III. Analysis Of Long Term Procurement Plans

A. Do The LTPPS Integrate The Commission's Direction From Other Related Proceedings And Meet The Criteria Established In The ACR/Scoping Memo?

1. General Assessment

PG&E, SCE and SDG&E each used its resource plan to inform the procurement decision, rather than to select a deterministic set of resources or to identify specific procurement actions. The IOUs interpreted the directions for scenarios as preparing background for, and illustrations of, their procurement strategies. The resource scenarios demonstrate the impact of key uncertainties and how resource plans can be structured to deal with these risks. The utilities request procurement and cost-recovery rules, not a preferred resource list. The utilities vary, or are unclear, about adoption of a Residual Net Open level as a floor or ceiling for procurements.

After existing resources and policy preferred resources have been compared to load and necessary reserves, the result is the amount of energy and capacity which a Load Serving Entity must still acquire. This is called either "need" or the "net open" position, sometimes subdivided into "net short" and "net long." Actual forecasts of net open capacity and energy were contained in confidential filings, so discussion in the testimony and hearings is both limited and general.

2. Load Forecast

The load forecasts of all three utilities was largely unchallenged. Appendix A to the June 4, 2004, ACR directed the IOUs to prepare resource scenarios as follows:

The Medium-Load Plan Scenario. The medium-load plan is to be the preferred resource plan of each utility that meets the needs identified in its Alternative Base Case load-forecast scenario or, if the utility does not choose to file an Alternative Base Case load-forecast scenario, its IEPR-CEC base case scenario. This Plan is to be a utility's best estimate of how it would prepare to meet the needs it believes ultimately will come to be. Though it is not necessary, or even possible, for utilities to specify in detail the placement of new generation facilities that may be needed up to ten years in advance, nor is it possible to indicate the specific paths of transmission additions or upgrades, it is appropriate that the utilities be more specific than they were in the Long-Term Plans submitted in 2003.

In addition, it would be appropriate to include alternative versions of this Plan reflecting different resource development options, reflecting differing expectations about the desirability of in-service-area generation, new transmission, and different fuel types.

High-Load Plan Scenario. The High-Load Plan is not to be an extreme case that has little chance of coming to pass. Rather, it should be a reasonable guess at how great the burden of service could become under high, but not unreasonable assumptions about future load growth. The Plan should be based on the assumption of greater than expected economic growth, resulting in higher load growth, assumption of a modest core-noncore load loss beginning only in 2009, and a modest development of CCA, also beginning in 2009. The utilities should assume that current levels of DA will continue throughout the time horizon.

Low-Load Plan Scenario. The Low-Load Plan similarly, is not to be an extreme example of conservation and changed priorities of Californians. Rather, it should be based on reasonable but pessimistic assumptions about the economy and on generous assumptions about the development of core-noncore impacts and CCA. Assume aggressive CCA development beginning in 2006, and an aggressive core-noncore scenario from the choices discussed above. Again, assume the continuation of DA service at current levels.

Although all three IOUs relied on different assumptions in modeling their medium case and in setting floors and ceilings for the high and low scenarios, for the most part the three LTPPs complied with the resource scenario request. The differing assumptions made cross-utility comparisons difficult, but each LTPP taken on its own provided a reasonable range of scenarios as boundaries of risk.

In reviewing the resource scenarios in the LTPPs, each intervenor brought a particular perspective to analysis of the plan that tended to color the evaluation. And to further complicate the resource scenario evaluation is the fact that it is sometimes difficult to integrate differing "load forecasts" with resource scenarios and to understand the interaction between the two concepts. For example, intervenors concerned with departing load are concerned that the IOUs are over resourced in general and that could lead to stranded costs if/when there is departing load. Other intervenors focused on whether the resource scenarios plan for sufficient RPS, EE, DR and DG.

PG&E asserts that it complied with the directions of the ACR, and that no party directly challenged PG&E's reference case (i.e. service area forecast) or its high, medium and low forecasts²⁰. In its medium case, "PG&E assumed that three percent of its current customers with load under 500 kW will begin to migrate to community choice aggregation in 2006, and the rate of loss to this market will increase by one percent annually, reaching 10 percent in 2013²¹. PG&E also assumed implementation of a core/noncore market structure beginning in 2007 and that 50 percent of noncore customers with load above 500 kW who are not already direct access service customers will depart from PG&E service²².

SCE contends that its forecasts are reasonable and that they comply with the ACR requirements. Since SCE's medium case, its preferred case, did not include any CCA or core/non-core, it was the focus of most discussion. SCE chose a forecast that was "consistent with Edison's current load forecast, without knowledge of what might come.²³" This is also the forecast used in SCE's 2006 General Rate Case, adjusted for expanded energy efficiency. SCE's low load case assumes low economic growth and aggressive departing load.

