Sparse Network Estimation in Nodal Pricing: Propagation Multipliers and the Incidence of Battery Storage
Abstract
We develop an econometric framework to recover a price-propagation network from high-frequency nodal electricity prices and use it to quantify the system-wide incidence of battery storage. Modeling locational marginal prices as a simultaneous system with an unknown sparse interaction matrix, we estimate marginal congestion linkages across nodes via adaptive elastic-net GMM and construct propagation multipliers mapping local shocks into network-wide price responses. Combining these multipliers with a dynamic storage adoption-and-operation model yields counterfactual system-wide impacts of incremental storage on prices, volatility, and welfare, supporting budget-constrained targeting of storage siting. The analysis clarifies identification limits when transmission constraints are always binding and the local, marginal nature of the recovered network.