Duke Energy's Winter Bet on Solar and Batteries Has a Timing Problem

On the morning of December 24, 2022, Duke Energy cut power to roughly 500,000 customers across the Carolinas as Winter Storm Elliott pushed demand beyond what its system could deliver. The rolling outages ran for hours. Duke had not ordered blackouts in the Carolinas before, and it did not order them at midnight or in a July heat wave. It ordered them at breakfast.

That detail matters more than the storm itself. WLOS reported this month that Duke is using new solar and battery storage to prepare for winter demand. Winter is the season that tests whether solar-plus-storage can support a utility through its hardest hours.

A winter peak arrives before the sun does

Winter system peaks in the Carolinas tend to land between 6 and 9 a.m., when heat pumps and furnaces are working against the coldest air of the day. Solar output at that hour in January is close to nothing. Solar output is low before sunrise, when morning demand can be rising.

So the resource Duke is adding is weakest exactly when the system is tightest. The evening peak, around 5 to 8 p.m., is friendlier: panels have charged the battery through the afternoon, and the battery covers the ramp as the sun drops. That is a real contribution. It is also the easier half of the problem.

What Duke has actually built

One example is the Hot Springs microgrid in Madison County, North Carolina. The draft reports that its roughly 2 MW solar-and-battery system operated independently during Hurricane Helene in September 2024. That is a local resilience example, not evidence that batteries can supply a utility-scale winter peak.

The larger numbers are elsewhere. Duke Energy Carolinas and Duke Energy Progress have been adding substation batteries in the tens of megawatts, and the company's carbon plan filings under North Carolina's 2021 law, which requires a 70% cut in carbon emissions from 2005 levels by 2030, pair batteries with new gas. Solar-plus-storage is also appearing more often in new US projects, as discussed in this review of recent installations.

The four-hour limit

Take a 50 MW battery with four hours of duration. It holds 200 megawatt-hours, enough to hold a 50 MW slice of the morning ramp for four hours before it is empty. Charging it costs energy: every megawatt-hour that goes in and comes back out is roughly 10% to 15% smaller. Storage moves power around inside a system that still needs enough of it.

Four-hour batteries can fit a daily evening-peak duty. Their relative cost depends on market, financing, fuel prices and how often the plant runs; the comparison is discussed in this analysis of batteries and gas peakers.

A peaker comparison is not an adequacy study. A cold snap is a three-day event, and no four-hour battery carries a system through day three. After the first discharge, the battery is a load, waiting for the next charge it may not get.

Facts, inference, and one forecast

The facts are narrow and checkable: Duke ordered rolling blackouts during Elliott; the Hot Springs microgrid carried its customers through Helene; North Carolina law sets a 70% reduction target for 2030; batteries are now a routine component of new solar builds.

Inference: batteries can help with short peaks and local grid constraints, but a four-hour unit cannot by itself cover a prolonged cold spell. The size of that contribution depends on when it can charge and how much capacity is available at the constrained location.

A forecast worth writing down: Duke and its peers will keep describing batteries as winter reliability assets while adding gas combustion turbines for the cold snap. That forecast would be borne out if future utility plans continue pairing new gas capacity with batteries and demand response, while assigning batteries a limited role in multi-day winter events.

Why gas still sits underneath

Elliott's lesson was less about megawatts than fuel. Duke pointed to generation outages and constrained gas supply, and the post-mortems across the Eastern Interconnection focused on plants that could not get fuel rather than plants that did not exist. PJM spent the following winters hardening that flank, a story covered in the longer fight over PJM's capacity crunch. Batteries do not solve a pipeline constraint.

What they do solve is local: a substation that cannot take another peak afternoon, a feeder that trips under load, a town at the end of a line. Those are distribution problems, and a 10 MW battery behind a constrained transformer fixes them faster than reconductoring does.

What to watch at 7 a.m. in January

Three indicators can test the argument over the next few planning cycles: Duke’s published winter capacity credit for batteries; whether planned projects exceed four hours of duration; and measured morning demand-response performance during cold weather.

The claim would weaken if Duke demonstrates that its storage fleet can sustain dependable winter capacity through a multi-day cold event, rather than only shifting energy across a few hours.