Corrosion of Materials Used in Geothermal

Power Production
Review of Materials and Treatment Technologies
3002007966

14017525
14017525
Corrosion of Materials Used in Geothermal
Power Production
Review of Materials and Treatment Technologies

3002007966
Technical Update, April 2016

EPRI Project Manager

A. Coleman

ELECTRIC POWER RESEARCH INSTITUTE

3420 Hillview Avenue, Palo Alto, California 94304-1338  PO Box 10412, Palo Alto, California 94303-0813  USA
14017525800.313.3774  650.855.2121  [email protected]  www.epri.com
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF
WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI).
NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY
PERSON ACTING ON BEHALF OF ANY OF THEM:
(A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH
RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM
DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED
RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS
SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR
(B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING
ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF
THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS
DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN
THIS DOCUMENT.
REFERENCE HEREIN TO ANY SPECIFIC COMMERCIAL PRODUCT, PROCESS, OR SERVICE BY ITS TRADE
NAME, TRADEMARK, MANUFACTURER, OR OTHERWISE, DOES NOT NECESSARILY CONSTITUTE OR
IMPLY ITS ENDORSEMENT, RECOMMENDATION, OR FAVORING BY EPRI.
THE FOLLOWING ORGANIZATIONS, UNDER CONTRACT TO EPRI, PREPARED THIS REPORT:
EGS, Inc.
Thermochem, Inc.
Lehigh University

This is an EPRI Technical Update report. A Technical Update report is intended as an informal report of
continuing research, a meeting, or a topical study. It is not a final EPRI technical report.

NOTE
For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or
e-mail [email protected].

Electric Power Research Institute, EPRI, and TOGETHER…SHAPING THE FUTURE OF ELECTRICITY
are registered service marks of the Electric Power Research Institute, Inc.
Copyright © 2016 Electric Power Research Institute, Inc. All rights reserved.

14017525
 ACKNOWLEDGMENTS
The following organizations, under contract to the Electric Power Research Institute (EPRI),
prepared this report:
EGS, Inc.
Waterfall Towers, Suite A112
2455 Bennett Valley Road
Santa Rosa, CA 95404
Principal Investigator
G. Suemnicht
Thermochem, Inc.
3414 Regional Parkway, Suite A
Santa Rosa, CA 95403
Principal Investigator
P. Hirtz
Lehigh University
P.C. Rossin College of Engineering and Applied Science
Energy Systems Engineering Institute
STEPS Building
Bethlehem, PA 18015
Principal Investigator
E. Cichonski

This report describes research sponsored by EPRI.

This publication is a corporate document that should be cited in the literature in the following
manner:
Corrosion of Materials Used in Geothermal Power Generation: Review of Materials and
Treatment Technologies. EPRI, Palo Alto, CA: 2016. 3002007966.

14017525 iii
14017525
 PRODUCT DESCRIPTION
Corrosion can quickly render newly installed geothermal well casings useless, causing damage in
the multimillion-dollar range. Mitigating corrosion has become more important as conventional
geothermal resource depletion progresses and reservoirs begin to produce more superheated
steam. Highly corrosive conditions related to near-magmatic temperatures may limit the future
development of deeper, hotter resources, or high-temperature enhanced or engineered geothermal
systems (EGSs). To support the development of new technologies for managing corrosion in
geothermal systems, this report provides critical information about environmental chemical
conditions that corrode equipment in high-chloride (Cl), low-pH, and high-temperature
geothermal environments. It reviews laboratory and field testing of corrosion resistant materials
and chemical treatments that mitigate corrosion.
Background
Geothermal systems are an established renewable energy resource with a long history of
adaptable, reliable baseload generation. Long-term production withdrawals, changing
temperature and pressure conditions, and differing fluid sources can alter reservoir chemistry
over time to yield more corrosive fluids. Expanding exploration to include intensely active
volcanic systems, deeper resources, and deep EGS stimulation experiments is purposely
targeting more challenging reservoir conditions at greater depths and higher temperatures.
Operations such as deep drilling in Kakkonda, Japan; magma drilling in Iceland; step-out drilling
in Larderello, Italy; and EGS stimulation experiments at The Geysers in northern California in
the United States have already experienced reservoir characteristics that far exceed the
temperatures and geochemical conditions common to conventional, convecting hydrothermal
systems.
Because geothermal fluid chemistry varies considerably, even within individual fields, a wide
array of heavy metals and salts may contribute to metallic decay (corrosion). Reservoir
conditions contributing to corrosion depend on depth: conditions between the surface and a depth
of 5 kilometers (km) vary greatly, while these conditions may contrast sharply with those at a
depth of 10 km. Conditions in a production well may be single-phase (water- or steam-
dominated), or two-phase (transitioning from water- to steam-dominated). In a two-phase
system, the water flashes to steam (at the flash point) as it rises to the surface. Conditions in a
production well can be most challenging near the surface as pressures initiate phase changes or
minor amounts of condensation occur near the well head creating a chemical environment
particularly hospitable to corrosion.
Carbon steel has commonly been used as a corrosion-resistant material in geothermal
applications because of its low cost and availability. But in lower-pH geothermal fluids, alternate
materials or coatings are required to prevent material degradation and failures. During the 1980s
and early 1990s, various studies evaluated the causes of material failures in geothermal
environments to improve geothermal power production. Laboratory and field testing revealed a
pattern of corrosion resistance among high-alloy materials, and rapid degradation of low-alloy
materials or plain carbon steel. In field testing, commercially available carbon steel, low-alloy or
chromium-molybdenum (Cr-Mo) steels, martensitic and ferritic stainless steels, high-nickel (Ni)
alloys, and titanium (Ti) were evaluated in geothermal production settings at high temperatures
(> 260 °C).

14017525 v
Only field testing under actual production conditions can validate the reliability of a given
material in a geothermal environment. But, no two geothermal wells have exactly the same
chemistry. In response to this fact, Wilson and Lichti (1982) proposed standardized materials
testing. Since then, corrosion resistant materials have faced increasing challenges as geothermal
wells are drilled deeper to obtain maximal heat output and power generation. Deeper wells also
tend to have lower pH and higher salinity.
Objectives
• To review chemical data from past and present research in a variety of geothermal fields with
temperatures approaching 400 °C to assess comparative conditions in high-temperature
environments.
• To review technologies, downhole chemical treatments, and metal alloys used for casings in
these reservoir systems from 1990 to the present.
Approach
A literature review identified the causes, types of damage, and mitigation practices associated
with corrosion of metals in geothermal applications.
Results
This review updates open source information on the development and testing of traditional and
innovative geochemical mitigation and monitoring methods. Managing the geochemistry of
geothermal fluids plays an important role in mitigating materials degradation in differing
geothermal environments. In sections of a field that produce the most corrosive superheated
steam, caustic (sodium hydroxide or NaOH) and condensate are injected at the wellhead to
control corrosion. This technique has also been used successfully to minimize corrosive effects
downhole. Available logging and sampling tools have been used in wells at temperatures
reaching about 330 °C, but tools suitable for use at 400 °C are still needed to monitor downhole
conditions and identify zones of corrosive fluid production in the hottest wells.
Applications, Value, and Use
The results of this research can be used to assess the unique geochemical conditions of corrosive
geothermal reservoirs under construction, especially in Japan and Indonesia, and under
development in Latin America. It offers a basis for comparing corrosive conditions at The
Geysers with superheated high-temperature conditions in other geothermal reservoirs. This work
also points out the need for more analysis of the environmental and chemical attributes
associated with the transition from a normal hydrothermal reservoir to a superheated steam
reservoir. Additional studies may be needed to model high-temperature zones at EGS sites, with
expanded evaluation of the mechanisms of chloride generation operating at high temperatures.
Keywords
Engineered geothermal systems (EGS)
Hydrothermal
Chlorides
Corrosion
Nickel-alloys

14017525 vi
 ABSTRACT
Corrosion can quickly render newly installed geothermal well casings useless, causing damage in
the multimillion-dollar range. Mitigating corrosion has become more important as conventional
geothermal resource depletion progresses and reservoirs begin to produce more superheated
steam. Highly corrosive conditions related to near-magmatic temperatures may limit the future
development of deeper, hotter resources, or high-temperature enhanced or engineered geothermal
systems (EGSs). To support the development of new technologies for managing corrosion in
geothermal systems, this report provides critical information about environmental chemical
conditions that corrode equipment in high-chloride (Cl), low-pH, and high-temperature
geothermal environments. It reviews laboratory and field testing of corrosion resistant materials
and chemical treatments that mitigate corrosion.
A review of published studies identified and compared chemical data from past and present
research in a variety of geothermal fields, including The Geysers in northern California, whose
temperatures approached 400 °C. The review also identified the causes, types of damage, and
mitigation practices associated with metal corrosion in these reservoir systems from 1990 to the
present.
The results show that managing the geochemistry of geothermal fluids plays an important role in
mitigating materials degradation in differing geothermal environments. In sections of a field that
produce the most corrosive superheated steam, caustic (sodium hydroxide or NaOH) and
condensate are injected at the wellhead to control corrosion. This technique has also been used
successfully to minimize corrosive effects downhole. The choice of corrosion resistant materials
involves tradeoffs between effectiveness, as demonstrated in laboratory and field tests, and cost.
Chemical injection combined with materials of lower cost and decreased corrosion resistance is
currently the best option. Available logging and sampling tools have been used in wells at
temperatures reaching about 330 °C, but tools suitable for use at 400 °C are still needed to
monitor downhole conditions and identify zones of corrosive fluid production in the hottest
wells. The results of this research can be used to assess the unique geochemical conditions of
geothermal reservoirs where sources of actual or potential corrosion require identification and
mitigation.

14017525 vii
14017525
 EXECUTIVE SUMMARY
Geothermal systems are an established renewable energy resource with a long history of
adaptable, reliable baseload generation. Dealing with corrosive fluid geochemistry can be one of
the most difficult challenges in developing and producing geothermal systems. Corrosion and
scaling issues have been addressed in a limited number of hydrothermal settings; however, the
majority of conventional hydrothermal systems initially produce relatively benign geothermal
fluids. Corrosion problems have become more prevalent as more geothermal systems are
discovered, developed, and produced for long periods of time. Mitigating corrosion in existing
geothermal fields can become an urgent need as resource production progresses and reservoirs
produce increasing volumes of superheated steam.
An enhanced or engineered geothermal system (EGS) attempts to access and utilize geothermal
heat by drilling into hot dry rock at depth. Development and implementation of successful EGS
technologies has the potential to expand the geothermal industry by siting geothermal power
plants in more locations and expanding the world’s potential geothermal resources. The creation
of an EGS involves hydraulically fracturing the deep rock to extract its geothermal heat for
power generation. Subsurface injection wells are drilled to allow the injected fluid to heat up and
transfer the maximum heat from the hot rock before being extracted by a production well.
Deeper drilling, exploration in near-magmatic volcanic environments, and projected deep EGS
resource development have shown that the presence of volatile hydrogen chloride (HCl) and
other corrosive conditions can occur in geothermal areas all over the world. Research efforts in
the late 1990s were motivated in part by the increasing importance of mitigating corrosive
geochemical environments and the recognition of the energy potential of deep, near-magmatic
systems. The feasibility of deep, high-temperature EGS projects—or drilling into near-magmatic,
highly corrosive environments as contemplated in the Icelandic Deep Drilling Project—depends
directly on addressing potential corrosion problems.
Research is ongoing to find cost-effective alloys for geothermal well casings. Although many
metals adequately survive corrosive conditions, the cost of some materials is so high that they
significantly impact geothermal development costs and project economics. Austenitic stainless
steels are not as corrosion resistant as titanium (Ti) and nickel (Ni) alloys such as Hastelloy
C-276 and Inconel 625, but they are more affordable. Titanium and high-nickel alloys are
currently too expensive for most deep-seated geothermal targets, and the price volatility of these
specialty metals further complicates project cost estimates. Although composite coatings, spray
applications, and electrochemically deposited or explosion cladded materials have been
successful in laboratory testing, there are few published studies on their use in large-scale
projects at temperatures typical of high enthalpy geothermal systems, and some coatings may
require thicker applications to resist corrosion in low-pH environments.
In many geothermal fields, chemical injection to adjust the fluid chemistry is the most cost-
effective solution to corrosion. Sodium hydroxide (NaOH) injections increase pH and can
improve operating conditions to reduce the immediate need for corrosion resistant alloys.
Chemical injection combined with a lower cost and decreased corrosion resistance is currently
the best option, until nickel and titanium costs moderate enough to justify manufacturing clad or
solid piping. Material and treatment technologies should be chosen based on laboratory and/or
field testing at conditions similar to those at a specific field, or even well site, if practical.

