10. Discussion

The purpose of the summer hearings, as directed by the ALJ, was to evaluate the system benefits and costs of adding transmission capability to the Southwest under multiple load and generation scenarios. For reasons described above, the Joint Parties only presented scenario analysis related to the need for transmission from a reliability perspective, i.e., the need to import power when in-state generation is insufficient to meet "physical" loads. We discuss that analysis below. At this time, we have no information on the record as to whether or when it would be advantageous to build new transmission to the Southwest from an economic perspective, i.e., to make less expensive power available to ratepayers.

The record in this proceeding shows that the ISO has never assessed this type of economic need for transmission projects since its inception in 1997. The last time the utilities came before the Commission with a transmission project designed to provide economic benefits in the form of cost savings from less expensive out-of-state energy production was in 1988, with SCE's amended application to construct Devers-PaloVerde No. 2.¹⁶ One of the issues we consider today is how best to obtain and evaluate information on the economic impacts of new transmission projects in the future. As discussed further below, we believe that this information is critical to California ratepayers and should be evaluated in an open, public forum, with an evidentiary record.

We agree with the Joint Parties' conclusion that at least for reliability purposes, California most likely will not need additional Southwest transmission capacity before 2008. However, we have strong reservations about the analysis by which the Joint Parties reach that conclusion. Below, we explain our reservations, and what we did in this proceeding to satisfy them, to ensure that the present conclusions have a sound basis and to guide future reliability analysis.

10.1 Conclusions from Reliability Analysis

The credibility of the Joint Parties' conclusions with respect to the need for new Southwest transmission capacity is dependent upon both the credibility of the model and information used in the reliability analysis. We have reservations about both.

The first reservation concerns the use of the Joint Parties' matrix model, which is a simplified planning tool compared to the power flow model. The validation ('benchmarking') of the matrix model against the power flow model depends on the derating method used by the ISO, since without derating the matrix model forecasts transmission need several years later than the power flow model. As we discuss below, the assumptions supporting the derating method are themselves open to question; in addition, the derating method was not applied consistently in the analysis. The second reservation concerns information used in the model. In some respects, key assumptions appeared to be out-of-date, and the Joint Parties did not thoroughly assess which sets of assumptions ("scenarios") are deemed more or less likely.

Consequently, we had additional model runs performed during the proceeding using updated assumptions and consistent use of the derating method relied upon for benchmarking. We then examined all of the results from the perspective of which scenarios seemed likeliest to occur. This examination convinces us that our conclusion about transmission reliability needs is valid for the 2001-2011 planning period. In the following discussion, we deal first with derating issues and then with information issues, in each cases detailing the adjustments we directed.

With respect to the validity of the matrix model, we note that the ISO and utilities have never used the matrix modeling approach presented in this proceeding to evaluate the need for new transmission.¹⁷ To test its validity, Joint Parties compared the results of one case (Very Low Internal Generation/Base Load) under the Planning Scenario against the results of the detailed power flow model used in the Southern CA Study.

The results of the matrix model and the Southern CA Study are the same for that case only if transmission capability is derated in the matrix model, as described in Section 4.8 above. Hence, the validity of the matrix model as a reasonable simplification of a power flow model for assessing reliability needs is called into question if the derating methodology is inaccurate or applied inappropriately. ISO Witness Le testified that up to a 500 MW excess of new load over new internal generation could be addressed by relatively minor internal "fixes," but above that amount, a derate would probably have to be applied on a one-to-one basis without proper reactive voltage support.¹⁸ However, upon further questioning, ISO Witness Le acknowledged that this one-to-one assumption would have to be confirmed with power flow studies.¹⁹

In addition, Witness Le testified that derating might not be required at all if there is sufficient reactive voltage support on the system.²⁰ Moreover, if projects to provide additional reactive voltage support were installed, Witness Le stated that the one-to-one ratio used in the formula would not hold.²¹ In fact, because reactive voltage support increased substantially since last year (due to the installation of shunt capacitors and other devices), the Western Systems Coordinating Council (WSCC) recently increased the transfer capability rating for Southern California, even while the gap between new load and new generation is expected to widen.²² While acknowledging that reactive voltage support is a factor in determining whether transmission transfer capability should be derated, Witness Le stated that he did not have a formula to capture that relationship.²³ We also note that the derating formula is not applied to models used to derive resource needs arithmetically in other ISO planning forums, as indicated in the most recent Study Plan for SCE's Grid Expansion Study.²⁴

If the one-to-one derate ratio is inaccurate, or should not be applied at all, the matrix model results will not match the Southern CA Study for the benchmarked case. As indicated in Table 1, if the benchmarked case is run without derating transmission capability, then additional transmission is not needed in 2008 (the year that need is shown under the more detailed Southern CA Study), but rather sometime after 2011. This represents a very large potential "bias" in the matrix model results if derating is inappropriately applied. In other words, we must consider the possibility that the matrix model will underestimate the need for new transmission by approximately 2-3 years.

