Electric transmission development is notoriously difficult, and these days, NIMBYism gets the brunt of the blame. But as data center loads surge and electricity prices climb, there’s a new roadblock: the messy world of multi-state cost allocation.
The Mid-Atlantic Resiliency Link (MARL) — a planned 100-mile, $960 million transmission line stretching across Pennsylvania, West Virginia, Maryland, and Virginia — was approved by PJM in 2022 under standard rules that spread costs across the entire region. But that plan was made before ChatGPT took off and data center forecasts shot upwards.
Fast forward four years, and now state consumer advocates are asking why local ratepayers should foot the bill for an infrastructure project designed to feed data centers in northern Virginia.
In this episode, Shayle sits down with Maeve Allsup, senior reporter at Latitude Media, to unpack her reporting on the project. They dive into how the rise of generative AI has disrupted traditional grid planning and explore why this challenge has proven to be such an impactful rate limiter for the AI boom.
[Editor’s note: In this episode, Shayle and Maeve refer to MARL as the Mid-Atlantic Reliability Line. The correct name is the Mid-Atlantic Resiliency Link. We regret the error.]
Shayle and Maeve discuss topics like:
- How a project approved in 2022 hit a vastly different policy and regulatory landscape by the time it reached state dockets
- Why data center growth breaks the historic assumption that regional transmission costs eventually “even out” between states
- How the Ratepayer Protection Pledge — a voluntary commitment signed by tech hyperscalers at the White House — is being harnessed by state advocates as a cudgel to demand data centers pay for grid upgrades
- Why the United States has gone from building thousands of miles of transmission a decade ago to just hundreds today
- How the intersection of local opposition and confusion over utility tariffs is delaying grid buildouts
Resources
- Latitude Media: How the ratepayer protection pledge became a transmission hurdle in PJM
- Latitude Media: FERC to grid operators: Connect large loads to transmission faster
- Catalyst: Looking for a turnaround in transmission
- Catalyst: The rise of flexible data centers
- Catalyst: AI scaling pathways: On grid, on edge, off grid, off planet
- Open Circuit: Grid utilization vs expansion: The 100 GW debate
- Open Circuit: A five-alarm fire for the grid?
Credits: Hosted by Shayle Kann. Produced and edited by Max Savage Levenson. Original music and engineering by Sean Marquand. Stephen Lacey is our executive editor.
Tune into Critical Capital, a brand new podcast from Crux and Latitude Studios. Hosted by Crux CEO Alfred Johnson, Critical Capital explores the interlocking forces powering clean and critical infrastructure. Join us every other Tuesday for in-depth conversations at the intersection of energy, government, finance, and global markets. Listen here, or wherever you get podcasts.
Catalyst is brought to you by FischTank PR, an award-winning climate and energy tech, renewables, and sustainability-focused PR firm dedicated to elevating the work of both early-stage and established companies. Learn more about their PR approach and how they can support your company’s messaging by visiting fischtankpr.com.
Catalyst is brought to you by EnergyHub. EnergyHub helps utilities build next-generation virtual power plants that unlock reliable flexibility at every level of the grid. See how EnergyHub helps unlock the power of flexibility at scale, and deliver more value through cross-DER dispatch with their leading Edge DERMS platform, by visiting energyhub.com.
Transcript
Shayle Kann: I’m Shayle Kann. I invest in early stage companies at Energy Impact Partners. Welcome to Catalyst. So electric transmission development is hard. We know this. I’ve talked about it many times here. When people ask why, they usually reach first for NIMBYism. And to be clear, NIMBYism is real and it kills plenty of transmission lines, but it’s not the only thing. And understanding the multiple factors that drive failure of new transmission, I think is key to understanding why power is, in my opinion, the rate limiter to AI growth and why upward pressure on electricity prices will probably sustain for a while. Anyway, consider a specific line, the Mid-Atlantic Reliability Line, or MARL. It’s about a hundred miles long, little under a billion dollars in expected cost. It would run from Pennsylvania to Virginia through Maryland and West Virginia along the way in a region of the country that needs more capacity than basically any other.
