V. Metering Requirements
There are two critical CSI implementation issues concerning meters: (1) whether to require separate metering of solar output, and (2) to what extent, to whom, and through what communications medium to relay solar system performance data from these meters. Subsidiary questions relate to the associated costs and benefits of the meters and ongoing communication functions, how the metering and communications mechanisms might be integrated into the proposed advanced metering infrastructure (AMI) plans of the utilities, and TOU tariff requirements.
An explanation of the current circumstances may help put this in context. First, with regard to total solar system output, both the Commission's SGIP program and the CEC's ERP program currently require a second customer-owned meter, separate from the main utility meter, to measure the gross solar output performance of the solar system. This meter is referred to as the net generation output meter. However, there is no requirement to connect any kind of communications device to this solar meter in order to deliver real-time or periodic reports about system performance to the owner, the manufacturer, or the utility. Some owners pay at their own expense for a reporting function. Smaller systems may have this meter installed as an integral part of inverter equipment, with accuracies in the range of plus or minus 5%.
Second, solar customers that elect to go on a net energy metering (NEM) credit system face specific metering requirements. An NEM customer is only required to have a standard "cumulative" NEM meter that spins forward and backward, registering just the net purchase of electricity from the grid. Customers who elect to take service on a TOU tariff typically install a two-channel time-interval meter that separately records the net inflow and outflow of electricity for each applicable time interval. Neither of these utility-owned meters collects information on the gross generation of the solar system, nor do the meters have a communication path to the customer-owned NGOM meter that measures gross solar system output.
The Commission has previously expressed the need for good metering to manage and monitor solar installations and the program generally. (D.06-01-024, p. 31.) Accurate solar metering helps ensure that ratepayer incentives result in expected levels of solar generation. In D.06-01-024, the Commission identified the need for greater specificity of metering solar performance, and urged exploration of approaches rewarding on-peak solar production, including the kinds and costs of meters used in relation to quantifying solar production, utility bill savings, and NEM credits.
In its April 2006 Staff Proposal, the Staff recommended measures to address both meter accuracy and system performance feedback. First, Staff proposed that all CSI incentive recipients must have a dedicated revenue-grade meter (i.e., plus or minus 2% accuracy) to measure solar system output. Staff reasoned that a dedicated revenue grade meter ensures accuracy in monitoring system output for PBI payments and can support communication of accurate system performance data to all solar owners. Staff also envisions administrative cost savings when a PBI system's performance data can be sent remotely to the program administrator for payment processing.
Second, Staff proposed that all systems larger than 30 kW, even those not receiving PBI, have not only a dedicated solar meter measuring gross output, but also the ability to communicate this information remotely over the Internet (for a web-based reporting system) or by a utility reading and reporting system. According to Staff's proposal, solar performance data can better inform system owners about their system performance than the customer's net-metered utility bill. A specific solar report also serves as a reminder to the customer to check on system maintenance. Further, if tens of thousands of small solar systems require remote meter communications, Staff envisions the solar market will develop affordable metering and communication devices incorporated into system designs.
Staff made no specific proposal concerning the entity that would process this performance information and report it to customers and program administrators, and suggested a working group examine the possibility of a third-party operating the performance data retrieval and reporting system. Along with this proposal, Staff invited comments on the feasibility of including solar performance data on utility bills by January 2007.
A. Metering Quality and Accuracy
There is little disagreement that revenue grade meters are required to ensure accuracy of PBI payments. Parties were split on whether the Commission should require revenue grade meters for all other CSI participants. FST, SCE, and Clean Power Markets support revenue grade meters (+-2% accuracy) for all CSI participants. FST contends California may need revenue grade meters for all sizes of systems to meet the measurement and accuracy rules required by Renewable Portfolio Standard participation, if Phase II of this proceeding requires solar output measurements for DG solar renewable energy credits. CARE advocates that all systems should have a TOU net generation output meter, which we presume would be revenue grade. Similarly, CFC supports "real time meters" (presumably revenue grade) if the utility will be paying for the meters, but otherwise believes the meter choice must be "cost-effective" relative to performing production measurement. Parties that support revenue grade meters for all CSI participants based their support on increased data accuracy on performance to help drive technological advancement, increased owner knowledge of system performance to foster adequate maintenance, and a meter industry ready to provide these meters at a cost-effective level.
