Data Centres, BESS, and Building the AI Backbone in Scotland

Giles Hanglin (00:02)
Scotland specific to the data centre market missed out when

Ireland grabbed the data centre markets of 20 years ago and they created all the tax incentives to go there. So we’ve been sitting down with Scottish government to try and create the case with them as a key stakeholder to say, right, here’s an opportunity. 5 GW of projects could bring 60 to 70 billion of capital construction value to the country, as well as thousands of job creations over sort of a decade period of time.

Adam Sroka (00:48)
Hello and welcome back to the Hypercube podcast. I’m honoured today to be joined by Giles Hanglin from Apatura. Giles, would you like to introduce yourself to the audience and maybe tell us a little bit about Apatura and what it is you’re up to.

Giles Hanglin (01:00)
Yes, certainly. Well, I’m delighted to be here, Adam. Thank you very much for invitation. Giles Hanglin, I’m CEO at Apatura. I’ve been there for about a year now. I joined from another battery storage developer based in East of England. And that’s kind of part of about 15 years I’ve spent in the renewable energy and storage industry. Before that decade at Savills, the real estate business, where I started off in the rural land management or the farming side of the business and then

migrated through the kind of solar era in 2009, 2010, then helped grow part of the Savills energy business from there. It went through the kind of the what we call the solarcoaster of, you know, tariff offers and tariff removals up to 2016 when we were looking at scaling around the large rooftops solar business across a lot of the industrial portfolio and across England and Scotland. And, of course, Amber Rudd.

Who was our favorite politician at the time pulling the plug in 2016, sort of kiboshed that. Took a couple of years out from the energy industry as it all refound itself and then got back into the storage area with Cambridge Power. 2018 and scaled that for five years. That’s a portfolio being sold at the moment. And then a year ago Apatura came along and said, we’d like you to lead the business and take us through into construction, operational, battery assets at the time, which was what the main state of business was. But things have moved on a little bit. Sure we’ll talk about that today. Yeah, definitely.

Adam Sroka (02:22)
I mean, the land real estate background, that’s got to be really useful in this

Giles Hanglin (02:28)
business, right? Absolutely. Yeah. And if you look at any energy developer, they all come at it from different angles. You get some that come at it from a grid angle, and kind of just base their business around grid to start with. Some of it, like ours, is around land and having a really strong land presence, because land is two thirds of the aspects of an energy project. You’ve got to get the land, you’ve got to lease it or own it, and then you’ve basically got to get planning on it as well, which is another real estate play. So having a strong aspect on that, once you’ve found the grid, we’ve found is the best recipe for success.

Adam Sroka (02:57)
So you’ve hinted at Aputura moving on slightly from the just battery development and it’s exciting stuff. Do you want to share some of the details?

Giles Hanglin (03:06)
Yeah, no, certainly. So came into a portfolio here at Apatura which in Scotland is around about 9 GW of projects, all secured across the central belt in the most part. And those were destined to be purely battery projects. I joined us, as I say, a year ago, there’s no hint that any of the use of those projects might come on, other than co-locating maybe some solar or some generation. We were looking at maybe some hydrogen projects at the time, which was quite buzzy. I think the appetite for that sort of wanes slightly at the moment. But about in the middle of last year, we started getting

interest from a couple of the data centre hyperscalers in two of our specific projects, which aligned with some interests they were taking in Scotland. And we’ve never been a seller of just under consented or non consented projects. So that’s something we didn’t want to do. That’s why I was brought in to monetise our projects properly. And so we took interest in this and decided to go and get ourselves upskilled on why certain leading global hyperscalers wanted our projects. And it turns out it was all part of the AI and the machine learning

growth revolution, which you’ve now heard so much about, and wanting to get some sort of large stakes north of the country where plentiful power is available. Of course, that’s what we sit on. And so we’ve now looked at repurposing part of our portfolio to be compliant with what those guys are looking for. And we spent the last nine months really getting up to speed on that, learning as much as we can, getting aligned with the right financial partners who can monetise that, the right technical partners who can assess sites and make sure they’re fit for purpose.

And so we’re sitting upon a position at the moment where we’ve got just over three GW of our projects we determined as being suitable for data centre development. And so we are across 10 sites taking those through a modification process with National Grid to change those type of collections. We are assessing and making more land available around those projects to create data centre campuses, which are about 200 acres each. And then we’re looking at the planning play, which hopefully should be a bit more of a shoe in maybe than some other sort of traditional projects.

Adam Sroka (05:02)
Does the network type of the fibre network connectivity add much complexity to like the planning bit. Cause obviously you’ve got to line up the grid connection, the site and all that, but then you obviously need connection. Is that hard or is that easy?

Giles Hanglin (05:20)
So everything’s driven by power to start with. If you go to any of these conferences, we went to a couple at the end of last year, they all tend to be in quite exotic locations. But what we heard is just power, power, power is the main thing that all of these campuses, it’s a global thing actually, if you look across the global market too.

