IV. Project Need

PG&E asserts that the Jefferson-Martin project is necessary for four reasons: (1) to reliably meet projected electric demand in the Project Area; (2) to satisfy applicable planning criteria; (3) to diversify the transmission system serving the Project Area; and (4) to implement the ISO Board of Governors' April 2002 Resolution approving the proposed Jefferson-Martin project for addition to the ISO-controlled grid. In this section, we describe the reliability planning criteria that are applicable and then use them, along with other considerations, to assess need for the Jefferson-Martin project.

PG&E and the ISO use different geographical areas in assessing need for the Jefferson-Martin project. The area PG&E refers to as the Project Area consists of the City and County of San Francisco, Burlingame, Millbrae, San Bruno, South San Francisco, Brisbane, Colma, Daly City, Pacifica, and Hillsborough. The ISO evaluates need for a broader San Francisco Peninsula Area, which the ISO characterizes as the area north of Palo Alto or north of the Ravenswood substation.

A. Reliability Evaluation

1. Reliability Standards

The ISO's reliability criteria incorporate national North American Electric Reliability Council (NERC) and regional Western Electricity Coordinating Council (WECC) planning standards as well as local reliability criteria, in particular, certain modifications for the San Francisco Peninsula Area. The ISO Grid Planning Standards include reliability criteria for the forecasted operation of the transmission system for several scenarios or categories of system conditions, as follow:

· Category A (base case). Normal ratings of equipment must not be exceeded with all generators, lines, and transformers in service and with no loss of load.

· Category B. Emergency ratings of equipment must not be exceeded with the loss of (a) a single circuit, generator, or transformer or (b) a single circuit and a single generator. Loss of load is not permitted unless the ISO Board of Governors decides that a capital project alternative is clearly not cost effective.

· Category C. Emergency ratings of equipment must not be exceeded with the loss of (a) a single circuit, generator, or transformer, or (b) a single circuit and a single generator; with that loss followed by manual adjustments and then the loss of another single circuit, generator, or transformer. Loss of load is allowed unless the ISO Board of Governors decides that the capital project is clearly cost effective.

The ISO's San Francisco Greater Bay Area Generation Outage Standard⁸ modifies the Category A base case to require that normal ratings of equipment must not be exceeded with three units out of service: the largest single unit on the San Francisco Peninsula, one 50 MW combustion turbine on the San Francisco Peninsula, and one 50 MW combustion turbine in the Greater Bay Area but not on the San Francisco Peninsula.⁹

In addition, PG&E and the ISO apply grid planning criteria called the Supplementary Guide for Application of the Criteria for San Francisco. This Supplementary Guide, which pre-dates the ISO, requires that emergency ratings of equipment not be exceeded, with no loss of load, under four specific sets of conditions. PG&E and the ISO consider the Supplementary Guide to be a modification of Category C requirements.

280 CCC takes issue with the Supplementary Guide, stating that it was developed by PG&E, not the ISO, and is significantly more stringent than the ISO's planning standards in that the Supplementary Guide does not allow loss of load in certain contingencies under which the ISO's Category C would allow outages. PG&E rebuts 280 CCC in this regard, pointing out that under Category C, involuntary load interruptions are not acceptable if the ISO Board has decided that the capital project being considered is clearly cost effective. PG&E asserts that, in approving the Jefferson-Martin project, the ISO found it to be cost effective.¹⁰ In its view, loss of load is thus not allowed under either the ISO Category C criteria or the Supplementary Guide criteria.

280 CCC contends that the contingency event modeled by PG&E to determine compliance with the Supplementary Guide reliability criteria has a miniscule probability of occurrence, which it estimates to be less 0.0000000257. PG&E responds that it is required to meet the standard, regardless of whether the contingency is a low probability event. PG&E explains that the established criteria are not based on probabilities of contingency events occurring, but on the reality that if a cable failure were to occur, it could be out up to several weeks.

280 CCC point to the ISO's February 7, 2002 Planning Standards, which include a new transmission standard that considers the likelihood of certain contingencies occurring for planning purposes. PG&E responds that quoted standard deals with the preparation of annual transmission expansion plans and is inapplicable in this case. The ISO maintains that it and PG&E are compelled by § 345 to plan the grid in accordance with national and regional reliability criteria which are deterministic, that generally accepted probabilistic standards do not yet exist, and that 280 CCC does not propose any in this proceeding.

