02 Dec 2021
Alex Done

Symmetric DC participation - how much can you stack?

Since its launch in October 2020, low-frequency Dynamic Containment (DCL) has been a market entirely served by battery energy storage systems (BESS). Save some difficulties surrounding qualification for the service, DC has been a mainstay revenue stream for the GB battery fleet, with stable, high prices and low cycling rates.

In November 2021, National Grid Electricity System Operator (NG ESO) introduced a number of changes to the DC market, which notably included the launch of the high-frequency variant of Dynamic Containment (DCH). While these changes caused a step-change in market dynamics (which you can read more about here), the focus of this article will be on the interaction between DCL and DCH services. Specifically, we’ll take a look at:

  • What is stacking and what does it mean in the context of DCL and DCH?
  • The new ESO terminology surrounding the delivery of DC, and what it all means.
  • How state of energy management in DC works.
  • How much stacking you can do within the service terms of DC.
  • How this could impact optimisation strategy.

What is stacking?

‘Stacking’ is a term often used in the flexibility space to describe the delivery of multiple services simultaneously. Basically, stacking allows assets to earn revenues from multiple revenue streams at the same time.

Since the very first conversations about DC, it has always been the intention that the high- and low-frequency DC services would be stackable - that is, providers would be able to deliver (and earn revenues from) both services simultaneously. This arrangement is seemingly perfect for BESS, allowing them to monetise both their import and export capacity in providing frequency response services to the grid (just like in FFR).

New terminology

As with all good things in life, there are some caveats when it comes to stacking DCH and DCL, namely surrounding the energy-limited nature of BESS. Luckily for us, this information is neatly wrapped up in the ESO’s guidelines for DC participation. That said, it does come with some new jargon for us to get our heads around. For the purposes of this article, we will use the ESO’s preferred new terms (where possible).

What is an energy-limited asset?

The idea of ‘energy-limited assets’ was introduced by the ESO as a new way to classify assets, with the purpose of defining specific rules within DC. Energy-limited assets are defined as any asset that meets either of the following requirements:

  • an asset that derives its store of energy using power drawn from the grid; or
  • an asset whose state of energy at the start of the contracted period would be unable to deliver its full contracted capacity for the duration of the contracted period - in the case of DC, a full EFA block.

Translation: energy-limited assets = energy storage. And, in the case of DC (which requires subsecond response), battery energy storage specifically.

What are the state of energy management rules?

In the standard contract terms for DC, energy-limited assets are subject to certain rules around the management of their state of energy (SoE), more commonly referred to as state of charge (SoC).

The SoE management rules are focused around three key concepts:

  1. Delivery duration - This is the length of time a DC provider must be able to deliver 100% of its contracted capacity without interruption. For DC (both high and low), this is set at 15 minutes.
  2. Response energy volume - The volume (MWh) of energy that a DC provider should be capable of either importing or exporting before being unable to deliver the service. Response energy volume is calculated as contracted volume (MW) multiplied by the delivery duration.
  3. Minimum energy recovery - The minimum volume (MWh) of energy that a DC provider must be capable of recovering in a single settlement period to manage its SoE. The ESO mandates this as 20% of response energy volume.

Figure 1 (below) shows example values for the above based on an indicative 10 MW DC contract.

Figure 1 - Worked example for the calculation of response energy volume and minimum energy recovery for a 10 MW DC contract.

How SoE management works in DC

Let’s work through a quick example to explain how all these pieces fit together. Note that we’ve used a simplified DC response profile here.

Asymmetric service delivery

Consider an indicative 10 MW/10 MWh battery, with a 10 MW DCL contract for EFA block 1. Figure 2 (below) shows how the asset responds to a frequency deviation, in addition to the steps required to be compliant with the state of energy management rules.

Figure 2. Panel 1 shows the frequency in addition to the ESO’s operational limits. Panel 2 shows an indicative response profile of BESS (including SoC management actions) and panel 3 shows the associated SoC level of the site.

Let’s talk through the points on the above graphic to explain what’s going on.

  1. At the start of EFA block 1, the asset has sufficient SoC to cover the response energy volume requirement for its DCL contract - 10 MW x 15 min = 2.5 MWh. Put another way, it has at least 25% SoC as it begins the contracted period, as indicated by SoC falling within the ‘acceptable range’.
  2. During settlement period (SP) 48, the frequency falls and the asset responds in accordance with its DCL contract, injecting 3.5 MWh into the grid.
  3. At the end of SP 48, the DC provider recalculates the volume of stored energy remaining and SoC. In this instance, SoC has moved outside of the acceptable range - that is, the site does not have sufficient stored energy to meet its response energy volume requirements.
  4. Since the provider does not meet its response energy volume requirement, it must take action to replenish SoC at the earliest possible opportunity, in line with the minimum energy recovery requirements. That is, the site must recover at least 20% of its response energy volume or 0.5 MWh (20% of 2.5 MWh). This equates to submitting a -1 MW baseline for delivery in SP 4 (the earliest available opportunity when factoring in gate closure time).
  5. For the duration of SP 4, the site imports in accordance with a -1 MW baseline (equivalent to the 0.5 MWh recovery volume), thereby recovering SoC and meeting the minimum energy recovery requirements of DC.