So what we sit on is the power and then the secondary items include a scalable land, a favourable planning regime. And then you write, you then get into the type of the kind of the equally important, but you kind of consider them later on, fibre connections, water availability, that sort of thing. Thankfully Scotland and the central belt is very well connected anyway between Glasgow and Edinburgh. And you also sit on the kind of the big highway of information between England and the UK and the Nordics.

And that big fibre connection goes offshore somewhere in Northeast Scotland, and it hits across. And again, that’s a massive information superhighway that we can tap into. Latency is not such an issue though for AI machine learning from what we’ve been made to understand. You don’t need to be a couple of microseconds outside London for that availability, which is where the main sort of throw of data centre development, in fact, the whole of the world really in terms of concentration is in that West London area. So now we can be up to sort of 10 milliseconds away, which basically

brings in the whole of the UK really from our point of view.

Adam Sroka (06:35)
No,

not too many trips to slough in the future. No, and totally in the direction of travel, like it’s the reason we exist. I always joke that the pitch deck for Hypercube is like, there’s a big hockey stick growth in BESS and renewables. And then there’s a big hockey stick growth in data and AI and you just slap out the logo in the middle and that’s the pitch deck. Like I was at Data Day Texas back in January and

Giles Hanglin (06:38)
look for us.

Adam Sroka (07:02)
They were talking about the announcement about Stargate or whatever it’s called came out like $500 billion. And I was saying, look, 30% of that’s going to be power infrastructure because that’s the whole game effectively. To develop that volume of AI capacity in new data centres in the US, the grid’s not there for it, the production needs scaled, the networking and things like that. it’s a really

interesting time. I just think doing, having expertise in both BESS and data centres, just makes total sense. it’s like every good idea, once you see it, it just makes total sense. After the fact.

Giles Hanglin (07:43)
There’s a lot of comparisons to the energy business in the UK from the part that I’ve been through. When we first started solar park development back in 2009-2010, we did some joint ventures with Solar Century and BELECTRIC at the time, which have now been developed on. But at the time, the limits on those projects were 5 MW to start with and then 10 MW. That was a 25 acre or 50 acre site, which clearly was a challenge at the time. It seemed to be big infrastructure. Now you’ve got projects being pushed through. We had one of our battery projects consented for 700 MW at

Port Glasgow earlier this year, another one this week of 400 MW. And so you then look at the solar side of things and you’ve got the likes of Arden Green, Paranelkin pushing through six, 700 megawatt projects. There’s another one being mooted in Norfolk, which is going to be a 4,000 acre project over 800 MW. And the scale of everything is just taken off. And so that doesn’t surprise me. So when you then talk to the data centre guys who’ve been used to projects up to a hundred MW in size, which for

a cloud-based system in West London. It’s been a very big project. If you look at the US market, which tends to trail on the DC side from here by sort of a year 18 months, and they’re doing GW scale projects out already, led by Tesla and others, it doesn’t surprise us. And I guess it’s less of shock to the energy industry that scale is there, that it is to the DC side, which is still getting their head around what does the architecture of a GW scale project in the UK look like.

Adam Sroka (09:06)
And there must be like a pride or like that warm feeling of actually quite a number of jobs getting created for developing and building these sites and not quick short-term jobs either high value high salary good technical jobs in locations that maybe haven’t seen them for yeah a couple decades it but it must be like a warm fuzzy feeling for that for Apatura.

Giles Hanglin (09:13)
state law.

Absolutely. Yeah. And think the whole thing about scaling our business, I was employee number 20 about a year ago, we’re now up to just over 40 people. So we’re creating a lot of high value jobs there. And as I was saying earlier on, it’s difficult to get those exactly in the same geography as you are. So we’ve had to be very flexible about the location of those. So a lot of our people are spread remotely across the country because we want their specific skill base. It’s secondary as a need to be in an office the whole time long. But looking at Scotland, I think Scotland specific to the data centre market missed out when

Ireland grabbed the data centre markets of 20 years ago and they created all the tax incentives to go there. So we’ve been sitting down with Scottish government to try and create the case with them as a key stakeholder to say, right, here’s an opportunity. 5 GW of projects could bring 60 to 70 billion of capital construction value to the country, as well as thousands of job creations over sort of a decade period of time.

Then the ongoing job creation, because these things, much to the…

to sort of dispel some of the myths. They’re not just empty sheds full of racks. Basically, there’s a lot of people that work in these, a bit like the distribution centres. Everything’s automated distributions, just full of robots. About 300 people a site, that are working in any of these. So it will be job skill creation, again, a high skill base, like you say, to maintain these. I mean, some of these data chips and things now are highly complex. And so if you bring that sort of engineering piece of soil up to Scotland and create bases for it, that’ll create jobs. That’ll then create the education stream to

deliver it as well with universities backing that up and creates like an ecosystem which is generating of that. And I think Scotland’s been missing somewhere particularly on the side of some jobs which we’re finding very difficult to procure up here. And we have to go further afield because it simply doesn’t exist and it hasn’t had the backing for it.

