7. Costs
7.1. Meter Devices
In its application request, PG&E forecast $402,656,000 for incremental meter and equipment costs.14 This amount covers HAN devices and load limiting switches for all customers, as well as the incremental costs associated with an advanced meter. PG&E indicates that it was, at that time, evaluating integrated meter devices proposed by a group of selected vendors and subsequently began to pursue an aggressive bidding process to obtain the best end-point technologies at the lowest possible price. In its May 14, 2008 Supplemental Testimony, PG&E indicated that it was then in the final stages of that process and had received "best and final" pricing from the remaining vendors in consideration. Due, in part, to the refined bids from these vendors, PG&E reduced its estimate for incremental costs associated with integrated meter devices to $342,789,000.15 As opposed to its original estimate, this amount also covers the costs of retrofitting solid state meters deployed in 2008 without a HAN device (Ubiquitous HAN or HAN Upgrade) and the cost of HAN repeater devices (HAN Connectivity). According to PG&E, this also reflects a price structure that includes the option for a substantially better warranty on the end-point technologies.16
There are a number of issues related to meter devices including DRA's estimate of meter device costs, the HAN retrofit, the electromechanical meter retrofit (also known as the Kern County retrofit), and HAN connectivity.
The only party to analyze the entirety of PG&E's proposed meter and equipment costs is DRA. Since DRA is supportive of the HAN and service switch, it recommends funding costs associated with this increased functionality. DRA estimates $267.3 million in incremental meter device costs derived from its own cost estimates for advanced solid-state meters that would have the same functionality as proposed by PG&E. DRA's consultant ultimately relied on confidential bids at his disposal from seven vendors. Having signed non-disclosure agreements to receive this information, DRA's consultant could not divulge the sources of this information or the underlying terms and conditions.
DRA notes that its consultant specifically used the lowest three bids amongst his sample set of seven, and that the average of the whole sample of seven produced a number in the same general range as PG&E's proposed cost. Knowing it could not produce enough benefits to justify PG&E's meter costs, DRA directed the consultant to use the lowest three to generate a "barebones" estimate. DRA also notes that the meters on which its consultant received quotes may have a lower level of functionality than do those that PG&E assumed in its presentation, however DRA states that it is unclear from the record what increased functionality PG&E's meters provide, or why this functionality is necessary.
From its cost estimate, DRA subtracted the funding that PG&E already received in A.05-06-028 for new or retrofitted meters. DRA also excluded all labor and network costs that were previously funded in A.05-06-028 except for labor costs associated with the Kern County retrofit. DRA included the labor costs for the Kern County retrofit because revisiting those meters would have been necessary anyway to provide the enhanced functionality.17
With regard to network costs, DRA's consultant states that further cost savings are available by using a single network for gas and electric meters in each geographical area. DRA was however unable to quantify these savings.
With regard to the determination of what meter costs were already approved in A.05-06-028 and should be subtracted from the cost of the advanced solid-state meters, DRA notes that in PG&E's May 2008 supplemental testimony, it assumed funding for a basic Tier 0 solid-state meter for all customers, while A.05-06-028 had only provided funding, for the residential sector, for replacing roughly one-third of the existing electromechanical meters, and merely refurbishing the rest of those meters at a fraction of the cost of a new one. PG&E's supplemental testimony includes a $61.1 million adjustment to its baseline costs for end-point technologies to reflect the estimated cost of the project decision to change totally from electromechanical to base solid state meters.18 DRA states it did not adequately understand this evolution in PG&E's thinking, and its consultant merely followed what had been authorized in A.05-06-028, which provided funding to replace only one-third of the existing electromechanical meters rather than providing solid-state meters to everyone. DRA believes it would be appropriate to modify its figures to put them on a comparable basis with PG&E's revised numbers, suggesting in errata that PG&E's $61.1 million reduction be used as a proxy for the effects of putting its numbers on a comparable basis.
DRA stresses that the $61.1 million is only a proxy of this reduction, and that a larger reduction can be achieved by directly substituting a blended cost for a Tier 0 basic solid-state meter, for the cost of new and retrofit electromechanical meters in its Table 2-1. According to DRA, doing this would more than compensate for other errors that PG&E alleges. However DRA indicates that it will refrain from further changing its estimates because there are compensating changes that could be made in both directions.
In response, PG&E states that DRA's original analysis is riddled with errors, which required DRA to make a number of corrections, one of which totaled nearly $200 million. Several additional errors were corrected in errata. PG&E indicates that it pointed out other errors to DRA that went uncorrected, including one that shorted PG&E about $10.5 million.
PG&E states that most importantly, after it pointed out DRA's errors, DRA changed its approach for this cost category and based its new recommendation on confidential pricing data from third parties that were never disclosed to PG&E. According to PG&E, DRA's unwillingness to disclose this third-party data -- on which it based its analysis -- deprived PG&E of its due process rights to examine such data and compare it to the data provided by PG&E.19 PG&E quotes the following from DRA:
...If you are asking me should PG&E know the other terms in order to effectively evaluate whether the product they are proposing to purchase is more cost-effective from their perspective than the alternatives I've proposed? I would say, yes, they need more information... .20
For the above reasons, PG&E argues that DRA's cost testimony should be given no weight.
DRA's recommended incremental cost for meter devices (the meter, disconnect switch, HAN gateway device and AMI module) is approximately $206 million, while PG&E's proposed amount is approximately $310 million.21 DRA's total cost estimate is approximately $471 million as opposed to PG&E's estimate of $575 million. With the evidence before us, we have little choice but to adopt PG&E's estimates of meter device costs. It is unfortunate that non-disclosure barriers prevents any detailed analysis of DRA's recommendation, but without some idea of what the differences are and whether those differences appropriately consider PG&E's situation and needs, we cannot adopt costs that are so different from that proposed by PG&E.
PG&E's estimate is based on costs derived from an RFP process. Based on responses to that process, PG&E conducted an evaluation of the integrated meter devices from certain vendors to help identify vendor and meter device technologies best suited to serve PG&E and its customers. According to PG&E, the vendors selected for further consideration were selected following a rigorous vendor selection process in order to ensure that the vendor ultimately selected has sufficient resources, credibility, and expertise to supply the necessary equipment and services to complete their work within an appropriate timeframe and budget. For a project of this magnitude such evaluation is prudent. However such evaluation cannot be performed with respect to the vendors and devices related to DRA's projected costs, due to the non-disclosure restrictions.
DRA's data, which according to DRA shows the average of the bids considered by its consultant as being in the same general range as PG&E's proposed cost, provides some additional assurance that PG&E's RFP approach is reasonable.
It would be inappropriate to impose DRA's proposed costs on PG&E without assurance that the related meter devices provide the necessary functions, without assurance that the vendors are capable of providing the equipment when needed, and without knowledge of the type of warranties that are associated with the costs.
For these reasons, we adopt PG&E's estimate of the incremental costs for meter devices. However, we will require that PG&E provide quarterly reports on the implementation progress of the Upgrade to the Commission's Energy Division and any interested parties. PG&E should consult with the Energy to determine what information PG&E should provide.
7.2. HAN Retrofit
As described in PG&E's testimony, the HAN retrofit22 involves PG&E deploying 288,000 upgraded meters with load limiting switches and upgrading these meters with HAN gateway devices at a later date. PG&E stated that one of the key principles guiding the company during its transition from electromechanical meters under the existing SmartMeter Program to the upgraded meters proposed in this proceeding was the objective of beginning deployment of solid state meters, preferably with load limiting switches and HAN devices, at the earliest strategic point in its deployment schedule. In its May 2008 supplemental testimony, PG&E indicated that it had recently learned that its preferred HAN devices were scheduled to become commercially available in the fourth quarter of 2008. Therefore, PG&E planned to install solid state meters that have a load limiting switch -- but that do not have a HAN device -- during the limited period between the time that PG&E completes the installation of the remaining electromechanical meters (e.g., summer 2008) and the time the HAN devices become available. To support real-time pricing, dynamic pricing, and opt-out programs for all customers, PG&E stated it will be necessary for PG&E to then retrofit these above-described solid state meters with HAN devices. PG&E estimated the net cost increase of such a retrofit will be approximately $30 million.
In support of its decision to proceed with the HAN Upgrade, PG&E's consultant, Lechner, performed an analysis of several meter deployment scenarios comparing lost benefits to reduced costs, if PG&E had suspended meter deployment until HAN devices became available. According to PG&E, the analysis indicates that lost benefits exceed reduced costs, and PG&E acted reasonably in moving forward with meter deployment without the HAN devices.
DRA excludes all costs associated with the HAN retrofit except those directly associated with enhanced functionality. DRA believes that PG&E could have merely suspended the deployment of solid state meters without a HAN device and avoided the additional costs that PG&E includes. DRA also criticizes PG&E's suspension analysis, stating that the cost-benefit analysis is distorted by three problems: (1) it ignores the present value cost savings of delaying the deployment of the subsequent five million meters; (2) it artificially truncates the stream of foregone benefits for all scenarios to 2011; and (3) it includes different numbers of months of foregone benefits for the four scenarios evaluated.
Regarding the first problem, DRA asserts that Lechner ignored the cost savings from delaying the deployment of some five million meters apparently because he did not find them to be important enough to include. According to DRA, the particular studies that led him to this conclusion are not in the record, but, because of this decision, the only endpoint costs Lechner includes in his analysis are those associated with the 288,000 meters, which he then compares with the foregone benefits associated with over five million meters. DRA states that the result is predictable - the benefits dominate the analysis.
The second problem, according to DRA, is that Lechner truncated the period of analysis such that it would end in 2011, in spite of the fact that the benefits persists for the projected 20-year life of the endpoints for all four scenarios he considered. DRA asserts that had Lechner not truncated the benefits streams, the benefits in nominal terms for each of the four scenarios would have been identical. The only difference would have been in the timing of the benefits.