SDG&E asserts that its (area) load forecast was unchallenged in this proceeding. "The medium-load plan represents SDG&E's best estimate of the resources needed to reliably serve its customers, and it is based on a load forecast that does not show any loss of load to a core/noncore split or CCA implementation. Given the uncertainty surrounding the timing and magnitude of emerging rules for CCA, core/noncore, or reinstatement of direct access and the potential resulting outcomes, the medium-load plan is best suited to meet the expected need absent firm, enforceable commitments and other final details to assess departing load models." ²⁴

All three IOUs included current levels of direct access throughout the planning horizon and did not plan for the return of self-generation customers.

B. Position of Parties on Load Forecasts

ORA conducted a thorough review of the service area load forecasts, noting that the growth rates are similar to those in the CEC's 2003 IEPR, but adjusted to fit the higher actual growth in 2002 and 2003. ORA found the service area load forecasts reasonable²⁵. ORA also examined the differing departing load scenarios and recommended that the IOUs use ORA's common set of departing load assumptions.

CMTA/CLECA found the IOU medium case differences in the treatment of departing load sufficiently troublesome to ask that the Commission direct parties to rerun their scenarios using a common set of assumptions. They also recommended that the Commission should have a low level of confidence in the medium case scenarios.²⁶

Calpine recommended that a 1-in-10 peak weather planning standard be used for all demand forecasts, as is required for local reliability transmission studies. This would add about 6 percent to the demand forecasts.

CCA asked for and received assurance from all the IOUs that existing load served by large co-generation was assumed to be continued to be served by self-generation and had not been included in the demand forecasts.

Several parties, such as Modesto Irrigation District, South San Joaquin Irrigation District, and the City of Chula Vista, asked that the load in their jurisdictions be removed from IOU demand forecasts, because they intend to serve the load themselves.

C. Discussion of Load Forecasts

The "service area" or "reference" medium forecasts presented by the IOUs in their LTPPs indicate reasonable growth trends and levels. The utilities use similar growth factors and are generally consistent with the IEPR forecast trends, except the levels are higher because they are updated from a 2001 baseline to a 2003 baseline. This update reflects the unanticipated economic recovery in 2002 and 2003 that was not reflected in the IEPR forecast.

The most obvious disparity between the IOUs' forecasts was in the area of assumptions about departing load for DA, core/non-core and CCA. PG&E does include departing load projections in its baseline forecast, where SCE and SDG&E do not. Potentially, PG&E's baseline could be too low, whereas the other IOUs' baselines could be too high. Parties representing potential departing load, and the energy marketers hoping to serve the departed load, questioned whether SCE and SDG&E's medium load scenario included sufficient assumptions about departing load.

The ACR required that the medium load forecast be the utility's preferred case and its best estimate of how it would prepare to meet the needs it believes ultimately will come to be. Since CCA has been set in statute and is the subject of an on-going CPUC implementation proceeding, it is reasonable that some CCA will start to occur in 2006. But, there was not sufficient evidence in this proceeding that CCA alone will have a material effect on IOU resource needs in the next few years.

The future of expanding direct access or creating a core/non-core market is more speculative. Direct access is currently suspended by legislation until the last DWR contract expires, currently scheduled for 2013. There is no record on which to base a choice on the probability that more retail competition will emerge.

As a consequence, we should take these demand uncertainty factors into account as one of the uncertainties affecting the level of acquisition and the need for flexibility in the resource plan. However, we are not going to adopt a fixed assumption regarding a most likely set of departing load. We acknowledge that the IOUs face considerable load variability risk, and will set policies accordingly.

We will not set a procurement cap based on the low cases, since this could seriously under-resource California's service areas during the planning period. Instead, we will rely on a portfolio approach and allow justification of specific contract types as the need arises. This will allow us to balance between obtaining adequate resources and not over-procuring in the case of departing load or crowding out of preferred resources towards the end of the planning period. We will monitor how the IOUs are doing on obtaining resources to meet their resource adequacy requirements on a forward looking basis.

We disagree with Calpine that all demand forecasting should switch to the 1-in-10 peak weather standard used for testing the robustness of local transmission systems. Existing resource planning uses average weather (1-in-2) and then adds a reserve margin which, in part, provides the cushion should hotter than average weather occur. This is the approach we adopted to implement our resource adequacy requirements.²⁷ Calpine's concern is already accounted for in existing practice.

