September 21, 1998

Mr. Bruce Kaneshiro, Project Manager

c/o Environmental Science Associates

225 Bush Street, Suite 1700

San Francisco, CA 94104

Dear Mr. Kaneshiro:

PG&E hereby submits the attached comments on the Draft Environmental Impact Report (Draft EIR) for PG&E’s proposed divestiture of its Potrero, Contra Costa, Pittsburg, and Geysers Power Plants. The comments are provided in two sections; the first section contains general comments and the second section contains specific comments by page number.

PG&E appreciates the opportunity to comment on the Draft EIR, and urges the Commission to certify the EIR as final as soon as possible.

If you have any questions, please call me at (415) 973-1595.

Sincerely,

Cecilia F. Montana

Acting Director, Divestiture and Gas Ratemaking

Attachment

1. Conservative Nature of the Analytical Methodology. PG&E would like to underscore the first sentence in the section of the Draft EIR that describes the analytical methodology employed throughout the document. The Draft EIR states, "In reading this EIR, it is important to understand the uncertainty involved in predicting the future behavior of PG&E or new plant owners; the legal framework in which this divestiture proposal exists; and the conservative nature of and bases for the assumptions used throughout the document to evaluate the environmental impacts of the project" (page 3-1).

Due to the uncertainty inherent in predicting the behavior of PG&E and new owners in the new restructured market, the analysts created hypothetical scenarios. Some scenarios predicted what PG&E's behavior might be in the future and other scenarios predicted the behavior of unknown future plant owners.

While creating the scenarios, the EIR drafters had a wide array of assumptions from which to choose. The drafters consistently selected the most conservative assumptions, that is, the assumptions that would result in the greatest potential for environmental impacts. This process was repeated for each of the scenarios so that the analysis would represent a reasonable worst case. The analysis therefore overestimates the potential for environmental impacts.

[End L1]

This consistent leaning toward the worst case resulted in scenarios that the EIR characterizes as unlikely. For example, the fundamental basis for the analyses is that new owners might operate the plants at higher levels than might PG&E. The EIR cautions that "the degree to which generation would increase at the plants slated for divestiture is highly uncertain" (page 3-8). Further "it is not possible to determine with any precision at which plants operations would increase, or the degree to which operations would increase at any particular plant" (page S-8).

One of the most influential assumptions in the analysis is that, for new owners of the fossil-fueled plants but not PG&E, "natural gas could be purchased in unlimited quantities at a 25 percent discount from the least expensive supply of gas assumed to be available to fuel California power plants" (page 3-12). This assumption is an artificial construct. It was used to force the computer model to move in the desired direction, that is, toward a greater potential for environmental impact. As the EIR points out, "The purpose of this assumption was to remove, to a great degree, the cost of fossil fuel from the new owner's decision whether and when to generate power. Although it is extremely unlikely that such a reduced gas price could be obtained, this assumption further strengthens the conservative nature of the impacts analysis" (page 3-12).

[End L2]

It is indeed unlikely that a new owner would ignore the price of fuel in its decisions. It is also highly unlikely that a new owner, and only a new owner, would be able to purchase natural gas for 25 percent less than its California competitors. It is more likely that similar gas prices would be available for all generators and the generator with the most favorable economics would tend to generate more than one with less favorable economics. In any case, total generation would be limited by overall demand. Nevertheless, PG&E recognizes that the Draft EIR analysts employed this conservative approach in order to examine the greatest potential for environmental impacts. As such, the potential impacts identified in the EIR are indeed only potential, and are potential only under an extreme set of circumstances. The projections should not be used or relied upon for purposes of reasonable economic forecasting.

[End L3]

2. New Owners of the Geysers Geothermal Plant. Alternative 3 in the Draft EIR evaluates whether there might be an environmental benefit if the Geysers plants were sold to the steam field operators. The analysis concludes that to do so would be environmentally preferable, as it "would allow for greater coordination of generating unit operations with steam field characteristics and may reduce steam stacking" (page 6-28). This is an inappropriate "conclusion" for two reasons.

First, the alternative presumes that there would be "greater coordination". This is an assumption embedded in the design of the alternative (page 6-23), not a conclusion derived from analysis. No evidence is presented that current operations are not now coordinated nor that they could be more coordinated.

