7. Discussion of Contested Issues
There were a large number of contested issues in these proceedings. During the course of these proceedings, the parties have moved from their initial positions on numerous issues. A summary of their pre-Settlement positions is attached hereto as Appendix C.
The Settlement Agreement is sponsored by parties representing a range of interests but is not supported by all parties. Certain provisions are opposed by DRA and Fielder who represent ratepayer interests. We appreciate the fact that the Settlement reflects a range of divergent interests, including those of the utilities and of residential customers. In addition, we have also reviewed and considered the objections of those parties that did not join in the Settlement. As discussed below in detail, we find merit in some of the objections raised by these parties, and we reject the proposed adoption of a rebuttable presumption of reasonableness for decommissioning costs for activities, other than Phase 1 of SONGS Unit 1, as not in the public interest nor reasonable in light of the whole record. (See Section 7.7 below.) Therefore, we reject the Settlement as a whole and now consider whether there is a reasonable basis for approving the proposed cost estimates, past expenditures, proposed trust fund contributions, and other policy matters based on the final positions of the parties after five days of full evidentiary hearings, settlement negotiations, an evidentiary hearing on the proposed Settlement, and post-hearing briefs filed by the parties.
7.1. Compliance With D.07-01-003
During the 2005 NDCTP, which was resolved by adoption of a settlement, the Commission ordered the utilities to serve testimony in the 2009 NDCTP in three areas: 1) the use of qualified and experienced personnel, 2) a conservative forecast of costs for LLRW storage, and 3) a conservative and appropriate contingency factor for inclusion in each utility's decommissioning revenue requirements.
Each utility provided information about its own process for assuring that only qualified and experienced personnel are used for decommissioning activities planned or occurring at SONGS Unit 1 and HB3.18 The utilities also jointly retained a consultant to perform an analysis of representative LLRW disposal rates available throughout the industry and used the identified base rates to develop a projected rate for use in the 2009 NDCTP. The utilities used the results and the evidence supported that the forecasts were conservative.
Lastly, PG&E developed and submitted a "Technical Position Paper for Establishing an Appropriate Contingency Factor for Inclusion in the Decommissioning Revenue Requirements" which included a review of available literature and reports, use of a contingency factor by other related industries, and recommended cost engineering practices from established professional organizations. The paper concluded that a 25% contingency factor for all nuclear decommissioning costs should be applied. SCE agreed based on its own independent research that the 25% contingency factor was conservative and appropriate. Both the original applications and the settlement proposal in these proceedings apply a 25% contingency factor to the cost estimates for all nuclear units.19 Fielder objected to 25% factor as inadequate because it excluded financial and regulatory risk and changes in scope. However, there was evidence that to the extent such risks were not included in a utility's contingency, they were otherwise accounted for in the cost estimates.
We find that SCE, SDG&E, and PG&E are in compliance with prior decisions applicable to decommissioning, including the Ordering Paragraphs 6, 7, and 8 of D.07-01-003 described above. We confirm that SDG&E may reasonably rely upon SCE, the majority owner of and exclusive operating and decommissioning agent for SONGS Units 1, 2, and 3 to make reasonable efforts to comply with the Commission's directives in D.07-01-003.
7.2. Approval of Decommissioning Cost Estimates
The utility cost estimates contain a degree of speculation by nature, partly due to persistent uncertainties about the key component of future storage and disposal costs for radioactive waste, and partly because detailed engineering studies are not completed until decommissioning is imminent. Over time, the Commission has seen substantial increases to the cost estimates brought forward by the utilities for review and approval. That trend continued in these proceedings and led to a high level of scrutiny by parties and the ALJ during the evidentiary hearings.
On balance we find that the cost estimates proposed in the applications for each nuclear generating unit, although developed somewhat differently by the retained experts, are supported by the evidence. We adopt these cost estimates subject to a few changes in assumptions as discussed below reflecting agreed terms in the proposed Settlement, and which are bound by the evidentiary record.
7.2.1. SONGS Units 1, 2, and 3
The remaining work scope for SONGS Unit 1 consists of Phase 2 which will end when all Spent Nuclear Fuel (SNF) is removed from the site and Phase 3 which is mostly dismantling and disposing of the ISFSI. Phase 3 is scheduled to occur concurrently with Phase 3 for SONGS Units 2 and 3 and projected to be completed in 2053. The estimated costs to complete decommissioning of SONGS Units 1, 2, and 3 were developed by ABZ, Inc. (ABZ), a recognized expert in nuclear decommissioning costs, using data provided by SCE based on its experience with SONGS Unit 1 and tested against ABZ's database of decommissioning costs at other nuclear sites.20 SDG&E conducted its own independent review of the ABZ cost study.
