12. Cost Recovery
12.1. General Proposal
Regarding cost recovery of the Upgrade, PG&E proposes the following ratemaking treatment:
· Rates will be set initially to recover forecasted project costs, including the incremental costs and benefits of the SmartMeter Program Upgrade; with true-up to actual costs achieved through the existing SmartMeter Balancing Account - Electric (SBA-E).
· The Commission will review forecasted incremental costs in this application and, as a result of that review, these forecasted costs will be deemed reasonable and will not be subject to after-the-fact reasonableness review. If actual costs exceed the forecast, then PG&E proposes to file for recovery of the difference through a traditional after-the-fact reasonableness review filing.
· Costs associated with the SmartMeter Program Upgrade incurred prior to a Commission decision of this application and recorded in a memorandum account, upon approval of the advice letter filed concurrently with this application, will also be reviewed in this application, and as a result of that review, these incurred costs will be deemed reasonable and will be transferred to the SBA-E for recovery.
· Incremental benefits or cost reductions will also be reviewed in this proceeding, and specified pre-approved forecasted benefits will be incorporated into rates through the SBA-E as associated project milestones are met.
· Rates covering the SmartMeter Program Upgrade, including the incremental costs and benefits, will be revised annually in the Annual Electric True-Up advice letter, or as otherwise authorized by the Commission.
As ordered in D.06-07-027,105 PG&E indicates that it will present testimony in its next GRC concerning the continuation of the balancing accounts as an alternative to traditional ratemaking treatment.
No party has challenged PG&E's general cost recovery proposal as described above. It is reasonable and will be adopted. However, parties have challenged certain aspects of PG&E's allocation methodology, as well as the benefits recognition proposal, as discussed below.
12.2. Generation/Distribution Allocation
PG&E proposes to recover the SmartMeter Program Upgrade costs from customers in the same manner as adopted in D.06-07-027 for other SmartMeter Program costs. That is, the total revenue requirement will be recovered in the same manner as other distribution revenue, based on the distribution revenue allocation and rate design methods authorized by the Commission at that time.
Since PG&E justifies the Upgrade costs primarily on demand response and energy conservation benefits, DRA recommends that any Upgrade costs approved by the Commission be allocated by a generation allocator. According to DRA, savings due to peak load reduction and energy conservation typically flow through an energy resource recovery account, from which the account balance automatically flows to customer classes based on a generation allocator. This means that, if the potential benefits of the Upgrade do occur, the energy saving benefits would flow back to customer classes accordingly. For the residential class, the generation allocator is approximately 40.6%. DRA argues that, as the residential class would obtain 40.6% of potential benefits, it makes sense that they also pay 40.6% of the costs. According to DRA, PG&E's proposal to allocate AMI Upgrade costs by a distribution allocator would allocate 55.1% of these costs to the residential class. DRA states that PG&E is thus recommending that residential customers pay far more than they would potentially benefit from the Upgrade. DRA instead recommends that the Commission allocate any approved Upgrade costs by generation allocators that would allocate approximately 40.6% of these costs to the residential class.
In response, PG&E states that DRA's proposal is inconsistent with established practices of cost allocation. PG&E notes that DRA acknowledges PG&E's proposal follows the method already being used to recover those costs authorized by PG&E's Original AMI Case. PG&E also notes that its proposal is consistent with the method adopted by the Commission in SDG&E's recent AMI case, as well as DRA's settlement with SCE on its recent AMI case.106 Furthermore, PG&E is not aware of any cases where distribution infrastructure costs have been allocated on a method other than to distribution-level EPMC.
