Loading summary
Transcript116 lines
- [00:00]
A
I'm your host, Ed Porter. Welcome back to Transmission. You might think that the grid connection queue in Great Britain is first come, first served. It isn't. Connections of form was meant to bring clarity. Instead, projects that were declared protected and pushed to the front are still missing their connection dates. As new issues have arisen with contract disputes now storing projects. This really matters. It hits what gets built and if we can deliver close to clean power. 2030. Katherine Cleary, a two time guest electrical engineer and grid connection consultant from Roadknight Taylor, has watched that misconception cost developers time and money and has the insight into what's actually determining who connects and when. If you want to find out more about how grid connections are affecting battery build out globally, sign up for free to access CO Moto Energy's AI analyst. Let's jump in. Katherine, welcome back. It's been a year and a half since you were last on transmission and we're excited to have you back again to talk about grid connections.
- [01:13]
B
Thanks very much for having me, Ed. It's a pleasure to be back.
- [01:15]
A
Okay, let's dive straight into it. So what is one thing that people get wrong about the grid connection queue?
- [01:20]
B
I think that the. Oh, am I allowed two things?
- [01:24]
A
You're allowed two things.
- [01:24]
B
Okay, so the first one is a really quick one. I think people sometimes think you join a grid connection queue, you submit your application, get an offer and that's it. And so someone else magically takes care of it and your connection gets delivered. And I think that that illusion has been shattered over the last couple of years. People, people get now that what can go wrong with a connection, the sort of delays that the work involved in actually delivering it.
- [01:42]
A
Because people often say, I've got a
- [01:43]
B
great connection, I've got a great connection. What they mean is I've got an offer, you know, I've got an offer that says in 2033, you know, if a lot of things go right, we'll, we'll have something and I get through
- [01:51]
A
all of these and I provide the capital, then it can actually help.
- [01:54]
B
So I think kind of oversimplifying it is probably a pretty common, common error. I think the other thing which has really cropped up in the last sort of year or two is that I think people have sort of assumed that a grid connection queue is a list of projects in the order they connect and it's not. And this is a really important point, you know, if, if you apply for a nuclear power station that's going to be located in Orkney and I apply for a 20 megawatt battery project that's going to be located in Somerset. You know, I can have applied two years after you and I should be getting a connection date well in advance of you. You know, we're going to need to build an awful lot of infrastructure for that nuclear power station. So actually, you know, the connection queue is an order in which you applied, but it's not necessarily going to be the order you connect. And that used to be really obvious, and I think we've lost that over the process of the queue becoming so enlarged. We've got so many people in the queue that actually we've ended up with this quite bizarre scenario where maybe the 20 megawatt battery has to wait for some of the same reinforcements as the nuclear power station.
- [02:53]
A
Do you think that's just a British thing? Do you think we're just like. So we're so embedded the concept of Q so much that we just. We will happily wait behind the nuclear power station until it's our turn?
- [03:03]
B
I think. I think there is something in that. And even the phrase first come, first served, which we used to describe queues, kind of suggests that, you know, I'm getting first wait for me, you know. Yeah, yeah. And that's. That. That is not technically true, you know, So I think. I think hopefully we'll see an output of connections reform that beginning to be reflected a bit more, you know, that actually how quickly projects can be accelerated or maybe just not delayed is going to depend on what type of project they are and where they are in the country.
- [03:28]
A
And there are some funky loopholes in this. Right. So if you have like a. Is it 4.9 megawatt like project, you could add that to an existing site and you don't have to kind of go through the very formal part of the queue.
- [03:39]
B
So I guess we have a transmission queue, you know, which is what we're really talking about when we're talking about connections reform, the big stuff and the big delays, you know. So, you know, If I'm a 20 megawatt battery project, I still at the moment have to wait for. For maybe upstream transmission reinforcements. That 4.9 megawatt number that you mentioned is the kind of the golden limit. If I'm less than 5 megawatts, if I round to less than 5 megawatts, then I don't need to wait for those transmission reinforcements. So, yeah, there is a threshold, a de minimis threshold.
- [04:05]
A
You can kind of sneak in a little bit with those smaller projects. Okay.
- [04:08]
B
There are some caveats. So it's not a golden rule for
- [04:10]
A
everything, but okay, okay, great. And then last time you were on connections, reform was still a design exercise and you laid out what you thought a good outcome would look like. Now TM04 or connections of form is live and the gate 2 offers are going out or have gone out. How close is what's happening to the version you hope for?
- [04:30]
B
So, I mean, I think my words last time possibly were about sort of speed and pragmatism. So let's get something okay and let's get it out quickly. It hasn't been fast, so I think that's quite an understatement. Connection to form is taking a long time. It's very complex and, and it's been really, really hard, to be honest, for parties who rely on grid connections as their kind of like, route to market. It's been a really, really rough couple of years with a lot of uncertainty and that uncertainty is still there. So I think there are definitely huge things that Grig reform has achieved, but it hasn't been a kind of panacea and it's not over yet. So we're still feeling it. So I suppose if you'd asked me that 18 months ago, I think I would have hoped we were out the other side by now.
- [05:18]
A
Yeah. So we're not being quite as pragmatic and quick as you'd hoped for, but there has been some progr. So people are getting offers. Like we have moved on to some degree.
