The impact of the transmission network on distribution demand applications

00:00:48 – Pete Aston
Hello and welcome to another episode of the Connectology podcast from Roadnight Taylor. I’m Pete Aston and I’m joined by Philip Bale and Kyle Murchie to discuss the impact of the transmission network on distribution demand applications. So, hello to you both.

00:01:04 –Philip Bale
Hi Pete.

00:01:04 – Kyle Murchie
Hi Pete.

00:01:05 – Pete Aston
We’ve, all of us, all three of us have worked in DNOs before and done demand applications. Many demand applications we processed from the other side of the table, and we’ve also been playing with demand applications for the distribution network as consultants as well. So we thought we’d bring you an episode on how the transmission network is starting to impact on applications for demand connections at transmission level. So it’s becoming an increasingly difficult situation, but before we look at the impact that the transmission network is having now on demand applications, we thought it was worth having a look at the history and looking at how demand applications have traditionally worked on distribution networks.

Firstly, Philip, if we could just come over to you, can you just talk us through how demand applications have traditionally worked on the distribution network in relation to its impact on the transmission system?

00:02:03 – Philip Bale
Yeah, absolutely. So, I think for the vast majority of networks in the past, transmission hasn’t been a bottleneck that ultimately the DNO has assessed the demand connection on their network, and then ultimately, they’ve taken a view as to whether it has a material impact on the transmission system or not and ultimately, they submit any forecasted demand growth in their week 24 demand forecast. Ultimately, they’ll then forecast out what they expect the demand to be at that specific GSP and no specific caveats have gone in offers to say that it’s subject to any transmission reinforcement. It’s effectively a – we’ve accepted your project and we’re telling National Grid as an FYI in terms of going through. So that’s historically how it’s been for most GSPs.

00:02:51 – Pete Aston
Could you just unpack that a little bit more. Can you explain what the week 24 demand forecast is? Because there might be some listeners who don’t really know what that’s all about.

00:03:01 – Philip Bale
It’s a process the DNOs as part of CUSC have to go through and forecast out the demand for each of the different GSPs. It’s called the week 24 demand forecast because that’s when it’s actually done and submitted across to transmission and it’s a forecast of what they expect the demands to be on the networks at those GSPs going forwards.

00:03:23 – Pete Aston
And, from my experience, those demand forecasts that the DNOs did, and there’s a 10-year forecast that goes into the week 24 – so it’s a 10-year view across all of the GSPs across a particular license area. Those demand forecasts typically include an element of accepted to connect demands. It’s the DNO going, we know that there’s a certain amount of schemes that are accepted to connect, and we believe that this batch of accepted demands is going to result in a certain growth over that 10-year period.

00:03:55 – Philip Bale
Yes, and each of the different DNOs have got different methodologies that they use. There’s two main different methodologies that’s used for demand forecast elements and that has a aspect of existing demand in there and also demand increase. It’s probably also fair to say that the demand forecast that goes forwards to national grid doesn’t just add on all of the accepted demands the DNO has, there is a certain element of diversity that comes from them that goes into the network, so it’s not just the case of – we’ve had an extra 10 megawatt connection, I’m going to increase the demand on the GSP by 10 megawatts, going forwards.

00:04:33 – Pete Aston
So yeah, so these demand forecasts are building in a certain amount of diversity, they’re building in a certain amount of attrition, assuming that some projects are going to fall away. So it’s very much a sort of finger in the air basis, but historically that’s only been, it has been a forecast, hasn’t it, this? It wasn’t scenarios, it was a, literally a – this is the forecast for what the demand is going to be across the 10-year period, and I know from experience that wasn’t always very good. So, I think for most of the time when I was in the DNO, it was always forecast to be a demand increase, but for years and years the demand just kept reducing, even though increases were forecast; so definitely wasn’t a foolproof process.

00:05:13 – Philip Bale
No, and I think that’s something that we will have come from different areas, where my region that I was looking after did continue to have demand increases. Obviously, it gets affected by things such as COVID and energy efficiency and big customers dropping away, but on the whole, there was still demand increases occurring in those regions and a lot of that was being driven by new demand connections industrial, commercial customers.

