Triad Triumph

Triad Triumph

At the end of March, National Grid published the Triad dates for 2015/16, confirming that Utilitywise successfully provided a warning of all three Triads. Yet again, our Triad alert service gave businesses the tools to cut their energy bills.

Triad Triumph

At the end of March, National Grid published the Triad dates for 2015/16, confirming that Utilitywise successfully provided a warning of all three Triads. Yet again, our Triad alert service gave businesses the tools to cut their energy bills. Crucially, many suppliers and providers of a similar service failed to predict all three of the Triads accurately.

Triad Settlement Graph Utilitywise

The Triad missed by many suppliers on 25 November provided the second ‘shock’ result of the last few years. In 2013/14 we had the first ever Friday Triad, and many suppliers also missed this. We’re proud to say we called it. Last winter, the shock was not so much caused by a Triad occurring on a historically unexpected day, but rather a result of demand trends and data transparency.

Flatter, lower demand

The big surprise for the Triad on 25 November was caused by the very small difference between demand on this day and that on the 23 November, the day largely expected by the industry to be a Triad. This was because in November, National Grid’s initial demand estimates, showed the difference in demand between the two days was around 300MW. While there was less than a 1% difference between the two days, it was enough for the majority of alert providers to not issue a Triad alert on the 25 November.

When National Grid confirmed the Triad dates at the end of March, the difference between the two days had dropped to just 40kW. Considering the electrical requirements of a microwave oven, it’s possible that the evening dinner of around 50 homes across the UK could have accounted for this difference resulting in 25 November being a Triad and not the 23 November.

In fact, there were many days that were within 100MW of this day. Therefore, only a small shift in consumption on any of these days could have replaced the 25 November as a Triad. The small difference in demand from day-to-day was not just contained to November. Demand was generally flat throughout the period.

The flatness in demand is partly a result of a general trend of demand reduction over the last decade. The highest Triad last winter – 50.6GW – was the lowest since the start of the millennium, with winter demand falling 15%, or nearly 9GW, in the last ten years.

Demand-Utilitywise

This is more than just an issue of increased energy efficiency in homes. Both the falling demand and its flatness are also a reaction to businesses using energy more effectively. National Grid has estimated that so-called Demand Side Response (DSR) sometimes cuts demand in the early afternoons by between 1.6 and 2GW. This is around half the output from the planned Hinkley Point C nuclear power plant. To find out more about the impacts of DSR on demand, and the challenges this is causing for Triad forecasting click here.

Data Trust

Given the nature of the UK’s electricity market, it is understandable that National Grid cannot provide a fully accurate demand figure on the day. The system operator only has immediate data on half-hourly (HH) meters around the country. The demand from the smaller non-half hourly (NHH) consumers is initially just an estimate. However, given recent demand trends, the accuracy of these initial demand forecasts for NHH consumers may be having a stronger impact on determining the Triad days than in previous years.

Based on National Grid’s initial demand, Operational Metering Data, it predicted three days which would be the Triads. In the end, only one of these forecasts has proved to be accurate. Even with the initial Settlement Metering Data, which is a delayed and a more accurate assessment of demand, only two were correct. Demand figures are then revised further by National Grid’s demand as more data becomes available. This leads to the final demand assessment used to set the Triad dates. The demand revision of 23 November to 25 November that many in the industry missed did not become public until after these assessments had occurred, with the data used not available until the Triad dates were published.

Crucially, even when the Triads are published, the demand levels being assessed are a mix of data from various stages of the settlement process. The 25 November Triad had already gone through two reconciliations, or reforecasts, when the Triad dates were confirmed. However, the Triad on 19 January had only gone through one reforecast and the latest Triad, on 15 February, had only got to the Settlement Final stage. While called ‘final’, there are actually four more reconciliation levels after this before we get to the actual final demand figure.

Despite the challenges with lower, flat, and large revisions in initial to final demand levels, we succeeded where many failed by providing alerts on all Triad days. Click here to learn more about how our Triad Alert service can help your business, or please contact 01527 511 757.

Paul Anderson

Posted by on Friday, the 15. April at 15.54

Paul Anderson joined Utilitywise in 2000 as a member of the Market Intelligence team.As a Senior Analyst, Paul is responsible for the production of bespoke strategic consultancy projects as well as developing analytical models. This has included a market-leading Triad forecasting model and a fully inclusive delivered energy pricing forecast tool. Paul has an Honours degree in Film and English from the University of Kent.