In fact, PG&E's contract with UNT to supply steam for the units in Sonoma County and PG&E's contract with Calpine to supply the steam for units in Lake County both require the companies to coordinate operations. To do so, the companies have instituted operational procedures and monitoring equipment specifically to prevent unabated steam releases, or "steam stacking".

Steam stacking can occur when generation is reduced. Such reductions may be scheduled or unscheduled. In the case of scheduled reductions, such as for unit maintenance, the steam field agreements require PG&E to notify the steam field operators at least 48 hours in advance. In the case of unscheduled reductions, monitors installed on PG&E's generating units and the steam suppliers' steam gathering systems immediately transmit changes in generation and steam pressure. The signals from the monitors are automatically relayed to Unocal’s and Calpine’s control centers, which are staffed 24 hours a day. When a signal is received, the steam suppliers automatically throttle the wellhead valves, which redirects the steam to other generating units with emissions abatement equipment. This monitoring system has been successfully operating for several years.

[End L4]

Secondly, the computer modeling for this alternative assumes that "the existing steam supply contracts would be inapplicable (because the contracting parties would be merged into one steam supplier/plant owner for each unit) and the price of steam to the plants would decline" (page 6-6). The modeling results show that the Geysers plants would operate more under Alternative 3 than under the other scenarios. The Draft EIR concludes that, due to this increase in operations, there would be less steam stacking. While the computer modeling results show a significant increase in year 1999, the increase is slight by the year 2005. In 2005, the modeled annual plant capacity factors under the proposed project (Cumulative Analytical Maximum Scenario), continued ownership by PG&E (Draft EIR Alternative 1), and ownership by the steam field operators (Alternative 3) are 87, 90, and 93 respectively (page 6-7). Thus, according to the model, the increase would be 6 percent averaged over 14 units. Due to the infield systems to prevent unabated steam releases, such an increase in operations would not result in a significant benefit, if any.

[End L5]

As the Draft EIR notes, the analysis of the proposed project does not indicate that the project would result in any change related to steam stacking. "Therefore, this alternative would not be necessary to reduce project impacts. It may, however, beneficially reduce steam stacking and release events" (page 6-24). In PG&E's opinion, Alternative 3 has not demonstrated that selling the Geysers units to the steam field operators would be any more or less environmentally preferable to selling the units to a third party. It should also be noted that the steam field operators have a right of first refusal to purchase these units, as stated elsewhere in the Draft EIR.

[End L6]

The CPUC has no responsibility to choose the environmentally preferable alternative. The California Environmental Quality Act (CEQA) Guidelines require that the lead agency identify an environmentally superior alternative among other alternatives analyzed. However, CEQA does not require the agency to choose the most environmentally desirable alternative if the agency has reduced the project's environmental effects to an acceptable level through mitigation. The CPUC does not have authority to order the sale of the plants to a particular buyer without PG&E’s consent.

[End L7]

3. Cumulative Analysis. In Section 5.2.2, the EIR describes the assumptions made in the cumulative analysis regarding new generation to replace the Hunters Point Power Plant. The EIR notes on page 5-4 that:

PG&E would like to underscore that the facility assumed in the cumulative analysis is a hypothetical facility. The generation and/or transmission facilities that will replace the power and reliability now provided by the Hunters Point plant are not known at this time. It will be incumbent on the proponents of the new facilities, and the agencies that must approve the project(s), to evaluate, in compliance with the California Environmental Quality Act, the potential for project-specific and cumulative environmental impacts associated with the new facilities and to address any significant environmental effects.

[End L8]

Page S-4, Figure S.2. Please note that PG&E will also continue to own the Diablo Canyon Nuclear Power Plant, located in San Luis Obispo County. (The same figure is used as Figure 2.1 on page 2-3.)

[End L9]

Page S-5, paragraph 3. The paragraph states that "PG&E will retain facilities and equipment at each site that pertain to transmission or distribution operations." Please note that PG&E will divest a small portion of the transmission lines at the Geysers Power Plant, as stated elsewhere in the document (e.g., pages 2-7, 2-28, and 4.12-17).

