Nodal pricing is a way of determining the price of electricity that varies locationally. It is sometimes called locational marginal pricing (or LMP). In Part One, we looked at the implications of nodal pricing for the whole energy system in Great Britain. Here, we explore what it might mean for battery energy storage systems (BESS). Later, in Part Three, we’ll model what a battery energy storage system might do across the course of a day in a simulated node.
Figure 1 (below) shows possible scenarios for how we might price by location. On the left, we have the current system - a single, GB-wide, wholesale electricity price. In the middle, we can see GB split up by its 14 Grid Supply Point (GSP) groups. There are already differences in network charges - such as DUoS and TNUoS - between these 14 groups. On the right, we’ve highlighted the 362 individual Grid Supply Points. This gives us a picture of how we could locationally price in the future - i.e. at a much more granular level.
Spoiler alert
In Part Three, we’ll use modelled data to look at what a battery energy storage system might do in more detail. In this article, we’ve looked at how nodal pricing would likely affect battery energy storage at a high level. Here are the key takeaways:
- Generation and demand profiles will differ massively between nodes.
- The need for frequency response services - and subsequent prices - is likely to vary according to the characteristics of each node.
- For, optimisers it would be simpler to navigate a unified wholesale market.
- Day/night price differentials will incentivise co-location at certain nodes.
Under nodal pricing, we could see 362 potential scenarios. Each will have differing solar, wind, baseload, hydro, total embedded generation, demand profiles, import and export capacities, loss of load factors, inertia, weather patterns - the variables are endless. This could all become very complicated. However, for optimisers (or those pulling together business cases), a unified wholesale market would be simpler to navigate, as mentioned above. Nodal pricing will mean the network runs more efficiently. It should mean a reduction in (currently) skyrocketing balancing costs.
All of this will impact the eventual price signal - not to mention the investment case - for battery energy storage.
How different could these nodes be?
The Energy Systems Catapult and National Grid Energy System Operator (NG ESO) have indicated that more granular nodes would help us meet carbon targets. It would do this by enabling the integration of large amounts of renewables without the need for major network reinforcements and excess balancing costs.
The ESO’s 2021 Future Energy Scenarios (FES) documents give us an idea of what these potential nodes might be: the 362 Grid Supply Points (GSPs). Figure 2 (below) shows the modelled make-up of embedded generation in each of the GSPs by winter 2025/26, according to the Leading the Way scenario. (As outlined here, Leading the Way is the most ambitious FES scenario - the “fastest credible [route to] decarbonisation”.) We have overlaid projected average demand (the purple line).
- Embedded solar capacity dominates this picture. This will produce lots of power during the day in summer months, but contribute little in the winter.
- In 62% of the GSP points, winter demand is projected to exceed the capacity of embedded generation.
Load profiles
While the above picture looks very healthy, with generation outweighing demand in a lot of these GSPs, we still need to take load factors into account. To model the generation vs. demand balance in these GSPs more accurately, we have applied the following load factors: 38% for wind, 4% solar, and 60% for hydro and other generation (which is likely to look like baseload generation).
Figure 3 (below) shows the average embedded generation after taking these load factors into account. As above, the projected average demand is overlaid.
- In 87% of GSPs, demand outstrips embedded generation in that node. This will result in these nodes net-importing, probably from transmission-connected generation.
The current landscape for battery energy storage
Before we dive deeper into what battery energy storage could do under a nodal pricing system, let’s remind ourselves of how it operates currently, in our nationwide system. Figure 4 (below) shows the revenue stack for battery energy storage for Q1 2022 (excluding Use of System charges).
- Frequency response services make up most of the stack.
- There is some merchant trading when the markets offer enough of a spread to warrant coming out of frequency response.
- The Balancing Mechanism offers another opportunity for flexibility.
- It’s not shown on the above graph, but there is also some (admittedly limited) opportunity for capitalising on location via TNUoS charges. We detailed this in our recent piece on the 2021/22 Triads.
What are the current locational requirements for frequency response services?
- There are no locational requirements for aggregation for legacy services (which are being phased out). FFR requires a minimum of 1 MW per unit, regardless of location.
- The new dynamic services - Dynamic Regulation (DR) and Dynamic Moderation (DM) - do have a locational requirement. Assets in DR and DM are (or will be) aggregated at GSP group level. A minimum of 1 MW of availability is required per GSP group to participate in the service.