2. Generation Capacity

There are currently three major generation facilities in the Project Area: PG&E's Hunters Point power plant, Mirant Corporation's (Mirant) Potrero power plant, and United Airlines' cogeneration facility. Hunters Point and Potrero both have steam turbines (Hunters Point Unit 4 and Potrero Unit 3) and combustion turbine units. Current generation capacity in the Project Area includes the following:¹¹

Generating Unit Net Rating

(MW)

Hunters Point Unit 1 50

Hunters Point Unit 4 163

Potrero Unit 3 207

Potrero Unit 4 50

Potrero Unit 5 50

Potrero Unit 6 50

United Co-gen 28

Total 598

In evaluating need for the Jefferson-Martin project, all parties include the existing Potrero units in the resource mix; they exclude the previously-planned Potrero Unit 7 since Mirant has withdrawn its Application for Certification at the CEC. Parties disagree regarding the continued operation of Hunters Point units and inclusion of planned CCSF combustion turbines.

Hunters Point. In assessing the reliability need for the Jefferson-Martin project, PG&E and the ISO assume that both units at Hunters Point will be shut down by the end of 2005. The ISO has stated, however, that if Jefferson-Martin is not operational by the end of 2005 or if new generation has not come on line, it would require Hunters Point units to remain on-line under a "reliability must run" (RMR) contract.

Under the terms of a 1998 settlement agreement with CCSF, PG&E is obligated to "permanently shut down the Hunters Point Power Plant as soon as the facility is no longer needed to sustain electric reliability in San Francisco and the surrounding area and the Federal Energy Regulatory Commission (FERC) has authorized PG&E to terminate PG&E's Reliability Must Run contract for the facility." The Commission approved this agreement in D.98-10-008. The ISO and PG&E assert that inclusion of Hunters Point in the supply forecast would defeat the intent of the settlement agreement, because its inclusion would delay the perceived need for an alternative resource. The ISO maintains that new resources are built to attain planning goals, which, in the ISO's view, include closure of both units at Hunters Point.

Another concern with Hunters Point arises because the Bay Area Air Quality Management District (BAAQMD) will implement decreasing nitrogen oxide (NOx) emission limits beginning on January 1, 2005. For Unit 4 (constructed in 1958) to continue operations, PG&E must either undertake a $15 million retrofit to install Selective Catalytic Reduction equipment or obtain Interchangeable Emission Reduction Credits (IERCs) from BAAQMD. PG&E has received IERCs for Unit 4 for use through 2005 and states that, if necessary, it will seek additional IERCs to keep Unit 4 operational beyond 2005. If Unit 4 continues to operate, it is expected that all available IERCs would be consumed by the end of 2008. Unit 1 (constructed in 1976) will meet the new NOx standards, although other BAAQMD regulations limit its operation to no more than 877 hours per year.

The ISO argues that both Hunters Point units should be excluded from the supply forecast for environmental, economic, and mechanical considerations, in addition to the settlement agreement. It states that Hunters Point Unit 4 is at or beyond the useful life of generating units of similar vintage and type and is six times as likely to suffer a forced outage than the general generation portfolio in the ISO control area, while Hunters Point Unit 1 is approximately three times more likely than average to be offline. The ISO expects that Hunters Point would require significant and increasing investment to continue operations.

CARE supports the closure of both units at Hunters Point. CARE submits that Hunters Point disproportionately affects the health and well being of San Francisco's Bayview Hunters Point neighborhood. CARE explains that this neighborhood has the highest pollution emissions in the city and the highest asthma hospitalization rate-twice the citywide average. CARE maintains further that the Hunters Point plant has degraded the Bay ecosystem and is a contributor to light pollution in the area.

280 CCC states that it shares the goals of other parties in this proceeding to shut down Hunters Point Unit 4. 280 CCC maintains, however, that generation should continue to be available from Unit 1 and, if necessary, could be available from Unit 4 through 2008. ORA believes that it is reasonable to assume that at least Unit 4 will likely not operate beyond 2005.

CCSF turbines. The State of California received four 45 MW gas turbines as part of a settlement with Williams Energy Company and has made these turbines available to CCSF for siting within the San Francisco/Peninsula areas. ORA and 280 CCC argue that the CCSF turbines should be counted among the available resources in analyzing need for the Jefferson-Martin project. PG&E, the ISO, and CCSF disagree.

