Details for: SDG&E Advice Letter 3546-E.pdf

Click on the image for full size preview

Document data

Clay Faber – Director
Regulatory Affairs
8330 Century Park Ct
San Diego, CA 92123

May 21, 2020
(U 902-E)

AND E-5016

San Diego Gas & Electric Company (“SDG&E”) hereby submits for approval by the
California Public Utilities Commission (“Commission”) a proposed methodology to collect
and report voltage excursion event data to estimate impacts on customers with inverterbased distributed energy resources (“DER”) interconnected with the electrical distribution
Decision (“D.”)16-06-052 required SDG&E, Pacific Gas and Electric Company (“PG&E”),
and Southern California Edison Company (“SCE”) – collectively the investor-owned utilities
(“IOUs” or “Utilities”) – to submit a Tier 3 Advice Letter (“AL”) with proposed revisions to
Electric Tariff Rule 21 to set forth technical requirements and effective dates for Phase 3
advance inverter functions.1 SDG&E submitted AL 3106-E on August 17, 2017, which
included a proposed date after which the technical requirements associated with smart
inverter Phase 2 communications Phase 3 advanced functions would need to be activated
on newly interconnected DER systems.


D.16-06-052, Alternate Decision Instituting Cost Certainty, Granting Joint Motions to Approve
Proposed Revisions to Electric Tariff Rule 21, and Providing Smart Inverter Development a
Pathway Forward for Pacific Gas & Electric Company, Southern California Edison Company,
and San Diego Gas & Electric Company, Ordering Paragraph (“OP”) 9.


- Page 1 -

Public Utilities Commission 2 May 21, 2020 On April 26, 2018, the Commission adopted Resolution E-4898, which incorporated the eight Phase 3 advanced functions recommended by the Smart Inverter Working Group (“SIWG”). The resolution established an effective date for Function 6 Volt-Watt mode2 of nine months following approval of the SunSpec Alliance Communication Protocol Certification Test Standard. On May 22, 2018, the SunSpec Alliance Communication Protocol Certification Test Standard was approved, and Function 6 was therefore activated nine months later, on February 22, 2019.3 The Commission found that the impact of Function 6 Volt-Watt mode activation was expected to be de minimis while the benefits to the electric distribution system could be realized. The Commission stated that: The Volt Watt Mode function will simply respond to these voltage events and keep the grid safe. The Volt Watt Mode function will facilitate higher penetration of DERs because it is the DERs interacting in a cumulative fashion (with everchanging load conditions) that can cause system voltage excursions […] If the activation of Function 6 is harmful to a particular system owner, it will only be because the voltage levels on the circuit are regularly outside acceptable Rule 2 limits.4 Finding that Function 6 was warranted and valuable, Resolution E-4898 directed the IOUs to 1) work with stakeholders to develop standardized reporting methodologies to monitor the frequency and amount of voltage excursions and 2) in consultation with the Commission’s Energy Division, each submit a Tier 1 AL on the proposed methodologies by October 1, 2018. SDG&E and the other IOUs timely submitted these required ALs.5 The IOUs’ October 1, 2018 submittals proposed the following process: • • • • • 2 3 4 5 Utilize each IOU’s respective voltage complaint process to help monitor the frequency and amount of voltage excursions experienced by DER customers who have installed a smart inverter with the Volt-Watt function activated. Monitor the voltage excursions by estimating the amount of energy reduction due to the activation of the Volt-Watt function. Identify and report on voltage issues caused on the utility side of the meter. Utilize Advanced Meter Infrastructure (“AMI”) data in conjunction with PVWatts production curves to estimate DER curtailment. Perform an energy loss calculation when the results of the investigation would reveal that the voltage issues were due to the customer’s equipment, for no more than 20 customers per year. The function modifies active power from DERs based on predetermined voltage ranges to prevent the local voltage on the distribution circuit from rising/dropping outside of allowable levels. A letter from Edward Randolph, Director of the CPUC’s Energy Division dated July 11, 2018, clarified that February 22, 2019, was the effective date for Function 6. Resolution E-4898, p. 27. PG&E 5395-E, SCE 3872-E, and SDG&E AL 3283-E.
- Page 2 -

