The California Rule 21 Conundrum: DER End-End Assurance
Led by California and the IEEE 1547 Work Group, the electric utility industry is rapidly developing a standardized method to communicate with and manage the growing penetration of DER assets at the distribution level. The California PUC, in conjunction with the IOUs and vendor community, has established a set of procedures for ensuring that the smart inverters meet both performance and communications requirements between utilities and the smart inverters.
This article looks at the current plans and how well they will ensure the desired performance of smart inverters under the direction of a utility DER management system.
While the test and certification procedures being put in place will make a huge difference in how well systems will interoperate and meet performance requirements, some additional testing will be required to ensure end-end interoperability and performance.
The CA Rule 21 Test and Certification Plan
CA Rule 21 specifies how distributed energy resources (DERs) such as Solar PV and battery storage interconnect to the grid. To address smart inverters, the updated Rule is organized into three implementation phases that correlate to three distinct parts of the testing and certification of the inverters and communications systems for California.
Two certification phases are currently available and mandated and the 3rd is coming in 2019.
- Phase 1 is already in place and specifies a set of “autonomous” inverter functions that are tested and certified according to the UL 1741SA procedures.
- The second phase is the communications requirement and is tested and certified according to the SunSpec CSIP Test Procedures and Program.
- The 3rd phase includes additional smart inverter functions requiring more intense communications and these will be tested and certified based on the pending IEEE 1547.1 test procedures for both the functions and the communications about these functions.
The overriding goal of these programs is to reduce the costs and time associated with adding and managing DERs for the benefit of the distribution system and its customers. Key to accomplishing this is the standardization of the functionality of smart inverters and the communications used to manage the DER assets. The goal of the testing and certification process is to ensure the intent of the utility is communicated to and performed correctly by the smart inverters.
Let’s dig a bit deeper into the current process to understand the likelihood of achieving these goals.
Tackling the End-End Problem One Step at a Time
While the goal is end-end validation of the behavioral intent of the inverters, the testing takes a building block approach. This makes sense in that there are so many potential use case scenarios and combinations of equipment, systems, aggregators, etc., that validating even a fraction of actual end-end possible implementations is an unmanageable task.
The building block approach is the way we test most complex systems today. For instance, we test printers so they can communicate over Wi-Fi and wired networks, but we separately test that computer systems can communicate using Wi-Fi. The industry has refined the tools, testing and certification so that you can take any Wi-Fi certified device and it will almost surely “plug and play” with any other Wi-Fi certified device even if the two devices have not been tested and certified together.
The Building Blocks to End-End Interoperability and Performance
The CA Rule 21 approach is to test the inverter functionality separate from the communications protocol. Inverter functionality is tested and certified by Nationally Recognized Test Labs (NRTL) accredited for certifying inverters to the UL 1741SA test specification. This covers the CA Rule 21 Phase 1 functions but only 2 of the 8 Phase 3 smart inverter functions. UL 1741SA does not address any of the required CA Rule 21 communications capabilities. Its purpose is to validate that the tested functions do indeed operate as intended.
Once the updated IEEE 1547.1 test standard for smart inverter functions is approved, smart inverters will be tested by the NRTLs using that test specification . IEEE 1547.1 will be more comprehensive and test additional Phase 3 functions (though not all).
While IEEE 1547.1 is primarily a functionality test comparable to (but more comprehensive than) UL 1741SA, it also adds for the first time a required communications capability using one of three designated protocols:
This is a huge step forward in that for the first time a hardware certification program will validate that sending instructions or information to a smart inverter in an industry standard protocol will indeed achieve the desired behavior of the system.
The significance of this for managing DER assets cannot be over-emphasized. It is a great starting point for gaining confidence in the interoperability AND performance of a smart inverter in a single certification program.
Does the CA Rule 21 Plan Guarantee End-End Performance?
