Mitigating the Variability of Wind and Solar Energy Through Pumped Hydroelectric Storage

Abstract

Renewable power production is both variable and difficult to forecast accurately. These facts can make its integration into an electric grid problematic. If an area’s demand for electricity can be met without using renewable generation, the addition of renewable generation would not warrant a further increase in generation capacity. However, to effectively integrate large amounts of additional renewable generation, it is likely that a more flexible generation fleet will be required. One way of increasing a generation fleet’s flexibility is through the adoption of pumped hydroelectric storage (PHS, see the glossary for definitions of select terms). Like traditional hydropower generation, PHS is capable of quickly varying its power output but it is also capable of operating in reverse to store excess energy for later use. This paper will address many of the operational aspects of combining pumped hydroelectric storage (PHS), which is currently used to store excess energy from traditional generators, with wind and solar power generation. PJM, a grid operator in the Middle Atlantic States, defines capacity value for renewable generation as the percent of installed generating capacity that the generator can reliably contribute during summer peak hours. Existing wind generators inside PJM have an average capacity value of 13% and existing solar generators have a capacity value of 38%. The chief reason for these capacity values is that the renewable power production does not usually coincide with the hours of peak electricity demand during the summer. If PHS were used to firm renewable power generation, it would translate into increased utilization of the renewable generation that would displace the least efficient/most costly generators. A computer model with one minute granularity is constructed in order to study the operational requirements of PHS facilities. PJM electricity demand, power prices, and wind power production data for 2010 were used in conjunction with NREL simulated solar power production as input to the model. Currently, various PHS operational strategies are being tested to ascertain their effectiveness at firming and time shifting renewable generation. Preliminary results show the profound effects of increased penetration of renewable energy on an electric grid. The results also demonstrate a niche for even greater PHS operational flexibility, i.e. variable speed or unidirectional ternary machine (UTM) PHS.