Vermont Legislature SmartTranscripts:

[Speaker 0]: Alright.

[Speaker 1]: Welcome back everybody to House Energy and Digital Infrastructure. It is Thursday, February 26, and we are here to take testimony on H seven twenty seven, an act relating to sustainable data center deployment. We're gonna hear from BPRG, and we are going to hear from VNRC. We're gonna hear from the Lake Champlain Committee. And we have until lunch ish. So I'm representative Kathleen James from Manchester.

[Speaker 2]: Scott Campbell from Saint John's Barry. Richard

[Speaker 3]: Bailey, Lamoille too. Chris Morrow, Windham, Windsor Bennington. Michael Smith, Woodonia two. Christopher

[Speaker 4]: Howland, Rutland four.

[Speaker 1]: Dara Torre, Washington two.

[Speaker 5]: Graham Kleppner, sitting in 13, Burlington.

[Speaker 0]: Laura Sibilia, Windham two.

[Speaker 1]: Great. And in the room.

[Speaker 2]: Dana Lee Perry, Bennington Michael Ellis, Sterling College.

[Speaker 1]: Serena Knight, BNRC.

[Speaker 2]: Jared Carpenter, Lake Champlain, if you may.

[Speaker 5]: Linda Julie Walsh with BPR.

[Speaker 2]: Nathan Eisen, finance journalist. Media, NBC five News, WPTC.

[Speaker 1]: Super. Alright. Who's first? Ben?

[Speaker 3]: You mean? Trying

[Speaker 5]: to join the Zoom.

[Speaker 3]: Okay. Morning.

[Speaker 5]: For the record, Ben Andrew Lee Walsh with the Vermont Public Interest Research Group. I'm gonna join this audio because nobody wants feedback. And Great. So, thank you again, for the opportunity to to come in and speak to the issue of data centers and how if and how Vermont should be regulating them. Obviously, this is an incredibly rapidly emerging area and one with potentially incredibly significant impacts to the state in terms of water, in terms of energy, in terms of toxics. My wheelhouse is energy, so that's what I'll be speaking to. I won't be getting into water or toxics or active 50. I I will say I'm not sure to what degree the other witnesses will be able to get into the issue of toxics, but it is certainly, something that should be considered, and we could potentially bring additional testimony or recommendations on that after the time we break the best interest to the committee. So I wanted to start just by talking a little bit about scale. We often hear that data centers are very large. I mean, physically, they take up a lot of space, but in terms of the energy footprint, of a data center and I wanted to try to get my head around that a little bit more. And I thought that would be helpful for the committee as well. And so what I did was I pulled together a couple of examples, from, legislation, the 20 megawatt data center that's represented in H 727 on the left. And then here is the senate bill, the 100 megawatt data center. That's the moratorium bill. I also did a little bit of research on, you know, how big is a typical data center these days. And the Boston Consulting Group came out with analysis that said right now, they're about 40 megawatts on average with obviously many larger and many smaller, and that number is growing so fast that it'll be 60 megawatts by 2028. Now I didn't dive deeply into, like, what's included and what's not. Certainly, there are some older style data centers that are much smaller. I don't know if that's included in that analysis. The point isn't 60 is a magic number so much as 20 megawatts is not a large data center in the scheme of things as this technology evolves. So just from left to right quickly, 20 megawatts in this bill, 60 megawatts national average as per the Boston Consulting Group. The green bar here is the total electric usage annually for all customers in Vermont Electric Co op territory. So roughly the same as a typical data center will be in a couple of years. Wow. A 100 megawatt data center, that's the Senate bill. 450 megawatt, this is a data center that's been proposed. It is planned. Seems like it will be built in, I think it's Lansing, New York. That would be 2,800, megawatts or essentially half of all the electric usage in the state of Vermont. That's the largest one I could find that had proposed for the Northeast. I should note, I didn't put on here, because it seems like it's a little bit iffy whether it's gonna actually land to be built, but there has been a 150 megawatt data center proposed in Massachusetts, which obviously would be sort of between those two sizes. All of Vermont's electric usage is here on an annual basis. I should note that retail sales, the total amount of electricity that sort of purchased by Vermont utilities is a bit higher than that because of line losses and their own use. And this is a meta, formerly Facebook data center that's proposed for Louisiana. That is, as you can see, about three times the total electric usage of the entire state of Vermont on one campus. Oh, Rutland, I'm gonna try and restrain myself from asking a lot of questions because this looks like it's

[Speaker 0]: gonna be pretty rich. Thank you, Ben. But planned, is that just kind of they've gotten themselves in the cure or are they actually, like, permitting and Oh. In the process?

[Speaker 5]: My the latter. I didn't dive deeply on exactly where each of these is, but, this 2,000 megawatt one, for instance, is one that Mark Zuckerberg talked about as something that is going to happen in an Instagram post. K. And they put out press releases and such.

[Speaker 1]: Ben, I'm sorry. I know we don't wanna bog down. But is there any kind of nexus between the megawatts and the actual size of the building?

[Speaker 5]: Oh, I'm sure there is. I did not attempt to try to pin that down

[Speaker 1]: at all. Okay. Yeah. You're so messy.

[Speaker 0]: I I

[Speaker 5]: thought about it, but I I do the line somewhere. Have land use, not my wheelhouse.

[Speaker 1]: So Yeah. You said that. Alright.

[Speaker 0]: Yeah. Okay.

[Speaker 5]: The other couple of things that I just wanted to mention briefly before I move on to recommendations, this is my one sort of, like, context slide is Well, there's one other example and then one sort of point of reference for how these data centers operate that's different than many other loads. So the other point of reference is we often talk about electric vehicles and heat pumps as other things as like re data center a few years ago. Mostly what we were talking about when we're talking about the grid potentially taking more electricity on in Vermont was EVs and heat pumps. So a 20 megawatt data center is roughly the same electric usage on an annual basis as 35,000 EVs, which is about, I think, twice as many EVs as the state has at this point.