Public Utilities Commission 3 May 21, 2020 Sunrun, Inc., (“Sunrun”) and the Interstate Renewable Energy Council, Inc., (“IREC”) protested the IOUs’ advice letters on October 22, 2018. Sunrun represented that the IOUs’ advice letters were missing detail regarding how the IOUs would proactively monitor, collect, report on, and respond to voltage excursions from AMI data. IREC expressed that the proposed methodologies lacked sufficient detail to ensure that the appropriate voltage data is obtained in compliance with Resolution E-4898. SDG&E replied to IREC’s protest on October 29, 2018. The IOUs’ advice letters were escalated to Tier 3 on July 22, 2019, requiring Commission action. On December 19, 2019, the Commission adopted Resolution E-5016 (“Resolution”) that denied the IOUs’ advice letters and proposed methodologies. The Resolution ordered the IOUs to 1) hold at least two meetings with the SIWG beginning within 45 days, and 2) resubmit Tier 1 ALs within 150 days of issuance of the Resolution. SDG&E submits this advice letter in accordance with the Resolution. Smart Inverter Working Group (SIWG) Meeting Overview The Resolution ordered the IOUs to meet with the SIWG to discuss whether PVWatts or another modeling tool or profile is best-suited to estimate photovoltaic (“PV”) production curtailment as a result of function 6 activation. In addition, the Resolution ordered the IOUs to provide additional information to the SIWG regarding what AMI data and capabilities are available to determine PV power curtailment, more specificity on the amount of data that would need to be processed under different monitoring approaches, and the feasibility and costs involved in using alarms and/or alerts in lieu of widespread data processing. The IOUs held four formal meetings with the SIWG during the period January 23 through March 19, 2020. In addition, the IOUs held multiple separate conference calls with the National Renewable Energy Laboratory (“NREL”) and with IREC and Sunrun. A summary of these meetings is provided in Appendix A. SDG&E highlights in greater detail specific discussions held on the topics required in the Resolution and highlighted above. During the SIWG call held on January 23, 2020, the IOUs discussed with the SIWG various PV modeling tools and profiles that could be utilized within energy loss estimation methodologies. The discussion focused on estimation method involving the use of PVWatts, typical PV outage profiles, and other NREL methods. During the February 6, 2020, SIWG meeting, the IOUs also discussed in greater detail the use of AMI data, individual IOU capabilities and limitations, and how each IOU could currently obtain the voltage data to estimate PV power curtailment. The IOUs provided details on AMI data related to voltage information, such as hourly voltage data, 15-minute interval voltage data along with average voltage data, and instantaneous voltage data. The IOUs also described the feasibility to develop systems, tools, and processes that manage AMI data, such as the use of alarms and/or alerts in lieu of widespread data processing. The IOUs each have differing capabilities and costs to implement a fully AMI solution to gather voltage data. Due to the expected de minimis energy loss values and the costs to
- Page 3 -

Public Utilities Commission 4 May 21, 2020 implement widespread data-gathering over AMI, the IOUs maintain the customer voltage complaint process as the most prudent way to gather data and report on voltage excursion events. Below, SDG&E describes various modeling tools for estimating power curtailment, what AMI data and capabilities are available, and the feasibility and costs of various methods for gathering voltage data. Power Curtailment Modeling Tools In accordance with Resolution E-5016, the IOUs considered multiple available tools for estimating power curtailment in consultation with NREL and the SIWG. SDG&E elaborates on the specifics of the available modeling tools, including the pros and cons of each methodology and an assessment of the technical feasibility of each. NREL Hawaii Method On a select feeder in Hawaii, NREL staff deployed irradiance monitors, voltage and current measurements, and a communication system to remotely measure the output of several PV systems. With this method, they were able to accurately determine the curtailed power production due to the volt-watt algorithm. This method directly provides the solar irradiance values as well as voltage and current. When coupled with the PV system power production and Volt-Watt curves, this method produces accurate value for curtailed power production. However, this is a costly and labor-intensive approach to determining PV curtailment. It took NREL staff several months to design and construct, making it impractical for largescale deployment. Therefore, this method was not reviewed further in the SIWG as a possible approach to compliance with the Resolution. NREL PVWatts® The PVWatts calculator is a web application developed by NREL that estimates the electricity production of a grid-connected PV system based on variable inputs. To use the calculator, information about the system’s location, basic design parameters, and an average annual retail electricity rate must be inputted. PVWatts then estimates of the system’s annual and monthly electricity production as well as the value of such production. SDG&E believes the PVWatts system can be useful for very preliminary studies of a PV system that uses modules (panels) with crystalline silicon or thin film PV cells. • Pros: This software has been utilized extensively and produces hourly data. It is also available at no charge from NREL’s website. It has the most flexibility and also considers seasonal production effects. Moreover, the PV profiles may be generated based on location-specific information (such as project type, installation angle, and zone within the service territory), which may increase the accuracy of the energy loss estimation. • Cons: PVWatts power output data is averaged over time and does not represent the specific output of a given system at specific time. In some cases, it could lead to energy loss estimation inaccuracies by predicting PV output intermittency that may not actually occur, as shown in Figure #1 Prominent PV Production Profiles. In these cases, using PVWatts could result in lower energy loss estimates than what would
- Page 4 -

