As it stands, the 2017/18 Triad season has been very unusual, with a few potential firsts that have made forecasting particularly challenging. They also point towards a potential evolution in electricity demand, indicating both increasing energy efficiency and demand management.
Why is this important?
The Triad season runs from the start of November to the end of February every year. In this period, National Grid identifies the three highest half-hour periods of demand. Each Triad needs to be at least ten clear days apart from the others. These three Triads form the basis of National Grid’s electricity transmission charges. For half-hourly (HH) consumers with direct pass-through of transmission costs, these Triad points are of particular importance. If such consumers can foresee when Triads will occur, and react by reducing their energy usage at these times, their final transmission costs can be significantly reduced.
Find out more about the Utilitywise Triad Alert service here.
At present, based on initial outturn data, the 2017/18 Triads are estimated to have occurred at the following times:
|Potential Triad dates|
National demand (MW)
|For the half hour ending|
Utilitywise alert level issued
|11 December 2017|
Red – take action now
|5 February 2018|
Red – take action now
|26 February 2018|
Red – take action now
The final Triads may differ through a process called Settlement, as further data will continue to be submitted to National Grid during this time. This could have a significant impact on the results.
In recent years Settlement has shifted demand by as little as 100MW, equivalent to the power used by less than a million kettles, yet this still led to a change in the final Triad. The Settlement data has already changed since the Triad period ended, with one of the Triad days shifting from early January date to the start of February, and resulting in a 300MW change in demand.
Utilitywise has provided clients with a call for all three potential Triad dates. In addition, the overall number of alerts issued during the season has also fallen, meaning that the disruption to our clients’ operations was minimised, and potentially secured transmission cost savings.
We’re expecting to see the lowest level of Triad demand since records began in the early 90’s.
Levels were lower by 13%, representing a 7GW fall in 10 years – the equivalent of more than two new Hinkley nuclear power stations! Triad demand was also 3% lower than 2016/17 winter.
Overall energy consumption has been falling for the last decade, and Triad data shows that peak demand has also been falling. There are a number of drivers for this. A large part has been a general improvement in energy efficiency, with homes and businesses investing in new TVs, fridges and other equipment. The switch away from incandescent lighting is also a factor.
However, we have also seen a growth in businesses taking a stronger role in demand side management. To reduce energy costs, businesses are cutting or shifting their consumption away from the points in the day where energy prices are higher. National Grid indicated last year that businesses reacting to these price signals – including Triad alerts – had the potential to cut peak demand by as much as 2GW. This is equivalent to over 1.5 million microwaves operating at the same time.
The prediction paradox
For predicting Triads, the evolution in energy demand complicates matters, as it creates uncertainty around at what level the system will peak. This is further complicated when the very act of reacting to a possible Triad, by reducing demand, could paradoxically mean the Triad does not occur.
It’s clear from the Triads that the peak demand swing (the difference between the peak and average demand) is dropping too. The swing was close to 5GW around a 10 years ago and is now estimated at just over 3GW, the same amount of power that would be used to run 10 million modern computers.
In essence, demand is getting flatter, at the same time as getting lower, making it much more difficult to pinpoint the Triads before they occur.
Flatter demand extends the period in the winter when a Triad might occur. In the past, seasonal weather trends, along with the length of the day, have meant that the vast majority occur in either December or January.
In recent years, the Triads are starting to spread out, appearing more in the ‘shoulder’ months. Indeed, of the six February Triads going back to the start of the century, half have appeared in the last 3-4 years.
Some of this can be attributed to differing weather patterns from year-to-year. Cold snaps can appear at any point, and cause unseasonable conditions. Of course that was the case this year, with the recent ‘Beast from the East’ causing a very late Triad. If the current dates turn out to be correct, it appears to have led to the latest occurring Triad on record. However, with the swing generally falling, it’s becoming increasingly easy for small changes in demand patterns to affect whether a Triad occurs or not.
National Grid’s forecast error is also a contributing factor in adding complexity to the prediction of Triads. On average in the last 4-5 years, the level of difference between forecast and outturn has been around 1.5% on Day-ahead demand (close to 500MW). This year, the error was closer to 2%, or 800MW. If National Grid is also struggling to predict demand this highlights how the changing market is having an impact on predicting Triads for everyone else.
Stay informed with Utilitywise
Once Triad dates have been officially confirmed we’ll provide an update. Our Market Intelligence team keep a close eye on the energy markets and industry updates, keeping our clients informed at a frequency to suit them.