DRA's third problem has to do with Lechner truncating the benefits of all scenarios to the end of 2011, which resulted in a five fewer months being used to calculate the benefits for the five-month scenario relative to the non-suspension scenario. According to DRA, had he allowed the benefits streams to continue for the lifetime of the equipment, the benefit streams for all the scenarios would have included the same number of months. The only difference would be the point in time when they would have occurred.
In response, regarding DRA's allegation that Lechner's analysis ignores the present value cost savings of delaying the deployment of the subsequent five million meters, PG&E states that Lechner specifically considered the cost implications of suspending five million meters and the analytical result was the basis for his conclusion, and cites the following cross-examination:23
DRA Counsel: Mr. Lechner, in your analysis did you include or consider the impact of delaying the cost of deploying 5 million meters?
PG&E witness Lechner: During the course of my analysis and analyzing the implications of the cost, I considered that, whether that would have an impact on the end result.
Q: What was your conclusion?
A: The conclusion is ... as I refined the model on the cost side by contemplating the time value of money under various different delay scenarios, in conjunction with additional escalation, in conjunction with additional inefficiency costs, in conjunction with the additional costs that would be incurred, each scenario that I looked at had no implications, no impact on the overall result, I drew the conclusion that the cost side of this model really isn't driving the equation. It's the benefits side.
Thus, PG&E asserts that, counter to DRA's allegation that Lechner ignored the cost savings from delaying the deployment of some five million meters apparently because he did not find them to be important enough to include, the record shows that Lechner specifically considered the cost implications of suspending five million meters and the analytical result was the basis for his conclusion. PG&E emphasizes that the only cost "savings" from a suspension scenario are related to the time value of money associated with deferral, and notes that Lechner specifically considered these "savings," but, unlike DRA, Lechner also considered the significant additional costs associated with suspending endpoint deployment.
PG&E states that DRA's second allegation -- that Lechner's analysis "artificially truncates the stream of foregone benefits for all scenarios to 2011" and that this "inflates" the differential between the lost benefits between PG&E's business case and a suspension scenario -- is wrong both in theory and application, with the following explanation:24
From a theory standpoint, Mr. Lechner properly pointed out during cross examination that in doing a comparative analysis between a continued vs. suspended deployment scenario, it is necessary to compare the same period of time. By comparing the stream of benefits generated by continuing deployment with the stream of benefits generated by a suspended deployment over a defined period of time, Mr. Lechner was able to determine the present value of "lost benefits" caused by a delay scenario. As Mr. Lechner also pointed out during cross-examination, extending the period of time to evaluate lost benefits caused by a suspension scenario does not change the fact that benefits accrue at a faster rate under the continued deployment scenario than they do under a suspension scenario.
From an application standpoint, DRA erroneously attempts to link the benefits associated with meter deployment to the estimated 20-year life of the endpoints and fails to consider the compounding nature of benefits over time. The estimated 20-year life for endpoints is not relevant for purposes of analyzing the economic impact of a deployment suspension scenario. Benefits begin to accrue when an endpoint is installed and activated. A large percentage of the operational benefits created by this endpoint activation are due to PG&E's ability to avoid the labor costs of meter readers on activated SmartMeter routes. When an endpoint reaches the end of its useful life, the meter will be repaired or replaced and the benefit stream will continue, uninterrupted (e.g., PG&E will not re-hire its meter readers at the end of the estimated life of a SmartMeter). This is another reason why it is essential to use the same end date for all scenarios in a comparative analysis of benefit streams.
Regarding DRA's third allegation that Lechner's benefits differential is inflated because he used "five fewer months" to calculate the benefits for the five month suspension scenario than for the non-suspension scenario, PG&E states that DRA misses the point of the comparative analysis. It is the timing of endpoint deployment that drives the magnitude of realized benefits, and suspending deployment of endpoints would delay the realization of benefits that would be obtained under a non-suspension scenario. PG&E states that Lechner's analysis properly modeled the stream of benefits associated with PG&E's endpoint deployment plan without a suspension scenario and compared this to the stream of benefits that would result from suspended deployment plans, and comparing the present value of these various benefits streams provides a clear quantification of the impact of suspension benefits realization.
It is TURN's position that PG&E could avoid this increased cost if it simply waits to deploy its solid state meters until (a) its preferred HAN technology is commercially available or (b) a final Commission decision on this application. TURN states that PG&E has chosen to prematurely move ahead with a large number of solid state meters by the end of 2008, even though PG&E intends to scrap or retrofit all of the meters later, requiring at a minimum, a duplicative expensive field visit from a PG&E employee or contractor, and argues that the ratepayers should not be saddled with the cost of PG&E's unreasonable management strategies.
TURN states PG&E' suspension analyses are flawed for many reasons and should be disregarded. First, the analysis was not completed before this application was filed in December 2007, so TURN states it could not have been used to justify the project management decisions. Second, TURN asserts analytical flaws render the analysis useless. According to TURN, a correct analysis would have taken all recorded costs and benefits up to July 2008 and then analyzed a delay (recognizing all recorded costs and benefits) compared to an updated forecast of remaining costs and benefits, something PG&E did not do. In addition, TURN criticizes PG&E's assumption that all meters are activated and providing O&M and demand response benefits in the same month they are installed. TURN notes that PG&E currently has over 534,000 gas meters installed but only 67,000 activated, and there are no demand response benefits currently and PG&E has been installing meters for at least a year and a half.
CCSF states that it agrees with TURN's reasoning for rejecting the HAN retrofit and TURN's position that ratepayers should not have to pay the additional $34.8 million (with risk allowance) requested by PG&E.
In response to TURN, PG&E states that TURN's suggestion that Lechner's analysis should be rejected because it was performed after PG&E's initial Upgrade filing, ignores the record, noting that PG&E witnesses Corey and Meadows both testified that PG&E had considered the potential costs and benefits of delaying deployment while PG&E evaluated the emerging technology. When PG&E submitted its Application in December 2007, it was in the middle of negotiations with its Upgrade vendors and was continuing to refine its specific technology selections and deployment alternatives. According to PG&E, this was an appropriate time to analyze the detailed implications of various deployment scenarios, including potential suspension of endpoint deployment depending on the availability of PG&E's preferred HAN device identified as a result of the ongoing vendor bidding. PG&E further states that its May 2008 update to its Upgrade Application included the results of its ongoing vendor negotiations and that, had the results of Lechner's analysis been different and concluded a suspension scenario was indeed preferable to continuing deployment, PG&E would have included such a result in its May update.
PG&E states TURN's suggestion that "[a] correct analysis would have taken all recorded costs and benefits up to July 2008 and then analyzed a delay ... compared to an updated forecast of remaining costs and benefits" ignores the fact that the costs incurred prior to the starting point of a comparative analysis (and recorded benefits) have no impact on the result of the comparative analysis because they are exactly the same for all scenarios being compared.
With respect to TURN's argument that Lechner's assumption regarding the timing of benefits relative to endpoint installation is wrong, PG&E states that the identification of benefits with endpoints in the month they are installed was a simplifying assumption applied to each scenario. While this does not calculate the precise timing of benefits realization, it is an appropriate approach to compare the benefit stream of a continued deployment scenario with various suspension scenarios, provided the assumption is consistent among the scenarios, which according to PG&E, it was.
PG&E's suspension analysis of the HAN Upgrade appears reasonable. Its consultant compared lost benefits due to suspension to reduction in project costs resulting from the suspension relative to the base case. Relative to the base case, the only cost that would not be reduced due to a suspension is the cost of the HAN gateway device. All of the other costs are associated with the retrofit of the meter. PG&E's consultant added the cost of the HAN device to the suspension costs25 to quantify the total costs that should be subtracted from the reduced costs due to the suspension before being compared, on a PVRR basis, to the lost benefits due to the suspension. In all three suspension scenarios (three, four and five-month suspensions), the analyses showed the lost benefits exceeding the net reduced costs.
We have evaluated the criticisms made by TURN and DRA with respect to PG&E's consultant's suspension analyses along with PG&E's responses. In general, we find that PG&E has adequately explained and defended the analyses, and we are comfortable in using the analyses as a basis for determining the reasonableness of PG&E's actions.
In particular, we agree that the estimated 20-year life for endpoints is not relevant for purposes of analyzing the economic impact of a deployment scenario. If deployment is suspended for five months, benefits for those five months are lost. At any point in time beyond 2011, when the base and suspension scenario are compared, the five-month suspension scenario will have five months fewer benefits. That is simply because the benefits go on indefinitely and do not end when the meter has been in place for 20 years and is retired and replaced or is refurbished.26 We also agree that the costs incurred prior to the starting point of a comparative analysis (and recorded benefits) have no impact on the result of the comparative analysis because they would be the same for all scenarios being compared.
Lechner's conclusion that the cost of the 5 million meters had no impact on the overall results in his analysis was based on his examination of his model outputs and appears reasonable. DRA had access to Lechner's model and has not indicated that the outputs that Lechner relied on are erroneous in any way.
Also, PG&E has provided sufficient explanation as to why its consultant's suspension analysis was performed after the filing of the application. What is important is that it was performed before this aspect of meter deployment began, and was thus available for PG&E's project management to use in determining whether or not to go forward.
