Sep 05, 2019 |
Read time: 4 mins
As power grids continue to face an increasing built-out of renewable and intermittent power sources, such as wind and solar, while large scale energy storage systems are still lacking, the impact on power prices is evident.
A good example are the states of Texas (US) or South Australia (AU). Both states have faced a continuous build-out of wind energy in recent years. Understanding the impact of such build-out of intermittent power sources on power prices is important for ensuring and understanding the profitability of your investment in renewable energy. Especially as renewable penetration reaches new heights in more and more markets and subsidized tariffs are replaced by merchant operations.
A visible consequence of the impact which intermittent generation capacity has on prices is the “duck-curve” effect imposed by the generation pattern of wind and solar on power prices. The effect is evident in the power prices around the hour 08 to 16, with peaks observed in the hours around 07 and 17.
The duck curve reflects both the impact of the cyclical pattern of the resources available for renewable generation (wind and sun), as well as the load in the system. Understanding the wider impact of the intermittent wind and sun generation on the grid, and thereby power pricing, is obviously important for projects seeking to operate fully or partially merchant, as well as the downside risk imposed by financial products, such as bank hedges for wind projects imposing fixed delivery commitments.
A correlation between wind speed and power prices
In addition to depicting prices through the day, the above figures suggest a correlation between the wind speed and the power prices which can be realized when operating in merchant markets with significant renewable build-out. We investigated this correlation using satellite data and local wind measurements correlated to a hub height usual for the two states investigated.
We looked at data for the last five years, and the results are clear. For ERCOT in Texas, the correlation is 76% and negative, with an R-squared of 0.6. In other words, when the wind is blowing and your turbines are spinning with a high production, the prices would be low, and vice versa. This correlation is depicted below in Figure 3, which shows the correlation between windspeed and price for the two markets, ERCOT in Texas (US) and the zonal price for South Australia (AU).
However, more importantly, when we look at what has happened in the last 5 years, it is clear that the correlation between wind speed and prices has only picked up. In statistical terms, in 2014 wind speed explained 48% of the impact of power prices in Texas. In 2018 this had increased to an astonishing 91%. Meaning that the wind speed explains 91% of the difference in observed power prices. This has huge implication for investors operating under merchant conditions, or if you are imposing fixed delivery commitments on your project.
The table describes the mean and standard deviation of the price and windspeed for the last five years. Its evident for the increasing correlation between windspeed and electricity market price in Texas.
Designing your wind farm for higher revenue
Firstly, your investment case needs to consider the impact of such significant price intermittency. Secondly, it leaves ample room for optimizing your investment case’s wind farm layout, not only for higher Annual Energy Production (AEP), but also for higher realized prices and revenue. In simple terms, the further it is possible to push the effect park power curve to the left and thereby upwards the price curve, the higher the captured merchant prices, as illustrated in the figure below.
This requires a different approach and thinking around turbine selection and micro-siting. However, the potential for optimizing your wind farm design and power curve selection based on the price pattern in the market is clear. We already shed some light on this potential in our analysis of the impact of the 3.0 MW V138 EnVentus power curve on realized power prices.
To further demonstrate this, we repeated our analysis of turbines from three of the world leading OEMs, Siemens Gamesa Renewable Energy (G), General Electric (GE), and Vestas (V). By analysing turbines in the 3-5 MW range, we wanted to test how different the captured price would be for each turbine variant. The two following graphs depict the intermittency for each turbine variant from 2010-2018. The lower the intermittency, the closer the power price realized by the specific turbine type is to the average power price in the market. The box plot depicts the spread from year to year, while the black line indicates the average intermittency across the period. Showing the advantage of a high rotor to generator ratio in not only reducing intermittency to increase merchant revenue, but also reducing the spread in intermittency from year to year, and thereby contributing to reducing the variability of the merchant revenue.