14017525 ix
14017525
 LIST OF TERMS
Al aluminum
(aq) aqueous phase
bar unit of pressure
bara unit of absolute pressure
Ca calcium
Ca(OH)2 calcium hydroxide (hydrated lime)
Ca2+ calcium ion
CaCl2 calcium chloride
CaCO3 calcium carbonate (limestone or calcite)
CaO calcium oxide (lime)
CaSO4 calcium sulfate
CCPA Central California Power Agency
Cl chloride
Cl- chloride ion
(Mg,Fe)3(Si,Al)4O10(OH)2·(Mg,Fe)3(OH)6 chlorite
CMFs corrosion mitigation facilities
C-N carbon-nitrogen (bond)
CO2 carbon dioxide
CO32+ carbonate
Cr chromium
CRA corrosion resistant alloy
Cr-Mo chromium-molybdenum
Cu copper
DHS downhole sampler
DOE U.S. Department of Energy
DSS dry steam scrubbing
EGS enhanced geothermal system

14017525 xi
F- fluoride
Fe iron
FeCl2 ferrous chloride
FRP fiberglass reinforced plastic
(g) gas phase
g/mol grams per mole
H20 water
H2S hydrogen sulfide
HCl hydrogen chloride or hydrochloric acid
HTZ high-temperature zone
IDDP Icelandic Deep Drilling Project
K potassium
K2CO3 potassium carbonate
KCl potassium chloride
kg/m s kilograms per meter per second
kg/m3 kilograms per square meter
kg/s kilograms per second
KHCO3 potassium bicarbonate
kJ/kg kilojoules per kilogram
kJ/mol kilojoules per mole
km kilometer
KOH potassium hydroxide
kPa kilopascal
lph liters per hour
2MgSO4·Mg(OH)2 magnesium hydroxide sulfate
m meter
m/s meters per second
Mg magnesium

14017525 xii
Mg(OH)2 magnesium hydroxide
mg/kg milligrams per kilogram
mil one-thousandth of an inch
mm millimeter
MMT montmorillonite
Mo molybdenum
MPa megapascal
Na sodium
Na+ sodium ion
Na2B4O7·10H2O sodium borate
Na2CO3 sodium carbonate (soda ash)
NaCl sodium chloride
n-alkyl normal-alkyl
NaOH sodium hydroxide
NCG noncondensible gas
NH3 ammonia
NH4+ ammonium ion
NH4Cl ammonium chloride
Ni nickel
n-octadecyl normal-octadecyl
O2 oxygen
ºC degree Celsius
OH- hydroxide
(org) organic
ppb parts per billion
ppm parts per million
ppmw parts per million by weight
PPS polyphenylsulfide

14017525 xiii
(R)3(R)NH2 tertiary-alkyl primary amine
(R)3(R)NH3+ Cl- tertiary-alkyl primary amine chloride
(s) solid phase
SiO2 silicon dioxide or silica
SiO2·nH2O silicic acid
SO2 sulfur dioxide
SO3 sulfur trioxide
t-alkyl tertiary-alkyl
Thermochem Thermochem, Inc.
Ti titanium
U.S. United States
Zn zinc
Zn3(PO4)2 zinc phosphate
Zr zirconium

14017525 xiv
 CONTENTS
1 INTRODUCTION ..................................................................................................................1-1
2 ACIDIC GEOTHERMAL FLUIDS..........................................................................................2-1
Causes and Occurrence.....................................................................................................2-1
Occurrence of Hydrogen Chloride in Steam at The Geysers ..............................................2-3
3 CORROSION RESISTANT MATERIALS .............................................................................3-1
Failure Mechanisms ...........................................................................................................3-1
Materials ............................................................................................................................3-3
Current Research for Deep-Seated Geothermal Wells .......................................................3-4
Cost of Materials ................................................................................................................3-6
4 CORROSION MITIGATION TECHNIQUES ..........................................................................4-1
Early Volatile Chloride Corrosion Mitigation Techniques ....................................................4-1
Downhole Volatile Chloride Scrubbing and Neutralization ............................................4-3
New Volatile Chloride Corrosion Mitigation Technologies ...................................................4-4
Hydrogen Chloride Mitigation Strategies—Detailed Process Examples ..............................4-5
Conventional Water and Caustic Scrubbing .................................................................4-5
Dry Steam Scrubbing ...................................................................................................4-9
Hybrid Steam Scrubbing .............................................................................................4-14
Steam Polishing .........................................................................................................4-20
Comparison of Treatment Options for Hydrogen Chloride Corrosion Control ..............4-22
5 SUMMARY ...........................................................................................................................5-1
6 REFERENCES .....................................................................................................................6-1

14017525 xv
14017525
 LIST OF FIGURES
Figure 2-1 Southwest-to-northeast cross section through the areas of Units 5 and 6 at The
Geysers. Red ticks represent steam entries to wells. Produced chloride concentrations
are shown next to bottom-hole locations of wells. Dashed line represents an upwelling
of high-chloride steam originating in the HTZ. ....................................................................2-4
Figure 3-1 Schematic showing gaps in geothermal materials research [25] .............................3-5
Figure 3-2 Cost comparison of materials for tank construction as performed by Ashland
Engineering in their 2015 report [52] ..................................................................................3-8
Figure 4-1 Water and caustic process flow diagram—8 bara ...................................................4-6
Figure 4-2 Water and caustic process flow diagram—8 bara, pre-heated scrub water .............4-7
Figure 4-3 Water and caustic process flow diagram—20 bara .................................................4-8
Figure 4-4 Calcium carbonate dry steam scrubbing process flow diagram—20 bara .............4-12
Figure 4-5 Potassium carbonate solution superheat vs. temperature .....................................4-15
Figure 4-6 Potassium carbonate solution superheat vs. pressure ..........................................4-16
Figure 4-7 Hybrid potassium carbonate process flow diagram—20 bara ................................4-17
Figure 4-8 Calcium chloride solution superheat vs. temperature ............................................4-18
Figure 4-9 Calcium chloride solution superheat vs. pressure .................................................4-19
Figure 4-10 Hybrid calcium chloride process flow diagram—20 bara .....................................4-20
Figure 4-11 Silica partitioning between steam and water as a function of
temperature [68] ...............................................................................................................4-22

14017525 xvii
14017525
 LIST OF TABLES
Table 3-1 Cost comparison of materials for fabrication of pressure vessels and heat
exchangers ........................................................................................................................3-7
Table 3-2 Price comparison of stainless steel plate and carbon steel plate clad
with C-276 ..........................................................................................................................3-8
Table 4-1 Calcium carbonate dry steam scrubbing bed design and operating
parameters for example well ............................................................................................4-11
Table 4-2 Steam scrubbing process comparison for example well .........................................4-23

14017525 xix
14017525
1
INTRODUCTION
Geothermal electricity production provides a renewable source of energy and baseload power
that other renewables, including wind and solar power, cannot provide. Conventional geothermal
power is produced in areas with naturally occurring hydrothermal reservoirs or in stimulated
dried-out geothermal reservoirs. Conventional geothermal systems are being explored in deeper
and hotter environments where handling corrosive fluids is challenging. Finding reasonably
priced materials to minimize corrosion in fluid production and power systems is a significant
constraint faced by project developers as exploration extends to these more challenging
environments. At the same time, producing conditions in maturing developed geothermal fields
are changing as superheating becomes the dominant reservoir condition, resulting in corrosion
problems that affect project economics.
An enhanced or engineered geothermal system (EGS) attempts to access and utilize geothermal
heat by drilling into hot dry rock at depth. Development and implementation of successful EGS
technologies have the potential to expand the geothermal industry by enabling geothermal power
plants to be built in more locations, thus enlarging the world’s potential geothermal resources.
The creation of an EGS involves drilling deep injection wells and hydraulically fracturing the
rock in order to circulate fluids needed to extract the geothermal heat for power generation.
Subsurface injection wells are drilled to allow the injected fluid to absorb the maximum heat
from the hot rock before the fluid is extracted by a production well. Wells in recent EGS
experiments are typically between 2,000 and 3,500 meters (m) in depth, but scientists and
engineers are looking to drill deeper to reach higher temperatures and produce more power [1].
EGS technology could be the future of geothermal energy; according to the U.S. Department of
Energy (DOE), if enough plants are successfully established in the United States, geothermal
energy could support 10% of U.S. energy consumption [2].
Corrosion of production wells, gathering lines, and turbomachinery by geothermal fluids (steam
and water) has been a problem with some geothermal reservoirs since the industry’s inception
and is a significant impediment for project developers to overcome. Carbon steel has been
commonly used for production wells and gathering lines because of its low cost and availability,
but it corrodes easily when exposed to acidic geothermal fluids. The industry has turned to
alternate, more expensive corrosion resistant materials—or coatings and treatment technologies
in some cases—to mitigate rapid material degradation and failure.
Geothermal fluid is highly variable in composition, as shown by the wide array of heavy metals,
salts, and dissolved gases that contribute to its corrosive nature and cause significant fatigue and
failure of metal components. A wide variation in fluid chemistries has been observed among
different geothermal fields, among wells in a single field, and even within individual wells where
the chemical composition of steam or water inflows can vary dramatically. Fluid compositions
have been known to change over the life span of a geothermal field, while injection of water for
mass replacement introduces fluids with different compositions, resulting in complex fluid
interactions and dynamic geochemistries. Reservoir conditions can be single-phase (water- or
steam-dominated) or two-phase. In the case of two-phase, the flash point—where the water

14017525 1-1
“flashes” to steam as it rises to the surface in the well and pressure is reduced—can be a
particularly corrosive environment for steel well casings. Furthermore, as time has passed, the
geothermal industry has drilled deeper to obtain maximum heat output and power generation,
revealing a new set of challenges. In addition to higher temperatures, deeper wells tend to have a
lower pH and higher salinity.
The combination of highly variable fluid compositions, temperatures, pH, and dynamic reservoir
conditions makes it difficult to predict material performance a priori. Therefore, the geothermal
industry has developed field-, and in some cases, well-specific solutions to corrosion, taking
complex site conditions into consideration when selecting materials and treatment strategies.
This report summarizes a review of the open literature on corrosion resistant materials and
treatment technologies deployed by the geothermal industry. Section 2 identifies geothermal
reservoir conditions leading to metal corrosion at The Geysers field in northern California in
United States, and other locations. Section 3 reviews corrosion mechanisms, identifies corrosion
resistant materials including solid and cladded piping, and summarizes costs. Section 4 identifies
effective chemical treatment technologies that have been used to mitigate corrosion, including
the treatment of highly corrosive geothermal fluids at The Geysers that result from the
co-production of acid gases like hydrogen chloride (HCl). Section 5 summarizes the findings and
makes recommendations for future research.

14017525 1-2
2
ACIDIC GEOTHERMAL FLUIDS
Causes and Occurrence
There are several major causes of geothermal fluid acidity, but hydrogen chloride is the most
prevalent and severely corrosive chemical species found in geothermal fluids associated with
high-temperature geothermal systems. Volcanic gases containing hydrogen chloride, sulfur
dioxide (SO2) and sulfur trioxide (SO3) may infiltrate a benign geothermal fluid system, causing
the liquid to become highly acidic. For this and other reasons, development of geothermal
reservoirs with strong volcanic influence is typically avoided. Immature geothermal systems may
contain residual acidic fluids of volcanic origin that have yet to be completely neutralized by the
rock. These systems have been avoided in most commercial developments.
Acidic brine may be caused by the influx of shallow acid-sulfate fluids into a deeper, benign
geothermal reservoir. The acid-sulfate fluid is often formed by the reaction of oxygen in
meteoric water with hydrogen sulfide (H2S), or by water contacting acidic minerals. The
associated corrosion of geothermal wells often manifests as external casing corrosion when the
exterior portion of a well is exposed to the shallow acidic fluids. These shallow fluids may or
may not directly affect the produced fluids, depending on the well completion design and casing
integrity.
Acidic brine is also created by heating seawater to temperatures of 300 degrees Celsius (°C) or
more. At these temperatures, magnesium in seawater precipitates as magnesium hydroxide
(Mg(OH)2), chlorite ((Mg,Fe)3(Si,Al)4O10(OH)2·(Mg,Fe)3(OH)6), and/or magnesium hydroxide
sulfate (2MgSO4·Mg(OH)2), causing the pH to drop. This process was believed to be responsible
for low-pH brine produced in early wells at Puna, Hawaii and at Reykjanes, Iceland.
In most cases, acidic brine can be treated by downhole injection of sodium hydroxide (NaOH), in
the same manner (described later in this report) as hydrogen chloride is treated downhole.
Downhole injection has been successful at several fields in the Philippines, but the only field
where acidic brine is currently being treated downhole is at Miravalles in Costa Rica [3].
Volatile chloride—in the form of hydrogen chloride and/or ammonium chloride (NH4Cl)—is
commonly produced from vapor-dominated geothermal fields such as The Geysers in the United
States and Larderello in Italy, and is typically associated only with superheated steam wells.
Severe corrosion problems related to volatile acid-chloride species in geothermal power facilities
have occurred at these fields. At Larderello, 1–10 ppm (parts per million by weight, or mg/kg)
chloride in steam has etched turbine components—mostly at the base of stationary and moving
blades in the wet stages of the machines—while 10–100 ppm chloride in contact with condensate
has caused severe corrosion of carbon steel steam lines. Chloride concentrations measured in
dry, superheated steam produced by wells at The Geysers range from less than 0.010 ppm to 200
ppm. Corrosion damage due to volatile chloride at The Geysers includes extensive well casing
corrosion, casing head perforation, well surface piping failures, and gathering system piping
perforations. The corrosion products exfoliated from wellbores and gathering systems impact

14017525 2-1
power plant equipment by plugging steam strainers and eroding turbine blades; in some cases
these impacts are more severe than direct corrosion of the components.
Geothermal steam containing volatile chloride has been reported in other fields throughout the
world, including Tatun, Taiwan; Krafla, Iceland; Saint Lucia, Windward Islands; Coso, United
States; and the Salton Sea, United States. Geothermal systems producing volatile chloride vary
greatly in reservoir characteristics; they may be vapor-dominated, liquid-dominated, or volcanic-
related fields. However, all of these fields have produced measurable levels of chloride (Cl) in
dry steam from certain wells without associated quantities of alkali metals such as sodium (Na)
and potassium (K), or alkaline earth metals such as calcium (Ca) and magnesium (Mg). In
Krafla, Iceland, hydrogen chloride in steam was first suspected 35 years ago in well KG-4. In the
1980s, well KG-12 was reported to produce about 100 ppm chloride in dry superheated steam,
while results from 2008 show chloride levels above 800 ppm in steam condensates from well
KG-36. The presence of hydrogen chloride fluids appears to be widespread throughout the Krafla
field [4–7]. The IDDP-1 well drilled in Krafla—as part of the Icelandic Deep Drilling Project
(IDDP)—with the target of reaching near-magmatic conditions and supercritical fluids, produced
superheated steam containing 100 ppm hydrogen chloride from a 400 °C subcritical reservoir.
Continued encounters with hydrogen chloride are expected as deeper and hotter geothermal wells
are drilled around the world.
It be conclusively stated for only a few geothermal systems that hydrogen chloride is the volatile
chloride species, although the presence of hydrogen chloride has been well documented for high-
temperature volcanic gases. In the condensate of chloride-bearing steam at The Geysers, there
are usually more ammonium ions (NH4+) than chloride, making it unclear whether the chloride
was initially hydrogen chloride that reacted with ammonia (NH3) in solution, or was transported
as ammonium chloride in the vapor phase.
Several existing mechanisms can produce the observed range of volatile chloride in geothermal
steam. Hydrogen chloride does not partition significantly from hot (300–350 ºC), high-salinity
(~2 molal), near-neutral sodium chloride (NaCl) brines [8, 9]. For most geothermal reservoirs
that are not in direct contact with volcanic fluids, hydrogen chloride is most likely generated
from superheated steam by reaction with chloride salts, quartz, and silicates upon brine dry out
[10]. An example of this occurred at Coso Geothermal Station in central California in the United
States, where a neutral brine-producing well that dried out and became superheated due to low
permeability and silica scaling began generating hydrogen chloride at 10 ppm. The reservoir
temperature was 270 ºC, brine salinity was 10,000 ppm and pH was 6 [11]. Simonson [12] has
shown conclusively that only about 8 ppm hydrogen chloride can be generated from pH 4 brine
containing 70,000 ppm chloride ion (Cl-) at 300 to 350 ºC, and less than 1 ppm hydrogen
chloride can be generated under the same conditions at pH 5. These values are an order of
magnitude lower than the values proposed by Truesdell and Haizlip [13] based on a deep boiling
brine theory for The Geysers.
Since hydrogen chloride is ubiquitously found in superheated volcanic fumarole discharges, it is
likely to be of volcanic origin for geothermal wells in contact with volcanic systems and
outgassing magma. Direct partitioning of hydrogen chloride from acidic, hypersaline brine is
known to occur at the Salton Sea, where brine (4 molal sodium chloride) is deliberately acidified
to pH 3–4 for silica-scale control, generating up to 50 ppm hydrogen chloride in steam at 220 °C.