While PG&E's decision to proceed with the HAN retrofit appears to be reasonable, the magnitude of the retrofit cost estimate ($32,026,000 plus a 10% risk based allowance) has not been fully supported and justified. There is little support for PG&E's quantification of the number of meters that would necessarily be installed without a HAN device. Also, the record does not include detail and substantiation of all of the various cost components of the retrofit. For instance, while the costs include that necessary to physically retrofit a meter with a HAN device, there is no detail as to what that particular cost is, what it was based on, and why it is reasonable. Also, it is not clear whether the communication module that is replaced has any salvage value and if so whether that was factored into the costs. To account for uncertainties and attempt to ensure that ratepayers only fund appropriate costs, we will reduce adopted funding for the HAN retrofit by $5,500,000 (plus $550,000 for the related risk based allowance).
7.3. Electromechanical Meter Retrofit
At the time of the application filing, PG&E had already procured 230,000 electromechanical meters intended for its Kern County region. Approximately 123,000 of these meters had already been installed and the rest were to be installed by mid-2008. Considering the availability of the improved meter devices and the continued ability to achieve the benefits of SmartMeter Program deployment, PG&E believed it would be reasonable to make the transition from electromechanical meters to solid state meters as early as practicable to minimize the potential retrofit of installed electromechanical meters with upgraded meter devices pending the Commission's approval of PG&E's request in this application. PG&E decide the time to make the transition was after completing deployment of the Kern region.
Once all customers have received an advanced meter (i.e., in 2011), PG&E proposes to upgrade the estimated 230,000 electromechanical meters with the new solid state meters so that all of PG&E's electric customers can participate in the new service offerings and increased functionality available with the upgraded meters. PG&E estimates that it will require approximately six months to upgrade these electromechanical meters installed prior to the SmartMeter Program Upgrade. PG&E has forecast $37,312,000 in costs relating to the retrofit of meters deployed in the Kern region. These costs would provide labor and material sufficient to replace the 230,000 meters deployed in the Kern region without a HAN device or load limiting switch, with a complete advanced solid state meter, integrated load limiting switch and a HAN device.
DRA states that it is supportive of the enhanced functionality associated with the HAN and the integrated service switch, as well as the advanced Tier 1 solid-state meter required for both these functions. Thus, DRA includes these costs in its business case even for the electromechanical meter retrofit. It also includes the labor costs for the Kern retrofit because a second visit to these meters would have been required anyway to install this new functionality. Unlike PG&E, DRA adds that it did not include the cost of new communications modules and network costs for the Kern retrofit, because it believes that the choice of the DCSI system was questionable to begin with.
DRA's argument for disallowing most of the Kern County retrofit costs is not based on the idea that the Kern County deployment could have been delayed, it is based rather on DRA's belief that PG&E came to the Commission prematurely with its original application, A.05-06-028, in the first place. DRA states that its support for that application must be qualified, in that such support was based on representations that PG&E made that have turned out to be wrong. Transcript evidence shows that DRA witness Abbott had expressed concerns to PG&E at a meeting in December 2005 about whether the DCSI system would have sufficient bandwidth to handle the signals in an urban area with high density. He was assured by PG&E that it had developed workarounds to this problem. Therefore, he gave PG&E the benefit of the doubt, and in his testimony in A.05-06-028, stated that PG&E's technology choice is "generally reasonable." According to DRA, representations also had been made by PG&E about the ability of the DCSI technology to support the HAN technology, and these did not pan out either. It is because of PG&E's decision to "jump the gun" that DRA does not even include the cost of the base meter in the Kern retrofit.
TURN recommends that the Commission disallow all the costs related to the electromechanical meters in the Kern region by (1) disallowing the $41.03 million requested in this application27 to retrofit the installed electromechanical meters; and (2) removing $23.2 million from PG&E's original AMI budget, thus making it less possible for PG&E to indirectly recover some of these costs through contingency allowances. TURN recommends the removal of the meter costs from the original AMI budget, because they were stranded by poor management decisions regardless of the outcome of this Upgrade application.
TURN states that despite the fact that PG&E filed a request for authorization of over a half a billion dollars to "upgrade" its AMI project, it persisted in installing meters in the Kern region that it knew it would strand in only four years. While PG&E claims that it did not finally decide it would change its AMI technology until the date that it filed this application in December of 2007,28 TURN argues that PG&E indicated that it began the process of evaluating solid state meters, integrated load limiting disconnect switches, and the availability of home area network technologies in early 2007,29 and, by May 2007, PG&E indicated that it was interested enough in the new technologies to adjust its meter procurement plan and tell its electromechanical meter supplier that it intended on terminating the contract for buying electromechanical meters.30 According to TURN, PG&E was not forced to strand this investment. It proactively chose to do so and did so while requesting additional funds to fully deploy an entirely different technology. In TURN's opinion, PG&E's decision to continue installation of electromechanical meters in the Kern region was unreasonable and imprudent, and the Commission should not insulate PG&E from the consequences of its decisions.
In response, PG&E expressed its understanding that DRA would allow about $18.8 million of the requested costs, by adding $6.3 million in labor costs to about $12.5 million for the incremental costs of an advanced solid state meter, the integrated load limiting switch and the HAN device.31 DRA would not allow funding for the "base" cost of the meter itself or the communications module that would need to be replaced. PG&E further understands that TURN estimates the installation costs of the Kern deployment for its proposed disallowance of $23.2 million.
It appears to PG&E that, despite the proposed disallowances, both DRA and TURN want the retrofit to be performed. According to PG&E, what intervenors debate -- and the issue on which their proposed disallowances depends --is whether PG&E should have installed (in the first instance) the DCSI power line carrier (PLC) equipment on electromechanical meters in Kern. For the three reasons described below, PG&E asserts that it was right to do so.
First, PG&E indicates that its deployment of the electromechanical meters in Kern followed the directives of D.06-07-027 to the letter and was strongly supported by DRA in that case. PG&E points out that (1) the meters deployed include the technologies approved in D.06-07-027, (2) no party has alleged that PG&E has somehow strayed from the letter or intent of D.06-07-027 in deploying these meters, and (3) the deployment has been successful and the meters are working as intended, generating operational benefits as meters are activated.
Second, PG&E states that the argument that PG&E should have delayed installing the Kern meters, as an alternative to incurring the proposed retrofit costs has no merit, because it ignores the evidence in the record that continued deployment was beneficial for ratepayers. PG&E explains that when it became apparent the Upgrade technology might be becoming commercially feasible, PG&E considered a short-term suspension of electric meter deployment, but determined this would not be in the best interest of its customers. PG&E concluded that delaying implementation would serve to increase overall costs as vendor commitments had already been made and a suspension would result in further delays to the benefits.
Third, regarding DRA and TURN suggestions that if PG&E had installed solid state meters in Kern, then a retrofit to accommodate the HAN device would not be necessary,32 PG&E states that a retrofit would still be necessary and the original deployment would have been more costly. This is because the use of electromechanical meters in the original deployment plan resulted in approximately $36 million in cost savings when compared to using basic solid state meters. PG&E also states that these basic solid state meters that were available for deployment at the time of the original AMI case would not support a HAN device and thus would need to be replaced now anyway, a point that DRA conceded during hearings.33
Electromechanical meters have been deployed in the Kern region, and, as a result of PG&E's Upgrade request, the electromechanical meter costs will become stranded once these meters have been replaced. We see the fundamental issue to be whether these stranded costs should be addressed as part of the costs of the original AMI program or as part of the costs of the Upgrade. As discussed further in this decision,34 we determine that the stranded costs related to the electromechanical meters should be considered as original AMI program costs, specifically under the risk based allowance for the original AMI project. Therefore, for purposes of this proceeding, we need not determine whether PG&E should or should not have deployed electromechanical meters in the Kern region, or whether PG&E came prematurely to the Commission with its original AMI application.
Our result is similar to that of DRA in that we include costs for the upgraded system, but exclude costs related to the original meter and communications device. Based on PG&E's representation of DRA's recommended cost for the electromechanical meter retrofit, we will adopt, as reasonable, an amount of $18.8 million for that purpose.
Because of the manner in which this issue is resolved, it would not be appropriate to remove $23.2 million from the original AMI budget as proposed by TURN. It appears that amount represents the stranded costs that should be absorbed through the risk based allowance or contingency.
7.4. HAN Connectivity
PG&E states that one challenge in effectively deploying HAN technologies is the variety in configuration of customers' premises. In some residences, the signal from the HAN device may need to travel long distances because of a meter located away from the home. Even for homes with attached meters, it is possible that appliances and devices such as thermostats, pool pumps or water heaters may be placed in locations that are difficult for the signal to reach. For example, water heaters may be located in basements or garages and pool pumps could be in external structures.
According to PG&E, currently, there are two predominate HAN gateway technologies in the marketplace, PLC technology and RF technology. Each of these technologies has strengths and weaknesses in dealing with the challenges created by the diversity of structure types and distances. For example, PLC technology is better at traveling long distances and has the ability to communicate with some devices that are not plugged into an electrical outlet such as a thermostat, while RF technology is better able to reach devices that may not be able to receive PLC communications.
To compensate for the variations in functionality of different HAN gateway technologies and to take advantage of the best available solutions, PG&E proposes a combined RF and PLC solution. This combination of approaches will serve more types of homes than one approach or the other. PG&E would likely deploy a PLC-based solution to customers living in multi-dwelling units. This is because the HAN signal travels into the home through the electric wiring instead of via radio signal that can frequently be blocked or attenuated. Therefore, for the HAN gateway, PG&E proposes to use a combination of Homeplug (PLC) and Zigbee (RF) devices - whereby the PLC solution would be used to enhance reliable connectivity for large, multi-storied and multi-unit dwellings, and the RF solution would likely be deployed to other types of residential electric customers.
Based on ongoing research and discussions with DRA, PG&E believes that it is prudent to deliver a standardized and common RF based HAN signal into all customers' premises.35 According to PG&E, this means that for the approximately 40% of premises that were expected to receive a Homeplug device, all of those premises will require some type of bridging or augmentation device to bring an effective signal from the meter location to an interior wall of the customer's premises.
