رونوشت 🔗
WEBVTT
00:00:00.080 --> 00:00:00.640
Good afternoon.
00:00:00.800 --> 00:00:05.599
Welcome to our podcast for the Northeast Ohio Regional Sewer District, Clean Water Works.
00:00:05.759 --> 00:00:12.960
We're excited to welcome you to this conversation about data centers and their current and future impacts on water resources.
00:00:13.439 --> 00:00:15.679
Today's virtual program is the first for us.
00:00:15.759 --> 00:00:20.079
We usually conduct interviews in our offices here in Midtown Cleveland.
00:00:20.399 --> 00:00:21.679
Donna, how's it going?
00:00:22.079 --> 00:00:22.480
It's good.
00:00:22.559 --> 00:00:23.920
I'm excited to be here today.
00:00:24.000 --> 00:00:27.039
You know, this has been a long time coming and a long time planning.
00:00:27.120 --> 00:00:32.159
We've been talking about a data center podcast for a long time, but it's finally here.
00:00:32.560 --> 00:00:34.719
Today we are hitting a somewhat hot button issue.
00:00:34.960 --> 00:00:37.600
Artificial intelligence seems to be everywhere lately.
00:00:37.759 --> 00:00:48.320
Whether we're using a chatbot or saving files to the cloud or asking Google a question, we are relying on infrastructure that is largely invisible to us, data centers.
00:00:48.640 --> 00:00:56.479
As demand for AI and cloud computing accelerates, so do questions about the resources needed to support the digital economy.
00:00:56.719 --> 00:00:58.079
Resources like water.
00:00:58.479 --> 00:01:02.640
Here at Claim Waterworks, we're very interested in bringing some light to this topic.
00:01:02.960 --> 00:01:09.599
I don't know about you all, but a lot of the conversations that I have with my friends or family, it's just very high level.
00:01:09.760 --> 00:01:21.040
And there's like a lot of like fear and uncertainty when it comes to data centers, a lot of like random statistics about amount of water per Google search and things like that that can be really daunting.
00:01:21.280 --> 00:01:24.000
And so I'm really excited for the panel that we have today.
00:01:24.159 --> 00:01:32.319
We really brought together a group of research, um, utility operations policy to help us unpack some of these questions.
00:01:32.719 --> 00:01:33.680
So some introductions.
00:01:33.840 --> 00:01:34.079
Dr.
00:01:34.159 --> 00:01:41.680
Landon Marston is associate professor at Virginia Polytechnic Institute and State University, also known as Virginia Tech.
00:01:41.840 --> 00:01:45.519
And he holds a PhD in civil and environmental engineering.
00:01:45.760 --> 00:01:55.359
His research group looks at how and where water is used within the economy, as well as the environmental consequences associated with economic production and consumption.
00:01:55.760 --> 00:02:06.159
Some of our out-of-state listeners may not know, but Ohio is actually a national hub for data centers with over 200 facilities, with most of those concentrated in the Columbus area.
00:02:06.319 --> 00:02:12.479
So today we have John Newsom is an
Load full transcript
administrator of the Division of Water at Columbus Water and Power.
00:02:12.639 --> 00:02:16.879
He served with the city's Department of Public Utility since 2006.
00:02:17.039 --> 00:02:23.280
And we've asked John to give us CWP's perspective on this emerging technology and water demands.
00:02:24.240 --> 00:02:30.719
And then also with us today from Columbus Water and Power is Sarah Bloom Anderson, Assistant Director of Policy.
00:02:31.199 --> 00:02:38.479
Sarah will share some information about the wastewater side, what happens to the water that's used to cool some of these servers and these data centers.
00:02:39.280 --> 00:02:42.800
And then bringing us back to the deep science of it all, we've got Dr.
00:02:42.960 --> 00:02:57.280
Peter Gravat, CEO of the Water Research Foundation, a nonprofit education organization in Washington, DC, which funds and publishes research on the technology, operation, and management of drinking water, wastewater, and stormwater systems.
00:02:57.439 --> 00:03:06.400
We really look forward to hearing about current research projects and how you're really supporting that work through water use, reuse, and efficiency.
00:03:06.639 --> 00:03:06.879
Dr.
00:03:07.039 --> 00:03:08.639
Marston, kicking it over to you.
00:03:08.800 --> 00:03:18.000
I wanted to start with you because I think most people, you know, they know they're using data centers every day, but maybe they don't actually understand what that means.
00:03:18.080 --> 00:03:22.000
And so I'm hoping you can help us demythify data centers a little bit.