On the other hand, the Joint Solar Parties, ASPv, SDG&E/SoCalGas, PG&E, and CCSF believe revenue grade meters should be applicable to PBI participants only, while a lesser meter with plus or minus 5% accuracy should suffice for systems receiving EPBB incentives. While the cost for external revenue grade meters may be only slightly higher than standard accuracy meters, parties supporting an exemption from revenue grade meters for small systems argue revenue grade quality is simply unnecessary for smaller systems receiving the up-front EPBB incentive, since these do not require the same measurement function as larger systems receiving PBI incentives paid on measured performance. PG&E is not opposed to a revenue grade requirement for smaller systems, but would exempt residential systems unless generation data is used to calculate the EPBB incentive. SDG&E commented that it already meters all solar systems larger than 30 kW.
SDG&E offered detail on a range of revenue grade meters and their costs as indicated in the table below. 29
Table 8: SDG&E Data on Meter Costs30
PV System Size |
System Installed Cost |
Meter Cost as a Percentage of System Installed Cost31 |
2.5 kW |
$15,000 |
$25 meter = 0.2% $175 meter = 1.2% $750 meter = 5% |
10 kW |
$60,000 |
$25 meter < 0.1% $175 meter = 0.3% $750 meter = 1.3% |
30 kW |
$180,000 |
$25 meter < 0.1% $175 meter < 0.1% $750 meter = 0.4% |
100 kW |
$600,000 |
$25 meter < 0.1% $175 meter < 0.1% $750 meter = 0.1% |
A few parties made the distinction between the cost of external meters and less expensive meters integrated with the solar system inverter. Others state that internal meters are not accurate enough to rely upon for program needs. PG&E states that internal meters may be sufficient for small residential customers, but large systems participating in PBI should have a separate revenue grade meter. Joint Solar parties state that internal meters should be satisfactory if they are revenue grade. SCE and SDG&E/SoCalGas state that they are not aware of an internal meter that is revenue grade.
We find SDG&E/SoCalGas' comments particularly helpful as they explain different types and costs of revenue grade meters. Staff proposed an unspecified "revenue grade" meter to confirm solar production levels for all sizes of solar installations. SDG&E/SoCalGas reveal that revenue grade metering of solar system production is achievable at a variety of prices. Essentially, the meter price depends on the degree of time interval detail and the method for communicating meter information.
In its January CSI decision, the Commission expressed the desire for system performance metering that permits the customer to identify potential system problems requiring adjustments or repairs. We will require revenue grade meters for all systems paid incentives through CSI, either through the PBI or EPBB mechanism. We continue to believe that it is in the ratepayers' interest to have accountability for solar generation output under the EPBB incentive structure even though the incentive mechanism itself does not require metered output. Accurate measurement of performance for all system sizes is of paramount importance to ensure optimum value for both solar owners and ratepayers, and has the potential to better inform the solar industry and utilities about technology performance. Moreover, such accuracy preserves options when we later turn our attention to the treatment of renewable energy credits in Phase II of this proceeding. Using the cost data for revenue grade meters provided by SDG&E/SoCalGas, we find that requiring some level of revenue grade meter would not add a significant cost burden to CSI participants. Thus, we find it reasonable for all PV owners participating in the CSI program to install revenue grade meters at their own cost, regardless of the type of incentive payment received. While mutual benefits exist, we believe it is fundamentally in the interest of solar owners to include meters and communication technologies in their solar system designs. Thus, the metering and communication hardware and software shall be installed at customer expense as a condition for receiving the CSI incentives.
In summary, we set minimum requirements for revenue grade meters as follows:
Table 9: Metering Requirements
Size of System |
Minimum Revenue Grade Meter Required |
< 10 kW |
Basic electro-mechanical meter |
10-29 kW |
IDR meter |
30+ kW |
IDR meter |
To the extent that internal meters are certified as "revenue grade" based on national metering standards, these are equally acceptable to stand-alone external meters for systems smaller than 10 kW.
