Data Centre Boom: catalysts and constraints

Hello listeners and welcome to Plugged In, the Energy News podcast from Montel, where we bring the latest news, issues, and changes happening in the energy sector. Artificial intelligence is driving an unprecedented wave of investment in data centers across the world. Governments are competing to attract new facilities, developers are searching for reliable and affordable power, and grid operators are grappling with what could become one of the biggest sources of electricity demand growth in decades.

But can Europe's power system keep up? Are grids and generation expanding fast enough to support the digital economy? And could data centers become part of the solution to Europe's energy transition rather than simply another challenge? To set the scene, Montel's Italy correspondent Alina Trabottoni joins me first before we hear from three experts at the centre of the debate.

So I'm absolutely delighted to be joined by Alina Trabottoni, our Italy correspondent. A warm welcome to you, Alina. Hi, Richard. Thanks.

Thanks so much. You've been covering data centres, primarily Italian ones, but also more general as well. You've been watching the investment pay sort of close up. So please set the scene for us.

Why have data centers suddenly become such a strategic prize for governments and energy companies alike? Well, absolutely, Richard. As you said, almost overnight, data centers have gone from being a sort of niche real estate story to uh pretty much national strategic assets across Europe. If we just think back a few years ago, electricity was the cheap abundant side of the equation, and the conversation was all about, I don't know, land, fibers, taxes, and that's actually flipped completely.

AI and uh and cloud growth, I mean, I think, you know, that's what everyone's really talking about now, have turned these sites into enormous round-the-clock energy consumers, power consumers, you know, causing a lot of concern as well. A single large one can draw as much power from a small city, and it draws it constantly, it just doesn't stop day and night. And Europe's demand from data centers could actually reach something like well, you know, experts are saying two and a half times today's level by 2035.

Though I'd treat that with, you know, with a pinch of salt. It's indicative rather than precise forecasts. And for governments now it's about, you know, digital sovereignty, AI competitiveness, and of course, tens of billions in investment. So no capital around Europe wants to be seen falling behind.

So are there any concerns that data centers could introduce new risks to power system stability if they're not integrated properly, Alina? Yes, yes, and that concern is now backed by fresh research, not just intuition, Richard. And there are broadly three risks. First and foremost, behavior during faults.

When the grid wobbles, a data center's backup system can disconnect instantly, all at once, basically, rather than taking off the way older industrial loads do. So like factories and things like that. So if enough drop together at the same time, that can actually deepen a grid event and make it, you know, very serious. And the Iberian blackout was effectively a live stress test to this, and it showed resilience came down to electrical design, not to the quality of the quick kids.

And that's really interesting. Secondly, the loads themselves have become active. AI training can swing power consumption rapidly and autonomously. And huge swings can disturb frequency, they can disturb voltage, even when there's no blackout.

And last but not least, the grid has become more renewable and inverter-based and has less natural inertia. So basically, I mean, what that amounts to, what that means, what it translates to is that it has a thinner buffer to absorb these swings. And the encouraging part though is that regulators are very closely monitoring all these developments, and experts are saying there could be some changes. So, for example, towards rules requiring large loads to ride through disturbances rather than trip off, so you know, switch off, or on-site batteries with grid forming capability that could turn a data center into a stabilizer rather than a stressor of the system.

And these are all huge game changers. And the real shift is actually from treating these data centers from being passive demand to transforming them into active participants in the system. And I'm sure we'll be hearing more about that in this episode. But you've been reporting on data centers for a while now, Alina, and your reporting suggests that it's not only price that's holding back development in data centers, but it's also simply getting grid connections, right?

Richard. You really hit the nail on the head because that's the heart of the issue now. For years the assumption was that, you know, cheapest power wins, but access is a completely overtaken price as the real constraint today. You know, if you look at the legacy hubs, Frankfurt, London, Amsterdam, Paris, Dublin as well, connection queues now run to seven to ten years.

That's a phenomenal amount of time to be waiting for a connection, sometimes even longer. And there's actually a structural mismatch. You can you can build a data center in one or two years, but then the grid reinforcement to feed it takes, you know, anywhere between five to ten in most markets. So at the end of the day, investment flows wherever you can actually plug in, really.

