Transmission Network Use of System (TNUoS) charges recover the cost of installing and maintaining the transmission network in Great Britain.
There are two ways that TNUoS charges apply to assets on the Leaderboard:
Transmission-connected battery energy storage systems (BESS) are liable for fixed TNUoS charges.
Embedded-export tariff charges apply to distribution-connected assets according to their import or export during triad periods.
Fixed TNUoS charges
An asset-specific TNUoS generation tariff (£/kW) is calculated for all transmission-connected assets. This generation tariff is made up of a wider tariff and a local substation tariff. The local substation tariff is collected in our Modo database, and corresponds to an asset's Grid Supply Point (GSP) and the connection type found there. The wider tariff is based on locational zones, detailed by National Grid and calculated using:
Annual Load Factor (ALF): currently BESS are classified as a 'Conventional Carbon' generation type.
'Year Round Shared' and 'Year Round Not Shared': these represent the proportion of transmission network costs shared with other zones, specific to each locational zone.
Adjustment Element: a flat rate for all generation zones.
These variables are used in the equation (shown below) to calculate the wider tariff which is then added to the substation tariff to calculate each asset's generation tariff.
Conventionally, TNUoS is charged annually. This annual figure is calculated by taking the generation tariff (£/kW) and multiplying it by each asset's Transmission Export Capacity (assumed to be their rated power). This is then shown on the Leaderboard as a monthly average of this estimated annual figure.
Embedded export revenues: triads
Export Transmission Network Use of System (TNUoS) charges (or revenues) can be an important addition to a battery site’s annual profit and loss statement. Calculated using the average of a site’s import or export during the triads periods of the winter, they are location-dependent, relying on the embedded export tariff of their Grid Supply Point (GSP) group.
Batteries in regions with strong potential triad revenues (ie they have a non-zero export tariff) may make optimization decisions based on the triad opportunity. For example, they may not secure a peak DC contract on a day when demand is forecast to be high, potentially missing out on revenue. As a result, they might have lower-value ancillary contracts over the month – but could have significant triad revenues instead. We would currently miss this in the Leaderboard.
For winter 2021/2022, we calculated triad revenues for the battery storage fleet (BM units only) in April 2022, once the triads were announced. This winter 2022/2023, we will estimate triad revenues each month as part of the Leaderboard. This is to have greater visibility on optimization decisions around triads.
The actual triads won’t be known until April, and they have been known to move around once the final settlement metering is in. We will produce a similar report to assess the performance of the fleet’s BM units across triad periods at that point.
Triad Revenue Estimation: Overview
Modo will estimate the revenues a site achieves from TNUoS by:
Estimating if there has been a triad in the preceding month, by evaluating the peak demand periods across the month (and the rest of the winter) and comparing to expected triad demand. More on this below!
Applying the locational import and export TNUoS tariffs to any possible triad periods (depending on the Grid Supply Point group of the site), and using power flows from the site (taken from P114 data).
As we get new data through the winter, the estimated ‘triad’ periods may change. For example, the triad period at the end of November could move to the start of December. So, each month, the preceding month’s triad revenue estimations may also change and will be updated.
Modo will do a final reconciliation of the TNUoS benefits once triads are announced in April.
How will we estimate if there has been a triad period?
Triads are the three highest half-hour periods of demand across the winter, separated by 10 days. To estimate if one has occurred in the previous month, we evaluate if demand has been high enough to be a candidate triad. And, set some rules on how many triads to expect.
What will the triad demand be in Winter 2022/2023?
a. GB peak demand has been decreasing in recent years.
b. We determine a derating factor to estimate the decrease for this winter, compared to last winter, based on the last 5 years, shown in figure 1, below.
c. Using Indicative National Demand Out-Turn (INDO) values for the previous triad periods, peak demand has decreased year-on-year average by -1.9%.
Taking the lowest triad value from 2021/22 (an INDO of 44,083 MW) and decreasing it by 1.9%, we set a threshold of 43,262GW to be the minimum demand for an eligible ‘potential triad’ for Leaderboard calculations.
A couple more conditions then determine whether these high-demand periods are eligible to be a potential triad:
The demand value is higher than 43,262 MW.
Any peak demand days must be separated by ten days.
There are no more than two triads in each month and no more than three overall.
For BM units, the revenues will be calculated using P114 data to determine the power (MW) and multiplied by each asset’s TNUoS tariff.
We may also change our minimum demand value for determining a triad depending on how demand varies over this winter. As the winter progresses, potential triad periods are likely to change - with the lowest triad moving around as new data arrives. So, TNUoS revenues may also change throughout the Triad season and we will re-calculate these each month once more data is available.
What's shown in the TNUoS column in the leaderboard?
The TNUoS column in the leaderboard will show the sum of fixed TNUoS (for transmission connected sites) and embedded export TNUoS revenues (for those who have a non-zero export tariff).
Because the TNUoS revenues are likely to change throughout the winter, as we get more data, the TNUoS revenue estimates will not be included in the Modo Benchmark.