[End L10]

Page S-5, paragraph 4. The Summary states that "The Purchase and Sale Agreement for each plant requires a deed restriction that prevents the new owner from using the site for residential or other sensitive uses." This is true for the fossil-fueled plants. However, for the Geysers plant, the Purchase and Sale agreement requires the purchaser to agree to a land use covenant that will require the purchaser (and its subsequent transferees, if any) not to use the property for residential or other sensitive uses. (See also pages 4.1-13 and 14.)

[End L11]

Page S-21, Table 2.6. Please note that the levels of significance indicated for Impacts 4.5-4, 4.5-5, and 4.9-3 for the proposed project should include an "(M)" to indicate that the EIR has identified supplemental mitigation for each of these impacts.

[End L12]

Page 2-5, paragraph 2. Please replace the last sentence regarding PG&E’s hydroelectric assets with "PG&E is currently considering various options for market valuation of its hydroelectric assets".

[End L13]

Page 2-20, paragraph 2. All three boilers associated with Units 5, 6, and 7 are capable of burning residual fuel oil.

[End L14]

Page 2.42, paragraph 2. The Draft EIR states that "Early reinjection of wastewater from the Southeast Geysers effluent pipeline have been encouraging. There has been about a 7 percent (60 MW) increase in capacity in the Southeast Geysers field as reported by staff at the Geysers Power Plant." Upon closer examination of the peak output achieved by the four PG&E plants that are in the Southeast area of the steam field, PG&E believes that 7 percent overstates the increase. A more accurate estimate would be that there has been about a 5 percent increase in capacity. In addition, given the design capacities of the four Southeast Geysers Units (Units 13, 16, 18, and 20) listed on Table 2.1 (pages 2-8 and 9), it appears that the 7 percent increase resulting in a 60MW increase is a miscalculation

[End L15]

Pages 2-44 and 45, Table 2.3. (1) There is also a lease from the State Lands Commission for public land associated with the Contra Costa plant. (2) Note that the Aboveground Petroleum Storage Tank requirement applicable to the Geysers plant is not a permit per se. The requirement includes a biannual statement, registration fee, and SPCC plan.

[End L16]

Page 4.4-5, paragraph 1. The last sentence states that "Sodium hypochlorite is regularly applied in the condensers to minimize growth of biological organisms and is then discharged". Please note that, in accordance with the NPDES permit for the plant, the sodium hypochlorite is neutralized with sodium bisulfate before discharge.

[End L17]

Page 4.5-14, paragraph 3. Note that since the beginning of 1994, PG&E has essentially ended using fuel oil. However, fuel oil may still be used in limited circumstances. Under BAAQMD Regulation 9, Rule 11, only natural gas is to be burned in these units, except during force majuere natural gas curtailment and very limited testing.

[End L18]

Page 4.5-24, paragraph 2. The paragraph states "Unit 3 is coupled to a single boiler, which is capable of burning natural gas or fuel oil; however, since 1995, only natural gas has been burned because of Regulation 9, Rule 11". The sentence could leave the misimpression that the switch to natural gas occurred because of Regulation 9, Rule 11. However, the fuel switch was made prior to when the regulation became effective in 1997. The sentence could be clarified by stating, "since 1994, PG&E has only burned natural gas and Regulation 9, Rule 11 requires that natural gas be burned except in very limited circumstances".

[End L19]

Page 4.5-34, last sentence. Please delete the reference to the mobile combustion turbine at the Contra Costa plant. It is no longer at this plant.

[End L20]

Page 4.5-39, last paragraph. The paragraph incorrectly states "While two of the seven boilers are permitted to burn either natural gas or fuel oil, all of them currently burn only natural gas." All seven boilers are permitted to burn fuel oil as long as permit and rule conditions are followed.

[End L21]

Page 4.5-47, paragraph 4. The paragraph refers to measurements of radon at the Geysers plant "that indicated levels of radon ranging from 3 to 5 pico-curies per liter of air, which is above typical background levels of 1 pico-curie per liter (1998, personal communication with Lake County APCD)". This range is substantially higher than the measurements reported as part of Geysers Air Quality Monitoring Program (GAMP), which measures radon at two monitoring sites in public areas of the Lake County geothermal area that are downwind of the plant. The most recent quarterly GAMP report of April 1 to June 30, 1998, which reported data for January, February, and March 1998, reported a radon range of 0.1 to 0.5 pico-curies per liter (pCi/l), with an average of 0.2 to 0.3 pCi/l. These measured levels are about 1/10^th of the levels identified in the Draft EIR.