I mean, in Italy, the market, which is, you know, the market I cover most closely, connection times are of around three years, and and there's roughly 30 gigawatts now sitting in the queue. Most of it registered just in the last couple of years. And and Spain and the Nordics are the other winners, of course, for the same reason, room on the grid. Ireland is a cautionary tale.

It froze new Dublin projects, then in December reopened on tough terms, very tough terms, with you know heavy renewable obligations attached, which of course is very good. But the new map that is being drawn around data centers in Europe relies on grid headroom as the attraction point, not on postcode prestige anymore. And and that's really the open question that's the recurring theme: whether Europe can build the grid fast enough to keep this investment here. Absolutely fascinating stuff, Alina.

Thank you so much for setting the scene on the Plugged In podcast. Thank you. So, data centers are no longer just a technology story, they're becoming an energy story too, raising questions about grid investment, electricity demand, system resilience, competitiveness, and the pace of Europe's energy transition. I'm delighted to be joined by three experts in this field.

So Janne Peljo of Finnish Industries, a warm welcome to you. Thank you, and happy to be here. And Elisabeth Cremona of Ember. Great.

Thank you so much for the invitation. And last but certainly not least, Liam Newcomb of GreenScale. And likewise, thank you for the invitation, Richard. Liam, I'd like to start with you really saying and a general question.

Is Europe falling behind in in the race to build data centers? It depends on the viewpoint you take. So we are not building them as fast as some other regions. The question is whether you think that's falling behind or not.

What's most interesting is very recently the European Commission have started to strengthen their position on sovereign capability for cloud and AI. And my suspicion is that that'll probably be the area where we will measure are we falling behind on whether the European um entities are able to build out the capacity that we want. There is concern over whether we have access to the frontier models. And I can see that from a geopolitical and national security perspective.

But from a commercial perspective, that may only put us a few months behind in model development. So I'd be less concerned about model development and more concerned about the smaller sites where we actually deploy AI and support the European economy directly with product that has revenue rather than RD. Janne, what's the situation in Finland? What you know, is there a lot of demand, is there is grid connection cues for data centers?

What's the situation in Nordic it maybe maybe even wider in the Nordic region? Yeah, I think both Finland and and the Nordic region as a whole is is experiencing experiencing a bit of a gold rush in terms of data centers. And I think the boom has pretty much surprised many of us. For example, just only in Finland, there are already over a hundred data center projects that are in various stages of planning.

And of course, not all of them will materialize, but I think that gives a good example of how fast the portfolio of data center projects has been developing over just a period of two or three years. Currently, there are a bit less than 20 data centers being built in Finland, so we're expecting to see large data centers becoming operational within the next two, three years, and and of course, we are hoping to see them also to scale up quite rapidly from their initial capacities.

So, in that regard, to me, it seems like Europe is uh an interesting destination for data centers, and within Europe, especially those locations that can offer competitive electricity that has low emission profile, and especially the grid connectivity that is quite easily attainable, or you can get a crit connection within two or three years, those destinations in Europe are quite flourishing in this trend right now. And Finland's one of them, then Finland is of course one of them.

I would like to say that of course the Nordics are are um an interesting area as of total, but there's one thing that's making Finland stand out. Of course, Finland uh beats other Nordic countries in in ice hockey, but also in grid connectivity. So RTSO Fin Grid has been investing kind of they have front-loaded investments in grid build-out for to account and cater for a massive build-out of renewables. And now that's also something that's that's helping data centers and other electricity-intensive industries to tap into the grid.

We'll we'll come back to that, but Finland is certainly an example of a of a stable, resilient as well as a clean power system. We'll come back to those points maybe later, Janne. But Elisabeth, if I can turn to you. You know, we see this you talk of of a boom, if you like, a boom in demand, but is supply keeping up in parts of Europe?

Is it is there enough supply coming onto the grid to meet the demand for both data centers and other forms of demand? Yeah, I think that's a great question. And it's something that's at this stage only applicable to the US, actually. I think in Europe we don't have a power supply problem, it's an access problem.