Public Utilities Commission 5 May 21, 2020 occur under actual system operating conditions. Finally, the utilization of typical clear day PV output profiles may lead to inaccuracies in the energy loss estimations, because the profile assumes that all clear days in the year and fails to account for inclement weather. NREL Method 1 This method assumes that the power output of a PV system is 100% between the hours of 9:00 a.m. and 3:00 p.m. everyday. It was developed by NREL to provide a reasonably accurate calculation that is comparatively simpler than the more rigorous Hawaii method discussed above. NREL Method 1 provides an 8760 voltage profile, and it uses the approved volt-watt curve along with the simplified production profile to calculate the estimated PV curtailment. This method has been demonstrated and validated through NREL’s prior work with a different partnering utility.6 In Method 1, the voltage measured at the customer premises through the AMI sensor (i.e. “smart meter”) is used to estimate how much PV output would need to be curtailed from the PV inverter located at the customer’s premises. The curtailment amount would be in accordance with the volt-watt curve and is calculated using the following equation: = . . max( − 1.06 , 0) 1.1 − 1.06 Where: • • • • Ecurtailed is the maximum possible curtailment due to volt-watt, in kilowatt hours (kWh), over the time period of interest, PPV is the rated alternating current (“AC”) power of the PV system, in kW tAMI is the period of the AMI measurements, in hours vAMI is the set of AMI voltage readings for the time period between 9:00 a.m. and 3:00 p.m. in per-unit (“pu”) Method 1 requires the following data as inputs for the PV system where the curtailment is to be estimated: time-series AMI voltage data at the customer premises and rated AC power of the PV system at the customer premises. • • 6 Pros: This is a relatively simple method to implement, is more broadly understandable, and can be used within an energy loss estimation methodology without complexity. It provides a value that is conservative, as a worst-case curtailment when compared to other methods. Cons: As described above, by relying on 100% daytime production, this method overestimates actual curtailment, does not reflect any seasonal changes, and has no adjustment to reflect the actual PV system orientation. Therefore, depending on the PV system’s specific characteristics, NREL Method 1 may produce lower or higher energy estimate values. “Simulation of Hawaiian Electric Companies Feeder Operations with Advanced Inverters and Analysis of Annual Photovoltaic Curtailment,” available at:
- Page 5 -

Public Utilities Commission 6 May 21, 2020 NREL Method 2 The second method (referred to as “Method 2”) overcomes the limitations from Method 1 by using real data for modeling and simulation of distribution feeders to estimate potential curtailment in any desired scenario. In this method, the validated planning models for utility feeders are used for generating synthetic AMI data. The planning models can be augmented to represent future penetration scenarios, different PV inverter settings, and different voltage control device setpoints. This method is represented in the figure below: Curtailment Estimation based on Distribution Network Modeling and Analysis As shown above, the planning models for selected representative feeders are used to generate “synthetic” AMI data for different combinations of PV penetration scenarios, smart PV inverter settings, and irradiance profiles. The synthetic AMI data is used to estimate PV curtailment. By comparing the PV curtailment from different combinations of parameters like smart PV inverter settings, PV penetration levels, etc., the PV curtailment due to each of these parameters can be estimated better. • Pros: This method can be used when AMI data is unavailable, since it leverages circuit models and irradiance data to generate synthetic AMI data that can be utilized for estimating curtailment under a variety of scenarios. • Cons: This is a costly and labor-intensive approach that first starts by converting existing circuit models into an open source modelling software package before generating synthetic AMI data. Given the high-performance computing requirements necessary to solve the equations and then the time to compute per circuit, this methodology may be impractical to scale, even at a national laboratory, beyond a small group of feeders that may or may not be representative. Evaluation of Proposed PV Output Estimation Methodologies At the first SIWG meeting, SCE provided example calculations utilizing three methods: 1) PVWatts generated profile based on typical installation parameters; 2) Typical clear day PV output profile based on historical performance data; and 3) NREL Method 1 proposed PV output profile that, although not realistic, provides a simplified profile for estimating energy losses with adequate precision.
- Page 6 -

Public Utilities Commission 7 May 21, 2020 This graphical representation of each production profile is shown in Figure 1 Prominent PV Production Profiles. Figure 2 Calculation Comparison shows illustrative curtailment calculations for each production profile. Figure 1 – Prominent PV Production Profiles Figure 2 – Energy Loss Estimation Methods Recognizing that the estimation methods discussed above all inherently involve a level of uncertainty due to estimation assumptions, the SIWG discussed which method would be the most appropriate, accounting for the complexity and accuracy associated with each method. SCE presented Figure 3 Comparative Results to the SIWG to facilitate a comparison of energy loss method estimations that would assume a representative AMI meter voltage profile, as shown in Figure 4 Representative AMI Hourly Voltage Profile. As shown in Figure 3, the estimated energy loss values from each method are generally similar to each other and within the margin of error. The SIWG concurred that all three methods could yield a similar energy loss estimation.
- Page 7 -