[Speaker 1]: And

[Speaker 5]: then the the last sorry.

[Speaker 1]: Go ahead. 20 megawatt equals 35,000 EVs, twice as many as are on the road in this

[Speaker 5]: Yeah. I didn't double check our current deployment numbers, but it's something like two x total EV deployment at this point. The other thing that's just important to note about data centers and how they operate that's a little different than most other loads is they have a very high, what's called a load factor. Load factor is essentially the percent a percentage of their theoretical maximum usage. So if you had a 100 megawatt load of some kind and it had a 100 load factor of a 100%, that would mean that that data center, that whatever, is operating at a 100 megawatts of usage every hour of the year, never dialing it back at all. If you operate at a 100% half a year and turned off half a year, you'd be at a 50% load factor. If you were always a high 50% and never turned down or up, you'd also be at a 50% load factor. So it can mean different things, but broadly, the higher the number, the more consistently it's operating at or near its capacity. And data centers, Dominion, Virginia put out numbers for real world load factor for the data centers in their territory. That's the territory in The United States with the largest deployment. They're 82%. Duke Energy is planning for 80%, so very close to full capacity all the time. The typical load factor for the grid as a whole in The United States is about 60. I believe ISO New England's a little bit lower than that. And if you're talking about, like, residential load, it'd be 50% or lower. So that's not necessarily a bad thing. If you are able to turn those down, either through curtailment or energy storage on-site at peaks, what you're doing is you're selling more kilowatt hours for the same amount of use of the system. We have an electric system, a grid that is very often not fully utilized. That's intentional. If you can spread the kilowatt the the infrastructure cost over more kilowatt hours, that can drive down electric rates. But you gotta be careful when you're talking about that because the potential for peak increase is so significant with this. You know, one, Vermont's peak electric usage is about 900 megawatts. The peak ever was just over 1,100. That was in 2006. And so if you added a single 100 megawatt center, you're talking about a 10 to 12% increase from one site in the month's total peak if you're not curtailing or otherwise dealing with that?

[Speaker 2]: Yeah. Before you repeat this Yep. You said this, but wanna make sure I understood. This this chart is based on a 100%?

[Speaker 5]: No. I'm sorry. I did not say that. Oh. All the data center numbers on here, I couldn't find any like, you know, Meta is not saying, here's how much electricity we expect to use. They're just talking about the total capacity. This is based on that 80% Oh, it's based on that number. Eighty, eighty two. I rounded. You know, these are not Yeah. Very precise. Yes. That's fine. Okay. So I just wanna I wanna get into our thoughts or recommendations on this build given this context, and then I know you have witnesses on the very important topic of water as well. Mean, don't wanna give them short drift. So, actually, before I everyone starts reading that, I should just say, and I apologize for this. Normally, I try to do a red line that's very clearly, like, page number and all that from the bills introduced. You have language from TJ Poor from the department, Peter Walk from efficiency Vermont. And so what I tried to do is sort of merge these concepts into this language, but that means that I can't give you, like, exactly what I'm proposing you, This is just sort of our thoughts on what the language could end up ultimately look like, and the yellow is mostly my language with a little

[Speaker 0]: mister porn, mister walks folded in. So just making sure I understand, you are attempting to include all versions that all red lines that I

[Speaker 5]: For the little chunks that I'm talking about. Yes. And and the reason I did that was sort of twofold. One, I think, and I'll caveat this in a second, that the contract construct that mister Poor brought forward, and I believe the piece you mentioned as well, as opposed to a tariff that could flow into a contract, is a pretty good one with the caveat that I'm not an attorney. I'm certainly not an attorney with any expertise in contract law. So I would wanna talk to some of my colleagues at other organizations that are attorneys to just ground truth. Is there any sort of thing that we're missing here? And

[Speaker 1]: that's a conversation you can have ASAP?

[Speaker 5]: Yes. I've already started with

[Speaker 1]: Okay.

[Speaker 3]: Yes.

[Speaker 5]: The other reason is his language works fairly well for the concepts that I was thinking about being incorporated. So that's what I'm Okay. So this is again jettisoning the tariff, the large load service equity contract shall. And my suggestion is dropping the language you currently have on the renewable energy standard. In the interest of time, rather than reading all of this, which you have, what I'd like to do is just describe the effects that I'm aiming for, which is for a data center, a 100% of all of the power that they're being contracted with would have to be new renewable energy as, you know, meeting the same definitions as in tier four of the renewable energy standard. So new renewables 2010 or later from New England or the immediately adjacent regions. And there's there's there's two reasons for this. One, from a greenhouse gas standpoint, this is the way that from an energy standpoint, you can get flow I mean, there's caveats about exactly when the renewables are produced and all that, but essentially get to zero greenhouse gases by requiring all of their electricity to come from new renewables. This is additional renewable energy, not just signing a contract for something that already exists and isn't sort of forcing more renewables to get filled.

[Speaker 1]: So I went I missed the last thing you said. Say that again.

[Speaker 5]: The idea is this this language would like tier four of the renewable energy center force new renewables to be built in the region.

[Speaker 1]: Okay. So instead of, because what we were talking about yesterday, and I was trying to think this through last night, was I I think that in the language that you recommend striking now, it's it's more like, okay. You're buying your power from a utility. The utility needs to be compliant with the rest, and that's it. You're you're suggesting, I think, a step. I I would I would think in a better direction, which is that they is this kinda bring your own power? It needs to be new in state?

[Speaker 5]: It doesn't

[Speaker 1]: You didn't say in state.