So it's something that is at the top of every data center developer's concerns when it comes to where they are going to locate their facility. And as Jano captured it, it's can I connect to the grid and how long is it going to take me to do that? So it's absolutely not a supply issue. I think Europe has enough supply for the pipeline of projects that we have in terms of data centers.

The question is how fast can you connect those data centers and where can you actually do that? And that's an absolute crucial question here. But Liam, I mean, I think we're talking here mainly about AI that's triggering this huge wave of interest in data centers globally. So how much of the current growth in Europe is being driven by AI specifically, and how much is it fundamentally changing the industry's outlook for future electricity demand?

It's an interesting topic. So underneath AI, we still have a growing and customer-sustained demand for the cloud workloads underneath. So while the focus is on AI, that ramp rate we had for cloud is still there underneath, and that that's all driven by existing profitable revenue-earning services. So they're they're very investable.

With AI, we need to break it down into sort of the model development, which is typically occurring in countries that have two positive factors. So one is politically they're seen as good places to develop the IP, and that the enforcement around IP laws is supportive of that. And the second, as Janne pointed out for the Nordics, is that they have cheap power. You want a lot of cheap power to do this.

Elsewhere, what we're seeing is the build of inference starting. So as we go from developing models to actually having paying customers and positive revenue services with real long-term supportable demand, that's where the inference models um arrive. What we're seeing with those is that those are not well suited to going into those sort of 10 to 50 megawatt nearer city cloud availability zones, but rather they're sort of 100 to 300, maybe a little more megawatt, say a medium-sized data center, as Google um interestingly puts it.

How our times have changed. Those are those are simply not well suited to being placed next to a city. So we we tend to go and look for a suitable place in Finland, Norway, Spain, or elsewhere. And and that's creating sort of two distinct development paths because the requirements for those buildings are quite different.

We go to where the power is, we will pull connectivity to them if necessary. And for these inference buildings, the tenants are a lot more willing to discuss uh things like flexibility, whether they need generators, etc. These do not look like the Five Nines availability cloud availability zone buildings in the cities. That's very interesting.

And what about Bitcoin miners and blockchain? Are they out of the picture completely, Liam? From what I've seen, I was I was never in that business, but from what I've seen, certainly some of the big US neo clouds have discovered that their Bitcoin farm is vastly more profitable doing um AI model development or inference. So they've bought some different chips to go in their shed.

Absolutely. So Jannne, I think obviously Finn Finland is is a very important centre here, and and Liam's just mentioned it and and for various reasons that you've highlighted. Governments are trying to compete across Europe. So, how how do you think policymakers should balance economic growth with climate and environmental objectives?

I think the right answer to the difficult question is that we should be able to capture them both. And I think Finland is a great example of doing that. As we have already pretty much decarbonized our electricity system, now we are seeing economic growth actually fueled by electricity and power-intensive industries. Such as data centers in in the first stage, but also hopefully kind of power-intensive industries such as hydrogen and and its derivatives or metals manufacturing.

So we're seeing that once we are able to produce competitively priced, clean electricity, it's a stepping stone and a foundation for for building other sectors that then can generate much more prosperity to the economy itself as well. Absolutely. Elisabeth, I think Ember's analysis has highlighted the scale of grid investment needed for the energy transition in Europe. With AI and data centers now emerging as a major new source of of power demand, how does this change Europe's long-term power system planning assumptions, would you say?

Yeah, I think we're starting with a fundamental paradox here, actually. So when it comes to infrastructure such as power grids, we often speak about long-term planning simply because it takes 10 to 15 years to build a large power line, let's say. But data centers, you're looking at one to two years to establish a new facility. And kind of just to contextualize the numbers that Leo was referencing before, a 300 megawatt data center consumes as much as 300,000 households.

So we are saying that essentially a small village could come online in one to two years in a remote area with no access. So we're talking about really big numbers that can come online very fast in very localized areas. So it's not so much long-term planning that has to adapt, but also the short term. What short-term solutions can we implement if Europe does want to encourage this kind of growth?