There are myriad technical and procedural details yet to be resolved related to the guidance provided by this decision on meters. These include specifications, issues of standards and certification, communication protocols and platforms, eligible recipients of information, and appropriate parties to execute these arrangements. We make general policy conclusions here, but need utility and industry metering experts to work out the technical details and to advise us further before we make decisions on further technical and procedural issues. In Section V.C below, we discuss the process for this further work, and we encourage the appropriate parties and technical personnel in the solar, utility, and metering industries to create a metering and data committee as part of the CSI Handbook process and on-going CSI Program Forum to address these issues.
B. Communicating Solar Performance
Having required revenue-grade performance meters, next we address what happens with the data collected. Specifically, we must resolve: (1) whether to move ahead now with reporting system performance information or wait and coordinate this effort with AMI rollout, (2) who will performing the monitoring and reporting function, i.e., what entity will receive the information and consolidate it into a report, and (3) which entities will receive the report once it is produced.
Although the Staff Proposal recommended revenue-grade meters and communication functionality, it made no specific proposals on these three issues. Parties provided comments on the issues, and we take each issue up separately below.
First, we address whether to require performance reporting and communication functionality now, ahead of AMI roll-outs by the utilities.
A number of parties, namely ASPv, CCSF, SDG&E/SoCalGas, PG&E, and FST, generally support the goal of using remote communication to carry out solar system performance feedback. Most of the utilities recommend the Commission not make a decision on requiring solar system performance monitoring until such time as AMI is decided. PG&E and SCE argue there is a potential for stranded costs if CSI meters are not compatible with AMI meters. PG&E comments that it expects a five-year roll out once it receives Commission approval for its proposed AMI plan.32 It comments that remote communication capabilities "would be helpful in providing general information on system operations for smaller installations," and that the specific meter and monitoring arrangement would need to be cost-effective. SCE believes all metering and communication technology should be AMI-compatible, and the utility should determine the "best fit" choice of meters and their placement. SCE states this requirement was not urgent and should be optional until AMI plans are resolved. SCE notes it would start its AMI rollout in 2009. SDG&E observes there needs to be flexibility of approach to fit with individual utility circumstances. SDG&E recommended the Commission go no farther than requiring an IDR meter at this time, reserving action on requiring a remote communication meter package until AMI is decided.
A few parties indicate that remote communication requirements should be applied to larger systems. ASPv and PG&E suggest remote communication for PBI only, while SDG&E, CCSF, and FST recommend this for all systems above 30 kW.
Offering a different view, FST supports immediate use of remote communication, indicating such methods are cost-effective now for systems 30 kW and above and amount to under 1% of system costs. FST believes this could be extended to systems greater than 10 kW in 2008, and to all size systems in 2009. To guide the pace of expanding remote communication and performance reporting, FST suggested the Commission require immediate remote communication capability where the combined cost of a package of hardware, software and the first five years of monitoring service does not exceed 1% of the total system installed cost up to 100 kW, 0.75% for 100 - 500 kW, and 0.25% for systems over 500 kW. For example, this would be up to $200 for a $20,000 solar system, and $1,250 for a $250,000 system. FST suggests individual customers can always upgrade to higher functionality. FST also believes a general requirement of communicating meters can reduce M&E and administrative costs.
The earlier sections of this decision have discussed in great detail the redesign of CSI incentives to incorporate a performance dimension and reward solar system output. We consider a performance feedback loop critical to achieving our goal of high performance solar technology. Therefore, we will not delay action pending the completion of the AMI proceedings. We will require that all solar systems receiving a CSI incentive, either PBI or EPBB, have some form of communication reporting capability. Options include remote communications via telephone, cable, modem or wireless transmission, or utilizing a utility's existing meter reading system. As discussed more fully in Section V.C below, the parties participating in the CSI Handbook Process can refine and recommend the exact details of this minimum communication function, within cost limits.