[Speaker 5]: I didn't say in state, and I I have a note on that. It's not necessarily bring your own power. It's making sure the utilities are selling them only renewable power. So the idea here is if you have a data center, that's a slice over here, you have the renewable energy standard, which governs all the electricity in the state. Now the data center comes in and you're saying, that data center needs to be a 100% renewable as per tier four of the res new renewable. That's totally separate. The and and the math would be totally separate. The calculation would be separate from the res. You the utilities would still have to get to all the numbers in the renewable energy standard, But then the the data center itself would have to purchase from the utility and pay fully for a 100% new renewable energy. It's a higher standard to your point than the renewable energy standard. It requires more renewable energy than the than the, renewable energy standard. But the idea by notwithstanding the requirements of 30 VSA eight thousand and four and five, the res, it essentially is saying data centers are in the separate, higher, more stringent category and everyone else still has to continue along the res as it existed before them.

[Speaker 0]: Okay. And we've got questions. Well, just I, want to speak in support of this concept because of the effect that the proliferation of these large loads is having on the markets and greenhouse gases and potentially the peak. And so this feels like an important affordability context piece as well, potentially. So I need to continue to understand it. Who is next?

[Speaker 2]: So as a practical matter, how would a data center comply with this? How would they know where the electricity is coming from?

[Speaker 5]: So this would be inserted into the provision about the contract they need to sign with the utility. So as I read, mister Poor's proposal with data data center comes in, and they then need to sign a contract with the utility that meets all of these criteria in the bill. The PUC would then need to make sure that contract meets all the criteria in the bill. And so that's how this would be enforced, that they would basically not be able to get power from utilities in Vermont unless they signed a contract that met, among other things, this requirement. Well, I guess what I'm wondering is so some of

[Speaker 2]: the electricity that a utility provides or or obtains and and then provides to customers comes from long term contract. Mhmm. And some of it comes from smart market, whatever the whatever the system mix is. So they would so presumably, the idea here is that then the data center would have to buy RECs to cover that power.

[Speaker 5]: The utility would have to, essentially on their behalf, have to purchase enough tier four eligible RECs to cover all of it. Yeah, just like the renewable energy standard itself.

[Speaker 2]: Well, the utility has to buy the RECs now. Yeah. But in effect, they've they've passed them the the cost onto the the data center. Right?

[Speaker 5]: Yeah. And they would have to buy more RECs than they would under the RECs as it stands. The the renewable energy standard peaks at 40% new renewable energy in 2035, some in state, some regional. This would say from day one when the data center comes on, their contract has to be providing essentially REX Yeah. Or power with REX Yeah. For a 100 of their power being renewable.

[Speaker 2]: And and I don't know if this if this if this matters, but I'm trying to understand how a a data center or how a utility is going to know where that electron comes from. But

[Speaker 5]: Yeah. Just I mean, there's contracts for power and REX just like any other contracts with power and

[Speaker 2]: moment to moment, where where the power is monitored. Is it is it not? Maybe maybe this is too much of a

[Speaker 5]: The last thing I can it's a a much bigger conversation right now the way Vermont's Renewable Energy Standard works is on an annual basis. So you're absolutely right. Electrons are second to second, millisecond to millisecond, but the contracts are so the law reads that it's on an annual basis. And that's this.

[Speaker 2]: I see. Okay. Makes sense.

[Speaker 5]: It follows that annual Right. Thanks. So this would force the utilities to source new renewable electricity faster than they would have otherwise. Yes. And given all the hurdles to building new renewables, one wonders what the effects are going to be. I'm worried about the timing issue. This going to require, if someone builds a data center, that they leave it turned off until the utility has gone and found developers to build a bunch of solar and wind and stuff? It's a fair question. I'd say a couple of things to that. One, any data center that's being proposed in Vermont certainly would be on a multi year time frame between initial proposal and construction. And so there is a fair bit of runway there. Two, these data centers are enormous, but relative to the New England grid as a whole, they are the individual 20 or 100 megawatt data center is not so dramatically changing the total demand for new renewables in New England that it, by our estimation would be impossible to find that new renewable energy on a two year timeframe. I mean, you could hear from the utilities, from the department, but that's our take on this. Thank you. The other thing I just say briefly on this side, and I am very aware that I'm running probably gonna run well over $11.30, but tier two in state that was mentioned. This does not have any language on in state. I thought a lot about that, and I do have a suggestion for how to deal with it. What I was struggling with is what I just said about the region is not necessarily true for renewable development in Vermont. If you had a 100 megawatt data center come online, you would and you and it's like in the out years in the twenty thirties, and the requirement for tier two at that point is like around about 20%. And you had to, over the course of one to three years, bring online enough new solar for them to get 20% of their power from new solar, that would be about a 100 megawatts of new solar. Right now, we're building about 50 megawatts of new solar in the state a year. So you double for a couple of years the amount of solar that needs to be built in the state, and that could drive up prices for everyone. And so simply saying it's 100 a percent new renewable and 20% of that needs to be from in state tier two power seemed like it could have unintended consequences. So my suggestion here, and it's not in this slide, it's something like a reasonable and increasing percentage of which must be tier two qualifying, something like that, but leaving it up to the PUC to decide exactly what that means on a contract by contract basis. So you're you are getting some new renewables in the state, but not blowing up the the market for tier two power in the process. Okay. I have a a few more language suggestions. I think that might be the most complicated. Okay. So that was about renewables. This is about peak and energy efficiency. So this language does a few things, and I should say some of this is language that is essentially Peter Waugh's.

[Speaker 1]: Okay, I wondered about that. So this incorporates Peter's suggestions.