Adam Sroka (11:12)
Yeah, and I think you get beyond a critical mass whereby it then starts to feed itself and then draws it becomes like a destination to come and do that kind of stuff. What an exciting future, something we like to ask people as CEO, what’s a day to day look like for you? Like you must be quite a busy guy.

Giles Hanglin (11:31)
I’ve quite a varied day to day and quite a varied week as well. So I have no real set routine. I live down in Cambridge and that’s kind of background for me, family down there, boys at school there. So for me that’s home. But our offices, our main HQ is in York and then our secondary office in Edinburgh, which has very quickly become this sort of equisized as we build that out as our key deliverable given all our projects are in Scotland. So all of our new recent employees on the construction and delivery side, on the data centre development side have all been up there.

Head of Legal’s up there and it’s important, obviously, having somebody in the right jurisdiction. And so I tend to split my time if I do travel kind of between York and Edinburgh, but this week, for example, a couple of days in Dublin, then here in Glasgow with you, back home in London tomorrow. That’s where some of our funding conversations are going on. So I have to be very flexible and do that travel. And that’s fine. That sits well with me. I don’t mind that so much, but it’s in my previous roles, I’ve always been a lot more project-based to where I’ve originated projects and been aligned with those.

The role here with Apatura to come in and take over from Adrian Hill, the owner, was very specific to lead us through a funding round to get funded for running the business to start with and then the capital into projects to build them, which were only batteries at that time. And now we’ve obviously pivoted to data centres as well. And then also to take those through construction, build construction teams and its operations. So you have to a much more strategic role in the business overseeing all of that.

Managing shareholders’ expectations, managing team growth, managing growing pains like your offices and things like that as well. And then also growing the presence of the business across the industry so that when you go out for capital and go out to contractors, they know who you are, they know the scale of you, know you’re not just another fly by night, and you’ve actually got the means and the purpose behind you to deliver the projects which is why it’s worth them spending the time on you.

Adam Sroka (13:16)
It’s interesting to hear because you think like, always think in what Apatura are doing, it’s complicated. Like it’d be hard to describe to someone 100 years ago exactly how you’ll make money and where the value is. But there is a lot of value. There’s a lot of money there. And I think actually to be able to achieve what you are with 20 to 40 people in the timeframe to become a key stakeholder, like in Scotland, doing all that.

Do feel like you’re punching above your weight a little bit? It certainly sounds that way.

Giles Hanglin (13:43)
Yeah,

Absolutely. I think if you look at us alongside the other people who hold significant portfolios connection to the grid connection queue, it’s the likes of SSE, it’s RWE, it’s all of those big corporate utility businesses, but they’ve got a mandate for delivering energy to customers through those utility connections. So they can’t pivot away and use those projects or anything else other than what they were destined for, which would be generation or storage projects. So they aren’t a threat to us as a business. So hence why

our portfolio projects in Scotland sits there kind of finally unrivaled. We’re still 100% privately owned. We haven’t yet had to divest any capital to any major institutional owner. That will come down the line, I’m sure, when we take some big capital inputs into the business. And it might be strategic at that point that giving away a certain percent to the company then leverages its value by basically adding more value on top. So at that time, we’ll be very open to doing so. But the right time to hold onto that share cap until now has been the way to grow the business and

Makes it more tricky from a funding point of view because you’re having to go out and raise private capital rather than relying upon an institutional partner. Believe me, that comes with its challenges. So I think I’ve spent all of my time since I’ve been here, constantly fundraising either for running the business from a private point of view or capitalising the projects and finding capital to build them out with parties. Particularly in the battery industry, that was challenging last year. Given downturn in revenues, been a big uptick this year and it’s helped with the likes of the listed funds

reporting better revenues and so therefore the sentiment’s back in the market again and it makes it much easier to go to market as we are at the moment and seek capital to build those projects out.

Adam Sroka (15:19)
Yeah, it’s funny. We feel the effects of that. We’re kind of downstream and a lot of our core customers are in the wind of the battery space. And we feel it where everyone’s really excited. Guns blazing, they want to do AI and then down-turning revenues and everything gets put with the buffers. Then again, now into this year, it’s been like absolute havoc since sort of the end of December through to now where everyone’s back on

the war path to try and accelerate and do more stuff. And it’s funny that there’s that bit of lag, but then like we do kind of follow those same trends and same cycles. I guess I wanted to reach on… yeah, recently passed a GW storage. We’re looking at Neilston, BESS in Scotland. You’ve recently had the announcement this week. Is there anything you’d like to share about the actual significance of this and the growth in the Scottish market and

Just expand a bit on the natural fit of why Scotland is so well suited. We touched on Ireland and missing out. Why Scotland is so well suited to this application.

Giles Hanglin (16:25)
So I often get asked whether I prefer to have a portfolio back still in England or develop a portfolio in Scotland. Of course, as an Englishman, coming up to Scotland is sometimes harder to do business, but thankfully, I know. But thankfully, we’ve got a homegrown talent up here and we’ve got a full entity in the office here to take projects for, which is important. But I would 100% always want to develop a Scottish portfolio at the moment. I think the regime up here has been very supportive.