Public Utilities Commission 8 May 21, 2020 Figure 3 – Comparative Results Figure 4 – Representative AMI Hourly Voltage Profile During the February 20, 2020, SIWG meeting, PG&E presented an analysis of AMI data and the calculation of PV curtailment utilizing the NREL Method 1. The findings for voltage outside the Rule 2 limits was based upon examining data from all meters.7 As shown in 7 Per Rule 2, the IOUs are allowed limited excursions beyond American National Standards Institute (ANSI) Range A (i.e. normal conditions) that may result from some temporary grid configuration of equipment problems.
- Page 8 -

Public Utilities Commission 9 May 21, 2020 the Figure 5 below, the analysis demonstrated that the impact of Volt-Watt function would be de minimis. PG&E also conducted an analysis of the NREL Method 1, which is shown in Figure 6 below. Again, the results demonstrate the de minimis impact of the Volt-Watt implementation and the ability of all the IOUs to conduct this consensus analysis.8 Figure 5 – PG&E Analysis of Voltage Readings from AMI 8 SDG&E’s analysis results are discussed in the Voltage Monitoring Application Group section below.
- Page 9 -

Public Utilities Commission 10 May 21, 2020 Figure 6 – PG&E Analysis of Curtailment Percentage using NREL Method 1 The SIWG determined that the preferred approach was NREL Method 1 along with NREL PV profile. Although providing a higher estimate of the energy loss, given its simplicity of calculation, this approach will yield an appropriate level of estimation accuracy. Voltage Data from Advanced Metering Infrastructure To promote transparency, Resolution E-5016 orders SDG&E to identify the available AMI data (i.e. instantaneous, average, minimum/maximum) and whether it has the same data available as the other IOUs, or what data is available for the different customer classes. The Resolution also directs SDG&E to include more specificity on AMI data and monitoring methodologies available, and how alarms and/or alerts could be set up to notify SDG&E of voltage excursions. As discussed with the SIWG on February 6, 2020, meters within SDG&E’s service territory are programmed to record five-minute voltage intervals. These intervals represent 12 times the data of the 60-minute intervals that SCE and PG&E collect from meters within their territories. Moreover, unlike SCE and PG&E, voltage data is stored in the meter and not collected over AMI. Smart meters are currently programmed to track voltage data in fiveminute intervals; however, SDG&E’s head-end system does not collect this data in its daily interrogations. SDG&E currently brings back 60- and 15-minute consumption data for billing purposes that consumes the majority of AMI capacity for traffic and storage. Expanding SDG&E’s data collection to include voltage data competes with retrieval of billing data and other production processes, such as interrogations, gap-filling, remote connect and disconnects, and over-the-air (“OTA”) meter reconfiguration. Therefore,
- Page 10 -

Public Utilities Commission 11 May 21, 2020 SDG&E’s existing AMI is unable to gather smart meter data in real-time or nearly real-time without system-wide upgrades to capacity to allow more traffic and storage. Despite these limitations in SDG&E’s existing AMI, SDG&E does collect voltage data over AMI from a small subset of the meter population. SDG&E shared this information with the SIWG and summarizes below. Voltage Monitoring Application Group (“VMAG”) SDG&E presented the SIWG with information on its VMAG, which is a dynamic population of roughly 18,000 meters that SDG&E is now looking to expand.9 The original VMAG included roughly 4,300 meters along 21 circuits. The metering group randomly expanded this population to roughly 18,000 meters in 2019 to test network capability. The circuits selected for the VMAG were selected based upon input from district engineering, electric distribution planning, and power quality. The circuits were selected to be representative of the SDG&E distribution system and include particularly problematic circuits, including those with high penetration of solar PV systems. This study was intended to: 1) demonstrate the usefulness of AMI data for system analysis, 2) provide general statistical data on distribution conditions, and 3) identify and quantify distribution system problems. SDG&E collects five-minute interval voltage data from the VMAG meters that is downloaded and stored daily. Data shows that the general distribution of voltage is concentrated between -1% and +3%. Because this population of meters predates the Function 6 activation date, SDG&E does not presently have such near-real-time voltage data for any DERs that have Function 6 activated. However, SDG&E has analyzed the data relative to the potential impacts of the Volt-Watt mode for the purposes of recent discussions with the SIWG. During the February 20, 2020, SIWG meeting, SDG&E presented some of these findings for voltage outside the Rule 2 limits.10 The table below summarizes the voltage exceedance data that falls outside of ANSI range, when the Volt-Watt function would activate to curtail PV production. The data is divided between the total VMAG population of 18,234 meters and the subset of meters with PV (approximately 45% of the total). Most importantly, the data shows that voltage generally stays within Rule 2 limits (i.e. within 5% of ANSI range). 9 10 SDG&E is now working on plans to expand to approximately 52,000 meters covering 106 circuits. These circuits are the additional circuits that are connected to the same substation bus as the original 21 circuits. In addition to the 106 circuits, this population would include all primary meters and volt-squared hour meters that have smart meters. Per Rule 2, the IOUs are allowed limited excursions beyond American National Standards Institute (“ANSI”) Range A (i.e. normal conditions) that may result from some temporary grid configuration of equipment problems.
- Page 11 -

