Electric System Peak Prediction

Knowing if tomorrow is going to be a peak demand day for the electric system is important to anyone who manages energy costs.?Your building's electricity demand on the system's peak demand hour sets a little known component of your electricity supply bill - the ICAP tag.??Reducing your building's electricity demand on the peak day (or days in some jurisdictions) can provide energy cost savings throughout the year.?It also helps reduce everyone's bills, as total system investment is reduced.?Other benefits realized from peak reduction include lower emissions and lower negative impacts of emissions.

Strategain has developed a peak demand prediction algorithm that was provided on our website last summer.?The probability that tomorrow will be a system peak day in New York City is displayed in the upper right corner of the website, or on the bottom of the drop down menu on your smart phone.?Below are the probabilities that the algorithm predicted vs. the rank order of highest load days in NYC this summer:

As you can see, the top 6 days were correctly predicted by the algorithm.?If you implemented demand reduction solutions for every day that our algorithm predicted a probability above 30%, you accurately reduced on the 6 highest days of the year.?

(As a side note, it is interesting to see that the ninth highest load day of the year had a 0% probability.?This day was a Saturday.?This is an extremely odd occurrence that rarely happens.?In fact, our algorithm automatically sets weekend peak load probabilities to zero, as historically the highest yearly loads have never occurred on weekends.)

Another way to compare predicted results is to compare predicted load rankings to actual load rankings.?This is shown in the plot below.?If the prediction algorithm is perfect, all points should line up on a diagonal line.?You can see that our algorithm forms a tight cluster around the diagonal line and performs as expected.?The Spearman’s correlation coefficient is 0.86, a measure of how well the ranks are predicted.?( where 1.0 is a perfect prediction, 0.0 is a useless prediction )

As with all forecasts, we continue to improve our predictive capabilities and look forward to more surprises next year.?If you manage a large amount of energy consumption, you need this information as, peak load charges impact about 50% of your bill.?

As the recent past has indicated, we expect the future to be unexpected.?Are you prepared??Reach out to us now to get this level of certainty in your future.

? Copyright Lucas Finco 2024