The SONGS Units 2 and 3 cost estimates increased from the 2005 NDCTP by $124.5 million (100% share, 2008$) due to assumed higher energy costs and staff and separation costs arising from NRC-mandated security actions, additional five-year delay to 2020 before SNF is removed, localized labor rates, related staff separation costs, and the application of a 25% contingency factor to the staffing costs.21
TURN initially viewed the estimates as excessive, but modified its position during the evidentiary hearings.22 We also find reasonable the use of the LLRW burial rates from the joint utility study and application of a 6.93% burial escalation rate based on historical rates.23
Based on the foregoing, we find that the cost estimates for SONGS Units 1, 2, and 3 are reasonable.
7.2.2. Palo Verde Units 1, 2, and 3
Arizona Power Service (APS), the operating agent for the Palo Verde units, retained TLG Services, Inc. (TLG) to prepare a decommissioning cost study. TLG, also a recognized expert in the field of decommissioning costs, used drawings and inventory documents to estimate waste volumes and make other assumptions in the cost study. SCE concluded that some assumptions made by TLG were inconsistent with SCE's experience and risk tolerance and, therefore, SCE made substantial adjustments to the TLG estimate and then applied a 25% contingency factor to all costs.24
The resulting estimate of $708.7 million (2007$) for SCE's share of all three units is about 7% below the estimate adopted in the 2005 NDCTP primarily due to significantly reduced LLRW burial costs, even though additional waste volume was projected.
Based on the foregoing, we find that the cost estimates for SCE's share of decommissioning the Palo Verde units are reasonable.
7.2.3. Humboldt Bay Powerplant 3
PG&E has begun preparatory decommissioning activities at HB3 and intends to commence decommissioning of the plant in 2010 and act as its own general contractor. PG&E retained TLG to prepare a detailed cost estimate which assumed a delay in beginning SNF disposal until 2020 and applied the LLRW burial costs from the LLRW cost study, a 7.5% burial escalation rate, an employee labor escalation rate of 3.75% based on its union contracts, and a 25% contingency rate25 to all costs. The estimate of $499.8 million (2008$) excludes $385,520 that has been disallowed by the Commission, but includes $82.3 million in costs incurred or projected to be incurred in 2009. The primary reasons for increases to the estimate from 2005 are increased staffing levels, revised or added unit cost factors for some activities, and increased waste volumes driven in part by site-specific challenges.
DRA generally accepted the cost estimate, but thought SCE's lower burial escalation rate of 6.93% should be used. PG&E used the higher figure based on its use in prior NDCTPs, the unreliability of having few data points in the LLRW study, and uncertainties about future disposal rates. PG&E's explanation of its differences on these items was reasonable for purposes of these proceedings. In the Settlement, the parties agreed to a modified labor rate which reflected PG&E's union contracts through expiration in 2011. This is also reasonable.
Fielder argued that a composite figure be used for LLRW disposal rates. However, there is no evidence that this approach is superior to the graded rates developed in the LLRW study and may be contrary to the Commission's direction in D.07-01-003. Additionally, Fielder's requested 35% contingency factor seems excessive and lacks evidentiary support. PG&E's witness Sharp explained that the 25% contingency factor was not intended to include changes in scope or other conditions which should be factored into the underlying cost estimate prior to application of the contingency rate.26
Based on the foregoing, we find that the cost estimate for HB3 is reasonable. In addition, we find that PG&E's uncontested forecast for 2010 O&M expenses associated with maintaining HB3 in SAFSTOR, including attrition through 2012, is reasonable.
7.2.4. Diablo Canyon Units 1 and 2
PG&E retained TLG to prepare cost estimates for the DC units under two decommissioning scenarios which included the same labor and LLRW burial rate assumptions and 25% contingency factor described above for the HB3 cost study. Under the more likely "DECON" method, which provides for prompt removal and dismantling of the facility, the total estimated cost for both units is $1,828.35 million (2008$). Consistent with the current operating license, the 2009 cost study also reflects shutdown dates for Units 1 and 2 of November 2024 and August 2025, respectively.
No significant objections were made to the cost estimates except that DRA continues to argue that PG&E should use SCE's LLRW burial escalation rate. The difference in cost would be about $1.8 million27 but we find that PG&E reasonably justified its use of the 7.5% rate. However, we adopt the proposed modification of PG&E's labor escalation rates contained in the Settlement, which fall within the bounds of the evidentiary record: 3.75% for 2009-2010, 4.0% in 2011, and use of SCE's 3.14% after 2011.