[End L22]

Page 4.5-61, last sentence. The Draft EIR indicates that the air quality modeling was performed using the AP-42 emissions factors that the U.S. EPA released earlier this year (per U.S. EPA reference listed on page 4.5-84). The Draft EIR refers to these factors as "new, more accurate data reported by EPA for combustion units". PG&E questions whether the new factors are indeed more accurate. The new factors for PM-10 are based on a very small sample – only four tests nationwide. In addition, for the first time, the EPA has reported separate emissions factors for the filterable fraction of PM-10 and a condensable fraction. There is reason to believe that the condensable fraction is purely an artifact of the sampling procedure as a result of nitrates or sulfates present in the exhaust steam. The EPA has itself indicated little confidence in the factor for the condensable fraction. On a scale of A to E, with A being the highest confidence, the EPA rates its confidence in the factor for the condensable PM fraction for boilers as D. If only the filterable fraction is used in the emission factors, the modeled PM-10 results would drop by a factor of 4.

[End L23]

Page 4.7-30, paragraph 3. The fourth sentence states "The revised draft HCP (dated April 22, 1998) and associated permit application documents have been reviewed and deemed completed by USFWS and NMFS staff in June 1998." To date, PG&E has not received the certificate of completion. Please delete the sentence.

[End L24]

Page 4.7-30, paragraph 3. To update the paragraph, in the fifth sentence, please delete "intend to adhere to" and replace with "concur with", and delete "July 1998" and replace with "September 1998". Similarly, in the last sentence of the paragraph, replace "staff" with "proposed" and replace "October 1998" with "December 1998".

[End L25]

Page 4.7-31, second bullet. As required by the federal agencies, the funding is now to be maintained in "a special deposit account".

[End L26]

Page 4.7-35, paragraph 3. The federal agencies have emphasized that the Implementing Agreements and HCPs are not transferred, but are part of the permit and reissued along with the reissued permit. As such, in the fourth sentence, please delete "and transfer of the draft Implementing agreement" and substitute with "including the related Implementing Agreement", and delete "to the new owner".

[End L27]

Page 4.7-35, paragraph 3. To update this paragraph, please delete the two sentences that begin with "USFWS has proposed" and end with "within a 60-day period", and replace with "The USFWS has concurred with this language".

[End L28]

Page 4.7-35, paragraph 3. For clarification, please revise the last sentence as follows: "Accordingly, if the Section 10 Permits are issued to PG&E at least 60 days prior to closing, the permits should be reissued to the new owner at closing, and the new owner will be subject to the restrictions of such permits and the CESA MOU to the same extent PG&E would have been.

[End L29]

Page 4.7-36, paragraph 3. For clarification, please delete "As a condition of closing the sale, the new owner will be required to obtain" and replace it with "If the Section 10 permits have been reissued to PG&E prior to closing, the new owner will be required to seek".

[End L30]

Page 4.7-37, top of the page. Under "Timing", to capture all of the monitoring actions associated with the mitigation measure, please revise the sentence to read: "Documents should be provided to the CPUC at least forty days before the title transfer, copies of the letters should be provided to plant managers at the close of sale, and the Section 10 Permits should be provided to CPUC when obtained."

[End L31]

Page 4.7-37, Level of Significance after Mitigation. For clarification, the term "Operational Constraints" should be in lower case as it is no longer a defined term. Also, please insert after "operational constraints", the phrase "in PG&E's Section 10 Permits and the HCP".

[End L32]

Page 4.9-4, last paragraph. The section states that "No acutely hazardous materials are used or stored at the Potrero Power Plant." However, lead-acid batteries are used at the plant; the sulfuric acid in the batteries is classified as an acutely hazardous material.

[End L33]

Page 4.9-5, last paragraph. In the last sentence, for accuracy, please delete "adversely" and replace "throughout the entire plant area" with "in areas of the plant".

[End L34]

Pages 4.9-6, 9, 11, and 13, Phase II testing. PG&E has now completed Phase II testing for all four plants. Risk Assessments have been completed for the Potrero, Contra Costa, and Pittsburg plants. A Risk Assessment is currently being prepared for the Geysers plant.