PJM actually approved it in 2022 for reliability purposes under the standard rules where costs are spread across the whole region the way the transmission has basically always been paid for. At that point, ChatGPT had not yet been released. By the time it hit the siding dockets at the state level in 2025, the ground had shifted. Data center load had exploded, electricity bills had become a political problem and the hyperscalers had stood in the Oval Office and pledged to pay for their own power. So now state consumer advocates are asking a sort of obvious question, which is why should our rate payers cover a line that they argue might exist to serve data centers in someone else’s state, in this case, Virginia. Reopen that question though and you have to reopen cost allocation across all of PJM and that may be exactly what keeps this line and the grid we say we need from getting built.
Because here’s the backdrop. In 2013, we built nearly 4,000 miles of high voltage transmission in the United States. Over the past few years, we’ve averaged hundreds. We need to be building thousands, honestly, maybe tens of thousands. So this is important to understand. My guest is Maeve Allsup of Latitude Media, who reported this story. She’s coming up after the break.
///
Shayle Kann: Maeve, welcome.
Maeve Allsup: Thanks so much for having me, Shayle. I’m excited to be here.
Shayle Kann: All right. Tell me what the Mid-Atlantic Reliability Line is supposed to be.
Maeve Allsup: Yes. It’s a planned high voltage transmission line spanning about 107, I think a little over a hundred miles in PJM. It would run from Southwestern Pennsylvania to Northern Virginia crossing through Maryland and also a good chunk of West Virginia. It’s being developed by NextEra and is expected to cost around $960 million.
Shayle Kann: And how long have they been developing it?
Maeve Allsup: Well, the project was approved back in 2022 to address the reliability issue that PJM noticed in the early 2020s, which was that they needed more capacity, west to east capacity, largely to serve growing data center load in Northern Virginia. But Nexera really just started filing for state approvals for the project last year and then this spring. So there was just this kind of ocean of time in terms of years, but also in terms of kind of policy and landscape change between when the project was approved and then now where development is really getting started.
Shayle Kann: Right. And that 2022 approval obviously didn’t mark the beginning of development. NextEra presumably had been working on it for a long time. It’s just always important to remind everybody of like how long transmission takes if it ever gets built. They were developing it for years. They got it approved in 2022. Then it’s 2025 and 2026 before they start seeking state approvals. And then obviously it’s still not online today. So timelines are long in any case, but that timeline is particularly interesting because … So everybody probably knows northern Virginia was already the hub of data centers before the AI wave. And so that project got approved in part, as you said, to meet the sort of growing load that they were seeing in northern Virginia prior to LLMs and the project gets approved and then it is presumably still in development. NextEra is still working on it, I presume.
And you would think that what happens as a result of the AI wave and the fact that data center demand growth gets kind of supercharged in the meantime, that everybody would sort of collectively say, “Oh wait, we need this more than ever. If anything, we should make a bigger line, but at a bare minimum, we should build this line.” And so I think for me, this is interesting because the logical side of me would say, “Hey, this thing was already in development and now we should build it even faster and maybe even bigger.” And instead it seems like what your reporting has suggested is that for, I don’t know, for lots of good intentions, kind of the opposite has happened. And in fact, the rise of AI has kind of gummed up the works for getting this project done. So maybe walk me through the kind of timeline of what’s actually happened here.
Maeve Allsup: And it is really interesting because we are at a point now where everyone is paying attention to something like a transmission project where people kept pointing out to me in reporting for this story that all of the data centers already online in Northern Virginia are supported by energy infrastructure that everybody in the region paid for. It just wasn’t an issue before. People weren’t paying attention in a way that now PJM is really in the hot seat and this project has found itself under a microscope. So to do a little bit of a timeline on the project, the cost allocation model that FERC approved for this suite of reliability projects back in 2022 shares costs across PJM with the larger share assigned to the utility zone that’s going to benefit the most, greatly oversimplified, not a transmission cost allocation expert, but that model kind of assumes that growth in PJM is even around the region.