In summary, although each IOU prepared its own LTPP using its own assumptions, each IOU asserts that its service area load forecasts are comparable to the IEPR, adjusted to more current data, and that their medium, preferred case is a reasonable basis for resource planning. We find all three LTPPs consistent with the 2003 IEPR, are reasonable for planning purposes and that the medium, preferred case should be followed for making planning and procurement decisions.

Adding another layer of complication to the review of the LTPPs is the fact that the utilities can only propose what characteristics would best fit Commission and direction and current circumstances, but only market-tested bids will actually produce a portfolio of specific resources. In this setting, planning is largely indicative, not deterministic. Some parties are concerned that the utilities will over subscribe to long-term contracts that will crowd out future opportunities. If the utilities have a mixed portfolio of different contract terms and lengths, this should be a manageable issue.

1. Resource Scenarios

Before reviewing load and resource assumptions, we need to set the stage by discussing the overall role of resource scenarios as a backdrop to the procurement plans. There are four principal sources of guidance regarding what this decision should direct the IOUs to do as a response to their long-term resource plans: the EAP; D.04-01-050;the April 4 R.04-04-003 OIR; and the June 16, 2004 ACR/scoping order as amended on June 29:

"The OIR is clear that the major focus is to review and adopt long-term procurement plans. However, the plans must be based on an integrated resource strategy that is consistent with Commission policy, reflects reasonable assumptions, and covers a rational range of scenarios."

D. Implementing the Energy Action Plan

The EAP contains explicit direction regarding the state's preferences for meeting identified resource needs and the IOUs are to prioritize their resource selections accordingly. As discussed earlier in the decision, the EAP "loading order" is as follows: energy efficiency and demand-side resources; renewable generation resources; efficient, clean fossil generation resources; upgrades and expansions to the transmission and distribution infrastructure; and customer- and utility-owned distributed generation. Sections of this decision describe the objectives and direction for aggressive procurement of renewable generation resources, contain guidance for procuring clean fossil resources and discuss transmission and DG, respectively. The direction is clear: IOUs should implement the EAP loading order when soliciting resources as a result of this decision.

1. Discussion on Compliance

Parties disagreed on whether the resource scenarios complied with the Commission's direction in the OIR and Scoping memo. Lingering behind these perceptions is the memory of the detailed resource assessments and specific direction which took place when utilities were monopoly resource suppliers. In a hybrid market, the utilities can propose which characteristics would best fit Commission direction and current circumstances, but only market-tested bids will actually produce a portfolio of specific resources. In this setting, planning is indicative, not deterministic.

E. Net Open Position

1. Position of IOUs on Net Open Positions

PG&E asserts that development of its Net Open position is reasonable. Based on the three scenarios PG&E developed, PG&E estimated the energy and capacity it will need to fill its net open position.

"For the first five years of PG&E's medium load scenario, PG&E's energy and capacity needs show little change because anticipated load growth and resource attrition are offset by projected load migration to the community choice aggregation and core/noncore markets. PG&E's energy and capacity needs begin to increase in the latter years of the 10-year planning horizon as the DWR contracts allocated to PG&E begin to expire."²⁸

In PG&E's high-load scenario, PG&E's energy and capacity needs become increasingly greater throughout the planning horizon. In PG&E's low-load scenario, its net open position grows longer during the first five years of the plan, but becomes increasingly shorter during the latter half of the planning horizon. ²⁹ PG&E's net open position is not affected by transmission additions, because PG&E did not propose any economically-driven transmission lines in its LTPP.

"SCE's current supply portfolio is dominated by long-term and baseload resource commitments. Such a portfolio results in SCE having excess supply that must be sold into the market.³⁰ There is a need for additional load-following and peaking resources.

Due primarily to the suite of grid reliability resources approved in D.04-06-011, SDG&E is essentially "fully resourced" through approximately 2009. Combined with efforts to achieve 20% of the energy mix by 2010 from renewable sources means SDG&E will primarily procure only renewable power until 2010. Nevertheless, SDG&E asks the Commission to take precise and specific action to address future needs identified in SDG&E's medium-load plan. Increased grid reliability needs, for example, appear in 2010 in the medium load plan due to load growth and limited in-basin generation. In addition, the presence of the DWR Sunrise contract in SDG&E's portfolio means that SDG&E does not have `headroom' until after 2010 to obtain further local reliability contracts.³¹

F. Discussion on Net Open

In summary, all three IOUs have capacity needs throughout the planning horizon. Capacity needs expand considerably in 2011, due to the expiration of most of the DWR contracts. All three IOUs are long on energy, primarily in the off-peak and shoulder hours, through 2009 (PG&E) and 2010 (SCE and SDG&E) until the bulk of DWR contracts expire. Because resources are `lumpy', adding preferred resources upon existing resources somewhat exacerbates this long position, requiring utilities to be energy sellers in many off-peak and shoulder hours.

