0
bar – which is initially superheated by about 24°C – would decrease the temperature of
the steam to 168°C and the degree of superheat to 16°C. This option is illustrated on the
right side of Figure 5.

3. Apply a salt inhibitor or crystal modifying agent together with potassium carbonate to
inhibit the precipitation of KCl and/or modify the crystal structure of the precipitate in a
way that keeps the resulting slurry fluid above 16°C of superheat and keeps solids from
adhering to metal surfaces.

Commonly, some amount of boric acid is present in superheated geothermal steam. If the
amount of potassium carbonate applied exceeds the amount needed to neutralize the HCl, the
excess will react with boric acid to form potassium borate, with K/B ratio similar to potassium
tetraborate:

K2CO3(aq) + 4 B(OH)3(g) → K2B4O7(aq) + CO2 (g) + 6 H2O

A solution of potassium tetraborate will support more superheat than KCl (up to 34°C superheat
at T = 171°C) but concentrated solutions of potassium borate and KCl form separate liquid
phases at this temperature with limited mutual solubility (Figure 6). Thus, the maximum
superheat of a solution of KCl increases only slightly when potassium tetraborate is added to it.
However, addition of nirilotriacetamide, which is a potent salt inhibitor, delays precipitation of
KCl, and causes the KCl to precipitate in a manner which produces a more liquid slurry in
combination with potassium tetraborate. This effect is illustrated in Figure 6.

Further protection against corrosion

A volatile amine can be applied together with potassium carbonate. The volatile amine will
provide a second line of defense in case some HCl gets past the potassium carbonate and
threatens corrosion or turbine damage further downstream. A variety of volatile amines which
have been used for corrosion control in power plants; for example: 2-amino-2-methyl-1-propanol
(AMP), ethanolamine (ETA, also called monoethanoleamine or MEA), N,N-diethyl-2-ethanol
amine, morpholine and cyclohexylamine. Preferably, the amine employed should be one

7
Weres and Kendrick

characterized by a combination of strong alkalinity and relatively low volatility; for example,
AMP or ETA. This combination of properties will increase the tendency of the amine to
dissolve in the small amount of concentrated salt solution that will be present together with the
superheated steam.

Field trials

Application of potassium carbonate was tested and compared with application of sodium
hydroxide in four geothermal wells producing superheated steam that contains sufficient HCl to
cause corrosion problems. Two of these wells produce weakly superheated steam that would not
cause a solution of sodium carbonate to dry out. The other two wells produce steam which is
sufficiently superheated to dry out sodium carbonate, but not nearly superheated enough to dry
out potassium carbonate.

The well head temperature, pressure and degree of superheat of the steam produced by these
wells and the treatments applied are summarized in Table 4 and illustrated in Figure 7, which
shows the location of these wells on a plot of superheat vs. temperature.

If the steam is saturated or weakly superheated, the sodium carbonate will stay in solution and,
therefore, sodium hydroxide will have a protective effect similar to that of potassium carbonate.
Thus, Well D was protected by sodium hydroxide as well as Well C was protected by potassium
carbonate with no problems reported by the field personnel in either case.

However, the sodium carbonate produced by injecting sodium hydroxide into the strongly
superheated Well A dried out and failed to prevent corrosion; indeed, injection of sodium
hydroxide appeared to aggravate corrosion. Massive deposits of a white solid with dark streaks
accumulated in and about the control valve, and the stainless steel ball of the control valve was
severely corroded. Analysis showed the solid to be a mixture of sodium chloride and sodium
carbonate with a minor amount of iron also present. Also, thick deposits of an intensely black,
friable material accumulated in the steam header and steam lines. This black material gradually
dissolved in acid with evolution of hydrogen sulfide; iron sulfide is believed to be a major
constituent, as might be expected if a highly concentrated solution of sodium bisulfide was
indeed present at the surface of the metal.