00:03:22.159 --> 00:03:23.199
What is a data center?
00:03:23.280 --> 00:03:28.800
How does it to support our Google searches and all of our other cloud usage?
00:03:29.039 --> 00:03:32.400
And from what I've learned, not all data centers are created equally.
00:03:32.479 --> 00:03:37.280
And so if you could talk a little bit about that in terms of scale and size, that would be great too.
00:03:37.759 --> 00:03:38.080
Sure.
00:03:38.159 --> 00:03:38.800
Yeah, thanks.
00:03:38.879 --> 00:03:45.599
And as Mike kind of said in the opening, data centers are this invisible infrastructure that underpins our digital lives.
00:03:45.680 --> 00:03:54.080
And so when we sign on to Facebook or when we watch a YouTube video or log into our bank, all of these things are run through a data center.
00:03:54.159 --> 00:04:01.840
And so even if you might not realize that there's a physical place where that where all those requests go to, these things are all around.
00:04:01.919 --> 00:04:07.520
And in in the United States, this is where all almost half of the data centers globally are located.
00:04:07.680 --> 00:04:10.400
As uh Donna mentioned, a lot of these are located in Ohio.
00:04:10.639 --> 00:04:15.360
Where I'm at in Virginia, actually Northern Virginia, it's the global epicenter of data centers.
00:04:15.439 --> 00:04:22.399
And so we see uh hundreds of these data centers located uh throughout Northern Virginia,
continuing to add more and more.
00:04:22.560 --> 00:04:25.279
And these have been around in some cases for a couple of decades.
00:04:25.360 --> 00:04:28.160
But in fact, there's actually multiple different types of data centers.
00:04:28.319 --> 00:04:36.079
I think it's important to make that distinction, particularly when we're talking about water, because it can have very different uh meaning when it comes to overall water use and discharges.
00:04:36.319 --> 00:04:41.680
Going back, you know, three or four decades, you would see what was called an enterprise data center.
00:04:41.759 --> 00:04:43.920
And this is what most data centers were at the time.
00:04:44.000 --> 00:04:47.600
This might be like a room within an office building.
00:04:47.680 --> 00:05:05.040
Um, and so wouldn't be like what you think of now, which are going to be these hyperscale data centers, the data centers that are the size of maybe a small college campus, for instance, that comprised multiple, you know, maybe hundreds of acres, um, and in some cases now thousands of acres of land that house these server racks.
00:05:05.199 --> 00:05:09.360
Before, you know, going back several decades, it would just be one part of a larger building.
00:05:09.519 --> 00:05:14.879
That room, the data center room would rely on the same cooling system that the overall building would rely on.
00:05:15.040 --> 00:05:18.240
And no one really thought about the water consumption associated with those.
00:05:18.319 --> 00:05:26.879
It's really only in the last one to two decades where we've started to start pay more attention to the environmental footprint of these data centers.
00:05:26.959 --> 00:05:35.040
And that's mainly because of the scale of the data centers, moving again from these kind of enterprise closet data centers to co-location data centers.
00:05:35.120 --> 00:05:36.800
This is kind of like a co-location data center.
00:05:36.959 --> 00:05:39.839
You can think about like a you know, a storage unit, right?
00:05:39.920 --> 00:05:44.480
And you go and you store your personal belongings there, but you don't own the whole unit.
00:05:44.720 --> 00:05:46.800
Someone else owns that, you rent space from them.
00:05:46.879 --> 00:05:56.079
A co-location center can be similar to that, where maybe different companies will rent space, rent servers to house their data, to process different transactions.
00:05:56.160 --> 00:05:58.800
And now moving to these hyperscale data centers.
00:05:58.879 --> 00:06:07.920
And so these again can be the size of warehouses, and now more recently with the AI, hyperscale data centers can be the size again of you know small college campuses.
00:06:08.160 --> 00:06:14.800
Obviously, there's a lot more compute that happens with that, but there's also a lot greater energy requirements and a lot greater cooling demands.
00:06:14.879 --> 00:06:21.439
And this is really when you were talking about the intersection of data centers and water, this is really where that occurs is that cooling.
00:06:21.759 --> 00:06:28.879
When energy comes in, electricity comes into the server, and some of that's used for compute, but then some of that manifests itself as heat.
00:06:29.040 --> 00:06:31.519
That heat has to be dissipated in some way, right?
00:06:31.680 --> 00:06:38.000
Uh these operate within a certain threshold, temperature threshold, and if it gets too hot, they're gonna they're not going to work effectively.
