Duke's 2026 Carolinas Resource Plan: 18.5 GW of Solar, 14 GW of Gas, and a Coal Retirement Schedule That Just Slipped Again

Duke Energy filed its 2026 Carolinas Resource Plan with South Carolina regulators on August 14, and the headline number is large enough to demand a closer look: 18.5 GW of new solar, 14 GW of gas, 13 GW of energy storage, and roughly 4.5 GW of new nuclear - all targeted by 2041^[1]. That is more than 50 GW of new capacity layered onto a system that currently serves about 20,800 MW of load across North Carolina and South Carolina. The plan is ambitious. It is also, in several important respects, a plan that has not yet been built.

Here is what has financing, a permit, or a signed supply agreement - and what is still a line in a spreadsheet.

The Demand Case Is Real, Even If the Exact Numbers Are Contested

Duke's justification for the scale of this buildout starts with load growth, and the growth is genuine. Both North Carolina and South Carolina rank among the fastest-growing states in the nation, and Duke told South Carolina regulators that its territory "continues to experience significant growth."^[1] The drivers are economic development, population inflows, and accelerating electrification - with data centers and AI infrastructure cited explicitly as major contributors.

Duke forecasts nearly 8 GW of large-load demand growth in the Carolinas. CEO Harry Sideris has said the company is deploying more than $1 billion per month to meet what it calls record demand. Advocates in North Carolina have pushed back, arguing Duke has overstated growth projections to justify its capital program - a tension that will play out in the NCUC proceeding, where an order is expected by December 31, 2026.

The South Carolina proceeding moves on a different timeline: a PSCSC hearing is scheduled for April 2027, with an order expected by June 2027.

What Is Actually Being Built

The solar target - 18.5 GW over 15 years - is the headline, but it is a procurement goal, not a portfolio of signed projects. Duke has not disclosed PPAs or interconnection agreements for anything close to that volume. The plan calls for continued competitive solar procurements, with locations to be determined through annual bidding processes.

The gas side is further along. Duke has received regulatory approval for a 1,400 MW combined cycle facility in Anderson County, South Carolina, and has executed turbine supply agreements with GE Vernova.^[1] That plant has a permit and a turbine contract - two of the three things that make a project real. Construction is anticipated to begin in summer 2027, with the facility expected to serve customers by early 2031. The gas split across the full plan runs 8.2 GW of combined cycle and 5.8 GW of combustion turbines through 2041.^[1]

On storage, Duke is scaling execution: projects are in service, equipment is secured, interconnection activity is underway, and the company has issued an RFP for 400 MW of standalone storage in South Carolina. The 13 GW storage target is a 15-year number; the near-term pipeline is a fraction of that.

The Nuclear Piece: Optionality, Not Commitment

The plan calls for about 4.5 GW of new nuclear, which Duke describes as "an important resource to support long-term reliability and system needs."^[1] Duke is pursuing license renewals for its existing fleet - it has already received subsequent renewed operating licenses for two of its six nuclear sites, both in South Carolina, covering four of the fleet's 11 units.^[1] The company is also evaluating potential new nuclear options at sites in Cherokee County, South Carolina, and Stokes County, North Carolina.

The new-build nuclear target carries a 2037-2040s in-service window. That is optionality, not a construction commitment. No NRC license application has been filed for a new large reactor at either site. SMR technology at the scale Duke is contemplating remains pre-commercial at most vendors. The nuclear line in this plan is a placeholder for a decision that will be made in the early 2030s, contingent on technology readiness, regulatory timelines, and what the gas and solar buildout actually delivers.

The Coal Retirement Delay Is the Uncomfortable Subtext

The 2026 plan delays coal retirements that were previously approved. Duke had aimed to retire all of its remaining coal plants by 2035; the new plan extends several facilities into the late 2030s to "provide flexibility" amid near-term demand growth. The 2.24 GW Belews Creek facility in Stokes County, North Carolina, could continue operations until 2040.

Duke's position is that regulators have made clear that replacement generation must be online before additional coal units can retire. That is a defensible sequencing argument - but it also means the coal retirement schedule is now contingent on the gas buildout completing on time, which in turn depends on turbine delivery, permitting, and interconnection. Each of those has its own queue.

North Carolina state law originally required Duke to reduce carbon emissions 70% by 2030 and achieve carbon neutrality by 2050. A new state law passed in 2025 eliminated the 2030 goal, removing the near-term regulatory pressure that had been the primary forcing function on coal retirement timing.

What the Plan Actually Tells Grid Planners

Three things stand out for anyone tracking the Carolinas grid.

Gas is the near-term answer, and it has hardware behind it. The Anderson County combined cycle plant has a PSCSC permit and a GE Vernova turbine supply agreement. That is more than most of the solar and storage in this plan can claim right now. Gas is being positioned as the bridge, and the bridge has a construction start date.

Solar at 18.5 GW is a procurement ambition, not a project list. Duke's annual competitive bidding process will determine what actually gets built. The pace of interconnection approvals, transmission upgrades, and land permitting will set the real ceiling. The 2025 plan targeted 4,000 MW of solar by 2034; the 2026 plan's 18.5 GW by 2041 is a significant step up in ambition. Whether the execution infrastructure - interconnection queue, EPC capacity, supply chain - can support that pace is the question the plan does not answer.

The coal retirement delay is a sequencing problem, not a reversal. Duke is not adding coal capacity. It is extending existing units while replacement generation is built. The risk is that the replacement generation slips, and the coal extension becomes the default rather than the bridge. That is worth watching in the annual plan updates.

Duke recently completed a $1.75 billion equity offering to support its capital needs amid rising demand, signaling that the financial markets are treating the load growth story as credible. The harder question - whether 18.5 GW of solar and 13 GW of storage can be sited, permitted, interconnected, and financed across two states in 15 years - will be answered one procurement cycle at a time.