Based on the foregoing, we find that the cost estimates for DC Units 1 and 2 are reasonable, as adjusted.
7.3. Approval of Decommissioning Expenses
In the first NDCTP, the Commission adopted a settlement that authorized the commencement of decommissioning at SONGS Unit 1 and created a presumption of reasonableness for its decommissioning expenditures if kept within prior estimates. D.99-06-007 provides:
If the scope of SONGS 1 (Phase 1) Decommissioning Work completed and costs incurred to date are bounded by the most recently approved SONGS 1 Decommissioning Cost Estimate, the Utilities' conduct will be presumed reasonable. Any entity claiming the Utilities acted unreasonably would, therefore, bear the burden of proving the Utilities acted unreasonably. The utilities will be responsible for proving that material variances from the most recently approved SONGS 1 Decommissioning Cost estimate are reasonable.28
To be entitled to a presumption of reasonableness here, SCE is required to provide a comparison of the 2004 estimated Phase 1 costs at SONGS Unit 1 to the actual costs for the work completed between July 1, 2005 and December 31, 2008. SCE incurred a net cost of $207.2 million (2008$) for the completed work compared to the $221.3 million (2008$) estimated cost approved in the 2005 NDCTP.29 Actual costs were lower in nearly all categories. No party has contended the expenditures were not reasonable or prudent and SCE provided uncontested evidence the work was performed by qualified and experienced personnel. As a result, based on the settlement agreement adopted in D.99-06-007, SCE's actions are presumed reasonable and we find no evidence to suggest they were either unreasonable or imprudent.
PG&E is subject to the general reasonableness review rather than the presumption. The company provided a comparison of approved cost estimates and actual expenditures in connection with preparatory decommissioning activities at HB3. PG&E incurred a net cost of $63.4 million (2008$) for the work scope that was completed compared to an estimated cost identified in the 2005 NDCTP of $58.6 million (2008$). The biggest excess occurred in the primary category of "Licensing, Design, Fabrication, and Construction of ISFSI" due to new NRC requirements and both design and contamination issues related to the confined space. This expense was partially offset by lower than expected costs for shipment and burial of certain waste. PG&E provided uncontested evidence the work was performed by qualified and experienced personnel.
Based on the foregoing, we find the decommissioning expenses claimed by SCE and PG&E are reasonable and prudently incurred.
7.4. Rates of Return and Trust Fund Contributions
The Commission's adopted rates of return should capture a reasonably conservative growth trend over the life of the trust funds to match the estimated decommissioning costs. The recent economic downturn resulted in lower than expected returns in the trust funds during 2007 and 2008, initially prompting requests for significant contributions to some funds. Each utility developed its own forecast for rates of return on the equities and fixed income portions of its trust funds for the qualified and non-qualified trusts. The parties had different views about what benchmarks to use and how to interpret them. Inconsistent assumptions about the trust fund portfolios and management contributed to disparate results. As the proceedings progressed, the trust funds recovered some of their lost value and trust fund balances as of December 31, 2009 will be applied to calculate approved contributions.
SCE initially applied an 8.06% pre-tax return on equity, SDG&E applied 8.13%, PG&E used 8.5%, and TURN proposed 10.05% for all three utilities, each estimate based on nationally recognized indices.30 TURN's recommendation was significantly different because it limited the forecast of equity returns to the 14-year period in which SCE and SDG&E funds were anticipated to hold equities, i.e., pre-decommissioning of SONGS Units 2 and 3. Thus, the 10.05% reflects the shorter-term forecast of higher returns for market recovery between 2009 to 2022, while the longer-term forecast out to 2038 is preferred by the utilities to smooth out a reasonable "average" return.
The utilities argued it was neither reasonable nor conservative to focus on shorter-term projections. They also proposed lower equity turnover rates than adopted in 2005 which seems reasonable given current market volatility. Although TURN proposed a uniform rate of return, the utilities opposed it on the grounds that their own forecasts were appropriate because each trust fund was differently composed and managed. For purposes of these proceedings, we agree that overemphasis on short-term market recovery is not a conservative approach to the forecasted return and uniformity is less of an imperative than consideration of the actual composition of the trust fund portfolios.
The Settlement proposed different rates for PG&E than for SCE and SDG&E. The proposed pre-tax equity return of 8.75% for SCE and SDG&E is an increase over the rates they proposed, but not outside the evidence presented for a reasonable rate of return. Similarly, the 8.5% PG&E proposed would remain applicable to the HB3 trust funds which will eliminate equities by 2013 in order to finance the concurrent decommissioning. This is the same assumption adopted in the 2005 NDCTP and not unreasonable. For the DC trust funds, the Settlement states that after-tax returns will be adjusted on a pro-rata basis in order to yield the proposed $9 million annual contribution. (See below, Section 7.4.3.1.) PG&E's somewhat artificial calculation is that an assumed equity return of 8.13% return31 would yield the proposed contribution. This is somewhat low but it matches rate of return evidence originally presented by SDG&E.