At this point, we will continue the use of the allocation methodology that applies to PG&E's original AMI authorization. In general, it is reasonable to allocate distribution infrastructure with distribution level EPMC related allocators, and PG&E's methodology is consistent with how SDG&E's AMI related costs are allocated. We will not preclude DRA, or any other party, from raising the issue in PG&E's next Phase 2 GRC proceeding. In fact, that would be a more appropriate forum for proposing such an allocation methodology that is based on principles which differ significantly from existing principles.107
12.3. Streetlight Allocation
CAL-SLA argues that PG&E does not need a meter to determine street light energy usage.108 According to CAL-SLA, PG&E already has more than sufficient information to determine annual energy usage from streetlights, so a meter would be surplus. CAL-SLA also notes that while some other customers might use the SmartMeter to alter their energy usage pattern, it is not the case with street lights, since they only operate at night.
CAL-SLA's policy position is, since SmartMeters will not be installed on street lights because they are unnecessary, street light customers should not pay for SmartMeters. CAL-SLA points out that it has never contended that street light facility charges which are unique to street lights should be assessed against all other customers.
PG&E disagrees with CAL-SLA's position for the following two reasons. First, it is at odds with the Phase 2 GRC Settlement, of which CAL-SLA was a signatory. Second, CAL-SLA's position ignores the benefits that would accrue to streetlights customers from the Upgrade. According to PG&E, street light customers will receive benefits as a result of many of the improved operating efficiencies that will benefit all customer classes, such as reduced labor costs and improved cash flows. PG&E also notes that street light customers will benefit from the new peak load management efforts and energy conservation efforts that should result in lower overall generation and distribution revenue requirements.
12.4. Discussion
In addressing this issue, we agree in general with PG&E's position that, while street light customers will not receive any benefits directly associated with having an upgraded meter, there are likely to be some benefits to street light customers due to the Upgrade, in the form of increased operational efficiencies and reduced revenue requirements. For this reason, it is reasonable to allocate some amount of the Upgrade costs to street light customers. We also feel it is reasonable to use the settlement in PG&E's last rate design settlement to do so.
In its testimony, CAL-SLA states the following:109
PG&E states that in Exhibit C, Table 1, the revenue allocation methodology is to allocate distribution revenue to each class based on each class' total share of present distribution revenue. For the street light class, revenue from facilities charges is included in distribution revenue used for the basis of the allocation. The inclusion of facilities charges causes the percentage increase for the street light class to be higher than for other classes and the systemwide percentage change.
PG&E goes on to state that the revenue allocation methodology used in the SMU application is not what was approved in D.07-09-004 in Phase 2 of the utility's 2007 Test Year General Rate Case.
CAL-SLA recommends that the Commission use the revenue allocation methodology adopted in the Phase 2 GRC D.07-09-004. Street light facilities charges should be treated as non-allocated revenues and therefore excluded from revenue allocation. Under the Phase 2 revenue allocation, street light's increase would be reduced from 1.7% to 0.5%.
The use of the Phase 2 GRC decision revenue allocation methodology for allocating the Upgrade revenue increase is apparently a secondary recommendation of CAL-SLA, whereby the street light customers' increase would be reduced when compared to PG&E's proposal for the Upgrade. In rebuttal testimony, PG&E states, "Yes. PG&E agrees that D.07-09-004, as issued in Phase 2 of PG&E's 2007 GRC, sets forth the appropriate methods for changing rates that may result from a change in revenue requirements to recover the costs of the Upgrade project."110
There were a number of settlements in Phase 2 of PG&E's 2007 GRC, which addressed marginal costs, revenue allocation and rate design. In the particular settlement on marginal costs and revenue allocation,111 Section VII.3 addresses rate changes between GRCs. The Upgrade will result in a rate change between GRCs, so it is appropriate that the Section VII.3 principles in the marginal cost and revenue allocation settlement should be followed in determining the allocation of Upgrade costs to the various customer classes. PG&E should allocate the Upgrade revenue increases accordingly.
CAL-SLA indicates that its primary recommendation does not comport with the Phase 2 GRC settlement but adds that SmartMeters were never identified in that proceeding as a cost to be allocated to street lights.
