IV. Utility-Specific Issues

A. SDG&E 2&3 Decommissioning Cost Estimate

The utilities estimate decommissioning costs for SONGS 2&3 at $2.23 billion. ORA proposes a reduction of $15 million in reactor vessel and internals segmentation and removal costs, and a reduction of $77 million related to spent fuel wet storage costs. ORA also opposes the utilities' escalation rates and rates of return. These issues are addressed later in this decision under common issues.

The utilities' estimate of the decommissioning costs for SONGS 2&3 is based on a site-specific review of the decommissioning requirements for SONGS 2&3, and takes into account experience in decommissioning SONGS 1.

The utilities' SONGS 2&3 decommissioning cost estimate includes increased costs associated with a four-year schedule increase. The schedule increase resulted in additional fixed project costs of $96 million. These fixed project costs are associated with the base project staff.

ORA recommends that the Commission reject the utilities' $150 million reactor vessel internals segmentation, reactor vessel segmentation, and large component removal activities estimate for SONGS 2&3, and instead use a $135 million estimate. ORA contends that the very high costs for these activities for SONGS 1 were partially caused by the newness of the segmentation and removal processes. ORA also argues that there will be technological developments that will simplify these processes. Therefore, ORA recommends a 10% ($15 million) reduction from the utilities' estimate to account for future improvements in decommissioning methods.

Regarding spent fuel wet storage and additional LLRW volume disposal costs, the utilities' estimate includes $96 million for fixed project costs for a four-year schedule extension. ORA contends that the fixed costs that the utilities propose are 102% of direct costs for LLRW, in contrast to the 35% fixed cost share the utilities used for their overall project estimate. ORA maintains that the four-year schedule extension does not mean that the scope of general decommissioning work should expand proportionately by four more years at an additional cost of $96 million. ORA contends that the decommissioning work should be spread out according to the new 15-year schedule, and the staffing costs should also be spread out over that period. ORA recognizes that there are still fixed costs that need to be added for four years of additional wet fuel storage, including the corporate overhead and base project staff. Therefore, ORA proposes reducing the utilities' $96 million estimate, by $77 million, to $19 million. The $19 million includes 20% for overhead, plus needed staff.

Discussion

ORA recommends a 10% reduction in costs for reactor vessel internals segmentation, reactor vessel segmentation, and large component removal activities to account for future improvements in decommissioning methods. While there may be such improvements in the future, what they may be, and the effect on costs is unknown. ORA offers no specific reasons why such improvements, if they occur, will result in a 10% savings. Therefore, we will not adopt ORA's recommendation.

As to the proposed staffing changes, ORA proposes reducing the fixed cost estimate because it believes that the schedule increase results in doing the same amount of work over a longer period of time. However, that is not the case. The additional costs are for additional work that will be performed. Therefore, we will not adopt its recommendation.

B. Palo Verde Decommissioning Costs

SCE estimates its share of decommissioning costs for Palo Verde at $503 million. ORA proposes a $27 million reduction in staffing costs associated with schedule changes, and a $3.5 million reduction in large scale component removal costs. ORA also opposes SCE's contingency factor, escalation rates, and rates of return. These issues are addressed later in this decision under common issues.

SCE`s estimate is based on a study performed for APS by TLG Services, Inc. (TLG). When assumptions in the TLG study were inconsistent with SCE's understanding of industry decommissioning experience, or its experience decommissioning SONGS 1, SCE applied adjustments. SCE says its decommissioning cost estimate for Palo Verde is not as detailed or definitive as the updated SONGS 2&3 cost estimate.

Changes to site staffing expenses account for $27 million of the increased staffing costs for dismantling activities. The increase is for engineering, cost and scheduling, emergency preparedness, and security work functions as well as support functions such as health and safety, legal, and regulatory affairs. SCE believes these staffing increases are consistent with increases currently planned for other decommissioning projects in the United States reviewed by TLG.

The SONGS 2&3 and Palo Verde reactor vessels, reactor vessel internals, and large components are of similar design and size. Therefore, SCE used the same estimation methods for internals segmentation, vessel segmentation, and large component removal activities at both SONGS 2&3 and Palo Verde.

ORA recommends the Commission disallow $27 million in increased labor costs included in SCE's estimate because SCE provided no specific justification, such as the previously unanticipated tasks these additional personnel will perform. As with the SONGS 2&3 estimates, ORA recommends that SCE's estimates for internals segmentation, vessel segmentation, and large component removal activities for Palo Verde be reduced by 10% ($3.55 million) to account for future improvements in decommissioning methods.

Discussion

SCE's estimate of increased staffing levels is based on the staffing levels associated with other decommissioning projects in the United States. No decommissioning plan has yet been developed for Palo Verde. Therefore, we would not expect the level of detail that ORA would have us require at this time. As a result, we will not adopt ORA's recommendation.

ORA recommends a 10% ($3.55 million) reduction in costs for internals segmentation, vessel segmentation, and large component removal activities to account for future improvements in decommissioning methods. While there may be such improvements in the future, what they will be, and the effect on costs is unknown. ORA offers no specific reasons why such improvements will result in a 10% savings. Therefore, we will not adopt its recommendation.