CCSF reported that it was still in the process of identifying possible sites for the turbines and had not filed an Application for Certification with the CEC. Residents of southeast San Francisco oppose siting the turbines in their neighborhoods.¹² CCSF states that, particularly in light of this opposition, it is not certain that the turbines will be sited and constructed.

ORA and 280 CCC believe that the combustion turbines likely will be sited and operational by 2006, based on the value of the turbines, CCSF's power purchase agreement with the California Department of Water Resources, and CCSF's stated intent to file an Application for Certification at the CEC by March 2004. ORA points out further that CCSF's Electricity Resource Plan includes the addition of 150 MW of new generation by 2004 and 250 MW by 2008. The CCSF turbines could provide 180 MW of that capacity.

PG&E and the ISO assert that exclusion of the CCSF turbines from the resource mix is consistent with ISO Grid Planning Committee Guidelines, the Commission's Valley-Rainbow decision, and prudent transmission planning principles. According to the ISO committee, in five-year planning cases, only generation that is under construction with a planned in-service date within the five years should be considered. In ten-year planning cases, new generation projects that have received regulatory approval may be included. PG&E and the ISO point out that in the Valley-Rainbow proceeding the Commission refused to assume that proposed generation projects that had not completed the regulatory approval process would "come on-line for purposes of our evaluation of reliability" (D.02-12-006, mimeo. At 34-35).

3. Transmission Capacity

In its power flow analyses, PG&E included transmission projects in the Project Area that either have ISO approval or are minor projects that do not require ISO approval and that are expected to be in place by 2006. The ISO includes all transmission projects that have been approved by the ISO and are included in PG&E's annual expansion plan for 2003.

280 CCC notes that PG&E has re-rated some but not all of the San Mateo-Martin 115 kV lines and has not assigned emergency ratings to any of these lines, although other PG&E transmission facilities have such emergency ratings. 280 CCC suggests that the design carrying capacity of the overhead portions of each of these lines could be increased to 261 MVA using a four-foot-per-second wind speed and series reactors to balance the load. It suggests that the line ratings for the underground "dips" where the lines cross a highway could be adjusted to 231 MVA, or 261 MVA under emergency conditions. 280 CCC concludes that re-rating these lines would increase the calculated LSC into the Project Area and that assigning emergency ratings would further increase the LSC. PG&E responds that the use of series reactors to balance loads as 280 CCC suggests is experimental and problematic. PG&E also explains that the underground "dips" could not be given an emergency rating of 261 MVA because they are built with a high-pressure gas pipe inside another larger steel pipe and without sufficient ventilation to dissipate the added heat generated by a higher loading.

WEM joins 280 CCC in expressing concern regarding the capacity of existing lines in the San Mateo-Martin corridor. WEM asserts that, if emergency ratings were established at 115% of normal, the corridor transmission system is capable of serving 1225 MW of load. PG&E responds that the 115% figure is hypothetical and that WEM provided no evidence showing it is prudent to assume, as WEM does, a 100% utilization of all transmission lines.

PG&E submits that, even if the San Mateo-Martin lines could be re-rated as 280 CCC and WEM hypothesize, the suggestion that such re-rates should be assumed in analyzing need for the Jefferson-Martin project ignores the fact that no such re-rates are currently planned. PG&E cites D.02-12-066 as recognizing that prudent resource planning does not assume transmission upgrades that have not occurred and are not planned.

4. Distributed Generation, Energy Conservation, and Demand Response Programs

PG&E explains that its load forecasting methodology relies on historical load data, which reflects the absence of demand resulting from distributed generation, energy efficiency, energy conservation, and demand management programs. It maintains, therefore, that the effect of these factors based on growth rates consistent with past growth is reflected in the system load forecasts. Except for conservation, future programs are not explicitly included in the forecasting process. The ISO agrees that PG&E's methodology is reasonable.

280 CCC and WEM assert that distributed generation, energy efficiency, energy conservation, and demand response programs were not appropriately considered in assessing need for the Jefferson-Martin project. WEM advocates a rapid increase in locally-based renewables, which WEM states would avoid the need for new transmission.

280 CCC references goals in CCSF's Electricity Resource Plan of achieving more than 180 MW of new energy efficiency and distributed generation in 2004 and more than 420 MW by 2008, and asserts that, because these amounts are part of a new initiative, they would not be reflected in historical load growth data. 280 CCC also points to the Energy Action Plan's goals of increasing conservation and energy efficiency and meeting energy needs first by renewable energy resources and distributed generation.