00:05:37 – Pete Aston
Yeah, you came from the busy East Midlands patch. I was from the slightly more sleepy Southwest patch.

But in terms of, if you were, let’s say, a housing development, a large housing development, let’s say five megawatts or something, usually speaking, if you were going to put in an application as a housing developer of five megawatt increase, you wouldn’t have expected any transmission impact. Was it, is that a fair point?

00:06:04 – Philip Bale
I would say, apart from in certain areas, I had one of the GSPs that I was looking after that always had very tight issues and it was something the DNO was battling. Often the customers had no visibility of the battles the DNO was doing in terms of trying to resolve transmission borderings. But I agree with you, on the whole, a five-megawatt demand increase from a housing developer would have no visibility of the transmission interface.

00:06:28 – Pete Aston
So, from a point of view of a grid supply point needing reinforcement, there would just be gradual load growth in the background. At some point, the DNO would trigger the changing of the super grid transformers or request an additional super grid transformer to increase capacity at a particular grid supply point. And most distribution demand schemes wouldn’t see any impact, or cost, or time scales necessarily from the transmission network.
You were talking to me a bit earlier, Philip, about a few schemes that you were aware of where that wasn’t the case, where that, where there had been a transmission impact for a few distribution connections. Do you want to just go through a few of those examples?

00:07:07 – Philip Bale
Yeah, I can indeed. So, I think the general process the DNOs follow is when to charge reinforcement to customers and when to socialise reinforcement and ultimately, if you’ve got small domestic demand growth, it doesn’t feel like the right message to say that they should be funding a 12, 14-million-pound asset that would create usually 240, sometimes 460 MVAs worth of new capacity. Where you get very significant large demands, then it is the right thing that they should be contributing to new super grid transformers rather than being socialised.

In my past I had a large distribution customer who was a large rail site that came through that required a new super grid transformer. That was an asset that was funded fully from them because of the impact that they had, there wasn’t enough spare capacity in their networks. I did have another site where at one stage, we were exploring whether any demand above one megawatt would need to contribute towards an SGT and ultimately, fortunately, the SGTs could be re-rated to push the problem out slightly longer – but it is still a problem that’s coming through of how do you resolve the demand capacity at one or two GSPs.

00:08:29 – Pete Aston
And it’s interesting that the sort of lack of reinforcement at grid supply points over the years is just testament to the fact that a lot of GSPs have been overplanted at some point in the past. I guess you could either call that a bad thing, too much money has been spent in the past, or a good thing that people were just planners in the past that just built in lots of capacity predicting lots of load growth, but for I guess for decades there’s lots of places that have been living off the fat of capacity that had been installed. You know…

00:09:02 – Philip Bale
On the whole, agreed, I think if you take the East Midlands as an example, there was specific issues that the DNOs had with demand growth where the solution ended up being new GSPs coming through that resolved issues, two to be an example of Bicker Fen and Stoke Bardolph are two brand new, two transformer sites that were nominally put in to resolve demand non-compliances as a result of growing demand and that’s then effectively carved off demand chunks, freed up capacity across wider areas. One of those examples is very urban, the other is very rural, where it then just became, there was a hole in the network and an inability to supply the demand.

00:09:48 – Pete Aston
And are you aware of how they were funded? Was that DNO funded works?

00:09:53 – Philip Bale
Both of those were DNO funded projects in terms of going through. So, yes, they were trying to remember the old runs DPI, ERP…

00:10:01 – Pete Aston
DPCR.

00:10:03 – Philip Bale
DPCR. That was it.
Yes, DPCR project funded projects.

00:10:05 – Pete Aston
So essentially the DNO would have applied for funding from Ofgem within their price control submission, for whatever five-year period it was in question, and would have got funding for those and built those GSPs out in association with National Grid to provide that capacity. So it was effectively socialised reinforcement rather than …

00:10:24 – Philip Bale
Paid for over exit charges over a very long period.

00:10:27 – Pete Aston
So that’s the historic impact, historic situation of how sort of demand applications have generally gone in the past. I think the situation now is very different, I would say, and accelerating in that, in how our problems are coming on the system in terms of demand applications.