Based on the foregoing, we find that the equity rates of return proposed in the Settlement are reasonable and within the range of reasonable outcomes based on the evidence in the record.
7.4.2. Fixed Income Rates of Return
For the fixed income portions of the trust fund portfolios, SCE originally assumed a 4.69% pre-tax return, SDG&E assumed 5.34%, PG&E applied 4.11%, and TURN agreed with SCE. The disparity is the result of different indices and assumptions, primarily whether to assume a municipal bond yield in the portfolios. DRA concluded the fixed income returns forecasted by the utilities were reasonable.
TURN's recommended debt return was based on 10-year municipal bonds and works out to the 4.2% post-tax return applicable to SCE and SDG&E in the Settlement. PG&E retained its original forecast for the HB3 trust funds. As noted above, the Settlement presumes a $9 million annual contribution to the DC trust funds without reliance on specific debt and equity returns.
We agree that despite some variations between the utilities, the forecasted returns are reasonable and the small modifications provided in the Settlement are within the range of reasonable outcomes based on the evidence in the record.
7.4.3. Contributions and Revenue Requirements
The Commission requires the utilities to update the trust fund balances to December 31, 2009 when calculating their contributions. Each utility has submitted an exhibit which describes the contributions and revenue requirements using the updated balances and the settlement terms which we have adopted herein.32
PG&E originally sought approval for $33 million in total trust fund contributions resulting in the grossed-up revenue requirements for 2010 set forth below:33
Diablo Canyon $23.329
Humboldt $16.982
Humboldt SAFSTOR $ 9.218 (O&M)
Total $49.528 million
This represents a $25.7 million increase from the currently authorized revenue requirement. When the trust fund balances are updated to December 31, 2009, without any other changes to the assumptions in the application, the required total contributions would decrease to $18.69 million.
One controversial issue in the proceedings was the proposed annual contribution for the DC units where parties advocated for amounts ranging from $23 million to zero. The Settlement was no less controversial in its proposal that PG&E's annual contribution be $9 million in what PG&E called a "black box" settlement derived from negotiation rather than specific evidence. PG&E contends this is a reasonable and informed compromise based on the litigation risks arising from various assumptions and arguments, including inadvertently omitted costs. We have previously said we disfavor such settlements where underlying assumptions are not disclosed because of the lack of transparency by which to verify them.34 In contrast, this provision has some evidence to support it.
PG&E argued that it has good reasons for an increase to the DC cost estimate: $135 million in omitted labor termination expenses and use of a higher (35%) contingency factor. DRA disputed that there was evidence to support either argument and pointed out that using the updated trust fund balances and original application assumptions, PG&E would only need to make about $5 million35 in contributions to the DC trust funds. PG&E replied that if the revised costs are incorporated with updated balances, the required annual contribution would rise to $29 million.
DRA's suggested contribution level was $1.8 million based on the updated fund balance and Settlement assumptions, except for substitution of the SCE LLRW burial escalation rate.36 We agree with DRA that the evidence in support of a 35% contingency for the DC cost estimates is limited37 and the omission of the claimed (and untested) labor termination costs is PG&E's error. However, this does not end the analysis. The goal of these proceedings is to adequately fund the trust funds based on reasonably accurate cost estimates. PG&E presented uncontested evidence that its updated annual DC contributions would be $16.76 million38 if it included the omitted labor termination costs and accepted the adjustments to its labor escalation rate and a five-year ramp down of equities after decommissioning begins, as set forth in the Settlement and adopted herein.
The record shows that SCE included labor termination costs without dispute and PG&E could argue that it should also have included them as a relevant cost (subject to protest for late submission). Moreover, the Commission is charged with assuring that the trusts are adequately funded by the ratepayers who receive the benefits of the generated power. There have been zero or nominal contributions approved for the DC trusts during the last two NDCTPs at a time when no detailed engineering studies have been done to assess contamination and certain costs have been omitted. Based on our review of the cost estimates and experience with rising costs as decommissioning becomes imminent, we find that these trusts are now underfunded.
Given these various considerations, a contribution of $9 million is within the range of likely outcomes had the Commission arrived at its own figure from a range of $5 - $16 million. Therefore, we find that the $9 million annual contribution is reasonable and justified and within likely litigation outcomes.