The ISO replies that the goals in the Electricity Resource Plan are commendable, but wholly speculative. PG&E states that it is not prudent transmission planning to assume that load reduction will occur at levels not proven out in the historical data. PG&E argues in addition that it is consistent with the ISO's guidelines and the Valley Rainbow decision to exclude consideration of possible, but unknown, distributed generation and other projects from supply forecasts.

5. Load Forecasts

PG&E presented three load forecast scenarios for the Project Area: a "high" load forecast developed in September 1999, a "medium" forecast developed in December 2000, and a "low" forecast prepared in August 2002. These forecasts were based on 1-in-10 year weather conditions,¹³ and PG&E updated them during the proceeding based on 2002 peak load data. PG&E prepared a new "low" forecast and recalibrated the other two forecasts to the temperature-adjusted 2002 peak load. The updated "medium" and "high" forecasts contain the previously forecasted growth patterns for years after 2002. PG&E submits that it is prudent to use all three load forecasts in assessing need for the Jefferson-Martin project, and to assign equal probabilities to each of the load growth scenarios.

PG&E's updated "low" load growth forecast is the same forecast used in PG&E's March 2003 Electric Grid Expansion Plan. The ISO states that it independently validated the reasonableness of this forecast in its 2003 grid planning process. The ISO based its analyses and testimony on PG&E's March 2003 forecast results, but for the San Francisco Peninsula Area since it analyzes need within that broader area. The March 2003 load forecasts for the Project Area (Exhibit 4) and the total San Francisco Peninsula Area (Exhibit 171) are summarized in Table 1.

Table 1

Load Forecasts for PG&E's Project Area

and the San Francisco Peninsula Area

(MW)

Project San Francisco

Year Area Peninsula Area

2003 1243 1857

2004 1266 1882

2005 1285 1915

2006 1308 1949

2007 1329 1978

2008 1349 2005

2009 1365 2027

2010 1381 2050

2011 1396 2070

2012 1412 2092

PG&E's "medium" forecast exceeds the "low" forecast in the Project Area by 80 MW in 2006; its "high" forecast exceeds the "low" forecast for that year by 103 MW. The record does not contain comparable "medium" and "high" forecasts for the San Francisco Peninsula Area.

The ISO witness testified that PG&E's March 2003 forecast is the appropriate demand forecast to use in assessing Jefferson-Martin. In briefs, the ISO states that the March 2003 forecast may be conservative because it reflects potential load growth during a period of economic downturn and future load may exceed the forecast. The ISO now suggests that giving weight to the previous higher forecasts, as PG&E recommends, may be prudent.

280 CCC criticizes PG&E's load forecasts as routinely over-forecasting demand in recent years. To the extent the Commission relies on any of PG&E's load forecasts, 280 CCC supports use of the "low" forecast, since it is based on the most recent economic and household growth projections and longer-term effects of the energy crisis. 280 CCC maintains, however, that the "low" forecast still overstates future load growth, partly because it does not account for increases in distributed generation, energy conservation, and demand reduction programs. PG&E responds that, when the California economy recovers, the demand forecast will again change, with demand perhaps growing at or near the previous pace. PG&E points out that, if the 2003 peak is temperature-normalized, PG&E's "low" forecast actually under-predicted load growth in 2003 by 3 MW (which we note is about 0.2%).

6. Parties' Reliability Need Analyses

The parties used two different types of analyses in assessing the reliability need for the Jefferson-Martin project: LSC (load serving capability) studies and power flow analyses. LSC is the highest load level that can be served in an area by the electrical transmission system into the area and available generation within the area, without violating the relevant reliability criteria. Power flow analyses model the transmission system under specified load and contingency conditions to determine if any elements are overloaded and reliability standards violated.

In an LSC study completed in July 2003, the ISO examined the entire San Francisco Peninsula Area, not just the Project Area as defined by PG&E, explaining that transmission constraints "downstream" could limit the LSC regardless of the capability of the transmission system closer to the load. The ISO's LSC study applied the ISO's reliability criteria, including the San Francisco Greater Bay Area Generation Outage Standard and the Supplementary Guide as described in Section IV.A.1.