14017525 2-2
But to our knowledge, partitioning of hydrogen chloride from brine has not been documented in
natural geothermal systems where there is no direct influence from volcanic gases.

Occurrence of Hydrogen Chloride in Steam at The Geysers

Early production from The Geysers reservoir consisted entirely of saturated steam with generally
low concentrations of noncondensible gas (NCG) that varied systematically throughout the field
[14, 15]. Production of saturated steam was not associated with corrosion problems in well
casings and surface piping. As The Geysers underwent rapid development during the 1980s,
reservoir pressure began to decline and superheated conditions developed in the reservoir.
Corrosive steam with relatively high concentrations of volatile acid chloride has been produced
from many wells since the reservoir steam transitioned from saturated to superheated conditions
in the latter half of the 1980s. Hirtz et al. [16] noted that it is unclear whether the volatile
chloride is transported in the vapor phase as ammonium chloride or as hydrogen chloride.
Simonson [9] later demonstrated that ammonium chloride transport can occur, with ammonium
chloride partitioning (up to 100 ppm chloride) from near-neutral, high-salinity brines at 350 ºC.
Haizlip and Truesdell [17] and Walters et al. [18] reported that steam from the high-temperature
zone (HTZ) in the northwest Geysers tends to have elevated NCG and volatile chloride
concentrations. The top of the HTZ is defined by a temperature at or above 260 ˚C, measured
during drilling. Hirtz et al. [16] reviewed the various origins proposed for volatile chloride that
include reactions involving concentrated brine and/or solid chloride phases at temperatures
above 300 ºC. Since hydrogen chloride gas is a well-documented component of many high-
temperature fumaroles in volcanic environments [19], hydrogen chloride in The Geysers steam
may emanate directly from a magmatic heat source. Whatever the origin of the volatile chloride,
its occurrence in produced steam signifies a dry (i.e. superheated) path from its source to the
production wellbore. Otherwise, the volatile acid chloride—whether ammonium chloride or
hydrogen chloride—will ionize, form acid, and react with rocks in the reservoir.
Calpine Corporation, the operator of most of The Geysers wellfield and 16 of the 19 operating
power plants, has collected steam samples from wells producing high-chloride steam using a
downhole sampler (DHS) designed by Sandia National Laboratories and Thermochem, Inc. [20].
The DHS employs a eutectic material with a high heat of fusion to condense steam and allow
collection of a significant volume of both condensate and NCG gas. Samples collected with the
DHS show higher concentrations of chloride when the sample is taken in the wellbore
immediately above the deepest steam entries. The high-chloride steam is believed to emanate
from the HTZ. Production of high-chloride steam from wells that do not penetrate the HTZ is an
indication of vertical permeability connecting the HTZ with the normal reservoir. Figure 2-1
shows, in a southwest-to-northeast cross section, the distribution of steam entries (short red
“ticks” crossing wellbores and representing steam-bearing fractures) for producing wells in the
areas of Unit 5 and Unit 6. The surface concentrations of chloride in steam produced from these
wells are shown adjacent to their bottom-hole locations in the cross section. In this area of the
steam field, although wells have a wide range of depths and steam entries, none are believed to
penetrate the HTZ. The cross section indicates that, in general, the concentration of chloride is
controlled by an upwelling dome of high-chloride steam in the vicinity of the SB25 well. None
of the shallow wells in the area surrounding the SB25 production well produce steam with high
chloride. The implication is that high-chloride steam is migrating upward from the HTZ along a
steep fracture zone [21].

14017525 2-3
Figure 2-1
Southwest-to-northeast cross section through the areas of Units 5 and 6 at The Geysers. Red
ticks represent steam entries to wells. Produced chloride concentrations are shown next to
bottom-hole locations of wells. Dashed line represents an upwelling of high-chloride steam
originating in the HTZ.

14017525 2-4
3
CORROSION RESISTANT MATERIALS
Studies began in the 1980s to determine how and why materials were failing in geothermal
applications, so a solution to the corrosion problem in geothermal fields could be developed.
Wilson and Lichti [22] recognized a need to standardize materials testing because no two
geothermal wells had exactly the same chemistry. Laboratory and field testing revealed an
obvious pattern of superior corrosion resistance among high-alloy steel and rapid degradation of
low-alloy and plain carbon steel. However, results were still too irregular to make any
unanimous claims about the overall best material for geothermal well construction. Carbon
steels, low-alloy or chromium-molybdenum (Cr-Mo) steels, martensitic and ferritic stainless
steels, high-nickel (Ni) alloys, and titanium (Ti)—all of which are commercially available
alloys—were evaluated in geothermal brine at high temperatures (260 °C). [23]
The Geothermal Task Force of the International Energy Agency initiated research in the mid-to-
late 1990s, including laboratory and field testing, to evaluate corrosion at high temperatures in
geothermal wells and materials for geothermal power production [24]. Task Force Subtask C
focused on materials research in two areas: (a) collecting data and information on materials, and
(b) laboratory and field testing to investigate failure mechanisms. Principal organizations
participating in Subtask C included Materials Performance Technologies, New Zealand (high-
temperature corrosion and deep-seated geothermal wells developed in New Zealand and proven
to have integrity and reliability); Tōhoku National Industrial Research Institute, Japan (materials
for geothermal wells); and the New Energy and Industrial Technology Development
Organization (NEDO), United States office in California (geothermal well casing and pipelines)
[25].

Failure Mechanisms
This section provides a brief overview of the most commonly observed corrosion mechanisms
related to materials used in geothermal applications; it is not intended to be a complete treatment
of the subject.
Materials issues range from general corrosion (uniform material loss) to more complicated
failure mechanisms like stress corrosion cracking and hydrogen embrittlement—but may also
include scaling and sludge deposition, as well as biological damage from algae or microbes [26].
In the 1990s, a survey [49] across seven companies found general corrosion to be the most
common failure mechanism in geothermal power plants. Corrosion occurs because of a
difference in chemical potentials of metals that encourages the transfer of electrons—it is more
energetically favorable for an electron to migrate from the anode, where it is at a higher potential
(more reactive), to a lower potential at the cathode. The greater the difference in this potential
energy is between two metals (or between a metal and an oxide), the faster the rate of corrosion
will be for the more active (less noble) metal.
For electrochemical activity or galvanic corrosion to occur, an anode, a cathode, an electrolyte,
and a metallic pathway must be present. In the case of high-salinity brines found in many
geothermal systems, the electrolyte consists of naturally occurring salts. Many geothermal fluids

14017525 3-1
have high concentrations of dissolved ions, including sodium (Na+), calcium (Ca2+), chloride
(Cl-), and silica (SiO2) that contribute to the total dissolved solids content of the fluid and
electrical conductivity of the solution. The total dissolved solids content of the geothermal fluid
plays an important role in the transfer of electrons leading to material corrosion. Elevated levels
of dissolved solids contribute to erosion in addition to corrosion, synergistically interacting to
degrade materials in geothermal generation stations. Erosion from high-salinity brine generally
occurs at material bends or elbows. According to Povarov [27], typical damage in geothermal
power plants is related to general corrosion, pitting corrosion, contact corrosion, erosion-
corrosion wear, stress corrosion cracking, drop impact erosion, cavitation erosion, and abrasive
erosion.
When the fluid has neutral pH, a passivation layer consisting of a thin film of sulfide or oxide
will form on the surface of the metal, protecting it from corrosion. The passivation layer prevents
geothermal fluids from coming in direct contact with the metal; instead they must diffuse
through the passivation layer, slowing the rate of corrosion. Oxygen (O2) destabilizes the
passivation layer, causing further corrosion [28]. Tests performed in aerated and de-aerated
environments have shown that an aerated environment accelerates general corrosion and pitting
for carbon steel and brass [29]. The worst corrosion to materials in geothermal steam occurs
when the system is exposed to air, creating an oxidizing environment. Although geothermal
operators do their best to control these conditions, they are not always successful. Aeration may
occur during startup or shutdown, result from a packing leak, or occur for a number of other
reasons [30].
Geothermal fluids may also contain high levels of acidic or noncondensible gases—such as
carbon dioxide (CO2), hydrogen sulfide, chlorides, and ammonia—that lower the pH of fluids in
a water- or vapor-dominated geothermal system. Low-pH water increases the solubility of the
sulfide or oxide film, exposing fresh metal to the corrosive effect of the geothermal fluid, thus
increasing the rate of corrosion. The presence of gases and the phase of the geothermal fluid,
therefore, play an important role in corrosion [27].
Stress corrosion cracking is a common failure mechanism that occurs when metals are under
tensile or residual stress in the presence of chlorides, sulfides, or fluorides (F-) found in
geothermal brines. Corrosion and pitting may occur in any location, but areas with high residual
stress, such as heat-affected zones or areas directly adjacent to a weld, are particularly
susceptible to stress corrosion cracking in the presence of sulfides or chlorides [31]. Since most
welds impose an irregular geometry, irregular flow patterns will be created that may lead to
higher localized corrosion rates. To avoid this problem, welds may be ground flush to the base
metal; however, this may not be an option on the inside diameter of geothermal well casings,
making welds the weak link in geothermal equipment integrity [32].
Four common mechanical failure modes in geothermal well casings result from four unique load
conditions: buckling, collapse, shear, or tensile failure. Euler buckling has the potential to affect
shorter sections of unsupported well casing that are under compressive stress induced by strain
from thermal expansion of the casing material. Although well casings should be supported
laterally by annular cement, shale shrinkage due to desiccation, a poor cement job, or a void that
has been washed out by fluid flows may cause sections of well casing to be unsupported. Wagg
et al. [33] stated that at 550 megapascals (MPa) for 178-millimeter (mm) casing in a 254-mm
borehole, an unsupported length of approximately 5 m is required to allow elastic Euler buckling.

14017525 3-2
Longer sections of casing may buckle in a helical fashion. It is possible to reduce these types of
failures by calculating the strain from thermal expansion and selecting a material with a higher
yield stress than would be induced by the calculated strain. Collapse is a failure mechanism that
results from a difference in fluid pressure between the outside and inside of the well casing—a
high ratio between external and internal pressures causes the well casing to implode [33]. To
protect against mechanical failure such as buckling, a standard known as the Code of Practice for
Deep Geothermal Wells (NZS 2403:2015) has been developed by Standards New Zealand® [34].
Once material parameters in particular conditions are understood and more research is
performed, setting standards will become easier.
Wright [35] and De Waard and Milliams [36] concluded that various compositions of carbon
steels in anaerobic, CO2-dominant, moderately acidic solutions had variable corrosion results
depending on the phases within the metal or the presence of inclusions. Wright [35] lists
extensive research indicating that localized corrosion, or pitting, occurred due to multiple phases
within the material exhibiting different electrochemical potentials that create galvanic couples.
The greater the difference in potential is, the higher the rate of corrosion will be. Galvanic
couples are present in steels due to the different electrochemical potentials of cementite—the
carbide-rich phase of austenitic steel, which acts as the cathode—and ferrite, which acts as the
anode. At elevated temperatures, carbides may diffuse out of the crystal grains, causing a
reduction in corrosion resistance in regions with a high concentration of carbide precipitates [14].
When some alloys come in contact with another corroding material, the metallic alloy may
become charged with hydrogen, which induces hydrogen embrittlement [37]. In some wells
where the reservoir fluids contain high concentrations of heavy metals, galvanic corrosion may
occur beneath the metal scale precipitating on the casing surface, causing the underlying base
metal of the casing to corrode. This complexity illustrates the importance of considering
geothermal fluid chemistry and treating the range of conditions as dynamic parameters that
inevitably change with temperature and pressure variations or with the addition of injection
fluids.

