Researchers develop framework for national battery storage planning

A research team led by scientists from the University of Cambridge in the United Kingdom has developed a new framework to identify the optimal locations and capacities of grid-scale energy storage on national transmission networks with high shares of renewable energy. They demonstrated the operation of the system using the projected 2030 electricity system of the Czech Republic as a case study.

“The main novelty is the introduction of an algorithm that makes the national-level battery siting and sizing problem computationally tractable while preserving the full physics of the AC optimal power flow problem,” researcher Amir H. Keshavarzzadeh told ESS News. “The study demonstrates how optimally installing energy storage across the transmission grid, while considering the entire network as an integrated system, can reduce electricity generation costs.”

The proposed framework consists of two optimization layers. The outer layer uses the ε-NSGA-II evolutionary algorithm, which generates candidate battery portfolios by selecting whether batteries should be installed at each transmission bus and determining their optimal capacities. Each candidate solution is then passed to an inner layer based on alternating-current multi-period optimal power flow (AC-MPOPF), which simulates hourly grid operation over an entire year.

The simulation models conventional power plants, hydropower, solar and wind generation, battery charging and discharging, transmission constraints, voltage and reactive power, and hourly demand across the network. It also accounts for lithium iron phosphate (LiFePO₄) battery degradation, fuel and carbon costs, and transmission losses. The optimizer evaluates thousands of possible battery configurations to find those that reduce overall electricity costs while keeping the grid operating reliably.

The researchers selected the Czech Republic as a case study as it represents a coal-dependent system undergoing an energy transition. By 2030, they assumed the country would have 4 GW of solar capacity, 1 GW of wind capacity, and 20% electric vehicle penetration, while coal-fired power plants would still supply a significant share.

“In the Czech Republic case study, where power generation is dominated by coal, the optimal deployment of energy storage reduced generation costs by up to 30% and lignite-based generation by 23%,” Keshavarzzadeh said. “These results demonstrate that the optimal deployment of energy storage can have a considerable impact on both system costs and the generation mix.”

Keshavarzzadeh added that his team is considering a few follow-up research studies.

“Potential extensions include considering different regional electricity-pricing schemes when deploying energy storage at the national level and different renewable penetration scenarios, as well as conducting power-system security studies,” he said. “More broadly, the algorithm could be adapted to make other large-scale mixed-integer nonlinear optimization problems computationally tractable.”

The framework is described in “High-resolution siting and sizing of grid-scale energy storage for real-world transmission networks under high renewable penetration: a two-stage metaheuristic approach,” published in Applied Energy. Scientists from the University of Cambridge, University College London and the Alan Turing Institute contributed to the study.