Pricing
02 Nov 2022
Robyn Lucas

Signal vs. the Benchmark: October 2022

In early August 2022, we launched Signal, our three-year forward view of battery energy storage revenues.

Three months in, how do the projected figures compare to actual revenues? In this article, we compare October’s Signal projection with the Revenue Benchmark - from the latest Leaderboard.

Robyn compares October’s Signal projection with the Revenue Benchmark.

How accurate was Signal?

Figure 1 (below) compares the value we estimated for August - October 2022 against our Revenue Benchmark, by median battery energy storage fleet revenue.

Figure 1: Signal battery revenue projection as compared to actual fleet revenues.
  • For the October comparison, we’ve used the October 2022 update of Signal. The forecasted October numbers in that update were higher than the previous August 2022 release. We detailed the drivers of these higher numbers here.
  • The main factor was the forward gas price. At end of September, it was higher than it was back in August. This drove higher revenues in the projection.
  • The actual battery energy storage Revenue Benchmark for October 2022 is £100k/MW/a (the annualized median fleet revenue), shown by the purple cross.
  • This is 64% lower than our forecasted P50 value.
  • It also sits below the uncertainty bands of our forecast - given by the P90 value of £131k/MW/a.

Battery energy storage revenues by market

Figure 2 (below) compares the share of the total revenues across October from both the Benchmark and from Signal, for each market.

Figure 2: Actual and forecast percentage share of the various markets for energy storage. Mean revenues across the month are used.
  • The relative proportions of revenues in the Balancing Mechanism, Dynamic Regulation, and Dynamic Moderation were correct to within a couple of percentage points.
  • Dynamic Containment was predicted to be 6% higher than it ended up being.
  • The proportions of Firm Frequency Response and wholesale revenues were roughly reversed. Signal predicted a 15% higher share of wholesale revenues, and a 20% smaller share of Firm Frequency Response revenues.

Figure 3 (below) shows the absolute revenues from each market.

Figure 3: Comparing the breakdown of the revenue stack for each of the markets for Signal and the actual revenues taken from our benchmark.
  • Firm Frequency Response revenues were 1.6x higher than Signal projected.
  • Signal also predicted a higher wholesale opportunity in October than we saw in reality - and similarly, higher Dynamic Containment revenues.
  • In October 2022, 0.4% of revenues came from Balancing Mechanism actions (read more about this in our Leaderboard article). In this first version of our battery revenue projection, we omitted revenues from the Balancing Mechanism - because they’ve made up such a small percentage of the stack.

So why was Signal wrong?

Signal did well in August and September! So what went wrong in October?

Well, the weather’s been mild. Really mild. Sandals and t-shirts in late October, anyone? It’s also been really windy. Record-breakingly windy, in fact - the UK got 50% of its energy from wind power on 26th October, and National Grid ESO reported that we generated just under 20 GW of wind power at one point during that day.

As a result, gas prices - and therefore wholesale power prices - dropped across the month of October. They also flattened, with reduced daily spreads. You can see this in figure 4 (below).

Figure 4: The average daily power price (the Day Ahead hourly N2EX price) is shown across October, alongside the daily min and max prices.
  • The mean power price dropped by 40% from the first to the last week of October.
  • In that same period, the average daily spread fell by 60%.

So what did this mean for battery energy storage revenues?

This means the value of flexibility dropped over the month: there was less value to capture by charging a battery during the minimum price periods and discharging during the high price periods. This was not foreseen at the end of September, or even the start of October - at least not to the levels we saw.

Firm Frequency Response prices, which were set at the start of September for the month of October, were quite high. Reasonably high wholesale opportunity cost - due to expected cold weather - was priced into these tenders. However, the cold weather didn’t really materialize - and prices subsequently dropped and smoothed.

As such, there weren’t the expected wholesale opportunities for battery energy storage. So, those auctions which clear close to real-time (such as Dynamic frequency response services) were more reflective of market conditions at the time. Because of this, they cleared at lower prices than we predicted in the Signal October update.

Check back next month to see how we fare in November!

You can see our public roadmap for Signal here: what we’re currently working on, and what's in the pipeline. We’d love to hear your suggestions on what you’d like us to include next!

Copyright© 2024 Modo Energy. All rights reserved