You are on page 1of 8

Geothermics 39 (2010) 321–328

Contents lists available at ScienceDirect

Geothermics
journal homepage: www.elsevier.com/locate/geothermics

Performance history of The Geysers steam field, California, USA


K.P. Goyal ∗ , T.T. Conant
Calpine Corporation, 10350 Socrates Mine Road, Middletown, CA 95461, USA

a r t i c l e i n f o a b s t r a c t

Article history: The performance of Calpine’s Geysers steam field from startup in 1960 to 2008 is described in this paper.
Received 21 May 2009 Since October 2003, Calpine has received approximately 482 L/s of tertiary-treated reclaimed water from
Accepted 14 September 2010 the City of Santa Rosa. To accommodate and derive benefit from this water, Calpine has converted 20
wells (ten producers, six shut-in, two observation, and two suspended wells) to high-rate injection ser-
vice. Additional nine wells were also converted to low-rate injectors that receive 12.6 L/s or less. Annual
Keywords:
recovery factors (i.e., fieldwide increase in annual steam production divided by annual injection) for the
Geothermal
first 5 years of Santa Rosa Geysers Recharge Project (SRGRP) operation have been estimated at 17.6%,
Geysers
Geothermal power
26.1%, 37.1%, 39% and 44.6%, respectively; reasonably close to or slightly higher than the values, pre-
Jack shaft dicted prior to SRGRP startup. Using a revised definition that includes the amount of un-boiled water in
Injection recovery factor the reservoir, the annual recovery factors turn out to be 17.6%, 16.1%, 14.6%, 12.4% and 12.2% from year
Performance one through year five. Improvements in the wellfield, water injection, and power plant modifications
Injection from January 1995 through December 2008 are also discussed in this paper.
Puff © 2010 Elsevier Ltd. All rights reserved.
Production
Decline rate
SRGRP
SEGEP
Wellfield

1. Introduction sition of Geysers Geothermal Company from Freeport McMoRan in


1990, Calpine expanded its operations in The Geysers, and by 1999
The Geysers geothermal field is located in northern California became the largest operator of the steam field and power plants.
(Fig. 1). It began commercial steam production in September 1960 Currently, Calpine and NCPA operate The Geysers field with a total
with the start-up of Unit 1, an 11-MWnet power plant. The installed installed capacity of 1541 MWnet . Calpine owns and operates 15
capacity increased to a maximum of approximately 2000 MWnet power plants with an installed capacity of 1302 MWnet . The remain-
by mid-1989 (Goyal and Box, 1991). Since then, several units have ing 239 MWnet capacity comes from two power plants owned and
been retired due to age and/or insufficient steam supply. These operated by NCPA. The power generated by the 15 Calpine plants is
include Unit 15 (57 MWnet ) in 1989, Bottle Rock (52 MWnet ) in 1990, mostly sold under bilateral contracts to the California Independent
Unit 1 (11 MWnet ) in 1990, Units 2–4 (67 MWnet ) in 1992, and CCPA System Operator (ISO) system. Through June 2008, Calpine sold
(Central California Power Agency) Units 1 and 2 (124 MWnet ) in electricity to PG&E under standard offer (SO) contracts from four
1996. Unit 10 (53 MWnet ) and Unit 9 (53 MWnet ) have been kept plants with name plate rating of 80 MW or less. These plants were
in a suspended status since 2000 and 2001, respectively. During designated as “SO” units, and the remaining 11 plants were desig-
the height of The Geysers development in mid 1980s, several com- nated as “market” units. This plant designation was removed in July
panies were involved with the development of the steam field. 2008, when all plants started selling power under market contracts.
These included Unocal, Magma, Thermal Power Company, GEO Since this is a recent change, the old terminology of “market” and
Operator, MCR Geothermal, Geysers Geothermal Company, NCPA “SO” units is retained in this paper. Recently, the U.S. Renewables
(Northern California Power Agency), Occidental Geothermal, DWR Group has been redeveloping the Bottle Rock steam field to sup-
(Department of Water Resources), CCPA and Mission Energy. Util- ply steam to a mothballed 52 MWnet unit located in the northern
ities included PG&E (Pacific Gas and Electric), SMUD (Sacramento end of the field (Snedaker, 2007). This plant is currently generat-
Municipal Utility District), DWR, CCPA, and NCPA. After an acqui- ing 10–12 MWnet . Western Geopower Corporation also announced
plans to redevelop the abandoned Unit 15 lease area, located on the
western edge of the field, and build a new 25 MWnet power plant
∗ Corresponding author. Fax: +1 707 431 6148. (Sanyal et al., 2007). This company was acquired by Ram Power in
E-mail address: keshav@calpine.com (K.P. Goyal). October 2009. At The Geysers, the operators assign lease areas to a