So, one of the key impacts has been the growth of battery schemes over the last five years or so. So, do you want to just talk us through, Philip, again, how the growth of battery schemes has had an impact on the sort of final demand connections on distribution systems?

00:11:08 – Philip Bale
Absolutely. So, I think it’s one of the challenges that we still exist as the industry is that battery storage schemes are still being considered as having unrestricted import capability, whilst they are starting to be looked at as non-firm generation, which means when the network is operating abnormally, they can be switched off. Under normal network running, a battery storage scheme can charge whenever it wants, not restricted by its time level, which means that the DNOs and, as such, the transmission system operator, network operator, are studying these batteries having the capability of importing during system peak.

Part of the reason why they’re still doing that is because there are some examples where battery storage does charge during system peak and that’s typically come from battery storage operating in the ancillary markets, frequency response – there is a big dip in frequency during the system peak demand times, the battery storage exports to try and resolve that, but after it’s normalised frequency, it then recharges again to be ready in case there is another event that’s coming through. So there is a history of battery storage schemes charging during winter peak, and as a result there has been a significant increase in battery storage across distribution networks, and we are now finding that most GSPs are having new super grid transformers being triggered and a significant number of them are having super grid transformers triggered because of import reasons, because commercially, battery storage schemes have the capability of importing whenever they want under normal conditions.

00:12:56 – Pete Aston
And I guess the other impact as well Philip, is that whereas a lot of demand applications are typically less than 10 megawatts, aren’t they? There’s sort of applications that are coming in the one, two-megawatt size maybe a five megawatt housing estate, maybe a five to 10 megawatt commercial park or something but batteries are coming on in 10, 20, 30, 50 megawatt chunks aren’t they. So, it’s having a, that’s having had a really profound impact on using up that capacity really quickly.

00:13:25 – Philip Bale
Yeah, it’s very normal to see 99.9-megawatt battery storage schemes within GSPs. Some GSPs will have multiples of those going through. So, it’s a significant point load on a system, especially when you consider that a two transformer, 240MVA substation, it only has 240MVAs worth of capacity. A 99.9 MVA battery is using a significant chunk of that.

00:13:53 – Pete Aston
And we have heard of some DNOs starting to look at this a bit differently, starting to take some different assumptions now, to go, maybe not all of the batteries are going to be importing at system peak, so therefore maybe we can, if you’ve got 100 megawatts worth of batteries on the grid supply point, maybe actually you might only assess it as 90 megawatts import, or 80 megawatts import, or something under system peak, but that seems a little bit more hit than miss. I think DNA policies are probably in development at the moment as to how they really are viewing that.

So other tech Philip, so, apart from batteries, are we seeing any other trends in different types of demand that are taking up large chunks of capacity?

00:14:35 – Philip Bale
We’re seeing significant growth in the hydrogen electrolyser market, so people using electricity to often take water and split the water into a hydrogen atom, that then gets used for other purposes, which is classified as a final demand site.

There has been significant growth in the data centre market, both the traditional data centres and also the data centres very much focused for AI purposes. As well as that, we’re also starting to see some smaller but more significant numbers of EV charging capability. So, while still usually in the hundreds of kilowatts, megawatts scale, when you start adding those up in urban areas, they do start to increase, along with the industrial commercial site, the manufacturing sites as well. So, there is a strong forecast of demand growth, some of which linked to battery storage, some of which linked to the decarbonisation of other sectors. In terms of coming through, one of the big things that we’re also seeing with industrial commercial sites is where you’ve got historic manufacturing or commercial processes that use lots of heat, and they typically use gas as that mechanism, converting across to electric significantly increases their electrification import on the sites and can trigger significant demand growth, often in areas where their existing supply is relatively small.

00:16:05 – Pete Aston
Sometimes we’ve seen the sort of sites that are looking to decarbonise, their demand increase is going what? Five to ten times what it is at the moment.

00:16:14 – Philip Bale
I think yeah, we’ve seen a site where at the moment they’re an 11kV connected demand less than five megawatts. When we looked at their decarbonisation process it was over 100 megawatts worth of import – a dramatic change. And there are numerous examples of those.

00:17:02 – Pete Aston
Let’s move on to then the impact that’s now having on other types of demand connections, so the sort of typical housing estate. What are we now seeing Philip, in terms of those smaller demand connections, coming through and basically the impact on the transmission system, of those smaller demand connections?