To the extent that we are comfortable with derating transmission capability along the lines described by the modeling witness, we can accept the results of the matrix model as a reasonable simplification of more complicated reliability modeling efforts. However, as ISO acknowledges, the derating formula was not applied consistently in the scenario analyses presented on the record, even in the very low generation cases.²⁵ In its brief, the ISO also recognizes that capacity retired should be deducted from new internal generation in the formula for derating transfer capability.²⁶ We have recalculated the matrix model results using the derate formula applied consistently across all cases, and factoring retirements into the formula.

In addition, we have updated two assumptions regarding transmission transfer capacity. First, we have increased the Path 45 transfer capacity of 408 MW to 800 MW, effective in 2002, consistent with the record.²⁷ In addition, as noted above, the Operational Transfer Capability Committee of the WSCC recently approved an increase in the transfer capability of Southern California import transmission from 13,200 MW to 14,300 MW (or 8.3%) due to the installation of new reactive voltage support. As Witness Le acknowledged during hearings, had he been aware of the WSCC ruling when he prepared the matrix, he would have applied the derating to the higher level of transfer capacity throughout the planning period.²⁸

The results of these changes are presented in Tables 3 and 4 appended to this decision. As indicated in these tables, the net effect of increasing the import capability and applying the derate methodology to all cases defers the need for new transmission in most cases, relative to the results presented in Tables 1 and 2. This is because the increase in transmission capability (which will delay need) is in full effect early in the planning period, whereas the derate (which will accelerate need) is only in full effect at the end of the period, when load growth is at its peak.

In evaluating these results, we must consider which sets of assumptions capture the most likely range of possible outcomes for the future. The ISO characterizes the Joint Parties' "Planning Scenario" (Scenario 1) as representing the more plausible set of assumptions regarding outages and retirements, and Scenarios 2 or 5 as "unlikely."²⁹ However, the evidence in the record presented by the CEC witness for the Joint Parties indicates that Scenario 2 assumptions for outages are more credible. As CEC Witness Vidaver testified, historical data on outages indicates that a one or two day each summer outage level in Southern California would be on the order of approximately 3000 MWs, rather than the 1150 MW assumed in Scenario 1.³⁰

We are also not persuaded that the "no retirements" assumption used in Scenario 1 is more likely than one where some generation facilities are retired in the future, based upon the inefficiency of the units involved. CEC Witness Vidaver testified that all of the units retired in Scenarios 2 and 5 have very high heat rates, in the range of 12,000 to 20,000 Btu per kWh, and that current prices are not enough to sustain these plants.³¹ We find it more plausible to assume that these highly inefficient units will retire between now and 2011, rather than to assume no retirements.

With respect to the various load forecast assumptions, we agree with Joint Parties that the base load plus 10% or 20% load cases are unlikely.³² In fact, the record in this proceeding supports a conclusion that somewhere between the "base load" and "base load minus 10%" represents the most plausible range of load projections. SCE Witness Canning testified that, by early May, conservation efforts appeared to be stabilizing at about 5% below the October 2000 forecast of base load levels.³³ SDG&E Witness Jack testified that SDG&E's forecast of load is somewhat higher than CEC's projections used in the matrix model, so that accounting for more extensive conservation and other factors, would not bring it to 10% below the base load forecast, in his opinion.³⁴ CEC Witness Rohrer testified that although there had been reductions in demand on the order of almost ten percent, those occurred in April or May before the summer air conditioning season began. In his view, a 5% reduction below CEC's baseload projections would probably be more sustainable when customers had to choose between saving energy and "sweat[ing] in their houses."³⁵

Based on the record in this proceeding, approximately 1950 MW of new generation projects are under construction that will be available to Southern California (including 720 MW at High Desert), and more than 3000 MW of additional capacity has been approved and is in various stages of financing.³⁶ We therefore agree with Joint Parties that the "very low" generation case, where only 720 MW of new generation is assumed to materialize, is highly unlikely.³⁷ However, it is much more difficult to assess the likelihood of "low" (approximately 5,500 MW) to "maximum" (20,500 MW) new in-state generation actually coming on line, given the information available at this time. As SDG&E and SCE witnesses testified, the utilities have never before had generators of this magnitude wanting to build in California. This is a new phenomenon that started over the past year. Therefore, they do not have any history on what would be the likelihood that proposed projects will actually be built.³⁸

Similarly, it is difficult to assess the likelihood of how many resources will materialize for export from Mexico to California. There have been no exports from Mexico to Southern California in the past, and construction and financing on these new projects have not been completed,³⁹ nor have contracts or commitments for power delivery to Southern California been finalized in all cases.⁴⁰ SDG&E projects a maximum of 2550 MW of export from Mexico by 2005, based on its Mexico interconnection queue. Coral Power estimates that there will be an additional 1300 MW available from 1) a 500 MW Rosarita Project planned for construction in Mexico and 2) 800 MW of exports during the winter period by CFE.⁴¹ Hence, the record presents a range between approximately 500 MW (medium) and 3,850 MW (maximum) potential of exports from Mexico during the planning period.⁴² As in the case of in-state new generation, we do not have sufficient information to assign the likelihood of projects materializing within this broad range.