0375-6505/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geothermics.2010.09.007
322 K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328

2. Brief historical background


Nomenclature
The Geysers steam production and injection history from 1965
IDS injection derived steam to 2008 is presented in Fig. 2 (CDOGGR, 2008). The amount of steam
MW megawatt supplied to Units 1 and 2 from 1960 to 1964 was small and there-
NCG non-condensable gas fore, is not shown. Water injection in the field did not start until
RF recovery factor, “original annual RF” 1969. The steam production rose from 70 kg/s in 1965, to a peak of
3546 kg/s in 1987. In 1965, only two PG&E units were in operation,
Subscripts
generating a total of 24 MWnet . The development of the field con-
net after subtracting parasitic load
tinued until mid-1989, when the total installed capacity reached
2000 MWnet (Goyal and Box, 1991). The eastern part of the field
was mostly developed from 1980 to 1988. This pace of develop-
ment led to a pressure drop in the reservoir and high annual steam
decline rates in excess of 25% in some areas (Goyal and Box, 1991;
power plant based on its size. In this paper, a lease area assigned to Sanyal, 2000). The decline in generation at The Geysers caught
a power plant is also called a “wellfield”. the attention of the California Energy Commission (Sanyal, 2000)
The Geysers field has produced a total of 2398 billion kg of steam and resulted in the formation of a Technical Advisory Committee
from 1960 to 2008. For a steam usage factor of 2.4 kg/s per MW, this (TAC), a consortium of steam producers, utilities, researchers, and
production is equivalent to an electric generation of 278 million consultants. Funding for the work was provided by steam produc-
MWh. Using a conversion factor of 1.5 bbl of oil per MWh, The ers, utilities, federal, state, and local governments (County of Lake,
Geysers has eliminated the need to burn 417 million barrels of oil 1991). The main objectives of TAC were to find ways to (i) reduce
for power generation by year-end 2008. decline rates and (ii) increase the life of the field. TAC efforts led
Information about The Geysers’ performance has been pub- to the development of conceptual and numerical models for the
lished by various operators, and an industry-wide comprehensive entire Geysers field, as well as identification of tracers suitable for
report was published in early 1995 through the Geysers Geother- vapor dominated systems. Additionally, the TAC report suggested
mal Association (GGA, 1995). This paper presents the performance bringing water from outside sources for injection and finding ways
of the portion of The Geysers field operated by Calpine. After acquir- to improve plant efficiency (Sanyal, 2000). The steam production
ing steam fields from Unocal and power plants from PG&E in 1999, data from 1995 to 2008 in Fig. 2 show that the steam decline rate
Calpine invested significant capital to extend the life of this valu- has been substantially reduced, and the field continues to produce a
able renewable asset by improving plant efficiency, developing significant amount of steam. In a nutshell, both of the CEC objectives
untapped reserves and extracting additional heat energy through have been met.
water injection. The steam fields and power plants management Though steam production from The Geysers started in 1960,
strategies and the resulting improvements in steam production injection did not start until 1969, when steam condensate was
are described for the period of 1995–2008, with emphasis on the returned to the reservoir through injection wells. Between 1960
injection performance of SRGRP which began in October 2003. Also, and 1969, the steam condensate from the cooling tower was dis-
definitions and methodology for estimating injection recovery fac- posed in nearby surface drainages. Initially, injection wells were
tors are presented and recovery factors during the first 5 years of sited as far as possible from active production wells (Barker et al.,
SRGRP operation are discussed. 1995) to avoid the possibility of water breakthrough. As reservoir