The HB3 trust funds have declined in value,39 only non-qualified funds are at issue, and the overall contribution has increased by more than $3.5 million40 assuming no other changes. Since this decision adopts the proposed changes to labor escalation and equity ramp down proposed in the Settlement, the HB3 contribution would increase by another $23,000. There is no dispute as to either proposed contribution and, therefore, we find PG&E's revised contribution to the HB3 trust funds to be just and reasonable.
The SONGS Unit 1 and PV trust funds are adequately funded so that no contributions are required in this triennial period. SCE originally sought approval for $64.537 million in total annual contributions for SONGS Units 2 and 3, which results in a total revenue requirement of $66.430 million.41 This would have been a 43% increase over the requirements authorized in the 2005 NDCTP. However, the updated trust fund balances alone would cut that to about $47 million.42 For the reasons discussed below, we adopt an even lower contribution amount.
TURN originally said no contributions were necessary for the SONGS Units 2 and 3 trust funds if SCE adopted TURN's proposed changes to the cost estimates. By our adoption of TURN's revised equity rate of return for SCE, as well as the updated trust fund balances, SCE's necessary contributions are reduced by half to about $23 million.43
DRA did not dispute the proposed contributions but argued that surplus funds were available in the SONGS Unit 1 trust funds that should be considered available for SONGS Units 2 and 3 decommissioning. However, we believe this view is premature given the uncertainties about radioactive waste disposal which could increase SONGS Unit 1 costs in the later phases.44
Based on the approved cost estimates for SONGS Units 2 and 3, inclusive of the revised equity rate of return we have adopted, SCE's revised contribution amounts and revenue requirements that result are just and reasonable.
SDG&E originally sought approval of an annual $15.284 million contribution to the SONGS Units 2 and 3 trust funds for its proportional share of the decommissioning expenses, plus continued recovery of $0.959 million related to SNF storage costs. Rather than seek a rate increase, SDG&E proposed to instead use overcollections in its NDAM and other balancing accounts or regulatory accounts to offset the revenue requirement. As discussed in Section 7.4.1 above, we are adopting a higher rate of return for SDG&E's equity investments which results in a lower contribution amount needed from ratepayers. Based on the updated trust fund balances, the company's annual contribution request has dropped to about $8 million.45
Based on the foregoing, we find SDG&E's revised contribution amount and proposal to fund the resulting revenue requirements out of existing balances to be just and reasonable.
7.5. Other Policy Issues
There was no objection to SCE's request to terminate its Decommissioning Tax Memorandum Account because it is unnecessary, and SCE has agreed to explore the feasibility of a separate NRC license to operate the ISFSI at SONGS Unit 1. As part of the proposed Settlement, the parties proposed solutions to other policy questions, which we adopt here. For example, in the next NDCTP, the utilities will provide, for information only, estimates of changes to funding for decommissioning associated with prospective license renewals for the SONGS Units 2 and 3 and DC Units 1 and 2. Also for the next NDCTP, the utilities will report the amount of pro rata share of funds held to meet NRC standards for License Termination, including copies of their most recent funding assurance letters to the NRC. For this NDCTP, we also accept the parties' agreement to allow the utilities to use different treatment of unrealized capital gains and losses when calculating the liquidation value of the trust funds.
The question of whether utilities should consider or assume in future NDCTPs that the trust funds will contain cash or some limited amount of equity investment for a period after shutdown or commencement of decommissioning is referred to Phase 2 of these proceedings.
7.6. Independent Panel
The level of decommissioning funds accumulated by the utility trust funds in California is high when compared with other states. It is unclear whether this is a result of appropriately conservative estimates, excessive caution, or mistaken assumptions. Therefore we agree with the parties that it is time to explore in detail the technical aspects of how decommissioning cost data is developed and presented so that the public, ratepayer advocates, and the Commission can better understand, analyze, and compare factors within the cost studies.
We adopt, with some modifications, the proposal in the Settlement to create an independent panel for the discrete task of improving the external review of cost estimates presented in NDCTPs. The panel will be comprised of individual decommissioning cost experts that worked with the utilities and TURN in these proceedings and, therefore, are also familiar with California's specific nuclear facilities: Nick Capik of ABZ, Geoffrey Griffiths of TLG, and Bruce Lacy of Lacy Consulting.46 Lacy would sit as a representative of consumer interests. DRA is concerned that these experts will not be "independent" of the utilities, and seeks a role for Commission staff and non-Settling parties. However, DRA was more interested in being kept in the loop than in sitting on the panel. Fielder also argues that the panel would leave out important parties, although he admits he declined to participate.47
We disagree because these arguments miss the point and purpose of the panel's work. The Commission has an interest in having the data presented in a form that is useful and comparable. Here, it makes sense to identify the experts needed for a rarefied technical task who have also agreed to work together for the benefit of California ratepayers.48 The panel will review volumes of technical data and their own proprietary models to develop recommendations to the Commission about how to improve transparency in decommissioning cost estimates for the benefit of the Commission and public, including Fielder and the DRA. The result is advisory, relates to the presentation of cost data, and does not in any way substitute for the NDCTPs or limit future participation.
