Data Centres, grid reform and the race for power

Matt (00:02)

Hello and welcome to Energise2030, a podcast from TLT unpicking the forces reshaping the future energy market. I'm Matt Grimwood, a Future Energy Real Estate Partner at TLT, advising on the real estate aspects of future energy projects, including wind, solar, storage, electric vehicle infrastructure and data centres.

And today I'm delighted to be joined by John Howells, Partner at Newmark, where he is Energy and Data Centre Sector Lead for Europe. Hi, John.

John (00:39)

Hi Matt, pleasure to be here. How are you?

Matt (00:41)

Very well, thank you. Good. Great to have you here.

Now, John and I recently spoke at a data centre conference together on a panel discussing the rapid development and delivery of data centres, and it's great to be continuing that conversation here.

Data centres have become one of the fastest-growing infrastructure sectors in the UK. Demand is being driven by AI, cloud computing and digital transformation. But that growth is also bringing new challenges. Questions around power availability, grid capacity and location are becoming increasingly important, particularly within the context of wider regulatory reform, and the recent consultation from Ofgem has put all of these issues firmly in the spotlight.

In today's episode, we'll discuss what these developments mean for developers, investors, energy companies and policymakers, and explore how the sector can navigate the opportunities and challenges ahead.

So, John, let's start with the big picture. Data centres have become a major part of the UK's infrastructure conversation. Why are they attracting so much attention at the moment?

John (02:08)

I think it's to do with growth, in short, Matt.

Compared to other real assets, real estate and infrastructure, which are going through more normal growth trajectories and evolution, data centres are going through this tremendous period of surging growth.

Compared to other sectors, which might be moving at something like plus 3% or plus 5% a year, with data centres you're looking at compound growth of 18% to 20%. And that's even with the restrictions that we're going to talk about later in this episode.

There are certain things in Europe which are constraining that growth. Policy, planning and grid, in the main, not demand and not capital. So we think that growth could be a lot faster if it were more unconstrained.

Matt (03:02)

One of the biggest developments we've had in recent weeks has been Ofgem's consultation on demand-side connection reform. Do you want to take us briefly, John, through what the industry is trying to digest and what the key pillars of that consultation are setting out?

John (03:22)

Yeah, certainly. I mean, this has been a hot topic, probably only 10 days ago, I think. So it's relatively hot off the press. We're still trying to digest it all, thinking about the implications and talking to clients, some of whom are concerned.

I think it's important to say, though, if you stand back a little bit, why is it happening?

Well, that's because we've got a queuing system which is not fit for purpose. It's a legacy of times gone by. It doesn't reflect the huge build-out of green energy that you and I have been working on, and it certainly doesn't reflect the huge rollout in demand, which is really only a few years old when you're talking about data centres as demand customers. Previously, they were all distribution-connected.

So what's happened is people have effectively grabbed a position in the queue in order to allow them to develop a data centre project. The queue is now unmanageably large, around 120 gigawatts as at last summer, and the reform, which is known as "Curate", being led by Ofgem, is all about rationalising that queue.

I think the important point to recognise is that, in trying to do the right thing and get this country more competitive, and get the digital infrastructure built quickly and efficiently, Ofgem has brought in some suggested reforms. We should clearly note that they're still subject to consultation.

But in bringing in those reforms, we will find that there are winners and losers. So the aim is honourable and well understood, but there's a lot of debate around the implementation.

Matt (05:07)

And how do you think developers, investors and other stakeholders might see this consultation? Do you think this is a technical change or, perhaps, something much more significant?

John (05:22)

It's technical. It's commercial. It's monetary as well.

There's a lot in there, but the two most important points that our clients are looking at are, firstly, the need for an additional level of security against the grid connection.

At the moment, if you have a transmission grid connection, whether that's for demand or generation, you sign up to a connection agreement which effectively makes the customer, or in this case the data centre, liable if you cancel that connection. That's because the transmission owner is building out infrastructure to facilitate your connection. Actual physical infrastructure is being delivered and, therefore, if the customer cancels it halfway through, it's fair enough that they should have to pay the aborted costs.