In summary, treatment of Well A with sodium hydroxide was a failure which caused at least as
many problems as it was intended to solve. On the hand, superheated Well B, which was treated
with potassium carbonate, did not appear to have any problems with deposits or corrosion. Thus,
treatment with potassium carbonate was successful in application to superheated Well B.

In a further test, Well D was treated with potassium carbonate injected downhole through a
stainless steel capillary tube together with a small amount of carrier water at a depth of about
2,000 feet = 600 meters below the surface. At this depth, the steam was superheated by about
7°C, precluding application of NaOH. The corrosion rate at the wellhead as determined using
an electrical resistance probe was consistently less than 2 mils = 50 micrometers per year and no
pitting or solid deposits were observed.

8
Weres and Kendrick

Conclusions

Potassium carbonate is a superior substitute for caustic to neutralize HCl in superheated

geothermal steam. In addition to its superior performance as a corrosion inhibitor, potassium
carbonate is safer and therefore easier to handle than caustic because it is much less alkaline and
much less corrosive. Potassium carbonate will not cause the damage to the chemical injection
apparatus that has been observed with sodium hydroxide, and field personnel will have much
less objection to handling it. Brief contact of skin with a concentrated solution of potassium
carbonate will not cause the immediate, severe injury that an equally concentrated solution of
sodium hydroxide or potassium hydroxide would cause.

Acknowledgements

The senior author wishes to thank Ted Derocher for directing his attention to the unmet need for
downhole treatment of corrosive dry steam, and also to thank Bottle Rock Power LLC for the
opportunity to conduct the field trials reported.

9
Weres and Kendrick

References

P. Hirtz, C. Buck and R. Kunzman (1991). Current techniques in acid-chloride corrosion

control and monitoring at The Geysers. Proceedings, Sixteenth Workshop on Geothermal
Reservoir Engineering, Stanford University, Stanford, California, January 23-25, 1991. Report
SGP-TR-134, pp. 83-95.

I. P.-S. Kamps, E. Meyer, B. Rumpf and G. Maurer (2007). Solubility of CO2 in aqueous
solutions of KCl and in aqueous solutions of K2CO3. J. Chem. Eng. Data, 52:3, 817-832, Figure
6, quoting G.G. Aseyev, Electrolytes Equilibria in Solutions and Phase Equilibria, Calculation
of Multicomponent Systems and Experimental Data on the Activities of Water, Vapor Pressures,
and Osmotic Coefficients; Begell House, Inc.: New York, 1998, pp. 254-283.

R.C. Moore, R.E. Mesmer and J.M. Simonson (1997). Solubility of potassium carbonate in
water between 384 and 529 K measured using the synthetic method. J. Chem. Eng. Data, 42:6,
1078-1081, Table 1.

W.F. Waldeck, G. Lynn and A.E. Little (1932). Aqueous solubility of salts at high temperatures.
I. Solubility of sodium carbonate from 50 to 348°. J. Am. Chem. Soc., 54:3, 928-936. Tables I
and II.

O. Weres and L. Tsao (1986). Activity of water mixed with molten salts at 317°C. J. Phys.
Chem., 90, 3014-3018.

O. Weres (1995). Vapor pressure, speciation, and chemical activities in highly concentrated
sodium borate solutions at 277 and 317°C. J. Solution Chem., 24:5, 409-438.

10
Weres and Kendrick

Table 1. Concentration and superheat of saturated solutions of sodium carbonate

Temperature Solubility Superheat

(°C) (w%) (°C)

112.0 30.8 5.1

121.0 29.8 4.4

160.0 26.8 3.6

176.5 25.5 3.3

231.0 19.8 3.0

313.0 6.2 2.1

Data quoted and interpolated from Waldeck et al. (1932).

11
Weres and Kendrick

Table 2. Concentration and superheat of saturated potassium carbonate solutions

Temperature Solubility Superheat

(°C) (w%) (°C)

125. 64.2 25.

140. 66.7 28.

160. 69.2 36.

180. 70.8 46.

200. 71.4 50.

