Duke Energy’s Modeling Manufactures Injustice

In an unincorporated community in North Carolina, two smokestacks reflect on a nearby lake, a dimly lit building standing in the background. The lake is surrounded by an idyllic park and hiking trail, sharply contrasting the streams of toxic coal ash leaking into groundwater supply. 

Four months ago, Duke Energy extended the life of the Belews Creek Power Plant for three to four more years, prolonging a painful legacy of cancer, reproductive failure, and neurological harm. In addition to delaying the retirement of coal plants, Duke Energy also has slashed renewable development and increased the construction of fossil fuel facilities. Despite recently proposing a rate hike of 15%, Duke Energy justified their energy plan by modeling cheap prices for consumers and estimating the reliability of renewable and nonrenewable energy types. 

Just like North Carolina’s taxpayers, these estimates are victims of a flawed modeling process, controlled and created by Duke Energy. In the process, millions of Carolinians are deprived of clean energy sources and instead left with mounting energy bills as Duke Energy chooses to align with an increasingly fossil fuel–dependent future. 

Modeling Process

To begin their analysis, Duke Energy tested the fidelity of different energy sources to see their contribution to the power grid. When inputting reliability scores in the modeling software, Duke Energy operated under the assumption that natural gas and coal energy were perfectly reliable. In a shareholder meeting in July, Duke Energy admitted that they gave natural gas and coal a 100% reliability score. In reality, coal and natural gas reliability is deficient - outages caused by Winter Storm Elliott in 2023 impacted more than 300,000 customers across the Carolinas.

Current renewable facilities were also excluded from reliability calculations, although historical coal and natural gas plants were included. Maria Roumpani, an expert in energy planning from Stanford University, testified that the 2023 plan, which used similar modeling and assumptions to the one in 2025, created up to a 20% overestimation in nonrenewable energy capacity. 

Conversely, Duke Energy downplayed the reliability of cleaner energy. Renewable energy capacity scores were calculated separate from energy storage. For solar energy, that means energy is solely generated when the sun is shining. Only battery storage up to 10 hours was considered, although existing technologies can store energy for up to 100 hours. When asked about multi-day storage, Duke Energy argued that long term battery storage was “lower on the readiness level” and inefficient. However, Duke Energy implemented other developing technologies like small modular nuclear reactors (SMRs) in their calculations, which are not on the market yet. Futuristic technologies like SMRs make it appear that the 2050 net zero goal will be fulfilled while delaying the building of green energy sources. 

Solar energy has especially endured this delay. Duke Energy cut solar energy installations roughly in half compared to the already conservative 2023 plan, shifting an annual goal of 1,600 megawatts (MW) of solar energy per year to just 770 MW after 2026. This is because Duke Energy systematically limited the amount of renewable energy possible under its simulations. In its earlier climate plans, Duke Energy confined the amount of solar energy capacity in its modeling to just 2800 megawatts per year after 2032. Duke Energy has the ability to implement more than two times that amount compared to similar companies in California and Texas, according to the 2024 expert testimony of Michael Goggin, the Vice President of Grid Strategies and a Harvard graduate. Duke Energy argued limitations on solar were due to cost constraints as well as temporary outages to power lines during construction. Systematic limits inputted into the modeling software allowed Duke Energy cut solar energy roughly in half compared to the already conservative 2023 plan, shifting an annual goal of 1,600 megawatts (MW) of solar energy per year to just 770 MW after 2026.

To further inflate renewable energy costs, Duke Energy neglected to consider the full impact of tax credit policy. Duke Energy chose to model with the old version of a software called EnCompass, even though tax credits can be more accurately modeled in the new version. When questioned months before releasing the plan, representatives of Duke Energy responded that they were not “able to evaluate the updated version of EnCompass with enough time to incorporate it [in the plan].” Duke Energy’s subtle combination of numerical inputs fundamentally biased decisionmaking, forcing EnCompass to automatically choose higher levels of natural gas and coal. EnCompass had no objective carbon reduction criterion, making it unresponsive to long-term energy goals. 

Broader Implications

The Carolinas are no stranger to the suppression of clean energy technology. In 2024, with the passage of Senate Bill 266, a targeted 70% reduction in carbon dioxide emissions by 2030 was retracted. Environmental groups have long tried to contest Duke Energy’s biennial plans, with four experts testifying just last year about deep methodological flaws. The modeling software Duke uses is extremely time-consuming and expensive, though, so it remains largely inaccessible to nonprofits. 

Some alternate models have been able to use more widely available technologies like Excel, presenting the possibility of investing in wind energy, retiring coal units earlier, and building nuclear and solar power plants earlier in time to reduce emissions 70% by 2032. Nevertheless, a lack of capable technology means these experts’ claims are pushed aside in favor of corporate interests. 

The pattern of blurring modeling with shareholder goals echoes in not only Duke Energy’s 2025 plan, it remains a long established practice. Opaque softwares leave consumers with mounting power bills and an increasingly expensive future. Investing in a more resilient energy grid means protecting the independence of utility planning, whether that looks like more energy storage, nuclear, solar, or wind power, depends on the voices of constituents.