Ultimately, it's a policy decision at the end of the day. On the long-term side, I just want to say we know grid investment needs to be stepped up, and it's something that many grid operators are currently struggling with today. How do you kind of have this huge flow of capital while still maintaining network tariffs at a decent rate? So this is also made a bit harder when we are planning for a future in the longer term where we have no idea what's actually going to happen with the data sector, data center sector.

So are we going to see growth increase by three times, four times, twelve times? How do you justify those huge investments if you have no idea what's actually coming? No, absolutely. So Janne, you'd like to come in there.

Yeah, just an example from the the Finnish perspective. Basically, we are already seeing that electrification is the industrial trend of basically this century, or at least the the first few decades of it. So the advice to any poison maker at any country is to massively invest in electrification, which basically means first of all in investing in grids. If you can fix your grids to a good shape to accommodate for for kind of new production and new demand, then you're at a very, very good first stage.

We're currently, for example, the Finnish TSO Fingrid is planning a decade-long investment program which would enable Finland to double its electricity consumption within pretty much a decade. And that's of course a massive investment from the TSO side. But I think that it's the example of the kind of volumes and magnitudes that need to be in place in terms of the uh the ambitions. So countries in Europe will need to kind of update their grids to fit the modern electricity system in the first place and then be prepared to double or even even manifold electricity consumption in in the economies in the years to come as the electro electrification is is ongoing trend in within the industries and and other sectors as well.

I mean these are huge numbers, but you'd like to come in before I turn to Liam, Elisabeth, you'd like to comment on what Janne was saying. Yes, I actually just want to come to FinGrid, the Finnish TSO in particular, as I think it's one of the few, if not the only TSO in Europe that has fixed a fundamental problem within the planning system, and that is the lack of feedback between the grid planners and the policymakers. In Europe, it tends to be the policymaker sets the overall strategy of the country, and then the grid operator simply has to deal with it and find a way to create the infrastructure to support it.

Fingrid, on the other hand, I've actually seen their studies where they chart multiple pathways that the country can take. They discuss that with the policymaker and together come up with the strategy as opposed to it being a one-direction conversation. So I just wanted to point out that as a really great practice example. I think in recent weeks on this podcast, we've really highlighted how how what a perfect place Finland is for the power for the power system and for power builds.

But Liam, what what's your perspective here in terms of are grids keeping up both on the distribution and transmission level and then enabling this transformation of demand moving from traditional to the AI-driven demand? I think I would probably reframe the question slightly because I I don't think it's fair to say that the grids aren't keeping up with a change in their demand pattern that's happened in such a short time frame compared to the cycles over which they are designed and regulated to operate.

So if you have a 10 to 15 year policy and then capital investment timeline, saying you're not keeping up on a three to five year swing is perhaps not reasonable. What what I think we do need to do though, and this comes back to flexibility and so social license to operate, there is no reason why these large data centers, and this is what we're actively trying to do with the sites we're building, can't work with the transmission operators and And bring in large renewables generation, storage, dispatchable generation, other grid supporting assets.

And if we cooperatively plan, and we're doing this actively in Northern Ireland at the moment, we can go and put these assets in a place where they benefit. They improve capture factor from renewables, they significantly improve the business case to deploy additional battery storage. Air grid has really led the policy world with the requirement for your entire import capacity to be matched by dispatchable generation, which you place in the capacity market so you don't push up power costs.

If we get all of those things right, data centers can bring along the network support to meet the timeline rather than us saying the grid operators didn't keep up. We've changed the timelines on them, we need to make a contribution. But also we can be cost neutral or actually drive costs down by putting additional support technologies on the network. And a key thing that I'm not sure many people in the market have realized yet, when we look at the efforts of um Nvidia and the Open Compute AI hosting racks, the capabilities that are being put into the new infrastructures for DC power distribution, etc.

, in the data centers, much as NSOE has asked for, it is a software fix to provide radio change frequency support, demand flex, voltage support, and other key services into the grid. And as the grid is moving from analog to digital, I'm nervous and excited at the same time about moving to a world where it is a software problem, not a hardware problem, to deal with some of these constraints and stability issues. I'd really love to know what Ember thinks of that. Well, we'll come to Elizabeth in a minute, but just just to follow up.