Despite all the new noise that Labour have brought about since the election and smoothing the pathway for large infrastructure projects, making data centres on the critical national infrastructure list. Great. I mean, that’s been really helpful down south in England, but all our projects are here. So we’re under the SNP regime, but actually it’s been very supportive for what we do. So if you take those couple of examples, those planning successes we’ve had this year, Neilston about a month ago took us just over a GW.

400 MW of Denny was this week and that takes us after nearly a gig and a half. So that’s a gig and a half of fully consented projects. And that’s in addition to Devol Moor, had through 700 MW, which makes it the single largest connection on its own in Scotland. So there’s some other cumulative ones that have kind of trumped that with Alcemi and then the CIP projects of Coalburn, but as a standalone.

Then it’s great. That’s what we’ve got to scale this year on consented projects. And thankfully with Clean Power 2030 coming through and the need to have those projects with their planning by this spring, then those thankfully will be grandfathered through. And so they will make the connections list and it gives a lot more clarity. Like you talked about the noise of last year, there is at least now clarity as to which projects initially are going to be taken forward. We think those lists are still far too underbaked. There’s no way you can deliver 35 GW of new generation into Scotland over the next decade with only 5 GW of storage on the network. So

there needs to be more and that will commercially find a way in future. Hence why we’re developing a much bigger portfolio to back up for those projects that fall out or get called upon to come and supplement them in the future. But think that Scotland has been certainly for us a very successful ground. I think the way you deal with big projects through the Energy Consent Unit on a planning basis gives you a lot of clarity. I know there’s a lot of frustration of some people getting caught in the early kind of part of the consenting process and needing a screening, but the portfolio we pushed through, we’ve had some very clear

routes of communication into the top of ECU. They tell us pretty much on a weekly basis where we are and what’s coming forward. And it does give you that clarity. And I think certainly with the data centre side as well, the meetings we’ve had with Scottish government, very supportive, going back to the point about Ireland before, they don’t want to miss out this time. They want to do what they can. They want to create a key stakeholder group that helps push these through. We’ll be responding this week to the expressions of interest on the AI growth zones around the UK. They’ve been pushing Scotland hard with their support

which is a really key one. But we’ll do everything we can to make our projects attractive, but government needs to do their part as well to try and create Scotland as a destination for this if they want to capture all that capital investment and the ongoing revenues and job creation that we’ll create up here. So yes, I mean, from that point of view, Scotland has been good for us and I think certainly we’ll keep pushing Scotland in doing what we’re doing. We have hosted more projects which are about to go into planning, kind of resource allowing.

But if you look at the Energy Consent Unit as a standalone body, which is projects over 50 MW have to go there. They’re advertising, I think, for 24 jobs at the moment. So, I mean, they are seriously understaffed. If we go back to us talking about resource constraints, we find the planning and planning consultancy, particularly in Scotland, is something that’s not come through the education system, there hasn’t been a huge creation. It’s a massive shortage. And you’ve got some of big developers who’ve been poaching staff and taking them on to expand their projects and therefore, it leaves everybody short.

And so that’s the constraint. You can’t just smooth through that with any sort of policy if you don’t physically have the people to do the work.

Adam Sroka (20:06)
I want to just dive into a little bit about the appetite from the hyperscalers and the, like, why is, why is putting a big BESS asset on like next to the data centre? Why does that make so much sense?

Giles Hanglin (20:23)
So the hyperscalers have come to Scotland, for example, and they’ve been looking to try and get into the power market and they’ve realised they can’t bully NESO and then Scottish Power to get connections. They’re like, well, join the back of the queue, mate. It’s 2040 and lodge your own connection application. And so they’ve not liked that. So they’ve had to buy into connections and they’ve realised for what they require,

buying solar connections is no good for them because there’s only generation connection. They need a demand connection. So they’ve been after battery assets. So the battery assets have already by virtue of the way that National Grid used to model batteries from sort of two, three years ago, they’ve modelled batteries as 100% import, 100% exports asset. So helpfully it’s already got that 100% import there, which is what they’re after. It needs to be turned into what we call a final demand connection for a data centre. It needs to have more resilience around it.

It can’t have the sort of 5% containment against the connection because it needs 100% or 99.998% grid resilience in the event that grid fails basically in creating a catastrophic meltdown given the heat that’s generated in these things now. So grid resilience for them through by virtue of two feeds from the same substation or two separate feeds from separate substations is what they’re after. And so we’ve had to go about

looking to procure that extra bay or the extra substation connections to make our projects fit for purpose. So to do that, obviously that precludes a lot of people from looking at it and the hyperscalers have had to understand that that’s the way that these things come about. And hence why we feel that putting that together using our expertise and understanding grid and what they want and having we brought Michael Hunter in from Google who joined us two, two and a half, three months ago now. And feels like he’s been here a lot longer.