- Dynamic Containment has a stricter locational requirement - a minimum of 1 MW per GSP. However, NG ESO recently outlined its intention to transition DC to 1 MW per GSP group, bringing it in-line with the newer dynamic services. This will remove a barrier to entry for smaller assets to participate in the service.
How could nodal pricing affect frequency response requirements?
Under nodal pricing, we imagine that NG ESO - or a potential Future System Operator - will be responsible for maintaining system frequency around 50 Hz. But how will nodal pricing affect ancillary services (and the battery energy storage systems that provide them)?
NG ESO is currently developing a mapping tool. This will allow it to better identify the location of assets. In turn, this will help the control room manage constraints locationally. Under nodal pricing, it is possible that frequency response markets would have different requirements in different locations, according to the make-up of generation and network constraints on each part of the network.
Other factors, such as loss of load or inertia, also have the potential to change requirements by node. For example, Hinkley Point C brings with it a 1800 MW of potential loss of load - which might bring with it large Dynamic Containment requirements. Yet, it might have little need for Dynamic Regulation and Dynamic Moderation. This is because inertia will be high, due to all those huge, spinning turbines.
Alternatively, a node with very little synchronous generation but lots of solar or wind generation could have significant requirements for DR, as inertia would be quite low. Even though there’s no significant loss of load risk, the potentially large RoCoF - for example, when clouds move over the area - might drive a relatively large DM requirement.
How could nodal pricing affect merchant opportunities?
With 5-minute settlement and greater price volatility (as a result of the markets doing more of the balancing), there would be more arbitrage opportunities. This suits the flexibility that battery energy storage offers. Battery dispatches in the wholesale market might start to look more like bids and offers in the current Balancing Mechanism.
There will be one merchant market in which to trade flexibility. Currently, we have wholesale hourly, half-hourly, intraday markets (potentially across several exchanges), and the BM, as well as locational network costs. Under nodal pricing, it should be simpler for operators to navigate one unified marketplace. However, this one market will (potentially) offer 362 different prices.
Optimising battery energy storage under nodal pricing
In Part One, we looked at an example node on the Scottish network. It has a significant transmission constraint, and a wind farm is producing at full capacity across the evening peak (see figure 5, below). As this is a rural area, there is little demand.
So how could battery energy storage help ease this constraint? Well, much in the same way that behind-the-meter optimisation works. In a behind-the-meter optimisation, you might have: demand from a factory or office block, or from EV charging; generation from solar and/or wind; and a limited import and/or export connection.
The demand and generation of the node help shape the nodal price, as demonstrated in figure 5 (above). When demand outstrips renewable supply, the price is high. When there is a surplus of generation, the price is low (or even negative). A battery can be optimised around this cost, while paying respect to import and export constraints. At times, this might mean minimising imports from the grid (by discharging). At other times, this might mean charging up from local generation, or acting to manage the constraints of the node. We’ll explore this scenario further in Part Three.
In figure 6 (below), we can see how a battery might store excess wind generation, and then discharge it to satisfy local demand, or export it later to the grid outside of that node (acting as a sort of cross-nodal interconnector). The right-hand diagram shows that scenario.
Any optimisation for ancillary services could take place on top of this. A battery’s availability to charge and discharge - and its subsequent availability for ancillary services - might be limited by the import and export constraints at the node.
Co-location of battery energy storage and solar
Currently, co-locating storage assets with renewable generation (namely solar) has few commercial benefits, besides spreading the cost of the shared connection. Batteries on co-located sites are usually operated independently. Therefore, there is no real commercial incentive to put a battery close to a solar farm.
Under nodal pricing, there would be a significant price gap between day and night in areas with significant solar capacity. This would create arbitrage opportunities to capture these spreads. In turn, this would incentivise the building of assets capable of capturing these spreads (such as battery energy storage systems). These spreads could last longer than 1 or 2 hours, which would also drive a business case for longer duration storage.
Final thoughts
It’s impossible to know at this stage exactly how nodal pricing would affect battery energy storage in GB, particularly as we don’t know exactly what those nodes would look like. However, we do have some idea of how it might impact frequency response services, wholesale trading, optimisation, and the case for co-location - as outlined in this article.
Make sure to come back for Part Three. We’ll model what a battery could do in a simulated node across the course of a single day.