How to fix our nation’s growing electricity supply problem

In the recent 2025-26 capacity auction for PJM, the nation’s largest grid operator serving the mid-Atlantic and Midwest, the auction price hit a record high of $269.92/MW-day. That’s up 10 times from the $28.92/MW-day price for 2024-25.

Supply isn’t meeting demand

Why did prices go so high? A number of factors contributed to a mismatch between supply and demand.

Supply is low due in part to power producers exiting the generation market and in part to fossil-fuel-powered plant retirements. Since the 2024-25 timeframe, about 6.6 GW of generation in the region has retired or is expected to retire, and PJM expects up to 40 GW, or 21% of the region’s installed capacity, to retire by 2030.

That’s the same amount that PJM expects its load to grow over the next 15 years, with the grid operator’s peak load forecast having increased by over 3 GW for 2025-26. Energy use in the PJM region is projected to increase 40% by 2039 — from 800,000 GWh to about 1 million GWh — driven by increasing electrification coupled with the meteoric rise of data centers and AI.

Load growth and retirement of large fossil generation could have been mitigated with greater buildout of renewable energy resources. But that’s been slowed by challenges including a major interconnection backlog that’s limited new generation capacity. While this backlog is a problem across the U.S., it’s most acute in PJM, which got the lowest score of all grid operators on Advanced Energy United’s 2024 Generator Interconnection Scorecard. As of early 2024, PJM had over 3,000 projects, representing 286.7 GW of capacity, in its queue. What’s more, the average time for a project to get approval has increased significantly in recent years to more than 5 years. Whether the issues begin with the interconnection queue or emerge after it, one thing is clear: energy supply is not keeping pace with demand, which doesn’t bode well for energy affordability or energy reliability — and slows down our quest for a net zero grid.

A flexible, cost-effective solution

The good news is that we have solutions that are ready to deploy now — and that will benefit grid operators, utilities, and customers alike.

One key to unlocking supply is distributed energy resources (DERs), including solar and energy storage. DERs mostly avoid interconnection queue issues because they are not interconnected at the transmission level. They can be deployed much more quickly and cost-effectively than large power plants. While we need to bring considerable amounts of new renewable energy generation online, utility-scale deployments alone aren’t a panacea. Most projects in the PJM queue are not expected to come online before 2030; the timeframe will be much longer for projects yet to enter the queue. So, the problem is not the market sending the signals to build — it’s getting the approvals to build.

DERs can avoid these delays. But as powerful as DERs can be for surmounting PJM’s supply and demand challenges, they can’t do all the work on their own. DERs like solar and wind introduce their own challenges, such as generation intermittency and grid imbalances. While energy storage can significantly mitigate these challenges, it can’t completely surmount them without any intelligence driving dispatch.

An even more effective solution, in concert with deploying DERs, is coupling distributed energy storage with an energy management system (EMS). To be the most effective, an EMS should be powered by predictive control and real-time optimization software that enhances energy storage performance for optimum peak shaving, load management, and coincident peak reduction.

Peak shaving and load management: Installed behind the meter along with energy storage, an intelligent EMS can help maintain system reliability. Using load forecasting, it can reduce peak loads for commercial and industrial clients and reduce large loads on the distribution system. Large commercial and industrial customers benefit by reducing their demand charges, while distribution system operators benefit from peak management of large industrial loads.
Coincident peak reduction: Installed in multi-site configuration, value stacking site peak loads and system coincident peaks can bring C&I customer savings while adding coincident peak reduction for distribution system operators (DSOs). When combined with an intelligent EMS, DSOs can automate coincident peak shaving solution and monitor effectiveness at the distribution feeder.

Even more powerful is aggregating these DERs + EMS + energy storage deployments into virtual power plants (VPPs), which are proving their worth as key players in increasing grid capacity and balancing the grid.

As their name implies, VPPs can take the place of power plants, enabling much-needed energy capacity to be deployed more quickly and cost-effectively. According to research firm Wood Mackenzie, the VPPs that are planned or already deployed in the U.S will offset 33 nuclear reactors’ worth of power.

They’re also showing they can provide valuable grid services. Under the U.S. Department of Energy (DOE) as part of the $1 billion Puerto Rico Energy Resilience Fund (PR-ERF), Generac Power Systems will facilitate residential solar and battery energy storage for Puerto Rican households. When aggregated into a VPP, these will provide resilience for the region’s entire power system. VPPs have already supported California’s power grid during a July heat wave. In Texas, grid operator ERCOT has enrolled 7 MW of flexible demand in VPP programs, to avoid a repeat of the Winter Storm Uri outages.

Demand response, a key component of VPPs, has kept the lights on this summer during California and Texas heat waves. PJM values demand response enough to have included it in the capacity it procured for 2025-26, but at only 5% of that capacity. Adding demand response in the form of automated VPPs can ensure reliable electricity supply in Texas and other regions.

The time to act is now

Our nation’s energy supply is facing a crisis. This issue is flaring up in PJM, where prices could remain high; Morgan Stanley projects prices could jump 157% in the next capacity auction in December, to a cap of $695MW/day. But no region is immune. Every region stands to benefit by deploying more VPPs with intelligent EMS capabilities — with more reliable energy and lower costs.

California could avoid $755 million in traditional power system costs and save consumers $550 million annually by deploying about 7.7 GW of VPP capacity — five times the state’s current capacity — by 2035, according to a report from Brattle.

The U.S. Department of Energy determined that tripling VPP deployments in the U.S. could address 10-20% of the expected 200 GW of peak-coincident demand that must be served with new energy generation resources by 2030 — while avoiding about $10 billion annually in grid costs. Along with other advanced technologies, VPPs could help expand the nation’s grid capacity by 20–100 GW — at a quarter of the cost of transmission lines.

The urgency of the situation requires that we act now. We have the tools; now, we need to use them.

Matt Irvin is CEO and co-founder of Maplewell Energy. Irvin co-created the company’s optimization and predictive control technology to enhance peak shaving applications, load management, and coincident peak. He has spent his career serving utilities, with experience at General Electric and Stanley Consultants. Matt has been lead architect on deep learning forecasting models, cloud software architecture, and edge computing architecture and has helped countless solar developers, ESCOs, energy managers, and facilities managers value stack revenue streams on battery energy storage.

The views and opinions expressed in this article are the author’s own, and do not necessarily reflect those held by pv magazine.

This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.