The implementation of the smart inverter requirements spelled out in California’s Interconnection Rule 21 is like a streaming media mini-series, complete with the drama one expects from fictionalized life.
The drama started back in 2013 with the establishment of the Smart Inverter Working Group under the auspices of the California PUC. Season One of the series had its ups and downs, but the combination of utilities and vendors worked quite effectively to develop a model for how smart inverters should behave and how the communications with them should be standardized using IEEE 2030.5.
Season One: Harmony
Season One started off on a feel-good note built around stories of working together and solving challenging problems. Major industry players came to some remarkable agreements and produced several documents on how the industry should move forward. Phase One (2014) specified how smart inverters should behave in the absence of any communications and instructions from the grid. This ended up as UL1741SA so that inverters could be certified to the standard. And these functions also were adopted by IEC 61850-4-720, IEEE 1547-2018, IEEE 1815 (DNP3), SunSpec Modbus, and IEEE 2030.5-2018.
Phase Two (2015) specified how the communications would be standardized between the grid operators and the DERs and identified IEEE 2030.5 as the default protocol for these communications. IEEE 1547-2018 also added IEEE 2030.5 as one of the three optional protocols for 1547 certification testing. Phase Two was adopted, and a date set for when any DER system communicating directly with a utility had to be certified to the CA Rule 21 IEEE 2030.5 requirements. That date was initially February 22, 2019 (9 months after publication of the SunSpec Test Specification) and was moved out 6 months in January of 2019 (to August 22, 2019) when it became apparent that the effort to implement the software and the certification program itself was significantly harder than expected.
Phase 3 (2016) was the specification of additional smart inverter functions that by their nature did require communications. These include setting of schedules and monitoring the DER status, alarms, etc. These functions also were adopted in IEEE 2030.5-2018 and mostly in IEEE 1547-2018.
Season Two: Discord
Up to this point, the process and work group functioned rather well together. But now we are entering Season Two and the drama begins to unfold.
In February of this year (2019), the California Solar and Storage Association (CALSSA), the primary voice of the industry in the CPUC proceedings, filed a Petition to modify the plans of the utilities for implementing the requirements of Phase 2 and Phase 3. The Petition took the CA IOUs to task for issuing what they call Advice Letters that:
“relied on a consensus process to develop details that were not contained in the Advice Letters. When the investor-owned utilities issued more detailed implementation plans, it became apparent that parties were far from reaching consensus on certain issues. CALSSA requests that the Commission modify the Resolutions to require the advice letters to include more details and not to exceed areas of consensus…”
Where Does IEEE 2030.5 Stop?
One issue that CALSSA raised was about SDG&Es implicit requirement that all smart inverters support IEEE 2030.5 locally as opposed to the SIWGs agreement that communications with inverters managed by an aggregator or building EMS were out of scope of CSIP and therefore did not need to be certified for IEEE 2030.5.
Getting to End-End Assurance – But How?
A second issue was the plan by SDG&E to develop and conduct their own testing to validate end-end performance of smart inverters when controlled by an aggregator or building EMS. CALSSA took strong exception:
“A one-sentence test procedure stated in a utility document that appears to have been hurriedly drafted bears no resemblance to the consensus-based testing protocols that are painstakingly developed by a diverse set of experts. Also, the CEC Approved Equipment List contains 548 inverter models that have been certified to the latest mandatory standard. The IOUs have not demonstrated that they can handle that volume of testing.”
CALSSA suggested an end-end DER test procedure of their own to be used until a formal procedure can be developed and implemented by the NRTLs. In the meantime, they stated, it may be necessary to again delay the date specifying when systems must be SunSpec CSIP IEEE 2030.5 certified.
Not to be bullied by anyone, the IOUs came back with some strong pushback against the CALSSA Petition and raised the ante by claiming that since the new IEEE 1547.1 would solve the end-end testing issue (which it won’t), any certification date for communications should be pushed out at least 18 months until 1547.1 products are on the market.