[Speaker 5]: Yes, or at least portions of them. So as well as enhanced energy efficiency standards, so that's Mr. Waugh's language. I I should say, and I have not had a chance to talk to him about this, which is why I haven't incorporated anything more than this. I'm wondering a little bit here about pointing to specific energy standards, stretch codes, something like that. I'm not sure if it's a perfect fit because data centers have such specific requirements, and I'm not really sure how they're incorporated into energy code. But I could imagine this very general language being a little bit too loose. It's a general comment. I'm sorry. I don't have language to propose, but I can work on that. I also added here including efficient use of waste heat, for two reasons. One, as you know, there's been a lot of work in Vermont to develop thermal energy networks, and this is exactly the kind of large source of heat that would be perfect for a thermal energy network. Mhmm. The other reason is the heat from one of these plants is one of the most significant issues with it from both an energy and a water standpoint. So if you were able to have, like, a closed loop water system that is cooled at another factory that needs a lot of heat and then brought back in, you're potentially feeding two birds with one seed as a friend of mine likes to say. So it doesn't really it doesn't get to specifics there, but it speaks to that as a valuable thing to be working on. It requires them to work with their an EU and a distribution utility and coordinate with them. And then this piece is new. This is not a language that mister Porter, mister Walk brought, but speaks to energy affordability and what reps Sibilia was talking about in terms of peak. So these measures, this is the language I'm proposing, shall be sufficient to reasonably ensure that load from the data center will not increase Vermont system peak load or Vermont's peak load coincident with ISO New England system peak load. Essentially what this says is you're bringing a data center and you have to figure out with the utility, the EU and the PUC some way of making sure that when the grid is at peak demand, you are not requiring electricity from the grid. That could mean you turn off for four hours curtailment. That could mean you have if you're a 20 megawatt system, have 20 megawatts of batteries on-site that have several hours worth of storage. So you can turn those on, keep chugging away as doing your work as a data center, but you're not pulling electrons from the grid at that moment. The reason that's so important in our view is this is potentially an enormous cost driver. If we have a data center come on and, you know, year over year, our peak goes up 10%, it was very imaginable. That would contribute to additional costs for us paying for regional transmission projects. This is something Vermont's done a lot of work to reduce the cost of, and one of these data centers could, like, instantly over, you know, overturn years and years of work to reduce Vermont's peak. And, you know, I will say mister Port proposed other language to sort of to try to ensure that these projects pay their full rate. I think that's great. I think a lot of the proposals that you brought in that regard really strengthen the protections in this bill. I think something very specific like this would go a step further and really make sure that we're not seeing peak go up in Vermont due to development of data centers, which is one of the things that's driving rate up in other places that are seeing a lot of data centers. Three more. One, and this is one point that I strongly disagree with mister Poran, although I like a lot of his other language. He suggested that perhaps the PUC could be given the authority to waive the energy efficiency charge, which funds efficiency Vermont, and allow a data center to do its own energy efficiency work, a la GlobalFoundries, the the Sikh program, or, you know, something similar to that. There are few programs similar to that. I do not think the state should be in the business of giving special treatment to new large loads that come in. So what we're suggesting is language that very specifically says they have to pay the energy efficiency charge. It's in law. They should have to pay it anyway. But if we're saying gross receipts tax, which mister Ford proposed, I think we should be pointing out that they have to pay the energy efficiency charge as well. And and this bottom language is almost entirely just copy paste from mister Block's testimony, except for the tariffs pieces deleted, presuming that you may adopt mister Poor's tariff free construct. But this just says all these various programs we have that, like, give individual companies a little bit of a special carve out, data centers can't get them. They just have to pay the EEC, and they work with the EU, and that's it. And last but not least, mister Poor also brought up in his written testimony, I don't remember if it came up in his verbal testimony, the self managed utility concept. So this is what Global Foundries has. They, I think, are still in the process of leaving Green Mountain Power. They're gonna be their own little utility with themselves as their only customer or nearly their only customer absent a couple of people who are wheezing from them. And when that was proposed by Global Foundries, they had all these interesting interpretations of law that basically said, oh, well, because Vermont law uses things like retail electric provider, we're not gonna be a retail electric provider, so we don't need to follow the renewable energy center. We don't need to do this. We don't need to do that. Now ultimately, they didn't. They they are complying with all the all Vermont laws, they have they have some carve outs. But absent something like this that says you just can't feed your own utility, you have to have a contract with some other utility, and that's that. I think a lot of the other work you're doing in this bill could potentially be null and void because the rest of the bill isn't written with a self managed utility in mind. It's written like the contract language. You have to have a contract with the utility. What if you're not being served by a utility? You don't need to have a contract anymore. Magically, you get out of all of these requirements. So rather than rewriting a whole bunch of existing law and this bill to contemplate what if they're in self managed utility, just let them be a self managed utility. They can play by the. And that is my testimony. Thanks,

[Speaker 1]: Ben. That was really helpful. And I know you've submitted that, and it's posted, I think. Great. Oh, yeah. I'm sorry.

[Speaker 5]: Quick question. So, basically, it will be the utility that's gonna have to go out and get the renewable contracts, and you're suggesting that they'll be charging that to that? Yes. Yes. Yeah. That's cost of that contract under their rate, whatever rate. Yeah. Whatever they agree to in that contract would need to incorporate the cost of that renewable energy. Yes. Yep.

[Speaker 2]: Thank you very much. Thanks, Ben.

[Speaker 3]: Thank you.

[Speaker 6]: Oh, nothing to roll. Okay.

[Speaker 4]: No. That's

[Speaker 3]: fine. It takes time.

[Speaker 0]: Ben, if it's alright, we will go. Great.

[Speaker 4]: At the same time? Is that density?