One downside is that if an inverter is certified interoperable in one protocol, that does not mean that using a different protocol would work as well. For example, an inverter certified using SunSpec Modbus must have some form of protocol translator to convert an IEEE 2030.5 message into the SunSpec Modbus messages. Such a protocol adapter will need its own certification program at some point to ensure the conversions between IEEE 2030.5 and SunSpec Modbus produce the intended inverter results.
The CA Rule 21 Phase 2 mandated IEEE 2030.5 communications will be independently tested and certified using the SunSpec CSIP IEEE 2030.5 Test Specification and Program. This validates that the communications capabilities of the smart inverter, building EMS, or Aggregator can correctly exchange information and instructions with the utility DERMS systems using IEEE 2030.5.
However, this test and certification program does not validate that the actual inverter behavior conforms to the instructions in the IEEE 2030.5 messages. It validates the correct understanding of the message by the receiving party, but this assumes an accurate translation to the internal programming and behaviors of the physical inverter.
The theory is that if a utility DERMS is sending a 2030.5 message to a SunSpec CSIP certified inverter, building EMS or an Aggregator client, the message will be correctly interpreted, and the behavior of the controlled inverters changed to reflect the utility instructions. Thus, it should be possible for a utility to communicate with any UL 1741SA and SunSpec CSIP IEEE 2030.5 certified smart inverter, any SunSpec CSIP certified building EMS and any SunSpec CSIP certified Aggregator and the resulting inverter behavior will be as intended by the utility.
Interoperability challenges solved, right? More accurately, the interoperability challenges are being addressed and some of the necessary procedures put in place. While this greatly reduces the chances of interoperability issues, it does not guarantee that there won’t still be significant issues in the integrated system.
CA Rule 21 Testing: What’s Missing?
The CA Rule 21 plan is a huge step in the right direction and the SunSpec CSIP certification program will greatly increase the probability that certified systems will work together as intended. The elements in the building block approach are an absolute necessary step to achieve the end goals. Those building blocks start with the UL 1741SA testing (to be replaced with testing to the IEEE 1547-2018 when available) and then move to the SunSpec CSIP testing for the IEEE 2030.5 communications between utilities and inverters, building EMS or aggregator systems.
But there is not yet a formal testing and certification plan for the links between an inverter, aggregator or building EMS receiving an IEEE 2030.5 CSIP message and the actual inverter behaviors. And the actual test and certification programs, while extremely valuable, do not guarantee interoperability. Why is that?
The Certification Programs and Missing Elements
It is useful to think of test and certification programs as risk management tools. While the long-term goal of a certification program may be 99.99% interoperability, the reality of programs just starting up is that they are focused on the 20% of actual features and functions that make up 80% of the interactions between systems. This is because the amount of testing one could do to reach a 99% confidence level is usually impractical.
With unlimited time and resources, we could conduct comprehensive testing to certify a system, but the reality dictates finding the point of diminishing returns that let us balance costs versus quality.
A good example is the developing IEEE 1547-2018 interoperability testing. The interoperability testing is intended to demonstrate that one of the three named standards in the IEEE 1547-2018 standard (SunSpec, DNP3 or IEEE 2030.5) can be used to receive messages at the inverter and that those messages will result in the intended inverter settings and behaviors. A comprehensive approach would be to send messages that cover every possible function setting and combinations of settings and measure the resulting inverter behaviors. But this is more comprehensive than conducting a complete IEEE 1547.1 test and would take potentially months to complete.
Instead, recognizing that the vendors themselves conduct comprehensive testing and that testing a sample of potential inverter functions will increase the confidence that all functions are correct, the testing can be thought of as more of a spot check. If there are no problems with testing a sampling of behaviors using the communications protocol, then this should provide increased confidence that the inverter will interoperate successfully with other systems and perform per the instructions given it. If, on the other hand, some issues show up in the spot checks, then it won’t be certified, and the vendor will have information that allows it to do a deeper dive to understand specific and systemic issues with their implementation and fix them.
