Do We Still Need Demand Flexibility as Batteries Become Cheaper? A Levelized Cost Perspective

Abstract

Energy storage and flexible loads both help balance power systems with high shares of variable renewables, but they do so differently. A battery moves electricity from one period to another. A factory or data center instead moves production or computing activity, while meeting demand for its product or service. As batteries become cheaper, it is natural to ask whether demand flexibility is still needed. Many assessments treat that controllable load as already available and focus on enabling or participation costs. For a large load that normally runs near full capacity, however, shifting electricity use can require extra equipment, delay sales, tie up working capital, and disrupt the process. Here, we compare energy storage technologies with industrial and data-center load shifting on a levelized cost basis, measuring the incremental cost of flexible relative to baseline operation. We find that energy storage becomes more expensive per shifted kilowatt-hour as the flexibility period lengthens and assets are cycled less often. Load shifting does not share that rise in capacity cost under a fixed avoided-hour share, but new productive capacity can still be too expensive to justify. Under representative central technology costs, batteries compete well for short-cycle shifting but are more than 80 times more expensive than load shifting of aluminum smelting when providing seasonal flexibility. The answer therefore depends on the timescale, the load, and whether the needed capacity is new or already sunk.