However, at the present time, there are still a number of uncertainties regarding the best approach to extend the connectivity of the HAN devices at the meter to an interior wall of a customer's premise. PG&E states that, although much work in the industry and in standards development is occurring, there is not yet a standard approach to reliably deliver HAN connectivity on a universal basis, including translation or bridging devices. PG&E and the others in the industry are currently evaluating several approaches to address this challenge. Therefore, while it is premature to settle on a specific solution and lock in to a defined approach for an extended period of time, PG&E believes its recommendation is appropriate given the stated goals related to the home area network and reflects a thoughtful consideration of the known technical challenges of each HAN technology and the state and direction of the HAN standards and industry.
PG&E has developed its estimate of costs to extend HAN functionality from the electric meter location to an interior wall of a customer's premises using the following assumptions:
(a) 40% of customers' premises with installed Smart Meters will require a bridging, translation or another augmentation device to bring RF connectivity to an interior wall of the customer's premises.
(b) During the period covered by the revenue requirement request in this case, 15% of the above-described customers' premises will require a bridging, translation or another augmentation device to bring RF connectivity to an interior wall of the customer's premises during the project period, considering customers' demand for HAN functionality. The remaining customers would obtain the bridging, translation or another augmentation device in later years.
(c) PG&E set an allowance of $50 for each bridging, translation or other augmentation device for either the provision of such a device or to provide a rebate to customers seeking to install their own devices.
By doing the above, PG&E states that it would deploy a solution that would bring the highest probability of transmitting a signal from the electric meter to an interior wall of the customer's premises. However PG&E cautions that no utility can guarantee that the HAN signal would be available throughout all areas of the customer's premises or property. Under PG&E's proposal, additional signal enhancements within a customer's premises to extend the connectivity of the HAN device from an interior wall to other locations within the premises would be the responsibility of the customer or the provider of the HAN enabled device with which the customer desires to establish a connection.
For HAN connectivity, PG&E seeks $16,891,000 in incremental costs. In total, the HAN connectivity related PVRR amounts to $59,123,000 under PG&E's proposal.
DRA recommends Homeplug deployment be set at 30% rather than the 40% requested by PG&E. According to DRA, while PG&E's Homeplug estimate is based on the "nature of dwelling types in its service area," that is, the ratio of single family homes to multiple family homes, it does not take into account that some multiple family homes are duplexes that are not much larger than a single family home. DRA states that PG&E has provided no data on the typical broadcast footprint (in feet of dispersion) of the Zigbee interface, and PG&E has adopted the most conservative assumption possible, that is, that all multiple family homes will require a HomePlug interface. DRA likens this to asking for an extra cushion on top of its normal risk allowance.
In response, PG&E states that the net effect of DRA's recommendation would be to reduce PG&E's costs by approximately $4 million36 and argues that DRA's recommendation is not supported by any analysis or documentation and is made solely as a way of reducing project costs. PG&E cites the following from the evidentiary hearing transcript:37
PG&E Counsel: What does DRA want to do?
DRA Witness Levesque: 70/30. [Meaning 70% ZigBee and 30% HomePlug.]
Q: Did you do any analysis of PG&E's system to come up with that percentage?
A: The foundation for that change was in one sentence of what if it were 70/30. And the reliance upon would 70/30 make sense was based entirely on subjective opinion of number of households, number of apartments and small apartment buildings the size of a population of the City of San Francisco. And that was in a communication with DRA that gave me that information.
I have no supporting, specific documentation for the 70/30. And I don't know if there is empirical evidence in the marketplace today as to whether HAN will produce 62/38 or 70/30.
Q: When you say what-if scenario, was that an effort to get the price down?
A: It was an effort to understand the magnitude of what a change would be of -- if HAN were 10% more effective, what that might do for pricing.
Q: The effect of raising the percentage of [ZigBee] effectively reduces the amount of money PG&E gets; right?
A: That is correct.
Accordingly, PG&E asserts that DRA's recommendation has no proper evidentiary basis, PG&E's proposal for a 60/40 split in the deployment of ZigBee and HomePlug devices is the only proposal on record with a proper evidentiary basis, and PG&E's proposal is the most appropriate for promoting HAN receptivity for customers.
TURN argues that the request should be rejected, because extended HAN connectivity costs are directly related to PCTs associated with PG&E's Title 24 program, and PCTs will not be incorporated into the next round of Title 24 building standards. PG&E will not be recruiting customers until 2013, outside the forecast period for this application. Therefore, TURN asserts that HAN connectivity costs should also be excluded from the program.
TURN also argues that HAN bridging device technology is not well known at this time, and is in the infant stage of development. According to TURN, the Commission should therefore not authorize this request and expose ratepayers to further risk of stranded technology and costs. TURN also questions the efficacy of this type of investment given that customers in multi-family dwellings are the least likely customers to be able to take advantage of HAN to alter energy usage since they rarely have the ability to install HAN-enabled appliances. Furthermore, because these customers generally have a lower energy usage than residential customers that live in single-family dwellings, TURN asserts they have less energy to conserve, reduce, or shift and are therefore poor candidates for providing demand response.
In response to TURN, PG&E states that regardless of whether a landlord or tenant owns an appliance, the person who pays the energy bill - typically the tenant - has the incentive to reduce his or her energy costs through the information available from the HAN repeater device. According to PG&E, studies have shown that tenants may have even more to gain from the information available from the HAN. This is because such tenants are deprived of the ability to control their energy use through hardware choices and their best means of control is through their use patterns and the information available through the HAN.
First of all, we are in agreement with PG&E's general direction in attempting to deploy a solution that would bring the highest probability of transmitting a signal from the electric meter to an interior wall of the customer's premises. To do this, it is reasonable to use both RF and PLC technologies as proposed by PG&E.
With regard to whether the HomePlug or PLC technology should be applied to 30% of the residences as proposed by DRA or 40% as proposed by PG&E, we will adopt PG&E's 40% proposal. The basis for DRA's proposal stems from a hypothetical analysis involving cost sensitivity based on a 30% assumption. There is no evidence as to the reasonableness of using 30% to reflect what might actually occur.
With respect to TURN's argument that HAN connectivity costs should be excluded because PG&E will not be recruiting Title 24 PCT customers until 2013, we decline to do so, because HAN connectivity relates to not only PCTs but also to other devices such as in home displays. In PG&E's supplemental testimony, the proposal for HAN connectivity was expanded to all customers, not just to Title 24 PCT customers.38
Regarding TURN's argument that customers in multi-family dwellings are the least likely customers to be able to take advantage of HAN to alter energy usage and PG&E's response, the determination of who will use the HAN technology, and to what extent they will use it, is fairly subjective at this point. From a policy perspective, we feel it is important that customers that wish to use the technology are, to the most reasonable extent possible, able to do so.
We are however somewhat hesitant to authorize additional funds to provide a single or common RF based protocol once the signal is made available within the customer's premises. As PG&E itself acknowledges there is not yet a standard approach to reliably deliver HAN connectivity on a universal basis, including translation or bridging devices. TURN argues ratepayers should not be exposed to the risk of stranded technology and costs, and PG&E's request regarding HAN connectivity should be rejected. On the other hand, we believe HAN connectivity on a universal basis makes sense for such purposes as advancing and developing the HAN technology in an efficient manner. With the expectation that it may be necessary in some form, we will authorize PG&E's HAN connectivity request. We expect PG&E to adapt the implementation of HAN connectivity over time consistent with approaches and solutions that are being addressed and developed, currently and in the future, by those in the industry that are addressing these issues. It is PG&E's responsibility to achieve HAN connectivity in the most cost effective manner within the costs and risk based allowances provided by this decision. PG&E should understand that we will be extremely reluctant to saddle ratepayers with stranded assets and costs associated with any cost overruns related to HAN connectivity.
7.5. Information Technology
PG&E estimates that it will incur incremental information technology (IT) costs resulting from the additional scope functionality of the SmartMeter Program Upgrade. These include IT costs to support the PTR Program, HAN functionality, the AC Program, the Load Limiting Functionality and IT project management. Briefly,
· In order to accommodate its proposed PTR program, PG&E states that it will be necessary to modify its Customer Care and Billing (CC&B) and Customer Service On-Line (CSOL) systems. To estimate the cost of these efforts, PG&E used its standard four-phase IT model: pre-build, develop, test, and support. The estimated labor cost of this incremental scope increase is $4 million, which is based on PG&E's average, daily, internal and external labor rate of $1,200. PG&E expects to incur these PTR-related costs from mid-2008 to mid-2009.
· To support the HAN functionality, PG&E proposes to establish reliable and secure two-way communication between PG&E's network management systems and the HAN gateway devices. It will also confirm the ability to address an Internet Protocol (IP) addressable device behind the meter and receive a response. PG&E anticipates it will perform the HAN infrastructure and integration work in 2009 at an estimated cost of $23.1 million, which includes $4.6 million of non-labor costs and $18.5 million of labor costs.
· Starting in 2013, PG&E proposes to use HAN capability to provide AC Program functionality for Title 24 compliant programmable communicating thermostats (PCT) as part of the SmartMeter Program Upgrade, in order to enhance and expand PG&E's current SmartAC Program. PG&E states that operating the AC Program on the HAN network (likely in parallel to the current vendor-provided SmartAC Program) for all Title 24 PCTs requires PG&E to: (1) provide in-house services similar to those currently performed by vendors for the SmartAC Program (i.e., program enrollment, deployment, customer service, and load/event management); (2) utilize the two-way AMI network/HAN; and (3) integrate a PG&E-hosted load management system with the AMI infrastructure. To estimate the costs of using the HAN network to communicate with new PCTs, PG&E reviewed the program's current business and technical requirements and estimated the software and labor resource needs required to build the system internally. PG&E anticipates it will incur these incremental AC Program costs in 2011. PG&E estimates the incremental cost of the upgrade to be $14.8 million, which includes $2 million of software costs and $12.8 million of labor costs.