So if you’re a rate payer in Virginia and you’re subsidizing the cost of a line in New Jersey, eventually New Jersey is going to subsidize the cost of a line for you and it will all kind of even out in the long run. That is really complicated with highly concentrated load growth, which is where we are now.
The thing that has happened in the years in between, and I guess I’ll say like this, in 2022, ChatGPT hadn’t even been released yet. So it was the middle of the Biden administration. The conversations that we were having around energy infrastructure were totally different and in the intervening years we have just explosive load growth forecasts from data centers. Everybody spent the last year talking about capacity auction prices in PJM, the White House and DOE are all trying to figure out how we get these large loads online quickly. And then of course in March of this year is when all the hyperscalers sat down at the White House to sign the rate payer protection pledge, which has really become central to some of the pushback over moral in particular and I think an important lever for consumer advocates to kind of pick up and run with in this ongoing conversation around how to make data centers pay their own way on the grid, which is particularly complicated at the transmission level.
Shayle Kann: So MARL, of course, being the acronym for Mid-Atlantic Reliability Line. Yeah. So it seems like the crux of the issue here is cost allocation. Who pays for it? And this model that we had historically, and the problem is multi-state transmission lines. These problems are always because we have multi-state actors. And so the question is who pays how much for the billion dollars or so it’s going to cost to develop this line? What is the argument for which folks are saying that the previous cost allocation is no longer fair?
Maeve Allsup: Yeah. I mean, they’re basically arguing that one, the data centers themselves should be covering the costs, not rate payers and PJM, but for this specific project, states are arguing that it should be Virginia rate payers, if anyone, that really foots the bulk of this bill. And as we mentioned, NextEra set out at these four state PUCs earlier this year and they’re seeing pushback pretty much across the board in all of these states. And it’s interesting because a lot of the filings do cite the rate payer protection pledge, but there are also just all these kind of general concerns when it comes to property values and agriculture and all of the traditional issues that you would see come up with a project like this. So it is kind of this intersection of data center cost allocation and also it is just hard to get people to approve a huge project like this.
So it is interesting, I don’t know if I mentioned it in the intro, but back in 2024, soon after these projects were approved, Maryland went to FERC and complained about the cost allocation framework for those projects and basically said at that time that Virginia should pay more than had been allocated. FERC rejected it at the time. Maryland has now come back to FERC in 2026 in May is taking kind of a zoomed out approach and is basically asking FERC to assign the data center driven transmission costs to the specific zone. So I think in this case Dominion, they want to say assign all the costs to Dominion and then let Dominion figure out internally how to make sure that data centers pay for them. It’s a little broader than the 2024 complaint because I think it would apply to lines going forward, not just that bucket of lines, but all of the states across the board are unhappy with it and we’re seeing state legislators get involved and it’s becoming a little bit of a mess that I think it is early days but definitely has the potential to delay the project.
Shayle Kann: There’s like an interesting counterfactual here, right? Which is if LLMs had never come along and we had never seen this kind of AI boom that we’re in right now, it was already true as you said, that this transmission line was there in or was planned in part to serve growing data center driven load in Northern Virginia, in Dominion territory. And so had that played out and Nexera still had to go to all the states and to the PUCs and get approval individually, it’s possible they would have made this same argument that they’re making now, but it feels like what has changed in the interim is two things. One is just the magnitude of the data center growth and so it’s harder to make the case now that that transmission line is not disproportionately benefiting data center customers because they’re the ones who need it the most and they’re where most of the growth is coming from.
And then second, there is this more generalized backlash against data centers. And so I can imagine that at the state level and even at the local level with all these authorities that need to approve, this maybe is getting tied up into this broader thing of, hey, either why should we build more data centers or why should we pay for somebody else’s data center to get built? Do you have a sense of how much it’s that sort of like generalized anti-data center sentiment versus this specific Northern Virginia has a ton of data centers and so this transmission line needs to serve them and get paid for by them?