We do not know what was assumed by the settling parties, including CAL-SLA, when the marginal cost and revenue allocation settlement agreement was reached. Settlements generally represent a compromise among the Settling Parties' respective litigation positions, in order to agree on a mutually acceptable outcome. What may not seem to be fair, when viewing a portion of the settlement in isolation, may be fair, when viewing the settlement in its entirety. We can only judge issues such as this by the plain language of the settlement. Authorization of the Upgrade necessitates a rate change between GRCs. The settlement provides principles for rate changes between GRCs. There is nothing in that section of the settlement that limits the application of those principles, if the increase is driven by SmartMeter costs or any other specific costs. There is nothing that states that certain customers can avoid an increase, if the reason for that increase does not directly benefit those customers. In order to honor the settlement process, we have no alternative but to impose the principles for rate changes between GRCs, as identified in PG&E's TY 2007 Phase 2 marginal cost and revenue allocation settlement, in allocating the Upgrade related revenues to customer classes. In doing so, street light customers will receive an allocation of Upgrade costs, although that allocation will be substantially lower than what was originally proposed by PG&E.
By our determination today, we are not precluding CAL-SLA or any other party from raising the issue of how SmartMeter costs should be allocated in PG&E's next Phase 2 GRC proceeding. We expect such an issue would necessitate a fairly comprehensive analysis of what types of costs, beyond just SmartMeter costs, directly benefit or do not directly benefit the various customer classes and which of those costs should be assigned to particular customer classes.
12.5. Benefits Recognition
PG&E proposes to continue the current mechanism for recognizing benefits resulting from the Upgrade on a monthly basis as meters are activated and project milestones are achieved. Specifically, once the remote connect/disconnect functionality has been activated (expected in the latter half of 2009), PG&E would adjust the existing per electric meter monthly benefits calculation from $1.7722 per active electric meter per month by an additional $0.1821 per active electric meter per month, to be in effect through the end of 2010. Starting with 2011, these amounts would be subject to revision through PG&E's GRC or other applicable regulatory mechanisms. DRA and TURN dispute the timing of PG&E's benefits recognition proposal.
DRA recommends that PG&E track and report the differences between the AMI benefits actually credited to ratepayers and those shown in PG&E's business cases, for both the original and Upgrade applications. DRA recommends that PG&E should automatically credit ratepayers with the benefits of both the original and Upgrade projects eight months after meter costs enter into the rate base. This will ensure that ratepayer benefits are not delayed due to further deployment delays. According to DRA, continuing the benefits recognition proposal adopted in the original AMI decision unfairly allocates a disproportionate share of the financial risks to ratepayers.
PG&E states that adhering to DRA's proposed timeline would reduce PG&E's incentive and flexibility to actively manage and reduce project costs. For instance, PG&E indicates that its management currently has incentives to take advantage of volume discounts for purchasing materials during a certain period of time, and for taking advantage of tax rules that can provide benefits from accelerating the purchase of items during a certain tax year. In order to take advantage of these discounts, PG&E may need to buy items in advance of what would be needed for the deployment schedule. A mandate to begin crediting customers eight months from the booking of such costs into rate base would provide a disincentive to PG&E from taking advantage of these discounts, resulting in higher project costs. PG&E indicates this would also increase the administrative burden and therefore the cost of running the project. Hence, PG&E believes that it would be prudent to adhere to the current benefits recognition method under which PG&E commences recording benefits only after the meter is activated.
TURN states that PG&E's AMI pre-deployment and AMI deployment funding requests were both authorized, in large part, because the tangible operational cost savings flowing back to ratepayers were supposed to pay for approximately 90% of the project costs; and PG&E is significantly behind in crediting ratepayers with the per-meter operational benefits that were included in PG&E's originally authorized AMI program. TURN asserts that because PG&E's AMI project is so far behind schedule, for both gas and electric meter deployment, as compared to the deployment forecast authorized in D.06-07-027, only negligible operational cost savings have been credited back to ratepayers to date (less than 18% of total costs). TURN therefore recommends that PG&E be directed to credit at least $44.8 million in operational benefits back to ratepayers as part of this proceeding.
