Who should lead the way on flexibility?
The case for accelerating energy flexibility is clear. But should this be steered by a bold new blueprint, or by empowering the market to forge its own path? In the latest chapter of an ongoing correspondence in Utility Week, Stonehaven's Adam Bell and Ofgem's Marzia Zafar go head-to-head on the 'Code Wars', market architecture, and who really holds the keys to unlocking flexibility.
Adam Bell, director of policy at Stonehaven, writes:
I hate, hate, hate the term ‘flex’. It’s a term of art used in the power sector to denote things that you can turn up and turn down. Using it marks you as a member of a cool in-group and absolutely baffles the general public. Lamentably, it also lends itself really well to puns, so my hate is largely about public understanding but also jealousy at not being the person to come up with ‘flex appeal’.
This hasn’t prevented me from having opinions on flex as a part of the electricity system, of course, some of which were recently published on the virtual pages of Utility Week. Maria Zafar, Ofgem’s deputy director for digitalisation, innovation & decentralisation has penned a very well-reasoned reply.
We’ve gone from a power system in which big power stations in the Midlands shunted power down miles of cable into people's homes and businesses in line with when they wanted it. Now, we have much more generation at the edge of the system, and that generation is driven by the weather rather than consumer demand. So being able to respond to the weather becomes very important.
You can either do this with large centralised generation and storage – gas and batteries, essentially – or you can do it at the demand level behind the meter. A home or business could have clever kit that lets them move electricity demand around in time, helping to smooth peaks and lower the cost of the system for everyone. This could be a small battery, or when you charge your electric car, when you produce hot water, or many other options besides. This is ‘consumer led flex’ or CLF.
Highly demanding
Why might you want to move demand around? Electricity has a specific value at a time and at place. If there’s lots of solar and wind output available at a specific time, that value is low. If that output can’t get to where you are because there’s too few cables to carry the current, then the value is high. So if you can change demand for electricity, either through producing more of your own or by holding off on using it for a bit, you should get paid in line with the value you’re providing. Of course, if prices perfectly reflected value, you wouldn’t need to get paid for lowering demand – doing so would be reward enough in itself.
Flex is almost the example par excellence of why prices are important in operating electricity systems: it represents millions of potential agents each with private information about their preferences on demand and their willingness to act. Prices can internalise information about system requirements to which these agents can respond. This will almost certainly deliver a more optimal outcome than the Control Room deciding whether Gladys at No 42 Richmond Road can turn her water heater on or not.
Of course, ‘prices’ imply some kind of price discovery process, which in turn implies a market. In my initial conversation with Utility Week, I sought to land the point that we require some kind of Target Operating Model for flex, which I shall define later. Zafar’s argument in response to this was:
“My view is that the answer is not to define a highly prescriptive target operating model. As regulators and policymakers, there is a risk that we become too specific about the future market architecture and inadvertently constrain innovation, competition and business model evolution. It is difficult to know today whether future flexibility will be delivered primarily through suppliers, aggregators, Original Equipment Managers, AI-driven optimisation services, local energy platforms or business models that do not yet exist.
“Instead, the role of government and the regulator is to establish the common rules, governance arrangements and infrastructure that allow those future models to emerge and compete. The objective is not to define the winners, but to create the conditions under which the market can discover the most efficient and valuable ways of delivering flexibility.”
Given what I’ve said above about the role of prices and markets, it should be clear that I almost entirely agree with Zafar. We don’t want government or its regulator to decide what kind of business model is optimal for flex, we want to discover that through markets.
The problem is that our current market architecture is already too specific.
The architecture as it stands today – separate licences for retail, networks and generation - is derived from privatisation and the Utilities Act 2000. It is not some kind of natural law, but something defined for a specific purpose and with a specific view of how trading should happen. It now actively constrains the development of flex, and assuming that it is possible to ‘evolve’ from this model into a much more contestable space without fundamentally rethinking how it works is just incorrect. The fundamentals of a market – property rights, contracts and trading – are just not in place.
We can illustrate this with the current debate around the ‘code wars’: a series of internecine debates around how exactly value is allocated between suppliers and flex providers. For those who don’t assiduously follow the progress of items of industrial legislation with exciting names like, “P510,” let me explain what’s going on.
Your supplier is responsible for buying enough electricity to meet your expected demand, half hour by half hour. They either use traditional ‘demand profiles’ to estimate how much you’ll need or, if they are more sophisticated, they will develop a profile based on your smart meter data and other relevant information about your household. They have to have bought this electricity ahead of real time.
If you happen to have a battery or an EV, you may have also signed up to a flex aggregator. They can take control of your device and, through determining when it charges or discharges, trade on your behalf in wholesale and balancing markets. But when they make a trade they change the demand profile of your household. You know, the thing your supplier has bought electricity to satisfy.
This means that your supplier may now have more or less electricity than they thought they were going to need. They are now in imbalance, and they need to pay for that. The ‘Code Wars’ are about how this is paid and who pays for it.