00:06:38.160 --> 00:06:41.439
Well, it takes a lot of energy to cool a really warm system.
00:06:41.519 --> 00:06:43.199
And so there's a few different ways to do that.
00:06:43.279 --> 00:06:44.720
We'll talk more about that later.
00:06:45.040 --> 00:06:48.240
Many of the ways to do that, the most efficient ways to do that, require water.
00:06:48.480 --> 00:06:51.279
Most of it's going to be uh for that cooling purpose.
00:06:51.439 --> 00:06:58.000
And then we have this indirect water use as well, um, which is low, which is associated with the um the power plants.
00:06:58.079 --> 00:07:08.560
And so these data centers consume a tremendous amount of energy, and with that energy, there's a lot of water that that's required to generate that energy, especially when you're talking about coal and natural gas and nuclear power.
00:07:08.720 --> 00:07:14.800
And I think increasingly what's captured public attention are these ladder type, these hyperscale AI data centers.
00:07:14.879 --> 00:07:22.399
These are the ones that are really large and that again require a lot of energy and also require a lot of water.
00:07:22.879 --> 00:07:28.240
Listeners here in Cleveland might be surprised to learn that we already have about 24 data centers in the area.
00:07:28.480 --> 00:07:37.360
Here in Cleveland, uh, this May, the city rejected a permit application for one of these large hyperscale facilities in the Slavic village neighborhood.
00:07:37.600 --> 00:07:45.519
That particular data center would have used 150,000 megawatts of power each day, enough to power 100,000 homes.
00:07:45.759 --> 00:07:49.839
And Cleveland City Council has also put a three-month hold on new permit requests.
00:07:50.000 --> 00:07:52.639
John Newsome from Columbus Water and Power.
00:07:52.879 --> 00:07:59.040
In Columbus, there are over 130 data centers, making it the seventh largest data hub in the world.
00:07:59.279 --> 00:08:01.360
What does this look like for your utility?
00:08:01.439 --> 00:08:06.959
And what are the challenges and the conversations happening around the increased demands for water?
00:08:07.360 --> 00:08:11.199
Yeah, for from a water standpoint, we actually looked at this recently.
00:08:11.360 --> 00:08:15.600
Uh, we've got, as you said, uh a large number of data centers here in Columbus.
00:08:15.759 --> 00:08:18.160
Only a handful of those are hyperscalers.
00:08:18.319 --> 00:08:23.279
And a lot of those in our are suburban uh communities, which we provide water to.
00:08:23.600 --> 00:08:41.120
From a water standpoint, uh, not a huge use, what we're seeing in the overall scale of things, but we are still planning, looking for the future, looking at different ways that uh we can maybe take them off a potable water standpoint, uh, use maybe a recycled water, reuse water.
00:08:41.360 --> 00:08:47.279
Columbus is a big growing region, expect another million people here in the next roughly 30 years.
00:08:47.440 --> 00:08:59.519
And then we also have an Intel facility coming, so we got a lot of growth beyond data centers, but yeah, that's one misconception I think is out there that uh these data centers in Columbus specifically are using so much water.
00:08:59.600 --> 00:09:04.080
And we're at current case, we're we're not we're not seeing that, or at least at this point.
00:09:04.720 --> 00:09:10.080
So it's not having as much of an impact on water usage as some might think.
00:09:10.879 --> 00:09:11.279
Correct.
00:09:11.519 --> 00:09:11.759
Correct.
00:09:11.919 --> 00:09:17.279
Now we're in a we're in a hot period right now, so there could be some high peak demands.
00:09:17.360 --> 00:09:22.000
Uh, we don't know yet, but it's not a huge quantity of water required.
00:09:22.240 --> 00:09:27.600
One of the reasons locating in Ohio we'll say is um climate.
00:09:27.679 --> 00:09:34.799
That means they can use ambient air uh basically from October up to we'll say March for that cooling.
00:09:34.960 --> 00:09:38.480
So uh that's it helps them with their efficiency as well.
00:09:39.200 --> 00:09:40.559
We are a big system.
00:09:40.720 --> 00:09:41.919
We have hyperscalers.
00:09:42.000 --> 00:09:46.320
There's many systems that wouldn't even be able to serve a hyperscaler.
00:09:46.480 --> 00:09:52.639
Um, you know, we serve 1.4 million people and over 54 billion gallons of drinking water per year.
