The role of large-scale energy storage design and dispatch in the power grid: A study of very high grid penetration of variable renewable resources

Highlights

• Approximates the maximum threshold of the required storage system size.

• Examines backup capacity requirement corresponding to a given storage size.

• Compare the role of transmission increase to energy storage on high penetration.

• Show how energy dumping reduces backup needs via increased use of storage.

• Describe important factors to design a least cost large storage renewable grid.

Abstract

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.

Introduction

The existing grid is not yet optimized to accommodate very large variable renewable energy systems. Due to its ability to produce low carbon electricity, integrating variable generators to electricity grid has attracted significant worldwide research attention. The dominant question of interest is as regards to the ability of the existing grid to accommodate their variable output as we increase system size [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17]. In the low to high penetration of energy from these resources, such approach could help us understand the technical and economic value of these technologies [1], [2], [3], [4], [5], [6], [7], [8], [9], [10]. However, very high penetration will most likely require the capability to enhance the use of energy from the variable technologies. These have three important aspects. The first is regarding the possibility to achieve an optimal temporal match between the variable generators output to the demand profile. The second one relates to a set of technological requirements that enable these optimal matching capabilities while providing sufficient capacity to meet the demand at any time of the year. The third is the possible operational requirements to optimize the use of these resources in order to achieve carbon reduction.

Multiple studies have shown that reduced carbon emission could be achieved through an increased use of intermittent renewable energy sources [1], [2]. For the reported low to high penetration, it was shown that switching from the less flexible coal firing generators to a more flexible gas firing technologies could help the grid to handle the fluctuating output of the variable generators. However, generating very significant electric energy from intermittent renewable resources would require significant use of energy storage technologies [11], [12], [13], [14], [15], [16], [17]. Little is known about the nature of storage need in an interconnected grid and factors that can limit/enhance its potential benefit in increasing grid penetration of intermittent renewable energy resources.

Very high grid-penetration of variable generators with large-storage have been a subject of a few studies. Studies by Denholm and Margolis [11] have shown the possibility of about 70% PV penetration to the ERCOT – Texas grid. An independent series of reports by Solomon et al. [12], [13], [14], [15], [16], [17] have shown PV penetration of up to 90% of the annual demand to Israeli-grid using energy storage and by allowing 20% total energy loss. In these reports, the energy storage capacities were lower than the daily average demand. Unlike the former, the later study employed a computational algorithm that can calculate a storage design requirement based on the seasonal and diurnal profile of the electricity demand. This was one of the reasons for the reported very high penetration in the later case. These series of reports identified two most important information about grid mainly fed by PV-storage system.

The first one is that designing proper storage is a significant part of achieving very high penetration, and that the design should be based on seasonal and diurnal interaction of PV output and the demand profile to be met by PV [12]. Second an employment of proper grid operation strategy could significantly reduce the existing grid’s conventional capacity requirement and grid operation cost [17]. Based on the data from the year 2006 if appropriate PV-storage grid were built, the total conventional generator capacity required would have been at least 3 GW less than the 10.5 GW capacity operated that year. Technology wise, large coal power plants were unnecessary but units that serve for intermediate and peak demand times are generally needed. Moreover, as the consequence of the above findings, the economic performance analysis of storage should incorporate the engineering aspect of storage design and use [12], [16], [17].

In the present study, we investigate the role of energy storage to increase grid penetration of intermittent renewable systems in an interconnected grid. Furthermore, this paper will discuss the value of storage design and dispatch, the corresponding conventional backup and operational requirements, etcetera. In the following sections, we present brief description of our methodology followed by a detailed presentation of the main results. In the end, we will give the summary of the result and our overall conclusion.