The ISO's LSC study analyzed and provided LSC results for 37 different scenarios. Table 2 summarizes scenarios that are most relevant to this proceeding. These scenarios assume that certain base case upgrades¹⁴ as well as certain re-rates and upgrades south of the San Mateo substation¹⁵ are undertaken, unless indicated otherwise.

Table 2

Selected ISO Load Serving Capability (LSC) Study Scenarios

SF Peninsula

No. Description Area LSC

02 Hunters Point (HP) retired, only base case upgrades 1596

03 HP operational, only base case upgrades 1971

11 HP retired, Jefferson-Martin (J-M) operational 1536

12 HP & J-M operational 2081

14 HP retired, J-M & internal cable projects (ICP) operational 2101

15 HP, J-M & ICP operational 2121

26 Present day: HP operational, no upgrades 1876

28 HP #4 retired; HP #1 operational 1731

29 HP #4 retired; HP #1 and ICP operational 1811

33 HP #4 retired; HP #1 & J-M operational 1666

34 HP #4 retired; HP #1, J-M, & ICP operational 2106

The ISO's LSC study identified and discussed that the ability of the Jefferson-Martin project to contribute to the LSC of the San Francisco Peninsula Area is limited by current transmission constraints south of the San Mateo substation and within CCSF's 115 kV cable system.¹⁶ It concludes that reinforcements of both the transmission system south of the San Mateo substation and the 115 kV cable system within San Francisco are needed if the Jefferson-Martin project is to be used to reduce the amount of generation within San Francisco.¹⁷

To determine the maximum potential increase in LSC that could be obtained due to the Jefferson-Martin project, the LSC study undertook a separate analysis focused on only the San Mateo-Martin corridor. That separate analysis indicated that, with Hunters Point Unit 4 shut down, the Jefferson-Martin project could add up to 351 MW of capacity to the San Mateo-Martin corridor if all relevant transmission constraints to the north and south were removed. The ISO reports that solutions to these transmission constraints are being planned and have been incorporated in PG&E's 2003 transmission expansion plan. PG&E reports it is pursuing the re-rating of internal 115 kV cables within San Francisco and a new Hunters Point-to-Martin cable. In addition, PG&E has begun the preliminary step of asking ISO approval of a new Martin-to-Mission cable. PG&E states that this new cable project may not be needed until 2011, but that it plans to proceed with permit acquisition in case the project is needed earlier.

In its testimony in this proceeding, the ISO undertook and reported two additional LSC calculations. The additional analyses differ from the July 2003 LSC study in that they include two additional transmission upgrades south of the Martin substation that are in PG&E's 2003 Expansion Plan. Both of the new LSC calculations assume that Hunters Point is retired; one assumes that the Jefferson-Martin project is not built whereas the other assumes it is built. The ISO compares the LSC results without Jefferson-Martin (1862 MW) to the San Francisco Peninsula Area load forecast for 2006 (1949 MW) and concludes that the Jefferson-Martin project is needed by the end of 2005 to ensure that the projected load can be served reliably. The ISO's second calculation, which includes Jefferson-Martin and assumes that there are no transmission constraints within San Francisco,¹⁸ indicates an LSC of 2092 MW, which would be sufficient to meet the forecasted area demand through 2012 if all transmission constraints within San Francisco were removed.

PG&E ran power flow analyses for the Project Area using ISO reliability criteria and 2006's "high," "medium," and "low" load forecasts, both with and without the Jefferson-Martin project. PG&E assumed that the Hunters Point units would be retired by the end of 2005 and that the CCSF combustion turbines would not be constructed by then. PG&E reports that its power flow analyses indicate that the Jefferson-Martin project would be needed in 2006 to avoid overload conditions, for all three load forecasts.

ORA states that scenarios 32 and 34 in the ISO's LSC study mirror ORA's assumptions, including that Hunters Point Unit 4 is retired and Unit 1 remains operational. ORA concludes that the CCSF turbines would provide a solution to the reliability problem in the near term, through 2006-2008, while Jefferson-Martin would provide a longer-term solution beyond 2008. ORA is concerned that, should both the CCSF turbines and Jefferson-Martin come online by 2006, which it believes is likely, ratepayers will overpay for reliability since either project would meet near-term reliability needs in the area.