Fig. 1. Currently, operating plant locations in The Geysers geothermal field, California. Inset map shows the location of The Geysers geothermal field. PP: power plant.
K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328 323

Fig. 2. Steam production and injection at The Geysers. Data from California Division of Oil and Gas and Geothermal Resources.

pressure declined and superheated conditions developed, signifi- spaces (rock matrix) boils, and steam moves into larger aperture
cant steam gains were observed in production wells located near fractures, driven by pressure gradients within the reservoir.
injection wells. Consequently, every effort was made to increase
injection into the reservoir. Mass replacement through steam con- 3. Steam production and generation
densate injection rose from 6% in 1969 to 24% in 1979 (Fig. 2).
Since 1980, condensate injection has been supplemented with fresh Fig. 3 shows year-end steam production from 1994 through
water from Big Sulphur Creek (BSC) and from small stream water 2008 of the wellfields associated with market plants (19 con-
extraction facilities operated seasonally during periods of high run tiguous wellfields supplying steam to 11 power plants), with and
off. The creek water extraction increased field wide mass replace- without makeup drilling and injection. Calpine has been receiv-
ment from 24% to over 30% in 1994, as shown in Fig. 2. ing about 227 L/s of water from the SEGEP project for injection in
Being a vapor-dominated system with rather low porosity, The the southeast portion of the field since September 1997. By 2008,
Geysers has more than 90% of its heat stored in the rocks. The the southeast portion of the reservoir became saturated because
commercially productive area, as delineated by drilling, is approxi- of water injection, and as a result injection had to be reduced. The
mately 78 km2 . There is little or no natural recharge to the reservoir SRGRP started supplying water to The Geysers in October 2003.
(Isherwood, 1977). This fact became quite clear during late 1980s, Steam gains from makeup wells (new steam producers) that were
when the drop in the reservoir pressure was accompanied by an brought on line between 2001 and 2008, and the injection from
increase in enthalpy and superheat of the produced steam through- SEGEP and SRGRP, are shown in Fig. 3. The year-end steam produc-
out the field. To recover some of the heat stored in the rocks, the tion rate remained high during 2003–2008 (Fig. 3), suggesting little
operators sought external water sources to supplement injection decline in the field. The dashed line in Fig. 3 displays the extrapo-
at The Geysers. The first project, called SEGEP (Southeast Geysers lated steam production from the original wells if the SEGEP, SRGRP,
Effluent Pipeline), started in September 1997. It brought 341 L/s of and makeup well projects had not occurred. The SEGEP injection
lake water and secondary treated water from the treatment plants benefits for Calpine are estimated as two thirds of the combined
of LACOSAN (Lake County Sanitation District) to Calpine and NCPA. MW gain at NCPA and Calpine plants. The MW gains are converted
This additional water increased the field-wide mass replacement to flow rate using an average conversion rate (steam rate per MW)
from over 30% to almost 50% (Fig. 2). Another injection project, for a given year. The power plant activities can skew generation
called SRGRP (Santa Rosa Geysers Recharge Project), started in numbers and therefore, the gain attributed to SEGEP. However, the
October 2003, bringing in 482 L/s of tertiary treated water from error caused by this assumption is estimated to be too small (less
the treatment plant of the city of Santa Rosa. This additional water than 1%) to affect our conclusions. A gain of 397 kg/s in steam pro-
further increased mass replacement to approximately 85% in 2006 duction was realized by December 31, 2008, from makeup wells
(Fig. 2). A dry winter in 2007 and 2008 led to a reduction in mass and injection as shown in Fig. 3.
replacement to about 75%.
Cumulatively, The Geysers reservoir had produced 2398 billion 3.1. Wellfield update
kg of steam and received 946 billion kg of injection water by Decem-
ber 2008 (CDOGGR, 2008). This represents a cumulative net mass Calpine brought seven additional production wells on-line dur-
withdrawal of 1452 billion kg and a cumulative mass replacement ing 2001 and 2002 providing 60 kg/s of steam. Out of these, four
of 39.5% by 2008. For a 78 km2 extent and 5% porosity, steam pro- were flowing in 2008. One well was converted to injection to
duction of 1452 billion kg would require a reservoir thickness of replace an existing failed injector, one stopped flowing, and the
21 km to store it all in the vapor phase. This thickness is clearly other well was shut-in because of corrosion. One new makeup well
unreasonable, suggesting initial reserves were mostly in the liquid was brought on-line in February 2004 at 15 kg/s. A second leg was
phase, although no water table has been found in The Geysers so far, drilled into an existing producer in 2008 which produced a total
even in wells deeper than 3.66 km. A possible explanation (Pruess of 5 kg/s of steam − almost twice that of the original leg. These six
and Narasimhan, 1982) is that liquid water trapped in tight pore wells provided approximately 45 kg/s of steam or about 20 MWnet
324 K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328