00:17:23 – Philip Bale
So, I think we are basically saying that ultimately, the view of the transmission system as an infinite bus for small schemes has now stopped. We are starting to see different DNOs assessing their limits for their networks differently. Ultimately, there is a lot of published guidance on this, we understand from our engagement with the different DNOs, there are different internal guidelines that they’re using of when they’re writing in demand offers, saying that it could be subject to a transmission impact assessment, and some of those are as low as a megawatt. Others are tens of megawatts. But anyone now applying for a scheme that historically would have just hit on the network for the last 15, 20 years in 99% of locations. Now I think we’re starting to see that the vast majority of them will end up having impact assessments. Where there’s significant demands, we’re still not expecting the one or two houses to come through, but where it is thousands of houses, we’re going to start seeing offers come through with caveats of being subject to that transmission assessment.

00:18:33 – Pete Aston
And when we’re talking about transmission assessment, we’re talking about MOD Apps here, aren’t we, Philip? Which are chargeable to those customers who are triggering them.

00:18:41- Philip Bale
They are, and also then they’ll be considered at the same stage as everyone else. So it’s fairly normal for us, when we’re looking at battery storage schemes, to expect that those projects will have 2037 dates, 2038 in extreme scenario. If it is classified as a material scheme on the transmission system, you could realistically see projects having to wait for brand new GSPs, having to contribute to those new GSPs. An example of a project that I was looking at only this morning, the capital contributions for projects are in excess of £40 million being shared across projects and that’s one of the scenarios increasingly we are going to see – final demand customers industrial, commercial, large housing estates could be expected to contribute towards significant transmission works, SGTs.

00:19:40 – Pete Aston
And that’s highly significant. And I guess the other thing that follows on from that is the requirement to put up securities and have cancellation liabilities associated with those transmission works as well, which again, is not something distribution demand connections have often had to face and we won’t go into it now, but the final demand connections have a different way of having their cancellation liabilities assessed, which is far more onerous at present than if you’re a generation scheme. So those securities and cancellation liabilities can be very high and I think, Philip, sorry you were going to jump in there Philip.

00:20:17 – Philip Bale
Just to add, UKPN published recently that they had a scheme that had over a billion pounds worth of securities and liabilities for final demand projects to their network – so the numbers are extreme.

00:20:32 – Pete Aston
And there is a CUSC Mod out there CMP417, if you want to go and have a look on Ofgem’s website which is looking to try and CUSC Section 15, requirements, which is basically what the generators currently have applied, which is a sort of pro rata approach to securities to apply that same sort of methodology to final of demand connections, which probably isn’t due to kick in until towards the end of 2025 or the earliest, but keep an eye out on that one.
And I think one of the other interesting elements about final demand connections and transmission impact is the fact that generation schemes are able to utilise technical limits in lots of places to get an accelerated connection. So basically, the generators connect under active network management and can just have curtailment but can connect prior to the transmission reinforcements happening, whereas if you are a final demand customer, if you’re a data centre or a housing estate, or something, you are not flexible enough to take advantage of any sort of active network management. So you’re effectively penalised compared to generation schemes, by not being flexible enough to be able to accommodate that. I think the only exception do you think, Philip is hydrogen, which probably is, or possibly is, flexible enough to be able to do that sort of thing?

00:21:52 – Philip Bale
Yeah, absolutely, I think it is. There again, it’s still trying to get that over the line as being a solution that people can utilize in terms of going through. The only other point I was going to make is in the West of London there has been an example where, whilst they couldn’t partake in technical limits, there was the agreed one megawatt per year ramp up solution for demand projects, which was a way of basically trying to help some specific projects bigger than a megawatt less than 10 megawatts in total that had a modular increase in capacity that could go through the connect ahead of some very long, lengthy transmission delays.

00:22:30 – Pete Aston
And I think it’s just worth picking up on that example, because that’s the one that’s often rolled out of how the transmission network impacts on demand. It was across West London. Was it about three grid supply points? Three or four grid supply points either, GSP being the sort of main one where data centre connections in and around Slough had exploded, Slough being the data centre capital of the UK and resulting in massive delays. I think at one point I think it was SSE had said nothing more than 250 kilowatts can connect to their system. Mayor of London got involved and then after that, SSE and National Grid changed their tune and then implemented that one megawatt option that Philip was just talking about.