While the Commission would like data for all solar systems to be accessible remotely to both support solar technology improvement and to support monitoring and evaluation data requirements, we are concerned that requiring this capability without limits could become a cost barrier. While parties generally did not comment on who should pay for the reporting hardware and software, existing rules for SGIP and NEM make it clear that the customer typically pays for any expenses beyond providing the minimum utility revenue meter. A dedicated solar system meter goes beyond this minimum.
To ensure reasonable balance between customer cost and value received, the metering subgroup developing the draft CSI Program Handbook should develop minimum standards and functional requirements within an overall cost constraint for inclusion in the Handbook. We will rely on the comments of FST to specify that the total cost of the minimum metering, communication, and reporting system over the first five years for each solar installation size grouping shall be less than 1% of total installed solar project cost for systems up to 30 kW, and less than 0.5% for larger systems. If the communications requirements should fall outside the cost cap, we urge the metering subgroup to find some effective solution for performance feedback to solar owners.
With respect to issues of coordination between CSI metering requirements and AMI, it is vital that performance monitoring be available commencing in 2007. While we appreciate the potential value of integrating such a performance reporting system with AMI in the future, we do not want the prospect of future AMI decisions and not yet developed technical parameters to hold back solar performance monitoring. The metering committee working to develop the initial draft CSI Program Handbook should address the tasks necessary to establish a minimum performance monitoring capability in advance of AMI. Systems of 100 kW and larger shall have reporting capabilities as part of the incentive payment mechanism. Systems between 30 kW and 100 kW that receive EPBB incentives shall have reporting capabilities as soon as the protocols are established by the metering subgroup. Proposed protocols should be included in the initial draft CSI Program Handbook, which will be developed according to the schedule in Section IV.B.4. Wherever possible, standard data communication protocols and other specifications should be selected to preserve greater likelihood of AMI integration in years ahead.33
Turning to the issue of what entity carries out the performance data collection and reporting function, Staff noted that in addition to the solar owner or installer, a utility or other third party could perform the role of system monitor.
In response to Staff's proposal, FST explains these services can be provided by independent third parties who may be preferred to avoid potential bias from solar owner or solar manufacturer/installer performance reporting systems. FST contends that if the Commission later decides that renewable energy credits will be available for solar system owners, the renewable energy credit rules require independent third-party verification of renewable production using revenue-quality meters.
SDG&E maintains the utility must have access to the solar system meter, although it adds that its Rule 2534 is a good starting point for defining a possible role of third-party meter providers and services. FST agrees with this in their reply comments. PG&E states that even prior to AMI resolution, it could produce performance reports through its Alternate Billing System.
We find the entity responsible for administering the performance reporting system(s), should be an independent party - either the existing program administrators or one or more third-parties not affiliated with solar system manufacturers or installers. We will require parties to include a proposal for independent performance monitoring as part of the initial draft CSI Program Handbook, as discussed in Section IV.B.4. We agree with SDG&E and FST that Rule 25 regarding metering for direct access may serve as useful guidance for this effort.
Staff made no specific proposal on who should get access to the metered information beyond customers and program administrators.
Two parties address this issue. ASPv advocates a "data accumulation service" should be available for customer use in January 2007, and that data should be made available to solar market participants as soon as possible. FST argues that in the case of residential solar systems, performance data is far more useful when provided to solar industry stakeholders, i.e., installers and panel manufacturers, who have a business interest to ensure their systems are performing.
We will require that performance information be communicated to customers and program administrators as soon as feasible, and we direct Energy Division to ensure this issue is addressed in the initial draft CSI Program Handbook. In addition, we agree with FST that the information could prove useful to the solar industry in their design of components and integrated systems. We also see value to providing the information to the general public for general consumer research on prospective solar investments. The CSI Program Forum should consider the concept of broader release of program information, and accompanying privacy or data confidentiality concerns, and make a proposal through the process described in Section V.C below.
C. Further Work in CSI Handbook Process and CSI Program Forum
There was uniform support among the utility parties and FST for initiating a CSI meter and communication technology work group. These parties recommended the group could be comprised of solar and metering industry representatives, utilities, and Commission staff. The group would be tasked with establishing metering and data communications standards and coordinating details with unfolding AMI efforts.