· PG&E estimates it will incur additional costs to integrate the load limiting connect/disconnect switches for all its single phase residential meters with a maximum of 200 amps. Modifications and interface changes will be required to create new credit/collection templates, start/stop algorithms, and partial Load Limiting Functionality. To estimate the cost of these efforts, PG&E used its standard four-phase IT model: pre-build, develop, test, and support. The estimated labor cost of this incremental scope increase is $3.7 million, which is based on PG&E's average, daily, internal and external labor rate of $1,200. PG&E expects to incur these costs from mid-2008 to mid-2009.
· PG&E states the Upgrade will require additional IT project management efforts to support the additional IT work discussed above. PG&E anticipates it will need three additional FTEs from mid-2008 to mid-2011 at an estimated total cost of $2.8 million.
As discussed further on in this decision, DRA opposes consideration of the PTR program as part of the Upgrade, because DRA feels the PTR program could be implemented in conjunction with PG&E's originally authorized AMI system. For this reason, DRA excludes all PTR benefits and the majority of PTR related costs including $4 million (PVRR) in IT costs associated with the PTR program. DRA states that, if the PTR program is funded in another proceeding, the associated IT cost could be considered there.
DRA also notes that an unnecessary duplication of IT costs has occurred because of PG&E's choice to implement a communication system as part of its SmartAC program that is duplicative of the HAN communication system. However, because DRA is supportive of the HAN technology, it did not exclude the IT costs associated with HAN communication.
With respect to DRA's exclusion of $4.0 million in PTR related IT costs, PG&E states that the adjustment is a corollary to DRA's position that benefits for the PTR program should also be excluded from the cost/benefit analysis for the Upgrade, and accordingly, if the benefits of the PTR program are included - as PG&E believes they should be - the IT costs for the PTR program should be included as well.
Similar to DRA, TURN asserts that PG&E's original AMI technology was capable of implementing PTR on a wide scale, and reduces both costs and benefits as they relate to the Upgrade. This includes exclusion of the $4.0 million in IT costs for the PTR program.
TURN also excludes $14.8 million in IT costs requested by PG&E in conjunction with the proposed use of the HAN functionality to communicate with Title 24 building standard compliant PCTs. TURN states that PG&E itself has withdrawn other costs associated with the Title 24 PCT program. Specifically, PG&E assumed in the application that the CEC's proposed Title 24 building standards would begin in 2009, but the CEC later postponed its recommendation. As indicated in its supplemental testimony, PG&E now assumes the standard will be implemented in 2012 and that PG&E will begin recruiting customers in 2013. TURN states that PG&E reduced its Title 24 PCT program cost request by $5.0 million39 because 2013, the year PG&E begins the program, is outside of the forecast period for this application and argues that the Commission should similarly reduce PG&E's request for the related IT costs.
TURN states that PG&E's current Smart AC Program is the result of a settlement with PG&E, DRA, and TURN that was adopted by the Commission in D.08-02-009. That settlement provided PG&E with sufficient funds to implement a 305 MW direct load control program by 2011. The settlement directs PG&E to come back to the Commission in the second quarter of 2009 with an additional application to extend the program to 2020 - after PG&E has completed and reported certain measurement and evaluation studies required in that settlement. According to TURN, any funds used to supplement the program or change recommendations to that program are supposed to be contained in the application PG&E is directed to file with the Commission in the second quarter of 2009. TURN states that the Commission should require that PG&E honor its end of the TURN/DRA/PG&E settlement and reject any costs for the Smart AC program that conflict with that settlement.
Finally, TURN asserts that PG&E requests ratepayer funds to duplicate processes that it readily admits are already being provided by its vendors. As stated in its application, PG&E wants to "provide in-house services similar to those currently performed by vendors for the Smart AC Program" and operate the program "in parallel to the current vendor provided" program. According to TURN, this is operating a redundant program and a wasteful use of ratepayer funds.
In response, with respect to TURN's Title 24 PCT related adjustment, PG&E states that TURN's primary argument, that since PG&E has delayed incurring approximately $5 million in administration and marketing costs associated with the Title 24 PCT program until 2013 or later -- due to the delay in the expected date of the new regulations from the CEC -- so too the IT costs should be removed, has no merit. According to PG&E, the administration and marketing costs associated with the A/C program are distinct from the IT costs. They are for different purposes and are to be expended at different times. PG&E states that under its proposal, the IT work for the A/C program would be performed in 2011, which is still prudent due to the fact that the CEC Title 24 regulations are now expected to be implemented in 2012.
Regarding TURN's other arguments on this issue, PG&E states that first, there is no conflict with the SmartAC settlement, in that, at the time of the settlement, PG&E had notified parties of the possibility that it might file an upgrade to its SmartMeter Program and the settlement expressly envisioned this fact. On this point, the Commission explained,
[T] he settlement requires PG&E to analyze how to fully integrate the AC Program with its AMI. Integrating the AC Program with AMI will likely increase the value of both programs and expand opportunities for customers to engage in demand response. Therefore, 90 days after the Commission acts on PG&E's pending AMI application (A.07-12-009), PG&E should provide a report to Energy Division, DRA and TURN explaining how PG&E intends to integrate the AC Program with AMI.40
PG&E argues it is disingenuous for TURN to suggest that there is conflict with the settlement when the settlement itself expressly envisioned that the AC program could be integrated with the Upgrade. PG&E adds that integration of the AC Program with AMI is what this IT expenditure is designed to do and the costs are neither redundant nor wasteful.
CCSF states that PG&E may well have underestimated the true cost of the Upgrade. While the hardware to be installed is the most visible element of PG&E's upgrade, it is common practice in joint development efforts of this kind that hardware engineering often leads software engineering. According to CCSF, many of PG&E's chosen hardware components reflect relatively early stage technology, and some of these components do not yet have software necessary to drive them, or to coordinate their individual functions into the larger web of grid and data management systems. To CCSF, this absence of the necessary software suggests that there will likely be significant systems integration challenges, the complexity and cost of which PG&E may well have underestimated. CCSF is concerned, therefore, that PG&E will at a later date seek to recover even more than the nearly $3 billion the Commission will have approved if this upgrade is authorized.
In response, PG&E states that CCSF makes no acknowledgement of the substantial amount of testimony that PG&E has submitted in the area of IT, which addresses not only the IT hardware, but also the software and system integration needs associated with the Upgrade. PG&E states that it understands and has already articulated the types of risks that CCSF purports to have discovered.
As discussed further in this decision, we have included the benefits of the PTR program in evaluating the cost effectiveness of the Upgrade.41 For that reason, it is also appropriate to include the $4.0 million in IT costs related to the PTR program in rates, as requested by PG&E.
Regarding TURN's proposed adjustment for Title 24 PCT related IT costs, PG&E's argument -- that assigning the costs to 2011 is still reasonable because the CEC Title 24 regulations are now expected to be implemented in 2012 -- is not persuasive. In its application filing, PG&E proposed to spend $6,728,000 in 2010 and $8,105,000 in 2011.42 Also, it expected to begin recruiting AC customers starting in 2011 and estimated the number of customers for that year to be 16,000 with increasing amounts thereafter (e.g., 47,000 new customers in 2012).43 In its supplemental testimony, PG&E indicates that it now expects to begin recruiting AC customers in 2013 and estimates the number of customers for that year to be 18,000 with increasing amounts thereafter (e.g., 52,000 new customers in 2014).44
PG&E has provided no specific reasons to justify why the IT related costs need to be incurred prior to or in 2011 and why they cannot be shifted commensurate with when the expected recruitment of Title 24 PCT customers is expected to begin. Without such justification, we conclude it is reasonable to shift the costs. We will do so by shifting these costs to 2013 and 2014, principally to remove such cost recovery from this decision. There is significant uncertainty as to when this program will begin,45 and we prefer not to authorize related costs at this time. The Title 24 PCT program costs have already been moved by PG&E to 2013, outside the timeframe for cost recovery authorized by this decision. Those costs will have to be recovered in a separate proceeding. PG&E should seek recovery of the related IT costs at the same time.
We do agree with PG&E regarding TURN's allegations of conflicts with the SmartAC program. It is clear that, in D.08-02-009, the Commission expected the SmartAC program would be integrated with the Upgrade. Also, in that decision, the Commission welcomed PG&E's commitment to incorporate Title 24-compliant PCTs into its project and expressed a concern regarding the settlement's 40% limitation on PCT installations.46 Further in this decision, we address issues related to the inclusion of the Title 24 PCT program in determining costs and benefits associated with the Upgrade.
Finally, we understand CCSF's concerns regarding what may be significant systems integration challenges. However, while nothing is certain, we feel that PG&E's IT proposal is a reasonable means for overcoming any related problems. This is consistent with our authorization of the same advanced metering technologies, with the same integration challenges, for SDG&E and SCE.
7.6. Title 24 PCT Program Costs
PG&E explains that customers with Title 24 compliant PCTs will need to be identified and recruited for participation in the SmartAC Program and there are costs associated with that activity.47 In addition, the initiative will be reaching out to customers with existing air-conditioning systems for an early change out of the thermostat with a Title 24 compliant PCT. Administrative costs and minor other costs for software and call center support are also included in incremental costs for the program.
Some of the outreach activities considered by PG&E include using new customer connect records for identification of likely new construction sites and purchasing permit records to target market to permitted retrofits. Customer acquisition costs of $53 per participant and $25 sign-up incentives are based on the current SmartAC Program estimates.