So that is a present issue. That’s not theoretical, that is actually happening. Now I’m aware of another scheme that I’m working on, which is a seven megawatt sort of commercial development, high-tech commercial development, looking to get to a GSP being told that it would have to wait till 2037 for a connection because of accepted battery capacity at the GSP in question. So these things definitely aren’t hypothetical, they are happening right now.
Kyle, I know you’re sitting in the background. Maybe talk briefly around the situation in Scotland because you’re doing some quite a lot of work in and around Scotland and the situation there’s a bit different with the different voltage level of transmission can talk us through what that’s looking like that?

00:24:00 – Kyle Murchie
Yeah, absolutely. I think, as we’ve just discussed, there has been a little bit of headroom, a little bit fat in the system for decades from a demand perspective and while that might be true as well, in in the north and Scotland you’ve had a lot less in terms of natural requirements and natural growth just because of the, the population density etc. But with that increasing and in certain areas just general load growth plus, plus a need to decarbonize industry, of which there are quite a number of significant industries kind of positioned in relatively remote locations in many cases that are definitely driving up the needs.

And of course, as you just said, with 132kV being transmission voltage, it does mean that GSP capabilities are using up quite quickly. There’s another aspect as well, as further north you go there are an awful lot more single transformer type GSP arrangements and the firm element is provided through a bit of cross-linking across the networks. Of course by doing that you inherently have a bit of a limitation quite quickly in being able to provide kind of firm connections for much larger demands very quickly.

So I know one thing that’s being looked at is whether substitutions that may be going and going in at transmission level for wind farms could potentially be utilized. Could that become a new shared infrastructure site? That’s definitely an aspect that’s being considered, trying to utilize the infrastructure that’s already gone in. But another challenge, and maybe one that’s driving some of those conversations, would be the island networks. The island networks are pretty limited. Isla, for example, is, I think it’s 10MVA in a total capacity which in any in anyone’s scale is pretty small or anywhere, never mind, but not an island that has industry. And you’ve got those traditional situations where you don’t have any transmission network on those islands to begin with. So the question isn’t just what you’ve structured you need at distribution or transmission level. It’s do you bring transmission to these islands or do you stick with distribution, but what, effectively, is the most effective solution?

00:26:22 – Pete Aston
Absolutely and, Philip, from your point of view, are there things that can be done by developers looking to put in final demand schemes to minimize the impact of their applications on the transmission system, to avoid it in some situations?

00:26:41- Philip Bale
That’s a good question. I think, ultimately, what we’ve been advising people to do is to try and be aware of the current restrictions in specific areas, aware of which areas have got greater restriction than others and ultimately, the easy answer at the moment is to try and fly under that radar, try and avoid being part of that issue. Where there are sites that have issues, that it’s not possible to try and stay below that radar, ultimately, we’re having to work with the DNOs to try and find ways of getting those projects connected as much as possible. I think there is a greater industry conversation needed to be had that ultimately, I think we could probably all agree getting to a scenario where battery storage has triggered new GSPs in half the country, a generation triggering new GSPs in the other half of the country and every existing GSP having a minimum of one new GSP plant, which is unfeasible for National Grid to go through and fund as well as plan and construct and operate. It does need to be a very serious conversation around SQSS, around types of demand that don’t need that level of security.

So we’re talking here about battery storage and around hydrogen generation and ultimately working out how to assess it and not holding it to the same standards that we would for housing estates and industrial commercial sites, because ultimately, having a network that is constrained for demand, having a network where demand cannot connect when it needs to, will hurt GB PLC. It will stifle the economy and that’s not good for the country, and also it will reduce our decarbonization.

If people are trying to decarbonize their site, weeding themselves off of gas only to be told that they can’t connect until 2037, 2038. And ultimately to decarbonize our energy system means around increasing the demand which will then utilize the renewable energy that’s coming onto the system. So a lot of it is chicken and egg also. So I think there is things that can be done, that we are doing, but there is also a much bigger conversation around how to try and fix this at a much bigger level, which I’m not certain is being had at the moment so.