Public Utilities Commission Voltage Exceedance 12 Total Meters: 18,234 May 21, 2020 PV Meters: 8,382 Five-Minute Reads Percentage of Total Five-Minute Reads Percentage of Total ≥ 6% 2,346,123 0.135% 634,000 0.078% ≤ 6% 1,940,325 0.112% 674,016 0.083% ± 5% 1,733,114,113 99.753% 812,265,057 99.839% SDG&E analyzed and presented information on the meters experiencing voltage exceedances for at least one hour. In the entire meter population, only 132 meters, or 0.72%, experienced this exceedance for 0.00059% of the time. In the PV meter subset, only 55 meters, or 0.3%, experience this for 0.00016% of the time. Finally, SDG&E presented a deeper analysis reflecting the number of meters experiencing at least hourly voltage exceedances at the times of day that a PV system would normally be expected to be producing power – roughly 7:00 a.m. to 7:00 p.m. Assuming a PV system capacity factor typical for San Diego (i.e., 0.2 kWh per kW) and 100% curtailment at every hour, the data shows approximately 0.003% curtailment for the PV systems associated with these meters. SDG&E also analyzed PV production curtailment impact on the one meter with the largest number of voltage exceedances described above and is shown in Figure 7. Three methodologies were utilized: SCE’s typical profile,11 PVWatts, and NREL Method 1. Refining the data further into curtailment by hour, there are 68 events between the hours of 9:00 a.m. and 3:00 p.m. The first group of numbers are the yearly production values without any curtailment. The second group of values are the yearly production values with curtailment by methodology. Finally, the third group of values present the percent calculated curtailment values by methodology. The two lower PV output production curves demonstrate the additional complexity that would be required to produce a more accurate production loss estimate absent actual production data from a net generator output meter. Consensus of the SIWG to adopt NREL Method 1, as shown in Figure 7, provided a conservative, higher value estimate on the PV curtailment. As anticipated in Resolution E4898, the data from the VMAG population shows the de minimis impacts that the activation of the Volt-Watt function would have on PV system production. 11 This SCE typical profile and calculations are identical to that presented by SCE in the first SIWG meeting.
- Page 12 -

Public Utilities Commission 13 May 21, 2020 Figure 7 – PV Curtailment Estimates for Worst Performing Meter Five-Minute Voltage Data over AMI Throughout SDG&E’s Entire Service Territory As SDG&E presented to the SIWG, SDG&E is in the process of increasing its monitoring of voltage data. SDG&E’s goal is to expand the VMAG population gradually over time to approximately 288,000 meters to provide coverage for all service transformers and primary meters, as follows: • • • • 1 meter for a service transformer with a single secondary customer 1 meter for primary meter services 1 meter for service transformer with multi-metered services at a single location 2 meters for service transformers with multiple customers at different locations using the meter closest to the station and the meter furthest from the transformer SDG&E anticipates this process to cost roughly $200,000, assuming two new full-time employees to develop analytics and operating practices. However, this approach cannot be implemented all at once. Rather, it must be implemented incrementally in phases to ensure that there are no negative impacts to normal business operations, including customer billing data. As SDG&E expands meter population subject to voltage monitoring, additional costs may need to be incurred to account for system upgrades. Modifications may be needed to change the interval lengths from five minutes to a larger interval (e.g. 10, 15, 30, or 60 minutes). Changing interval lengths requires many steps, including testing over a dozen separate meter configurations and then reprograming all meters in the service
- Page 13 -

Public Utilities Commission 14 May 21, 2020 territory over the air accordingly. In addition, this expansion may require additional investment for more Field Area Routers, application development and computing capabilities to handle the large volume of data, and storage, with an additional cost of up to $300,000. Threshold Voltage Data over AMI Using Alarms SDG&E has also considered using events or alarms to trigger the gathering of voltage data over AMI, as required in Resolution E-5016. Capability exists via AMI to set up notifications for specific conditions, such as 6% voltage exceedances, that could provide conditional voltage information for smart inverter services. When an exceedance is detected, this method targets and retrieves 48 hours of 5-minute voltage interval data from smart meters associated with inverters that have Function 6 Volt-Watt mode activated to record voltage excursion events without negatively impacting the retrieval of critical interval billing data. Any historical voltage data beyond the 48 hours would need to be collected manually. However, this approach has several drawbacks that increase costs and reduce the feasibility of the expanded VMAG approach outlined above. First, it targets a sample size with a custom interrogation process based on alarms. Second, SDG&E estimates that this approach could take about nine months to design and complete with a cost of roughly $700,000, requiring new computing infrastructure such as servers, storage, network, and dedicated resources that include a project manager, software developers, information technology quality assurance (“ITQA”), and multiple business analysts. However, this estimate does not account for additional costs that could be incurred as a result of SDG&E’s simultaneous effort to expand the VMAG population, which compete for the same resources. Therefore, SDG&E does not support this approach. Proposed Methodologies for Gathering Voltage Data and Estimating Curtailment Gathering Voltage Data Through Customer Complaints Consistent with the other IOUs and consensus in the SIWG, SDG&E proposes relying on the established customer voltage complaints process to collect voltage data to gather data on voltage excursions and estimate PV production curtailment in accordance with Resolution E-5016. Given what SDG&E knows from its VMAG population with the highconcentration of PV and SDG&E’s goal of expanding voltage data-collection throughout the service territory, SDG&E believes this to be the most prudent and appropriate method. Upon receiving a customer complaint, SDG&E would send an employee to the field to manually retrieve available voltage data using a laptop and an optical probe. However, the amount of historic voltage data available at each location is limited, due to meter storage constraints that vary by location and meter type. The file containing the interval voltage data would be returned to office for follow-up investigation. SDG&E’s smart meter system is not
- Page 14 -