00:09:52.799 --> 00:10:05.759
You know, if you put it in perspective, we aren't feeling this yet, where maybe other utilities that are much smaller and have a much smaller daily average, you know, they could probably feel it at a different level.
00:10:06.159 --> 00:10:13.759
Landon, I think you had an interesting statistic about that that I heard you mention on another platform about national use versus local use.
00:10:13.919 --> 00:10:15.120
Can you talk about that a little?
00:10:15.519 --> 00:10:16.080
Sure, yeah.
00:10:16.159 --> 00:10:34.799
So if you look at the national scale and how we use water across the US economy, um, the vast majority of the water that we withdraw and consume is dedicated to two sectors, energy, so electricity that we produce and the houses and you know fuels our homes and our businesses, and then the food that we eat, so irrigated agriculture.
00:10:34.879 --> 00:10:41.519
Um, together those those constitute about 80% of water withdrawals and probably closer to 85 to 90% of water consumption.
00:10:41.759 --> 00:10:51.759
If you look at how much uh data center is of the total water usage across the United States, at a high end, the estimate is probably around 0.04%.
00:10:52.320 --> 00:10:54.000
That's total direct water usage.
00:10:54.159 --> 00:11:09.759
Now, if you consider the indirect water usage, so how much of that electricity, um, if you were to attribute that that water usage or that water consumption, associated electricity that data centers use, it increases maybe to 0.1, 0.2%, something like that, um, but still fairly small.
00:11:09.919 --> 00:11:18.559
And so, why then is there all this concern amongst the general public about data centers if it constitutes such a small percentage of overall water demand?
00:11:18.720 --> 00:11:26.559
Even though at a national level the impacts are pretty small and almost almost negligible, at a local level, they can be pronounced.
00:11:26.639 --> 00:11:29.600
And I think that's where the tension needs to be to be laid.
00:11:29.840 --> 00:11:35.120
Larger utilities don't face the same challenges that maybe smaller utilities would.
00:11:35.279 --> 00:11:54.960
When you know, talking even to like Prince William Water and Loudenwater, it's not a negligible percentage of their total demand, but it's in the case of like Prince William Water, I think their most recent statistics I saw, it was about 3.3%, 3.4% of their total average daily demand and about 10% of their peak daily demand.
00:11:55.120 --> 00:11:59.440
Still fairly small, even for one of the largest players in the industry.
00:11:59.600 --> 00:12:09.120
That's perhaps not as big a concern, but what I think might be a bigger concern is you're seeing now these hyperscale data centers come going more and more in rural areas.
00:12:09.200 --> 00:12:11.919
And when I talk about hyperscale, I'm mainly talking about AI data centers.
00:12:12.159 --> 00:12:18.399
A lot of the data centers have been located uh near major city hubs for things like fiber, access to labor.
00:12:18.559 --> 00:12:28.159
And probably one of the most important is latency because if you're streaming on Netflix, for instance, um, you want to be able to ping that data center and that information to come back to you quickly.
00:12:28.320 --> 00:12:32.960
If you have a data center way out in the middle of nowhere, that latency is going to slow things down a bit, right?
00:12:33.039 --> 00:12:37.279
And so they try to put data centers close to the end users for that purpose.
00:12:37.440 --> 00:12:42.480
Well, for AI data centers, it's a bit different because they're using these data centers to train their models.
00:12:42.639 --> 00:12:44.320
Latency isn't as important, right?
00:12:44.480 --> 00:12:54.240
And so they can put these out in more rural areas where land's cheaper, resources are cheaper, but it also typically means there's less people there, which implies that the utility is often going to be smaller.
00:12:54.320 --> 00:13:05.519
And so when you have a data center like the one not too far from me in Botatak County, right outside of Roanoke, this data center is
going to be requesting up to 8 million gallons per day peat demand, right?
00:13:05.679 --> 00:13:10.080
So for the utility that's that's serving them, it's actually based out of Roanoke, so they can handle it.
00:13:10.240 --> 00:13:16.960
But if it was a small country town like the one I grew up in, that would be more than their total demand, right, across their entire population.
00:13:17.120 --> 00:13:36.159
So I think it that nuance can matter, like where these things are located, both from a uh you know, infrastructure capacity and institutional capacity, and then also things like climate that John touched on earlier, but the ability to use free cooling versus uh evaporative cooling, um, that's going to have a big impact on how much water is being consumed.
00:13:36.559 --> 00:13:38.159
This is a good place to bring in Dr.
00:13:38.320 --> 00:13:42.720
Gravat with the Water Research Foundation looking at these issues nationally.
