X. Cost-Effectiveness Conclusion

As discussed above, we have adopted the following changes to PG&E's modeling assumptions to be used in our cost-effectiveness calculations:

· SGRP cost of $706 million (base case), and $815 million cap.

· Base capital additions of $87 million for 2016 and after.

· 4.5% O&M escalation rate after 2011.

· September 5, 2003 and April 19, 2004, NYMEX closing prices for gas.

· 30-year facility life for combined cycle generation.

We first change the combined cycle facility life to 30 years. With this change, market prices for replacement energy are lower than combined cycle generation or combined cycle generation with 10% wind. Therefore, we will use market prices in our cost-effectiveness calculations.

PG&E performed steam generator tube inspections during the October-November 2004 refueling outage of Unit 2. We do not have those results at this time, and the results of the tube inspections during the Unit 1 refueling outage in early 2004 were not included in the record.²⁵ Therefore, we will include consideration of the possibility that the results of the inspections will indicate that the most probable date for Unit 2 to go out of service without the SGRP is one refueling cycle later (referred to as "1 unit refueling outage" in the table below). We will also consider the possibility that the most probable date for both units to go out of service without the SGRP is one refueling cycle later (referred to as "2 unit refueling outage" in the table below).²⁶

The following table shows the NPVs using PG&E's Monte Carlo simulation model, in 2003 dollars, of five scenarios illustrating the results of our cost-effectiveness analysis. A negative NPV indicates that the costs of the SGRP exceed the benefits. The term "High Gas" refers to replacement electricity costs based on the September 5, 2003 NYMEX closing prices for gas. The term "Low Gas" refers to replacement electricity costs based on the April 19, 2004 NYMEX closing prices for gas. The base case (first scenario) uses the above modeling assumptions and a $706 million SGRP cost. Subsequent scenarios incorporate additional assumptions. Each scenario is shown using the 90.6% capacity factor used by PG&E in its application, as well as an 85% and an 80% capacity factor.

Table of NPV Results

Scenario Assumptions Capacity factor²⁷ Low Gas High Gas

($ millions) ($ millions)

1 Base 90.6% 522 804

85% 313 578

80% 129 378

2 Base +1 unit refueling outage 90.6% 429 687

85% 226 468

80% 47 275

3 Base +1 unit refueling outage 90.6% 333 591

+$815 million SGRP cost 85% 130 372

80% -49 179

4 Base +1 unit refueling outage 90.6% 194 439

+$815 million SGRP cost 85% -1 229

+1-year outage²⁸ 80% -172 45

5 Base +2 unit refueling outage 90.6% 217 450

+$815 million SGRP cost 85% 21 240

80% -152 54

We have no reason to believe that a one-year outage of one unit is likely. In addition, we have no reason to believe that the tube inspections during the 2004 refueling outages will extend the most probable date for one unit to go out of service without the SGRP by more than one refueling cycle, or for both units by one refueling cycle. Therefore, we believe the third scenario is the most probable. Under this scenario, the SGRP will be cost-effective, even at the low gas price and the $815 million SGRP cost, as long as the capacity factor remains above approximately 82%.

Although we do not believe it likely, if we add a one-year outage in 2015 to the third scenario, the SGRP remains cost-effective at the low gas price and the $815 million SGRP cost as long as the capacity factor remains above approximately 85%, as shown in the fourth scenario.

We have no reason to believe that the tube inspections during the 2004 refueling outages will extend the most probable date for both units to go out of service without the SGRP by two refueling cycles. In that case, however, the SGRP will still be cost-effective, even at the low gas price and the $815 million SGRP cost, as long as the capacity factor remains above approximately 85%, as shown in the fifth scenario.

The above analysis assumes that, if the SGRP is not performed, there would be generation facilities ready and waiting to provide replacement power. If the SGRP is not performed, it would not be known for certain when either Diablo unit would shut down until it is relatively imminent. We expect that investors would be reluctant to build replacement power plants given this uncertainty. Therefore, it is possible that replacement power would not be available when needed, or that the cost would be high. In addition, large generating facilities of any kind, including any necessary fuel transportation facilities and electric transmission facilities, cannot be built overnight, especially given the need to obtain financing, an appropriate site, and the necessary regulatory approvals. For these reasons, the assumption that there would be generation facilities ready and waiting to provide replacement power is optimistic, and likely understates the SGRP's cost-effectiveness.

Additional benefits that derive from the SGRP are the increased likelihood that Diablo will remain in operation as a reliable energy source, reduced air pollution compared to fossil generation, reduced dependence on fossil fuel, and diversity of electricity resources. These unquantified benefits increase the cost-effectiveness of the SGRP. We also note that there are additional unquantified costs that result from risks associated with additional spent nuclear fuel that will be generated by the continued operation of Diablo due to the SGRP. Such costs would decrease the cost-effectiveness of the SGRP.

Based on the above, we preliminarily determine that the SGRP will be cost-effective.