- [05:25]
B
Yeah, absolutely. So I guess what has been successful is the kind of TMA4 idea, the sort of ideology was worked up into methodologies. Those methodologies have now been implemented and we've got a kind of process by which customers are now receiving their gate to offers. And the first ones of those sort of started coming out early this year. NISA are very happy because they've kind of got to the end of their first tranche of offers, which was projects with connection dates in 2026 and 27. But I think we're, you know, we're still seeing some, some hiccups kind of even there. So.
- [05:59]
A
So are those a golden ticket? Like if you get one of those offers, is it just like, okay, everything's going to go smooth sailing from here, or are there, like, are there, are there problems?
- [06:07]
B
Well, yeah, golden ticket is a good way of phrasing it, isn't it? Because the 2026 and 2027 connection dates, they gave you the highest level of protection. So you were a protected clause 2B project. And the sorry 2A and 2B. I guess the reality of what's happened, people really expected, therefore, their Gate two offer to look exactly the same as their grid connection offer had before. They were saying, look, I've built half the thing. We're expecting it to energise it next year. Don't change anything. Thanks very much.
- [06:35]
A
Just have the same thing back again.
- [06:36]
B
Exactly. That's not quite what we've seen. So some Gate two offers that have come out are very simple and are basically copies of what was there before. We've also, though, seen a lot of Gate two offers come out with technical errors in them. So that means that NISA have done a great job of issuing the offers, but they need to be able to address the kind of technical queries.
- [06:56]
A
What does a technical error mean?
- [06:57]
B
Well, so I mean, that could be anything from, you know, perhaps like just a typo that says, oh, sorry, did you want a thousand megawatts? We gave you 100. You know, they're quite, they're quite important things in these contracts, so they can be kind of purely typographical errors. Sometimes they're sort of errors in the technical assumptions that have been used. Someone's picked up the wrong data and that's literally given you the wrong technical solution.
- [07:19]
A
And surely this is an easy fix, right? Surely that's like, I'm sorry, I applied for, for 1,000 megawatts. Like, you've typed in 100. Like, let's just get this whip round and get it fixed.
- [07:26]
B
Yeah, I think. And that's probably the nub of a lot of frustration, is that in isolation, all of these things are pretty straightforward things to fix. And from an engineer's perspective, you just look at it and go, yeah, that's fine, I know what's gone wrong here. I can reissue your offer. Absolutely fine. What happens is, I think when we've seen, when you try and do these exercises en masse and you're issuing several hundred offers, actually that kind of quality assurance process isn't there, and that means that you're suddenly, you know, you've issued 250 offers, great, but 240 of them have got errors on and now you're having to come back and kind of reprocess that. So that hasn't been great. And I think it's something that hopefully NISO will kind of have a bit of learning from for the next ones around. Actually, it's really worth putting in place that good QA process. You don't want things to have to come back to you.
- [08:12]
A
You could go a little bit slower and it could be faster in delivery. Even though, yeah, you kind of do have to wait for that QA process to run.
- [08:18]
B
Yeah. And I think, you know, obviously in particular with things like Transmission offers. So Transmission Office and Distribution Gate 2 offers are both kind of coming out at the moment. Distribution offers tend to be delivered by a sort of slightly tighter team. You know, those teams within those DNOS have done that job for a long time. There's been slightly less kind of changeover of staff. Whereas NISO, as a kind of like new organization have grown hugely. We've got lots and lots of staff who don't have that kind of deep history and connection. So it's perhaps understandable that there's been a little bit more error prone kind of contracts that have come out. But I think, yeah, resolving those issues quickly. We're still seeing customers, you get three months to accept an offer at Transmission, but we're seeing most customers who raise these kind of technical queries needing that three months to be extended because NISO can't resolve the query within three months.
- [09:06]
A
And so big picture, right? We're pushing towards Clean Power 2030, a very good goal, but perhaps something that there's a question mark around sort of how deliverable it is. But with the Connections reform, are we starting to sort of quietly fall behind the pace we would need to hit in order to do clean power 2030?
- [09:23]
B
Yeah, well, I suppose, I mean, I guess it's no secret that there was quite a lot of publicity around this statistic that came out earlier this year, which was that 62% of those 26 and 27 projects at Transmission couldn't be delivered by their protected date. So the TOS basically wrote to OFGEM and said, we don't think we can do this either because we're slow, the developer's slow. Maybe we're both slow because this connection to before process has taken so long. So, I mean, I think that should give you an indication. Those are the projects that are the most advanced and the majority of them are not on track to hit their dates. So, you know, I don't think that's kind of a secret. I think 2030 is a real challenge and we've allocated these projects to 2030. We've said, you know, you are in the Clean Power 2030 plan, you're strategically aligned, but that doesn't guarantee you a, a date, a connection date pre2030. So no, I'm expecting we will see even more slippage over the next few years.
- [10:19]
A
Okay. And is it purely kind of the. It kind of, it feels easy perhaps to point at connections reform and say this is the problem. But at the same time is it also that actually to do the work of clean power 2030 and to get this many connections live just actually, even if it was perfectly organized and perfectly technically planned, we would still really struggle to deliver it because it's a huge amount of work.
