This analysis focused on how integrating renewable energy sources affects grid stability and impacts electricity prices for end users. We utilize real-time electricity pricing data from Texas, the U.S., and ERCOT spanning from 2010 to 2023, accessible through their data products page: https://www.ercot.com/mp/data-products/data-product-details?id=NP6-785-ER
Renewable energy sources like solar and wind are known to introduce variability into the electrical grid, primarily due to their dependence on environmental conditions like wind speed(wind farms) and sunlight availability(morning and evening peaks). This variability can decrease grid reliability, particularly during peak demand times in the mornings and evenings.
Let's look at the data and statistics behind it to prove or reject it.
1. First analysis
Statistical Examination:
One common misconception is that annual/monthly average or median electricity prices provide a comprehensive view of grid stability. However, these metrics don't give us a complete picture obscure the real issues that typically occur during brief periods. Problems can happen, Maybe several hours during several months of the year. Hence, relying solely on median or mean electricity prices may be misleading.
We approach this issue statistically by examining the properties of symmetric normal distributions where, ideally, the mean should align closely with the median.
That is how it looks:
In practice, when we have grid instability, electricity prices gradually increase. Peak prices can be, for example, 500 USD/MWh, when the mean price will be 20 USD/MWh. Significant deviations between mean and median price two can indicate underlying instability, as when sudden price spikes create a right-skewed distribution.
The right skew distribution looks like this:
Skewness Coefficient:
The skewness coefficient measures the asymmetry of a distribution. A skewness greater than 1, where the mean significantly exceeds the median, indicates a right-skewed distribution with numerous high outliers—suggestive of grid instability.
The higher the number of skewness coefficients, in our case mean/median electricity price – the more the grid is affected by instability.
Why not use more easy-to-understand metrics, like a change of mean price over the years or peak price during month/average month price? Other metrics don't give reliable results and can be significantly affected by extreme outliers that corrupt results.
Visual Analysis:
The visual component of our study is laid out as follows:
- The x-axis represents the years from 2010 to 2023, with the mean value between 2010 and 2016 years at the beginning
- The y-axis is a dual-axis; the left (grey) axis plots the mean/median coefficient, and the right (orange) axis plots the percentage of renewables in ERCOT's energy generation mix.
- A grey line represents the mean/median coefficient, with a trend line indicating its progression over time.
- An orange box plot shows the renewable energy share, with a trend line highlighting its growth.
From the graph, it is apparent:
- The coefficient remained stable at approximately 1.29 from 2010 to 2016, when renewable penetration was minimal.
- Post-2016, as renewable integration increased, the coefficient rose to 2.31 by 2023, driven by greater frequency and severity of pricing anomalies due to grid congestion and instability.
- The notable spike to a coefficient of 6 in 2021 coincides with a storm in February of that year, which led to prolonged grid outages and dramatically high electricity prices.
2. Second investigation:
From our initial study, we observed an increasing difference between mean and median electricity prices as more renewable energy sources are integrated. To investigate this trend further in a more visual way, let's use the following methodology.
Methodology:
To minimize the impact of short-term price spikes, we'll calculate the monthly average of the maximum electricity price during the day using a two-hour window each day:
- Every day, we identify the two consecutive hours with the highest average electricity price. Given that ERCOT updates Real-Time Market (RTM) prices every 15 minutes, this method covers eight intervals within two hours (2h/15 min = 8)
- This analysis will be conducted for each price hub, which varies in price.
- We will then compute the average of these two-hour peak prices each month, smoothing out day-to-day volatility.
- This process yields the maximum two-hour rolling average monthly prices across various zones.
If this sounds too complex, this visual part will help to understand:
2023 Data Review:
On the x-axis, we plot the months of the year.
On the y-axis, we display different price hubs in Texas.
Each cell in our graph represents the daily maximum two-hour rolling average price.
Observations:
- During the winter months, the maximum average prices ranged from $40-50/MWh.
- In contrast, the summer months saw prices skyrocketing to $1000/MWh, illustrating significant seasonal fluctuations. Wow. It's a lot.
Comparative Analysis with 2016:
While the median spot price increased by only 10% (from $18.3/MWh to $21.3/MWh), possibly due to inflation, the peak prices escalated dramatically—from $110/MWh to $1100/MWh, marking a tenfold increase.
These spikes, lasting more than two hours, underscore the volatility in electricity pricing under increased renewable penetration.
This comparison highlights the critical moments when energy storage systems become essential to mitigate pricing volatility and maintain grid stability.
And it is precisely the moment when energy storage kicks in.
P.S. If you have read so far, you can look at this nice video to see the change year by year from 2016 to 2023, to feel this change:
https://www.youtube.com/watch?v=bu9m8dDUQ18