Fig. 3. Steam production from Calpine market units.

in 2008. Except Aidlin, Bear Canyon, and West Ford flat, the steam run off. Three pump stations are used to extract water from Big Sul-
field is now fully interconnected through the addition of numer- phur Creek. Pump station BSC#1 started in 1980 and has a pumping
ous new pipelines (crossties). New crossties between the fields for capacity of 442 L/s. BSC#2 started in 1983 and has the capability to
Units 9/10 and Sonoma, Sonoma and Calistoga, Sonoma and Unit pump 568 L/s. BSC#3 started pumping water in 1994 and can pump
20, and Unit 13 and Unit 18, were installed after the 1999 acquisi- up to 95 L/s.
tion by Calpine. Units 9 and 10 have been mothballed since April The SEGEP project has provided approximately 341 L/s of sec-
2001, due to their poor efficiency and high operations and mainte- ondarily treated water and lake water to Calpine and NCPA since
nance costs. The steam from these areas is flowing to nearby plants September 1997. Calpine injects 227 L/s (two thirds) and the rest is
through crosstie pipelines. utilized by NCPA. The capacity of the SEGEP system was increased
Additional injectors were added after the startup of SRGRP to 448 L/s in early 2009. Out of this total delivery, Calpine is eli-
in October 2003. A total of 20 wells (ten-producers, six shut-in, gible to receive 259 L/s. The SRGRP has provided tertiary treated
two observation, and two suspended) were converted to high- wastewater from the Santa Rosa regional sewage treatment plant to
rate (31 L/s or more) injectors between 2003 and 2008, to accept The Geysers through a 66-km, 76–122 cm diameter pipeline since
increased injection water and to replace failed injectors. Injection October 2003. Per revised contract, the city is obligated to sup-
wells fail because of downhole problems such as collapsed casing, ply an annual average of 553 L/s. The city pumps water from a
water breakthrough to nearby producers, formation sloughing, and treatment plant to the base of the Mayacamas Mountains using
corrosion of the casing and/or liner. Injection wells in The Geysers two 746 kW pumps to overcome an elevation of about 91 m. Three
accept water under vacuum. Additionally, nine wells (three produc- pump stations, each housing five pumps (three constant speed and
ers, five shut-in, and one observation) were converted to low-rate two variable speed) of 746 kW each, provide a total vertical lift of
injectors during 2007 and 2008. These injection wells are com- approximately 914 m to a storage tank located near The Geysers.
pleted without any liner and most are receiving less than 12.6 L/s Each pump is capable of pumping 175–185 L/s. Calpine then pumps
of water. All the injection wells were selected to distribute water this water to its storage tank that is 168 m higher through a pump
over a large area to achieve efficient boiling and maximum heat station which houses five 447 kW pumps. Calpine distributes this
recovery. water to 8–12 injection wells through 29-km of 20- to 76-cm diam-
A portion of The Geysers field associated with the 115 kV trans- eter pipeline network. The SRGRP pipeline is designed to handle
mission line was curtailed from November 6, 2004, to April 8, 2005, approximately 876 L/s of the tertiary treated wastewater to allow
and again from January 31, 2006 to February 26, 2006, due to work for additional water delivered to The Geysers. During the winter
on the Cortina substation. Some curtailments, due to the availabil- months, four out of five pumps, installed at each station, are needed
ity of inexpensive hydro-power, were also experienced during the to pump approximately 740 L/s of treated effluent to The Geysers.
spring of 2006. Only two to three pumps are needed to pump water to The Geysers
Calpine monitors groups of production wells surrounding injec- for the remainder of the year. The benefits from SRGRP include (i)
tion wells on a routine basis to identify water breakthrough and reduced wastewater discharge into the Russian River by approxi-
possible wellbore problems that may develop. Wells that indicate mately 7.6 billion liters per year, (ii) production of renewable green
lost production from scale deposits or bridging in the wellbore are energy of approximately 85 MW at The Geysers, and (iii) water
routinely cleaned out to recover the lost steam production, using reuse for irrigation purposes would conserve Russian River water
surging, venting, and bridge-busting techniques. (City of Santa Rosa, 2003).
Since the startup of the SRGRP in October 2003, Calpine’s annual
3.2. Injection update average total injection from all sources has ranged from 1135 to
1451 L/s depending upon the weather and the reservoir manage-
Injection in The Geysers started in 1969, when steam conden- ment considerations. In 2008, Calpine injected approximately 37.9
sate was returned back to the reservoir through injection wells. billion liters of water into the reservoir. This volume of water
Since 1980, injection has been supplemented with fresh water from represented a mass replacement of 73.5%. This injected water con-
Big Sulphur Creek (BSC) and small streams during periods of high tained 34% steam condensate, 18% SEGEP, 43% SRGRP, and 5% creek
K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328 325