280 CCC submits that PG&E's own calculations indicate that, if Hunters Point Unit 4 (but not Unit 1) is retired and the CCSF turbines are operational, Jefferson-Martin would not be needed under PG&E's "low" forecast through at least 2012. Additionally, if both Hunters Point Units 1 and 4 remain operational (but the CCSF turbines are not built), there would be no need for the Jefferson-Martin project until 2014 under PG&E's "low" forecast, 2009 using the "medium" forecast, and 2008 using the "high" forecast. 280 CCC emphasizes that these conclusions are reached using PG&E's planning contingencies and load forecasts, both of which 280 CCC contests.

280 CCC also conducted its own power flow analyses for the Project Area using PG&E's planning contingencies and "low" load forecast for 2006. 280 CCC' power flow studies indicate that, without the Jefferson-Martin project, PG&E would not experience overload conditions in 2006 if Hunters Point Unit 4 is retired and either Unit 1 remains in service or the four CCSF combustion turbines are put into service. PG&E and the ISO point out, however, that 280 CCC' power flow studies show a 99.7% loading on one 115 kV circuit in the scenarios assuming Hunters Point Unit 1 remains in service, which they assert is not a reassuring margin of safety.

B. Delay or the No Project Alternative

The scoping memo required PG&E to describe its plan of action should the Jefferson-Martin project not be completed by September 2005. PG&E responds that its plan of action would depend on whether the Commission approves the project and there are simply construction-related delays, or denies the CPCN. If there are construction-related delays, PG&E expects that the ISO would require PG&E to delay shut down of Hunters Point until the Jefferson-Martin line becomes operational. If the Commission declines to issue a CPCN for the project, PG&E expects that the ISO would require PG&E to continue running Hunters Point until new generation is constructed within the Project Area. PG&E states that it would have to consider whether to attempt to obtain IERCs available for Unit 4 up through 2008 and/or commence the process to retrofit Unit 4's emission control equipment.

In accordance with CEQA requirements, the EIR evaluates the No Project Alternative in addition to route alternatives. In essence, this alternative examines environmental impacts if a project is not approved and built.

The FEIR states that its No Project alternative is based primarily on an April 18, 2003 letter from the ISO to PG&E and CCSF, which identified future requirements that would allow retirement of Hunters Point Unit 4. The requirements in that letter do not include construction of the Jefferson-Martin project. Consistent with that letter, the components of the FEIR's No Project Alternative include the following:

· Generation. The four CCSF turbines would be installed and Hunters Point Unit 4 would be closed.

· System Upgrades. Re-rating and upgrading of certain transmission lines, and installation of a new transformer would occur.

· System Improvements. PG&E system improvements would be made, including the conversion of San Mateo-Martin #4 from 60 kV to 115 kV and the installation of a Potrero-Hunters Point 115 kV underground cable.

· System Management and Planning. PG&E and the ISO would continue to implement an Interruptible Load Program, demand-side management would be encouraged. Curtailment of electric service could be required in worst-case demand growth scenarios.

· Special Protection Schemes. Continued and increased use of Special Protection Schemes would be needed to provide for controlled involuntary load curtailment during "high load" operating conditions.

The FEIR concludes that the environmental impacts of the No Project Alternative would result primarily from operation of gas-filed turbine generators, and would include substantial air emissions and ongoing noise near the generators as well as visual impacts of the generators depending on their locations. In addition, the No Project Alternative has the potential to result in electric service disruption and curtailments, which would increase use of back-up diesel generators, resulting in additional pollutant emissions. The FEIR concludes that the project alternatives it identifies as environmentally superior alternatives are preferred over the No Project Alternative.

CARE and 280 CCC address the FEIR's No Project Alternative. CARE agrees with the FEIR's conclusion that the No Project Alternative is environmentally inferior to the Jefferson-Martin project. It maintains that the CCSF combustion turbines would be a less efficient alternative to the Jefferson-Martin project. In addition to environmental concerns cited in the FEIR regarding the No Project Alternative, CARE cites thermal pollution in the Bay caused by continued operation of power plants in San Francisco. 280 CCC disputes the FEIR's analysis that the No Project Alternative would result in significant impacts due to emissions, arguing to the contrary that the new CCSF turbines would replace the polluting Hunters Point Unit 4.