Not worth the paper
So we have a consumer who has signed a contract with a supplier for a specific service and then has signed another contract with a flex aggregator that actively impacts that service, and then rather than having a market where there is no interference, the regulator is instead trying to manage cash flows and licence conditions so that consumers don’t have to worry their poor little heads about signing conflicting contracts and so suppliers don’t start suing their customers.
It should be immediately obvious to a normal person not steeped in this world that this is insane. You can’t have a market where contracts are merely suggestions and where you can sell the same thing twice. These are basic conditions of trading.
Ultimately we require a Target Operating Model because a solution for this will necessarily involve moving away from a 2000s-style arrangement. If we don’t move away from such an arrangement then Zafar’s over-prescriptiveness test is not satisfied, as we can guarantee that suppliers will become the dominant flexibility providers. They have the existing consumer relationship, the existing relationship with the regulator and a deep enough bank of regulatory specialists to ensure that the Code Wars turn out in their favour.
There is a very good argument for the current arrangement: we want to incubate multiple kinds of flex providers so we can use future markets to determine the optimal flex business model. But we can’t use current markets to do that.
“Ultimately we require a Target Operating Model because a solution for this will necessarily involve moving away from a 2000s-style arrangement.”
TOM-TOM
I’ve said ‘Target Operating Model’ a few times and I should probably say what I mean. When you’re a regulator and changing markets, you can either make incremental reforms aimed at solving current problems, or you can make sweeping changes to achieve wider objectives. An example of the former is the Code Wars, an example of the latter is the Electricity Market Reform process that the erstwhile Department of Energy and Climate Change undertook in the early 2010s. That implemented Contracts for Difference and the Capacity Market, among other changes.
But those changes were additions to the current market, not measures that would instantly change how trading happens. When you’re changing the fundamental structures of a market, you can’t do it overnight, especially with something as important as electricity. Instead, you need to establish the model that you’re aiming at and implement phased changes so you can get to your end state without causing anything to fall over.
An example of this is the current market structure itself. Initially implemented via ‘shadow’ arrangements in the 1980s to enable system operators to gain experience of price formation in wholesale generation markets, the Utilities Act – which implemented retail competition – represented the implementation of a TOM that was defined as early as 1983.
I am not going to pretend that I can copy that level of ambition in a Substack post. What I can do is look at why current market arrangements don’t satisfy the kinds of requirements for a flexible market architecture that Zafar laid out and sketch out the characteristics of a TOM:
- All consumer demand must be contestable, hour by hour, kilowatt by kilowatt. Currently only consumers are contestable, but they are the purveyors of something that is now an asset in a more flexible world, and they need to be able to benefit from that. They should be able to choose any configuration of businesses to satisfy their demand and manage their appliances.
- Daily demand profiles should be scheduled and allocated in line with that contestation. If a consumer has picked you to run their battery charging cycle, you should be responsible for buying any relevant energy to satisfy that.
- Consumer demand must be exposed to price signals that reflect both time and locational scarcity. SAY HELLO, REMA’S REVENGE.
What should be clear is that there are multiple market architectures that satisfy the above. You could manage (1) by giving every consumer an AI agent who trades on their behalf and manages their appliances. You could do it through meter splitting. You could do it by splitting suppliers into billing agents and wholesale traders. The point here is that what I would like to see is a statement by Ofgem or DESNZ that this is what their reforms are pointing towards.
What we have instead is a set of very comprehensive plans for change that don’t yet state what trading arrangements would be necessary to underpin the vision. They are an answer to the question of what government can do to make flex happen, rather than the question of what market architecture would make flex fly. Here is where I disagree with Zafar: this is a question that demands an answer, or the outcome will be less flex for all.
Ofgem responds:
Marzia Zafar, deputy director for digitalisation, innovation & decentralisation, Ofgem
My view is that GB’s challenge is not simply one of market design. It is also one of culture.
We operate in a highly decentralised, consensus-driven system. That brings many strengths. It allows different perspectives to be heard, promotes collaboration, and helps build confidence in change. But it can also mean that every challenge attracts another forum, another process, another governance layer, and another group that quite reasonably wants a seat at the table.
The result is a paradox. We collectively create complexity, and then collectively complain about it.
That is why leadership matters. When we developed the Market Facilitator concept, our ambition was never simply to rewrite rules. It was to create an institution capable of aligning stakeholders, coordinating reform and providing direction where responsibilities are dispersed.
The Market Facilitator is still in its infancy. Its immediate focus is appropriately practical: establishing itself, building trust, and addressing barriers that inhibit wider participation in flexibility markets. That is important work and a necessary foundation.
But over time, Ofgem would like to see the role evolve further. Questions around baselining, demand turn-up, interoperability, market participation and future market design will require more than isolated code modifications. They will require a sector-wide conversation about where we are trying to go and how we get there.
In that sense, perhaps the answer to the “TOM for the TOM” is not simply another blueprint. It is creating institutions that can help the sector collectively define and navigate towards that future.
We have every confidence in Elexon and its chief executive Peter Stanley to help develop that capability. If we want a clearer direction of travel, we need to be prepared not only to challenge Elexon, but also to empower them to do so.
“We operate in a highly decentralised, consensus-driven system ... The result is a paradox. We collectively create complexity, and then collectively complain about it.”