TLG and ABZ are among the few nationally recognized experts in the field of decommissioning costs. They have prepared the cost estimates for the utilities in prior NDCTPs and, consequently, are among the best informed persons about past practices and current trends. Lacy was TURN's expert witness on decommissioning costs on behalf of ratepayers and is familiar with the ABZ and TLG studies used in these proceedings. We agree with TURN and the utilities that this is a vitally important task best tackled by experts familiar with nuclear decommissioning costs and experiences nationwide, as well as the unique characteristics of California's individual sites. Notably, neither DRA nor Fielder offered similar witnesses at the evidentiary hearings.
Moreover, we adopt several steps to assure the panel's work is useful and comprehensible. Similar to what the Settlement proposed,49 we require the panel to discuss the status of its work, listen to comments, and answer questions to be sure the resulting recommendations improve public review of cost estimates. Documents used in the development of the report would be available for review. The following opportunities for Commission staff and the parties to be included should occur:
· Within 30 days after adoption of this decision, the panel shall conduct a briefing about the panel's initial work plan.
· The panel shall conduct a briefing when it has completed the bulk of its work and considers its findings to be ready for presentation in draft.
· The utilities shall provide reasonable notice of the briefings to the parties in these proceedings.
· Upon notice to the ALJ, a workshop will be scheduled within these proceedings where the panel will present the Report for review and comment by all parties and Commission staff, including response to questions and feedback.
· The panel will issue a final report with recommendations which shall be filed in the consolidated proceedings by March 1, 2010, unless the ALJ extends the date.
Although Fielder rather rhetorically describes the panel as a "star chamber" which would "hijack" the NDCTPs,50 we think he misunderstands the limited nature of the assigned tasks. All of the identified technical issues were raised during the proceedings, in part due to frustration of the parties and the ALJ when trying to test, analyze, and compare bits and pieces of the cost estimates.51 The differing cost formats, assumptions, and definitions made it quite difficult and sometimes impossible. We are concerned that going forward, as more decommissioning costs and expenses are submitted for approval, we will lack clear benchmarks and comparables by which to make fully informed judgments of reasonableness.
We find the scope of activities set forth in Section 2.2 of the Settlement Agreement to be appropriate. This is a unique opportunity to get information about decommissioning activities in other states, determine what cost and financial assumptions can be applied on a common basis, identify state-of-the-art ideas about how to reduce costs, and, importantly, to find a common format for cost estimates to improve the quality of future scrutiny, analysis, and public participation.
The panel will limit its focus to PV, DC and SONGS Units 2 and 3 because these units are of similar size and design, still operating, and nowhere near commencement of decommissioning.52 Fielder objected to the exclusion of HB3, but HB3 is a unique facility in many respects and is already into the decommissioning process.53 Therefore, its exclusion does not diminish the usefulness of the panel's recommendations.
Finally, we adopt a $275,000 budget cap, instead of the proposed $250,000 budget cap, to funding of the panel's work, because of additional assigned tasks. The Settling Parties proposed that the costs be paid by the three utilities through the NDAM accounts and we agree that this nominal cost is an appropriate decommissioning expense. The actual allocation is based on the nuclear generating capacity of the DC Units 1 and 2, SONGS Units 2 and 3, and PV Units 1, 2, and 3.54 It is our expectation that the panel's recommendations will enhance the Commission's ability to exercise its statutory review obligation, likely lead to decommissioning cost savings, and assist the public in its analysis of future decommissioning cost estimates. The nominal impact on rates should be readily recovered in the value of these probable results.
7.7. Reasonableness Review
We reject the proposal to extend the form of reasonableness review applied to Phase 1 of SONGS Unit 1 decommissioning expenditures to Phases 2 and 3 and to all phases of HB3. It is neither in the public interest nor reasonable in light of the whole record. The rebuttable presumption method was accepted as part of an unopposed settlement in the first NDCTP prior to any actual decommissioning activities. It employed a model drawn from another purpose (i.e., Energy Cost Adjustment Clause reviews)55 that was not subject to close examination. Based on the knowledge and experience since gained by the Commission, it is clear that this is an important review process, influenced by speculative cost estimates and safety concerns, not suitable for an abbreviated method of oversight. At this time, we find that a full after-the-fact review of both costs and conduct best serves the interests of ratepayers and the public.