The impact of these decisions is to reduce the amount of capacity needed in the 2010 medium case scenarios by 800 MW for PG&E and 1,500 MW for SCE, while increasing SDG&E's resources above the minimum reserve margin by 280 MWs.

This Commission favors openness in its decisions and in the information that market participants have in dealing with each other. Another section of this decision discusses specifically how we are responding to legislative direction on confidentiality matters. In this section we note that it is not the intent of the Commission to provide the means by which market power could be exercised against the LSEs and, hence, against electric service customers in California. Therefore, this decision does not present information about the current net open positions of the utilities. Nor do we provide the elements from which that information can be calculated. However, we will provide simplified tables based on projections of future resource balance information for the years 2007-2014 after those numbers have been refreshed from their initial filing back in July.

G. Implications of the Three Resource Scenarios

As set forth in the April 1, 2004 OIR, the purpose of the three resource scenarios was to "help the Commission understand how each utility intends to respond to a wide range of load scenarios. The focus is not on forecasts, but rather on the adoption of long-term plans that can accommodate many possible outcomes."³²

The IOUs filed their LTPPs, with resource scenarios, on July 9, 2004, almost four months before the Commission issued its decision on Reserve Margin Requirements/Resource Adequacy (RMR/RA) in D.04-10-035 on October 28, 2004. At the time the LTPPs were prepared, the IOUs and many intervenors were concerned with the utilities overprocurring resources-especially in the short-term. However, pursuant to the direction given by the Governor Schwarzenegger and President Peevey, and adopted to by a majority of the Commission, the current focus is on maintaining and enhancing grid reliability through accelerated reserve margin targets. When this goal is integrated with the directive from D.04-07-028 issued by the Commission this summer ordering the utilities to concentrate on near-term reliability, it is evident that the IOUs must increase and retain supply for the near future. We will try to balance grid reliability with our other primary public duty of protecting ratepayers from excessive charges and also be mindful of potential departing loads and stranded costs.

Therefore, in capsulizing the IOUs resource scenarios, we note that the IOUs did not have the benefit of the RA decision when the scenarios were prepared in July 2004 and it may be necessary to direct the IOUs to revisit and update their LTPPs to comply with the new reserve margin targets.

As an appropriate segue, in its LTPP, PG&E states that it is most concerned about the resource risks associated with customer load uncertainty and the risk of stranded costs due to excess procurements. "IOUs devoutly wish to avoid being "over-resourced," but procurement strategies based on short-term procurement and dependence on external suppliers have even greater risk, as the energy crisis demonstrated. PG&E's plan is designed to recognize these tradeoffs. The full implementation of Assembly Bill (AB) 57, as called for in last April's letter from Governor Schwarzenegger to President Peevey; the assumptions and conditions suggested in PG&E's integrated resource plan, including a request for long-term procurement based on the low case scenario and the possibility of a nonbypassable charge if long-term procurement commitments are stranded; and the "hybrid market structure" already approved by the Commission, would all facilitate competition by ensuring that: (1) LSEs share the costs of long-term commitments; (2) bundled customers are indifferent to the departure of load to competitors; and (3) new resources are developed. ³³

PG&E urges the Commission to approve its resource assumptions, medium case load forecast scenario, and portfolio strategy, which implement the EAP loading order cost-effectively and fills PG&E's projected net open position with "preferred" resources and a mixture of short, medium, and long-term products. The utility argues that the Commission should ignore self-interested proposals from other parties that could force the utilities to procure resources that are unneeded or would not be cost-effective. "The Commission should find that PG&E may procure 1,200 MW of long term peaking resources by 2008 and an additional 1,000 MW of long term shaping resources by 2010. . ." ³⁴ These levels are based on net open needs identified in PG&E's low load scenario. PG&E also requests that the Commission re-authorize short- and mid-term contracts, in order to have a robust portfolio. Additionally, depending on resource need, PG&E may enter tolling contracts with existing resources and bilateral agreements with generators after they are no longer needed for reliability must run (RMR) support as well as with generators whose current contracts with DWR expire within the planning period.³⁵

SCE states that under its scenarios:

· SCE's expanded demand-side portfolio is cost-effective in every scenario, but must be adapted based on SCE's bundled customer needs.