Due to PG&E's revised assumed timing of the Title 24 PCT program from 2009 to 2012, costs will occur outside of the time period that PG&E is requesting the related rates as part of this application. For costs through 2030, PG&E estimates costs with a PVRR of $37,906,000.
DRA and TURN have not forecasted the PVRR of any Title 24 PCT program costs, not because of any differences in what the estimated costs should be, but because of their positions that neither Title 24 PCT program costs nor benefits should be included in the cost effectiveness analysis of the Upgrade. As discussed elsewhere in this decision, we have included the benefits of the Title 24 PCT program in evaluating the cost effectiveness of the Upgrade. For that reason, it is also appropriate to include an estimate of the costs through 2030 on a PVRR basis for use in the cost effectiveness analysis. However, consistent with our adjustments for reduced participation to the expected benefits of the program, as discussed in Section 10.4.3 of this decision, we reduce the costs by related marketing and incentive amounts. We adopt Title 24 PCT program costs of $26,174,000 on a PVRR basis, as opposed to PG&E's estimate of $37,906,000.
7.7. Peak Time Rebate Program Costs
The PTR program48 does not require customers to enroll, however awareness of a critical peak event (the day and time period that PTR as well as CPP will be in effect) is critical to achieve both customer bill rebates and DR resources. PG&E estimates that approximately 50% of residential customers will need to be aware of critical peak events in order to achieve anticipated PTR benefits. According to PG&E, awareness is not an indication of a committed effort. Instead, it provides a proxy for "participation" in the determination of average benefits. PG&E has developed a general strategy for an estimated $7.5 million annual marketing campaign to achieve an average of 50% residential awareness rate of an event without any enabling technology. The media strategy calls for two phases to achieve the objective:
1. Education phase: This includes a pre-summer media and PR effort to raise general awareness of the program; and
2. Event phase: Media and PR during events focused on immediately notifying customers an event is in effect.
The day of the event activities will include newspaper, spot radio, TV and geo-targeted online efforts. The level of media available is constrained by the fact that events are not known more than 24 hours in advance.
PG&E will begin the PTR program in 2010 and will not have the SmartMeter Program Upgrade technology and features, including interval billing, fully deployed in the PG&E service territory that year. As a result, the marketing campaign will be limited geographically in 2010 and is estimated to cost $3.4 million. Years 2011 and 2012 are estimated at the full $7.5 million annual cost for the two-phase education strategy. Years 2013-2030 have a lower annual estimated cost of $1.8 million due to the assumption of a transition to a more direct method of event notification through in-home displays and enabling DR technologies the customer will choose to install.
DRA and TURN recommend no PTR program costs, not because of any differences in what the estimated costs should be, but because of their positions that neither PTR program costs nor benefits should be included in the cost effectiveness analysis of the Upgrade. As discussed further in this decision, we have included the benefits of the PTR program in evaluating the cost effectiveness of the Upgrade. For that reason, it would also be appropriate to include the $18.3 million in PTR program costs, in rates, as requested by PG&E. However, since this decision approves a two-tier PTR incentive structure that will be detailed by PG&E in a November 2009 rate design window filing,49 it would be more appropriate to address the costs of such a program at the same time, and we will order PG&E to do so.
While PG&E's current PTR program cost estimate of $18,342,000 is for a single tier PTR incentive structure, we will use the related PVRR of the PTR program costs, which amount to $27,592,000, for the purpose of evaluating the cost effectiveness of the Upgrade.
7.8. Project Management Costs
PG&E has forecast $15.3 million in additional project management costs associated with the Upgrade. According to PG&E, these costs are associated with additional project management efforts that will be required as the industry continues to evolve and offer new technologies. PG&E specifically cites additional project management efforts that will be required to deal with the added technological complexity of the HAN, ubiquitous load limiting switch and the advanced solid state meters and to manage additional vendors and the associated issues in contract administration and management of warranties, supply chain issues, costs and benefits realization, and performance metrics.
DRA excluded incremental project management costs completely from its business case, because it believes that what PG&E received in the original case was sufficient. DRA explains that while PG&E asserts that there is additional complexity associated with managing multiple technologies, in its original case, PG&E argued for the need for multiple technologies, one for gas and one for electric, and included the cost to manage the deployment of and operation of these multiple technologies. Since the Upgrade proposes to eliminate the PLC technology, deploying only the Aclara RF technology, and PG&E anticipates introducing a second technology, Silver Springs, DRA asserts that PG&E would still be managing only two technologies as proposed in its original case.
TURN argues that PG&E has not adequately justified its request to increase its project management costs, and the Commission should reject PG&E's request. According to TURN, while PG&E states that the additional funds are supposed to pay for in-house labor costs associated with the increased costs of dealing with more vendors resulting from this "AMI Upgrade" and external professional services to help with in-house project management, risk assessment, and evaluation of PG&E's program management process, with the exception of retrofitting meters with yet unavailable HAN devices and re-deploying solid-state meters to replace stranded electromechanical meters, in general, PG&E is installing the same number of gas and electric meters that were authorized in A.06-07-027. TURN further states that PG&E may have a handful of additional vendors to administer but PG&E has not met its required burden of proof demonstrating that there is a linear function between administering a few more vendors and its proposed increase to program management costs. TURN adds that the rate at which PG&E has been spending its project management and risk allowance funds without installing many meters has led TURN to believe that PG&E's request is premised on the fact that PG&E has squandered its original budget.50
In response, PG&E states that it has provided substantial evidence regarding how the additional complexity of the industry and the new project technology will add to its project management costs, and intervenors cannot legitimately ignore the evidence presented by PG&E - that clearly shows a correlation between project management costs and increased numbers of vendors within an increasingly complex industry -- and instead rely on alternate theories that would correlate project management costs with the numbers of meters or networks being deployed.
As discussed further in this decision,51 we determine that PG&E's project management costs associated with the Upgrade should be considered as original AMI program costs, specifically under the risk based allowance. Therefore, for purposes of this proceeding, we need not determine an appropriate measure or theory to guide our determination of incremental project management costs, or whether PG&E's project management to date has been imprudent.
7.9. Operation and Maintenance Expense
PG&E has forecast $5.1 million in operation and maintenance (O&M) costs. These costs include O&M costs related to the load limiting switch, the HAN device and IT. The only category of these costs challenged by intervenors is that relating to expected calls to PG&E's call centers concerning the HAN device. These call center costs - forecast at $455,000 per year through 2010 - are tied to expected rates of HAN adoption.52 That is, the higher the rate of HAN adoption, the higher the expected call center costs.
DRA's benefit calculations reflect the use of a lower HAN adoption rate than assumed by PG&E. DRA modified PG&E's annual HAN technology adoption rate by a ratio of 21 to 30, which is equivalent to a scalar adjustment of 0.7. This adjustment results in the projected annual adoption rate increases from 0.1% in year 2012 to 21% in 2024. DRA recommends reducing PG&E's call center costs by 70% to reflect the fewer calls that will be received as a result of DRA's lower HAN adoption rate.53 DRA's adjustment results in a $319,000 reduction in O&M costs.
As discussed further on in this decision,54 we have adopted DRA's proposed HAN adoption rates, which were derived by applying a 0.7 scalar to PG&E's proposed adoption rates. Therefore, we will apply the same 0.7 scalar to PG&E's proposed call center costs, resulting in an adopted call center estimate of $319,000, which is $136,000 less than projected by PG&E.
7.10. Technology Assessment Costs
In PG&E's original AMI decision, the Commission stated:
While we recognize that PG&E's AMI deployment meets our functionality requirements as set forth, new technology may emerge that offers PG&E and its customers increased reliability and performance enhancements. We expect PG&E to monitor market place developments so, whenever feasible, it can upgrade its AMI system and offer its customers technology upgrades. (D.06-07-027, p. 52.)
In response to this statement, PG&E states that it has closely monitored the advancements in AMI technology advancements. In its application, PG&E proposed technology assessment and pilot costs of $15.4 million through 2012. These costs include approximately $9 million in staffing and other recurring costs and $6.4 million for a pilot test of new technologies.
Considering recent technology developments in communication networks supporting the transfer of information between a utility and its customers' premises, PG&E indicates that it has embarked on a program to identify, evaluate, and test the latest emerging technologies that it may be able to incorporate into its SmartMeter Program Upgrade.
In its May 2008 Supplemental Testimony, PG&E included additional technology assessment costs of $22.5 million for HAN standards development. This consists of $12.5 million for demonstration facility/laboratory testing environment, $5 million for labor for HAN standards support, and $5 million for devices that would enable home computers to function as in-home display devices.
PG&E states that it will continue to work with the other utilities in California and throughout the United States to establish standards for HAN technology and applications and encourage customers to take advantage of the benefits supported by HAN-enabled functionality.
The total of PG&E's technology assessment request is $37.9 million.
DRA states that given that PG&E's technology assessment request came in response to a Commission directive to monitor the market, DRA has proposed that this program be partially funded. DRA recommends an amount of $9 million (direct nominal dollars). DRA indicates that this figure would allow for the monitoring of emerging technologies. DRA excludes the cost of a technology laboratory, a demo facility for HAN devices, HAN standards work, development of a Zigbee device that can be plugged into a computer, and an ongoing pilot test of the Silver Springs Network.
DRA does not believe there are sufficient benefits in PG&E's business analysis to cover these costs. If the Commission disagrees, DRA would suggest moving up to a figure of $15.4 million, which is what PG&E included in its initial application and testimony in December 2007. That figure would only cover the monitoring of new technologies and the Silver Spring pilot, which is currently being carried out by PG&E anyway.
According to DRA, much of the added work that PG&E proposes is more properly done by organizations such as the Electric Power Research Institute, by national research laboratories, or by consortia jointly financed by several utilities. Furthermore, no other California utility has received an authorization to perform AMI-related research and development work at the same level as what PG&E has requested.55 DRA states that while SCE may have received more pre-deployment money than PG&E, adding $37 million will clearly put PG&E higher than SCE.