[Speaker 3]: Water people like to do everything together. Yes. Exactly. Energy people, they fly so We

[Speaker 4]: don't have specifically, you know, slideshow or anything like that. We have notes, obviously, testimony, and John and

[Speaker 3]: I are gonna go back and forth

[Speaker 4]: a little bit. For the record, my name is Jared Carpenter. I am an attorney in the water protection advocate with the Lake Champlain Committee, and this is.

[Speaker 3]: I'm John Grumman. I'm the policy and water program director for Vermont Natural Resources Council, also, an attorney. Worked on water issues for decades, I think.

[Speaker 0]: Yes. Just to since you don't have a presentation, I would just ask you to be explicit about which version you are talking about, if you are making recommendations, if it's as as introduced or if you're looking at the red line from the

[Speaker 3]: So I think our recommendations are more more general. Okay. I mean, I I looked at the as introduced, though, but I think that

[Speaker 0]: So you're not

[Speaker 3]: interested this. Would apply.

[Speaker 5]: Yeah.

[Speaker 3]: We have some suggestions for language.

[Speaker 2]: But they could

[Speaker 4]: go into any version because of where you're dealing with with with water consumption, which isn't really addressed in any of the versions quite yet. Mhmm. Committee's discretion of where it would be in food and so to You got it. Now John's been working on the on water issues for a long time. I've been, you know, in this building for ten years starting with the the Montclaim Water Act and then working on water policy sort of all the way through. So this is a new issue for us just like with everybody else. We're still wrapping our head around it, and we're working on on getting some some suggestions, and we have some here today. It's it's very concerning. The size of these size of these data centers, even a small one uses 300,000 gallons of water a day intake. It evaporates 80%, and then the rest would be discharged as as as thermal discharge as wastewater. You would need a a fairly decent sized water source. You're not gonna put one, I would think, on the Winooski River out here because you'd overwhelm it. So one would think that this

[Speaker 3]: would have to go somewhere on

[Speaker 4]: the Connecticut River. This committee has discussed the burner location where Maumayankee was potentially, or on Lake Champlain, Lake Memphremagog, Wallace Pond, the Connecticut Lakes, one of our bigger waters bodies of water, thus it being a concern for the Lake

[Speaker 3]: Champlain Committee. We Or or groundwater.

[Speaker 4]: Or ground we can talk about groundwater as well.

[Speaker 0]: Okay. And just yesterday, we heard discussion from doctor Wen about these being connected, like, largely being connected to public. I see that systems.

[Speaker 3]: We do I like to submit to the committee. I just didn't have time, but some general good resources, some pretty reputable resources on data centers and water. And they do discuss using public water supplies. I mean, in Vermont, that would be a challenge, you know, except for maybe Lake Lake Lake Champlain, which raises a host of other issues. But they could it really is and we have some experience because we've had the cooling issue we've seen with energy plants. Right? We've seen it with certainly Vermont Yankee, and we've seen it with biomass facilities in the state. So and all the biomass facilities, some of looked at groundwater, most looked at surface water. None of them have looked at water supplies. And Vermont Yankee was obviously Connecticut River taking the water. And I just wanna I just wanna emphasize the the thermal discharges have serious impact, and we learned that with Vermont Yankee. We wrestled with Vermont Yankee, as you know, for decades, on a host of issues, including the water discharge issue. And Since Rimonton Yankie has been closed down, we have seen the fish populations come back in tremendous numbers. We always had this debate about what was the impact of this thermal pollution. We knew there was an impact, and thermal pollution is in the Clean Water Act. It is in the water quality standards. You need a Clean Water Act federal discharge permit for thermal pollution. And it's just a matter of one of the limits. But it's really incredible to see once we turned off that heat, how the water quality and

[Speaker 0]: the fisheries will come back. So, mister Groveman, we're trying to figure out gaps. So I think you just said that thermal pollution is something that would be considered any

[Speaker 4]: They would

[Speaker 3]: need a federal clean water act. Yeah.

[Speaker 4]: And this committee yesterday, you kept talking about dual jurisdiction of, you know, would be DPS or PUC. But there's really another I mean, ANR has jurisdiction over this as well in terms of the in terms of the water quality impacts, including thermal discharge, the water intakes. There could potentially be, you know, storm water permitting, shoreland impact permitting. Right. Yeah. Know, if you're moving Earth around, you could have army corps permitting. So there's a there'll be a a a number of reviews of this that are already in place,

[Speaker 3]: but we would also you know, we'll suggest some some additional reviews as well. Right. And and you have experience with so under section two forty eight, you know, you're it seems like the bills are, you know, replicating those the environmental criteria, including incorporating the active 50 criteria. And with regard to whether it was Vermont Yankee, whether it was, a biomass facility, so you have the ANR permits, but then you still have to go to the PUC and there's this broader umbrella undue adverse effect on the environment. The permits are used as evidence to address issues, but the PUC with regard to these energy issues with regard to thermal pollution could go further than that. And so we have experience with that, and Act two fifty does serves the same role. If Act two fifty is involved, the ANR permits are evidence. They're strong evidence. They legally create presumptions of compliance with some of the criteria, but, the state or other advocates or or concerned citizens could say, even though they have their permit, there are issues not addressed in the permit and urge the PUC to go. So we have all that experience with those regulatory processes. Other states don't have Act two fifty. I don't know what the other states citing laws are, like section two forty eight, but we do have those good umbrella protections, which we have experience using. That is positive, but you wanna make sure that they're in the bill and and, you know, be explicit about, you know, how the how how they're gonna be applied.

[Speaker 2]: It's question. A Oh, sorry. That's it. You're Yeah.

[Speaker 1]: Go ahead.

[Speaker 2]: You're gonna mix it.

[Speaker 0]: I didn't raise my hand.

[Speaker 1]: Oh, okay. He's made a sound.