Maeve Allsup: Yeah. I think it is all of the above and it’s interesting because this kind of conundrum that the MARL line finds itself in is not unique to this project, right? It just happens to be in this weird interim where we haven’t yet figured out exactly how to make sure data centers pay for transmission. So we don’t have a system in place and they are trying to build really quickly. They want to start construction in 2029 and the systems to make sure that data center customers can pay for that just don’t exist. So it’s kind of building the plane while flying. I’m not sure if that’s quite the right analogy, right? But they find themselves in this really interesting place where all of the data center opposition, this is a project that probably wouldn’t have had this level of scrutiny several years ago, but now has become kind of this embodiment of all of the problems that people have with data centers concerns over electricity rates, everything else playing out at PJM and at FERC kind of piling on top of this one project that just is in the right place at maybe the wrong time.
Shayle Kann: Just to add to the complexity of it, you mentioned that we haven’t quite figured out yet how to enable data centers to pay for transmission and one of the reasons for that is that it is clearly a socialized benefit, right? In very few cases, is it true that the transmission line, and certainly not in this case, it is not true that the transmission line is being built for the benefit of a single data center. And so what happens is now, so all the hyperscalers sign this rate payer protection pledge and basically what they’re saying is, okay, when I’m going to go cite a new data center, I’m going to negotiate, I’m going to maybe install some infrastructure on sites and behind the meter generation. But mostly what I’m going to do is I’m going to negotiate a tariff with the local utility wherein I will cover the cost of anything incremental that needs to get built in order to serve my demand.
And so mostly what that means is new generation capacity. And in theory, if there’s a clear upgrade that is required to serve that specific data center, you can add that in, but this is a very different case. This is a multi-state transmission line that is going to benefit lots of customers, including a bunch of data centers in Virginia. And so how do you facilitate the joint payment of those data center operators or developers to cover the cost of this transmission line? It’s not obvious to me how you do that. There’s a weird sort of tragedy of the commons problem.
Maeve Allsup: Yeah. It’s not obvious. And again, like something that makes this line particularly complicated is this question of is this a data center line? It was approved as a reliability project in its approvals, PJM did specifically cite large load growth, including from data centers in the Northern Virginia region, but that was before the AI boom happened. For me, I think of the kind of kickoff of the AI frenzy as being that, was it November, December 2022 New York Times article where they’re talking to the disastrous Bing chatbot, Sydney, this was well before that. So even if it had data center relevancy, it was approved at a time when we just were not talking about this at all. And it’s a question that keeps coming up in conversations that I have, which is, what is the date, what is the time at which projects approved after X date, we assume are largely for the benefit of data centers, and projects before this date go forward under the prior existing framework?
I don’t know where MARL falls on that and NextEra doesn’t seem to know either. They’re kind of just in this weird unfortunate limbo.
///
Shayle Kann: Yeah. And then looking at it from the other angle, the playbook now for the hyperscalers, which is I think emergent, but becoming clearer and clearer is when they announce a new data center, they announce it alongside a bunch of investment that they’re making in the community of one kind or another and they announce it alongside a tariff with the utility wherein they can say, “And we’re supporting X megawatts of new generation and this will have no detrimental impact on rates.” I think what’s trickier to say is something like, “And also we are contributing 17% of the cost of a transmission line that’s going to be developed and the rest of it is going to go to other”…It’s just harder to tie that in from a timing perspective and you have the timeline thing. As you said, this is supposed to be fast for a transmission line, but they’re beginning construction, you said in 2029.