But he used to be part of the site origination team at Google for five years. So he’s been able to screen and kind of oversee our site development to design it in the way that Google would have wanted to see it. And every hyperscaler is different, of course, and they’ve got different products. But generically, if you can create something that’s open to the AI market, it makes it a lot more advanced than somebody just throwing a single

feed connection with some content on top of it, just because it’s a 400 KV, maybe in the central belt, a hyperscaler is saying, here you go, how about my project, take it forward. It’s not fit for purpose. It’s not what they want. They want speed to market and they want basically something that’s put together.

Adam Sroka (22:41)
Yeah, okay, I hadn’t thought about the grid resilience bit but that, yeah, of course it all makes total sense.

Giles Hanglin (22:45)
So the battery part, so if we come back to original question, what does a battery form part of that? Well, obviously the battery connection’s there, we’ve applied to National Grid to have these grids connections, some call modat, a modification application, to make it into a co-located battery and final demand connection. So it means it can serve a data centre, but it can also have the battery, it’s got the generation side still on it. Our ethos is to try and now put a battery on each feed into the data centre, both those feeds

will be off at either independent supplies or off two bays. And then that can form two parts to the grid resilience. It can form the initial part of what’s called the uninterrupted power supply, the UPS. So for the first 20 minutes in the event of power failure, given the heat that’s generated in these things now from these data chips, it’s worse than the battery case of kind of the thermal runaway in the risk. It’s massive. And so before your main UPS, which is normally a 72 hour supply of gas or diesel on site kicks in,

battery can kick in a sub-second response and create that backup for power whilst you’re making UPS and then your grid supply comes back. So that’s one part of it which they really like. So you’d always have to have 20 minutes of supply on your battery. The second we see is a massive OPEX saving. So it’s being able to procure all of the curtailed wind or use curtailed wind off the network at the moment of which there’s massive amounts. There’s a huge piece as you’ve probably seen in the press over the last week. Scotland’s biggest wind farm being paid 65 million last year alone in

curtailment payments and only being operational 29% of the time. I mean, that’s just crazy for me. But they’ve been given the kind of the lifeline to carry on through these curtailment payments, which were just over a billion pounds last year from national grid. Given the volume of wind that’s going to be built out for 35 gig over the next decade, those curtailment payments are only going to get more, more, more significant. And so therefore if we can put batteries on the network that mop up that power typically overnight and then feed it back into the data centre during the day.

You are effectively using green free electrons that could be powered in and it ticks a lot of the boxes of the hyperscalers in terms of the ESG requirements of what they’re doing. And I think also National Grid will love you as a customer because you’re creating new demand on the network, mopping up generations. They don’t need to spend it on the transmission upgrades going south. And also of course then putting really what that generation was created for, putting it on the map where demand is co-located alongside it. But unfortunately,

the southeast of England is still the biggest use of power in the whole of the UK. Power needs to find its way there somehow. There’s not much generation located around the southeast and that’s just a factor where we are now.

Adam Sroka (25:16)
Does that like, so there’s a lot of chat out there just thinking about that. And cause the regional aspect, like putting demand where the generation is, it’s great for Scotland because no shortage of generation, no shortage of land. Like let’s go. Does the regional pricing, like conversation bump into that. Does that help or hurt that use case?

Giles Hanglin (25:38)
So it’s a proper marmite subject, the regional pricing. You’ve probably seen this week there’s been a big outflow of protesting from a lot of the industry about this and the fact that obviously pricing in the North, particularly Scotland, going to be the cheapest and pricing in the South is going to be most expensive. And of course that massively impacts a lot of both the generation of storage projects up and down. For the data centre market and in Scotland particularly, it massively helps because having the cheapest pricing where

Scotland’s basically where we’re locating these is completely going to add to the OpEx savings. So if you think about the billions of kilowatt hours these projects are going to use each on a lifetime, that’s a hell of a saving across that time. But you can see why that might then impact our battery projects by having a shorter spread of pricing and the ability to sort of leverage that. So there’s two sides to this. Octopus have come out and have been very public about supporting regional pricing. And of course they see it looking at what is good for the customer across the country.

But of course, a lot of those people who’ve got vested interests in projects and assets that are subject to it, are clearly very much against it. We see it as a positive clearly for data centre side and it’s a negative for the factory standard and battery sides. But given that hopefully we’re trying to push more of our resource into batteries that support DCs, probably on a private wire basis into the back of them as well, it’s a positive for what we’re doing. Is it coming much further down the line? I mean, David Wildash, who joined us from National Grid

was former head of connections there, joined us three months ago, has been another really key addition to our team, came in just at the right time to take us through kind of grid reform on the MODAP process. I mean, luck and judgment here. But having designed all of that at National Grid and having all those contacts there, it’s been great really to understand the way that he foresaw it running. I think going forward, he says, look, it’s something that’s likely to come in probably the end of the decade, 2030, something like that, but it’s still a little way away, but still.

It’s come back on the agenda and it’s been very topical in the last six weeks, I suspect, in the press. We’ve seen a lot of chat about it and clearly you can argue both ways. You haven’t got huge demand in Scotland. You’ve only at the moment got five GW of demand in Scotland. You’ve already got 10 GW generation there, forecast to go to 35 GW the next decade. That’s all just going to come flowing south. And I guess, like a toll road, you charge monies at the moment through the TNUoS transmission user system charges

to send power from the north to south. So TNUoS charges in Scotland are the highest, the lowest down south. But of course, the most energy is created in the north. So there’s always a balanced argument for this. And so I kind of sit on the fence with it because it can benefit both and negatively impact both of our projects as well.