In terms of the assumptions concerning potential exports from Arizona/Nevada, we note that there was apparent miscommunication between the CEC witness who prepared the CEC project status list and the ISO matrix modeling witness who used this list to derive inputs for the matrix model. ISO Witness Le testified that he thought he was receiving figures relating to projects only under CEC status 1-3 (projects under construction, approved or under CEC review), which are generally used in CEC load and resource assessments.⁴³ As described in Section 4.4 above, 50% of the project totals were used for the maximum case and 20% for the medium case runs. Instead, Le applied these percentages to the resources listed under CEC status levels 1-5, which include projects identified in press releases, or just starting the application process with CEC. In fact, projects listed under CEC status levels 1-5 include approximately 10,000 more MW for Arizona and Nevada than levels 1-3.⁴⁴ Because of this miscommunication, we consider it unlikely that the maximum case scenarios for external resources available from Arizona and Nevada will materialize.

The results of the matrix scenario analysis, when derating is consistently applied and transfer capability is updated, indicates no need for new transmission to the Southwest until 2009 or beyond in all cases except those run under the most "unlikely" assumptions: 1) very low new internal generation (720 MW) and 2) low or medium internal generation with a 10% or 20% increase in base load demand. In fact, the only two cases in which need is indicated before 2011 in scenarios with low, medium or maximum internal generation is the "low generation" scenario with a 1) 10% increase in base load demand, assuming no retirements after 2004, and 2) base load demand assuming post-2004 retirements. (See Table 4.) Hence, the preponderance of cases run with the updated transfer capability assumptions indicate that new transmission for reliability reasons will not be needed until 2011 or later. These results assuage our concerns over the possibility that the matrix model contains a bias towards underestimating the need for new transmission. As discussed above, if the ISO's derating concept is questionable and the matrix model actually underestimates need by 2-3 years based on the benchmark run, the need for new transmission still does not surface before 2008, under all but relatively unlikely combinations of load and internal generation assumptions.

We will monitor the reliability modeling efforts conducted through the ISO's Grid Coordinated Planning Process in order to update and confirm these results with the detailed power flow studies conducted during that process. To this end, we direct Energy Division to report to us on an ongoing basis if the power flow studies indicate a need for reliability purposes earlier than 2008. This report should take the form of a letter to the Assigned Commissioner and ALJ, with service on all parties in this proceeding, or its successor.

The results of the reliability analysis in this proceeding indicate that we have a sufficient window of time to further update planning assumptions and consider the need for new transmission to the Southwest from an economic perspective. However, we do not believe that decisions concerning the economic need for major transmission projects, which could cost ratepayers over a billion dollars, should be left to the discretion of the ISO management personnel or Board, given that the ISO does not have the mandate or statutory authority to protect ratepayers' interests, and lacks an open, evidentiary process to scrutinize the methodologies and assumptions used to reach such decisions. While we appreciate the ISO's efforts to facilitate a resolution of the economic need issues through an RFP process, we believe that the public interest is best served by evaluating the economic need for new transmission projects, and the appropriate allocation of costs among beneficiaries, in this proceeding-- where we can ensure that a public record is fully developed. To that end, we direct SCE, SDG&E, and PG&E to jointly file the results of the ISO/stakeholder RFP process within 15 days from the date that the consultant's final report is completed. The assigned ALJ will hold a PHC as soon as practicable thereafter to schedule evidentiary hearings on the economic need for new transmission to the Southwest.

10.2 Other Issues

With regard to Coral Power's testimony concerning needed in-state transmission upgrades, we have recently scheduled a separate set of evidentiary hearings on the net economic benefits to ratepayers of relieving two potential in-state transmission constraints in Southern California, including alternatives to address potential congestion west of Miguel.⁴⁵ On the issue of upgrades to Path 45, we note that SDG&E is already moving ahead with adding a second circuit to the La Rosita-Imperial Valley component of Path 45, which will increase the capacity of that path from 408 to approximately 800 MW this fall. There are already general discussions underway between SDG&E and CFE to consider additional upgrades to Path 45.⁴⁶ We direct SDG&E to submit information on the status of those discussions and any upgrade proposals that would involve additional ratepayer funding in the monthly status reports ordered by D.01-03-077.⁴⁷ To the extent that significant ratepayer funding is involved to further upgrade Path 45, we may include this issue in the evidentiary hearings on economic need for new transmission to the Southwest. However, if private developers or the CFE fund the additional upgrades, then we will not need to review the issue further in this proceeding.

Finally, with respect to SSRC's position in this proceeding, we concur that no conclusions can be made from the record in this proceeding regarding the adequacy of the in-state transmission grid in Southern California.