[Speaker 0]: I did. Okay. He's made So

[Speaker 2]: About thermal pollution, and and it was interesting what you mentioned about about fish populations returning in in Connecticut River. Do do do you have I imagine we have data about about how much pollution thermal pollution. How much did the oh, my Yankee raise the temperature in their pumps?

[Speaker 3]: Well, I'd like to give you all that data. Yeah. I I don't have those atomic heads.

[Speaker 2]: Roughly two degrees, five degrees,

[Speaker 3]: and, you know I honestly just don't

[Speaker 5]: I don't remember how to look

[Speaker 3]: at that. But was fairly significant, that was the debate Yeah. About where is the and there's a thermal plume. It's not just raising the temperature Right. Like, at the discharge point. Yeah. A lot of the debate was like, well, how far to anticipate? You know? And there was a lot of debate about that thermal plume. Obviously, you know, Vermont Yankee say, well, it doesn't go it's not it doesn't go that far down.

[Speaker 2]: That's right.

[Speaker 3]: And then the advocates would say, Dean was You you could have David Dean, he's the expert. As a former legislator, he has studied this upside down and sideways as he was not only the chair obviously of the Fish and Wildlife and Water Committee, but he was the Connecticut River water keeper. And So the Pluto- Okay.

[Speaker 2]: Just curious. Thank you.

[Speaker 3]: Yeah. Think somewhere I suggest, the ANR permitting, so the federal Clean Water Act thermal permitting, I think that is pretty set and they'll have to go through that. But the debates are, as experienced with the Yankee, there is disagreements about how much needs to be done. So that will, I'm sure, happen if we get a giant data center. Then there is like, where's the water gonna come from? If it comes from groundwater, we have, we put in place in 2008 a very stringent groundwater withdrawal permitting program, and we declared groundwater a public truck resource, which I think we're we're one of two states that have done that. And so if you're if you're withdrawing more than basically 50,000 gallons per day of water, and as Jared said, these for cooling is the low end is 300,000 gallons per day, and the high end seems to be 5,000,000 gallons per day. You would need one of these permits, and they're not easy to get. I don't think anyone's tried, actually, since we put in the law in 2000. It was really set up for water bottling. I don't know if you remember there there was a big issue back at that time in Vermont. There were all there there was a Vermont pure water bottling and there were a number of other water bottlers that were looking. And so then we we haven't had one come in since that long. If a data center wanted to use groundwater, that would be a good protection that we have. Jared could talk to, in addition to the Clean Water Act, we're actually developing water withdrawal laws.

[Speaker 4]: We are. It was actually a little it was a little timely on our part. So back in 2022, this body passed act one one thirty five. It's an act relating to surface water withdrawals and interbasin transfers. And the thought was, along that time, is that in times of drought, there would be competition for water, and therefore, you know, water would have to be divided up and and allocated to the users. So for the past two years, surface water users have been required if they withdraw a certain amount of water to register and report to DEC. And if you're a farmer over to AFM, you know, the amount of water you're being used you're using to sort of establish a baseline. Rule making is going to affect this summer. The rule is supposed to be in effect 07/01/2026. I think the rivers program is running a little behind in it, but their plan is to have it in place this summer. You know, and it's so I have some notes on it, and it's based on the potential impacts to surface waters and other factors. They have to follow the Vermont Water Quality Standards, and the rules require efficient use and conservation of surface water, establish withdrawal limits based on low flow or drought, and require assessment of reasonable and feasible alternatives to propose withdrawals. So what they are proposing is that anybody who has been to withdraw water from a surface water, so from a river or a lake, is gonna have to get a permit if they withdraw above de minimis standard of water. And de minimis is defined as, for water terms, 5% of the of the seven to q 10 flow. Seven to q 10 drought level is the low flow of a river in drought conditions as the lowest seven day average flow that occurs once every ten years. So it's pretty low. So if you draw 5% of that, you're gonna have to get a permit in order to withdraw surface waters. You have to meet water quality standards based on acute cumulative effects of other withdrawals. So it's not just you, it's everybody else who's withdrawing water around you because obviously water withdrawals is cumulative. And then which is key is you have to provide an alternative analysis of, during drought, an alternative analysis of what your alternative water what's your backup plan? So if the river gets down to a certain level, if the lake gets down to a certain level, you know, where are you gonna get your backup? Where are you gonna get your water from? And this is something that Ben touched on about which, you know, this is John and I are still learning about this. But if if the river drops to a certain level like it did last August Right. Can you turn these things off? I mean, can you just sort of turn the data center off and it doesn't get its water anymore? Because you're gonna have to meet the other priority uses of the of the designated uses of the river, which is public water supply, aquatic habitat, different forms of recreation. There are there are in rules certain certain qualifications called designated uses that have to be met. This would not be one of those. So

[Speaker 0]: Is ag one of those?

[Speaker 4]: Yes. Irrigation. Okay. Irrigation for crops and other another water use. So ag is the seventh one.

[Speaker 3]: And ag is not subject to, you know, to the withdrawal, we they're don't need to get a permit? No. Ag doesn't. They do have to report.

[Speaker 0]: Yep. They

[Speaker 3]: do. But but they don't have to get a permit. And we wanna make they are priority use as is water supply for if it's for drinking water if it's used for that. Importantly, this is not in place yet. So the timing so we I think we feel validated. We argued for this a couple of years ago saying, hey. We don't have Yes, thermal discharge? Yes, but we don't have a comprehensive program for withdrawals. Unless you go through Act two fifty, really it's more of an Act two fifty issue than Section two forty eight, you wouldn't even question the withdrawal. So we wanted a comprehensive program and then have the criteria, as Jarrod described, to have a rational way to make sure we were looking cumulative to all the users. And then we had the drought this summer, which sort of proved the point that we really it just the government doesn't have rules yet about how to manage those drought situations.