Adam Sroka (28:15)
I wanted to touch on the debate around the surplus heat usage and that kind of peripheral benefit. Could you expand on that and how the heat from data centres can be used for local businesses, communities and things?

Giles Hanglin (28:26)
Yeah,

For sure. If we all think back to the Paris Olympics, there was the well-documented example of the main Olympic swimming pool being heated off a data centre. So, Paris, it’s a great story. And that’s how things like that, something like a local leisure centre that has a huge heat demand is a great paired up asset to go with a thermal producer of heat. Down in Cambridge, where I live, we’ve got a lot of biodegestors around, it’s a very agricultural area.

And there’s been co-located glass houses and other off-takers of that heat that have sprung up purely off the back of that surplus heat. They’re growing mushrooms, tomatoes, strawberries, the lot really very successfully under controlled conditions. Look, that is a great use. I think we can look at pairing heat off-takers with the data centres. It works in two ways. One, it creates a tertiary use of a byproduct, which of course is great rather than just releasing into the atmosphere. But secondly, also…

reduces the cooling requirement of the data centre. So if you’re simply extracting heat rather than cooling in situ, and that’s what the power requirement of these things is really is just the escalating heat within these coming off these chips that needs to be cooled and powered. If you can take away that HVAC requirement and basically extract that heat and use it somewhere else, it’s a perfect partner. Thankfully where we are in central belt with quite urban still in some of our areas. And so we have got some very large potential municipality heat offtakes.

And we hope that we might be able to promote more land within the campuses to make it available for people to come in and look at third party or third arrangements, such as building glass houses and looking at maybe the food demand thing as well, really creating another requirement there to do that. So yes, absolutely is something that’s kind of on our list. It is always a secondary kind of consideration to all the endeavours we’re doing, but we think as a part of a political stakeholder play it’s very important.

Adam Sroka (30:13)
So I sit on the board of The Data Lab, Scottish government sort of innovation centre for AI hub here. And I was hosting a session where one of the talks was about how actually for every pound spent on like green initiatives in a city creates like up to £37 worth of like secondary and tertiary benefits. And you could totally see how you could tie

that story in where, okay, we’ve created a data centre, but then there’s a carbon element to saying, actually we’ve created like a low carbon costs, yeah, leisure centre or like you said, low food growth. And you could see how then, especially in the central bit of Scotland for people that don’t know geography, like everyone up here is used to driving anyway, or getting from one place to another. It’s not an enormous geography actually in the central belt. So you could see how, yeah, springing up new

communities and business locations that make use of all this stuff is totally feasible for where we are. So really exciting proposition.

Giles Hanglin (31:16)
Yeah,

One of our sites has got a prison next door to it, the other one’s got a big hospital next door to it. So we’re in discussions with them already about the potential for the heat offtake. Believe it, there’s a whole lot more infrastructure, heat exchangers and all sorts of things that are required, a lot of tonne in the piping really. Transport of heat is a lot more tricky clearly than transport of electricity, but it is still viable, it’s still worth looking at. I think the cooling aspect is one thing that really struck me when everybody knows that

Scotland’s cooler than the south of England, it’s typically if you look across the temperature graph, it’s three to four degrees on average all the way across the year. And each degree centigrade relates to 4% energy cost saving. So being four degrees up here, you can already have 15, 16% energy use as an efficiency case versus having a data centre in the south of England. So you’re already creating yourself an opportunity here from doing that. And if you can then further add to that, it’s only a win-win.

Adam Sroka (32:12)
What about the kind of arguments against them. Do you think, I feel like we’re making a sales pitch today on data centres because it is such a genius use case. What about the flip side? Like, do you think there’s a risk of overdevelopment and people, I think this happened in batteries where a few people sort of carved a path, did something really interesting. Everyone saw the revenues and then lots of big pots of money followed and just kind of copied, right? We got a little bit of the, we’ve got the grid connection problem.

Do you think we run risk of overdevelopment of data centres or this use case in the next decade?

Giles Hanglin (32:46)
Absolutely, yeah, there’s a huge rush at the moment over supply. I think if we look at the grid queue, let’s go back to energy projects for the time being and Clean Power 2030, the way it’s been designed to try and thin down that grid queue, which was getting towards terawatts, I think it was, of projects really, I mean, it was sort of 900 GW went by, absolutely plenty. And it was always four times over subscribed for what we needed to get to net zero by 2050.

And everybody’s got to remember that, which is why coming along and we’ve seen some of the big developers, including ourselves, we’ve had our portfolios significantly trimmed by having projects now that don’t fit the criteria being fully consented to fit in the pots that are being ascribed to each technology around the country. And so I think that’s done a job and clearly we think some of those are undercooked, particularly South of Scotland, which I alluded to earlier, and that you’re going to need more storage to facilitate 35 gig of new generation up here. Other parts of the country I can see are probably about right.