Public Utilities Commission 15 May 21, 2020 capable of bringing back this amount of data over-the-air.12 This approach would leverage existing staffing and is not expected to add incremental costs to implement. The reporting methodology is consistent with the approach proposed by the other IOUs and consensus with the SIWG. Estimating Curtailment SDG&E’s proposed methodology is consistent with the approach that SCE and PG&E propose, based on collaboration with the SIWG and input from IREC and Sunrun. The IOUs discussed the methodology below with SIWG stakeholders on March 19, 2020. Stakeholders did not raise objections to the majority of the eight steps proposed below. One stakeholder stated that additional internal review was required on certain items (namely, 3.d and 7). However, no additional feedback was provided following the conclusion of the SIWG discussions and follow-up communication. Only one stakeholder commented against applying parameters to the duration of events. Because Rule 2 allows for limited excursions beyond ANSI Range A (i.e. normal conditions) that may result from some temporary grid configuration of equipment problems, it is appropriate to limit the duration of events, as reflected in step 2.c. Moreover, Resolution E-4898 recognizes that some limited voltage excursion can occur without violating Rule 2: “If the activation of Function 6 is harmful to a particular system owner, it will only be because the voltage levels on the circuit are regularly outside acceptable Rule 2 limits” (emphasis added).13 Below is the complete eight-step process that the IOUs propose. 12 13 Itron’s OpenWay® Collection Engine provides the interface between the metering system and utility processes, such as meter data management, billing, outage management and load control. Resolution E-4898, p. 27.
- Page 15 -

Public Utilities Commission 16 May 21, 2020 Eight-Step Voltage Complaint and Estimation Process Step 1 2 3 4 5 6 7 8 14 Activity The IOU voltage complaint process will be the basis for voltage reporting to satisfy CPUC requirements. The IOUs will provide curtailment calculations for Volt-Watt-enabled customers in the voltage complaint process with voltage readings over 106% of nominal voltage with the following characteristics: a) Include calculations for all utility-caused, customer-reported issues per reporting period. b) Report on up to 20 of the first customer-caused and reported issues on an annual basis. c) When the event duration is less than 48 hours, it will be considered a limited anomaly and will not be reported. IOUs will use the NREL Method 1 process, using currently available AMI data, to calculate estimated curtailment utilizing the following processes for estimation calculation: a) Voltage intervals of 1-hour for residential or 15-minute for commercial, as available b) Average voltage readings utilized over the voltage interval for SDG&E and SCE and instantaneous voltage readings for PG&E, based on existing capabilities c) NREL Method 1 is only being approved for generalized/anonymous reporting but not for potential individual customer claims d) For three-phase service points, the average voltage per phase will be used for curtailment estimation e) Any missing or removed data due to data quality will be removed from numerator and denominator of % curtailment calculations Report will provide yearly % energy loss calculation using the previous 12 months of the available voltage data per customer14 Report will provide 1 month % energy loss calculations using the previous 30 days of data as available per customer* Report will provide the utility mitigation used in the voltage complaint process or indicate if the reported complaint was a customer-caused issue Report will summarize the customer curtailment studies by the number of customers with curtailment estimates in the following buckets: a) ≤ 2%, b) 2% ≤ 4%, c) >4% Provide voltage histogram for % of total hours in Volt-Watt curtailment range (>106%) for the worst case customer with energy loss >5%. Consensus Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Will use available data within these date ranges from data stored in centralized AMI databases (PG&E/SCE) or the limit of meter memory (SDG&E). The yearly and 30-day reports are triggered when the voltage complaint is received.
- Page 16 -

