‘Cost-optimal’ solutions don’t always provide best mix for power generation, study finds

As industries, utilities and regulators consider the best ways to accommodate our increasing need for power generation, cost concerns weigh heavily on their decision-making.

New research, however, shows that choosing the least expensive option isn’t always the best solution, and even a little wiggle room on cost can provide a much more socially, environmentally and politically coherent outcome.

A recently published paper in the journal Joule relies on research from Binghamton University Assistant Professor Neha Patankar, who uses a technique called modeling to generate alternatives (MGA) to systematically map out economically and technically viable planning strategies and their trade-offs.

Collaborators on the new paper include researchers from the Technical University of Delft (the Netherlands), UiT—the Arctic University of Norway, Technical University of Denmark, Princeton University, Technische Universität Berlin and the University of Oslo.

With the growing array of electricity generation options—from fossil fuels and nuclear to solar, wind and hydropower—determining the right mix of technologies is a complex challenge that goes well beyond simply picking the lowest-cost option.

“Even a small relaxation of total system cost, as little as 2%, can lead to radically different technology portfolios for meeting growing electricity demand,” said Patankar, a faculty member at the Thomas J. Watson College of Engineering and Applied Science’s School of Systems Science and Industrial Engineering. “It highlights how the so-called ‘cost-optimal’ solution is highly sensitive to uncertain assumptions and may provide only a false sense of certainty.”

The strictest price-conscious models driven by artificial intelligence and machine learning don’t make recommendations based on variables that may be more important in the long term, such as ecological, social, political or environmental effects.

“Using MGA to show options that are near-optimal cost can reveal strategies that align with unmodeled objectives such as social viability, resilience to sudden supply disruptions or hedging against policy shifts,” Patankar said. “Stakeholders can see practically viable consensus solutions hidden by the insistence on cost optimality.”

As climate change accelerates the shift toward renewable energy, researchers like Patankar and her collaborators are working to map out effective strategies for navigating the complex tradeoffs of the energy transition.

“Our main conclusion is that MGA is now accessible and versatile enough to become a standard in improving the reliability and usefulness of the analyses shaping urgent energy transition decisions globally,” said Francesco Lombardi, an assistant professor at TU Delft and the lead author of the new paper.

“The many organizations that directly use energy planning models for their strategy can immediately pick up our recommendations to enhance the quality of their analyses and ensure that they deliver reliable, practically viable advice.”