Pub. Util. Code § 8325(c) allows the utilities rate recovery for "reasonable and prudent decommissioning costs." In D.99-06-007, the Commission authorized the commencement of SONGS Unit 1 decommissioning and a form of expenditure review that applied a rebuttable presumption of reasonableness to decommissioning activities based solely on completing work within an approved cost estimate. SCE was required to submit cost estimates and expenditures, along with its explanation of "material" differences in future NDCTPs. Absent "material" cost variations, the burden to show unreasonableness was shifted to other parties.
Unlike that presumption, the Commission described in the 2005 NDCTP its standard of reasonableness review for other decommissioning expenditures:
[W]e define reasonableness for decommissioning expenditures consistent with prior Commission findings; i.e., that the reasonableness of a particular management action depends on what the utility knew or should have known at the time that the managerial decision was made.56
Going forward, we affirm this is the appropriate review to apply to actual decommissioning expenditures.
PG&E argued that it wanted "a process in place" by which it could evaluate how it would conduct decommissioning. It said that making advice letter estimates "and then having the completed projects reviewed, really isn't appropriate for this phase of the proceeding."57 Essentially, PG&E contended that it was far better for the company to move review into the estimate phase instead of questions being raised after the fact. No actual review would be lost, said PG&E, because the presumption is rebuttable. We disagree.
The crux of PG&E's concern seems to be that the Commission would retroactively micromanage the decommissioning process. Its concern is somewhat misplaced because the Commission is not in the business of managing the decommissioning of a nuclear facility. Yet, the Commission is charged with assuring that ratepayers are not liable for unreasonable costs and that decommissioning activities are prudently undertaken. The utility wants to assure the Commission solely through its cost estimates that they will hire appropriate people and spend appropriate amounts doing the right things safely. This is a leap of faith we are not prepared to take. We now know that cost estimates keep growing, unexpected things occur, the extent of contamination is unknown until it is removed, and that not all those expected to be hired have been hired at the time of the cost estimate.
SCE's arguments in support of the proposal centered on the claim that the presumption "worked well" for the Commission's review of its Phase 1 expenditures for SONGS Unit 1 and has been approved in each successive NDCTP.58 The utility emphasized that the cost estimates were highly detailed and accurate and any party could challenge the costs even if within the estimate. Whether it "worked well" for SCE is not the same question as to whether it "works well" for the public. Cost estimates for remaining phases at the SONGS sites grew dramatically since the last NDCTP. SCE admitted learning a lot in Phase 1 as costs rose and it continued to grow the estimates for SONGS Units 2 and 3. Neither past use of the presumption, nor assurances of the reliability of a cost estimate, are persuasive reasons to alter the more complete, after-the-fact review set forth in D.07-01-003 for the benefit of ratepayers and the public.
SCE disputes Fielder's view that the presumption creates a "lighter burden of proof" and contends the utility has made the same evidentiary showing of expenses necessary to sustain a finding of reasonableness, notwithstanding the applicability of the presumption. SCE further notes that no one has disputed their costs, nor did Fielder even ask a question about them. This may be so, but it does not change the fact that the prudence review has been subsumed by the cost analysis, nor does it address whether the presentation of the data is functionally penetrable by the parties and Commission staff in the time available during the NDCTP.
We have related policy concerns with application of a presumption, albeit rebuttable, to the most important part of our review of the decommissioning of California's nuclear facilities. For example, cost estimates are not as reliable as the utilities claim, nor are they the final word as to what activities are conducted and by whom. The fact there is wide agreement that cost estimates are opaque, inconsistent between utilities, and rely on disputed assumptions, underscores the limited reliability of an estimate even as decommissioning approaches. That is why the work of the independent panel is so important for improving future review of cost estimates. It also illustrates why the Commission and other parties may have difficulty reviewing the expenditures within the time available and matching them to work scope in order to test the presumption.
Another concern is that SCE and PG&E are acting as their own general contractors for the decommissioning. This is uncharted territory which may yield cost benefits to ratepayers but includes risk of myopia from exclusion of third-party perspectives about operational practices affecting costs. Fielder called it a "conflict of interest" and said, "[O]nly the utilities will know what they did and when they did it...."59 Similarly, at the evidentiary hearings, TURN's counsel said:
Essentially, they're asking the Commission to decide that that money belongs to the utility, not to the ratepayers, and they want an upfront guarantee that they can spend these funds irrespective of what facts may come to light in the future or how the utilities actually behave, and perhaps most importantly, whether actions that the utility has taken are contributing to the increase of those costs.60
TURN dropped its opposition to this proposal as part of the Settlement, presumably because it gained agreement on the independent panel and other changes in utility assumptions. However, that does not eliminate the importance of these concerns for the Commission.