- [10:43]
B
Yeah, yeah, I think that's a really good point. Like we shouldn't necessarily sort of like point at connections reform and say well that's the problem. You know, actually delivering transmission infrastructure is hard, you know, and it always has been. You know, we've, we've kind of, I guess grown quite used to needing to kind of proactively manage those sorts of delays and the delays happen on both sides, don't they? So you know, you might, might be that it's a, to struggling to get consent for their overhead line but it might equally be a generator, you know, who's got loads of opposition to some of their, you know, cabling works or something like that. So I think we're, we're kind of used to that process at and what we're trying to do is I guess accelerate those connections against a backdrop of maximum kind of process uncertainty due to connections reform, really high volumes of connections and probably realistically increased opposition from things like the planning space and so on. So yeah, I think it's a bit of a perfect storm really of factors
- [11:33]
A
of things popping up and maybe just with that perfect storm, let's add in an extra bit of storm because we've got long duration energy storage that just came through. So we have extra projects now that are going to be squeezed into the queue somewhere. A little bit of a Q jump maybe because they need to come online by 2030, 2033 of their sort of pumped hydro and that already feels tight. So how does that Q jump work?
- [11:55]
B
I think that's, it's really interesting actually and really naively I looked at the list of projects which, I mean it's just a minded two position at the moment, isn't it from ofgem's perspective? But that kind of minded to and my first thought was well, where are they? On a grid, you know, on a grid kind of deliverability readiness level and actually there's a complete spread. You know, some of those are gate one projects. They don't have a place in the queue. That really surprised me. I thought wouldn't that have been one of the criteria, you know, actually that you need to show you could deliver grid by 2030. I think we should be really cautious about assuming that we can squeeze projects in or accelerate projects. I think, you know, let's be realistic. We want to try our best to deliver the projects that we've got in the new queue, you know, as close as possible to the dates that they were originally intending to meet. Knowing that we're going to slip a bit, you know, but I think the idea of sort of saying, oh well, let's, you know, chuck some more in and I'm sure we can deliver those in the Next, you know, 2030 is four years away, that's not long.
- [12:50]
A
But somebody's going to get bumped. Right. And they've been in that queue. They were ahead of those projects and they have to get bumped if you're going to get these projects online by 2030, I mean.
- [12:59]
B
Yeah, that's an interesting, you know, you're assuming they definitely get bumped to get the projects online by 2030. I think I'm perhaps assuming it's pie in the sky that we're going to get those LDES projects online by 2030.
- [13:09]
A
There's going to be a loosening of the rules, you know, the finish line is going to get a bit of
- [13:13]
B
a. I think you'd need different criteria, you know, to say actually, you know, well, maybe we want ldes project that are, you know, I guess, yeah, good economically but also really deliverable because they're a bit more progressed or something.
- [13:24]
A
I love that your first reaction on seeing the Elders outcome was like, what is the deliverability of these from a key position.
- [13:29]
B
I'm very, I'm an unabashed, good engineer, I'm afraid.
- [13:32]
A
Yeah, I love that. Okay, let's then stick on with, with storage. So we have around, let's say we're trying to build 30 gigawatts of battery pipeline. In theory, I think we think the number by 2050 is a higher number. But let's just take a rough government number of 30 gigs. Do we have the kind of engineering capability to deliver it and should we be looking at sort of shortcuts, perhaps things like bay sharing or other workarounds that would allow us to put battery storage onto other existing connections to sort of buy a bit more headroom to go a bit faster?
- [14:07]
B
Yeah, so I guess battery storage has had loads of attention because of the, I guess over subscription compared to The Clean Power 2030 kind of predicted figures from the government. And I mean that's sort of a bit of a problem of connections reform zone making, you know, we came up with these protection clauses which were to try and give investors and industry certainty. And you know, and then we, then we've sort of said as an industry we don't really like the outcome of this. You know, we've protected too many battery projects. This must be a problem. I think, you know, my answer to that from a grid perspective is that that's only a problem if those extra battery projects are driving reinforcements. You know, the customer is going to have to pay for or that are going to delay other projects or you know, or that we think maybe just aren't need. So there's a really important look at what's the engineering implication of adding more storage. And I think it's quite an interesting question because some of that storage is co located for example, so it doesn't need another bay. It's not going to have a kind of adverse system impact necessarily. Some of that battery storage is maybe embedded in the distribution network. So I think the big broad numbers have been bandied around a lot. We've got 60 gigawatts, we only need 30. We've got another 20 that might be protected in future Windows. But actually I think to work out whether that's a problem, you have drill down into the of those. Are any of them really triggering extra system reinforcement?
- [15:29]
A
This is just it. I think it's a framing problem. Right. So if you said you need to get. So the way people currently think about it is we need 30 gigawatts. So we need to get 30 gigawatts of grid connection bays or slots that are available. So we need to add in these projects. I think if you reframed it and said Catherine, I'm going to try and I'm going to ask you, can you sneak 30 gigawatts of batteries onto the system without adding any new connections, would you be able to do it? I think you might have a good go at it.
