The Benefits of Dynamically Resizing Residential Storage.

Today's solar and storage systems are typically sized at the time of deployment to match historical loads. However, loads do change over time, which may make one-shot sizing sub-optimal. Modular storage systems allow storage capacity to change over time to better match changing loads. We investigate how a modular battery system, in combination with a solar photovoltaic (PV) in a grid-connected system, can be dynamically resized to reduce overall costs. We develop an algorithm for dynamic resizing and evaluate it in multiple scenarios using both real and synthetic traces of solar and load. We find that our algorithm is able to closely approximate the offline optimal algorithm. Moreover, in our case study, we find that overall costs are not significantly reduced due to resizing of storage, and this conclusion is insensitive to types of load changes and the choice of resizing interval.