· SCE will meet the Energy Action Plan's accelerated renewables target in every scenario, and under the low load scenario SCE has no need for additional renewable generation until 2012,

· SCE has no need for baseload resources until at least the end of the decade and later, under the core/non-core scenario;

· SCE's current resource portfolio is overweight with long-term resources (greater than 5 year commitments) whether measured by capacity or energy. This situation is even more pronounced under core/non-core scenarios;

· SCE's current resource portfolio is overweight with baseload resources in the near-term and require s balancing with peaking resources,

· When compared to today's resource mix, SCE will require more peaking and intermediate resources and less baseload resources in the future.³⁶

SCE seeks to minimize the financial risk of such [excess baseload] resources to bundled customers by committing only to short- and medium-term peaking and intermediate resources. The multiple scenarios SCE presented in its LTPP all indicated that SCE would follow this strategic path forward regardless of the changes to its load.³⁷

Originally, SCE had requested authority only for short- and mid-term contracts of 5 years or less, but in its reply testimony it outlined a proposal for a 10-year contract if there could be off-ramps for specific purposes, such as greater than expected departing load. This option was added, in part, due to requests by parties that SCE enter some long-term contracts. SCE proposes to pursue this option in a future application to amend its procurement plan.³⁸

SDG&E claims that its resource plan does not assume that the exact size, timing, and sequence of each specific future resource addition be etched in stone through approval of this plan. Instead, SDG&E argues that approval of its resource plan, tested under a variety of scenarios, provides a critical first step to subsequently bringing forward specific resources for Commission approval. Adoption by the Commission of SDG&E's long-term plan would therefore constitute the Commission's agreement that the portfolio of resource types identified in this long-term plan represent desired outcomes for customers, and that SDG&E's moving forward to further study and permit the additions shown in the plan is consistent with Commission policy.³⁹

SDG&E argues that the Commission should approve its medium-load plan because it is SDG&E's best estimate of how it can prudently and reasonably prepare to meet its customers' needs over the next ten years. SDG&E's medium-load plan fully reflects the Commission's preferred loading order that first takes into account cost-effective energy efficiency, demand response, and renewable sources of energy before consideration of supply side resources and transmission.

For SDG&E a key component of its long-term resource plan is it proposed 500kV transmission line and it is seeking Commission support on this concept as part of its LTPP approval. The utility does acknowledge that it will still have to file a CPCN application for the transmission line. The CPCN proceeding will, among other things, consider the trade-off between transmission and generation, which was an analysis that numerous parties specifically mentioned. SDG&E argues that this analysis need not have to be done at this stage, however, and it does not prevent the Commission from concluding now that new transmission is a key component of SDG&E's long-term resource plan that needs to be further analyzed. ⁴⁰

H. Position of Parties on Implications of Resource Scenarios

To summarize, each party reviewed the IOUs resource scenarios under the microscope of their own perspective and ask for Commission action to promote that viewpoint. Those concerned with reliability and marketing power to the IOUs tended to argue in favor of more resources; those concerned with the environment, renewables, conservation and generally reducing demand for power wanted the IOUs to concentrate more on EE, DR, DG and renewables and less on fossil-fuel resources; proponents of brown sites advocated giving more priority to aging power plants; potential departing load parties worried the IOUs were over procuring, and consumer/ratepayer groups, while advocating reliability, question "at what cost?" Parties recommended a mix of contract terms from short-term ones to reduce the possibility of stranded costs, to long-term contracts to capture some certainty for prices in the future.

The IOUs complied sufficiently with Commission direction in preparing their resource scenarios so we will not require the preparation and resubmission of LTPPs at this time. What we glean from deficiencies in these LTPPs can be addressed by requesting updates as the Commission gives new direction or clarification in other resource/procurement proceedings and can direct us in giving guidance for the next LTPP proceeding.

In general, the three IOUs and the more than twenty-seven intervenors recognized that the resource scenarios represented "best guesstimates" and there is no way to predict the energy demand/supply situation with any certainty, especially in the face of changing load situations. A mix of resources, fuel types, contract terms and types, with some baseload, peaking, shaping and intermediate capacity, with a healthy margin of built-in flexibility and sufficient resource adequacy, is the best the IOUs can do at this point in time. It is also important the IOUs have room in their plans to procure resources as directed by the Commission in the areas of EE, DR, DG, renewables, and soon QFs. The IOUs need to balance expiring DWR contracts with required targets in EE, DR and renewables, so they are not fully resourced for the ten-year planning period with no head room for new resources.

Following is guidance on meeting the identified IOU needs in accordance with the EAP loading order and the carbon adder adopted in this decision. When executing procurement plans in response to the direction below, each IOU is to take the following steps:

· Procure the maximum amount of cost-effective energy efficiency and demand-side resources;

· For further resource needs, procure the maximum amount of renewable generation resources via all-source RFO, and be prepared to defend any selection of fossil over renewable resources; and

· Employ the GHG adder, described in this decision, when evaluating fossil generation bids.