Fig. 4. Effect of SRGRP injection on steam production of all Calpine wells in The Geysers.

water. The benefits from the SEGEP and SRGRP have been estimated Calpine continues to carry out optimization studies aimed at
at approximately 36 MWnet (88 kg/s) and 105 MWnet (252 kg/s) conserving the resource while maximizing plant output. Calpine
of additional steam by year-end 2008 in the Calpine area. These uses an integrated reservoir, pipeline, and power plant numerical
estimates are based on actual production or generation and the model to investigate the long-term benefit of major plant improve-
extrapolation of pre-project decline trends. A detailed summary of ments, injection projects, additional pipelines, and production of
The Geysers injection experience is available in Goyal (1999). steam from untapped areas.

3.3. Power plant update 4. SRGRP injection recovery factors

Because of decline in the reservoir pressure and steam deliv- Stark et al. (2005) presented forecasts and early results of SRGRP,
erability at The Geysers, the power plants are not running at full concluding that the project was providing increased steam flow
capacity, and thereby are less efficient than when operated at and generation approximately in line with the expectations. Fig. 4
the design capacity. Several power plants have been modified to shows total steam flow of all Calpine wells from January 2000 to
improve their efficiency at a reduced steam flow. For example, sin- February 2009 that were flowing before the start-up of SRGRP in
gle rotor (jack shafted) Units 12 and 17 were each replaced by two October 2003. Additional steam, produced by activities, such as new
tighter rotors (with a smaller steam flow area compared to the wells brought on-line or increased flow due to improved gathering
original design) during 2000 and 2001. This was done to accom- system facilities, has been subtracted from the total steam. In other
modate steam from the closure of Units 9 and 10. Unit 20 also words, in Fig. 4 the steam production shown before and after SRGRP,
received two rotors in 1999 similar to those in Units 12 and 17. is under similar conditions, so as to reflect the effect of SRGRP.
Units 11 and 14 each received new tighter and more efficient rotors The production data without SRGRP is fitted with a 4.5% harmonic
in 2002, designed for prevailing reservoir pressure and steam flow decline rate starting in November 2000, and represented by a solid
conditions. line in Fig. 4. This decline rate is consistent with the 6% harmonic
The gas ejector capacity of several power plants was also opti- decline rate in January 1995 estimated for Calpine’s market units
mized to handle larger amounts of non-condensable gases (NCGs), (Goyal and Box, 2004). Areas between actual steam production data
as well as to remove NCGs efficiently by increasing the number and the extrapolated curve are then calculated for each year from
of stages. The efficient removal of NCGs reduces backpressure in November 2003 to October 2008. These areas represent mass (kg)
the condenser that, in turn, increases power generation. All these of additional steam recovered from SRGRP injection. Similarly, the
changes reduced the steam usage from 2.48 kg/s per MW in 1998 amount of water injected during these annual intervals is converted
to 2.39 kg/s per MW in 2008. For a steam supply of 1663 kg/s, a net into mass of water. The ratio of the additional steam amount, pro-