- [15:54]
B
Yeah. I mean I think, you know. Well, I think definitely, you know, there's a role for hybrid projects there, you know, so adding battery storage behind the meter of another connection. So you're not actually adding any more local connection assets. It doesn't need another bay. And NISA have made some quite big strides in terms of their connection modelling assumptions to sort of say, well actually we don't think that batteries most of the time will sort of add to system peak loads. So you don't expect batteries to kind of be triggering a big new 400kV line for example. So yeah, I think that is a big opportunity there Specifically with hybrid project bay sharing. So that would be a solution that would say you're not a hybrid project. You know, I own the battery, you own the solar farm, you know, but we're five miles apart. Could we share a bay? And I think that's, that's something which at the moment it's, realistically, I think it's kind of getting lip service. It's sort of been put in the this is slightly hard to do pile. But it is important that we do it, I think, you know, realistically. So a 400kv bay, so that's sort of, you know, your switch gear bay within a substation is rated for. We could connect up to 1,800 megawatts to one bay. And we do when we have large offshore wind farms, for example. So the idea that you're going to use one bay for one 200 megawatt battery project is really inefficient from an engineering perspective. So we have a bit of a kind of, I think we have a bit of a societal duty to say that's not a very good engineering outcome. We should try and make these things share bays. And even if that needs a bit of. The reason it's complicated is it does need some changes to the commercial ownership boundaries. So I'm not saying, gosh, why hasn't anyone done this already? I definitely respect the challenges, but it's an important one.
- [17:32]
A
Okay, so low. I think the thing if you're hearing this is kind of low utilization grids are generally quite expensive. If you can bump up your utilization, then you should get lower costs to consumers and things like bay sharing get quite exciting. But we haven't finished with grid geekery yet because we're going to go to CMP470, which is the oversubscribe technology commitment fee, which is just the snappy title.
- [17:57]
B
One in a long list of fees.
- [17:58]
A
Yes, one in a long list of fees. So what is that doing and what does it mean for developers?
- [18:06]
B
So this was a proposal which didn't actually come from niso that should be recognized. So this was anyone who's a cusk party. So any generator or demand who's connected to the network can raise a cusk modification. And this came from a battery storage developer and it was saying, look, we think that NISA have said they've got a problem. They've got over 60 gigawatts protected projects in the gate 2Q. They only think they need 30. This is I guess a financial mechanism, a market mechanism for saying if we make those battery projects or they've said technologies which are oversubscribed compared to their Clean Power 2030 targets, if we make them pay an additional premium for staying in the queue, then that would help thin the. Thin the numbers out. Right. You're only going to pay if you've got a real project. And so anyone with speculative projects or so on or projects that are no longer looking so good for development might pull out of the queue. And that helps kind of thin it down. I think I'm a bit sceptical as to whether that's just a sort of easy numbers answer to a problem, which I'd like us to go back and look at that engineering impact. Do we need to thin it down? If we can actually, if it turns out that 60 gigawatts, 50 gigawatts of it, you know what we can get on with base sharing through hybrids, which doesn't have a adverse network impact, then why have another financial metric? And my worry is the financial metric kind of grabs the headlines and everyone says, oh, that'll do the job. Let's not do the hard engineering bit.
- [19:37]
A
Yeah. And also you could see the world where we go, the system suggests it needs 30 gigawatts. We slim down the queue from 90 to 30 and we go, great job done. And then we reassess the position, say, oh, actually, we needed 50 gigawatts.
- [19:51]
B
Well, I think that's.
- [19:52]
A
And then we go, Christ, where's the 20 gigawatts spare? We've lost it.
- [19:57]
B
I'm sure there are loads of philosophers, I can't remember, but the only thing you know about forecasts is wrong. Right. It's just a forecast. So the idea that we're saying, oh, it's definitely going to be 30 gigawatts, that seems very presumptuous to me to. Not. The error bars on that must be quite large. And I think we do need the queue to reflect that. We can't just have a queue of. I was listening to your Kate o' Neill episode, by the way, and I think the term she used was sort of a queue of kind of rotten hype projects. You know, like the idea that the projects in the queue are all kind of like perfectly ready to be connected. I think that would be a disaster because I think what you'd find is them going, I'm ready to be connected, you know, And Tia would go, okay, yep, if you could just wait for seven years, you know, so that's not really how it works. You know, the queue has to gradually move as the. As the Connections become available and we build the infrastructure. But over that, over those kinds of time periods, we see really significant changes in the market conditions. People will fall in and out of the queue. So I think that idea that we've got a number of and we need to come up with as many mechanisms as possible to get exactly that. 30 gigawatts seems a bit flawed.
- [20:58]
A
So you're. I'm trying to read the room. You're a fan of it as a concept or you're a fan of it doing some of this work, but you think it might be an overly blunt mechanism and actually if given the free choice, you'd rather it being done from engineering first principles than I think we
- [21:13]
B
should do the engineering first. Okay, so I'm not really a fan.
- [21:17]
A
Not a fan. Not really fast of cmp.
- [21:20]
B
I say it like I. But yeah, I think that we're probably prioritizing the wrong solution there.
- [21:26]
A
Okay, makes sense. And you snuck cusk into the conversation, which is the connection and use of systems code. It is, yes. Sorry, what is that? What is that mythical thing?
