Last week’s Grid Edge Innovation Summit 2019, Greentech Media’s annual conference, featured two days of in-depth analysis and frank discussions about the challenges and opportunities that the rise of distributed energy is bringing to utilities and regulators seeking to integrate them into the grid.
We’ve already covered some of the highlights of the two-day conference, including the latest data from Wood Mackenzie Power & Renewables on grid edge investment trends, and presentations from utilities involved in cutting-edge projects involving rooftop solar, energy storage, electric vehicles and other distributed energy resources (DERs).
But as with all of our grid edge conferences, there were plenty of news announcements and shared insights to catch up on after the fact. And for our GTM Squared members, we have archives of the videos of the on-stage presentations for those interested in hearing more.
Here are some of the most interesting developments from last week’s event.
AutoGrid’s view on the world’s (currently) largest virtual power plant in Japan
Over the past few years, we’ve seen a bit of a competition over the title of “world’s largest virtual power plant.”
More specifically — and to differentiate what we’re talking about from big multi-megawatt behind-the-meter battery-based VPPs — this competition is for the world’s most aggressively scoped plans for battery-backed solar systems deployed in the thousands, enough to make them a potential replacement for traditional power plants in terms of system capacity and flexibility.
To date, the two biggest contenders have been the residential solar-storage projects being rolled out by Vermont utility Green Mountain Power and Australian utility AGL, which are aiming at 2,000 and 1,000 homes participating, respectively. To be sure, neither project has yet reached those goals, and both have had their reported share of challenges in getting things rolling with their groundbreaking efforts.
But now Japan has taken claim to the title of world’s largest VPP, at least on paper.
The plan, announced last week by Japanese energy trading company Eneres Co. and Silicon Valley startup AutoGrid, is to build the “largest storage-virtual power plant in the world by asset volume,” with more than 10,000 behind-the-meter DERs, including solar panels, energy storage systems, combined-heat-and-power units, and electric vehicle batteries, in 2020 and 2021.
Eneres is already developing commercial-industrial demand response assets that will form the foundation of the project’s first phase. But the second, more DER-focused phase will incorporate DERs at scale, using AutoGrid’s Flex platform, to create a “dispatchable virtual resource to participate in Japanese wholesale energy and capacity markets.”
AutoGrid’s software is managing more than 5,000 megawatts of DERs for more than 50 leading energy companies including Total SA, National Grid, CPS Energy, NextEra Energy, Schneider Electric and Xcel Energy. Most of its work has been in North America providing demand response optimization software to utilities. But it’s also orchestrating large-scale industrial VPPs in Europe, and is playing a role in projects in India, Australia, and now, Japan.
The companies haven’t released more details on their record-breaking VPP project, such as where it will be located or how much it will cost. But Adam Todorski, AutoGrid’s senior director of product technology, did offer some thoughts on the project during an afternoon panel session at Grid Edge Innovation Summit (GEIS) 2019, including its value as “a nice case study of one way this regulatory environment driving a business opportunity can work.”
Japan is in the midst of a foundational shift in its energy markets and infrastructure, from its nuclear power plant shutdown in the wake of the 2011 Fukushima disaster and the rapid adoption of solar, wind, batteries and other clean-energy replacements (alongside imported natural gas), to the coming reform of its energy regulations, meant to open the vertically integrated system to more competition among energy providers.
AutoGrid’s partner Eneres has adapted to these changes by expanding the scope of its energy services efforts from traditional demand-response capacity to more flexible reserves and regulation capabilities, Todorski noted.
“We’ve had scales of hundreds of resources in pilot scenarios doing precisely that,” he said. “And the system we’re setting up for Eneres will scale into the tens of thousands — because that’s the scale that makes sense.”
OpenDSP: The apps developer platform for grid edge distributed intelligence
For the past six years, we’ve been covering an effort among U.S. utilities and technology vendors to create a distributed intelligence operating system for the distribution grid — an effort commonly summarized as equivalent to building a “smartphone” for the grid, and allowing developers to write “apps” to run on them.
The technology behind this effort is called Open Field Message Bus, or OpenFMB, and has been championed on the utility side largely by Duke Energy, Texas utility CPS Energy and Pacific Northwest utility Avista. Duke has hosted the first real-world microgrids running on OpenFMB specifications, while Avista has been seeking ways to leverage its capabilities for its Spokane, Wash. smart city project with locally headquartered smart meter vendor Itron, among other uses.
In terms of standards development, the OpenFMB effort has been handed from the Smart Grid Interoperability Panel industry organization, through its merger with the Smart Electric Power Alliance, to the Utility Communications Architecture International Users Group, which is handling the next stages of commercializing the standard.
At this year’s DistribuTech conference, Duke and partner Sierra Wireless demonstrated a wireless grid router with hardware and firmware enhancements making it ready to run OpenFMB in future commercial applications.
OpenFMB covers the hardware development side of things. But what about writing the apps to run on OpenFMB-enabled devices? To spur that effort, Duke and Avista made a joint investment in February into Open Energy Solutions, a startup headed by Wade Malcolm, the former chief of Siemens and Accenture joint venture Omnetric Group.