The policy problem is amplified by the fact that neither PG&E nor SCE officially submitted their decommissioning plans to the NRC for substantive review because such submission is not required unless in connection with a license termination. Absent NRC oversight, the NDCTP seems to be the only regulatory review of their actual decommissioning plans. Therefore, the Commission is the front-line agency in position to examine whether the decommissioning is done prudently. Adoption of the reasonableness presumption would inappropriately submerge the character of the activities within a cost test that fixes the burden of proof.
We are not comforted by the utilities assurances that the data is submitted for review regardless of whether there are material cost differences, and parties have the ability to challenge costs and prudence even if the presumption applies. If the presumption does not alter the evidentiary showing, then it seems of little benefit to the utility. More importantly, we find that the Commission's duty to review decommissioning activities to assure the costs were prudently incurred, in addition to being reasonable, is too significant to lump into a presumption solely based on cost. Furthermore, the inclination to overestimate costs could arise.
Based on the foregoing, we conclude that it is not in the public interest nor reasonable in light of the whole record to provide, going forward, a presumption of reasonableness for decommissioning activities which are completed within cost estimates. This finding is sufficient to reject the Settlement as a whole.
18 Exhibit SCE-1 at 11; PG&E Supplemental Testimony at 3-1 through 3-3.
19 As discussed in more detail below, there is inconsistency between the utilities as to whether this factor covers only engineering contingencies or other unknown risks.
20 The cost studies intended to account for recent changes in technology, regulation, and economics and also account for the unique features of each facility.
21 The SONGS units sit on land owned by the United States Department of the Navy and there are significant uncertainties about the required standards for final site restoration and site lease termination. Therefore, SCE and SDG&E have made very conservative assumptions about the amount of contamination they must remove.
22 Reporter's Transcript at 565.
23 After a new LLRW burial site becomes available to California nuclear generation facilities, we expect the utilities to review the escalation rates using then current data.
24 For example, APS assumed that it would incur no costs for disposal of non-contaminated materials or final clean-up following United States Department of Energy disposal of SNF.
25 D.07-01-003 adopted a 25% contingency rate for HB3 in the 2005 NDCTP.
26 Reporter's Transcript at 201-203.
27 Reporter's Transcript at 849.
28 86 CPUC2d 604, 620 (Attachment A Settlement Agreement § 4.2.2.2(c)).
29 Exhibit SCE-1 at 14.
30 The forecasted rates of return are adjusted for management fees, taxes, and equity turnover rates.
31 Reporter's Transcript at 853.
32 Exhibit PG&E-20, Exhibit SCE-15, and Exhibit SDG&E-20.
33 Exhibit PG&E-1 at 8-2, Table 8-1.
34 D.88-02-030, 1988 Cal PUC 100 at 32-33.
35 Exhibit PG&E-20.
36 Exhibit DRA-10.
37 When asked, PG&E's expert said he would not object to 35%. Reporter's Transcript at 203.
38 Id.
39 The HB3 non-qualified trust funds are predominately in fixed income investments.
40 Exhibit PG&E-20.
41 Exhibit Utilities-3 at 24, Table III-12.
42 Reporter's Transcript at 851.
43 Exhibit SCE-15.
44 There are also unresolved tax implications arising from fund transfers because California has not adopted certain changes in federal tax law relative to Internal Revenue Code § 468-A.
45 Exhibit SDG&E-20.
46 Reporter's Transcript at 779.
47 Reporter's Transcript at 808.
48 Reporter's Transcript at 805.
49 Opening Brief of SCE, SDG&E, PG&E and TURN at 2-3.
50 Fielder's Post-hearing Reply Brief at 3.
51 Reporter's Transcript at 780.
52 Reporter's Transcript at 800.
53 Reporter's Transcript at 800-801.
54 PG&E's allocation would be 44.78%, SCE's allocation would be 46.62%, and SDG&E's allocation would be 8.60%. See Attachment A to the Settling Parties' Post-hearing Opening Brief.
55 86 CPUC2d 604, 615.
56 D.07-01-003 at 7-8.
57 Reporter's Transcript at 504-505.
58 Reporter's Transcript at 478, 480.
59 Intervenor Scott Fielder's Reply Brief at 5.
60 Reporter's Transcript at 499.
