UC Berkeley

Electric power systems are currently facing substantial changes in their operating frameworks on a wide variety of fronts. With increasing deployment of grid-connected wind and solar generators, power system operators are concerned that the challenge of following variable generator outputs may degrade system reliability. Additionally, as temperature changes associated with climate change increase, the magnitudes and frequencies of peak loads will correspondingly grow, thus compounding the variability problems introduced by intermittent generation. The fossil fuels that are currently used to run most of the system are changing in cost as new methods are developed to extract them, even as new policies are being proposed that restrict or tax carbon emissions. Additionally, much of the current electricity system infrastructure is nearing the end of its life, and will soon need to be decommissioned or replaced.

Energy storage has been proposed in several different venues as a solution to many of these problems. If energy storage can be deployed appropriately, it is possible that variability could be reduced, fossil fuel dependency could be lessened, and additional investments in new plants to handle increased demand could be avoided. Though these benefits could be large, the benefits to adding storage to the current grid have not been fully characterized.

We seek to better characterize the potential for storage to change overall grid operations. To do this, we use a 240-bus model based on the US Western Interconnection. In this model, we first optimally locate storage devices in the network and then dispatch them along with conventional power generation units by using a unit commitment model with DC load flow. Storage in the model can provide frequency regulation, load following, and arbitrage for each hour of a study year, and we investigate a range of scenarios for fuel price and renewables penetration. We use this model to investigate the demand curves for added storage and added TCL aggregations that function as thermal storage, as well as the extent to which carbon taxes may have an effect on the overall benefits that storage may provide. We find that storage and demand response are most valuable operationally and economically when they are providing high power services like frequency regulation and demand response. We also show that the relationship between carbon taxes and the benefits that energy storage resources could provide is not linear.