And so that is generally beyond the timeline under which data center developers want to come online. If this is going to begin construction in 2029, I assume it’s going to be an operation in 2032 or something like that, 2033 and they’re all focused on speed to power. So you have these multiple layers of dynamics that just make it difficult. Everybody I think wants the same thing ultimately, maybe not everybody. Some of the local folks might just want no more transmission anywhere near them, but at a broad level, it’s this quandary wherein everyone agrees we need more transmission. And in fact, you might have some parties that are willing to pay for it. I imagine that the data center companies would do it if it would get more capacity online faster, but actually enacting that is super complicated.
Maeve Allsup: Yeah. Very complicated. And again, I think it’s this interesting kind of tangible example of this macro trend we’re always talking about, which is the conflict between the pace of AI, the pace of the tech industry, and the much slower pace of energy infrastructure build out.
Shayle Kann: So where does it stand today? You have all this state by state pushback. You said Maryland went to FERC to complain about it. It’s getting pushback from other states as well. Where are we in the journey? Is this project going to happen? Is this project not going to happen? Do we have any idea?
Maeve Allsup: It is still very early days because I think three of the four states next era started the filing process in 2026 this spring. Some of them are further along than others. They’re all kind of playing out in these four parallel processes. I know that in West Virginia, as I mentioned, they’ve started the public hearing process this month. So it is very early days. I think there’s a long way to go. You I’m sure know more about the development and approval process than I do, but everybody that I’ve talked to has said there is a really good chance that this line gets delayed and no one is quite sure whether FERC and PJM will open that broad can of worms on whether to change how PJM allocates transmission costs. And I know you’ve talked about this quite a bit on the show, but the other thing going on in the background right now is FERC through its advanced notice of proposed rulemaking is considering this complicated question you’ve outlined of how to make sure that data centers can cover 100% of the grid upgrade costs that they need.
I think we might have a decision or some type of ruling in that proceeding in between today when we’re recording and when the show publishes, but there’s a lot going on all at once. Again, to go back to the question of timing, I don’t fully understand when, again, the line at which those decisions apply retroactively to this line that was approved and started development in 2022, something that’s decided by FERC today, does it apply backwards? I really don’t know the answer to that question.
Shayle Kann: Yeah. And there’s this other dynamic that this is an example of that I think is worth monitoring, which is the rate payer protection pledge, right? The US government gets together all the hyperscalers and gets them to sign this agreement that has no teeth in it, but is a pledge that says we’ll cover the cost of anything that’s required to build our data centers on the power side. So great. So that seems like it to me at least is like, they should be committing to that and I’m glad that they are and they all have to go do it. But what’s interesting is you noted that in their complaints to FERC, for example, a bunch of states have cited that pledge and I think that creates, it’s an opening for anytime anyone wants to argue against however a tariff is getting structured or a new transmission line or whatever it might be, they can sort of try to say, well, Google or Meta or Microsoft or whoever it is sign this rate payer protection pledge, they have to cover the full upgrades, this doesn’t cover the full upgrades.
And so it’s going to get used as a cudgel a lot I think in these battles. Now maybe for good reason, like they should have to justify that they’re adhering to it, but on the other hand, it may just further slow down new development of new stuff.
Maeve Allsup: Right. Yeah. Ironically, even though I think it was kind of designed as a speed to power and also kind of general public sentiment around data centers play, at least in this case, it is definitely becoming a slow down, a barrier to development.
Shayle Kann: Yeah. What is your take on the emergent negative sentiment about data centers and the degree to which it has or hasn’t extended into the power infrastructure for data centers, right? There have been people talking about moratoria or temporary pause at a lower level. There are some states that are suspending the sales and use tax exemptions, things like that. You’re seeing this bubble up a fair bit now. It seems mostly to be focused on the data centers themselves, but are you seeing it apart from, in this case, extend to all the power infrastructure generation transmission distribution storage that is getting built or needs to get built to serve the data centers?