gain of approximately 26 MW was obtained from improvements in vided by injection, to the amount of water injected over the same
efficiency. period of time is defined as the “recovery factor” (Goyal, 1995).
The Aidlin power plant underwent a condenser peroxide wash This recovery factor (RF) is also called the “original annual RF” in
in 2007, and cooling tower fill replacement in 2008. Chemical opti- this paper to differentiate it from the “revised annual RF” discussed
mization at the Ridgeline plant reduced caustic consumption by below.
50% in 2007. A turbine acid wash at Big Geysers in 2007 improved The SRGRP RFs achieved for the first 5 years are shown in Fig. 5.
plant efficiency, as well as reduced turbine inlet pressure, resulting An original annual RF of 17.6% was realized in the first year, 26.1% in
in a net increase in generation of approximately 10 MW. Regular the second year, 37.1% in the third year, 39% in the fourth year, and
maintenance and overhauls at a rate of two to three per year keep 44.6% in the fifth year. These numbers are close to or slightly higher
the average steam usage factor essentially unchanged, i.e., from than the forecasted RF of 13.6%, 27.2%, 43.1%, 42.4%, and 41.8%,
2.37 kg/s per MW in 2002, to 2.39 kg/s per MW in 2008. respectively. Forecasted SRGRP recovery factors were based on a
326 K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328

Fig. 5. SRGRP recovery factors for first 5 years.

combination of geochemical data (primarily stable isotopes in pro- period to obtain cumulative RF. Thus, after 5 years, increased steam
duced steam), tracer studies, and numerical model results (Barker production has amounted to 33.2% by mass of the SRGRP water
et al., 1995; Goyal, 1999). Calculated injection derived steam (IDS) injected. This suggests that one third of the injected SRGRP water
using tracer tests was found to be higher than that calculated by has already been produced as steam, and the remaining two thirds
using a decline curve method (Goyal, 1999). This discrepancy can is still in the reservoir.
be explained by the fact the IDS may have displaced some normal Based on the regression analysis of the data in Fig. 5, the fol-
reservoir steam that would otherwise have been produced by the lowing relations can be used to calculate original annual RF and
wells in its area of influence. Additionally, IDS provides an instan- cumulative RF for a given year.
taneous value for the time when the tracer test was conducted,
unlike RF that considers the ratio of the mass of water injected and Original annual RF = −0.0107T 2 + 0.1309T + 0.0533 (1)
the mass of additional steam produced over a given period of time. Cumulative RF = −0.0041T 2 + 0.0634T + 0.1159 (2)
Thus, IDS percentage may be significantly higher than the RF.
Cumulative recovery factors are also shown in Fig. 5. The cumu- where T is time in years
lative RF of 17.6%, 22.4%, 27.3%, 30.3%, and 33.2% were calculated
for the first, second, third, fourth, and fifth years, respectively. For 5. Revised definition of annual SRGRP recovery factors
the 5-year cumulative recoveries, the increased production after
each year is combined to calculate recovered steam. This recov- The recovery factors discussed above are defined as the ratio of
ered steam is then divided by the cumulative injection for the same mass of steam gain per year divided by the mass of water injected

Fig. 6. Comparison of annual recovery factors calculated by two methods.