But what this is bringing in is effectively a reservation charge, which is additional and different. It doesn't reflect the level of build-out of those costs or that infrastructure. It's effectively just a "use it or lose it" charge against the grid connection.

So the analogy we were sort of kicking around the office is more like a popular restaurant where you have to put a deposit down for each of the diners, which you will lose if you don't show up. And if you lose it and they manage to resell the table, you don't get it back again. It's just gone. You don't effectively get anything for your money. You're effectively being penalised for accessing it.

You might take the view that that's kind of normal business, and we've got an oversubscribed system. We need something like that to deter speculation. I'm sure that's where it's coming from, but where you're looking at figures of between £200,000 and £700,000 per megawatt, we're talking enormous sums of money. Some of these schemes are up to a gigawatt in this country now, and that will be quite polarising in terms of who's able to make those payments and who isn't.

So that's the first big flagship issue.

The second one is around the commitment from customers. The reforms are also looking for readiness to be demonstrated by customer commitments. We've seen something similar in the energy sector. You and I have done a lot of work on that over the last couple of years with our clients around readiness.

But customer commitment here, what we're really talking about is someone who's ultimately going to consume the power. So the entity that is effectively using that compute and is going to need the energy to power their operations.

Now, there are a lot of data centre projects in the pipeline which are a very long way from having a customer commitment. And that's deliberate because the people promoting those projects aren't yet ready. The product's not yet ready.

There are potentially unforeseen consequences of bringing in customer commitment milestones in terms of how we allow projects to come forward in the future and who is developing them.

So the way we see it is having lots of buckets of promoters. They range from renewable energy companies who are looking to diversify, through to strategic land promoters, through to propcos. A lot of the big UK PLC propcos are in data centres, as you know, particularly from the industrial background.

You've then got a lot of our energy clients. So I'm talking about big PLC energy clients who own land and assets around transmission grid supply points, and they want to co-locate data centres with low-carbon energy.

You've got data centre operators, co-lo developers, which is a large group of very well-capitalised players, but they participate in slightly different ways depending on which one you're talking about. And then lastly, of course, you've got the hyperscalers, the big American tech giants who we're all familiar with.

So, if you think across that broad cross-section, that's a lot of participants in the market, a lot of whom are doing things based on the rules today, in the right way. They're trying to bring projects forward in order to meet demand, and they're now going to have to wrestle with potentially losing their projects if they can't put up the money and/or if they can't find a customer.

Matt (09:48)

I think you're right, John. There are a lot of different stakeholders with different capital profiles, shall we say, among those that you've listed, which of course therefore gives rise to a huge diversity of interests and considerations. So it's going to be very interesting to see how that plays out.

I just want to pick up on one of the themes that you pulled out there from the previous piece, around co-location.

One of the big themes I think we've both seen in recent years is that idea of a data centre being co-located, located very close to a power generation site, which might be wind, solar and, indeed, storage, where there is a multi-technology aspect, perhaps. So we're hearing lots about co-location, but why do you think that's become such an important topic?

John (10:42)

It's certainly been a huge trend over the last year or so over here. We've been a little bit behind, I think, some other countries in terms of co-location of generation with demand.

But if you think about it, this was happening way back in the Industrial Revolution when those power-hungry industries set up shop where the power was being generated. And we think, as a house, that trend will continue, and you will see data centres moving out to where the abundant green energy is being produced. We're already seeing that in certain locations around the coast of the UK.

But to answer your question, the reason it's got particular precedence in the last year is around two things. And these two things we hear all the time.

Number one, time to power on, or time to energisation. So how quickly can I get my data centre built, up and running?

Number two, cost of power. How many pence per kilowatt hour, or pounds per megawatt hour, is it going to cost in a country where, as we all know, power is very expensive?

Matt (11:48)

That's it. Yeah.

John (11:50)

So those two things are absolutely crucial.

It's sad to see that ESG has probably fallen back a little bit in terms of that list of priorities. I think most of the clients developing data centres will want to see green energy used where possible, but not at the expense of power-on date and not at the expense of cost of power.

And of course, you've got redundancy as well, but that's a slightly different point in terms of resilience and uptime.