Generation and demand-side commitments with start dates after 2010 may be deferred until the next procurement cycle. The winding-down of DWR contracts will materially affect the magnitude and nature of choices available, and we will be able to take advantage of two years of experience in implementing policy-preferred resources.

We find reasonable PG&E's strategy of adding 1,200 MW of reserve capacity and new peaking generation in 2008 and an additional 1,000 MW of new peaking and dispatchable generation in 2010 through RFOs is compatible with their medium resource needs, does not crowd out policy-preferred resources, and is a reasonable level of commitment given load uncertainty. Those commitments may need to be increased or expedited for PG&E to meet its 2006 resource adequacy obligations.

Depending on the nature of the bids obtained, PG&E is authorized to justify to the Commission why higher levels might be desirable. Nothing in this decision precludes PG&E from offering local reliability contracts, should they become necessary, pursuant to D.04-10-035.

We find SCE's LTPP resource plan is reasonable, subject to the compliance requirements covering its demand forecast, demand response, energy efficiency, QFs, and other factors set forth in this Decision and other Commission decisions in those designated proceedings. SCE has demonstrated that its primary residual resource need through 2011 is for peaking, dispatchable and shaping resources. SCE has considerable need for peaking and shaping resources, which should be obtained through a short, medium- and long-term acquisitions.

SCE's strategy of relying primarily on short- and mid-term contracts during this planning period is reasonable, but it may be prudent to add some long-term resources. SCE is authorized to present such a case to the Commission as an implementation of its LTPP by way of an application following an RFP.

SDG&E's resource scenarios were the most complete and useful in understanding the impact of differing loads, risk strategies, and the complex process of compiling a portfolio that meets reliability, adequacy, policy preferences and cost moderation goals. We find SDG&E's resource plan reasonable, subject to the modifications required for the compliance filing. SDG&E is essentially fully resourced through 2009, other than needed investments in renewable resources to meet RPS targets. Because SDG&E is fully resourced, SDG&E's resource plan is vulnerable to departing load and the utility is still obligated to meet its renewables, energy efficiency and DR goals. Since SDG&E's estimated reserve margins, which exceed 17% in some years during the planning period are the result of prior Commission decisions. There should be no finding of unreasonableness if they exceed 17%.

One critical element of SDG&E's LTPP that we are not approving is their request for a 500kV transmission line. As we discuss elsewhere, we do acknowledge the lengthy process that is needed to plan, license and construct transmission, and thus encourage SDG&E to continue its planning efforts and move forward with evaluating these transmission alternatives for meeting a local resource deficiency by 2010.

For this round of procurement filings, we find that the IOU filings are EAP-compliant if they included the EAP targets established in the RPS, DR and EE proceedings; included, at a minimum, the DG forecasts in the 2003 IEPR, and added transmission and clean central-station generation to meet remaining energy and capacity needs.

We will direct a compliance filing of annual energy and capacity resource accounting tables, consistent with directions on baseline load forecasts, EE, QFs and DR as explained elsewhere in this decision, but we will not require refiling of whole resource plans. We do expect the IOUs to make incremental improvements in their next round of analysis to be filed with the Energy Commission in 2005. Procurement resulting from the plans should comport with the direction, above, regarding obtaining the maximum feasible amount of renewable generation.

We concur with the CA ISO that the transmission elements of the plans were insufficient to meet our goals and accept their recommendations that future plans should include conceptual scenarios that illustrate the impact of potential generator location. We also concur that when an IOU proposes a major transmission line, it should include a companion scenario without the line. To the extent an IOU believes that the range of need identified in the 2005 IEPR is sufficient to justify a transmission project then it may be identified as a specific proposal to satisfy need in the 2006 procurement proceeding filings.

I. Natural Gas Price Forecasts

1. Regulatory Background

The May 2003 EAP, D.04-01-050, R.04-04-003 and the June 4, 2004, ACR all informed the IOUs on the subject of natural gas price forecast issues. R.04-04-003 reiterated the EAP's message that the IOUs were to "[f]irst seek to optimize all strategies to increase conservation and energy efficiency in order to minimize increases in electricity and natural gas demand." ⁴¹

This was building on the direction from D.04-01-050, and further emphasized in the ACR that "Long-term plans should reflect the most recent fuel-price forecasts available at the time of the plans' preparation and should include fuel-price variation as an element of the plans. We are not convinced that the actual degree of potential variation in fuel costs was reflected in the cost scenarios presented in the long-term plans. Therefore, we caution the utilities to consider seriously the degree of volatility that should be expected in fuel prices when developing high percentile scenarios for procurement costs particularly. We direct that future long-term procurement plans should reflect fully the expected range of prices of fuel and costs of purchased power at least up to the 95^th percentile of the expected distribution."⁴²(p.98)

