Tesla’s launch of its new residential battery system, the Powerwall, represents a major step toward affordable distributed energy storage. In promoting the Powerwall, Telsa has emphasized its ability to time-shift solar generation to hours of peak demand. However, it is not entirely clear how customers would benefit from this time-shifting given that net-metering already allows them, in effect, to use the electricity grid as a battery. For time-of-use rates with a large delta between peak and off-peak hours, energy storage could prove to be viable, but the vast majority of residential utility customers are not on such rate structures and many utilities do not even provide this option.
The benefits of residential energy storage appear to be limited under net-metering, but these storage systems may be well positioned to dominate markets in which net-metering is limited or nonexistent. Several U.S. states, such as Tennessee, Alabama and Mississippi, as well as large parts of Texas, do not currently allow net-metering. But perhaps more importantly, net-metering has increasingly come under attack from utilities in many states where the solar industry has historically been successful. In parts of Hawaii, high solar penetrations have led utilities to halt or significantly delay new interconnections. In some cases, this is leading customers to install non-exporting systems, which do not send excess generation to the grid. APS, the largest utility in Arizona, has pushed the state’s Corporation Commission to impose large fees on net-metered customers, and while this has so far been largely unsuccessful, the battle is sure to be fought again.
In markets where net-metering is not allowed or is heavily penalized, cheap energy storage may provide an alternative by allowing solar customers to store their excess generation rather than feeding it into the grid. A year ago, I authored a White Paper titled Residential PV-Storage System Optimization Under Self-Consumption, which investigated the very question of whether residential energy storage can provide an alternative to net-metering. In order to examine the economic viability of such a setup, I developed a nonlinear optimization model that determines the optimal sizing and operation of a residential PV + lithium-ion battery system under a variety of cost assumptions and utility rate structures.
Based on this model, it does not appear that the Powerwall, with costs of around $350/kWh can serve as a replacement for net-metering in most locations. However, in markets like Hawaii, where electricity costs are extremely high and net-metering is already being curtailed, these batteries are already looking attractive for solar customers. And with battery prices projected to reach $200/kWh by 2020 and to continue declining for the foreseeable future, much of the U.S. is likely to reach this same threshold in the next decade.
Beyond these high level findings, I found that the economics of solar+storage are highly dependant on how utilities structure their rates, and that this relationship is not always intuitive. For instance, time-of-use rates that charge more for electricity during the day than at night can lead to relatively small optimal system sizes and a trade-off between a larger PV system and a larger battery system. Given the important role that rate structures will play in shaping the future of distributed energy storage, more research is needed by regulators and utilities, academics and renewable energy providers to ensure that distributed storage realizes its full potential to enable a clean and reliable electricity system.
For more detail, please see the full paper: Residential PV-Storage System Optimization Under Self-Consumption by Graham Provost.