The modern, renewable-energy-rich power grid is an increasingly complicated machine to model. Maybe technology born of the massively complicated world of designing computer chips can help.
That’s the genesis of Pearl Street Technologies’ approach to analyzing the complexities of power flows and planning parameters for the transmission grid. Its Suite of Unified Grid Analyses with Renewables (Sugar) software platform can take minutes to find solutions to highly complex problems that can take traditional modeling methods weeks to derive — if they can pull it off at all.
On Tuesday, the Pittsburgh-based startup raised an undisclosed pre-seed investment from Powerhouse Ventures, following an earlier investment in Incite Ventures, the investment platform created in part by Nest co-founder Matt Rogers. The new investment will add to $1.8 million in Small Business Innovation Research grants from the National Science Foundation and the Department of Energy, as well as an award from DOE’s ARPA-E research program for grid optimization software.
The new funding will allow the six-person company to ramp up pilots and hire new employees to support its first moves into commercial deployment, CEO and co-founder David Bromberg said in an interview.
“We closed deals with major clients last quarter, and our pipeline continues to grow,” he said.
Since its 2018 spinout from Carnegie Mellon University, the startup has tested Sugar with a transmission grid operator and Southern Company, a vertically integrated utility that operates large-scale transmission, he said. Not only has the platform proven to be about 200 times faster at engineering analysis compared to industry-standard modeling tools, but it can also often find solutions that existing methods can’t, he said.
That may be because grid operators are facing a much more complex planning environment with the emergence of renewable energy, which rises and falls based on weather conditions rather than grid operator dispatches. As we’ve noted in previous coverage, the U.S. transmission system is failing to keep up with the growth in low-cost renewables, which could threaten to undermine efforts to decarbonize the electricity system.
Multiple studies indicate that the U.S. will need tens to hundreds of billions of dollars of transmission investment in the coming years to integrate the wind and solar power that’s become the most cost-effective new generation resource. Emily Kirsch, founder and CEO of Powerhouse, estimates that the market for grid planning and operations technology to support this growth stands at about $4 billion per year globally.
Compared to the “costly and time-intensive workarounds and brute-force number-crunching” needed to plan for these grid expansions today, Pearl Street Technologies “is reinventing grid planning software from the ground up,” Kirsch wrote in a Tuesday blog post.
From computer chip design to power grid optimization
Sugar’s capabilities in solving these hyper-complex modeling challenges stem from a 2014 project Bromberg started with his doctoral adviser, Larry Pileggi, who is the head of the Carnegie Mellon Electrical and Computing Engineering Department and Pearl Street Technologies' co-founder and chairman. The research centered on applying integrated-circuit simulation methods to power system analysis.
These computer chips are tiny, but contain “tens of billions of electronic devices on them that exhibit complicated behavior,” he said. “If you want that chip to work, you need to have really good models and robust simulation tools to understand what the behavior and performance of those devices will be.”
That’s a different approach than the power flow analysis software built by Siemens, General Electric and a handful of other providers for the world’s transmission utilities and grid operators. “Existing grid software can be good at modeling small changes over a base case, but if you try to model large changes, that can take days or weeks of time,” he said. In worst-case scenarios, the modeling process may crash, which may mean either that it couldn’t yield a safe solution to the problem it was given or that it couldn’t handle the computational complexity, he noted.
Fixing this often involves engineers adding new devices to the grid model, representing infrastructure that would have to be built to handle large or complex changes, leading to solutions that are “suboptimal and likely not very attractive from a cost standpoint.”
After a stint as a senior scientist with software startup Aurora Solar, Bromberg returned to Pittsburgh to spin out Pearl Street Technologies from Carnegie Mellon in 2018. “It quickly became apparent to us that the robustness we were building into the software [meant it could] simulate and optimize for complicated events,” such as integrating large amounts of renewables, retiring large amounts of generation, and “extreme contingencies like you might see in a natural disaster.”
“If any of these scenarios represent grid collapse, Sugar identifies and quantifies the root cause, giving the user exact locations and mitigation solutions to bring the model to a safe, reliable operating point,” he said.
Breaking down the barriers to the integration of renewables onto the grid
Pearl Street Technologies’ work with real-world utility use cases has given it a chance to prove Sugar’s outputs are accurate by running them through existing grid modeling systems and deploying its first commercial applications in grid planning, he said.
There’s no doubt that the country’s independent system operators (ISOs) and regional transmission organizations need help integrating renewable energy. Interconnection queues for wind and solar projects are backlogged across the country, and planning processes lag well behind the need to bring these clean energy resources online, grid experts contend.
Greening the U.S. power grid will require an overhaul of how these transmission operators plan and allocate the costs of upgrading their grids, these experts say. Renewables projects often bear the burden of paying for grid upgrades triggered by their interconnection, while more holistic approaches to growing grid capacity to accommodate the rapid rise in renewables are bogged down in outdated planning processes.
“The ISO partner we’ve been working with was testing Sugar as part of a policy study on renewable integration in their footprint and how to carve a path from where they are today to ultimately serving a majority of their load with renewables,” Bromberg said. So far, its efforts have yielded solutions that involve far fewer grid upgrades and less additional equipment.
Sugar can also reveal hard-to-find details like the ideal interconnection locations for new renewable energy projects that can reduce rather than increase congestion, he said.
A number of research projects and startups are working on technology to better integrate renewables into the grid, ranging from systems that can more accurately reveal capacity on congested circuits to advanced computing techniques to route power across transmission networks.
“There are a lot of groups out there trying to leverage advances in computing power” to increase the capacity of existing planning technologies to solve for complex problems, he noted. “[But] I don’t think anyone does it from our angle — it’s a fundamental reformulation of how the electrical physics of the grid are expressed.”