QualityLogic, SunSpec, and the Commission Staff pushed back on both CALSSA and the IOUs, arguing that their premise was incorrect and that delaying further certification requirements will only cause uncertainty and confusion in the industry and cause a major setback to the goal of integrating DER resources into grid operations in California. The major disconnect is that the utilities are looking for assurance that a message they send will result in the behavior changes of the target inverters (which the current CSIP IEEE 2030.5 testing does not do). And the IEEE 1547 only assures that very specific messages in one of three protocols (DNP3, SunSpec Modbus, or IEEE 2030.5) result in the intended inverter behavior. An inverter need not be (and probably won’t be) tested with an aggregator or building EMS system. So, there is no concept in 1547 interoperability testing of an end-end test.
QualityLogic, in its own comments, elaborated that:
The hope that IEEE 1547.1 will address the end-end testing issue is misplaced. What IEEE 1547.1 will not do is:
- IEEE 1547.1 will not require that one of the standard protocols is used in the installation and operation of an inverter. While it ensures a “capability” to use one of the protocols, the requirement to use it will be a vendor, utility, or policy specific decision.
- It does not ensure that the local interface for IEEE 1547.1 testing will be IEEE 2030.5. That will be up to the inverter vendor. If they already have a SunSpec or DNP3 local interface, that may well be the protocol used for IEEE 1547.1 certification.
- There is no “end-end” testing in IEEE 1547.1. The certification only validates that a correct message in one of the protocols from a simulated aggregator, utility, cloud-based adapter, EMS, etc., will result in the desired performance. There is no testing with a specific EMS, aggregator system, utility DERMS, or any other source that may be sending real instructions to the inverter.
- The IEEE 1547.1 Interoperability test is not a protocol test. While it ensures that the IEEE 1547 functions can be managed via a specific protocol (including monitoring and scheduling), it does not validate that the rest of the protocol is functioning correctly. That is what is done in a protocol test such as the SunSpec IEEE 2030.5 CSIP test. This means that an inverter can pass a 1547.1 interoperability test but still not communicate correctly with a production server for that protocol.
Season Three: In Progress
Season Two ended in a classic cliffhanger and Season Three is currently in production. The big question Season Three must answer is, “what will be the CPUCs decision on the CALSSA Petition?” The California Public Utilities Commission (CPUC) recently issued a Draft Resolution regarding the CA Rule 21 proceeding on June 6, 2019.
The Draft Resolution, in response to the Petition of the California Solar & Storage Association for Modification of Resolution E-4832 and Resolutions E-4898, proposes to move the compliance date from the current date of August 22, 2019 to January 22, 2020. It further re-affirms that individual inverters that are behind either an aggregator or a building EMS do not need to be IEEE 2030.5 CSIP certified.
The IOUs had proposed 18 or more months delay but the CPUC did not agree and chose instead a 5-month delay.
As we watched the various comments filed with the Commission, we felt compelled to weigh in and our comments were instrumental in some of the proposed Resolution. For example:
“QualityLogic further argues that a delay in the compliance deadline for Phase Two and Functions 1 (Monitor Key Data) and 8 (Scheduling) of Phase Three would be counter to California’s renewable energy targets. QualityLogic states that, in order to avoid hindering the State’s climate goals and avoid undermining faith in the CPUC’s Rule 21 process, the August 22, 2019 deadline should not be delayed.”
Our primary intent is to support our customers who have been investing in meeting the CA Rule 21 CSIP requirements. We want to make sure that their investments are recognized and valuable by minimizing delays in the requirement date.
Unfortunately for inverter vendors there is still significant ambiguity about what they need to do for protocol support. If they anticipate any direct communications between their inverter and the utility DERMS, they do need to have a certified interface (could be local or cloud based). But IEEE 1547 will require demonstration of local communications with one of three named protocols (DNP3, IEEE 2030.5, or SunSpec Modbus). So, implementing IEEE 2030.5 insures conformance to CA Rule 21 for direct communications and IEEE 1547.
There is another 30-day comment period but once the CPUC issues a draft it is likely to become the formal Resolution with minimal changes. QualityLogic is in the process of a deeper analysis of the 40-page draft Resolution. There are additional testing requirements that vendors and labs will have to pay attention to as well as a directive to leverage IEEE 1547.1 within a specified period after the approval of the standard.
Season Three should end when the final CPUC Resolution is approved. Season Four will delve into the latest twists to the plot lines introduced this year.