- [21:35]
B
So that's the set of commercial rules, effectively, which govern how generators demand users, network operators all interact commercially on the transmission system. So it governs charges, connections. Yeah, processes.
- [21:49]
A
So if you hear cusk now, you
- [21:51]
B
know you'd already dropped CMP for. So I thought that was, that was, you know.
- [21:56]
A
Okay, right. And those are the, I think the code modification proposals.
- [22:01]
B
They are code modification, yes.
- [22:02]
A
Okay, very good. Let's move on. We're going to go on to colocation. So we see colocation a lot in Australia, Germany, Spain as a way of getting more out of the same connection. And we've touched on this a little bit in base sharing, but it feels like we're not necessarily pushing hard on this. Where do you stand on those colocated projects?
- [22:21]
B
Yeah, so colocation has been a bit of a hot topic recently because there was the original TM04 plus proposals. I think, you know, it's fair to say, with the benefit of hindsight, had some. Had some sort of unintended consequences and potential shortcomings in particular for hybrid projects. So if you're a hybrid project, so you've got two different technologies, those two technologies were assessed completely independently as to whether or not they were ready and whether or not they were needed. So you could have a solar farm and a battery that were co located and they were both ready. So you had land rights for both of them. But when they went through the kind of churning the handle deciding are they strategically aligned with the Clean Power 2030 plan, maybe it came out that your solar was, but your battery wasn't. And what that means is that you get a gate two offer for your solar, only your battery falls away to gate one and that kind of breaks the hybrid model. Lots of people would have been like, well, you know, but the business case for this project relies on there being a solar and a battery. So now I'm a bit stuck. Maybe I'm not even going to develop the solar because I haven't got an offer for the battery.
- [23:23]
A
And we see that in regions like Spain at the moment where people are looking for at solar projects and saying, well, unless I've already committed to this, I'm putting the brakes on this, I don't want it. Unless it's got storage on it, I'm not going for it.
- [23:31]
B
And I mean, it sounds a little bit bad, but then you begin to see the actual sort of project specifics. Had one that a colleague mentioned last week, which was a very large scale solar farm and battery. They were going to be D.C. coupled, so that's where they share the inverters. So it's even more than just your business case relies on the other technology. Your fundamental technical connection arrangement relies on there being two technologies. I think they'd received planning consent for the battery, they'd gone through their kind of DCO process, so quite an involved, significant process. I think they received consent for it a week after the window closed for connections reform. And so needless to say, what ended up happening was they got a gate to offer for the solar and they didn't get a gate to offer for the battery. And you think, well, actually, is that kind of the right outcome? We've got something which is not. That battery wouldn't have had a kind of adverse technical impact because it's sharing the same inverter technology. It doesn't have any kind of additional system impact. So I think there's a lot of legitimate frustration that we've shot ourselves in the foot here and we've got loads of good hybrid projects that are either half in the queue and half out or they've just gone, well, I can't deliver under these kind of conditions.
- [24:46]
A
There are probably lots of photos that could be taken of sad concrete plinths on solar sites that are ready for a battery at some point in the future.
- [24:52]
B
At some point in the future, yeah, exactly.
- [24:54]
A
I feel like I've heard that expression about a thousand times.
- [24:56]
B
Yeah. So NISO have kind of heard that criticism, I think, and they have come out with some tweaks to the methodologies. So I guess that's worth saying. TMA4Plus was the starting point. But Connexions reform is evolving and this kind of process, this enduring grid connection process is still evolving. And they've sort of said, well, we think we probably need to do something about hybrids. We recognise maybe we didn't get it right the first time around. So what they've proposed is, is that they will in future windows, so the next time there is a Gate two window, which we're still waiting on that one, they will prioritize. A project which is hybrid will be prioritized compared to a sort of standalone project within its own kind of category. So you've got, you know, 10 wind farms, they've all got planning consent, but one of them's got a battery on it. Okay, that wind farm plus battery will go to the front of those 10 projects.
- [25:47]
A
Hold on, isn't this pointless?
- [25:49]
B
They're going to the front of the back of the queue.
- [25:50]
A
They're going to the front of the back of the queue, which is already, already too long anyway. Like this. That. Yes.
- [25:55]
B
Your words, not mine.
- [25:56]
A
Yeah, I would be quite frustrated with that. I think.
- [25:58]
B
I think, I think for a lot of people that's sort of too little too late.
- [26:02]
A
Yes.
- [26:02]
B
The other thing that NISO have done is they've introduced what they're calling a parent child concept. So that's where you, if you've got two technologies, you can tell NISO when you're applying. Okay, okay. This technology is the parent technology and what that means is it could stand alone. So if you're only going to give me an offer for the solar farm and the solar farm is the parent technology, then I'm still interested, so give me that offer. But the battery technology is a child, for example, and therefore, if I'm only going to get an offer for the battery technology, don't bother giving me the offer. It doesn't work commercially, so that's quite a good thing. And they've been quite flexible. They said, look, this is so that you can tell us as much about your commercial viability as you want to. You could have two parents, you could say they could both stand alone and then let us make the decision. But I think that kind of concept, it's a good one, it's come from developers in the kind of feedback to consultations, but it probably needs a bit more fleshing through and probably a lot of explanation to people.
