11. Management of Known Geothermal Resource Areas

An ongoing issue in the geothermal power industry is the prudent management of finite geothermal resources. Known geothermal resource areas (KGRAs) in the United States with resource conditions sufficient to generate electricity are rare, occurring domestically only in the Western United States and Hawaii, with an estimated electricity generating capacity of 27,400 megawatts, which is believed to be sustainable for 40 years. Of the currently identified resource base in the United States, around 3,000 megawatts of capacity, or about 10 percent of the estimated total, was installed as of 1995. The successful planned management of one U.S. KGRA, Coso Hot Springs, is described here.

The major industry concern about KGRAs is, and for the past decade has been, the declining production at The Geysers, located along the San Andreas Fault in Sonoma County, California. The Geysers KGRA produces more electricity than any other geothermal field in the world. This chapter presents historical information on the geology, sale, corporate structure, and financial aspects of The Geysers project, specifically addressing the production decline, plans for a pipeline that will carry wastewater to one corner of the field for injection as a strategy for mitigating the decline, and lessons learned from the production decline.

Coso Hot Springs

Coso KGRA Development

Coso Hot Springs (Table 34) is in the middle of the Mohave desert of California, closer to Death Valley than to any metropolitan area. It lies within the boundaries of the China Lake Naval Air Weapons Station (NAWS). The U.S. Navy operates the NAWS as a site for testing Navy weapons.

The Navy’s purposes in developing the Coso Hot Springs KGRA were threefold [209]:

• To manage the electricity generation component of NAWS expansion
• To provide an alternative energy source, allowing the Navy shore facility to reduce overall Navy dependence on fossil fuels, especially imported fuels
• To provide a $40 million cost saving on utility bills over the first 10 years of operation.

To meet these objectives, development and operation of the field were contracted out, with the electricity being sold to the utility servicing the area, Southern California Edison. The Navy contracted with CalEnergy Company, Inc. (CECI) to support the development of the field. At this point, CECI, the Nation’s largest geothermal corporation, owns or operates:

• About half of the Coso Hot Springs project
• Seven operating geothermal facilities in the Imperial Valley, including the Nation’s most recent geothermal power plant, Salton Sea IV
• All of the operations at Desert Peak in Nevada
• The well field at Roosevelt Hot Springs in Nevada.

Advantages to the Navy

From 1987 through 1993, the Naval Weapons Center (now NAWS, China Lake) received direct reductions in its electricity bill totaling $24.2 million as a result of electricity production from Coso Hot Springs. The saving for calendar year 1993 alone was $4.2 million, which equates to a one-third reduction in electrical energy cost.

In cooperation with private industry, NAWS China Lake has promoted the development of its geothermal resource in a way that ensures continued Navy control and supports Navy goals. The production from Coso was equivalent to 16 percent of 1993 electricity consumption by Navy shore facilities. This is a significant step toward meeting the directives of the Chief of Naval Operations regarding the conversion of Navy shore facilities to operating on alternative energy sources.

Other benefits to the Navy are derived from the sale of excess electricity by the contractor, CECI, to the local utility company, Southern California Edison. The project allows the Navy to manage the resource, including administering the contract, funding an exploration program, and supporting energy cost avoidance projects. All of this is done without expenditure by the Navy for the plant construction or plant operation.

Benefits are also realized in the local economy through taxes, jobs, and economic stimulation. Taxes paid to Inyo County by the project have amounted to more than 20 percent of the county’s income in recent years.

Total generating capacity at the Coso geothermal field amounts to more than 250 megawatts. If 1 megawatthour of electricity will meet the hourly needs of approximately 1,000 households, and assuming an average of 2.5 people per household, the output from Coso can provide enough power to serve approximately 625,000 people. The week of peak production in 1987 averaged 266.4 megawatts per hour, generating roughly 6,400 megawatthours (1.9 billion Btu) per day, sufficient for 640,000 residential consumers per day, assuming that each person’s electricity consumption was equal to the 1995 national average [210].

In addition to a reliable source of power, Coso provides a significant benefit to air quality in the environs. Coso’s geothermal power plants have sulfur emissions rates that average only a few percent of those from fossil fuel alternatives: less than 1 pound of carbon dioxide per megawatthour of electricity generated, as compared with 14.47 million metric tons of carbon per quadrillion Btu (328 pounds of carbon dioxide per megawatthour) for a plant fueled by natural gas and 25.71 million metric tons (583 pounds per megawatthour) for a coal-fired power plant [211].

The Geysers

The Geysers produces more electricity than any other geothermal energy field in the world. At the end of 1996, the installed nameplate capacity of the 22 generating units at The Geysers field was 1,837 megawatts. Of this capacity, 14 units totaling 1,224 megawatts were operated by Pacific Gas & Electric Co. (PG&E). The remainder was owned by five other companies, including the Calpine Corporation.

