In a winter season marked by weak wholesale power prices resulting from mild temperatures, low natural gas pricing and record wind generation, the variability of wind generation is a significant driver of wholesale price volatility. As recently as Tuesday, February 23, 2016, a miss in ERCOT’s wind forecast resulted in five-minute North Hub prices spiking over $1,000 during the evening demand peak.
As new wind generation continues to assume a greater percentage of the generation supply stack there is increased opportunity for market participants to capture this volatility through Genscape’s superior forecasting and fundamental supply stack analysis.
Wind Generation in ERCOT and the PTC
Late 2015 saw a number of new wind farms added to ERCOT, ahead of the expiry of the renewables Production Tax Credit (PTC) at the end of the year. The PTC was extended in December for another five years, galvanizing wind projects already in the development pipeline for the next few years. Under the PTC, utility-scale renewable power resources are able to claim a tax credit of $23 per megawatt hour for the production of electricity from wind, solar, and other renewable means. In a wholesale power market, this enables renewable power production to remain profitable even in normally uneconomic situations such as negative price dispatch. This tax credit coupled with the meteorological conditions in West Texas has had a substantial impact on the ERCOT generation stack.
The growth in wind generation as a means of power production within ERCOT over the past 15 years can be observed below in Figure 1. This graphic shows the annual proportions by fuel-type of new generation interconnections as measured in MWH capacity according to data from the most recent ERCOT Seasonal Assessment of Resource Adequacy (SARA). As evidenced in the figure below, wind generation (in blue) has comprised a growing percentage of new generation interconnections on an annual basis over the last 15 years. This growth has been particularly notable since 2012 with wind generation comprising the majority of new generation interconnections in nearly all of those instances.
ERCOT Wind Variability and Impact on Power Prices
Over the past several months, days with the highest North Hub locational marginal prices (LMPs) were associated with errors in the wind forecast. More specifically, when wind generation verified weaker than both the day ahead and real-time forecast, correlations were noted with higher prices. Figure 2 illustrates such an example with actual wind generation verifying 13.25 percent weaker than ERCOT’s bal-day forecast for hour ending (HE) 18. This helped buoy North Hub LMPs to over $200 despite actual load verifying weaker than forecast by 1.89 percent (Figure 2). On this particular day, high pressure quickly built into the footprint from the West, causing wind across West Texas (where the bulk of ERCOT wind farms are located) to abruptly weaken. This high pressure system worked in tandem with the typical afternoon dissipation of the LLJ to yield a 7 GW drop in wind only 12 hours. Interestingly, the significant over performance of wind generation over the early afternoon hours drove very weak real-time pricing, thinning out the supply stack and making the grid very sensitive to supply and demand inefficiencies over the evening hours.
Another situation involving both wind forecast errors and sharp increases in North Hub’s LMPs occurred on January 8, 2016. In this particular example, wind generation verified over 3,000MW below ERCOT’s day-ahead forecast (Figure 3 below). The resulting under-commitment of generation resources led to sharp increases in North Hub LMPs through the morning demand peak. While not directly related to volatility of the low-level jet, the cause of this major forecast blunder was the result of errors in the predicted location of low pressure near the panhandle of Texas. Despite only a narrow miss in the predicted location of this area of low pressure, the tight clustering of wind farms in western Texas makes wind generation very sensitive to spatial forecast errors. As this region will continue to be the focal point for new wind farm projects, the sensitivity of wind generation to spatial forecast errors will continue to increase.
Pricing Impacts of Increased Wind Generation
As evidenced in the example from January 8, 2015, the necessity to build out wind generation facilities in locations that favor strong and consistent wind patterns has led to a tight clustering of wind generation facilities in West Texas. As a consequence of this tight grouping of wind generation facilities, tight bands of wind that deviate from forecasted patterns by as little as 50 miles can lead to large misses in the wind generation forecast. Despite the obvious advantages of harnessing renewable resources for power generation, there are inherent disadvantages when actual wind generation strongly deviates from forecasted levels. In a review of the 500 strongest North Hub real-time power prices in ERCOT since 2013, it was found that roughly 80 percent of those instances occurred on days when wind generation verified below ERCOT’s short-term wind power forecast, or STWPF (see Figure 4).
Not surprisingly, this robust buildout of wind generation has had obvious impacts on wind generation within ERCOT, the first of which has been a large increase in the monthly average of hourly wind generation between May 2009 and December 2015. The increase in the monthly average of hourly wind generation over this period exhibits a strong positive trend of nearly 4 GW. Mirroring that trend is an increase of the standard deviation in the monthly average of hourly wind generation. This trend indicates that as wind generation totals have increased, the average variance in hourly wind generation has also increased. This variance in hourly wind generation can and does drive significant deviations in wind generation from one day, or even hour, to the next and has had notable impacts on the ERCOT market.
The cumulative implication of these trends is that as we can expect to see continued growth in wind generation facilities in the ERCOT footprint, we can also expect to see increased volatility in wind generation output. This wind volatility will have a strong impact on both bullish and bearish pricing events within the ERCOT system.
Genscape’s team of meteorologists enhances Genscape's Power Market services by providing on-demand, up-to-date forecasts and a unique perspective on how weather conditions impact electricity demand, with accurate load forecasts up to 15-days out and analysis of weather conditions that drive demand for the major power markets in the contiguous United States.
Genscape has the most complete, real-time monitoring of the U.S. power grid, with accurate and timely data on capacities, flow, and utilization. Genscape's Power IQ Market Intelligence Service provides access to market-specific analysis and price forecasts, allowing for more informed trading decisions.
Wind is one of the fastest growing and most difficult to predict forms of electric power generation. It is also a significant factor in driving congestion and price volatility in many regional markets. Now power traders in key wind regions can receive the most comprehensive and skillful forecasts thanks to WSI and Genscape's market intelligence unit. WSI WindCast IQ helps traders gain a clearer understanding of the expected wind power generation within an ISO. Click here to learn more or request a free trial of Genscape's Power Market Services.