Building on that, the ACR stated "[i]n addition to providing estimates of the resulting increase in cost of meeting load under these assumptions, the utilities should provide gas prices and market prices that correspond to the 95^th percentile. The utilities should submit a simple comparison of these price series to the base case assumptions. For gas prices, these should include monthly average prices."⁴³ (p. 16)

J. Utilities And Party Positions

PG&E developed gas price forecast using gas commodity prices based on the April 19, 2004, closing price of forward contracts traded on the NYMEX plus location basis obtained from broker quotes for gas delivered at AECO, Topock, Malin and PG&E Citygate for the period through February 2009, which marks the end of NYMEX availability. For March 2009 and beyond, PG&E extrapolates gas prices using monthly energy prices and maintaining the same monthly relationship as exhibited in the prior 12 months to March 2009. As required by the June 4 ACR, PG&E states it estimated its 95^th percentile portfolio risk using thousands of natural gas and electricity price scenarios in a Monte Carlo simulation.⁴⁴

PG&E includes Table 5-4 in its rebuttal testimony, Exhibit 36, at pp. 7-8, that presents gas prices resulting from their representation of volatility. Widths of the probabilities are substantial and grow as the delivery period is further into the future. Monte Carlo analysis was not driven by a set of fundamental variables; natural gas prices were simulated directly.

PG&E argues that its forecast shows substantial volatility, measured using standard deviation of probability distribution of prices, widths are substantial and grow in future exhibiting substantial deviation.⁴⁵

SCE's gas forecast relied on a gas price forecast prepared by Global Insight (GI), an international consulting firm and noted expert in gas forecasting, for all the major pricing points in the WECC and provided SCE with first and second standard deviation gas price forecasts. GI developed its gas price forecast using global and local factors which impact gas prices in the WECC. Global impacts include the price of oil and importation of LNG into the U.S. Local impacts include LNG facilities and supply basin and pipeline development in the western U.S. Standard deviation forecast developed using fluctuating variables such as U.S. economic growth, LNG imports, California economic growth and weather. SCE provided an analysis between GI's forecast and CEC's. CEC's forecast is higher than GI's and assumed to be due to the impact of future LNG supplies. ⁴⁶ SCE acknowledges that forward gas markets have risen since April 2004, however the magnitude of the gas price forecast is not a major factor for SCE in determining what proportion of resource additions are gas or non-gas fired.

SDG&E developed a gas price forecast of $4.70 based on San Juan basin prices said to be the dominant supply resource at the California border. The forecast was designed following a five step process using the Gross Domestic Price inflation index, basin differentials and adding various costs for transportation from the basin to the border. SDG&E provides comparisons with other gas price forecasts and SDG&E asserts that its forecast is in line with other gas forecasts. Variations among the other forecasts are due to assumptions about LNG and outlook about other supply conditions.

The average cash price of gas at Henry Hub is $5.92. The $1.22 difference between SDG&E's forecast and Henry Hub is statistically insignificant since it is within one standard deviation of historical monthly prices. SDG&E argues that it is inappropriate to use NYMEX futures as a forecasting tool, since it is a one day sample of the market.

Monthly prices at San Juan basin are not "adjusted", but calculated from historic month-to-annual ratios. In response to criticisms of its gas forecast, SDG&E contends that it is erroneous to state that many charges are added to San Juan Basin prices. It is reasonable to expect LNG supplies to continue to grow and moderate prices in out years. Gas price forecast applies to base case scenario and does not reflect "year-to-year" volatility.

UCS asserts that none of the utilities/ provided enough information (e.g., description of inputs or relationship to end results) in filings or confidential work-papers to allow UCS and other intervenors to determine exactly what their assumptions were in conducting computer simulations of expected future gas prices.

In general, UCS alleges that the IOUs gas price forecasts were deficient as follows: PG&E did not discuss how it would manage gas price risk associated with gas-fired resources apart from its DWR and QF contracts or whether PG&E designed its portfolio options in order to minimize gas price risk; SCE did not say what its preferred portfolio was and included no discussion on how it would manage gas price risk or provided any alternative portfolios designed to minimize that risk; and SDG&E, intentionally or unintentionally, minimized its gas price risk through 2010 by choosing a portfolio that would not require it to procure conventional resources before then and then failed to indicate whether gas price risk will be a consideration in procuring power post 2010.

UCS recommends that the Commission mandate that utilities account for gas price risk when determining how they plan to buy power; provide details of all variables and ranges used in simulations; and results of simulations should be used to create a portfolio least susceptible to future expected gas price risks. In addition, the Commission should require the IOUs to supplement their forecasts using different price scenarios and clearly detail the variables and range of values assigned to each variables used in simulations and use results to create portfolios that mitigate future gas price risk.