- [26:54]
A
It still feels a bit too late. It just feels like we have this roadmap Almost yes to 2030, but let's say we're not hitting 2030 and this ends up being the roadmap 2035 or 2040 or quite a long way into the transition. So the back of the queue is a long way away. And so I think I'd worry about that just to go back on a few things. We had AC and DC coupling. So we have alternating and direct current. Most co located sites are alternating ac. So that means that the solar has its own inverter, the battery has its own inverter and then you connect it to inverter. That's pretty standard in lots of places, D.C. direct current. So that's essentially the solar and the battery connects before the inverter and you have one inverter for the two technologies that's much more common in places like Australia do a lot of DC coupling. But just, yeah, just so for people sort of thinking about that and thinking, okay, how do these work? And then you also mentioned dco, that's obviously the sort of the direct consent that comes through for people to get planning to move their project forward quite quickly. So yeah, just to clear up a bit of acronym, a bit of jargon, okay, and let's move on from colocation into the demand side because if generation was tough and we found a big queue, demand must be easy. So we are now getting lots of large new loads like data centers trying to get connected. Is this a harder problem than connecting generation? It feels like actually what we need is demand. So shouldn't it be quite straightforward to get them on?
- [28:16]
B
Yeah, it's interesting that, isn't it? I think, I guess from a system operation perspective, like some more demand sometimes might be useful, some more controllable demand would be incredibly useful. But you know, again, stepping back, that kind of engineering hat on ironically, connections is really quite separate from system operation. You know, in the connections world we think how do we plug you in? We don't really care what you do, you know, once, once you're plugged in. So and demand, when it comes to plugging in demand, you know, it's simple things like actually the security of supply. So how resilient does your connection need to be For a data center is different to a wind farm, you know, like if you're Microsoft and you're building a massive data center, you really don't want it to go off. So that probably means you're going to have at least two points of connection. So two bays, you might want that resilience to be spread between different substations. So we build more infrastructure per megawatt for demand than we do for generation. So that sort of needs to not be underestimated, that when we say we've got, you know, 170 gigawatts of demand in the queue, if we were to actually try and connect all of that,
- [29:17]
A
we'd be using a lot. So one gig of demand is harder than one gig of generation.
- [29:21]
B
It requires more assets, more local assets to be built from the grid side. So that's one thing, I think. The other thing is that again, perhaps sort of some of the quite relatively simplistic ways we assess the network when we look at someone wanting to connect. So again, you take that 1 gigawatts of demand and maybe it's located in West London, someone wants to put a data center in that kind of Slough hub, then what you're actually looking at is if I add that one gig of demand to the system, there are going to be times when effectively what that does is, is use more onshore wind. That's happening in Scotland, right, Which is sort of good from a system perspective, but I still need to get it from Scotland to Slough, so that demand, especially demand in the south of England, can still trigger really quite significant upgrades. So ironically, you'd probably have your one gig of wind in Scotland and your one gig of demand in Slough, and they'd probably both have the same, what we call enabling works, so the kind of transmission works needed to facilitate those power flows on the system. System.
- [30:21]
A
So if somebody, let's say you've got two data centers, both one gigawatt one applies in Slough and the next one applies in Aberdeen the next day, but is behind it from a sort of a queue perspective, does the system operator have the ability to say, oh, hold on. It actually would be pretty useful to have a bit of demand in Scotland and having one right next to London is going to be. It's going to trigger all of these reinforcement works across all these boundaries. Does it have the ability to kind of. Of just slightly rejig things in order to make the system run slightly more smoothly, or is it sort of bound by the rules?
- [30:54]
B
It's a really interesting question at the moment. The rules don't sort of say, okay, well we're going to queue jump this one, which might be helpful to the network. There are ways you could do that. Project designation, things like that are kind of routes where you could begin to start looking at strategic demand and perhaps ways of accelerating that. But at the Moment we don't generally do that. But instead I guess it comes back to that fundamental what is the queue? The queue is just the order that you are applied and we studied you in. But we can still take that data centre in Scotland and it might get an earlier connection date, for example, than the data centre in Slough. We haven't rejigged their queue positions. We just looked at what engineering works
- [31:29]
A
we need to do. Yeah, okay, that makes sense.
- [31:31]
B
I think the one thing is, I would say is that, you know, the grid queue for demand is really significant and I mean the grid queue for demand even in Scotland is really significant. So a lot of people have had this idea and I think what would probably be great is if you put a data centre in Scotland that didn't mind not operating when it wasn't windy.
- [31:48]
A
That would be fantastic. That's a real sort of.
- [31:50]
B
But I haven't found that kind of Netflix like punter crowd yet.
- [31:53]
A
Yeah, yeah, I mean it's. We had a really interesting episode on thermal storage recently around the potential to sort of heat up concrete, concrete blocks to like essentially be us, be, be a, be a demand load and reduce our reliance on gas during times of the year. And then at times when the electricity grid is scarce or it's really tight, you say, well look, we're just going to stop running on electricity. We still have some of our gas system here, run the gas through. The same energy is coming into the. But we are, we are reducing our demand in peak winter but we're increasing our demand in the middle of summer, let's say when we have industrial loads. And I kind of, I really like that concept but I'm not sure that that would get bumped in the queue just because of how we think about these projects.