Public Utilities Commission 17 May 21, 2020 Effective Date Pursuant to Resolution E-4898, SDG&E believes that this filing is subject to Energy Division disposition and should be classified as Tier 1 (effective pending disposition) and respectfully requests an effective date of May 21, 2020, the date submitted. Protest Anyone may protest this Advice Letter to the California Public Utilities Commission. The protest must state the grounds upon which it is based, including such items as financial and service impact, and should be submitted expeditiously. The protest must be made in writing and must be received by June 10, 2020, which is 20 days from the date submitted. There is no restriction on who may submit a protest. The address for mailing or delivering a protest to the Commission is: CPUC Energy Division Attention: Tariff Unit 505 Van Ness Avenue San Francisco, CA 94102 Copies of the protest should also be sent via email to the attention of the Energy Division Tariff Unit ( A copy of the protest should also be sent via email to the address shown below on the same date it is mailed or delivered to the Commission. Attn: Megan Caulson Regulatory Tariff Manager Email: Notice A copy of this filing has been served on the utilities and interested parties shown on the attached list including interested parties in R.11-09-011 and R.17-07-007 by either providing them a copy electronically or by mailing them a copy hereof, properly stamped and addressed. Address changes should be directed to SDG&E Tariffs by e-mail at /s/ Clay Faber ________________________________ CLAY FABER Director – Regulatory Affairs
- Page 17 -

Public Utilities Commission 18 May 21, 2020 Appendix A Summary of SIWG Meetings Side Meeting – December 16: SDG&E discussed the draft resolution to alert NREL of the discussion regarding PVWatts software, determine what alternative methodologies are available, and ensure NREL support in the SIWG effort to review alternative methodologies. Side Meeting – December 19: NREL presented alternative Method 1 for review. Side Meeting – December 30: Meeting occurred with NREL and SCE the purpose of this meeting was for SCE to gain a greater understanding of the NREL energy loss methodologies Side Meeting – January 9: Meeting held between NRELs and IOUs and centered upon discussions surrounding available software/options for estimating DER curtailment in response to voltage excursion events with Function 6 Volt-Watt mode activated. Side Meeting – January 17: Meeting with IREC, Sunrun, and IOUs regarding areas of the IOU Advice Letter filings and related protests that should be focused upon for SIWG discussions. Meeting 1 – January 23: SCE and NREL presented various energy loss estimation methodologies, including a comparison across the energy loss methodologies. The discussion found that that the various energy loss estimation methodologies would yield approximately the same estimated energy loss value. There was consensus within the SIWG to use NREL’s simplified curtailment calculation, which produced the highest estimated curtailment but was the easiest to implement. During the meeting, IREC and Sunrun indicated that their protests were triggered by lack of clarity on the implementation details and the limitations of the IOUs’ AMI systems. Meeting 2 – February 6: Each IOU gave presentations on the capabilities and limitations of their AMI systems. There was some discussion of the manufacturers providing voltage data to the IOUs to assist in vetting voltage violations, since the voltage at the inverter terminals is a known value. Meeting 3 – February 20: This meeting was focused on discussing the upcoming IOU advice letters and reporting – and what these items could look like along with confirming if prior outstanding issues and concerns had been resolved. Consensus appeared to coalesce around the NREL simplified curtailment calculation. Additional discussion ensued regarding the customer complaint process and what potential enhancements could be made including manufacturers directly submitting a complaint to each utility’s customer complaint process on behalf of the customer. The main focus of the meeting was to discuss preliminary voltage analysis based on data provided by SDG&E and PG&E. Both presentations demonstrated that the estimated energy curtailment would be expected to be de minims under the NREL energy loss methodology (Meeting 1).
- Page 18 -

Public Utilities Commission 19 May 21, 2020 Side Meeting – March 4: The IOUs met with IREC and Sunrun to walk through the IOU proposal for the May advice letter filing. PG&E shared a presentation that summarized at a high level the process. The meeting focused on trying to confirm the areas of consensus and address the points of disagreement. Side Meeting – March 11: The IOUs met again with IREC and Sunrun to walk through all eight points, identifying where there was consensus and where there was disagreement. Proposals to resolve the disagreements were discussed with the understanding that this process would be presented at the upcoming fourth SIWG meeting. Meeting 4 – March 19: The IOUs presented the outline of their proposed methodologies and reporting approach to stakeholders. Consensus has been reached on most items, but the IOUs and protesting parties (IREC and Sunrun) are not in agreement regarding application of a cutoff for the duration of voltage events for running calculations. No thirdparty stakeholders have offered any formal proposals at this point, and the IOUs will now start drafting their advice letters. The foundation of this advice letter was presented at this last SIWG meeting.
- Page 19 -