In the sections above, we have directed various metering issues to be addressed either through the CSI Program Handbook development process described in Section IV.B.4, or through the CSI Program Forum that will convene in 2007. If the parties find it beneficial, they are free to organize a metering and data communication committee of either group so that the appropriate technical representatives of utilities, solar installers and manufacturers, metering and remote data communication providers, and customers can address these issues. In summary, we direct the parties to address the following metering issues in the CSI Program Handbook process:
1. Propose agreed upon meter standards and data transfer protocols, within 1% of total installed cost for systems up to 30 kW, and less than 0.5% of total installed costs for larger systems, for the requisite hardware, software, and performance reporting services, with the goal of standardization and widespread utilization in California.
2. Propose the kind of solar performance data to be included on the owner's solar system report or energy bill and the options for providing this information.
We direct the following issues to be considered by the CSI Program Forum:
1. Whether and how solar system manufacturers and integrators/installers should have access to performance data about their components and systems. There should be consideration of how to use data as potential for general consumer research for those considering buying a solar system, and how solar industry might use the data to improve performance of component products and/or integrated solar system designs.
2. AMI coordination issues once each utility's AMI plans, schedules, and any associated fee-for-service offerings become clear.
D. TOU Tariffs
A related dimension to solar system performance meters is whether we should require all CSI participants to be served on TOU tariffs to the extent participants' default meters and tariffs do not already have time differentiation.
Both CFC and CARE commented that the Commission should consider having all solar customers use either real-time or TOU meters, respectively. PG&E commented that "currently about half of PG&E's net metered customers take service on a TOU rate." (Reply comments, page 14.) No other parties commented on this topic.
The Commission has a long history of supporting TOU tariffs for customers, where they are cost-effective. Moreover, we understand that a large portion of solar capacity is already served by time-differentiated meters and tariffs, either because large customers required to be on TOU tariffs, or smaller customers, have opted for a TOU tariff to capture the financial advantages in bill savings and NEM credits from solar's day-time availability. Thus, most solar customers not already required to be on a TOU tariff voluntarily choose a TOU tariff to capture these benefits.
In the case of smaller solar systems for smaller customers not choosing a TOU tariff, the EPBB incentive structure, which pays based on a system's design relative to optimal south-to-west orientation for on-peak production, fulfills the goal of providing incentives for on-peak solar production. This is achieved without imposing the additional customer costs of IDR utility revenue meters and their associated monthly meter reading costs35.
To properly consider the possibility of requiring small solar customers to use a time-differentiated tariff, we need to look at the overall economics of such an action for the solar owner and ratepayers, including how this interacts with the value of bill savings, net energy metering, and avoided energy supply costs. We may also need to coordinate such a decision with other proceedings involving AMI and demand response tariffs. We intend to address this tariff question together with the cost-effectiveness issue scheduled for Phase II of this proceeding, as outlined in the Scoping Memo.
29 See SDG&E/SoCalGas Comments, 5/16/05, p. 22.
30 SDG&E explains that a simple (electro-mechanical) meter costs $25, an interval data recording (IDR) meter costs $175, and a meter with remote communications for collecting historical time series data costs approximately $750. (SDG&E/SoCalGas Comments, 5/16/05, p. 22.)
31 SDG&E's percentages of solar system installed cost are based on an assumed PV system cost of $6000/kW. Other parties' comments indicate systems today cost considerably more than that.
32 PG&E's AMI proposal (A.05-06-028) was approved by the Commission on July 20, 2006.
33 We realize that the metering protocols may require additional time. If proposed performance monitoring protocols cannot be established by the deadlines for the initial draft CSI Program Handbook, the ALJ may determine a separate schedule for the metering committee to file its proposals.
34 Rule 25 pertains to Direct Access third-party meter and data rules. This is Rule 22 for PG&E and SCE.
35 We require an IDR solar production meter for systems above 10 kW. The cost for these meters can be substantially lower than the utility charges for rate-based utility TOU meters and utility meter reading.