K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328 327

ciency, developing untapped reserves, and extracting additional


heat energy through water injection. From May 1999 to December
2008, these improvements resulted in a total increase of approx-
imately 190 MWnet in the field capacity. Of the latter amount,
plant efficiency improvements, developing additional reserves, and
the injection of SEGEP and SRGRP provided a gain of 26 MWnet,
20 MWnet , 36 MWnet, and 105 MWnet, respectively.
Calpine began receiving 482 L/s of SRGRP water for injection
in October 2003 and, by December 2008, recognized a net gain
of 105 MWnet . The calculated gain from SRGRP indicates annual
injection recovery factors (original) of 17.6%, 26.1%, 37.1%, 39%, and
44.6% for the first, second, third, fourth, and fifth years, respec-
tively. The corresponding cumulative recovery factors are 17.6%,
22.4%, 27.3%, 30.3%, and 33.2%. The revised annual recovery factors
for these years are 17.6%, 16.1%, 14.6%, 12.4%, and 12.2%, respec-
Fig. 7. Steam production decline trend in Calpine market units without SEGEP,
SRGRP, and other improvements.
tively. Without the SEGEP and SRGRP injection, and with no other
improvements, the estimated annual decline rate for all Calpine
market units would have been 6% harmonic starting from January
1995, or approximately 3.3% in 2008. The combined benefits of
per year. This definition is revised to include the effect of un-flashed
injection projects and improvements in the steam field and power
water left in the reservoir. Therefore, the revised RF is defined as the
plants have reduced the decline rate to essentially zero for the past
ratio of mass of steam gain divided by the mass of water injected
6 years (2003–2008).
in a given year, plus the total un-boiled water left in the reservoir
Prior to 1995, annual decline rates of steam production were
up to that year (starting from year zero). The annual RFs calculated
much higher, exceeding 25% during the late 1980s. The dramatic
using both definitions are shown in Fig. 6. The RFs based on the
moderation of decline rates can also be attributed to the following
original definition shows an increasing trend from 17.6% in year
factors in addition to the abovementioned improvements. These
one, to 44.6% in year five. The revised RFs exhibit a declining trend,
factors include: plant retirements, effective field management, and
from 17.6% in year one to 12.2% in year five. A decreasing trend in
operating efficiencies related to the consolidation of competing
the “revised annual RF” suggests overall boiling is decreasing with
interests from as many as 11 separate companies to two (Sanyal,
cumulative increase in the un-flashed water in the reservoir.
2000).
Based on the regression analysis of the data in Fig. 6, the follow-
ing relations can be used to calculate revised annual RF for a given
year. Acknowledgements
3 2
Revised annual RF = 0.0017T − 0.0136T + 0.0163T + 0.171 (3)
Authors would like to thank Dr. Subir Sanyal for reviewing the
where T is time in years. paper, Eloise Leonard, Corky Bracisco, and Linda Kalmar for their
helpful comments, and Calpine Corporation for permission to pub-
lish this paper.
6. Decline rate of Calpine wellfields (market units)