Materials
Corrosion protection is the most important damage mitigation mechanism for geothermal well
casings and pipelines [38]. Early proposals for corrosion protection included the use of polymer
coatings and austenitic high-alloy nickel-free chromium manganese steels [39]. Metallic zinc and
aluminum sacrificial coatings were also tested, since the cost of manufacturing hundreds to
thousands of feet of piping made from corrosion resistant alloys was not an economically viable
option at the time. Various application methods were used, including galvanic, vacuum, thermal
diffusion, and spray coating. Zinc- and aluminum-based coatings were applied to pipes, which
were then field tested in high-temperature water and steam. Zinc (Zn)- and aluminum (Al)-based
coatings were recommended as an economically viable solution to corrosion [39]. Polymer and
polymer matrix composite coating systems were tested, in combination with metallic zinc
phosphate [Zn3(PO4)2] compounds or nickel-aluminum alloys, as an intermediate layer to
increase abrasion resistance. The only drawback to using these coatings is that they may not meet
the thickness requirements for corrosion resistance in low-pH geothermal conditions.
Thermal spray coatings provide corrosion resistance; however, they are typically applied to a
thickness of a couple mils (a couple thousandths of an inch). Based on corrosion rates observed
by Lichti et al. [40] in the extreme near-magnetic environment of White Island, New Zealand,

14017525 3-3
well casings in severe geothermal environments would require a very thick (e.g., several hundred
mils) coating of high-nickel alloys, such as Inconel 625 and Hastelloy C-276, to achieve a
service life of 25–30 years typically seen in developed geothermal fields. Solid pipe would be a
better choice under these conditions. In addition, applying a thick thermal spray coating may
create residual stresses within the coating, leading to cracking and subsequent failure of the
casing.
In the mid-to-late 1980s, powder metallurgy offered another option for metallurgically bonding
(cladding) a corrosion resistant alloy to a mild carbon or low-alloy steel well casing. Clad steels
are generally classed as roll bonded, explosion bonded, and overlaid or laser clad [41]. Roll
bonding uses pressure to deform and bond the metals, and sometimes involves heating the two
metals to bond them together. Explosion welding creates a strong bond between the corrosion
resistant alloy (CRA) cladding and the base metal, which is generally a carbon or low-alloy steel
[42]. The bond is created by chemical explosives that fuse the cladding and base material
together. In laser cladding or overlay cladding, a powder coating is applied and fused to the
surface of the base materials with a laser. During laboratory testing, alloy 625 cladding
performed better than other materials in terms of strength and corrosion resistance [43].
From the 1980s on, carbon steel was commonly used in steam gathering systems and power plant
piping because it made economic sense. However, chromium-manganese steel performed better
during corrosion testing in thermal mineralized waters when alloyed with molybdenum (1%–
2%), copper (Cu) (1.5%), and silica in low-temperature geothermal fluids (60 °C) [12]. In
Kurata’s 1992 survey [44] of seven geothermal power generation companies, wells and pipelines
were either replaced, exchanged for a higher grade material, or patched with undefined other
material. Few companies reported patching pipes, but 25%–30% of the companies replaced,
exchanged, or took other action.
From these data, Kurata et al. [44] developed an equation to calculate corrosion rate based on
variables for pH, temperature, and activation energy. According to experimental data collected
from the Fushime, Matsukawa, Hatchobaru, Kurokawa, and Onikoube geothermal fields in
Japan, actual measured corrosion rates exceeded calculated corrosion rates. Further research
through the mid-1990s led Lichti [45] to develop a computer-based program for materials
selection based on environmental input parameters, such a temperature and pH. These analyses
have the greatest potential for successful materials selection for geothermal environments
because geothermal fluids have unique, highly variable chemistries and parameters such as pH,
salinity, temperature, and dissolved gas concentrations that must be taken into account when
selecting materials. Additional research is needed to enable the use of these tests and their
outcomes as data points in a computer-based materials selection program. Recently, Kurata’s
corrosion resistance equation was revised to reflect environmental parameters of a well in
Indonesia, where materials were selected for testing based on equation predictions [38].

Current Research for Deep-Seated Geothermal Wells

From early 1990s research, Lichti et al. [30] determined that for fluid pH less than 3 and
temperatures greater than 300 °C, pipe material mass should be more than 30% chromium—
potentially requiring duplex stainless steel or high-nickel alloys. More recent testing has been
done on high-chromium, nickel, and/or molybdenum materials, all of which aid in corrosion
protection. Flow velocities also influence corrosion rate of various materials. At a pH of 3,

14017525 3-4
low-alloy and carbon steels showed a direct relationship between fluid velocity and corrosion
rate, while Type 304 and Type 316 stainless steels did not. At a pH of 2, stainless steels (and
even some high-alloy stainless steels) had corrosion rates over a hundred times greater in flowing
fluid than in a static fluid. Hastelloy C-276 and titanium were found to have low corrosion rates
under both static and flowing conditions, making them acceptable materials for corrosion
reduction [46]. Sanada et al. [46] identified gaps in geothermal materials research that need to be
addressed before the industry can make headway with deep-seated geothermal wells, as shown
below in Figure 3-1 [25].

Figure 3-1
Schematic showing gaps in geothermal materials research [25]

Deep-seated geothermal investigations are underway in Italy, Japan, Mexico, Indonesia, New
Zealand, the Philippines, and the United States. The DOE has compiled a database on materials
performance under different geothermal conditions as part of the Geothermal Legacy Collection,
a searchable public access platform on the Office of Scientific and Technical Information
website (http://www.osti.gov/geothermal/) [31]. The database was compiled to increase
information exchange and to aid scientists and engineers in selecting the optimal material(s) for a
geothermal well, based on unique reservoir conditions.
Corrosion may also be prevented by using corrosion inhibitors that interfere with corrosion
processes by forming a thin coating (with no effect on heat transfer), or by coating pipes with a
“polyethylene coating, epoxy coal tar, or fiberglass reinforced plastic (FRP)” [26]; however, FRP

14017525 3-5
does not hold up well in extreme thermal conditions. Newflex pipeflex, or other anti-corrosion
coating materials may also have a future.
Steam washing is used to help prevent corrosion in the steam turbines. Corrosive chemical
species are minimized by injecting alkaline solutions and caustic soda into the header [26].
Chemical injections of sodium hydroxide that affect the pH of geothermal fluids have recently
had great success in reducing corrosion. Cost-effective injection of chemical additives has been
demonstrated to help control corrosion in geothermal reservoir fluids with pH of approximately 2
[47].
Coatings have not been researched as extensively as cladding or solid alloys. On their own,
polymer coatings do not have the wear-resistance to endure geothermal conditions and their
melting point is generally too low for high-enthalpy geothermal conditions [48]. Brookhaven
National Laboratory researched geopolymers, composite materials made from clay particles and
polymers [49]. These composite materials provide better wear and corrosion resistance. In
laboratory testing at 300 °C in a simulated hot geothermal brine environment, the
polyphenylsulfide (PPS) with montmorillonite (MMT) filler was discovered to provide adequate
corrosion protection. This composite material had improved mechanical properties, a higher
melting point than PPS alone, and increased crystallization energy. Instances of hydrothermal
oxidation related to sulfide-sulfite linkage transformations were eliminated for this composite
material [49]. Tests of geopolymer compounds also found that the coating adhered better when
an intermediate layer of zinc was applied. Multi-layer coating systems are often necessary for
successful corrosion protection. In some cases, combining multiple layers enhances the
properties of the coating system, as with self-advancing coatings demonstrated by Sugama [50].
These self-advancing coating systems—which are a combination of styrene acrylic latex and
calcium aluminate cement—exhibited good corrosion resistance in laboratory testing of the
conductivity of corrosive electrolytes in an autoclave at 250 °C.
Geothermal power stations inspect their equipment more frequently than is customary at
conventional thermal generation facilities. Baseline data collected during these inspections help
power plant engineers and inspectors track the corrosion rate of a material and assess the
condition of the material over time [51]. In more recent years, failures have been recorded and
root analyses performed to determine the corrosion mechanism and to find a way to prevent
corrosion in the future. Recently, failures have been studied in greater detail to identify specific
locations in the system and pieces of equipment that are being degraded. By using computer-
based technology to estimate the dynamic environmental parameters in geothermal wells, well
models can predict the areas of highest corrosion where casing strings designed with CRAs are
deemed necessary.

Cost of Materials
Material costs for corrosion resistant alloys are high because these materials generally are a
blend of high-cost metallic elements. Plain carbon steel is comparatively inexpensive and has
commonly been used in geothermal plants. But its poor corrosion resistance makes it a cost-
ineffective choice, as more corrosive surface system conditions are encountered or geothermal
reservoir conditions become more corrosive over time. The service life of a geothermal well is
typically 25–30 years, so well casing materials should have a similar life expectancy. Even
though the cost of corrosion resistant alloys can be anywhere from 3 to 7 times the cost of carbon

14017525 3-6
steel, their improved corrosion resistance could pay for itself compared with plain carbon steel
that requires replacement after only a few years of service.
Typically, cladding would be the most cost-effective way to manufacture well casing with
corrosion resistant characteristics without paying for an expensive CRA. However, market
variance for specialty metals pricing has currently brought the price of nickel down to a six-year
low, making it more cost-effective to manufacture well casing entirely out of a CRA, instead of
cladding the interior of a carbon steel pipe. The current price is a reflection of market variability
and may change rapidly to higher values, making it difficult to develop long-term plans for
projects that require nickel-based alloys [52]. Blakely and Olivas [42] suggest that significant
cost savings could be achieved if a thin clad stainless steel, monel, zirconium (Zr), or titanium
were applied to piping. Thin titanium cladding is impractical because of problems with metal
workability, while high-nickel (alloy 625 or C-276) cladding is supposedly more expensive than
manufacturing the same length of solid alloy pipe [53].
Research performed on CRA for vessels and tanks for geothermal power plants shows similar
pricing of materials (Table 3-1) [54].
Ashland Engineering recently compared price ratios between alloy 2205 duplex stainless steel
and alloy C-276 (Table 3-2) [52, 55]. These values were originally compared to prices for fiber-
reinforced plastics, which provide great corrosion resistance but cannot be used in high-
temperature applications.
Table 3-1
Cost comparison of materials for fabrication of pressure vessels and heat exchangers

Alloy 316L 2205 Titanium Titanium 825 625 C-276 Zirconium

Grade 2 Grade 12 702
Cost 1.0 1.01-1.2 1.16-1.5 1.43-1.6 2.2 1.92-3.6 2.04-3.6 4.0
Factor

14017525 3-7
Table 3-2
Price comparison of stainless steel plate and carbon steel plate clad with C-276

Material Cost Cost Ratio

Mild carbon steels (P110, N80, P235GH, P265GH) > 97.5% Fe 1
Stainless steel (430F) 16–18% Cr 1.7
Alloy 2205 (9.5-mm steel plate) $1,050/m2 1.9
Alloy C-276 clad carbon steel plate $1,548/m2 2.9
Duplex (318LN) 21–23% Cr, 4.5– 6.5% Ni, 2.5–3.5% Mo 7.1
Stainless (Alloy 316L) 16.5–18.5% Cr, 10–14% Ni, 2–2.5% Mo 8.3
Titanium alloy (Grade 2) 99.2% Ti 16.2
Special Austenite (Alloy 914L) 19–21% Cr, 24–26% Ni, 4-5% Mo, 19.4
< 2% Cu
Superduplex (Alloy 31) 26–28% Cr, 30–32 %Ni, 6–7% Mo, < 1.4% Cu 33
Fe—iron
Cr—chromium
Ni—nickel
Mo—molybdenum
Cu—copper

Figure 3-2
Cost comparison of materials for tank construction as performed by Ashland Engineering in their
2015 report [52]

14017525 3-8
Ashland Engineering also provided data for cost comparison of materials for tank construction,
which show the drastic increase in price difference between carbon steel and high-nickel alloys,
such as alloy C-276 (Figure 3-2).
Love et al. [56] calculated that, over a 30-year period, the cost of carbon steel pipe would be
more than five times that of Beta-C titanium because of the frequent replacement of carbon steel
required by rapid corrosion rates of 1,000 to 2,000 mils per year.

14017525 3-9
14017525
4
CORROSION MITIGATION TECHNIQUES
The Geysers has more than 350 producing wells, most of which produce noncorrosive steam
containing less than 0.5 ppm chloride. About 20% of the wells produce steam with 1.0 ppm
chloride or more; these wells are corrosive if not sufficiently superheated. A few wells currently
produce steam containing as much as 100 ppm chloride, and several wells have produced steam
with up to 200 ppm chloride. This steam can be corrosive at a superheat of 25 °C or more, due to
the dew point effect of ammonia. An excess of ammonia relative to chloride in steam lowers the
vapor pressure of volatile chloride, requiring much higher superheat to prevent initial
condensation compared with pure hydrogen chloride in steam. Corrosion of casings, wellheads,
and pipelines occurs where hydrogen chloride-bearing steam encounters films of moisture on
steel surfaces or surfaces sufficiently cool to initiate dew point deposition of ammonium
chloride. At those points, hydrogen chloride (or ammonium chloride) partitions from the steam
and concentrates in the moisture, transferring up to thousands of ppm chloride into the liquid.
Liquids collected from pipelines at The Geysers have contained 40,000 ppm ferrous chloride
(FeCl2), the product of hydrogen chloride reacting with steel well casings and pipelines.
Typically, the excess of ammonia relative to chloride in most chloride-bearing steam
condensates results in pH above 5 at 20 °C, although systems without excess ammonia may
produce condensates below pH 2 at 20 °C. Ammonia is a very weak base at steam line
temperatures (~175 °C); therefore, the measured pH of condensate containing ammonia at
ambient temperature can be much higher than the measured pH of condensate in the pipeline at
high temperature. Condensate films with high-chloride contents and excess ammonia typically
have a pH of about 4 at steam line temperatures, sufficient to cause the very high corrosion rates
observed at The Geysers.

Early Volatile Chloride Corrosion Mitigation Techniques

High-chloride steam is handled in several ways to prevent corrosion. Many moderate-chloride
wells (~10 ppm) that are superheated (≥ 20 ˚C) flow to pipelines that are also superheated. In this
case, all the steam can be passed through the gathering system and scrubbed immediately
upstream of power plants by desuperheating and steam washing using hotwell condensate. In
these steam lines carrying high-chloride, high-superheat steam, prevention of cold spots is
critical at pipeline heat sinks such as valves, rupture disks, and pipe anchors. In some cases, cold
spot prevention is impractical because a few wells feeding the same part of the gathering system
may be saturated, effectively quenching the superheat needed to carry the steam safely to the
plants.
Many wells have been damaged when conditions required curtailing the flow of the steam field.
Lower flow rates allow wellhead pressures to rise, thereby decreasing superheat. Slower steam
flow through wellbores and pipelines also allows more conductive heat loss, increasing the
probability of dew point conditions along the flow path. A partial solution to this problem at The
Geysers is to preferentially curtail wells that have low chloride while keeping high-chloride
wells as close to full flow as possible. This requires monitoring well chloride levels to identify
which wells can be safely curtailed.