Electricity production had been on a slow but steady decline since the late 1980s, when it declined sharply in 1995 (Table 35 and Figure 25). Of the available capacity, some was not taken due to the relatively expensive nature of the electricity, and some was not taken due to reduced steam pressure within the KGRA.

Production Decline at The Geysers

Economic Curtailment

One of the major reasons for the production decline at The Geysers is economics. In 1996, California had surplus gas transmission capacity in the range of 1 to 2 billion cubic feet. With this much surplus capacity, the cost of off-peak gas-fired energy is essentially the cost of spot market gas. With recent spot prices roughly in the range of $1.50 per million Btu, the fuel cost component of surplus gas-fired energy from a conventional steam plant (10,000 Btu per kilowatthour), is $15 per megawatthour. If the marginal generation (not capacity) comes from a combined-cycle plant of only average efficiency (8,000 Btu per kilowatthour), the fuel cost component of the surplus gas-fired energy could be as low as $12 per megawatthour. Marginal transmission charges off-peak are essentially zero. The small marginal cost of operations and maintenance for a steam or combined-cycle gas-fired power plant are insignificant.

Since the late 1980s (when steam production started to decline) surplus hydroelectric, coal-fired, and gas-fired electricity generation from throughout the West and Northwest have been available at prices much lower than those specified in some of PG&E’s power or steam purchase contracts (see Appendix E). In the past few years, surplus energy costs and prices have declined to the lowest levels in recent history. For example, average production costs for the Western Systems Coordinating Council averaged about $16 per megawatthour in 1993 and 1994 and less than $14 per megawatthour in 1995. In August 1996, nonfirm off-peak energy at the California-Oregon border was priced at about $12 per megawatthour. Nonfirm on-peak energy was less than $18 per megawatthour. (Prices vary daily by $1 to $2 per megawatthour or more.)

In contrast, a representative long-term geothermal electricity contract currently provides for firm capacity payments in the range of $156 to $167 per kilowatt per year and $130 or more per megawatthour or more for associated energy—about 3 times and 8 to 10 times higher, respectively, than avoided costs for capacity and energy purchased separately. In these cases, PG&E chooses maximum contractual or economic curtailment, whichever is greater.

Because of the availability of low-cost surplus energy and low-cost on-system generation, PG&E has been renegotiating its geothermal steam supply contracts, which currently run from $130 to $167 per megawatthour. In August 1995, PG&E and three of its steam suppliers at The Geysers entered into an agreement that lowered the price of generation from those steam supplies above the 40 percent of annual field capacity for which PG&E has take-or-pay commitments. As a result of the discounted price, PG&E increased generation at The Geysers over what it would have taken at the higher price. That agreement expired on December 31, 1995. The parties entered into a similar agreement for February 1996 and negotiated a discounted steam price agreement for the remainder of 1996.

Geothermal power that is generated near avoided cost or purchased on an avoided-cost basis would be curtailed only due to steam conditions and reservoir maintenance and stability requirements.

Reduced Steam Pressure

Venting of steam instead of reinjection of the condensed steam has led to the current reduced steam pressure at The Geysers. Figure 26 shows that steam production averaged about 240 billion pounds for 1987-89 and about 195 billion pounds for 1992-94. On a peak-to-trough basis (through 1994), steam production declined from 246 billion pounds in late 1987 to about 188 billion pounds in 1994. At the same time, the injection-production ratio increased from about 70 percent to about 90 percent. Without an increase in injection, production might have fallen more than it has.

PG&E’s consolidated capacity factor for The Geysers is projected to be approximately 35.8 percent of installed capacity in 1996, which includes economic curtailments, forced outages, scheduled overhauls, and projected steam shortage curtailments. The actual capacity factor in 1995 was 37.3 percent. Without take-or-pay contracts and contractual limits on curtailments, electricity production from The Geysers would have been considerably lower in recent years than was the case.

The effect of reduced steam flows on the economics of curtailment is twofold. First, reduced steam flows generate an absolute increase in fixed operations and maintenance costs and an increase in unit variable costs. Costs increase for reasons including reconfiguration of turbines, condensers, and gas control devices for low-load or low-pressure operations; extra maintenance to address condensate and evaporative problems; and water injection system development. Unit variable costs increase because of water injection, control of noncondensible gases, fixed or semi-fixed staffing at reduced power levels, and other factors. Second, since steam flows are clearly declining, it makes no economic sense to extract the steam for use in nonfirm, low-value, off-peak generation.