00:29:19 – Pete Aston
So it sounds like what you’re saying, Philip, is that demand is crucial to decarbonization, or maybe as important as the generation side of things. But the generation side of things are being given a bit more prominence at the moment, and that brings us on smoothly to our last point that we want to talk about, which is Connections Reform. As we all know, Connections Reform is going full steam ahead. At present, Kyle is sitting on work group after work group looking at how things are changing week by week, but Connections Reform is due to start – the current start date is now June 2025. Maybe it will be after, who knows, delayed again, but at present, Connections Reform, Kyle, is mostly focused on generation. Certainly, at transmission, it’s generation and demand isn’t it, but at distribution level, it’s just generation?

00:30:12 – Kyle Murchie
Yeah, that’s absolutely right. Probably worth saying it’s technically Q2 2025, but we are expecting, based on what that timeline looks like, that’ll be June, as you say. Yeah, so one of the reasons why demand at distribution level was excluded is a simple one that is quite complicated and it wasn’t felt that it had any impact on the modifications themselves. The modifications were transmission process modifications, so it had to include demand at transmission level, generation at transmission level, and the thinking was that generally the issues that we see at grid supply points with new infrastructure required are being triggered by generation, embedded generation. Therefore, the embedded generators and storage included in that need to be included. So that was the initial reasoning for excluding demand at the distribution level and in theory there are certain benefits because by excluding demand at generation level then does that allow those projects to move forward more freely. They can get a DNO connection. They don’t have to worry about going through the gated process. If they do trigger upstream works and have to go through a Mod App. It’s going through a Mod App as per the current process. It’s not going through a gated process as it’s been set out under reform.

But there are a few challenges because, while hypothetically at a really high level, that all sounds great when you start looking at some scenarios which, as we’ve been discussing before, these are the sorts of things that we are we’re looking into quite a bit of detail on, and you do have a number of unintended consequences with the process as it is and as it’s been finalized in legal text and then trying to apply that to real projects Today, you’re going to have effectively, from a distribution perspective, two sets of modification applications, therefore two queues – a demand one and a generation one. A battery would fall under both, but they’ve got different timelines in which they operate and slightly different rules. The gated process, the gate one, gate two idea would only apply to the battery’s generation element, not the demand. So could we get into a situation where projects that are wanting to move forward in industrial decarbonisation are sitting behind batteries that effectively aren’t moving in the queue? That’s been rectified from a generation perspective, but it’s, but there’s no mechanism to do that in the demand side.

32:49 – Pete Aston
So if, so are you saying that if a large number of batteries were moved out the queue and effectively moved into a gate one position, couldn’t meet gate two – as we’re expecting, quite a lot of batteries won’t be able to meet gate two, or, if they aren’t needed, at least under the Clean Power 2030 targets – are you saying, then, that if they’re not in the transmission queue in terms of generation point of view, they might still be taking up capacity from a demand point of view, and so therefore, even though they don’t have a place in the queue, they’re still blocking demand connections from coming forward?

00:33:28 – Kyle Murchie
Yeah, so based on the two Mods CMP434 and CMP435 and the methodologies that will be coming out alongside those. That particular example would never be covered because it’s a distribution situation. So, it is possible that the DNOs, alongside the ENA, can resolve that by themselves putting a link in between the gated process and the non-gated process for demand, so they could effectively manage that. But at this moment in time there’s no mechanism put forward that’s going to achieve that.

00:34:01- Pete Aston
Okay, it feels like there’s a long way to go on this issue, but it’s been really interesting, Kyle and Philip, to talk through this issue of how the transmission network is affecting demand connections on the distribution network. I think that we’re going to see a lot more of this coming to light over the next six to 12 months, I suspect, with further restrictions on grid supply point capacity. It’s going to be really interesting to see how it all pans out. But, as ever, thank you, Kyle, thank you, Philip, for your input. Thank you everyone for listening and we hope you join us for our next episode of the Connectology podcast. Thank you very much. Goodbye

00:34:40 – Kyle Murchie
Thanks Pete. Thanks everyone.