It is TURN's position that, given that so few operational benefits are being provided as planned, combined with the time value of money where costs and benefits in earlier years are weighted more heavily than in the outer years, PG&E's original 90% operational cost-effectiveness will no longer be achievable unless the Commission orders a crediting back to ratepayers.
In response, PG&E provides three reasons why it believes TURN's proposal should be rejected.
First, according to PG&E, the values used by TURN to calculate the level of expected benefits were forecast estimates and never meant to be-nor did they become-required targets set by the Commission. TURN's recommendation to, in essence, require PG&E to record benefits in accordance with such a schedule is contrary to the method adopted by the Commission in D.06-07-027. That method requires PG&E to record in the balancing accounts revenue requirement costs and agreed-upon benefits only after meters are activated, not in accordance with some prescribed schedule. The Commission stated:
We find PG&E's proposed balancing account mechanism, with a per meter benefit credit, to be reasonable because PG&E recovers its new AMI-related costs on an actual basis and it ensures ratepayer benefits are captured as meters are activated. (D.06-07-027, p. 51.)
PG&E notes that in adopting this mechanism, the Commission expressly rejected a competing ratemaking proposal from TURN that would have levelized costs and benefits according to a prescribed schedule somewhat analogous to that proposed here by TURN. The Commission rejected TURN's proposal stating that it was not persuaded by TURN "[T]hat such a method is reasonable for either ratepayers or shareholders."112
Second, PG&E states that TURN's argument ignores the fact that recorded costs have also trended behind the original forecasts; and while TURN argues that benefits are trending $45 million behind schedule, the costs of the project are trending $161.9 million behind the original schedule. PG&E argues that this "delay" in expenditures dwarfs the value of "delayed" benefits, a fact that benefits ratepayers under the ratemaking scheme adopted by D.06-07-027.
Third, PG&E states that TURN's argument ignores the fact that PG&E's current deployment schedule still reflects an overall completion timeframe of five years as per the original timeframes within the AMI case; and any "delay" in benefits or costs will be short-lived with project benefits accelerating during the later years of deployment.
We see no compelling reason to change the benefit recognition procedures adopted in D.06-07-027 and will not adopt DRA's proposal. We recognize that DRA's proposal is similar to the benefit recognition procedure that was included in SCE's AMI decision. However, it is not clear from the record that, over the long term, the DRA proposal will be more beneficial to ratepayers. Consistency is important, but being consistent with the benefit recognition procedures previously found reasonable in D.06-07-020 is just as valid as being consistent with the settled procedure adopted for SCE. We have not been presented with evidence that suggests PG&E is mismanaging funds, and recognizing benefits when the meter is activated is reasonable, if only because no benefits can be realized until the meter is activated. Also, as PG&E indicates in responding to TURN, while benefits are trending $45 million behind schedule, the costs of the project are trending $161.9 million behind the original schedule. For that reason, we do not see any harm to ratepayers by continuing the existing procedures.113
Also, PG&E's reasons for rejecting TURN's $44.8 million ratepayer credit proposal are persuasive, and we will not adopt that proposal.
105 D.06-07-027, Ordering Paragraph 15.
106 The SCE AMI settlement defers consideration of the allocation methodology to SCE's GRC, and uses a distribution allocation for any interim period.
107 In this proceeding, the record on this issue is limited. Viewing it in the context of all of PG&E costs would provide a venue for considering all costs and applying the proposed principles in a consistent manner across all costs, if adopted.
108 According to CAL-SLA, out of the approximate 45,000 streetlight accounts taking service from PG&E, 1,000 are metered under Schedule LSD-3.
109 Exhibit 301, p. 8.
110 Exhibit 8, p. 5-3.
111 See D.07-09-004, Appendix B.
112 D.06-07-027, p. 54.
113 While rates will be set initially to recover forecasted project costs, including the incremental costs and benefits of the SmartMeter Program Upgrade; a true-up to actual costs will be achieved through the existing SmartMeter Balancing Account.