One Step Towards End-End Guarantees
An even better approach would be to use one of the named protocols to conduct all the specified IEEE 1547.1 functional tests. This would provide confidence that both the functions and the communications about those functions using SunSpec, DNP3 or IEEE 2030.5 behave as intended. The only downside is that this doesn’t validate that use of either of the other two non-tested protocols results in the same level of performance.
Assume we have a certified IEEE 1547-2018 inverter that includes validation of the IEEE 2030.5 messaging; a SunSpec CSIP IEEE 2030.5 certified inverter client and a SunSpec certified DERMS IEEE 2030.5 server. We plug them together and start sending instructions. If the instructions are only those that have been tested and certified, there is a high probability of successful interoperation.
However, there are cases that are not tested in the certification processes and may cause issues. For instance, if interoperability testing does not include tests for complex programs that include multiple inverter function setting changes in one message, there is a risk that such a message would result in an unexpected behavior.
When building EMS and aggregator systems are added to the mix, even more interoperability risks are created. A SunSpec CSIP certified Aggregator validates that it can correctly exchange information and instructions with a DERMS CSIP server, but it does not test at all that the instructions and information are correctly translated into whatever protocol it uses to communicate to an inverter. Even if the inverters are certified for IEEE 1547-2018, unless a specific test and certification step validating that the aggregator-inverter interface delivers the intent of the DERMS server message, a new risk has been introduced. Every time an IEEE 2030.5 message is translated into any other protocol (whether a standard or proprietary one), new possibilities for errors are created. And as of now, we don’t have a process in place or planned to address this issue.
What Can A Utility do to Insure End-End DER Performance?
It makes little sense for a utility to replicate all the testing that is already being required for CA Rule 21 acceptance. First and foremost, a utility can absolutely mandate that any DER inverters, building EMS systems or aggregators, and their own IEEE 2030.5 server pass the UL 1741SA/IEEE 1547-2018 and/or SunSpec CSIP IEEE 2030.5 certifications before even being considered for demonstrations, pilots or deployments of DER management systems.
Starting with these “building blocks”, the level of effort to reduce risk of interoperability problems becomes manageable. The new IEEE 1547.1-2018 test procedures include commissioning tests. These make sense for larger, one-off installations but are not very practical for any sort of large-scale deployment of smaller inverters. In this case, the most useful next step is to design and execute a use-case specific acceptance test process to be conducted internally or using a 3rd party lab with appropriate equipment and skills.
The primary focus of any utility testing program should be to ensure that its unique planned deployment scenarios are tested for the end-end system. If the deployment is focused on using behind-the-meter storage to manage excess solar PV through an aggregator, then the test design and test lab should be designed to validate that DERMS 2030.5 messages are implemented correctly in the end inverters. This would require:
- A clear specification of the use case including the types and nature of the messages and instructions – e.g., requests to the aggregator for storage and solar PV status updates; new ramp rates for storage; schedules of when to store and when to discharge into the grid; schedules or instructions to the PV inverters to change curves and settings.
- A grid simulator and a PV simulator along with measurement devices to capture both the inverter settings and electrical behaviors.
- A set of tests that implement the use case specification focused on the more complex scenarios such as multiple DER functional curves and setting changes in one message and scheduling of storage and PV inverter behaviors.
With such a test system in place, it would be relatively straightforward to validate each of the candidate aggregator and DER systems from an end-end perspective.
Over time, this task should shrink as the IEEE 1547.1 certification is put in place and future end-end certifications are designed and implemented by the industry.
Leveraging CA Rule 21 Certifications for Each Utility
There is no question that the CA Rule 21 test and certification requirements already being put in place will go a long way to reducing the cost and time to implement an effective DER communications system. This will benefit utilities, vendors, and consumers not just in California but globally since the standardization makes leveraging the investments being made a very attractive process.
For utilities with particularly unique or complex use cases, the CA Rule 21 procedures will make their specification and implementation tasks much easier. But there still will be cases where additional end-end testing will be required to gain the assurance required in relying on DERs as a critical part of the grid.