[End L35]

Page 4.9-17, paragraph 1. The Draft EIR states, "The appropriate lead agency at each plant will be selected by means of the Site Designation Process Under the Unified Agency Review of Hazardous Material Release Sites". It is likely that the lead agency for each plant will be selected by the Site Designation Process. However, other selection processes may be used to honor other processes and arrangements with local agencies in place at the time. PG&E recommends that in the quoted sentence, the words "will be" be replaced with "will likely be".

[End L36]

Page 4.9-19, paragraph 3. Please note that ammonia is not currently used in emissions abatement equipment at the power plants to be divested.

[End L37]

Page 4.11-16, paragraph 2. The Geysers Power Plant generates an estimated $920,000 per year in property taxes to Lake County (not $920,000 million).

[End L38]

Page 4.12-3, transmission section, paragraph 2. Note the transmission corridor that serves San Francisco and the former Skyline District is rated at about 750 MW assuming the underground 230 kV cable is not available.

[End L39]

Page 4.12-5, last paragraph. Regarding the transmission lines, please delete the clause "to protect the in-city on-line generation from potential crippling damage". The lines are automatically opened, not to protect generation, but to ensure service to the network load. The generators are equipped with their own underfrequency relays to protect them from damage.

[End L40]

Page 4.12-7, paragraph 2. (1) Note the CPUC is not served by the network load. (2) Please delete "and in BART trains under the bay". While the BART stations are served by the distribution network, traction power to run the trains is served from a dedicated substation in San Francisco.

[End L41]

Page 4.12-7, paragraph 4. (1) Note that the SFOC require the Dispatcher to load San Francisco generation at differing percentages based on the load. At 800 MW required operating capacity, 40 percent is required; at 700 MW, 30 percent is required. The sentence could be corrected by replacing "40 percent" with "certain portion". (2) In the paragraph’s last sentence, please delete "not". During off-peak periods, the SFOC is intended to sustain some, if not all, of the much-reduced downtown loads.

[End L42]

Page 4.12-9, paragraph 1. Please delete the word "single", as there are multiple lines between San Francisco and the outside power transmission grid.

[End L43]

Page 4.12-9, last paragraph. For accuracy in the first sentence, please replace "and replacement of 430 MVA transformers" with "and replacement with two 420 MVA transformers".

[End L44]

Page 5-11, paragraph 2. The section states that "PG&E is one of the key players in the Santa Rosa Wastewater Modified Geysers Recharge Project. However it is assumed that if PG&E were to sell its units, the new owners would simply assume PG&E’s role in the process." These two sentences overstate PG&E's involvement with the project. PG&E is not a party to the contract signed by the City of Santa Rosa and Unocal-Thermal, nor has PG&E entered into separate agreements with the steam suppliers as it did for the Southeast Geysers Pipeline Project. PG&E recommends deleting the two sentences.

[End L45]

Please see the General Comment above regarding New Owners of the Geysers Geothermal Plant.

[End L46]

PG&E questions Attachment C’s conclusion, and the underlying statements that lead to the conclusion, that new owners of the plants would tend to operate the plants at higher levels in the future than would PG&E. According to Attachment C’s authors, three primary factors could influence increased generation: the portfolio effect, fuel procurement practices, and the ability of new owners immediately to participate in the direct access market. The discussions of these factors unfortunately contain inaccurate and speculative statements. However, PG&E recognizes that the resulting conclusion of increased generation is embedded as an assumption in many parts of the Draft EIR analyses and serves to further the conservative nature of the analyses. Thus, PG&E is choosing to not submit detailed comments on Attachment C. Please note that by accepting Attachment C’s conclusion solely for the purpose of an assumption in the Draft EIR analyses that leads to increased generation, PG&E is not accepting Attachment C’s statements, analyses, nor conclusions for any other purpose.

[End L47]

Please see the comment for pages 2-44 and 45, Table 2.3.

[End L48]

There are several minor inaccuracies on these maps. Please note that while the inaccuracies do not affect the EIR analyses or conclusions, these maps should not be relied on for purposes other than this EIR.

[End L49]

- - - - - -