We use two methodologies to calculate wholesale revenues for assets in the Leaderboard. Which one we use depends on whether an asset is Balancing Mechanism-registered (a Balancing Mechanism Unit, or BMU), or not (a non-BMU). This is because BMU assets are required to submit physical notifications to ELEXON, which we then use to estimate wholesale revenues. Unfortunately, for non-BMU assets, we do not have the same visibility of the physical operation of an asset. Therefore, estimating wholesale revenues for these assets is more complex.
For BMU assets, revenues are calculated based on two things: energy throughput (MWh); and prices in power exchanges (£/MWh). Wholesale energy throughput is calculated using physical notifications submitted to ELEXON, on a settlement period basis. We then multiply this throughput by the corresponding price, which is taken as the average between Nordpool's day-ahead hourly auction and EPEX's day-ahead hourly auction.
For non-BMU assets, calculating wholesale revenues is more complex. The Modo team analyses these assets on a settlement period basis. When we have no public visibility of an asset's operation, we need to decide whether or not we believe they are pursuing merchant trading activity, or whether they are out-of-market.
We assume an asset is pursuing merchant trading activity during periods when it is not providing frequency response services and when there is a wholesale spread greater than £50/MWh (on the average of Nordpool's day-ahead hourly auction and the EPEX day-ahead hourly auction). We assume an asset is out-of-market when it is not providing frequency response services and when the wholesale spread is less than £50/MWh.
When we assume an asset is pursuing merchant trading activity, we estimate wholesale revenues using the following methodology:
We award assets the largest spread in Nordpool's day-ahead hourly markets, depending on site duration.
Revenues are then assigned with the assumption that each asset cycles once per day.
To calculate the energy throughput in a given cycle, we take an asset's highest contracted capacity in frequency response markets (from the previous five days) and multiply that by the asset's duration.
We review these estimated wholesale revenues against those given to BMU assets of the same duration. We then apply a weighted calibration to non-BMU wholesale revenues, in order to smooth out any anomalies or unrealistic discrepancies.
Trading activity in wholesale markets can be complex. Therefore, assessing wholesale trading can also be complex. At Modo we make assumptions to help provide users with the best estimations we can make. These include:
Non-physical trading: our estimations do not account for non-physical trading.
Power exchange prices: we take an average across both Nordpool's day-ahead hourly auction prices and EPEX's day-ahead hourly auction prices. This is because some optimizers will have access to either or both exchanges. However, power may be traded via other auctions in reality (e.g. EPEX intraday, or Nordpool's day-ahead half-hourly).
NIV-chasing: estimations do not account for any potential NIV chasing;
Non-BMU assets: assumptions are made as outlined above.
Failure of delivery: we have no visibility of assets failing to deliver on their physical notifications, as delivered to ELEXON. Therefore, we assume that all assets have delivered on their physical notifications.