14017525 4-1
The first full-scale, volatile chloride corrosion mitigation system at The Geysers was designed
and operated by Thermochem, Inc. (Thermochem) at the Aidlin project in 1989 [57]. This
system, still in use today, involves water and caustic (sodium hydroxide) injection at each
production wellhead with subsequent liquid removal by vertical separators, and a final two-stage
steam wash system at the plant inlet. Although steam washing with dilute sodium hydroxide has
been used successfully for many years at The Geysers and Larderello to protect steam gathering
and power generation equipment [16], it causes a reduction in steam utilization efficiency. The
wash water quenches steam superheat and condenses steam, reducing the total enthalpy and mass
flow rate. The efficiency loss in steam washing is greatest in vapor-dominated fields, due to the
high levels of superheat typically present. For The Geysers, it is estimated that up to 5% of
recoverable energy is lost as a result of steam quenching and increased steam utilization factors
when steam washing is employed. For the IDDP-1 well in Krafla, a loss of about 10% is
estimated with conventional steam washing [58].
Steam washing with sodium hydroxide is highly effective for corrosion control in pipelines and
for protection of the steam turbine downstream. Typically, volatile chloride is reduced to less
than 1 to 5 ppm at the outlet of the wellhead scrubbing process (95–99% removal efficiency),
and to less than 0.1 ppm after a second-stage water wash at the plant inlet. In conventional fossil
fuel and nuclear steam turbine power plants, chloride-induced stress corrosion cracking is
considered to be one of the major problems in turbine reliability. These units operate at much
higher inlet pressures and superheat than geothermal units, generating greater stresses on the
machines and increasing the likelihood of salt deposition upstream of the Wilson line
(condensation zone). Sodium chloride is often found in the turbine deposits of conventional units
and is linked with stress corrosion cracking. Hydrochloric acid and ammonium chloride have
also been identified in these turbine deposits and are considered highly undesirable due to their
extreme corrosivity, in addition to their tendency to cause stress corrosion cracking.
The turbines in use at the Cerro Prieto geothermal field in Mexico are exposed to mechanically
entrained sodium chloride, rather than volatile chloride, and have experienced negligible
corrosion. The turbine blade material in use at these units is 12% chromium steel, similar to the
blade material at The Geysers units. The Geysers’ caustic scrubbing process for removing
volatile chloride generates sodium chloride and sodium carbonate (Na2CO3), which are entrained
in the turbine steam at trace levels, but have turbine impacts similar to those experienced at
Cerro Prieto. Operating standards for steam purity at The Geysers require less than 0.1 ppm
chloride at the inlet to the turbine for most plants. Although units have run for many years with
volatile chloride in the range of 0.1 to 1.0 ppm, this concentration is not recommended for long-
term operation due to the potential for cracking and pitting damage to the turbine alloys in use.
At The Geysers, wells producing high chloride or a combination of moderate chloride and low
superheat such that corrosion has been identified as a problem are equipped with corrosion
mitigation facilities (CMFs). Near the wellhead, enough sodium hydroxide solution is sprayed
into the steam for desuperheating, as well as to absorb and neutralize the hydrogen chloride. A
separator removes the spent liquid from the pipeline downstream. The amount of sodium
hydroxide needed is stoichiometric with chloride, plus a slight excess, resulting in a separated
water pH of about 8.5. The scrubbed steam flows to the main pipeline at saturation and serves to
lower the superheat of the steam with which it is commingled in the gathering system. This
raises the probability that volatile chloride in dry, superheated steam from non-CMF wells will
mix with wet steam downstream of the CMF. Corrosion can then occur at the mixing point and

14017525 4-2
beyond, so this situation is carefully avoided. In some parts of the field, parallel pipelines are in
use—one carrying moderate-chloride superheated steam, the other carrying steam from CMF
wells with very low chloride and saturated steam.
The 130-MW Coldwater Creek Power Plant, built by the Central California Power Agency
(CCPA), was located in the northwest Geysers, where chloride and NCG concentrations are the
highest and the HTZ is the shallowest. The plant never ran at full load and was shut down in
1994 after only six years of operation, primarily because of corrosive steam. Much of the steam
produced to the CCPA plant originated in the HTZ. Although high-NCG steam was problematic,
chloride corrosion proved to be the terminal problem. Wellhead mitigation was eventually used
on all wells, but extensive gathering system damage and casing corrosion downhole would have
required an extensive replacement program that made the project uneconomical. However, near
the CCPA plant in the northwest Geysers, the Aidlin Power Plant has been running for more than
25 years without corrosion problems. The Aidlin wells produce some of the highest-chloride
steam at The Geysers, with one well at 160 ppm and another 200 ppm. The reason this project
has been successful is that properly designed CMFs were installed before startup and diligently
operated ever since, demonstrating that it is possible to utilize high-temperature corrosive
resources.
Most wells in Larderello have high superheat (> 50 ºC) and low ammonia compared with wells
in The Geysers, so volatile chloride can be allowed to pass through the gathering system and be
removed by scrubbing with water and caustic at the power plant. One of the main drawbacks for
wellhead water scrubbing in vapor-dominated fields is that each wellhead location scrubbed
produces saturated, wet steam. When this steam combines with untreated steam from other wells,
the superheat in the entire gathering system is reduced. Severe corrosion can occur when the
steam from relatively low-chloride wells (1–3 ppm), which previously did not require scrubbing,
mixes with the wet steam from treated wells. Ideally, superheat throughout the gathering system
is maintained at a maximum level to avoid the volatile chloride dew point.
Downhole Volatile Chloride Scrubbing and Neutralization
Steam washing has also been successfully employed downhole at The Geysers. Generally, a
minimum of about 20 ºC superheat is required to prevent severe corrosion damage to casings and
piping for wells at The Geysers producing a few ppm or more volatile chloride. Wells producing
chloride at more than 2 or 3 ppm under nearly saturated conditions must be treated downhole or
shut in. In some cases, steam from volatile chloride producing wells is saturated at the surface,
and may contain significant fractions of liquid due to wellbore condensation or intrusion of
shallow water through casing breaks and corrosion perforations. A solution of water and caustic
is injected at depths of 1,000 to 2,000 m to mitigate pH and protect the casing and wellhead.
Mitigating downhole corrosion by injecting a water-sodium hydroxide solution through coiled
tubing was successful at the CCPA project in the northwest Geysers, but such severe damage had
already been done to the well casings, gathering system, and plant that the project was
abandoned. At about the same time, the Aidlin project was brought on-line in the northwest
Geysers. Downhole corrosion mitigation was also employed at Aidlin, but was not needed long
term because injection was changed to return these wells to superheated conditions.

14017525 4-3
New Volatile Chloride Corrosion Mitigation Technologies
Alternatives to traditional steam washing for removal of volatile chloride from superheated
geothermal steam have been conceived and demonstrated over the last ten years. These
alternatives, termed Dry Steam Scrubbing (DSS), are designed to remove volatile chloride from
steam without significant desuperheating and subsequent reduction in steam utilization
efficiency.
Initially, Hirtz [59] discovered that solid-phase, alkaline compounds such as sodium carbonate
(Na2CO3) and sodium borate (Na2B4O7·10H2O) can absorb up to 90% of the produced hydrogen
chloride from superheated steam in wellbores at The Geysers. This phenomenon was observed
during testing of downhole hydrogen chloride mitigation systems (water washing with sodium
hydroxide in the wellbore) after the downhole injection was shut off and the alkaline scale in the
wellbore dried out.
Fisher and Jung [60] proposed dry scrubbing, oil washing, and hybrid cleaning as alternatives to
traditional steam washing. Dry scrubbing consists of adsorbing or absorbing steam impurities
with an agent, followed by separating the spent agent. The spent agent is reconditioned and
recycled into the process. Oil washing consists of spraying oil into steam, separating oil from
impurities, and recycling the oil. High-boiling point oils (that exist in equilibrium with highly
superheated steam) can be used, so there is no heat loss due to vaporization of the scrubbing
liquid. Hybrid cleaning is described as contacting steam with a liquid/solid mixture formulated to
react with impurities in the steam pipeline. The active components are separated from the steam,
regenerated, and recycled.
Gallup [61] investigated removing hydrogen chloride from steam by boiling a weak hydrogen
chloride solution and passing the vapor through small columns containing various solids and
liquids. The materials tested for hydrogen chloride removal included:
• Absorbents—calcium carbonate (CaCO3), anhydrite (CaSO4), sodium carbonate, and liquid
anion exchange amines.
• Adsorbents—zeolites, hydrotalcite, silica gel, activated carbon, carbonaceous adsorbent,
phenolic foam, and a solid anion exchange resin.
A U.S. patent was issued in 1999 on these process ideas (U.S. Patent No. 5,879,433) and a patent
was also secured in Italy, but no other foreign patents were filed. Thermochem has the exclusive
license to the U.S. patent, which covers many of the current DSS innovations.
Beginning in 2000, Thermochem conducted pilot testing at The Geysers, which demonstrated
that calcium carbonate was an effective DSS agent with 98% removal efficiency in wells
producing up to 160 ppm hydrogen chloride and 30 ºC superheat, the highest available test
parameters at the time. Mixtures of amines and oil were also effective, but were more costly due
to amine vapor-phase losses despite liquid-phase recycling and liquid-liquid regeneration [62]. A
full-scale demonstration of DSS using calcium carbonate was conducted in 2008 on a well
producing 100 ppm hydrogen chloride at 34 ºC superheat. Removal efficiency was in the range
of 92%. The lower removal efficiency was attributed to flow channeling through the bed due to
uneven distribution in the larger-diameter vessel (1,370 mm diameter in the full-scale
demonstration vs. 305 mm diameter in the pilot test).

14017525 4-4
A commercial water treatment company has reported that a solution of amines and dilute
potassium carbonate (K2CO3) [63] can be used to treat hydrogen chloride at The Geysers, with
minimal loss of superheat. But the treatment scheme results in complete dry out and potassium
chloride (KCl) salt deposition at superheat values above 16 ºC. In most cases, the volume of the
liquid phase is insufficient to effectively remove hydrogen chloride at the wellhead. The process
has some short-term applications where the reaction products of hydrogen chloride neutralization
can be allowed to remain in the pipeline, but is not useful or cost-effective to treat steam with
superheat above about 160 ºC.

Hydrogen Chloride Mitigation Strategies—Detailed Process Examples

The following section describes example process conditions and chemical data for treatment of
hydrogen chloride produced by a subcritical, high-temperature, superheated geothermal well.
These examples are intended generally to represent treatment of targeted production from deep-
seated, high-temperature, and EGS-type reservoirs approaching 400 °C.
The example process design conditions are as follows:
Wellhead temperature: 380 °C,
Wellhead pressure: 22 bars,
Steam enthalpy: 3,200 kJ/kg,
Mass flow: 30 kg/s,
Hydrogen chloride concentration: 100 mg/kg.

Conventional Water and Caustic Scrubbing

As discussed above, water and caustic scrubbing has been successfully employed to treat volatile
chloride at The Geysers and Larderello for many years. The main drawbacks of this treatment are
loss of superheat, condensation of steam, and mixing of saturated steam with superheated steam
from other wells with low to moderate levels of hydrogen chloride—resulting in corrosive
conditions that did not previously exist when the entire gathering system remained superheated.
A water and caustic scrubbing process is shown in Figure 4-1 for the example well conditions
listed above. With water at 40 ºC and 30% caustic (sodium hydroxide) used to scrub steam from
the well at 8 bara (absolute pressure), 5.5 kilograms per second (kg/s) water and 31 liters per
hour (lph) caustic (sodium hydroxide) would be needed to achieve a liquid fraction of 2% after
desuperheating and a neutral-to-alkaline pH. This will result in 34.8 kg/s saturated steam at
170 ºC, while the 0.71 kg/s residual scrub water will be sent for disposal. The clean water must
be oxygen-free (< 10 parts per billion [ppb]); the usual source for this process is hotwell
condensate from a surface condenser, which is typically 40 ºC. Direct-contact condenser water is
too high in oxygen and may contain undesirable treatment chemicals (oxidizing biocides) from
the cooling tower. With good mixing design, only a 2% residual water fraction (after
desuperheating) is typically required.
If the scrub water were preheated to 123 ºC using low-pressure brine in a heat exchanger, then
6.4 kg/s scrub water would be required, resulting in 35.7 kg/s steam (about 1 kg/s improvement
over unheated scrub water), as shown in Figure 4-2. Given the low residual scrub water flow

14017525 4-5
rates, additional heating using the 170 ºC scrub drain water results in a negligible increase in
efficiency (0.05 kg/s).
For maximum energy recovery from this example well, it would be better to scrub the steam at
20 bara if the well can sustain the same flow of 30 kg/s at this pressure (Figure 4-3). At 20 bara,
5.9 kg/s scrub water, preheated to 123 ºC, is required to reach a 2% residual scrub water fraction.
The caustic requirements would be the same as in the above cases. Under these conditions,
35.1 kg/s steam can be recovered at 20 bara. This is improved by only 0.1 kg/s if the scrub
injection water is further preheated using the drain water at 212 ºC.