ADVICE LETTER SUMMARY ENERGY UTILITY MUST BE COMPLETED BY UTILITY (Attach additional pages as needed) Company name/CPUC Utility No.: San Diego Gas & Electric Company (U902- E) Utility type: ELC GAS PLC HEAT ELC = Electric PLC = Pipeline WATER Contact Person: Brittany Malowney Phone #: (858) 637-3714 E-mail: E-mail Disposition Notice to: EXPLANATION OF UTILITY TYPE GAS = Gas WATER = Water HEAT = Heat (Date Submitted / Received Stamp by CPUC) Tier Designation: 1 Advice Letter (AL) #: 3546-E Subject of AL: San Diego Gas & Electric Company Proposed Reporting Methodology to Monitor the Frequency and Amount of Voltage Excursion Events Pursuant to Resolutions E-4898 and E-5016 Keywords (choose from CPUC listing): AL Type: Monthly Quarterly Annual One-Time Other: If AL submitted in compliance with a Commission order, indicate relevant Decision/Resolution #: E-4898; E-5016 Does AL replace a withdrawn or rejected AL? If so, identify the prior AL: Yes, SDG&E AL 3283-E Summarize differences between the AL and the prior withdrawn or rejected AL: Please see attached Advice Letter Confidential treatment requested? Yes No If yes, specification of confidential information: Confidential information will be made available to appropriate parties who execute a nondisclosure agreement. Name and contact information to request nondisclosure agreement/ access to confidential information: Resolution required? Yes No Requested effective date: 5/21/20 No. of tariff sheets: 0 Estimated system annual revenue effect (%): N/A Estimated system average rate effect (%): N/A When rates are affected by AL, include attachment in AL showing average rate effects on customer classes (residential, small commercial, large C/I, agricultural, lighting). Tariff schedules affected: Service affected and changes proposed1: N/A Pending advice letters that revise the same tariff sheets: N/A 1 Discuss in AL if more space is needed. Clear Form
- Page 20 -

Protests and all other correspondence regarding this AL are due no later than 20 days after the date of this submittal, unless otherwise authorized by the Commission, and shall be sent to: CPUC, Energy Division Attention: Tariff Unit 505 Van Ness Avenue San Francisco, CA 94102 Email: Name: Megan Caulson Title: Regulatory Tariff Manager Utility Name: San Diego Gas & Electric Company Address: 8330 Century Park Court; CP 31D 92123 City: San Diego State: California Telephone (xxx) xxx-xxxx: (858) 654-1748 Facsimile (xxx) xxx-xxxx: Email: Name: SDG&E Tariff Department Title: Utility Name: San Diego Gas & Electric Company Address: 8330 Century Park Court; CP 31D 92123 City: San Diego State: California Telephone (xxx) xxx-xxxx: Facsimile (xxx) xxx-xxxx: Email: Clear Form
- Page 21 -

cc: (w/enclosures) General Order No. 96-B ADVICE LETTER SUBMITTAL MAILING LIST Public Utilities Commission Clean Energy Renewable Fuels, LLC Office of Ratepayer Advocates (ORA) P. DeVille R. Pocta Clean Power Research T. Schmid Energy Division M. Ghadessi G. Novotny M. Salinas Davis Wright Tremaine LLP L. Tan J. Pau R. Ciupagea Douglass & Liddell Tariff Unit D. Douglass CA Energy Commission D. Liddell B. Penning Ellison Schneider Harris & Donlan LLP B. Helft E. Janssen Advantage Energy C. Kappel C. Farrell Energy Policy Initiatives Center (USD) Alcantar & Kahl LLP S. Anders M. Cade Energy Regulatory Solutions Consultants K. Harteloo L. Medina AT&T Energy Strategies, Inc. Regulatory K. Campbell Barkovich & Yap, Inc. EQ Research B. Barkovich General Braun & Blaising, P.C. Goodin, MacBride, Squeri, & Day LLP S. Blaising B. Cragg D. Griffiths J. Squeri CA Dept. of General Services Green Charge H. Nanjo K. Lucas California Energy Markets Hanna and Morton LLP General N. Pedersen California Farm Bureau Federation JBS Energy K. Mills J. Nahigian California Wind Energy Keyes & Fox, LLP N. Rader B. Elder City of Poway Manatt, Phelps & Phillips LLP Poway City Hall D. Huard City of San Diego R. Keen L. Azar McKenna, Long & Aldridge LLP J. Cha J. Leslie D. Heard Morrison & Foerster LLP F. Ortlieb P. Hanschen H. Werner MRW & Associates LLC M. Rahman General NLine Energy M. Swindle NRG Energy D. Fellman Pacific Gas & Electric Co. M. Lawson M. Huffman Tariff Unit RTO Advisors S. Mara SCD Energy Solutions P. Muller Shute, Mihaly & Weinberger LLP O. Armi Solar Turbines C. Frank SPURR M. Rochman Southern California Edison Co. K. Gansecki TerraVerde Renewable Partners LLC F. Lee TURN M. Hawiger UCAN D. Kelly US Dept. of the Navy K. Davoodi US General Services Administration D. Bogni Valley Center Municipal Water Distr G. Broomell Western Manufactured Housing Communities Association S. Dey Interested Parties in: R.11-09-011 R.17-07-007
- Page 22 -