So they’re not producing huge amounts of generation. And if they are, it’s in constrained areas where you’ve got heavy populations and putting large infrastructure projects in there. It’s just not viable. I think when it comes to data centres, it’s exactly the same, really. I mean, it’s been well talked about across the world and you’ve seen Microsoft already this week start to pull out of some of their surplus leases because they’ve obviously procured more leases or options for leases across the projects. And lots have been made about it. I read into the detail of it. I was not particularly fussed, it was not Microsoft pulling out to the AI market globally.

It was them just releasing some of the sites that maybe fringe and weren’t core to their products and wanting to maybe relocate some of their central AI to the US. And that’s fine because that’s, they’re a big part of that Stargate project as well. So I think over here, it is definitely speed to market, which is why we’ve heavily invested very quickly and got through and we’re making sure our projects are going to be fit for purpose so that when a hyperscaler or an operator looks at them, they’re going to look at them once and go, yeah, these guys know what they’re talking about, these projects are fit for purpose. They suit our

capability. We’ve got a customer with the demands that can absolutely be fit from 2028 onwards with scalable power up to say half a gig or a GW 2031. That’s going to fit market, brilliant, ready to go. Yes, we’ll secure it rather than pay the risk of something else coming down the line. So it is a speed to market thing. And again, there’ll be a lot of people I’m sure who will go through the MODAP process and that’s not a cheap process. I mean, that’s 40 to 50K per site to submit and modification applications to national grid for a transmission connection. Those are added up and they had a whole

fluster of those come in just before the deadline. I mean, they’ll be sitting on a lot of processing now, but a lot of revenues from projects are never going to go anywhere. They’re probably going to take six months to get processed, whereas ours are all underway thankfully. But that is going to be a product of oversupply and there won’t be all those projects needed. And that’s just a classic case going back to what I said earlier on about having projects in the right location, understanding all the other parts that are needed for a data centre project.

But we are still, from our point of view, still selling Scotland versus North England versus the Welsh coast versus Silicon Sands and Blackpool versus projects further south. That, the power is what basically drives it. And that power is what we have. So we sort of sit upon the number one asset.

Adam Sroka (35:51)
You kind of always want to see what’s going to happen. I guess for where you are now though at Apatura, what are the kind of biggest challenges you’re facing?

Giles Hanglin (35:58)
Biggest challenges? We all ride what’s called the solarcoaster, haven’t come up with an acronym for it yet really in terms of on the battery side of things, but that’s been one. And I think that being a professional within it, and I’ve got some very good people in my business, take a long-term view on development. You’ve got to look through the highs and the lows. Of course you celebrate highs, but you basically don’t get too low with the lows and you’ve always got to take a longer term view of where you’re going. And that’s, that’s just a characteristic of somebody who’s going to stick it and

not sort of lose all their hair or have a cardiac, you know, within a decade. I think that the challenges are clearly always going to be under-resourced planning, getting projects through basically to meet suitability. Resource at national grid level, resource at government level, relief for dealing with planning is always going to be short and it’s always going to be difficult to get that, but you’ve got to try and set yourselves up and try and navigate it as best you can. And you’ve got to accept sometimes you’re gonna get hit.

One of our projects in Scotland got brought forward five years about a month ago from being out of the money basically in the mid 2030s to now late 2020s. And that’s great because that adds another 500 MW basically into the charter. That’s just unheard of. You can’t allow for that. I mean, that’s just because they sort of suddenly start looking seriously at your grid connection, thinking about the build out, realising some of the projects have fallen away and yes, you can come up the queue a bit. Delighted with that.

On the flip side, we’ve had projects that get pushed back because they realise they can’t actually deliver the infrastructure they’re going to do. So that’s a big challenge is that unknown of the grid, but what we’re trying to do is stay as close as we can. David Wildash coming in clearly can help with that by looking at the TEC reg to the queues and understanding them with a bit more granularity compared to some maybe taking a high level view as to what infrastructure can be needed. And Clean Power 2030 to achieve it.

And let’s remember, it used to be 2035, the conservative census, the records, as the target. And now Labour have said 2030, they’ve trimmed five years off. National Grid are just going to choose the ease to deliver projects to do that. And so anything that needs major rewiring or transmission upgrades is just going to get parked into like a second pot. And hence why we have to focus and look at how that impacts our portfolio and make sure that the projects that we’re really investing in, which is what we’ve done as part of the kind of a screening on the DC side as well.

Is make sure projects that will easily get their grid connections because there’s minimal there to do to facilitate those connecting to the transmission network.

Adam Sroka (38:17)
And so, yeah, I guess what are your… to capitalise in all that and make it happen and like, what are your goals for the next sort of five years and the opportunities you’re going to go after?

Giles Hanglin (38:28)
after.

I’m always quite realistic about what we’re doing. We’ve got a very big pipeline, got over 10 GW of connected projects. That’s been cut by more than half because of Clean Power 2030, but we’re going to keep developing part of that portfolio certainly into the early 2030s because we think it will fill some of the gaps that we left from the commercial need after Clean Power 2030 and actually commercial need of that 35 gig. It’s going to come into Scotland.