Figure 4-1
Water and caustic process flow diagram—8 bara

14017525 4-6
Figure 4-2
Water and caustic process flow diagram—8 bara, pre-heated scrub water

14017525 4-7
Figure 4-3
Water and caustic process flow diagram—20 bara

Water and caustic scrubbing is a very low-risk approach, as the technology is well understood
and mature. The drained scrub water could be disposed of with the low-pressure injection brine
with no significant impact on silica scaling. A clean, oxygen-free water source is needed, as well
as a small heat exchanger to preheat the water using injection brine. The annual chemical cost
would be roughly US$200,000 based on current bulk pricing in the United States. At this high
level of superheat, a mixing run that is longer than normal would be required to desuperheat the
steam and completely scrub the hydrogen chloride. The mixing spool must be fabricated from
alloy 625, as discussed above. Typically, mixing spools are only 1 to 2 m long, but for this well a
spool at least 10 m long would probably be required. Residual hydrogen chloride after scrubbing
would be in the range of 0.1 to 1 ppm at the most; with careful mixing process design, the
residual hydrogen chloride could be less than 0.1 ppm.
Severe corrosion can occur immediately upstream of the desuperheating point in the scrubbing
process (the quench zone) due to conductive heat loss through the pipe. Condensation in the
quench zone does not have the benefit of chemical neutralization by scrub water, and is
extremely corrosive. The use of duplex stainless steel alloy 254 SMO for mixing spools has been
considered. This material was tested at The Geysers in the bulk steam flow and drop pots of
corrosive wells, but never in the quench zone [64]. Although 254 SMO performed well with
respect to stress corrosion cracking and pitting corrosion resistance, super-duplex 2507
performed better. However, no alloy except 625 has survived the quench zone in wells at The
Geysers. This is the only material routinely used in water and caustic mixing spools at The

14017525 4-8
Geysers. Wellhead tees and casing heads are also fabricated from alloy 625 for highly corrosive
wells that are not sufficiently superheated.
Dry Steam Scrubbing
The removal of hydrogen chloride and/or ammonium chloride from geothermal steam has been
thoroughly tested and demonstrated at The Geysers using calcium carbonate (calcite or limestone
pellets, ~ 40 mm diameter) as the absorptive material [62]. Calcium carbonate consumption in
the DSS process is governed by the reaction between calcium carbonate and hydrogen. This
process can occur directly between the superheated gaseous phase containing hydrogen chloride
and the solid calcium carbonate substrate, without the presence of liquid water:
CaCO3(s) + 2 HCl(g) = CaCl2(s) + H2O(g) + CO2(g)
There are no known competing reactions that could occur in geothermal steam exposed to
calcium carbonate, unlike caustic. Sodium hydroxide reacts rapidly with carbon dioxide to form
sodium carbonate, but is still effective in neutralizing hydrogen chloride. Due to the very low
solubility of calcium carbonate at high temperatures (retrograde solubility), virtually no calcium
carbonate is dissolved when the material does become wet during wellfield operations or
washing of the calcium carbonate bed to remove reaction products. This low solubility is
important in maintaining the mechanical integrity of the bed material and properties such as
reactive surface area, porosity, and permeability.
Reflux regeneration is the most effective means of removing reaction products from the calcium
carbonate bed and maintaining long-term performance. In this case, the reactive surface is
regenerated, but the calcium carbonate is still consumed in the process through reaction with
hydrogen chloride. Regeneration is accomplished by shutting in the outlet pipe from the vessel
for several hours while leaving the inlet open (pressurized with steam) and the drain valve
partially open to allow a slow steam bleed and liquid drainage. Calcium carbonate bed
regeneration is required about every 15 to 30 days, depending on inlet hydrogen chloride levels.
The process can be accelerated through the use of a water wash, with nozzles spraying clean
water down from the top of the vessel.
Previous research studies at The Geysers demonstrated that up to 160 ppm hydrogen chloride
could be treated effectively by the DSS process. Wells producing 50 ppm or more hydrogen
chloride were conservatively estimated to need a calcium carbonate bed depth of 1,800 mm. The
first full-scale DSS vessel was designed to accommodate a bed depth up to 2,440 mm. The bed
depth can be changed to optimize chloride removal and minimize pressure drop by simply
adding or removing calcium carbonate pellets.
Volatile chloride concentrations are typically reduced down to about 2 ppm (± 1 ppm) regardless
of the inlet chloride concentration range up to 160 ppm. This volatile chloride reduction is
sufficient to prevent corrosion in gathering systems where at least 20 ºC superheat is maintained.
This process has been tested to a maximum steam temperature of 210 ºC. Pressure drop through
the beds at The Geysers is typically less than 30 kilopascal (kPa), with superficial steam
velocities of 1 to 1.5 meters per second (m/s) through beds with a bulk porosity of about 50%.

14017525 4-9
The Ergun function [65] is used to predict the pressure drop as a function of steam flow,
pressure, temperature, bed pellet size, depth, and diameter through the beds:

where the variables represent the following parameters:

ΔP = pressure drop,
L = depth of the packed bed,
G = ρV = mass velocity,
V = superficial linear velocity,
ρ = fluid density,
μ = fluid viscosity,
D = effective particle diameter,
ε = inter-particle void fraction,
g = gravitational constant,
k = conversion factor.
Table 4-1 lists the initial design operating parameters for a calcium carbonate DSS bed installed
on our example well; the basic process flow diagram is given in Figure 4-4. To minimize the
diameter of the DSS vessels and facilitate maintenance, at least two vessels are employed in
parallel for a well of this size. It would be possible to use four vessels, with a reduced diameter
of 1,220 mm per vessel. To assess the relative value of these options, the cost of two large-
diameter vessels versus the cost of four smaller-diameter vessels with the additional cost of
piping would have to be determined and compared. For bed regeneration and refilling, it is best
to have at least two vessels in parallel so that only one at a time has to be taken off-line for
service. Multiple, smaller-diameter vessels may be more desirable as they would be easier to fill
and wash, and less likely to encounter flow channeling due to uneven distribution of the bed
packing.

14017525 4-10
Table 4-1
Calcium carbonate dry steam scrubbing bed design and operating parameters for example well

14017525 4-11
Figure 4-4
Calcium carbonate dry steam scrubbing process flow diagram—20 bara

A more advanced approach would be to use fluidized bed designs. The fluidized bed concept
uses smaller-diameter calcium carbonate pellets that become suspended in the upward steam
flow. Spent material would be separated by a conventional cyclone integral to the vessel, and the
calcium carbonate would be replenished using a slurry pump and an aqueous calcium carbonate
suspension for the feedstock. Fluidized beds can be highly efficient due to thorough contact
between the gas and the solid, but it may be hard to maintain the optimal pellet size in the
fluidization process during long-term operation. Computer modeling software is available to
simulate fluidized bed designs, and pilot testing could be conducted to prove the solids handling
and mechanical feasibility of the concept.
Based on a calcium carbonate (limestone) cost of US$0.50 per kilogram, the annual chemical
cost for the DSS process on this well would be US$65,000 per year. The DSS chemical cost
should be substantially less than that of the caustic scrubbing process.
The primary uncertainty in the calcium carbonate DSS process is the efficiency of hydrogen
chloride removal at 380 ºC. The limiting factor in hydrogen chloride removal may be equilibrium
vapor/liquid partitioning of hydrogen chloride from saturated films of aqueous calcium chloride
(CaCl2) on the surface of the calcium carbonate pellets, and/or sublimation of hydrogen chloride

14017525 4-12
from solid calcium chloride. Highly concentrated (50% by weight) aqueous films of calcium
chloride will exist in equilibrium with superheated steam at the conditions that have been tested.
At temperatures up to 260 ºC, calcium chloride is soluble to 77% by weight. The calcium
chloride will probably dry out completely in 380 ºC steam with 160 ºC superheat; it is not known
what effect this will have on the process. Partial desuperheating of steam with clean water to
300 ºC or 275 ºC may be required if efficiency is poor at the full superheat temperature.
Partial desuperheating could involve water injection directly into the bed to eliminate any
potential for dew point corrosion near the water injection point. However, it is known that dry
limestone efficiently removes sulfur dioxide and hydrogen chloride from flue gas at temperatures
of 300 to 1,200 ºC. Under these conditions, the relative humidity (water vapor saturation) is very
low, suggesting that it may not be necessary to maintain a liquid film on the calcium carbonate in
a highly superheated geothermal DSS process. This possibility can be tested under pilot
conditions.
To date, the evaluation of dry sorbent materials for DSS processes has focused primarily on
commodity chemicals and raw materials that don’t require recycling in order to minimize cost
and complexity. High-boiling point amine and oil mixtures were rigorously evaluated early on in
the laboratory and at The Geysers. Given the cost of these chemicals, recycling is required for
steam containing more than a few ppm hydrogen chloride. Recycling and regeneration of amines
is effective using liquid-liquid extraction with sodium hydroxide in a continuous loop process.
But the vapor-phase and entrainment losses on the order of 15% typically experienced at The
Geysers make the process uncompetitive with other techniques. Losses would be greater at
higher temperatures.
Crystalline molecular sieves, such as naturally occurring zeolite and activated carbon, have
previously been proposed as potential DSS solids. These materials have been demonstrated to
remove hydrogen from steam in qualitative laboratory experiments [61]. But it’s not clear how
the materials would be regenerated after saturation with hydrogen chloride, or what the cost
might be for such a process given uncertain requirements for mass and surface area, as well as
regeneration efficiency.
Other materials that have been considered, but not tested, for hydrogen chloride adsorption in a
once-through DSS process are sodium carbonate (Na2CO3, soda ash) and calcium oxide (CaO,
lime). Sodium carbonate should perform as well as, or better than, calcium carbonate in terms of
hydrogen chloride adsorption. But soda ash is much more water-soluble than limestone, so it
would cause problems with bed integrity during wet stages of operation such as startup and
regeneration. Lime should be even more reactive than soda ash since it is a very strong base and
also reacts with water to form calcium hydroxide [Ca(OH)2, hydrated lime]. Under saturated
steam conditions, lime would quickly convert to hydrated lime, which is still a solid. Under
superheated conditions, lime may remain primarily as calcium oxide and maintain the
mechanical properties needed for a packed bed. Lime is available in large lump (200-mm
diameter) and pebble (6- to 60-mm diameter) form. If limestone is found to be inefficient at
removing hydrogen chloride at temperatures above 300 ºC, then lime might be an alternate
material for a dry packed bed process, and should be pilot tested.
Both soda ash and lime could be used in a spray-dry process, where a slurry of the powdered
material, at roughly 70–80% by weight in water, is sprayed into the superheated steam. The

14017525 4-13
water carrier is immediately vaporized and the solid particles (and possibly highly concentrated
droplets) are carried with the steam to a separation device for removal. The particles would likely
still contain active soda ash or lime that could be recycled back to the slurry. The superheat loss
would be minimal as very little water is needed, but the mechanics of recovering the dry particles
without deposition on piping and vessels could be challenging. Dry scrubbing using solid lime
and spray-dry scrubbing with lime slurries is routinely performed in coal-fired power plants for
sulfur dioxide and hydrogen chloride removal. These processes are also used in municipal waste-
to-energy plants and secondary aluminum plants for hydrogen chloride removal. Process
temperatures are often well above 400 ºC, but pressures are very low—which may facilitate
solids handling (injection and recovery). Efficiencies are reported in the range of 95–99%,
comparable to geothermal wet and dry scrubbing processes for hydrogen chloride.
Hybrid Steam Scrubbing
Hybrid process chemistries involving highly concentrated inorganic liquids that can exist at
equilibrium with superheated steam have been demonstrated in the laboratory. The advantage of
a hybrid process over DSS is that it eliminates vapor-solid contact issues, large vessels, pressure
drop, regeneration, and mechanical systems for solids conveyance. A hybrid liquid scrubbing
process involves a carrier liquid that can be injected and recirculated at sufficient liquid-to-steam
ratios to provide good contact between liquid and vapor for hydrogen chloride mass transfer
(1–5% by mass liquid fractions). The liquid must be high-boiling (exist in equilibrium with
highly superheated steam) without drying out and/or precipitating solids. Finally the liquid must
be kept sufficiently alkaline to allow high-efficiency mass transfer of hydrogen chloride to the
liquid phase, and to minimize corrosion of steel or alloys in contact with the liquid. The
following chemistries have been identified with the properties needed in a hybrid scrubbing
process:

Potassium Hydroxide–Carbonate–Chloride system (KOH–K2CO3–KCl)

This process involves the injection of water and potassium hydroxide (KOH) into geothermal
steam to generate potassium carbonate (K2CO3) and potassium bicarbonate (KHCO3):

2 KOH + CO2  KHCO3 + KOH-  K2CO3 + H2O

At temperatures above 200 ºC, potassium carbonate solubility is at least 70% by weight, and the
liquid phase can maintain a superheat of 70 ºC at a 300 ºC solution temperature. Hydrogen
chloride from the vapor phase reacts with the saturated potassium carbonate solution to yield
neutral potassium chloride:

2 HCl + K2CO3  2 KCl + CO2 + H2O

Potassium chloride has much lower solubility than potassium carbonate, and will precipitate at a
concentration of about 2% in saturated potassium carbonate at 100 ºC. Thus the reaction product
of potassium hydroxide and hydrogen chloride can be removed from the system by cooling (and
then reheating) the liquid recirculated through a series of small heat exchangers. The overall
reaction is then:

HCl(g) + KOH(aq) KCl(s) + H2O

14017525 4-14
The superheat vs. temperature or pressure curves for a 70% by weight solution of potassium
carbonate are given in Figures 4-5 and 4-6. The values in the figures were actually measured at
Thermochem Laboratories—not taken from the literature or derived from calculations. The
solubility of potassium chloride in saturated potassium carbonate at 100 ºC was also directly
measured in the Thermochem laboratory. It is likely that higher equilibrium superheat is possible
at higher temperatures— to date, the limit for the laboratory tests by Thermochem is 300 ºC. At
20 bara, the superheat measured in the Thermochem apparatus was 63 ºC. Additional testing up
to 380 ºC and with higher potassium carbonate concentrations needs to be performed to
determine the maximum superheat possible for this process.

Figure 4-5
Potassium carbonate solution superheat vs. temperature

14017525 4-15
Figure 4-6
Potassium carbonate solution superheat vs. pressure

A conceptual process flow diagram for this system operating at 20 bara and 63 ºC superheat is
given in Figure 4-7. The annual chemical cost would be about US$450,000, based on bulk
pricing in the United States for 50% potassium hydroxide. Potassium carbonate solution could be
used instead of potassium hydroxide solution, but the total annual chemical cost for this choice,
per U.S. pricing, is even higher.