[Speaker 4]: Withdrawal. The bill also has the act also has something a provision in it that I think would provide precedent for for this purpose for large water withdrawals, and that's the interbasin transfers. So what the example was there was that Killington was drawing water off the Ottoquichi River, using it to make snow on Pico, and then that was melting and going into the Otter Creek. It's an interbasin transfer between watershed basins because the Ottoquiche goes into the White River, which goes into the Connecticut River, which goes into Long Island Sound, where Pico melts into the Otter Creek, it goes north into Lake Champlain. So you're you're taking water out and moving it. And so under this large withdrawal, in order to do this, they had to get what's called a water quality certificate, which is normally associated with let's see. It's associated with the Clean Water Act two four zero one,

[Speaker 2]: but it

[Speaker 4]: requires it requires a lot of studies are done. What are the impacts on aquatic habitat gonna be? What are the impacts on erosion, on withdrawal, on other uses? This is just something that's that's standard. It's usually under a a FERC standard for for dams and other other generators of energy that, you know, we would have to figure out how the language would work, but would be a good qualifier to say if you're gonna do this, you would need this volume of water. You know, this isn't some little business that's gonna draw off, you know, 500, 600 gallons a day. We're talking an amazing amount of water.

[Speaker 3]: Right. That's what we need to see. So this is a gap in the also, Vermont Yankee had to get the thermal discharge permit, had to also get this water quality certificate because it was an energy generating facility that was licensed by the Nuclear Regulatory Commission. So basically, the Clean Water Act, any facility that is licensed by FERC, Federal Energy Regulatory Commission, Nuclear Regulatory Commission, if you need an army corps of engineer permit, then it triggers, the requirement that you get this It's what Jarrod described as also this holistic water quality review. So Vermont Yankee had to get his discharge permit. We didn't have a withdrawal permit, but they had to deal with those issues, I guess, under February to some extent, and they had to deal with them through this water quality certificate. So one of our ideas is could be fashioned into the law that these data centers would have to essentially meet their criteria for a water quality certificate even though they don't trigger that review. They wouldn't otherwise trigger the work quality certificate review. Was one factor that came up yesterday, I don't know, John, I

[Speaker 4]: haven't really had a chance to talk about it, but I think it was TJ that brought up that some of these facilities generate their own power, and then they could sell it on tele access onto the grid. I think that might trigger a FERC review because they are generating power and putting it out there in the same way, you know, a hydroelectric facility or something else would. But I don't think you'd wanna wait for that possibility to happen in order to require this full review. We want it to happen for any data center no matter if they're drawing off the grid or creating their own power.

[Speaker 0]: I wanna clarify. So you're suggesting that we would include in the bill the language of the type of review that would be necessary if they were subjected to

[Speaker 3]: Exactly. Right. Exactly.

[Speaker 4]: As well as every other a and r you know, your standard a and r permits for stormwater and and wastewater discharge and thermal and everything else, this would be this would be this would be the belt and suspenders version, I believe.

[Speaker 3]: Yeah. This would be everything that certainly, like, Vermont Yankee or even a bombastic facility would have to comply with. Statistically, Chair said, making sure you're checking all the boxes. Otherwise, if it's not in the bill, it probably wouldn't apply except for the circumstance that Chair indicated that it would. And it can be a very it's a very helpful holistic water quality review. It kinda looks at all of the water quality issues together.

[Speaker 4]: At the end of the day, you're still allowed

[Speaker 3]: to withdraw the water. With conditions. With conditions. Lots of conditions, and that's a

[Speaker 0]: debate about It's not part of the typical activity No.

[Speaker 3]: Assess No.

[Speaker 0]: It's not that. Or okay.

[Speaker 1]: So it'll be I know we talked about this earlier, but for me, it it anyway, it would be really important if you guys could submit written testimony that we can

[Speaker 3]: Yeah. We're, like, doing this on the fly, like you said. But so the but we could we could work on language, and we did talk you know, we work with other war groups in the state. We we meet every week during the sessions to kinda follow the bills and kinda brainstorm. If something like this comes up, it's very helpful. People have different areas of expertise. So we can work so I assume you're not you're have town meeting break and they have you got something to you next week while you're on break so you can shape it a little.

[Speaker 1]: I think the plan is to try to wrap up our you know, the most immediate, all the kind of foundational testimony this week, and then get comments back to us ASAP so that we can have next week to

[Speaker 0]: do a draft meeting with Maria on Friday Yep. Tomorrow to try and get the

[Speaker 4]: next draft started. We could, I think, at least give you language. It's a it's a placeholder that we could then tinker with a little bit Yeah. Over next week. And like I said, it wouldn't be real I don't think it'll be reliant on the rest of the bill. This would be sort of it doesn't do with energy. It would be sort of its own that you could just sort of drop into whatever whatever bill you decide to pursue.

[Speaker 3]: Yeah. Whatever choices you make about the energy issues and the other issues like this should be plugged in or you could talk about plugging it in and what it should say. But these are the general This is the lay of the regulatory land and some of the gaps that we see. And yes, we can craft language that people then can look at in terms of well, we should make sure we're requiring this and we're requiring that. So yes, we we could we could certainly do that. I mean, we could get you something tomorrow, but it'd be better if we can have a little time, like, into the early part of next week.

[Speaker 1]: Yeah. I mean,

[Speaker 0]: if you can get just the title and, like, the

[Speaker 2]: piece of the shoulder.