If Adrian was sitting here, he’d be telling me, we’re going to deliver all of it, every single megawatt, and then I’ll have nothing less. I’ll be slightly realistic with him sometimes and sort of say, look, come on, look, we’re giving ourselves the best chance. If we had 10 gig and it’s given us five to take forward, that’s much better than having five and only two to take forward. So you always have to assume there’s going to be attrition. It always hits the team hard when a site they’ve been working on hard and delivering the news about Clean Power 2030 and the number of projects we’re going to have to park or drop.

Yeah, of course it’s hard for some people and we’ve had to let some landowners down. That’s always difficult knowing if being from a land background point of view, if you suddenly found some, let’s be honest on the Scottish revenues and farming compared to energy, nowhere near those. If you’ve been promised it’s all the golden egg and suddenly gets taken away because policy’s changed, well, we can’t help that, but it’s still a difficult message to deliver. I think that we are again setting ourselves up with a large portfolio of data centre projects. They’re not all going to be needed. There is not a three GW need for

data centre demand in Scotland or anywhere in the UK. That will scale. And it’s a case of creating the first project, which are easy to develop, creating capability for yourselves. We’ve got to remember as Apatura as a business, one of our weaknesses is we haven’t built a project yet. We delivered stuff that has been built for other partners, but we haven’t built anything ourselves yet. So we’ll start that this summer with a construction of our first battery projects just around Glasgow. We built the construction team, guys have all been there, done it. We’re going through procurement at the moment on the battery side.

But once we’ve done that, we’ve got the capability and then you become an easy partner to go back to because you’ve delivered what the market wanted that first time around. And that’s trust. And that’s exactly that. And that’s what we’ve been. My ethos has always been surround myself with people who know what they’re doing, been there, done it in a far better way than I have done. But ultimately as a team, therefore you build that capability and that confidence to a part where you can go and do it. Then you might get asked to go and do it abroad because clearly you can do it here. You can then go and scale it in other jurisdictions.

Adam Sroka (40:46)
Before we let anyone go, always just like to ask if anyone wants to find out more about you or Apatura or if there’s anything you’d like to promote or plug, where should we point them?

Giles Hanglin (40:56)
So a year ago when I joined, Apatura was a totally unknown business, including to myself. I had to go away when I was asked to come and join and do some proper research on who they were. And that’s because the name changed, Green Power Consultants before through to Apatura. But they’ve been very modest about talking about anything. We’ve changed that totally. We’ve had to make ourselves proper known customer to National Grid, going through our funding processes, the contractors we talk to as well. So we’ve now got a much better website. We’ve got some much better PR and comms. You’ll see us in the news press regularly.

Debs is our fantastic PR consultant, she came in and really revolutionised our marketing, our PR. So we’re much more proactive now, it’s aligned with our business plan. And so we have a mandate really to use the press and use marketing media. So come to the website, you’ll find us on there, look on LinkedIn. We’ve all got basically decent profiles on there. And offices as say York and Edinburgh, people more than welcome to come visit if they need to, and come see us. So yes, we have a pretty active profile now, which is fine.

I’m very happy to be so fronting that, encouraging the rest of the team out of their sort of more modest shells to be sort of putting their face behind the PR releases. Andrew, my head of development, loves having his face in front of every press release on the planning results, but there we go.

Adam Sroka (42:07)
Very good. Someone’s got to do it,

Giles Hanglin (42:09)
He’s

got a great face for it.

Adam Sroka (42:12)
Well, look, thank you again very much for coming and joining us here today. Yeah, best of luck over the next few years.

Giles Hanglin (42:17)
It’s been a pleasure, thank you very much.

Adam Sroka (42:19)
Cheers.

Adam Sroka (42:23)
What a discussion that was. Really interesting to talk to Giles about what they’re up to at Apatura. I just think like any really good idea, it seems totally obvious after someone’s shown you it. Capitalising on what you could do with the BESS asset, the existing grid connections, getting round to the bottleneck by co-locating huge data centres. Capitalising on the massive spike in demand for AI, something that’s very close to our hearts here at Hypercube.

You can just see how the stars have aligned and it makes total sense. The way that Apatura are doing it, bringing in really experienced people from across the industry and actually carving a bit of a path and fighting the fight for doing it up here in Scotland as well, where we’ve got an excess of curtailed generation. We’ve got land, like all the appetite’s there. Really, really good discussion.

And really looking forward to seeing like what it is Apatura get up to over the next few years, because I think it’s a really exciting journey and it’s up here in Glasgow with us. If you’re interested in finding out about more of the exciting goings on in the AI, data and energy space, I really do encourage you to join the Beyond Energy Community. It’s a really collaborative environment where we’re trying to collect all the people that care deeply about this stuff and how

Data and AI is really helping drive the energy transition. If you’re interested in that, please check the links in the description for the show or jump onto our website at wearehypercube.com to find out.