In response, with respect to HAN standards, PG&E cites the cross-examination of DRA's witness who stated it was not unreasonable of PG&E to request the funds for one pilot during the construction of this project. He indicated there might be value to a pilot but objected to the notion of establishing the timing and cost in this proceeding. PG&E argues that DRA has not provided any evidence regarding what timing or magnitude of testing is more appropriate than that provided by PG&E, and the only record evidence on this issue supports PG&E's proposal.
With respect to pilot testing, PG&E similarly cites the cross-examination of DRA's witness who stated that he agreed that PG&E should be involved in the HAN standards development process but does not agree that PG&E's cost estimate is the right number. PG&E again argues that DRA has not provided any countervailing evidence regarding what level of commitment is more appropriate than that proposed by PG&E, and the only record evidence on this issue supports PG&E's proposal.
Regarding PG&E's application request of $15.4 million, TURN recommends that the Commission reject the total amount.
TURN states that when the Commission authorized PG&E's full pre-deployment funding request in A.05-03-016 it did so in part because it felt that PG&E's AMI project was farther along than the other two electric utilities and that PG&E was past the technology assessment phase and required pre-deployment funding to essentially keep its AMI deployment on track. According to TURN, requesting the additional funds to evaluate AMI technology is akin to re-asking the Commission for pre-deployment funding, and PG&E is too far along in its AMI deployment to continue wasting ratepayer money to evaluate new AMI technologies.
TURN also states that D.06-07-027 already requires PG&E to regularly assess AMI technology and to report back to the Commission on its assessments as one of the requirements for receiving authorization of its proposed $1.7 billion funding request, and the Commission has therefore already funded PG&E's technology assessment activities with that $1.7 billion authorization.
Regarding PG&E's supplemental testimony request of $22.5 million, TURN recommends that the Commission authorize $2 million to provide input to and obtain information from private sector projects that will ultimately develop HAN standards.
It is TURN's position that developing HAN standards and functionality to enhance the commercial availability of home area networks is the job of private industry not the ratepayers. Private industry will benefit from selling HAN devices to customers and, therefore, private industry should have the responsibility of developing the technology. In addition, TURN asserts that HAN devices contained within a customer's home are the property of the customer and are not necessarily wholly devoted to managing the energy usage of appliance end-uses. TURN adds that, in the context of an application to redo a multi-billion dollar project a few years after it was authorized, the Commission should not fund extraneous exercises such as this.
In response, PG&E notes the cross-examination of TURN's witness who stated (1) he could not say he had the expertise to understand exactly what was going on in the HAN industry; (2) he did not know how a standard is developed for HAN; and (3) he did not know whether or not a pilot was necessary. PG&E asserts that TURN's recommendation for this cost category is arbitrary and put forth by a witness who acknowledged that he has no specific knowledge or understanding of PG&E's technology evaluation requirements, and, therefore, TURN's recommendation should be rejected.
In response, TURN states the depth of its witness's knowledge of HAN standards development is irrelevant, given that TURN does not believe any of the specific tasks related to its proposed disallowances are necessary for upgrading PG&E's existing AMI system with new meters.
PG&E's request has not been fully justified and appears to be excessive.
With respect to its application request of $9.0 million for staffing and recurring costs, PG&E indicates that it is actively evaluating broadband over power line (BPL) and medium-band over power line (MPL) network options along with Internet Protocol (IP) solutions as an approach to expand its network bandwidth and create a more open communications framework. In our previous discussion on network technologies, we gave PG&E latitude on the type of networks to be deployed, with the understanding that it would be within previously authorized budgets. It is not clear that these currently considered communication networks are deficient in particular respects. It is not clear how BPL, MPL or IP would be incorporated into the currently proposed AMI structure.
PG&E did indicate that the backhaul technology is in rapid development and there may be a time when new methods of data transport become commercially viable for deployment. However, while this may warrant continued monitoring, it does not necessarily warrant extensive evaluation processes as proposed by PG&E.
PG&E has not provided convincing evidence that its proposed technology assessment expenditures related to communication networks are necessary or reasonable. However, since there is potential value in having PG&E monitor market place developments, we will authorize $4.0 million for that purpose.56
With respect to the $6.4 million pilot testing request, it appears to be related to a network technology that is currently being considered and which may be deployed as part of the Upgrade. There is value in pilot testing to ensure that the proposed network can be integrated into the AMI and will work as intended. We will authorize the requested amount.
With respect to HAN standards development costs, we are in general agreement with the positions of DRA and TURN. Laboratory testing and product demonstrations should first be the responsibility of those in private industry who will in the end profit from the various HAN related devices. Also, some of the work might be done by organizations such as the Electric Power Research Institute, by national research laboratories, or by consortia jointly financed by several utilities. We see no justification for saddling PG&E's ratepayers alone with these laboratory testing and product demonstration costs. However, PG&E has alternatively proposed that for $21 million of its proposed costs, ratepayers would provide half of the amount and PG&E would obtain the remainder from other private or public sources to defray costs that exceed the ratepayer share.57 We see merit in PG&E's proposal as it relates to laboratory testing and product demonstrations. It is reasonable that ratepayers provide at least some of those costs related to protecting PG&E's system from such potential problems as security breaches, interference with bill reading and interruption of customers' service, which can be avoided by first testing devices in a lab that replicates PG&E's system. We will allow $6 million (plus the associated risk based allowance) for this purpose with the understanding that PG&E can use those ratepayer provided funds to the extent that it matches those funds from other sources. Any unspent funds should be credited back to ratepayers.
With respect to the $5 million for labor for HAN standards support, there is value in having PG&E provide input to and obtain information from private sector projects and to interact with developers and other utilities as HAN standards are developed, and we will provide funds to do so.
With respect to the $5 million for devices that would enable home computers to function as in-home display devices, the purpose of these costs is unclear. The funding is for a device that would enable IHD functionality on a home computer but it is included under technology assessment. We are not clear as to whether the device itself is being tested or whether the customers' use of the device is being assessed. If it is the former, we would exclude the costs as being the responsibility of those in private industry who will, in the end, profit from the device. If it is the latter, we see no reason why the device should be free or discounted when, under PG&E's Upgrade proposal, the cost of the IHD is the customer's responsibility. For these reasons, we will not adopt funds for this category.
In total, the adopted technology assessment costs amount to $15.4 million.
7.11. Training Costs
PG&E has included incremental training costs of $1,697,000 for installation vendor software training, Field Automation System training, and customer call center training. No party disputes any of these costs, and they will be adopted.
7.12. Risk Based Allowance
PG&E estimates $506,920,000 in Upgrade costs and on top of this adds an additional $65,533,000 as a risk based allowance or contingency. PG&E indicates that it followed the same approach in calculating its risk based allowance for the Upgrade as it followed in its original AMI application. In D.06-07-027, for that proceeding, the Commission authorized $128.8 million for a risk based allowance on top of $1,610.6 million of estimated project costs. In the Upgrade, the risk based allowance increases costs by 12.9%, while in the original AMI application, the risk based allowance increased costs by 8.0%.58
TURN recommends that the risk based allowance be limited to 7.5%, based on what was authorized in D.06-07-027.59
PG&E argues that its risk based allowance estimates are dependent on the category of cost and the specific risk associated with that category of cost. According to PG&E it followed the same procedure as in the original AMI application. That is, certain risk factors were assigned to specific cost categories based on PG&E's perception of what that risk factor should be. The 8% number is a result of assigning different risk factors to different cost categories and looking at the results in total. The overall risk based allowance percentage calculated for the Upgrade is higher than that of the original AMI request because the Upgrade has higher amounts of expenditures in the higher risk categories than did the original AMI request.
No party appears to object to the concept of a risk based allowance or contingency. Consistent with the outcome of PG&E's original AMI decision, we will adopt the use of such a factor for the Upgrade. We understand that elements of the risk profiles that were considered in determining the reasonableness of PG&E's contingency amounts were such things as "the types of equipment that PG&E is proposing to deploy; the maturity levels of the industries that will be providing equipment; vendor experience with similar projects; the timing and scope of the deployment efforts; the current phase of the different contract life cycles; the number and types of vendors that will be managed during the project; equipment failure rates; and other project based factors."60 We therefore consider these elements as the types of things that should be covered by the risk based allowance for both the original AMI project and the Upgrade.
Consistent with the manner in which the risk based allowance adopted in D.06-07-027 was calculated, we will adopt a risk based allowance for the Upgrade based on the risk profiles of the specific categories of Upgrade costs. That PG&E's estimated overall Upgrade risk based allowance factor of 12.9% is higher than the 8.0% allowance for the original AMI project is a result of PG&E's analysis of risk for specific categories of Upgrade related costs as opposed to its analysis of risk for specific categories of costs for original AMI project. We agree with PG&E's position that the analysis of risk for the Upgrade should consider the risk profiles specific to the Upgrade, rather than that of the original AMI project.
Because of the manner in which TURN's recommended risk based allowance factor is derived, there are no specific evaluations of, or agreements or disagreements with, the specific risk factors that PG&E has assigned to the various cost categories. However, it is not surprising that overall risk related to newer technologies included in the Upgrade, in particular the currently evolving HAN technology, and the information technology system integration might have higher risk factors than that for the more traditional technologies that were included in the original AMI project. A review of PG&E's proposed risk factors does not cause any specific concerns with the magnitude of the factors or with the cost categories to which they are applied. We will therefore adopt PG&E's proposed risk base allowance methodology along with the specific factors themselves and the categories of cost to which they are applied.