Maeve Allsup: Yeah. It’s interesting. We have started to do a little bit more of kind of a local level look in different parts of the country where data centers are being built and yeah, we are seeing pushback in part because the data center build out is so high profile, right? So if you’re in a community where a data center is getting built, you’re also hyper aware of all of the infrastructure that is going in around that. I’ve talked to people for a story I reported about a project in Indiana who are really thinking on kind of a minutia level about all of the roads that will need to be built to build these substations, not just the data center itself, but this whole ecosystem of infrastructure that has to exist. And I think there is sort of the negative sentiment around data centers and AI in general, but it’s all snowballing with all of these other concerns that folks have about costs and again, like property value and pace of life and the views out there front windows, all of these things kind of coming together into just a generalized opposition.
And I also think that there is, I don’t know if it’s misinformation so much, but there’s definitely confusion, right? Because these are really complicated topics. I cover energy and as I just admitted, transmission cost allocation, very confusing to me. There are a lot of things that are hard to follow. And so now you’ve got folks on a local level who are trying to track things like data center water use and the types of chips and the cooling proposals and there’s a lot of skepticism I guess and concern around all of those things.
Shayle Kann: Okay. So back to transmission, like I’ve said many times on this podcast and elsewhere that if I could wave my magic wand and change one thing for the better of, I don’t know, climate, electricity, certainly AI, it’s just like we got to be able to find a way to build new transmission in this country and we certainly have not. In your reporting on this particular project, did you come away feeling incrementally more pessimistic about our ability to solve it? Because I read it and I’m like, again, like I said before, this should be a project that should get built. It’s got everything going for it, that we need more capacity in PJM. We have a reliability problem. There is all the load growth. It should happen and yet it seems challenged. So I worry about the takeaways from that.
Maeve Allsup: Yeah. My takeaway for this particular project, again, not having spoken to NextEra and with the caveat that it is early days, the level of opposition and the multiple layers of opposition, right from a local landowner all the way up to state legislators in the case of Maryland, like their office of the consumer advocate taking this all the way to FERC, the level of opposition, the force of the opposition at such an early stage relatively I’m less than optimistic, I guess, about this particular project. I do think maybe a lot will depend on what FERC does in this parallel proceeding, but I feel like they’re running up against all of these macro trends in a way that feels really hard to overcome between now and 2029.
Shayle Kann: All right, before I let you go, any other projects that we should be monitoring as like a bellwether for what’s happening in the market? Weather transmission or otherwise, other stuff that you’ve got your eye on?
Maeve Allsup: Ooh, that’s a good question. I mean, the main thing that we are keeping an eye on is the data center projects themselves, how quickly they go up, the kinds of approvals that they are getting for generation. And we touched briefly earlier on the tariffs that individual utilities are building to kind of slave data center costs, infrastructure costs directly to data centers. There’s opposition, there’s pushback to those two and just kind of Seeing how those play out and how quickly the data centers themselves can actually get online is another big macro trend that we are trying to keep an eye on. But I do, from a narrative perspective, I really like looking at these specific projects because they’re just kind of an interesting lens for what these trends and slightly esoteric debates that we’re having about policy actually mean for immediate term projects.
Shayle Kann: Yeah. And if you don’t look closely at them, you just assume whatever the one-liner is. For me, the thing that’s most interesting about this is just like people know transmission’s hard to build. They think it’s because of NIMBYism. In this case, it’s not really NIMBYism. It’s like a whole other complicated set of things. And so you have to really understand. You have to look at it straight in the eye and try to understand what’s actually happening to figure out why it’s challenging and then hopefully what we can do to turn it around. So with that, thank you for walking me through it. Really appreciate the time.
Maeve Allsup: Thanks so much for having me, Shayle. Great to be here.
Shayle Kann: Maeve Allsup is Latitude Media’s founding reporter. This show is a production of Latitude Media. You can head over to latitudemedia.com for links to today’s topics. This episode was produced by Max Savage Levenson, mixing and theme song by Sean Marquand. Anne Bailey edits the video version of the show. Stephen Lacey is our executive editor. I’m Shayle Kann and this is Catalyst.