Those two things have been crucial. And you're seeing now in the market, in the role that we have in the market, people are coming to us and our peers in the industry saying, "we need to be on in two years, and we need power to be less than 16p" or something, and beyond that, we don't really mind very much.

So they're quite location-agnostic, even to the extent of being cross-border location-agnostic. We've seen requirements in the market that will go pretty much anywhere, provided that it's quick and cheap.

So that's the main reason that behind the meter has come to the fore, because, in theory, you can develop generation assets more quickly than you can get grid connections.

Matt (13:01)

Yeah, yeah. And we've seen, of course, not just with the data centre demand-side reforms, but also the generation-side reforms last year and Gate 2, and some of the issues and challenges that created within the sector as well. So it would be good if we could try and learn from some of those positions, no doubt.

Do you think the market now is moving towards a future where energy strategy becomes a deciding factor in where data centres get built?

John (13:30)

I think it already is, to a large extent. You're simply not able to develop a data centre without an energy strategy.

You've got different push and pull factors driving locational decisions at the moment, some of which are around what I just said: being able to find early power and cheap power.

You're seeing that with the eye-watering 10-gigawatt projects in Texas, which are using cheap Shell gas, five cents per kilowatt hour electricity costs, abundant land and thousands of acres in the middle of Texas or other states. Those projects are drawn there because of the energy strategy. So that is already the case.

I think it's going to become even more marked as we go forward. What I think you'll find is that, over the next few years, people are going to have to be quite innovative.

We're working on quite a few projects at the moment which are not at all straightforward from an energy perspective, but they're in good locations. You find that developers and investors are willing to be much more creative if it's in a good location.

So you might find you're in Paris, Frankfurt, London or Dublin, somewhere of that ilk, a top-tier market, and there is no grid power in the next maybe eight to 10 years. So what do you do? Do you just not develop, or do you try to find a solution?

So that's where we're seeing behind the meter coming to the fore. But everything I've just said there, using your phrase "energy strategy", is all about energy strategy.

I would like to see a world where we have more of a system-wide approach. We've seen, with generation connection reform that ran last year, NESO, with a reasonable amount of success, standing back from all of the noise as to how it happened and just thinking what is best for UK PLC when it comes to energy strategy.

I'd like to see a world where the demand side is looked at in a similar way, because you've got a lot of players in the market, all of whom have got strong views, and we all do, but we need to think about what's best for us. How can we be most competitive globally? And how do we use energy generation and demand to make that happen?

Matt (15:54)

I think you make a great point, John, about that competitiveness with other countries as well.

You alluded to that traditional sort of FLAP-D dominance within the data centre world and more emerging markets, maybe like Milan now, or other less traditional data centre locations. So it will be really interesting to see how that develops.

Just moving back into the UK and those traditional data centre hotspots that the two of us know very well, around Slough and West London, do we think that the geography of that market in the UK is beginning to change?

John (16:35)

It is, but that's not because people have gone off that area. It's purely because of the supply-side constraints that I touched on earlier on. There isn't enough power and there isn't enough land.

Even if there were both of those things, the way that particular market works is that most of those data centres in that location are going to be between 20 and 100 megawatts, that kind of scale.

There are a couple of projects being promoted around that West London corridor that are larger than that. But if you think about that kind of scale of project, and remember that there isn't that power, if there were, we would be talking about sites of maybe four to 10 acres, 12 acres, that kind of size.

You can sort of see, with a bit of a leap of faith, that it is possible to find sites with low planning risk, or ideally brownfield sites, maybe close to the Slough Industrial Estate or Hayes, which can facilitate that kind of development.

But if you're looking at AI, then we're talking about a completely different scale: 300, 500, 700 megawatts, up to a gigawatt, and we have had enquiries larger than that. So there we're talking about things that just can't realistically be delivered in those sorts of locations: hundreds of acres and hundreds of megawatts.

So that's what's leading to the locational shift. It's not because those areas are... that people haven't gone off them.

Matt (18:16)

Okay, and when we try to look forward then, maybe to the next 12 to 24 months, what do you think are the biggest questions facing the sector?

One thing I was going to briefly touch upon, although it's probably the subject of a podcast in itself, is from a planning perspective, where we've had the NSIP regime now a bit more aligned with data centres, in terms of seeing the NSIP DCO regime. But where do you see the biggest questions facing the sector over the next year or two?