14017525 4-16
Figure 4-7
Hybrid potassium carbonate process flow diagram—20 bara

Calcium Chloride–Calcium Oxide system (CaCl2–CaO)

This process recirculates a concentrated solution of calcium chloride (CaCl2) as the carrier
solution, with addition of lime (CaO) or hydrated lime (Ca(OH)2) as the base to neutralize
hydrogen chloride. The carrier solution is actually the reaction product, so only a blowdown
stream is required to maintain solution volume and concentration, rather than the selective
precipitation and separation required in the potassium hydroxide process. (Note that calcium
chloride is much more soluble than potassium chloride.) The overall reaction chemistry is as
follows:

2 HCl(g) + Ca(OH)2(s, aq) CaCl2(aq) + 2 H2O

The pH of a calcium chloride solution drops with increasing temperature due to complexing of
hydroxide (OH-) with calcium (Ca2+), so an excess of lime is maintained in the system. Lime
reacts with water to form hydrated lime, which has low solubility but provides free hydroxide as
needed to neutralize a strong acid such as hydrogen chloride. The bulk of the hydrated lime will
be present as a suspension in the calcium chloride solution. Based on chemical modeling, the
solution pH will be slightly alkaline at temperatures above 200 ºC (pH ~ 7.7 at 200 ºC and 7.2 at
300 ºC). At this pH, minimal carbon dioxide will be absorbed. As with the caustic and potassium
hydroxide processes, any carbonate (CO32+) that does form will also be available as a base to

14017525 4-17
react with hydrogen chloride. It seems unlikely that calcite (CaCO3) precipitation will be an
issue, but a calcite scale inhibitor could be added if necessary. Given low blowdown rates, the
inhibitor consumption would be very low.
At temperatures above 200 ºC, calcium chloride solubility reaches 80% by weight or higher, and
the liquid phase can maintain a superheat of more than 100 ºC at a 300 ºC solution temperature.
The superheat vs. temperature or pressure curves for an 80% by weight mixture of calcium
chloride in water are given in Figures 4-8 and 4-9.

Figure 4-8
Calcium chloride solution superheat vs. temperature

14017525 4-18
Figure 4-9
Calcium chloride solution superheat vs. pressure

A conceptual process flow diagram for the calcium chloride system operating at 20 bara and
113 ºC superheat is given in Figure 4-10. The annual chemical cost would be US$50,000, based
on calcium oxide (anhydrous lime) pricing in the United States. Given the low chemical cost,
high superheat retention, and simple operation, this currently appears to be the preferred process
for scrubbing hydrogen chloride from high-temperature, high-superheat geothermal wells
producing substantial amounts of hydrogen chloride.

14017525 4-19
Figure 4-10
Hybrid calcium chloride process flow diagram—20 bara

Steam Polishing
Steam scrubbing processes are not 100% efficient—the wet, dry, and hybrid processes all can
potentially leave 1 to 5 ppm hydrogen chloride remaining in the steam. At The Geysers, the final
steam polishing before the turbine is accomplished by second-stage steam scrubbing using
hotwell condensate only. Second-stage steam scrubbing removes residual hydrogen chloride
from the few caustic and water wellhead systems, and any dry steam scrubbing wellhead systems
in the field. It also removes low-level hydrogen chloride (1–5 ppm) from other wells without
wellhead scrubbing. At Larderello, most steam scrubbing is accomplished at the power plant
using hotwell condensate and caustic. Steam entering the turbines is saturated.
It is important to optimize any steam scrubbing system for maximum hydrogen chloride removal,
ideally to less than 0.1 ppm at the turbine inlet. For wet scrubbing systems, this can easily be
accomplished through second-stage polishing at the plant inlet. But for dry and hybrid scrubbing
systems, a steam wash at the plant would defeat the purpose of maximum pressure and superheat
retention for minimum exergy loss.
It might be possible to find turbine blade and rotor material that can tolerate hydrogen chloride in
steam at levels of a few ppm over the long term. Because hydrogen chloride is more volatile and
less acidic at high temperatures, it may not be a problem in the early stages of the machine use.
A number of alloys were tested in Italy in the 1980s as possible alternatives to steam scrubbing,

14017525 4-20
including X15CrNiMo13, XCrMnNiMoN2564, Ti6A14V, X12CrMo13, XCrNiMo126 [66].
Unfortunately, only the latter two were determined safe for the manufacture of turbine blades
and rotors, respectively. An important research effort will be to work with turbine manufacturers
to identify new materials for use in superheated steam with residual hydrogen chloride at high
inlet pressures.
Hjartarson [58] performed a thermodynamic study on various options to utilize steam from the
IDDP-1 well in Krafla. Some of the power cycles considered were: (a) two steam turbines in
series, with the first turbine remaining superheated and wet scrubbing being employed before the
second turbine, and (b) an organic Rankine cycle plant with no steam turbine, and (c) a
conventional steam turbine plant with DSS upstream. It was concluded that the DSS process
would be the most efficient.
A possible alternative to steam wash polishing or new turbine materials is the treatment of
turbine inlet steam with filming and/or neutralizing amines. Thermochem has extensively studied
amine chemistry for applications in dry steam scrubbing and corrosion control [67]. Although
amines are usually not cost-effective as a primary scrubbing process, they do have application in
a polishing process without loss of superheat. Hydrophobic, high-molecular-weight amines were
used for primary steam scrubbing with recycling, and would also be effective in a once-through
process. A water-soluble, lower-molecular-weight amine may also function well for turbine
protection. The overall reaction for the amine scrubbing process from vapor to organic phase is:

(R)3(R)NH2(org) + HCl(g)  (R)3(R)NH3+ Cl-(org)

There can also be significant differences in amine chemistry among amine classes. The primary
amine shown in the above formula and found to be most effective was an amine with a tertiary-
alkyl (t-alkyl) hydrocarbon chain of the general structure (R)3(R)NH2 [the straight-chain primary
amine is represented as (R)NH2]. In hydrocarbon solvents, tertiary-alkyl primary amines are one
to two orders of magnitude more basic than primary or secondary-alkyl primary amines. The
carbon-nitrogen (C-N) bonds in tertiary-alkyl primary amines are 100 to 200 kilojoules per mole
(kJ/mol) more stable than their primary amine counterparts, making them less likely to
decompose at high temperatures.
Another benefit of the tertiary-alkyl primary amines compared with normal-alkyl (n-alkyl)
amines is lower melting point. A tertiary-alkyl amine may still be liquid at 0 °C, but an n-alkyl
amine in the same weight range (C18) melts at 55 °C. A low melting point makes it easier to
handle the chemical in storage and use, but is not critical to the process. A drawback in a primary
scrubbing process for the t-alkyl structures studied is a lower boiling point—310 to 330 °C,
versus 350 °C for a normal-octadecyl (n-octadecyl) amine. But for a polishing process, the
amines would not be recovered and allowed to condense with steam through the turbine.
Therefore an amine with the optimal boiling point could be selected to target specific stages of
the turbine most vulnerable to attack from hydrogen chloride. For a once-through process of
amine protection at the inlet for a turbine using production from the example well, the cost
would be roughly US$75,000 per year. This cost is based on a 30 kg/s steam flow, 2 ppm
residual hydrogen chloride in steam, 1:1 mole ratio amine to hydrogen chloride, a molecular
weight of 325 grams per mole (g/mol), and a cost of about US$4.50 per kg. Amine pricing can
vary dramatically, so it may be possible to find lower-cost generic amines for this purpose.

14017525 4-21
For direct use of high-pressure steam from the example well, volatile silica, volatile salts, and
suspended solids carried in steam must also be considered. With brine production at separation
temperatures of 240–270 °C, silica partitioning to the vapor phase will be on the order of 0.2%,
resulting in several ppm silicon dioxide (SiO2) in the steam (Figure 4-11). Silica in turbine inlet
steam should be less than 0.05 ppm to prevent scaling.
Volatile silica levels may be even higher in the superheated steam at these temperatures,
depending on the source of the fluid and path to the production wellbore. Wet scrubbing to
saturation in the range of 200 °C or less will be capable of removing vapor-phase silica, salts,
and solids. Dry and hybrid scrubbing may also be effective in removing volatile silica and
entrained solids. Vapor-phase silica in steam is present as silicic acid (SiO2·nH2O) and is much
less volatile in the silicate form once neutralized by a base, analogous to hydrogen chloride
scrubbing from steam.

Figure 4-11
Silica partitioning between steam and water as a function of temperature [68]

Characterization of impurities in steam other than hydrogen chloride and their removal will be an
important research effort for utilization of high-temperature, high-superheat steam.
Comparison of Treatment Options for Hydrogen Chloride Corrosion Control
For the steam scrubbing options considered in this report, Table 4-2 summarizes the estimated
process parameters, operating costs, and potential drawbacks for each process treating the
example well. At this time, the DSS process using limestone or lime, and the hybrid calcium
chloride process look the most promising in terms of retained energy and chemical cost. Further
pilot testing will be required to determine process efficiencies and undesirable side effects, such
as scaling and residual impurities in the steam.

14017525 4-22
Table 4-2
Steam scrubbing process comparison for example well

14017525 4-23
14017525
5
SUMMARY
Research is ongoing to find cost-effective alloys for geothermal well casings. Although many
metals have been proven to offer adequate protection against corrosion, they remain impractical
for geothermal development because the cost of most materials is too high. Austenitic stainless
steels are not as corrosion resistant as the titanium (C-276) and nickel (625) alloys, but they are
more affordable. Titanium and high-nickel alloys are currently too expensive for most deep-
seated geothermal targets, and the price volatility of these specialty metals further complicates
project cost estimations. Although composite coatings, spray application, electrochemically
deposited, and explosion cladded materials have been successful in laboratory testing, there is
little published information on their use in large-scale projects at temperatures typical of high-
enthalpy geothermal systems. Furthermore, some coatings may require thicker applications to
resist corrosion in low-pH environments.
At many geothermal power stations, chemical injection to adjust fluid chemistry is the most cost-
effective solution to corrosion. Sodium hydroxide injections increase pH and can improve
operating conditions to reduce the immediate need for corrosion resistant alloys [69]. Chemical
injection, offering a lower cost but decreased corrosion resistance, is currently the best option—
until nickel and titanium costs moderate enough to justify manufacturing clad or solid piping.
Material and treatment technologies should be chosen on the basis of laboratory and/or field
testing at conditions specific to a given field, or even a given well site, if practical.

14017525 5-1
14017525
6
REFERENCES
1. Sanada, N. and Lichti, K.A. (1997). Prospects for the evaluation and development of
materials under EIA research collaboration program on deep geothermal resources, in
Proceedings, NEDO International Geothermal Symposium: Sendai, Japan, 192–199.
2. Kurata, Y., Sanada, N., Nanjo, H., and Ikeuchi, J. (1995). Casing Pipe Materials for Deep
Geothermal Wells, Geothermal Resource Transactions 19: 105–109.
3. Moya, P. and Sánchez, E. (2002). Neutralization System for Production Wells at the
Miravalles Geothermal Field. Geothermal Resources Council Transactions, Vol. 26,
September 22–25, 2002, 667–672.
4. Allegrini, G. and Benvenuti, G. (1970). Corrosion Characteristics and Geothermal Power
Plant Protection, U.N. Symposium on the Development and Utilization of Geothermal
Resources, Geothermics, Vol. 2, Part 1.
5. Ellis, P.F. and Anliker, D.M. (1982). Geothermal Power Plant Corrosion Experience - A
Global Survey, Materials Performance, Volume 21, No. 2, February.
6. Ármannsson, H. (2010). The Chemistry of the Krafla Geothermal System in Relation to the
IDDP Well, Proceedings World Geothermal Congress 2010 Bali, Indonesia, 25–29 April
2010.
7. Einarsson, K. and Pálsson, B. (2010). Acid Wells in the Krafla Geothermal Field,
Proceedings World Geothermal Congress 2010 Bali, Indonesia, 25–29 April 2010.
8. Andersen, G. (1989). A Thermodynamic Model for Predicting HCl Partial Pressure Above a
High Temperature Hypersaline Geothermal Fluid, Geothermal Resources Council,
Transactions, Vol. 13, Oct. 1989, 221–225.
9. Simonson, J.M. and Palmer, D.A. (1993). Distribution of Ammonium Salts Between Liquid
and Vapor at High Temperatures, Proceedings of the International Conference on Interaction
of Iron-Based Materials with Water and Steam, 1993.
10. D'Amore, F., Truesdell, A.H., and Haizlip, J.R. (1990). Production of HCl by Mineral
Reactions in High- Temperature Geothermal Systems, Stanford Research Engineering
Workshop, Jan. 1990.
11. Hirtz, P.N. (1992). Results of HCl Measurements at Coso. Thermochem Internal reports,
1992, Thermochem Inc., 3414 Regional Parkway, Suite A, Santa Rosa, CA 95403.
12. Simonson, J.M. and Palmer, D.A. (1993). Liquid vapor partitioning of HCl (aq) to 350 C,
Geochim. Cosmochim. Acta, 57, 1–7.
13. Truesdell, A.H and Haizlip, J.R. (1989). Hydrogen Chloride in Superheated Steam and
Chloride in Deep Brine at The Geysers Geothermal Field, California, Proceedings, 13th
Workshop on Geothermal Reservoir Eng., Stanford, CA., Jan. 1989, 93–100
14. Gunderson, R.P. (1989). Distribution of Oxygen Isotopes and Noncondensable Gas in Steam
at The Geysers. Geothermal Resources Council Transactions, Vol. 13, 449–454.