- [32:37]
B
Yeah, I mean, and, but it is coming. I think that there is some really good kind of noises that are happening now where we're beginning to see, I guess really because of how much pressure there is on transmission owners in particular, you know, to give people earlier dates. And so we're seeing more sophisticated sort of assumptions around how loads and how demands operate on the network. So whenever a customer applies the transmission owner, if they're a transmission customer, would do a network study to work out what their impact is, how many things need to be reinforced. And they do that using sort of, we call them construction planning assumptions, CPAs and actually there's a lot of focus now on what those CPAs are, how time based they are. You know, do they just pick a couple of, you know, we Used to do this basically would say, what does a windy day on the system look like? What does a not windy day on the system look like? And what does a sort of summer Sunday look like? You know, just pick those three. Whereas now actually we could do, if we've got more flexible demand, more sophisticated demand, we can do much more sophisticated studies.
- [33:37]
A
Okay. Now, as two engineers, we should never venture into the world of politics, but here we go in, in the demand queue, you have obviously data centers that want to get connected, but you also have these technologies that are of very much sort of public need. And I'm not to say that data centers aren't a public need. I'm already tripping up over this question. But we have a need for electrifying heat, we have a need for electrifying transport, and we have the need to electrify industry as well. And so you've got these kind of. Perhaps, let's say we've got four things that are all trying to get electrified all at the same time. Should we be thinking about politics and policy in this question or is it just. Actually we should just be pure grid engineers and it should just be done that way?
- [34:19]
B
Yeah, well, I suppose, Well, I am a pure grid engineer, I suppose so I'd probably steer away from the politics, but it is not, perhaps that's harder to do. Now we've got niso, ri, whole system energy system operator. So they're already kind of trying to more centrally plan and they've got things like their strategic spatial energy plan, the regional energy plans, the regional energy spatial plans, the rest, let's go for that. They're RESPs. Yeah, I remember the acronym. So I guess there is far more central planning and that covers both demand and generation. So. Absolutely. Things like electrification of heat, even like lots and lots of house building, for example, transport as well. I think my slight. It feels like a big system change to go from, you know what, you know, you guys, if you're building a hospital or some houses or a new train network, you come and tell us when you want a grid connection and we'll slot you in the queue kind of type thing. It feels like quite a big change to go from that to we think, you know, this rail improvement is more important than that hospital and all of them are more important than this Microsoft Data Center. That feels like a very, very, very
- [35:22]
A
sort of the headline's going to be bad, isn't it? I think it might be, you know, Data center gets three bays for 99.999 availability, yet hospital sits 100th in queue.
- [35:35]
B
Yeah, and I, and I think I, I'd really question the need to do that. So what I would say is a difference is that in the demand space, particularly when it does come to things like, you know, perhaps more sort of like societal needs like hospitals and so on, you don't tend to get the same amount of kind of gamification or speculative development. You never known an NHS trust think you know what, we'll stick in 10 grid applications and we'll flog them off to someone if, you know, if we decide not to develop. As far as I'm aware that doesn't happen. And so I think the response that we need is different as well. I think we can hopefully let those demand projects sort of come along when they are ready rather than requiring a top down system approach. The data centers I think are where we've seen much more of that sort of speculative behaviour. And therefore there are some proposals to look specifically at data centers and perhaps ways of, to raise the bar for entry into the queue so that we don't end up with them sort of pushing out that hospital that really needs it.
- [36:30]
A
Okay, so let's finish off with two more practical questions. So getting these projects online. It always used to be the case that the transmission transformer was the hardest thing to get your hands on and was going to be. People would talk about sort of the longest pole in the tent. So the thing that's kind of going to take the longest to do. Is that still the case? Do you still see that problem coming through?
- [36:51]
B
I mean lead times for key like very large scale plant equipment are still, still very long and that's a, that's a worldwide constraint. Right. So it's not just a UK market that we're struggling with. I'm not convinced that really that those have necessarily relaxed, but I think potentially they have stopped being the, the critical path, that longest pole in the tent because we've, we've started to see some other real challenges. I think the two I would highlight are, I mean one of them as sad as it is, is literally the contractual side just getting offers signed. You know, seeing those kind of projects in the really near term have got 2026 or 2027 connection dates, they've got queries on their offers, those typos we talked about or maybe more fundamental issues and they're really not being able to get over that final line. Maybe they're most of the way through construction but they're not actually going to be allowed to operate. They're not going to be able to go through the compliance process, for example, unless they've got a grid agreement which has their tech rights, their transmission capacity. So. So it sounds small, but it's happening to real projects. So we're seeing. I've got a couple of 20, 26 projects on my desk at the moment who don't have a Gate 2 offer yet. They're pretty close to being energized, but they still don't have it.
- [38:00]
A
And it's really bumpy. Right. Because if you miss your window, the team that are doing that work, they might be on other projects for the next two months and so you might miss it by a day, but you might end up two months further back.