The solubility values in Table 2 interpolated from Moore et al. (1997), Table 1. The superheat
values were calculated using vapor pressure extrapolated from the vapor pressure values quoted
by Kamps, et. al. (2007), Figure 6.

12
Weres and Kendrick

Table 3. Summary of Superheat Data

Salt First solid Condition when

appears at ΔTsup > ΔTsup,Max
ΔTsup,Max (°C)

Na2CO3 ~ 3.5 Dry salt.

NaCl ~ 13 Dry salt.

KCl 16 Dry salt.

K2B4O7 + 2 KCl 16.4 Incrustations form in the absence of a

salt inhibitor; fluid slurry persists to
significantly higher superheat in the
presence of a salt inhibitor.

2 KCl + 1 K Acetate 18 Fluid slurry.

1 KCl + 1 K Acetate 22 Fluid slurry.

KCl + ½ K2B4O7 23.5 Fluid 3-phase emulsion.

+ K Acetate

FeCl2 24 Dry salt.

K2B4O7 34 Glass.

K2CO3 45 Dry salt.

K Acetate > 63 Dry salt.

13
Weres and Kendrick

Table 4. Geothermal Wells Treated

Well P(bar) T(°C) ΔTsup(°C) Treatment Result

A 4.17 162.2 17.1 NaOH Corrosion and

accumulation of solids

B 3.90 152.8 10.1 K2CO3 No problem.

C 3.82 144.4 2.4 K2CO3 No problem.

D 3.97 145.0 1.7 NaOH No problem.

14
Weres and Kendrick

Figure Captions

Figure 1. Schematic illustration of corrosive conditions in a dry steam well and treatment with
caustic. “Surface Superheat,” calculated using the temperature of the casing, is about 3°C = 5°F less
than “Bulk Superheat,” based on steam temperature. The Surface Superheat governs whether a
liquid phase will form on the inner surface of the casing, and therefore determines the rate of
corrosion.

Figure 2. Degree of superheat of highly concentrated solutions of ferrous chloride and potassium
carbonate as a function of weight fraction water. In each case, the horizontal dashed line indicates
the degree of superheat at which solid salt begins to precipitate; when only one salt is present, the
solution will completely dry out at this degree of superheat. The vertical dashed lines represent the
limit of reliable data. To the right of the vertical line the two temperature probes in the pressure
vessel gave consistent readings, indicating that the material in the pressure vessel was still fluid. To
the left of the vertical line precipitated solids began to adhere to solid surfaces and the readings from
the two temperature probes diverged.

Figure 3. Degree of superheat vs. weight fraction water of a highly concentrated solution of
potassium chloride (solid circles) and a solution containing equimolar amounts of potassium chloride
and potassium acetate (open circles). With KCl alone, the charge dried out at constant superheat of
16°C. In the case of the salt mixture, KCl started to precipitate at about 21.5°C, and the solution
gradually became more superheated as more water was removed and more KCl precipitated from
solution, whereby the proportion of potassium acetate in solution gradually increased. Finally, at
32°C superheat the KCl crystals began to adhere to solid surfaces and readings from the two
temperature probes diverged.

Figure 4. The right side of the drawing illustrates treatment of the well using potassium carbonate.
Simultaneous application of potassium acetate or a salt inhibitor would allow the treatment to be
injected at greater depth, conferring additional protection against corrosion.

Figure 5. The right side of the drawing illustrates treatment with potassium carbonate with sufficient
water provided to decrease superheat to 16°C at the point injection. Complete protection against
corrosion is provided with no risk of solids accumulating in the well bore.

Figure 6. Superheat vs. weight fraction of water in solutions containing equivalent amounts of KCl
and potassium tetraborate, with the salt inhibitor nirilotriacetamide added at 0.2g per equivalent of K
in the solution (open circles), and without the salt inhibitor (solid circles). Without the salt inhibitor
(solid circles) KCl began to precipitate and adhere to the temperature probes at about 16.5°C
superheat. With the salt inhibitor added (open circles), initial precipitation of KCl was delayed until
about 19°C superheat, and the mixture remained fluid until enough water had been removed to
increase superheat to about 24°C. The flat portion of the plot is believed to represent an interval
where two liquid phases coexisted while solid KCl gradually precipitated, one liquid a solution
containing mostly KCl, the other a solution containing mostly potassium tetraborate.