So IGrid and Fingrid are doing many things right. So what what kind of examples should are being set here? What should other, both at the TSO and at the DSO level, what should other network operators do? So I think to Elizabeth's point, what we need is a two-way dialogue with not just the network operators but the government policy people.

So what we see in the UK is we have the AI growth zones, which are a central government policy, but probably due to the speed of development, it is not very well articulated in terms of what NESO should do about that. Should NESO go ahead and release capital or perform planning or create an energy hub and ask solar, wind, battery, data center, and everyone else to turn up and connect there. So we need that conversation running all the way from the tenants through the data center delivery channel, which is ourselves, the TSOs and policy makers, so that we can produce something of net social benefit rather than a series of not very well coordinated projects.

So AirGrid and the CRU in Ireland, Commission for the Regulation of Utilities have done that quite impressively in terms of setting a coherent policy position. Fingrid are way out in front in terms of communication and publishing data, being very Nordic. ENTSO-E are trying quite hard to bring um some common specification and requirement. If we look at things like fault ride-through, all of the TSOs will have it very shortly, but they will all have a slightly different version of it that makes it very difficult for me to produce designs that we can deploy.

So I think recognizing that data centers are an operating part of the network and not simply a customer. And we need to recognise that on both sides. Data centers need to recognise, and I think the TSOs do as well. Yeah, absolutely.

I think that's a very important point. Janne, you you'd like to come in, but then before I turn to Elizabeth, we all want to hear Ember's views on what Liam just said as well. Just to chip in on kind of the role of flexibility in the modern electricity system, that's of course one of the key benefits or ge key attributes that that determines how efficiently you can uh utilize the whole network. So in terms of planning ahead, the network operators should put a very strong emphasis not only on the communications with governments and kind of other stakeholders such as corporations and kind of corporate groups, but also they should be vocal about the demands of future clients and good quit neighbors in the future.

So we've kind of used to the model that there are only a few a few large industrial sites that are providing flexibility to the electricity kind of networks. But in the future, the more flexibility large consumers can provide, and the more efficient networks that you can build and and late the less costs it's per user. So it's not only kind of building long-term projections and communicating and discussing the plants, but it's also guiding investments to be facilitated in a way that it adds to to kind of the overall efficiency of the grid.

Elisabeth, over to you. Thanks. Well, emble up some numbers. So again, just to bring some numbers to the conversation, uh we've had IEA estimates when it comes to flexibility of data centers that say if you just provide 30 hours a year of flexibility, which is nothing, you can effectively double the grid capacity that is available to you.

So this is a let's say a relatively simple way for data centers to just get way more online, which is obviously a big competitiveness element for them. But I do think before we talk about the data centers as a grid helper or a grid asset, we need to talk about the potential grid hazard. And this is something that ENTSO-E has flagged relatively recently. Essentially, their concern is because these data center facilities have very high reliability standards, any minor disturbance on the grid means you can effectively take off at a single second or even less notice megawatts, large megawatts of demand, which would trigger further disturbance on the grid.

So there is this risk of hazard that needs to be resolved before we can see this safe operation, and therefore the second step of then being assets to the grid as opposed to any kind of hazard. So this could potentially, if you're saying this, if it could happen within you know less than a second, then that could trigger not on a scale as we saw on the Iberian Peninsula, but something similar. Is that what you're saying, Elisabeth? That is the concern at this point from ENTSO-E.

And I found it interesting that their conclusion from that was the only way to maintain reliability of the grid in such cases where you have huge data centers in a country, is you have to reduce the amount of renewables on your grid. So data centers are at risk of enforcing res curtailment. So this is certainly an issue that needs to be dealt with early on, especially in those countries where we see big data center clusters. So they're actually contributing, or there's a the concern that they could potentially curtail renewables rather than actually accelerate the generation of renewables.

Is that is will that be a fair summary, Elisabeth? Exactly, but only again because of the perception of risk, nothing in particular that they would be doing, just the risk that they pose because of the potential rapid disconnection. So this is definitely something that needs to be dealt with as soon as possible, as well, because it's a relatively, let's say, simple fix in terms of addressing the operating standards of these facilities. Absolutely.