Since low-frequency response requires a provider to export onto the grid, and SoC management for DCL requires importing power from the grid, baselining to replenish SoC does not compromise the provider’s ability to deliver DCL. Note, the vice versa applies for DCH.

So for the delivery of just DCL or DCH, BESS can bid at its full rated power and still be compliant with the ESO’s state of energy management rules.

Symmetric service delivery

The above story changes slightly when considering the delivery of both DCL and DCH simultaneously - imagine the above scenario, but now for a site providing both DCL and DCH.

If, during the period of time in which the site was replenishing SoC, a high-frequency event occurred that required a site to import power at 100% of contracted capacity, the site would be unable to do so, contravening the requirements of the DCH contract.

That is, baselining to recover SoC from one service can comprise the delivery of the service in the opposite direction. To avoid this ever happening, the ESO has stipulated in its participation guidance “an x MW capacity unit can [only] offer < x MW of symmetrical DC”.

This is precisely why the state of energy management rules are so important in the context of symmetric service delivery - they prevent a site from delivering a contract for the symmetric service against its full rated power.

How much can I stack?

The above raises an important question: how much capacity does a battery need to hold back from DC in order to be compliant with SoE rules? Or put another way, how much of a battery’s rated power can be used to secure a symmetric DCL + DCH contract?

The ESO has been a little cagey about this, largely down to the impacts of round trip efficiency on how effectively providers can manage their SoC. Their guidance states:

“We cannot prescribe how much headroom/footroom must be maintained to allow for adequate SoE management.” - ESO DC participation guidance document

That being said, we can make some simplifying assumptions to provide an estimate for symmetric DC capability. Let’s consider a 10 MW BESS site, unconstrained by grid connection for both imports and exports and assuming a round trip efficiency of 100%. Table 1 (below) shows the associated response energy volume, minimum energy recovery volume, associated headroom/footroom required and the subsequent capacity which is eligible for symmetric DC.

Table 1 - Required response energy volume, minimum energy recovery and associated equivalent recovery baseline. Calculation assumes 100% round-trip efficiency.
  • Based on the above, BESS must reserve 10% of its rated power for SoC management in order to be compliant with the state of energy management rules. Due to our assumption of 100% efficiency, we can expect this number to be higher in reality.
  • It should be noted that DC bids must be submitted as integer quantities, meaning some sites will need to hold back a larger proportion of their rated power for SoE management. For example, a 32 MW asset could theoretically deliver 28.8 MW (32 MW - 3.2 MW). However, the integer-value bid rule means that, in practice, the site would only be able to secure a contract for 28 MW.

The symmetric DC strategy

So with all the above taken into consideration, what’s the best strategy in DC? Fundamentally this is a decision between three options:

  1. Secure DCH at full rated power.
  2. Secure DCL at full rated power.
  3. Secure DCL + DCH, holding back 10% of rated power.

Let’s see how the commercial cases for each strategy stack up. Figure 3 (below) compares the three above strategies, for an indicative 10 MW site. Remuneration for both DCL and DCH is assumed to be in line with the (volume-weighted) market-clearing price since 01 November.

Figure 3 - A comparison of 3 monetisation strategies in DC. Pure DCH, pure DCL and a symmetric, DCL+DCH strategy.
  • Due to the higher (volume-weighted) average market-clearing price in DCL, participation in the low-frequency response service trumps that of DCH.
  • The DCL + DCH strategy is the most lucrative strategy, despite withholding capacity to uphold the state of energy management rules. The symmetric strategy provides an additional 37% on top of the pure DCL strategy.
  • One thing to consider in the symmetric strategy is the size of both markets, especially given the ESO’s low requirement for DCH over the coming months (see here for more info). When DCH requirement is low (and the likelihood of rejection increases), providers may find it optimal to solely secure DCL contracts at 100% rated power.

Key takeaways

The stacking of DCL and DCH has long been anticipated by the BESS community and opens the door for the monetisation of both imports and exports. With this new opportunity, comes more responsibility with a number of new rules for energy-limited assets.

While these rules cover many nuances of SoC management, the key takeaways can be summarised as followed:

  • For asymmetric participation in either DCL or DCH, sites can bid at their full rated power.
  • For symmetric participation in both services, BESS must hold back sufficient headroom and footroom for SoC management. Assuming perfect round trip efficiency, we estimate a site can secure symmetric contracts for at most 90% of its rated power.

With more decisions to be made each day, choosing a monetisation strategy for BESS is becoming more challenging. Current pricing levels in DCL and DCH suggest that a DCL + DCH strategy is the most lucrative, even though this means reducing contracted capacity.

That being said, this is likely to change in the not too distant future in response to the ESO’s latest forecast of DC requirements. Lower service requirements and increasing competition will make securing symmetric contracts more challenging - if not impossible - as we move into the new year.