Resource depletion at The Geysers KGRA continues to result from a series of decisions that generally were based on the assumption that the resource was infinite, or at least infinitely and rapidly replenishable. Starting as a hot spring resort and geyser tourist attraction during the California Gold Rush in 1849, the KGRA was surveyed and sized as a minable resource (i.e., an “open loop” system) after World War II. In the mid-1950s, legislation authorized the State of California to auction geothermal energy rights to the few square miles of the KGRA to competing firms, in a manner somewhat parallel to the current auctioning of oil and gas rights in the Gulf of Mexico and off the continental shelves.

The local electric utility, PG&E, contracted with each of the high-bidding corporations, agreeing to purchase a certain amount of electricity from each. Ownership of the various steam supply system components varied from contract to contract. This arrangement consistently gave neither the utility nor the corporations uniform incentives to conserve the steam, to use the steam efficiently, or to cooperate in preventing or curtailing resource depletion.

The first energy extraction systems were designed to be open, venting water vapor and heat. The condensed water vapor was allowed to flow into existing waterways or was evaporated. Reinjection began much later, and a few plants still do not reinject remaining fluids. There were and still are no consistent disincentives for inefficiencies; on the other hand, there have been incentives to tap as much as needed by some, but not all, firms, in order to generate the contractually agreed upon electricity. However, without a prior binding agreement by KGRA users to cooperate in the event of a resource shortage, and without a legislated solution, field depletion will continue.

In the early 1980s, it became increasingly apparent that The Geysers could be depleted simply by mining. A change in thinking occurred, to making 30-year estimates of resource utilization, rather than the amount that could be exploited in the near term.

Halting Pressure Declines by Well Injection

In 1989, Calpine and the Northern California Power Agency (NCPA) started a joint injection program. The two organizations control a total of 198 wells, 10 power plants, and 325 gross megawatts of capacity, all in the southeastern portion of The Geysers KGRA. The flow rate of wells surrounding Calpine CA956A-1, one of the wells converted from production to injection in 1989, had been declining at an exponential rate of about 26 percent during 1988 and 20 percent during 1989 until injection began in late 1989 [212]. Injection had the effect of slowing the decline in the remaining 12 production wells to an exponential rate of about 10.5 percent. Combining Calpine’s CA956A-1 and the NCPA’s nearby C-11 well, the two organizations are injecting roughly 1,500 gallons per minute. Within 5 months, steam flow at 25 nearby wells increased, resulting in 20 megawatts more power and a drop in noncondensible gases.

Because of the demonstrated value of water injection in slowing steam flow declines, several well owners have increased injection or have planned injection programs. The major injection program is the Southeast Geysers Pipeline Project, which involves the construction of a 20-inch-diameter, 29-mile-long pipeline with a capacity of 5,400 gallons per minute (7.8 million gallons per day). The pipeline will carry water from a wastewater treatment facility north of The Geysers for injection into the steam reservoir in the southeastern portion of the KGRA. It is designed primarily to support NCPA’s two plants (247 megawatts) and four of PG&E’s plants (495 megawatts).

Construction began in October 1995, and the pipeline is scheduled to begin operation in 1997. A sustainable increase in capacity of 50 to 70 megawatts is expected. The pipeline is estimated to cost about $45 million [213], including $7.2 million from the U.S. Department of Energy, but including neither the estimated $7 million required to move the water from the KGRA’s boundary to the injection sites nor the additional water treatment facility needed before the water enters the pipeline. While the cost of construction for the first pipeline and pumping systems may be high, the value of a small, but potentially environmentally benign, waste effluent disposal system is seen as a benefit and, therefore, a partial cost offset.

Halting Pressure Declines by Cycling Operations

In 1988, NCPA was the first operator at The Geysers to switch from baseload operations to cycling operations. The switch was designed to slow the decline in steam production from the field. NCPA’s generating units are rated at 247 megawatts, but production was reduced to 150 megawatts. Tests in June 1993 showed that the plants are capable of sustained operations up to 221 megawatts.

PG&E began cycling its units in August 1994. Between August 1994 and May 1996, PG&E deferred 3,500 gigawatthours of geothermal energy. PG&E’s reasons for cycling were partly technical (to maintain steam flow) and partly economic (to substitute lower cost hydroelectric, coal-fired, or gas-fired energy). In 1995, PG&E curtailed or completely avoided production from some of its units at The Geysers for as much as 5 months, because it could economically dispatch nongeothermal units instead. The combination of large winter snow runoff and low-cost natural gas led to a substitution of hydroelectric and gas-fired energy for almost 2,000 gigawatthours of geothermal energy. PG&E estimated that deferred geothermal generation would approach 2,000 gigawatthours in 1996. For a point of reference, a power plant rated at 228 megawatts and operating at 100 percent capacity factor can generate 2,000 gigawatthours per year.