In adopting PG&E's broad application of the risk based allowance methodology to its cost estimates, for both the original AMI project and the Upgrade, we feel it is vital to fully consider the implications of the risk based allowance concept. Specifically, we must consider if, and to what extent, it can be assumed that the risk based allowances for the original AMI project should cover specific requested Upgrade costs. Also, going forward, we must be vigilant in identifying future costs related to the Upgrade that should be covered by the risk based allowance that we are adopting today, rather than covered by additional rates adopted in another proceeding where such costs might be raised, such as in a future general rate case (GRC).
Regarding future costs that may be related to the original AMI project or the Upgrade and which are raised in separate proceedings for the purpose of additional rate recovery, they are only speculative at this time. We can only note that, in order to get such additional rate recovery, PG&E has the burden to show that such costs are neither covered by the specific costs adopted in either proceeding nor by the risk based allowances adopted in either proceeding.
Regarding requested Upgrade costs that should be covered by the risk based allowance adopted in D.06-07-027 for the original AMI project, two requested Upgrade costs are of concern. They are the incremental project management costs and certain of the costs related to the Kern County electromechanical meter retrofit.
For project management, PG&E requests additional cost recovery for activities related to the newer technologies and an increased number of AMI vendors mostly caused by the added Upgrade functionalities. However, PG&E itself, as described above, includes "the types of equipment that PG&E is proposing to deploy ... and the number and types of vendors that will be managed during the project" as elements of the risk profiles that were considered in determining the reasonableness of PG&E's contingency amounts for the Upgrade, and we see no reason why it should be any different for the original AMI. It follows that these activities are of the type that should be covered by contingencies such as the risk based allowance. It is reasonable that the additional project management costs requested by PG&E as part of the Upgrade should instead be covered by the risk based allowance adopted in D.06-07-027.61 The requested amount of $15.318 million ($17.914 million PVRR) will be excluded from the adopted Upgrade costs.
For the electromechanical upgrade, it is reasonable to include the incremental costs of the advanced solid state meter, the integrated load limiting connect/disconnect switch, the HAN gateway device and the installation cost as part of the Upgrade costs. These are the specific costs necessary to provide the functionalities of the Upgrade project and are reasonable. However, the electromechanical upgrade also includes the costs needed to install the approximate 230,000 electromechanical meters that are being replaced by the upgraded devices. The question to consider is whether the stranded costs related to the premature retirement of the electromechanical meters should be absorbed through rates established for the original AMI or through rates established for the Upgrade. The decisions to deploy the electromechanical meters were made by PG&E in conjunction with the original AMI authorization. It is appropriate that the consequences of those decisions should be reflected as part of that same authorization.
Also, as indicated above, PG&E has identified changed timing and scope as elements of the risk profiles that were considered in determining the reasonableness of PG&E's contingency amounts for the Upgrade, and we see no reason why it should be any different for the original AMI. Changed scope (i.e., advanced meters with higher functionality) is the driving factor that resulted in the electromechanical meters and associated equipment becoming obsolete. It follows that the costs imposed by the premature retirement of the electromechanical meters are of the type that should be absorbed through the risk based allowance. Those costs were imposed as part of the original AMI project and it is reasonable to assume the related stranded costs should be covered by the risk based allowance authorized by D.06-07-027 for the original AMI project. We will therefore exclude $18.5 million ($20.0 million PVRR) related to the Kern County electromechanical meter retrofit from the adopted Upgrade costs.
14 PG&E forecast costs of $606.575 million reduced by the costs approved in its original AMI project for electromechanical meters, remote connect/disconnect collars and real time output devices, which amounted to $203.919 million. The costs do not include that related to the electromechanical meter upgrade which is quantified and discussed separately.
15 In its supplement, PG&E forecast costs of $607,819,000 reduced by the costs approved in its original AMI project for electromechanical meters, remote connect/disconnect collars and real time output devices, adjusted to reflect the estimated cost of the project decision to change from electromechanical meters to base solid state meters, which in total amounts to $265,030,000. The costs do not include that related to the electromechanical meter upgrade which is quantified and discussed separately.
16 The costs set forth in PG&E's application included a five-year warranty on the end-point technologies, whereas the revised costs include an option to extend the warranty by an additional 15 years.
17 The Kern County retrofit is discussed in more detail elsewhere in this decision.
18 Estimated incremental Upgrade costs were reduced by the $61.1 million amount.
19 Because of PG&E's concerns over the process followed by DRA, PG&E filed a motion to strike DRA's meter and equipment cost analysis. The motion was denied. However, in his oral ruling, the administrative law judge conceded the difficulty of relying on the evidence provided by DRA and indicated that any use of this information by the Commission in this proceeding will take into consideration the possible ramifications of the confidentiality restrictions, and the evidence would be weighed accordingly. See 5 RT 612-613.
20 DRA, Levesque, 4 RT 553.
21 The number for DRA incorporates PG&E's $61.1 million adjustment to baseline costs that was reflected in its May 2008 supplemental testimony. The total baseline costs for end-point technologies from PG&E's original AMI decision is approximately $265 million. For comparison purposes, PG&E's number does not include HAN connectivity costs or HAN Upgrade costs other than the HAN gateway device itself and does include new meter devices associated with the Kern County electromechanical meter upgrade.
22 The HAN retrofit is also referred to as ubiquitous HAN.
23 2 RT 271.
24 See PG&E Opening Brief, pp. 21-22.
25 Suspension costs include the monthly suspension costs that PG&E is contractually obligated to pay for suspending the installation contract, the monthly costs for suspending PG&E project management office operations, and the labor escalation costs PG&E would incur by installing the meters with HAN devices months later than originally planned.
26 While any future AMI system may differ from the upgraded SmartMeter Program, the current benefits of the SmartMeter Program will likely be obtainable through any future new systems and will continue.
27 This number includes the risk based allowance associated with the electromechanical meter retrofit.
28 Exhibit 208, p. 12.
29 Exhibit 209, Attachment G.
30 Id.
31 According to PG&E's opening brief, the costs of the Electromechanical Meter Upgrade of approximately $37.3 million (confidential Workpapers Supporting Exhibit 7, WP A-2, line 7) includes approximately $12.5 million of incremental equipment costs. This includes $4.8 million of incremental costs associated with advanced endpoint functionality (230,000 x ($58 - $37)), approximately $5.2 million of costs associated with the integrated load limiting switch (230,000 x $23), and approximately $2.5 million of costs associated with the HAN Gateway Device (230,000 x $11), for the endpoints located in PG&E's Kern Division (Confidential Workpapers Supporting Exhibit (PG&E-7), WP 1-50).
32 PG&E cites DRA, Exhibit 108, Exhibit 2, Chapter 3, p. 3-3, line 11 and TURN, Exhibit 208, pp. 12-13 as the basis of the suggestions.
33 DRA, Abbott, 4 RT 463.
34 See Section 7.12.2.
35 For example, regardless of what technical solution PG&E uses for a particular HAN device in the meter (RF or power-line based), the customer would be provided a single or common RF based protocol once the signal is made available within the customer's premises.
36 By changing the percentage split of ZigBee/HomePlug from 60%/40% to 70%/30%, the weighted average cost of the HAN Gateway Devices would be reduced by $0.75 and result in a decrease of approximately $4 million.
37 Reporter's Transcript, p. 548, line 13 to p. 549, line 9.
38 See Exhibit 7, p. 8.
39 Reduced costs are related to program administration, marketing, customer incentives and the call center.
40 D.08-02-009, p. 13.
41 See Section 10.2.4.
42 Exhibit 3-4W, p. WP 4-1.
43 Exhibit 3-5W, p. WP 5-3.
44 Exhibit 7-W, p. WP 1-71.
45 See Section 10.4.3.
46 See D.08-02-009, pp.13-14.
47 A description of the PCT program and the associated benefits is provided in Section 10.4 of this decision.
48 Descriptions of the PTR program and PTR benefits are provided in Sections 10.1 and 10.2 of this decision.
49 See Section 10.1.2.
50 TURN cites evidence that indicates that, while PG&E has already spent 79% of its authorized project management budget, it has only installed 4% of its forecast electric meters and 11% of total gas meter installations. Further, it has only activated 2% of its electric meters and only 1% of its gas meters.
51 See Section 7.12.2.
52 In rebuttal testimony, PG&E revised its forecast of call center costs in outlying years, but the forecast through 2010 remains the same as set forth in the December 2007 testimony. See PG&E, Exhibit 8, p. 3-19, Table 3-1.
53 DRA does not explain the apparent discrepancy of recommending adoption of 70% of PG&E's HAN adoption rates but recommending only 30% of the call center costs related to the HAN adoption rates.
54 See Section 9.1.4.
55 DRA indicates that SCE received a total of $67 million in pre-deployment funding ($12 million in A.05-03-026 and $45 million in A.05-12-026). PG&E received $49 million, and when the $37 million in technology assessment costs are added to $49 million, the result is $86 million.
56 For technology assessment, there is no evidence as to what costs might be reasonable for monitoring purposes as opposed to evaluation purposes. The $4.0 million amount for monitoring purposes is based on the assumption that monitoring costs and possibly some evaluation costs would be substantially less than the $9.0 million proposed by PG&E for essentially evaluation purposes.
57 See PG&E Reply Comments on Proposed Decision of ALJ Fukutome, pp. 1-2.
58 This overall percentage is calculated by dividing the total authorized risk based allowance by the total authorized costs less the authorized risk based allowance.
59 TURN calculates the percentage by dividing the total authorized risk based allowance by the total authorized costs.
60 See Exhibit 8, p. 10-10.
61 This adjustment does not apply to information technology project management, which has been estimated by PG&E to be $2.8 million, plus the associated risk based allowance. That amount is included in this decision as an authorized cost.