C. Discussion

In D.98-10-029, the Commission approved a settlement agreement between PG&E and the CCSF which allows PG&E to shut down Hunters Point as soon as it is no longer needed to sustain electric reliability in San Francisco and surrounding areas. There also is a restrictive covenant attached to the deed that prohibits the site from ever again being used for power generation. In concert with legislative action (Budget Act of 1998)¹⁹, a City of San Francisco resolution (98-0181) and community interests, we find that the addition of Jefferson Martin and other transmission reinforcements will facilitate the eventual closure of the Hunters Point power plant. In the spirit of the PG&E/CCSF settlement agreement, we agree with PG&E and the ISO that Hunters Point should be closed at the earliest possible date. To facilitate this goal, we find that Jefferson Martin should be online at the earliest possible time, as soon as construction can be completed. With other transmission reinforcements in place, we believe that the sooner Jefferson Martin is operational the sooner PG&E will be able to decommission Hunters Point. Accordingly, we find that there is an adequate record to support a conclusion that the Jefferson-Martin 230 kV project is needed pursuant to § 1001. We conclude that we should grant a CPCN to PG&E to construct new 230 kV transmission facilities between the Jefferson and Martin substations. Section VII outlines the location and routing of the approved project.

One of the most contentious issues in this proceeding related to the need for whether or not the Jefferson-Martin project will allow the Hunters Point units to be retired. PG&E and the ISO, supported by CARE, view the closure of Hunters Point as essential. On that basis, they conclude that the Jefferson-Martin project is needed and should be operational by late 2005 in order to allow PG&E to provide reliable service.

We agree with both PG&E and the ISO that the Hunters Point Units 1 and 4 should be shutdown as soon as technically possible. The ISO has established that Hunters Point Unit and 1 and 4 are beyond their useful lives and are highly likely to suffer a forced outage. The ISO expects that Hunters Point would require significant and increasing investment to continue operation.

There are several environmental concerns that arise from the continued operation of the Hunters Point facility. For example, Hunters Point Unit 4 must comply with stricter emission limits at the beginning of 2005. While PG&E has received IERCs for Unit 4 for use through 2005 and could obtain additional IERCs if necessary, we believe that continued reliance on these power generation sources is not environmentally prudent or cost-effective. There are several environmental benefits that will occur when Hunters Point is decommissioned, such as reduced air, noise, and thermal pollution. Such an outcome will be consistent with the community values of the Bayview and Hunters Point neighborhoods, as CARE has testified. We take note that shutting down the power plant may produce beneficial health impacts for the Bayview Hunters Point neighborhood. We give these concerns great weight in balancing the competing interests in this proceeding.

We have established that both Hunters Point units should be excluded from the supply forecast because it is consistent with a past Commission decision, local government and community interests, and because of environmental as well as operational considerations. To meet the forecasted demand, we believe that the Jefferson Martin project (along with other transmission reinforcements) will need to be online prior to decommissioning Hunters Point (Table 2, scenarios 11 and 14; Exhibit 165, Attachment 2). Increased reliance on transmission rather than local generation to meet San Francisco's power needs would create transmission constraints within San Francisco. Thus, the Jefferson-Martin project by itself, although necessary, is not sufficient for closure of Hunters Point. The ISO's LSC study indicates that a combination of the Jefferson-Martin project and three internal cable projects when constructed may be sufficient to meet load reliably (Table 2, scenario 14). We encourage PG&E to construct Jefferson Martin as soon as possible; given the approximate two year construction period for Jefferson Martin, we fully expect the project to be operational by mid-2006. With other transmission reinforcements completed, we believe Hunters Point 1 and 4 can be retired and decommissioning may commence.

The Jefferson Martin project also is needed for reliability purposes because it will diversify the path and source of power brought into San Francisco. Currently, all major transmission lines importing power travel through a single corridor from the San Mateo substation to the Martin substation and receive power from the San Mateo substation. Thus, the system is vulnerable to events disrupting supply at the San Mateo substation and/or along the San Mateo-Martin corridor. We recognize that a Jefferson-to-Martin route for the project does not diversify the risk of loss of load due to equipment outages at the Jefferson substation, since all lines would continue to travel through that substation. However, to use 280 CCC's terminology, a Jefferson-Martin line would eliminate the "choke point" at the San Mateo and Martin substations, as well as the sole reliance on the San Mateo-to-Martin corridor. While the San Mateo substation brings in power from the East Bay, the Jefferson-Martin project will tap power originating from the region south of the Peninsula area, thus diversifying the source of power.²⁰ Hence, prompt construction of the project will allow these diversification benefits to be reaped sooner rather than later.

