This paper explores various questions regarding very high penetration of intermittent renewable to electricity grid. The study was performed using one-year hourly demand data of California's electricity grid together with the hourly-simulated output of various solar and wind technologies distributed throughout the state. The result shows that wind-solar complementarities carry significant multidimensional benefits to the future grid as compared to a stand-alone wind/solar based grid. Specifically at 20% total energy loss, it was shown that their optimal complementarities lead to very high renewable penetration with smaller storage and backup requirement. Our study of system dispatch shows that storage provides flexible dispatch strategy, which makes it applicable for various services depending on season of the year including reduction of summer backup capacity need. We have also found that the existing peak load based long term planning reserve criteria may no longer be applicable to very high renewable grid. Finally, comparing our result to literature data, we conclude that some of the pessimistic views of storage technology are borne out of our lack of understanding about its role and design requirement.

We present a result of hourly simulation performed using hourly load data and the corresponding simulated output of wind and solar technologies distributed throughout the state of California. We examined how we could achieve very high-energy penetration from intermittent renewable system into the electricity grid. This study shows that the maximum threshold for the storage need is significantly less than the daily average demand. In the present study, we found that the approximate network energy storage is of the order of 186. GW. h/22. GW (approximately 22% of the average daily demands of California). Allowing energy dumping was shown to increase storage use, and by that way, increases grid penetration and reduces the required backup conventional capacity requirements. Using the 186. GW. h/22. GW storage and at 20% total energy loss, grid penetration was increased to approximately 85% of the annual demand of the year while also reducing the conventional backup capacity requirement to 35. GW. This capacity was sufficient to supply the year round hourly demand, including 59 GW peak demand, plus a distribution loss of about 5.3%. We conclude that designing an efficient and least cost grid may require the capability to capture diverse physical and operational policy scenarios of the future grid. © 2014 Elsevier Ltd.