Fig. 7 shows the steam flow-rate data associated with the References
Calpine market plants from the dashed curve in Fig. 3. This fig-
ure is an extension of Fig. 3 of Goyal and Box (2004) to which the Barker, B.J., Koenig, B.A., Stark, M.A., 1995. Water injection management for resource
data from 2003 to 2008 are added. Data points in this figure reflect maximization: observation from 25 years at The Geysers, California. In: Proceed-
ings World Geothermal Congress, Florence, Italy, pp. 1959–1964.
steam production from The Geysers under base (normal operat- CDOGGR, 2008. Annual Report of the State: Oil & Gas supervisor. Division of Oil, Gas,
ing) conditions, but without the estimated benefit from sources of & Geothermal Resources, California, Department of Conservation, Sacramento,
water outside The Geysers. The data points do include effects of the California, USA, Publication No. PR06, pp. 249–262.
City of Santa Rosa, 2003. Connections – Santa Rosa utilities update. 1 (1), October, 2
injection of creek water and steam condensate. Loss of steam deliv- pp.
erability due to conversion from production to injection wells is also County of Lake, 1991. Agreement for TAC Supervision of GeothermEx Reservoir
included. Fig. 7 supports a harmonic decline rate of 6% starting in Modeling. Planning Department – Resource Management Division, Lakeport,
California, USA, September 10, 1991, 18 pp.
January 1995. The data points above the trend line in 1999–2001, Geysers Geothermal Association, 1995. An update on The Geysers, November 1994.
may reflect “puff” production following heavy seasonal power plant Geothermal Resources Council Bulletin 24 (1), 9–21.
curtailments which occurred from 1995 to 1998. The term “puff” Goyal, K.P., Box, W.T., Jr., 1991. Reservoir Response to Production: Castle Rock
Springs Area, East Geyers, California. Society of Petroleum Engineers Paper No,
is used to describe a rebound in reservoir pressure and a tempo- 21790, Western Regional Meeting, Long Beach, California, USA, March 20–22,
rary increase in steam production (Goyal, 2002), following reduced 14 pp.
steam withdrawal from the reservoir. Since 1999, Calpine has pro- Goyal, K.P., 1995. Injection recovery factors in various areas of the Southeast Geysers,
California. Geothermics 24 (2), 167–186.
duced the field with little curtailment. The 6% harmonic decline
Goyal, K.P., 1999. Injection experience in The Geysers, California – a summary.
rate starting in January 1995 translates into a base decline rate of Geothermal Resources Council Transactions 23, 541–547.
approximately 3.3% in 2008. This decline rate is quite small com- Goyal, K.P., 2002. Reservoir response to curtailments at The Geysers. In: Proceed-
pared to more than 25% observed in the late 1980 s, suggesting the ings 27th Workshop on Geothermal Reservoir Engineering, Stanford University,
Stanford, California, USA, January 28–30, pp. 39–45.
field is being managed sustainably. Goyal, K.P., Box Jr., W.T., 2004. Geysers performance update through 2002. In:
Proceedings 29th Workshop on Geothermal Reservoir Engineering, Stanford
University, Stanford, California, USA, January 26–28, pp. 5–10.
7. Conclusions Isherwood, W.F., 1977. Reservoir depletion at The Geysers. Geothermal Resources
Council Transactions 1, 149.
Pruess, K., Narasimhan, T.N., 1982. On the fluid reserves and the production of
Calpine’s efforts at The Geysers have focused on extending the superheated steam from fractured vapor-dominated reservoirs. Journal of Geo-
life of the field and increasing its output by improving plant effi- physical Research 87 (B11), 9329–9339.
328 K.P. Goyal, T.T. Conant / Geothermics 39 (2010) 321–328

Sanyal, S.K., 2000. Forty years of production history at The Geysers geothermal field, Snedaker, G., 2007. Re-commissioning non-operating power projects, the Bottle
California – the lessons learned. Geothermal Resources Council Transactions 24, Rock geothermal case study. Geothermal Resources Council Transactions 31,
317–323. 535–537.
Sanyal, S.K., Klein, C.W., McNitt, J.R., Henneberger, R.C., MacLeod, K., 2007. Assess- Stark, M.A., Box Jr., W.T., Beall, J.J., Goyal, K.P., Pingol, A.S., 2005. The Santa Rosa-
ment of power generation capacity of the western geopower leasehold at The Geysers Recharge Project, Geysers geothermal field, California. Geothermal
Geysers geothermal field, California. Geothermal Resources Council Transac- Resources Transactions 29, 145–150.
tions 31, 447–455.

You might also like