Figure 7. Two wells treated at the surface with potassium carbonate compared with two wells
treated with caustic.

15
 Fig. 1. Treatment with caustic.
Surface Superheat
o
Untreated ( C) Treated with caustic

Corrosion certain.

5 Wet steam.
Highly corrosive
NH4Cl solution
condenses. Calcite scale
deposited if injected
10 water contains
calcium.

Na Borate solids
15 deposited if
insufficient water
Increasing risk of
injected.
corrosion.

20 Severe corrosion
possible near injection
FeCl2 soluble. point.
0
Corrosive FeCl2 solution 25
dries out above 24oC Inject caustic and water
superheat. to desuperheat and
dilute, equal to 3-4% of
No liquid, no corrosion. steam flow.
30 30

Severe corrosion.
35 35
Corrosion.
Possible corrosion.
No corrosion.
40 40
 60

50
Superheat (oC)

40
K2CO3 @ 172oC

30

20

10 FeCl2 @ 172oC

0
0.2 0.3 0.4 0.5 0.6 0.7 0.8

Weight Fraction Water

Figure 2
 40

30
KCl + KAcetate 1:1 @ 174oC
Superheat (oC)

20

KCl @ 171oC

10

0
0.2 0.3 0.4 0.5 0.6 0.7 0.8

Weight Fraction Water

Figure 3
Fig. 4. Treatment with potassium carbonate, no water.
Surface Superheat
(oC)
Untreated Treated with K2CO3

Steam remains superheated.

Corrosion certain. 5
Highly corrosive
NH4Cl solution
condenses. No solids, no corrosion.
10
No additional water needed.

KCl completely dissolved.

15
Increasing risk of
corrosion.
All of HCl and some boric
Inject K2CO3 here.

20 acid neutralized, forming a

slurry of KCl crystals in a
concentrated salt solution.
FeCl2 soluble.

Corrosive FeCl2 solution 25 No corrosion.

dries out above 24oC

superheat. K2CO3 and
potassium borate
30 soluble.
Severe corrosion.
Corrosion.
Possible corrosion. 35
No corrosion. Injecting potassium acetate or salt
inhibitor together with K2CO3
allows treatment at greater depth,
40 conferring more protection.
 Fig. 5. Treatment with K2CO3 and water.
Surface Superheat
(oC)

Untreated Treated with K2CO3

and water
Highly corrosive NH4Cl
solution condenses.
No corrosion or deposition of
solids even if superheat goes
to zero at some point
downstream.

15 5
Increasing risk of
corrosion.
Steam remains superheated.

20 10

FeCl2 soluble.
Inject K2CO3 and limited
25 25 water to decrease
Corrosive FeCl2 solution
superheat.
dries out above 24oC
superheat.
Injected water equal to 0.7%
30 30 of steam flow.

Severe corrosion.
Corrosion. 35 35
Possible corrosion.
No corrosion.

40 40
 30

25 K2B4O7 + 2 KCl + nitrilotriacetamide

@ 170oC
Superheat (C)

20

15

10
K2B4O7 + 2 KCl @ 171oC

0
0.2 0.3 0.4 0.5 0.6 0.7 0.8

Weight Fraction Water

Figure 6
 50

Treatment applied:
K 2
CO 3
r
K2CO3 i mit fo
40 ilit yl
NaOH olub
S
Superheat (oC)

30

A
20
Deposits and severe corrosion with NaOH
B
10 No problems with K2CO3

Solubility limit for Na2CO3 C

No problems D
0
120 130 140 150 160 170 180

Temperature (oC)

Figure 7
View publication stats