Coupled together, the Jefferson-Martin project and the closure of Hunters Point will also provide economic benefits. In addition to a fixed revenue requirement for Hunters Point, ratepayers pay above-market variable costs in response to RMR dispatch notices. To the extent power imported over the Jefferson-Martin line is from more cost-effective generators, there will be an economic benefit from the project. In addition, continued operation of Hunters Point Unit 4 may require new emissions control equipment if IERC credits are not sufficient. PG&E anticipates there may also be other retrofit costs needed to keep Hunters Point operating much beyond the end of 2005. PG&E suggests additionally that the Jefferson-Martin project may allow the ISO to reduce the RMR requirement for Potrero. As the ISO points out, construction cost increases may also act to offset any purported cost savings from deferring construction.

We reiterate our support, as expressed in the Energy Action Plan, for increased reliance on conservation, energy efficiency, and distributed generation as high-priority ways to meet California's energy needs. At the same time, there is no convincing evidence in the record that the near-term development of such resources is sufficiently certain to affect the need analysis for the Jefferson-Martin project. We find that PG&E has reflected these resources satisfactorily in its load forecasting methodology.

The ISO reviewed PG&E's forecasting methodology in its most recent annual transmission planning process, and has found this forecast to be reasonable and sufficient for its own transmission planning purposes and for its testimony in this proceeding. We find it reasonable to use PG&E's March 2003 load forecast, contained in Table 1 above, in assessing need for the Jefferson-Martin project.

We adopt the ISO's method because it assesses the LSC based on the entire San Francisco Peninsula Area. The ISO has made a convincing showing that the ability to serve electric load within San Francisco is affected by transmission facilities within the entire San Francisco Peninsula Area and also by transmission facilities connecting the Peninsula to the greater Bay area. The ISO's LSC study and its testimony provide the only reliability assessments in the record that takes this broader view.

In conjunction with the San Francisco Peninsula Area load forecasts (Table 1), the ISO's LSC results for the San Francisco Peninsula Area allow us to make findings regarding reliability need for the Jefferson-Martin project. In consistent with D. 98-10-029, we find that it is no longer prudent or cost-effective to continue operation of Hunters Point (Units 1 and 4). This conclusion coupled with the projected load increases and the results from the LSC study allow us to arrive at a conclusion that Jefferson Martin is needed and can provide necessary reliability by mid-2006. We find further that there would not be enough time for other alternatives such as a trans-Bay transmission line, as some parties have suggested, to be planned, permitted, and constructed to meet this reliability need.

Inclusion of the four CCSF combustion turbines in the resource mix used to assess need for the Jefferson-Martin project would not be consistent with the ISO's guidelines for either five-year or ten-year planning cases, since they have not received regulatory permits. We take official notice of information on the CEC's website indicating that an Application for Certification was filed on March 18, 2004 (CEC Docket No. 04-AFC-1) for three of the four turbines. In light of the on-going controversy about the turbines and the early stage of their certification process, we do not have sufficient confidence that the three CCSF combustion turbines subject to that application will be constructed in a timely enought fashion to warrant deviation from standard industry practice and include them in the resource mix used to evaluate need for the Jefferson-Martin project. We have no information regarding the fate of the fourth CCSF combustion turbine.

⁸ The ISO explains that it uses the standard because of the unusually large concentration of generation units in the greater Bay area and the fact that historical forced outage rates for units in the Bay area are significantly higher than the industry averages for similar units.

¹⁰ Regarding the determination of cost-effectiveness, PG&E states that it submitted "decision quality cost estimates" to the ISO on April 4, 2002 before the ISO Board approved the Jefferson-Martin project. At that time, PG&E estimated the cost to be $173 million for the Proposed Project and $213 million for the AUA.

¹¹ 280 Citizens reported ISO data showing a combined capacity of 635.5 MW for these plants. However, that total represents maximum capacity under ideal temperature conditions, which are unlikely to occur during peak load conditions. We agree with PG&E and the ISO that the maximum capacity levels should not be used for transmission planning purposes.

¹³ The ISO's Grid Planning Standards require that transmission studies addressing local load serving concerns utilize a 1-in-10 year extreme weather load level, whereas studies focusing on regional facilities, i.e., major interties, may use a less stringent 1-in-5 year extreme weather load level. The ISO views more rigorous local area requirements as necessary because fewer options exist to mitigate performance concerns.