Wind generation and zonal-market price divergence: Evidence from Texas

The extant literature on wind generation and wholesale electricity spot prices says little about how wind generation may affect any price differences between two inter-connected sub-markets. Using extensive data from the four ERCOT zones of Texas, this paper develops a two-stage model to attack the issue. The first stage is an ordered-logit regression to identify and quantify, for example, the impact of wind generation in the West zone on the estimated probability of a positive or negative price difference between the North and West zones. The second stage is a log-linear regression model that identifies and quantifies the estimated impact of wind generation on the sizes of those positive and negative price differences. It is shown that high wind generation and low load in the wind-rich ERCOT West zone tend to lead to congestion and zonal price differences, that those differences are time-dependent, and that other factors such as movements in nuclear generation and natural-gas prices, as well as fluctuating non-West zone loads, also play a role. The results have broad implications for energy policy makers that extend well beyond the borders of Texas and, indeed, those of the United States.