Liam, I can see you're nodding your head. Well, absolutely. So the the fault ride-through and act post-fault active power recovery as the the common terms if anyone wants to go Google for it. Air grid, around 20% of their grid is currently data center load.

And as Elizabeth says, when there's a short transient event on that grid where your lights might flicker, hundreds of megawatts or gigawatts of data center will have the same voltage quality threshold set on their uninterruptible power supplies and will all just disconnect from the network in milliseconds. That can cause generation to be tripped as voltage rises, frequency rises. Unfortunately, 30 seconds later, those uninterruptible power systems decide the grid's stable again and they put all their load back just when you've emergency tripped.

So it is a significant issue. And so this is one of the areas where AirGrid has led. So both of the major sites we have in development at the moment are designed to be fault right-through compliant. They will stay on the network through the voltage fault curve and they will recover in the timeline specified by that local TSO.

This is one of the areas, though, that I think it would be very helpful if ENTSO-E is able to bring together um each of the local country TSOs and give us a little bit more standardization in terms of the requirement because we can then apply that back to our supply chain of vendors for electrical and mechanical equipment. But yes, absolutely. It is in our best interests to be connected to a stable, reliable grid. And because we are no longer small customers, but we are now large loads, we need to take a step up and join the club of the grid-connected players and act like a grid-connected player.

Absolutely, rather than just an off-taker. Absolutely. And I think that's that's obviously incredibly important. We don't want a repetition of what happened in the Iberian Peninsula.

No, absolutely not. Janne, I'd like to round off by a little bit talking about public acceptance. Now, I think there often there's been a lot of opposition, local opposition to data centers that are mainly you know driving AI and and cloud computing. How has Finland dealt with that?

Or what's the situation in Finland? I think the short summary is that a decade ago we were very excited to have the first Google data centers kind of you know flagship of of digitalization, more jobs and kind of economic activity. But then fast forward to today, there's an increased skepticism over data centers and whether those are actually beneficial for the economy or not. I think the standard joke is that that data centers are something that comes from the outside and they steal our electricity and they steal our electricity grid connections.

They don't bring any local jobs, and if they bring any local jobs, the local jobs are poor and bad quality. So why would we have them? But I think the popular discussion in Finland has been become luckily much more nuanced. People are starting to understand that these are actually quite large investment projects that are stimulating the local economy as they are being built.

And at the same time as data centers are increasing electricity demand, then new PPA contracts for electricity supply are being agreed on, which then again stimulates new electricity production investments, which has been stalling in Finland because of some of the overcapacity in electricity production. So I think we've gone from the one end to another, and now we're slightly going back to kind of the centrist tone of discussion. But I think what data center sector, at least those operating in Finland, what they need to improve in is the transparency and communications on what they're doing and what are the benefits of of their activity, what data are they processing and and what are the the kind of local and and also kind of global implications of of their activity.

Yeah, that's absolutely crucial, I think. And Elisabeth, you know, but there are other environmental concerns. I mean, that you know, for example, the use of water. How you know the data centers do consume a lot of water.

How do you deal with these kind of concerns as well in the public mind? Yeah, I think we've seen public acceptance, as Janne was saying, becoming a big concern, and it's certainly about consumption of resources in general. Ultimately, it's a policy decision if you want to encourage data centers in your country. We're here talking about it from a grid perspective, but there are many other elements that can attract or detract data centers.

When it comes to flexibility, which is one way you can just really increase the amount of data centers your country can physically withstand, I think that's something that the developers themselves want to see simply because it's an easy way to circumvent the grid capacity crunch that we have in Europe. So that is something I expect to see actually coming from them. But then what we would want to see on the policymaker side are things that can incentivize the best use of resources within a country.

So, for instance, at the moment the EU is discussing I've forgotten the exact word, communication on network tariffs. I can't remember exactly how that's called. But something that we can do is have connection fees that actually reflect the conditions in which you are choosing to locate your data center. So are you choosing to locate it in an area where there's a huge amount of grid capacity, great water resources, so on and so forth?