John (18:52)

There are many.

I think if we stand back initially and just think about where we sit in terms of international competitiveness, it's well documented that our cost of power is pretty unhelpful for all types of industry, but particularly digital industry. So we need to get better at that.

But we also need to encourage investment to come here. So in terms of having a supportive policy environment, we need to show the right amount of receptiveness and friendliness to these tech companies so that they want to invest here.

But we also need to think really carefully around sovereignty as well. We don't want our entire digital economy to be dominated by just a few players.

Getting a bit more local, you touched on planning there, Matt. We've got this whole debate around Town and Country Planning versus DCOs, data centres being nationally significant infrastructure or critical infrastructure, and that has implications in terms of consenting risk in a relatively small island with lots of AONBs and national parks and multiple objectors.

I think we're treading a bit of a tightrope here because if the policy position under DCO is a presumption in favour, as it is with some energy technologies, people are going to end up with very large digital infrastructure investments being built potentially quite close to sensitive receptors and sensitive communities, and having very little way in which they can influence that process.

I'm not saying that's wrong. Like I said, from a system-wide approach, we need to think about how we want to deliver this infrastructure in the country. We do need it. We can't just decide that we will go without. That is not an option, unfortunately.

But we need to balance this quite carefully because, as you've seen, even in some of those American states that I mentioned earlier, where seemingly, from over here, it feels like you can just pretty much build what you want, they've seen a huge amount of local opposition. And that, of course, can lead into a different political environment in terms of how people are feeling about those sorts of projects.

Matt (21:15)

So the consultation we mentioned from Ofgem is open until 16 September. So still five or six weeks to go at the time of recording. Decisions, we understand, will be taken later in 2026.

I think it's worth mentioning, as you alluded to earlier, that the demand-side queue went from approximately 40 gigawatts to about 120 gigawatts between 2024 and 2025.

So, keeping all of that in mind, what would you like to see, really, John, from industry participants and stakeholders as this debate develops through to the end of the consultation period, but also beyond that too?

John (22:04)

I think it would be really helpful if we could hear from, or rather if Ofgem could hear from, different participants along that value chain we talked around.

There will be louder voices than others, some of whom have got very well-rehearsed arguments around why it's important for their investment to continue to support UK growth. But there are quieter players as well, including some that are playing a really valuable role in this whole process.

We talked about energy companies, for instance, which are very used to putting consultation responses in when we're talking about energy market reform, but maybe haven't done it before when we're talking about demand-side reform.

I think it'd be useful for those companies to perhaps come together and discuss what they'd like to see, and make sure that they're able to make the representations they need to.

But there are other people in that whole spectrum of participants that we talked around who are also doing a really important role to get this country ready for the future. I'd like to see them all represented.

The other thing is really around that academic piece. This is big industry. There's lots of money flowing into it. Everyone's got a vested interest. That's kind of normal. That's just where we're at. People want their own organisations to succeed, of course. That's human nature.

But I think, in the same way that we've seen in the energy sector, I'd really like to see some more whole-system, academic-type thinking around these things, about how we make it work for the UK, particularly in terms of sovereignty, security and international competitiveness.

I'm concerned that if we've got too many disparate, loud voices and too many opinions, then that whole UK system-wide view might be lost.

Matt (23:59)

Yeah, I think that's a great way to conclude things, John.

I think, as you say, it is about making things as cohesive rather than disparate as we can. It's unquestionable, isn't it, that data centres sit at the intersection of digital infrastructure, energy and economic growth.

Grid reform, availability of power and location strategy will all play a major role in determining how the sector develops.

I think the challenge now is not whether we need data centres, but how we are, from a regulatory perspective and from an economic perspective, going to come together to avoid that disparate piece, as you mentioned, to form a cohesive whole, which will then engender confidence in investors, developers, policymakers and energy companies alike.

So absolutely a critical piece as we move forwards.

Unfortunately, that's all we've got time for today. A big thank you to John for joining me on the podcast. And thanks for listening to TLT's Energise2030 podcast. If you'd like to catch up on previous episodes in the series, you can visit the TLT website.