UCAN only addressed SDG&E's gas forecast and urges the Commission to reject it since it reflects prices significantly lower than current NYMEX prices. UCAN is concerned that low gas price projections may skew long-term resource plans and, if intended to be used as a baseline, then it may have the effect of triggering new procurement decisions and impacting hedging strategies. UCAN recommends that SDG&E use most up-to-date information available and that it update its natural gas price forecast at least monthly using NYMEX data and or broker quotes.

Given current gas spot and futures prices, Strategic Energy claims that all utilities forecasts appear too low. Strategic Energy asserts that unrealistic low gas price forecast would depress the wholesale power price forecast and may affect least-cost procurement and skew results of comparing bids between utility-build and third party procurement.

This group of intervenors is concerned that each IOU utilized a different method to develop its gas forecast making an across-utility comparison difficult. Forecasts are flat showing little price volatility and they want the Commission to direct the IOUs to develop forecasts with more volatility.

WPTF finds the gas price forecasts too low and fears that they could be used by the utilities to skew results to favor its own or an affiliate's offer. The Commission should require a utility to commit to a gas price forecast if the utility offers a long-term resource based on a specific gas price forecast. WPTF also recommends the use of third party independent review of competitive solicitations.

Parties have recommended that the utilities have used separate approaches toward developing their gas price forecasts, that their forecasts appear low or that they do not exhibit much volatility. Such concerns were the basis for the gas price forecast guidelines we adopted in D.04-01-050 and the June 4, 2004 ACR and, in particular, that the LTPPs should reflect a range of expected prices. These requirements adequately address the concerns raised by the parties and ensure that the LTPPs are responsive to the uncertainties of predicting long-term gas prices. To ensure that gas price forecasts submitted in future LTPPs remain robust, we will require that the utilities provide updated gas price forecasts using the same criteria set forth in D.04-01-050 and the June 4, 2004 ACR when subsequent long term procurement plans are filed with the Commission.

K. How the Utilities' Long-Term Plans Reflect Policies, Goals, And Outcomes From Other Umbrella Proceedings and Comport with the Energy Action Plan

1. Umbrella Proceedings

This OIR was designed to be an "umbrella" proceeding to coordinate and incorporate Commission efforts in the CCA, DR, DG, EE, Avoided Cost and Long-term Policy for Expiring QF Contracts, RPS, Transmission Assessment and Transmission Planning proceedings, as well as to address Resource Adequacy (RA) requirements. The June 4, 2004 ACR identified LTPP and RA as the "critical path" issues that need to be addressed in this proceeding.

2. Resource Adequacy

The Commission's decision in RA, D.04-10-035, issued October 28, 2004, among other things, established that all Load Serving Entities (LSE), including the IOUs, must have reserve margins of 15-17% by June 1, 2006. As part of meeting this reserve margin requirement, each LSE must have 90% of its next summer's requirement [May through September] fully resourced by September 30 of the year before. The decision also established a 100% forward commitment obligation for a month-ahead horizon for the five summer months, so each LSE must acquire the incremental remaining 10% of forward commitments needed to satisfy resource adequacy requirements. The IOUs are to plan to meet all RA requirements as set forth in D.04-10-035 as they go forward with their LTPPs.

²⁰ PG&E opening brief, p. 7.

²¹ Ex. 34, PG&E/Aslin, p. 4-7.

²² Ex. 34, PG&E/Aslin, p. 4-7, PG&E Opening Brief, p. 7.

²³ SCE/Whatley Tr. Vol. 11, 1602:16 - 1603:14.

²⁴ SDG&E opening brief, p. 11.

²⁵ ORA Testimony, EX 41C, pp 1-16. "C" after an exhibit indicates that it is a "confidential" exhibit and only parties who are members of the PRG groups or who signed the protective order have access to the confidential version.

²⁶ CMTA/CLECA Opening Brief, p. 3.

²⁷ D.04-01-050 and D.04-10-035

²⁸ (Ex. 34 and 35C, Tables 4-3 and 4-4.)

²⁹ PG&E Exs. 34 and 35C, Tables 4-3 through 4-8, PG&E opening brief, pp. 16-17.

³⁰ SCE Opening Brief, Appendix A, p. A-7.

³¹ SDG&E opening brief, pp. 16, 18.

³² R.04-04-003, mimeo p. 4.

³³ PG&E opening brief, p. 5.

³⁴ Id., p. 2.

³⁵ Id., pp. 20-21.

³⁶ Id., pp. 8-9.

³⁷ Id., p. A-7.