If so, then your grid connection fee is really low because that's where you want the data center to go. If you are instead choosing to connect it in a very difficult to access area with few resources, then your connection fee should be higher. And this is a way that you can send the right signals to these investors and effectively allocate your resources more efficiently. That sounds a bit like nodal or regional pricing by the back door in a sense.

Only on connection fees. So no. Only on the connection fees. Only on the connection fee side, a single one-off fee.

Perfect. Thank you for the clarification there. Liam, if we're talking public acceptance here, how how does one win over a sceptical public public? First off, on the water topic, because it has become a popular item of discussion for very good reason, but there is a disconnect between the US and Pacific rim designs and what we build in Europe for various reasons, largely Legionella limited land availability.

In Europe, evaporative cooling, where we actually evaporate water to do the cooling, to avoid spending power on compressor cooling, is really not very popular. One of our sustainability commitments is we're simply not going to build data centers that use water, consume water for cooling. The water we put in the pipe the day we commission the building should still be in that pipe 15 or 30 years later. It's very clean, it's well treated, it's filtered, it's expensive, we want to keep it.

So yes, it's a valid concern. No, we don't need to use water to run a data centre anywhere in Europe, and particularly not in the arid or water-stressed regions. It's simply not necessary. So there shouldn't be any argument that hotels can't fill their swimming pools in Spain, but data centres are evaporating water.

That simply shouldn't be happening. But the other elements are, as Elizabeth indicated, that they're around the broader local social benefit and acceptance. We've seen a number of countries move to a policy position where people living near large wind farms get cheaper power as part of the benefit of having that wind farm placed in their local area. Unfortunately, because of the fairness regulations that are baked into most TSO's um operating codes, it's actually very difficult.

For example, for a data center with large behind-the-meter renewables to spill any of that when there's an access to local consumers. We're not allowed to interfere in the market in that way. And you can only through circuitous processes in balancing markets and other things produce something like that. But what we need to do is understand and this comes back to the policy to TSO to data center developer connection and discussion, we need to understand for each region.

So if we're in a residential or commercial light industrial region in Europe, perhaps the smart thing to do is heat recovery and combine that with a local government policy to put the pipes in the road. We'll supply the heat for free. That that's not a problem from the data center. It's not a significant revenue stream for us.

And we can hopefully help make the case for a district heat network. In other places, though, that's going to be much less valuable. If we go to Norway, for example, the Nodges price is so low, it's very hard to make a case for heat pumps or heat recovery. But the way that Nodge's price is structured, where it's targeted back to domestic households, that's a pattern I think we could use in conjunction with the regulators.

So if the data center, AirGrid again are leading here and requiring 80% of the total megawatt hours consumed to be purchased from net new renewables. I'm a little concerned on the price impact and investability for other renewable projects from that. But if we're able to leverage that, when we push the price down, inevitably by doing that, to provide cheaper electricity to people in that region, then I think we're going some way to getting the social license back. And perhaps we're not the biggest industrial employer, but if we make energy-intensive industries substantially more economic and viable by feeding them free heat and then driving down domestic or commercial power bills, then we drive secondary social benefit.

And that's really where we're trying to get to as Greenscale, but we really need a connected policy, TSO, local government and data center approach and strategy to do that. I think the con connectivity aspect here is absolutely crucial, getting not policy makers, TSOs, data oper data center operators, but also local communities involved in these kind of decisions and this kind of communication. Unfortunately, we've run out of time. I could continue for hours talking about this.

This is a fascinating topic. So, but we have to close here. So, Elisabeth, Liam, and Janne, thank you very much indeed for being guests on this week's Plugged In Podcast. And to you listeners, thanks for listening to this episode of Plugged In.

If you enjoyed this discussion, please like, rate, and follow to make sure you get the latest podcast episodes as soon as we release them every Thursday. We'd also love to read your reviews of the podcast. It helps us to keep up to date with what you, our listeners, think of the podcast and what content you want to receive more of. Finally, you can head to montelnews.

com for more news and analysis from our team of journalists across Europe and beyond. See you next time.