Solar variability and its impact on photovoltaic generation

As the use of solar technologies for power production becomes more prevalent the sudden proliferation of wind and solar power plants has raised critical questions about the stability of the electric grid. Of special concern to balancing authorities, system and grid operators and utilities is the impact of variable generation on load balancing, reliability of the grid and power quality. The primary reason for variable generation from wind farms or solar power plants is the variability in the “fuel” source. Therefore a better understanding of how the “fuel” sources behave will result in better strategies to increase penetration of both wind and solar technologies as the source of power. This paper focuses on solar variability and their impact on ramps in power production from utility scale photovoltaic (PV) power plants. We focus on this technology as PV panels instantaneously respond to variability in solar radiation.

 As solar energy can vary both temporally and spatially it is necessary to make high frequency measurements from multiple fast response sensors. These measurements enable modeling and characterization of ramps in power production from power plants of various sizes. Ultimately an understanding of power production ramps will enable decision makers to make informed strategic decisions related to transmission and distribution.

This paper presents a detailed outline of the problem as well as a synopsis of ongoing work to address the problem. The National Renewable Energy Laboratory in Golden Colorado has deployed a multi-sensor network to measure solar radiation at high temporal and spatial resolution (e.g. Figure 1). A detailed analysis of this dataset will be presented. The analysis will provide an understanding of the time and space scales important for modeling production. This paper will also provide an understanding of the response of different sized PV plants to solar radiation.

Figure 1: Distribution of solar radiation modeled from a 17 sensor network.