The purpose of this investment is to develop an Open Distributed System Platform, or OpenDSP — an open-source software platform that can serve as the development platform for coders to write “apps” that run on OpenFMB devices. Specifically, OpenDSP will be a “micro-service-based architecture built on a specific Linux variant with packaged services,” as described by Duke and Avista engineers involved in the project.
Kurt Kirkeby, a fellow engineer at Avista who’s leading its OpenDSP efforts, spoke at last week’s GEIS about the challenges his utility faces in integrating DERs into its distribution grid.
Avista has made significant investments into distribution automation and grid-scale energy storage, along with the information and communications technology to support it. But these traditional SCADA systems and centralized command-and-control software platforms simply can’t coordinate tens or hundreds of thousands of DERs at the speeds necessary to solve edge-of-grid voltage stability or power flow problems, he noted.
Building a system that can handle these tasks requires treating DERs as individual computers in a network of like-minded computers, sharing data and making decisions independently of a central control system, he said. “We will never be take advantage of these distributed energy resources that can’t communicate with each other.”
But that will require a new approach to integrating DERs that treats them more like network nodes to be “discovered” and incorporated into the platform, rather than as traditional grid devices, he said.
“We’ve got this whole discovery problem, and then we’ve got [the issue of] how those discovered devices interoperate with other discovered devices. So you’re into a grid IOT platform. That’s what we need,” he said — and with OpenDSP, “we finally decided we wanted to do something about it and created an open-source platform.”
Kirkeby stressed that the work being funded by Duke and Avista at OES will be truly “open source for utilties — it’s not our thing; it’s a utility industry thing.”
By next month, the group expects to have their first reference applications, along with about four functional bundles of code being developed at Duke and Avista.
While the backers of OpenFMB and now OpenDSP have relied heavily on the smartphone-and-apps metaphor to describe their work to date, Kirkeby made clear on the GEIS stage that these “code bundles” represent real-world distribution grid capabilities that might be better served by a more robust description.
“I don’t even want to call them apps, because we’re trying to get away from apps,” he said. “They’re the first functionalities.”
Competing views on utility data agility
One of the central themes of Greentech Media’s grid edge events is the challenge of managing the massive amounts of data flowing from smart meters, grid sensors, DERs and other devices of importance to how we generate, distribute and consume electricity.
Smart meters, which now serve more than half of all U.S. electric customers, have represented the first wave of grid edge data for utilities to manage.
The first generation of smart meters lack the computing power and network capacity to do more advanced data-collection and analytics, but the current generation is capable of much more, Tim Weidenbach, senior vice president of customer operations for Landis+Gyr North America, said in a GEIS panel discussion.
“In the next phase of meter deployments, our sample rates will go to the point [where] you’re seeing waveform capture at a residential level,” he said.
This kind of detail can inform energy disaggregation capabilities like searching for signs in the electric waveform that indicate a failing air conditioner, or faults in an office building’s HVAC system, he said. It’s not as high-fidelity as the data being collected by phasor measurement units deployed across much of the country’s transmission grid, he said, but it’s pretty close.
Utilities that have deployed earlier-generation smart meters are still pulling lots of data from them, while also adding the latest technologies, Stephen T. Johnston, growth development manager at San Diego Gas & Electric, noted on a GEIS panel.
“We are an incredibly data-rich utility, and we use it everywhere from providing customers more insight on their bills, revenue protection and leak detection on the gas side, but also operationally, using that data to identify more efficient ways to do things in the field, and how to better mitigate risks.”
These efforts are also informing SDG&E’s groundbreaking wildfire detection and prevention efforts, including its home-brewed metrics for measuring fire risk, its practice of proactively de-energizing power lines to prevent them, and its efforts to build microgrids to support communities through the resulting outages, as Jonathan Woldemariam, the utility’s wildfire mitigation and vegetation management, noted on a GEIS panel.
At the same time, utility customers aren’t reaping the same benefits from this data, according to Devin Hampton, vice president of corporate development at UtilityAPI. His company builds software that translates myriad utility data formats and standards, mostly having to do with energy usage and billing, into data that can be used by utility customers and the solar installers, efficiency retrofitters and energy services providers that serve them.
From that standpoint, a good analogy for how customers want to access their data is a bank ATM, he said.
“You don’t care how the banks are talking to each other, you just want your money,” he said. That requires an approach to data-sharing “that’s both safe and secure, but also streamlined and standardized, so you don’t have to start over every time you move to a new utility.”
To date, the utility industry’s implementation of purportedly common standards has been mixed at best, Hampton said.
“There are a couple of standards in the marketplace,” he said, notably Green Button Connect, meant to standardize consumer smart meter data for electronic transfer.
But Hampton said that he did not have "the best opinion of what Green Button Connect can do,” based on his experience working with various utility implementations at UtilityAPI.
Implementation has varied significantly from utility to utility, and in some cases has required intervention by state regulators to move things along, he noted.
