Thursday 19th of October 2017 | Posted In: Generation Management, Utility Account Management

TRIAD update and historical analysis

Pre-winter 2018-2019

Changes to TRIAD Rates

The key development since the last TRIAD season is that the Government’s plan to reduce TRIAD income has now been finalised. The initial consultation on this faced several legal challenges, but they were all ultimately unsuccessful in reversing the decision to wind down the scheme over the next three years.

This year’s rates for exported power during TRIAD calls have reduced by an average of around 35%, with some areas experiencing a much higher reduction than others:









National Grid’s forecast of the upcoming five years’ of TRIAD rates makes the reductions in tariffs very clear. For the majority of areas, export rates reduce to levels approximately between £3/kW and 7/kW before coming to a rest. For some unluckier zones such as Scotland and the North of England, the tariffs will be removed entirely as of the 2020/21 season.









This places even more importance on chasing the TRIAD income this winter before the rates reduce to these levels. Whilst rates are lower than last year, there’s still good money to be earned, with the lowest (Northern Scotland) being worth £11.36/MW exported and the highest (London) being worth £39.96/MW. These rates are imposed as costs on importers, so evading these remains as much of a priority as hitting peak export power levels during them, if it can be done without interruption to normal site operations.

Pre-winter 2018 2019

History can help us to narrow down the days and times when TRIADs are most likely to happen. Whilst this tells us when a TRIAD has never occurred and when they are more likely, there are no guarantees. The information that follows is based on data published by the National Grid, covering all the TRIADs from the 1990/91 season onwards.

Analysis by month

Looking at the graph below, you can be almost certain that there will be at least one TRIAD in December and one in January. However, it is rare for there to be no TRIAD in either November or February in any given season, although this is what happened in 2016/2017.








Looking at November:

  • The earliest a TRIAD has been was the 17 November in both 1992 and 1998.
  • Last season there were no TRIADs at all in November.

Looking at December:

  • Generally, there is a fair distribution of TRIAD dates throughout December.
  • There have been none during the Christmas break; the latest in December was the 20 in 1999, 2006 and 2010.
  • Last season, there was a TRIAD on the 11 December.

Looking at January:

  • Generally, there is a fair distribution of dates throughout January.
  • The earliest being the 03 January in both 2002 and 2008.
  • Last season, there were no TRIADs at all in January.

Looking at February:

  • The latest a TRIAD has been during this month was 26 February 2018.
  • Last season, there were two TRIADs on the 05 and 26 February.

Analysis by day of the week

No TRIADs have ever been recorded on a Saturday or Sunday.

Until 2013, there were no TRIADs on a Friday (06 December).

Monday has a respectable lead for TRIAD dates, last year all three TRIADs occurred on Mondays. However, looking at historic results there is a high likelihood of a TRIAD occurring on any of these four days.









Analysis by time of day

Important: TRIADs are quoted by National Grid as the half-hour ending, whereas electricity meter data is generally quoted at the half-hour starting.

We have reported TRIADs as the half-hour starting as this is how most suppliers present the half-hourly meter data. The graph below shows that the majority of TRIADs (90%) are between 17:00 - 18:00.









  • In the last 10 TRIAD seasons (30 Triads in total):
    • 25 were 17:00-17:30
    • 3 were 17:30-18:00
    • 2 were 18:00-18:30
  • The last time a TRIAD was 16:30 - 17:00 was 17 December 1997.
  • The last time a TRIAD was 17:30 - 18:00 was 05 February 2018.
  • Last season all three TRIADs were at different time bands.

Key drivers

The key drivers that affect when there is a TRIAD are:

  • Combination of industrial and domestic demand
    • The overlap between some industry still running and people getting home from work gives us a reliable peak around 17:00 - 18:00.
    • A reasonable amount of industrial use still ends the working week at lunchtime on a Friday. Therefore explaining why there is rarely a TRIAD on a Friday.
  • Weather
    • Generally, if it is cold, electricity demand is higher.
  • Street lighting
    • As day-length increases towards the end of February, the overlap with domestic use thins out. Explaining, in part at least, why there is rarely TRIAD after the first week of February.

However, we are seeing a trend of more industrial users turning equipment off when there is a TRIAD warning, which is making TRIADs overall less predictable. There are also more embedded generators (connected to the local distribution network) hoping to earn TRIAD income that reduce the demand on National Grid. Combined these make TRIAD prediction increasingly difficult.

Your TRIAD strategy

So, armed with all of the above, what approach are you going to take this winter, or perhaps, what approach would we take?

  • Not worry about TRIADs too much until 07 November, possibly a week later if it is not cold.
  • Focus on Monday – Friday inclusive.
  • Focus on 16:00 - 19:00
    • Power prices during this period also tend to be high.
  • Relax over the Christmas – New Year holiday period.
  • Not worry about TRIADs after mid-February, unless it is exceptionally cold and/or winter has been very mild so far.

Pull this all together and there are only around 60 days out of a theoretical maximum of 120 days when it is likely there will be a TRIAD, and we can be confident they will happen between 16:00 - 19:00.

‘Expert’ TRIAD warning services typically issue 25 TRIAD warnings in a season. Therefore, after a bit of simple logic (above), they are issuing a TRIAD warning, on average, slightly less than every other ‘likely TRIAD’ day.