Cycling geothermal wells, however, causes some operational problems. Among the problems are thermal cycling of steam within well bores, water collection in steam gathering systems, water carryover to steam separators and turbines, increased wear and maintenance requirements, and an increase in noncondensible gases when generation is increased.

When production is curtailed or shut down, plant operators often have to close in wells to comply with air emission regulations. Closing the wells allows steam in the well bore to condense, resulting in thermal cycling of the wells (not the power plants). This thermal cycling can damage the well and create flow problems when operations resume. Modified operating practices, such as targeting minimum field-wide steam flows and setting limits on the duration that wells are in no-flow conditions, have lessened but have not eliminated this problem.

Condensation in low-flowing or nonflowing well bores allows noncondensible gases to build up near the well bore. When the wellhead control valve is opened and the gases pass through the gathering system to the generating station, the condenser or abatement system can quickly become overloaded, resulting in reliability or environmental compliance problems. Changes in noncondensible gas concentrations in The Geysers appear to be correlated with changes in injection strategy; injecting water correlates with decreases in noncondensible gases, while shutting in an injection well correlates with increases in noncondensible gases. In the southeastern portion of The Geysers KGRA, where noncondensible gases have been increasing despite aggressive injection programs, the decline in reservoir pressure appears to have been large enough to offset the beneficial effects of injection.

Condensation within the steam gathering system under low-steam or no-steam conditions creates a risk of water carryover to the steam separators and turbines when production resumes. This problem can be controlled to some extent by increased monitoring of the steam field and the use of extra moisture traps or drop pots, but the problem has not been eliminated.

Cycling operations also increase wear and tear on the equipment and systems (e.g., stuck valves). In some cases, evaporation from the cooling tower is greater than the amount of condensate from the steam flow. Running only a few cooling tower fans at a plant can ease this problem.

Lessons Learned

Some owners and operators of power plants at The Geysers have learned that the resource must be intensively managed rather than mined for steam. The notion of an inexhaustible thermal resource (within the bounds of existing extraction and generating capability) has been clearly proven incorrect. Since the late 1980s, when the steam decline became both noticeable and sustained, six plants, totaling approximately 200 megawatts, have been retired or suspended. Most other plants have been effectively derated due to declining steam production. Reinjection of consumed steam has been successful in slowing steam declines but thus far has not been shown to increase steam production to the levels that prevailed in the late 1980s.

Through the 1980s, production drilling activities were often designed to find new steam. Since that time, drilling activity has declined sharply, and the objective of drilling campaigns has shifted toward a goal of more economical exploitation of existing steam sources (e.g., via double- and triple-forked bottoms from a single hole). Older wells are also under consideration for deepening. Many of the oldest wells are relatively shallow, and extending the borehole by several thousand feet at an existing plant is a small expense in relation to the potential for a more robust steam resource.

In the past few years, the major plant owners have bought or are planning to buy new turbines and steam paths designed to operate at turbine inlet pressures less than 100 pounds per square inch gauged (psig), for which most of the units at The Geysers were originally designed. Most operators are considering main steam line pressures down to 50 psig. Operators are also modifying the design and operations of existing turbines, condensers, and gas handling systems for low-load and cycling operations. These changes promise to extend the life of the resource but at a higher cost than was the case in the 1980s.

Thus, the combination of rising costs, reduced steam flows, and abundant low-cost energy from conventional generation means that The Geysers will increasingly become a cycling resource with selective system and resource upgrades. The days of unlimited extraction at rated power plant output and of large-scale in-fill drilling are over.

Outside The Geysers, the geothermal electricity generation industry has watched this series of events intently, and has responded by constructing closed-cycle systems that reinject virtually everything that comes out of the ground, including the residual heat. Binary and dual-flash heat extraction systems are the only ones being installed anywhere in the world today; the paradigm of “steam mining” has been replaced with the recognition that these geothermal resources have finite flows and capacities but, with proper management, can be sustained indefinitely.

Also, development to the limits of the capacities estimated for other KGRAs is being approached more cautiously than before, so as to avoid a scenario similar to the one at The Geysers. Competition between corporations working within the same KGRA as a result of State or Federal auctions has eased as a result of ownership consolidation and changing auction strategies. Wells in new developments are being spaced further apart; each resource is being tapped to only a fraction of its estimated sustainable potential; and water resources are being examined carefully for sufficiency of flow and quantity, water chemistry, and tendencies toward brine and scaling.

A paradigm shift seems to have been implemented, though not completed. While vapor-dominated hydrothermal resources were once viewed and engineered as unlimited, they are now understood as finite resources requiring prudent management to sustain them. There is no evidence worldwide of the steam mining paradigm’s continued acceptance at developing sites.

Contact:

Mark Gielecki

mgieleck@eia.doe.gov

Phone: (202) 426-1141

Fax: (202) 426-1280