- [38:10]
B
Yeah, I mean, this stuff was already quite hard. Definitely. I think anyone who's been involved in sort of commissioning the grid compliance of large scale scale transmission projects, be they batteries or other technologies, will be aware of how stressful that is. And you really are kind of day to day trying to manage that. So if you add in contractual uncertainty, which we have done over the last year, suddenly you could have done the most amazing job on procurement and you're still actually going to be delayed by some pretty frustrating paperwork. So that's one thing, I think the other thing which has changed is we are seeing more and more delays on transmission owners themselves delivering reinforcement works. And sometimes, unfortunately, that that news might be communicated quite late to a customer.
- [38:50]
A
So people sort of upstream of the project.
- [38:52]
B
Yeah, so the project's done a good job, they've got their transformer in place, for example. But perhaps the transmission owner says, oh, actually there was a reinforcement work that maybe we thought we'd done a couple of years ago and it turns out we didn't actually do it, so we need to go back and do that now. And so those are kind of real examples of projects that are meant to be connecting now and that are delayed.
- [39:14]
A
Incredibly frustrating, I imagine, for those projects. And one natural response to that might be, well, well look, rather than relying on the current transmission owner, I'd rather get an independent transmission owner to do this. So I'd like to have my own sort of responsibility for this. Are we making headway on allowing sort of independent transmission owners?
- [39:34]
B
Yeah, actually. And so there is quite a lot of headway on this and I think it's sprung from two different places. Yes, probably a little bit of frustration perhaps with some of the current processes and that lack of control. Control also definitely from a new type of customer. So these data centers, for example, they're much more constrained in at the moment, what they can deliver themselves. So just for people's understanding, a generator at the moment in the UK tends to, especially in England and Wales, they tend to build back to the grid substation. So if they're given a point of connection at a particular substation, they would build their own sort of substation on their site, they'd cable back to that NGET substation and they would even populate the kind of switch gear at that substation. So actually they're delivering a lot of the local works themselves and they own and operate those assets as part of their business. If you're a demand customer because of the way the electricity sort of licensing works, you're not allowed to own what we call high voltage transmission assets, like 400kV assets, for example. And so it works slightly differently. And the transmission owner builds out to you. And now you really need to be worried about whether or not that transmission owner is taking the right procurement actions. They doing it? Are they designing it the way you'd want it to be designed? So you've got that kind of new emerging. Sorry, you've got that new kind of demand customer base that's saying actually we want control of the stuff, the generators do it. Why can't we?
- [40:57]
A
Yeah, I mean that feels perfectly fair. So interesting to see that's moving forward and we're starting to see sort of ITOs potentially coming forward.
- [41:04]
B
Yeah. So we've had, within the last couple of weeks we've had a letter out from OFGEMS as a publication that's effectively kind of setting a direction of travel, suggesting that they will, you know, specifically to facilitate demand, they would like to potentially enable an independent transmission owner regime. There are some other routes that could enable the same thing. The reason that generators can do it, they don't have a generator ito, they just have effectively within the kind of generation licensing, they have the ability to own some of these assets. So we could look at things like the licensing and there are class exemptions and stuff. So there's a few different work streams.
- [41:40]
A
Okay. There's a whole world we get into. This could be at least two hours. But what we're going to do is we're going to draw it to a close with a final question, which is what is a contrarian view that you hold about energy?
- [41:50]
B
Well, I think now maybe it's not that contrarian anymore. Now I think we've worked through our. I suppose for a grid engineer, I'm going to say what I didn't think I would say, you Know, for the last sort of 15 years of my career, which is that I think we're at risk of building too much grid, too much underutilized grid. I think we're really, I'm really surprised by some of the engineering outcomes that we're seeing, which are saying, saying, you know, we need to build a 20 bay substation here, you know, even though actually, you know, we've been through the gate two process and we previously had 10 customers and now we've got two, you know, but we're still gonna, we're standardization, we're still gonna build the 20 bay substation.
- [42:27]
A
So build it and they will come
- [42:28]
B
type thinking and it's that that's just feeling really wrong, you know, like you look at our kind of electricity bills now and you think, no, we don't want to build an expensive grid, we want to build a streamlined, efficient, quick grid.
- [42:37]
A
OFGEM should be red hot on this.
- [42:39]
B
I think, worryingly though, OFGEM have been sold, you know, from an engineering perspective. They've said standardise, you know, standardise everything, you know, make sure that you're going for low regret scenarios, make sure it's future proof. You know, they all sound like really good words but if you dig into them, they're really expensive words too. And I think that kind of we need to be a bit more agile. What I'm seeing at the moment is that we are still seeing designs for things like substations, especially at transmission, which are reflective of the pre connections reform queue and they haven't moved at pace to go. Actually. We might be making an expensive mistake here or building a white elephant.
- [43:17]
A
I'm going to give you lots of points for your answer. I'm going to subtract some of those points. I'm not sure it's contrarian. Hugely. I think a lot of people are starting to think it, but I really like it.
- [43:25]
B
You spent at least 14 years saying we need to build far more grid, Ed.
- [43:28]
A
Yeah, well, Duff, look, you got points, you should be happy with points. Well, Katherine, all's left me to say thank you very much for coming on again. It has been a. Another sort of tour de force on what's going on on the connections world. So thank you for your time.
- [43:40]
B
Oh, no, thank you very much, Ed, for having.