| Summary: | Accurate estimation of the balance of costs required to deliver a turnkey battery energy storage system (BESS) is highly important for decision making on optimal battery type and sizing for a given project. Such costs can be greater than the direct current (DC) boundary module costs for short duration systems. Despite this, very little data is publicly available for such costs at the commercial/industrial scale, and data decomposed into $/kW, $/kWh and fixed components is even scarcer. In the present work, cost models reported by PNNL and Lazard are tested against cost instances reported by EPRI. As system duration increases the Lazard model increasingly overpredicts balance of costs, whereas the PNNL model increasingly underpredicts them. This is due to the former placing more costs in the $/kWh category. It is shown that averaging the PNNL and Lazard models gives better agreement with the EPRI data than either does alone (10% discrepancy compared to 36/30%). The disagreement in cost scaling requires further attention, as it currently hinders the study of longer duration systems.
|
|---|