[Speaker 4]: Great. Yeah. And another and another suggestion would be that Ben talked about it a little bit. There's an open loop system and a closed loop system Yep. We're learning about. So open loop, my understanding is you lose 80% of the water to evaporation. So if you're drawing 300,000 gallons, you lose 80%. The rest would become thermal discharge, and you're open loop that's coming through. Under a closed loop system, I'm not sure how the technology works. You capture the evaporation and you reuse the water. Yeah. I'm sure you'd lose some, so you'd have to draw in some extra water, but it would

[Speaker 3]: It mitigates certainly. It really removes the water. We talked about I don't know if they and I some of the I'll send you the reports that I had, which they're from, like I said, from from national sort of water, you know, think tank type groups that have looked at data center issues. You know, one of the suggestions is requiring or incur like, you know, it's certainly encouraging, like, closed loop, you know, rather than going to, like, Connecticut River or certainly a groundwater withdrawal.

[Speaker 0]: Is that Conserve Mhmm. That promotes conservation. Yes. Does it reduce You're still gonna have to wastewater?

[Speaker 4]: That's a good question. I think I think you were still gonna need all those permits. You're still gonna need a withdrawal permit. You should still do the studies. You're still gonna discharge.

[Speaker 3]: Yeah. You're the it'll still be hot water, just not so much. But it shouldn't reduce at all aspects of it. Right. If you're keeping it flowing within the system and minimizing what's discharging out into the water, minimizing the water that needs to be added. It is something like it does not because evaporation, you're gonna need Yeah. Water, but you need less water. You discharge less water, and that means less impacts, obviously.

[Speaker 1]: When we heard from doctor Renden yesterday, he did show us a like, a dry a diagram of a closed loop, and it it showed some some water, you know, condensing out the top, basically. I But and we asked how much it like, what was that percentage? And I feel like he didn't I'm not sure if he knew.

[Speaker 4]: I don't think it could ever be a completely I don't know. I'm I'm I'm I'm guessing guessing here. I don't think it could be a completely closed system. I think he would lose over time through evaporation and other things. He'd have to he'd have to draw water off on a on a surface water or groundwater on a regular basis, but I think it would just

[Speaker 3]: And that's what these little little reports say, and I just you know, it's going on the Internet last night. I mean, there's a lot out there on and it's basically saying what we're talk we're talking about. It's not always you're gonna lose some water, but it does it minimizes all all of the need for additional water, and it reduces the dish the discharge.

[Speaker 1]: And he did say and I don't remember the exact stat, but he said that some particularly high percentage of these were on municipal water supply.

[Speaker 3]: Yeah. That's weird. Is I did see that. I agree. And it's such I don't know. Know, I don't you might wanna have A and R in, like, the water supply people. I guess I can't It have to be really large water supplies. Like, the smaller water supplies, even the medium ones don't have. Like, Montpelier, Burlington. I mean, this is not water. Lake Champlain? Okay. Yeah. I mean, I can see that you know, if you're gonna put it in Vernon, think they're gonna try to replicate what I would imagine whatever my Yankee did. I think so. The Connecticut River is there's a lot of water in that part of the Connecticut River.

[Speaker 4]: Well, Gray, you talked about it a little bit. I mean, you can get your water from different sources. I mean, they need to have, you know, bathrooms and everything else too, and that could come off the city water system within the meantime, they're drying off the surface water. So I you know? But I can't I couldn't I couldn't fathom where you'd be able to put these in just sort just sort of Yeah.

[Speaker 3]: It's nothing bad. I don't know what's happening in Virginia where there's a lot of these what the size of their water supplies are. I mean, Vermont, obviously, we're a different scale. We said, obviously, if if you put this in the Chinchicka County and Lake Campaign, it's a water source for drinking water source for people in Chinchickney County from Burlington, you know, down, you know, past Shelburne. So

[Speaker 1]: K. We have additional questions?

[Speaker 6]: No. If you're talking a closed loop system, you run it through some condensers, and condensers are gonna use more power. So now a 20 megawatt facility becomes a 40 megawatt facility because of the antigen itself.

[Speaker 4]: I have read that. I don't I don't know exactly how much percentage more, but there in there in lies the rough. You're gonna use more No.

[Speaker 6]: I right. I just look at the size of of the cooling system has a, effect on the load of the system to begin with. So a small system becomes a medium or or large.

[Speaker 4]: Yeah. And I've also read that there's, you know, there's systems that use air conditioning to I can't even can't fathom. Well, that's exactly energy that

[Speaker 3]: would Yes. I saw the air conditioning.

[Speaker 4]: Yeah. That's a really? Air conditioning. It all it all sounds like The closed loop water seems

[Speaker 6]: to make it all it all comes down to this Newton's laws.

[Speaker 4]: Then he's gotta come from somewhere. And he goes, right. Can't be created

[Speaker 3]: But but, again, I I I'm really glad that the committee's looking at it. This I mean, it is a fast emerging issue, but it is obviously very, very concerning. So This part suddenly has to go with every committee.

[Speaker 1]: Yeah. This

[Speaker 4]: part. This is the first

[Speaker 0]: time I've testified in front of this committee. Think John too.

[Speaker 3]: Yes. Yes. We we don't usually We

[Speaker 6]: spend a lot of

[Speaker 3]: allowed to come here.

[Speaker 4]: So Yes. So We spend a lot of time going to committees. But, no, Christopher Simon, great to see you all. Thank you for the opportunity to participate. Yeah.

[Speaker 1]: Yeah. Thank you much. Thank you so much for your time. We will look forward

[Speaker 0]: to getting to see you.

[Speaker 3]: Yeah. Send you those reports right away. We'll work on some some Yeah.

[Speaker 4]: We'll get it to you as soon we do.

[Speaker 1]: Alright. Thank

[Speaker 0]: you. Thank

[Speaker 4]: you, Sonia.

[Speaker 1]: We can go outside. Howland voting

[Speaker 6]: going?