14017525 6-1
15. Beall, J.J., Wright, M.C., and Hulen, J.B. (2007). Pre and Post-Development Influences on
Fieldwide Geysers NCG Concentrations. Geothermal Resources Council Transactions,
Vol..31, 427–434.
16. Hirtz, P.N., Buck, C., and Kunzman, R. (1991). Current Techniques in Acid-Chloride
Corrosion Control and Monitoring at The Geysers, Proceedings, Sixteenth Workshop on
Geothermal Reservoir Engineering, Stanford Geothermal Program, Stanford University, 83–
95.
17. Haizlip, J.R. and Truesdell, A.H. (1992). Noncondensable Gas and Chloride Are Correlated
in Steam at The Geysers. Monograph on The Geysers Geothermal Field (C. Stone, editor),
Geothermal Resources Council Special Report 17, 139–143.
18. Walters, M.A., Sternfeld, J.N., Haizlip, J.R., Drenick, A.F., and Combs, J. (1992). A Vapor-
Dominated High-Temperature Reservoir at The Geysers California. Monograph on The
Geysers Geothermal Field (C. Stone, editor), Geothermal Resources Council Special Report
17, 77–87.
19. White, D.E. and Waring, G.A. (1963). Volcanic Emanations, Data of Geochemistry (M.
Fleischer, editor), USGS Professional Paper 440-K. p. 29.
20. Beall, J.J., Walters, M., and Thermochem Inc., (2009). Geothermal Permeability
Enhancement, Final Report, U.S. Department of Energy, EGS Award: DE-FC07-02ID14405;
Amendment No. M006.
21. Hirtz, P.N. and Beall, J.J. (2013). Injection Study for Superheat Management and HCl
Mitigation of the Krafla High Temperature Reservoir, Thermochem Internal Report, 2013.
Thermochem Inc., 3414 Regional Parkway, Suite A, Santa Rosa, CA 95403.
22. Wilson, P.T. and Lichti, K.A. (1982). Assessment of Corrosion Performance of Construction
Materials in Geothermal Steam, in Proceedings, Pacific Geothermal Conference/New
Zealand Geothermal Workshop, Vol. 1 University of Auckland Geothermal Institute, 185–
190.
23. Syrett, B.C., MacDonald, D.D., and Shih, H. (1980). Pitting Resistance of Engineering
Materials in Geothermal Brines–I. Low Salinity Brine, Corrosion, Vol. 36, Issue 3, 130–139.
24. Kimbara, K., Muraoka, H., Kobayashi, H., Sanada, N., Fujimoto, K., and Ohsato, K. (2000).
A Summary of Results of the IEA Task Activities of Deep Geothermal Resources, in
Proceedings, World Geothermal Congress: Kyushu-Tohoku, Japan, 2653–2658
25. Sanada, N., Kurata, Y., Hiroshi, N., Kim, H., Ikeuchi, J., and Lichti, K.A. (2000). IEA Deep
Geothermal Resources Subtask C: Materials Progress with a Database for Materials
Performance in Deep and Acidic Geothermal Wells, in Proceedings, World Geothermal
Congress: Kyushu-Tohoku, Japan, 2411–2416.
26. Gazo, F.M. and Datuin, R. (1990). Geothermal Power Plant Operations in the Philippines:
Material Problems and Preventive Measures, Geothermal Resources Council Transactions,
Vol. 14, Part II, 1009–1014.
27. Povarov, O.A., Tomarov, G.V., and Semenov, V.N. (2000). Physical and Chemical Processes
of Geothermal Fluid Impact on Metal of Geothermal Power Plant Equipment, in Proceedings,
World Geothermal Congress: Kyushu-Tohoku, Japan, 3277–3282.
28. Lichti, K.A. (2007). Forgotten Phenomenon of Materials Selection and Use in Geothermal
Energy Applications, in Materials Issues Governing the Performance of Advanced 21st

14017525 6-2
 Century Energy Systems, Edited by Tana Levi, Jim Norton, Nigel Simms, lan Wright and,
John Nicholls, Wellington, New Zealand, February-March 2006.
29. Wells, D.B. and Lichti, K.A. (1985). A Corrosometer Study of Corrosion in Rotorua
Geothermal Fluids, in Proceedings, Seventh New Zealand Geothermal Workshop: Auckland,
New Zealand, 91–96.
30. Lichti, K.A., Johnson, C.A., McIlhone, P.G.H., and Wilson, P.T. (1995). Corrosion of Iron-
Nickel Base Titanium Alloys in Aerated Geothermal Fluids, in Proceedings, World
Geothermal Congress: Florence, Italy, 2375–2380.
31. Amend, B. (2009). Considerations in the Testing and Selection of Materials for Corrosive
Geothermal Environments, Geothermal Resources Council Transactions, Vol. 33, 633–637.
32. Amend, B. (2010). Welding Technology Solutions to Geothermal Energy Production
Challenges, Geothermal Resources Council Transactions, Vol. 34, 985–990.
33. Wagg, B., Slack, M., Kaiser, T., and DeGeer, D. (1997). Identification of Casing String
Failure Modes Using Caliper Data, Geothermal Resources Council Transactions, Vol. 21,
129–133.
34. Southon, J.N.A. (2005). Geothermal Well Design, Construction and Failures, in Proceedings,
World Geothermal Congress: Antalya, Turkey, 1–6.
35. McWright, R.D. (1980). Corrosion Behavior of Materials Exposed to Hypersaline
Geothermal Brine, Livermore California: Lawrence Livermore National Laboratory, 1–13.
36. De Waard, C. and Milliams, D.E. (1975). Carbonic Acid Corrosion of Steel, Corrosion, Vol.
31, 177–181.
37. Kurata, Y., Sanada, N., and Nanjo, H. (1995). Material Damage in a Volcanic Environment,
in Proceedings, World Geothermal Congress: Auckland, New Zealand, 2409–2414.
38. Eksari, N. and Marbun, B.T.H. (2015). Integrated Analysis of Optimizing Casing Materials
Selection of Geothermal Well by Using a Model for Calculating Corrosion Rates, in
Proceedings, World Geothermal Congress: Melbourne, Australia, 1–14.
39. Mandzhgaladze, S.N., Tavadze, F.N., and Knutsariya, E.M. (1978). New Corrosion-Resistant
Steels and Metallic Coatings for Equipment at Geothermal Electric Power Stations, in
Proceedings, All-Union Scientific and Technical Conference on the Use of the Earth’s Heat
for the Productions of Electric Power: Moscow, Russia, 73–74.
40. Swann, S.J., White, S.P., Sanada, N., Kurata, Y., Nanjo, H., Ikeuchi, J., and Christenson,
B.W. (1997). Corrosion in Volcanic Gases, Proceedings NEDO International Geothermal
Symposium, Sendai, Japan, NEDO, 154–163.
41. JFE Steel (2015). Corrosion Resistant Clad Steels, JFE Technical Report No.20 (Mar. 2015),
New Products and Technologies. Available at www.jfe-steel.co.jp
42. Blakey, M. and Olivas, J. (2010). Explosion Welded Materials for Geothermal Applications,
Geothermal Resources Council Transactions, Vol. 34, 991–994.
43. Justus, S.B., Stulga, J.E., and Brosius, K.S. (1986). Severe Service Metallic Clad Casings and
Wellhead Components, Georthermal Research Council Transactions, Vol. 10, 61–65.
44. Kurata ,Y., Sanada, N., Nanjo, H., and Ikeuchi, J. (1992). Material Damages in Geothermal
Power Plants, in Proceedings, 14th New Zealand Geothermal Workshop: New Zealand, 149–
164.

14017525 6-3
45. Lichti, K.A. (1995). Selection and Use of Materials for Geothermal Energy Applications
Using Computer Based Technology, in Proceedings, World Geothermal Congress: Florence,
Italy, 2981–2986.
46. Sanada, N., Kurata, Y., Nanjo, H., and Ikeuchi, J. (1995). Material Damage in High Velocity
Acidic Fluids, Geothermal Resources Council Transactions Vol. 19, 359–363.
47. Iwanaga, T., Watanabe, E., Hirowatari, K., Yamauchi, M., Itoi, R., Eddy, S.R., and
Alejandro, R.B. (2005). Field Experiment of pH Adjustment for Acid Geothermal Fluids in
Miravalles, Costa Rica, Geothermal Resources Council Transactions, Vol. 29, 693–696.
48. Scheiber, J., Ravier, G., Sontot, O., Hensch, C., and Genter, A. (2013). In Situ Material
Studies at the High Temperature Skid (HTS) Bypass Systems of the Geothermal Power Plant
in Soultz-sous-Forêts, France, in Proceedings, Thirty-Eighth Workshop on Geothermal
Reservoir Engineering: Stanford University, Stanford, California, 1–9.
49. Sugama, T., Butcher, T., and Ecker, L. (2011). Experience with the Development of
Advanced Materials for Geothermal Systems, Materials Challenges in Alternative and
Renewable Energy, 389–401.
50. Sugama, T. (2006). Hydrothermally Self-Advancing Hybrid Coatings for Mitigating
Corrosion of Carbon Steel. Brookhaven National Laboratory, Upton, NY.
51. Amend, B. and Yee, J. (2013). Selective Application of Corrosion Resistant Alloys Mitigates
Corrosion in pH-Modified Geothermal Fluids, NACE International, Houston, TX, Paper No.
C2013-0002416.
52. Lambrych, K., Johnson, T., and Kelley, D. (2015). Performance and Cost Comparison of
Materials Used In Design and Fabrication of Hydrometallurgical Equipment, web accessed
12/4/15:
http://www.ashland.com/Ashland/Static/Documents/APM/Material_Choice_for_Design_CO
M2015.pdf
53. Personal communication on November 17, 2015 with Robert McGehee of Schulz Xtruded
Products, Schulz USA.
54. Houser, R. (2010). Specialty Metal Solutions for Corrosive Geothermal Application,
Geothermal Resources Council Transactions, Vol. 34, 1045–1050.
55. Mundhenk, N., Huttenloch, P., Kohl, T., Steger, H., and Zorn, R. (2012). Laboratory and In-
Situ Corrosion Studies in Geothermal Environments, Geothermal Resources Council
Transactions, Vol. 36, 1101–1106.
56. Love, W., Cron, C., and Holligan, D. (1988). The Use of Beta-C Titanium for Downhole
Production Casing in Geothermal Wells, Geothermal Resources Council Transactions, Vol.
12, 49–53.
57. Hirtz, P., Miller, J., and Prabhu, E. (1990). Operational Results of a Dry-Steam Resource
Chloride Corrosion Mitigation System, Geothermal Resource Council Annual Meeting,
Transactions, 1990.
58. Hjartarson, S., Sævarsdóttir, G., Ingason, K., Pálsson, B., Harvey, W.S., and Pálsson, H.
(2014). Utilization of the Chloride Bearing, Superheated Steam from IDDP-1, Geothermics
49, 83–89.

14017525 6-4
59. Hirtz, P.N. (1992). Results of HCl Measurements on Down-hole Mitigation Wells.
Thermochem Internal reports, 1992. Thermochem Inc., 3414 Regional Parkway, Suite A,
Santa Rosa, CA 95403.
60. Fisher, D.W. and Jung, D.B. (1996). Alternatives to Traditional Water Washing Used to
Remove Impurities in Superheated Geothermal Steam. Geothermal Research Council
Transactions, Vol. 20, 737–741
61. Gallup, D.L. and Farison, J.R. (1996). Alternative Steam Scrubbing Process. UNOCAL
internal report, February 29, 1996.
62. Hirtz, P.N., Broaddus, M.L., and Gallup, D.L. (2002). Dry Steam Scrubbing for Impurity
Removal from Superheated Geothermal Steam; Geothermal Resource Council Annual
Meeting, Transactions, 2002.
63. Weres, O. and Kendrick, C. (2010). Corrosion by HCI in Dry Steam Wells Controlled using
Potassium Carbonate without Destroying Superheat, Geothermal Resources Council
Transactions, 2010.
64. Gallup, D.L. and Farison, J.R. (1998). Materials Testing in Corrosive Wells at the Geysers
Geothermal Field, in Proceeding of the 20th New Zealand Geothermal Workshop, 331–339.
65. Ergun, S. (1952). Fluid Flow through Packed Columns, Chemical Engineering Progress,
Vol. 48, No. 2, 89.
66. Bracaloni, M., Culivicchhi, G., and Fornari, B. (1995). Erosion and Corrosion Problems
Experienced during the Operation of Geothermal Turbines in Italy, in Proceedings World
Geothermal Congress 1995, Vol. 4, 2427–2432.
67. Hirtz, P.N. (2001). Alternatives to Traditional Steam Washing for Removal of Volatile
Chloride from Superheated Geothermal Steam, Final Report to the US Department of
Energy, Geothermal Power Initiative Grant No. DE-FG07-00ID13915, July 2001.
68. Ellis, A.J. and Mahon, W.A.J. (1977). Chemistry and Geothermal Systems, Department of
Scientific and Industrial Research. New Zealand. Academic Press.
69. Tamboboy, R.J.T., Aragon, G.M, and Reed, M.H. (2015). Evaluation of NaOH Injection in
the Neutralization of Highly Acidic Cl-SO4 and SiO2-Saturated Geothermal Fluids, in
Proceedings, World Geothermal Congress, Melbourne, Australia, 1–8.

14017525 6-5
14017525
14017525
 The Electric Power Research Institute, Inc.
(EPRI, www.epri.com) conducts research and
development relating to the generation, delivery and
use of electricity for the benefit of the public. An
independent, nonprofit organization, EPRI brings
together its scientists and engineers as well as
experts from academia and industry to help address
challenges in electricity, including reliability,
efficiency, affordability, health, safety and the
environment. EPRI members represent 90% of the
electric utility revenue in the United States with
international participation in 35 countries. EPRI’s
principal offices and laboratories are located in Palo
Alto, Calif.; Charlotte, N.C.; Knoxville, Tenn.; and
Lenox, Mass.

Together…Shaping the Future of Electricity

© 2016 Electric Power Research Institute (EPRI), Inc. All rights reserved.
Electric Power Research Institute, EPRI, and TOGETHER…SHAPING THE
FUTURE OF ELECTRICITY are registered service marks of the Electric
Power Research Institute, Inc.
3002007966

Electric Power Research Institute

3420 Hillview Avenue, Palo Alto, California 94304-1338 • PO Box 10412, Palo Alto, California 94303-0813 • USA
14017525800.313.3774 • 650.855.2121 • [email protected] • www.epri.com