00:13:42.799 --> 00:13:48.559
Can you talk about your organization's work and some trends that you're seeing for data centers across North America?
00:13:49.039 --> 00:13:49.759
Sure, Donna.
00:13:49.919 --> 00:13:50.320
Thank you.
00:13:50.399 --> 00:13:53.360
I'm so excited that you're putting together this forum.
00:13:53.519 --> 00:14:02.720
And for the Water Research Foundation, uh, some of your listeners probably know we're the largest, the world's largest water research collaborative.
00:14:02.879 --> 00:14:05.519
We're supported primarily by water utilities.
00:14:05.600 --> 00:14:13.919
And here we've got Northeast Ohio Regional Sewer District and Columbus Water and Power, both of whom are subscribers to the Water Research Foundation.
00:14:14.159 --> 00:14:22.879
We look at just a plethora of different issues, but data centers are one that is emerging in the conversation all across the country.
00:14:23.120 --> 00:14:28.879
In fact, all around the world, as I've been in other parts of the world, and a lot of people are talking about data centers.
00:14:29.120 --> 00:14:39.200
It's kind of a curious situation that it is so such a localized issue in many ways, and it plays out quite differently in some communities versus others.
00:14:39.360 --> 00:14:54.320
I think the way I think about this as well is that there's a tremendous amount of uncertainty around not only where these data centers are going to be sited, top on their list is not necessarily, well, where's water going to be available for cooling?
00:14:54.480 --> 00:15:00.639
You know, they're looking for space, they're looking for a good economic deal, they might be looking for tax incentives.
00:15:00.799 --> 00:15:10.240
And I think there's a lot of uncertainty as well in terms of how the dynamics around water use for cooling are going to change over time.
00:15:10.639 --> 00:15:18.159
We just finalized a study at the Water Research Foundation focused on the residential end uses of water.
00:15:18.399 --> 00:15:29.279
And that study found that over 30 years for residential use, water use has decreased by 44% in the United States over 30 years.
00:15:29.519 --> 00:15:41.840
And that is primarily as a result of technology and changes in standards, low flow toilets or low, low uh water use uh washing machines or dishwashers or shower heads.
00:15:41.919 --> 00:15:46.240
And you might wonder, well, why is he talking about what we're supposed to be talking about data centers?
00:15:46.399 --> 00:15:54.720
The reason I bring this up, those decreases in water use at the residential level were not driven by changes in behavior.
00:15:54.879 --> 00:15:57.519
They were driven by changes in technology.
00:15:57.759 --> 00:16:07.039
And I think we have every reason to assume that technologies for cooling or the ability for chips to operate at a higher temperature, all that may change.
00:16:07.120 --> 00:16:17.279
And so if you're a water utility that's supporting the siding of a data center that's using a particular paradigm now for cooling, that could well change.
00:16:17.440 --> 00:16:27.200
And if you're building infrastructure to support those needs, and then those needs change, you could have stranded infrastructure and saying, well, what do we do now?
00:16:27.440 --> 00:16:32.879
And then I think it's also interesting to see where data centers are choosing to cite.
00:16:33.039 --> 00:16:48.080
You know, and one of the things that's always been striking to me is when we look at a map of the most water-stressed regions of the United States, those are many of the same regions that are experiencing the most rapid growth of things like data centers.
00:16:48.240 --> 00:17:04.720
So if you think about a place like Las Vegas or a place like Phoenix or various places in Texas, even some of the issues that we've been talking about, like, say, the availability of water, there's not so much water right now in Las Vegas, right?
00:17:04.880 --> 00:17:12.480
Those aren't necessarily the primary drivers that are leading data centers to say, where do we want to set up shop?
00:17:12.720 --> 00:17:27.279
Columbus Water and Power, you're dealing with both of those issues around energy availability for data centers and water and the trade-off between energy use and water use, depending on what strategy for cooling you decide to deploy at a data center.
00:17:27.359 --> 00:17:30.160
So that's another really interesting set of issues.
00:17:30.720 --> 00:17:32.880
Yeah, I can weigh in on that real quick.
00:17:33.039 --> 00:17:36.799
We have three of the major hyperscalers here in Columbus area.
00:17:37.039 --> 00:17:39.359
They all have different cooling strategies.
00:17:39.519 --> 00:17:56.480
And one of the things that we're looking at, again, I referenced it earlier, is a recycled water, reuse water system, uh, get them off the potable, which then could have potentially impact their their equation when they look at the power versus the water use, which is their you know their their economic driver.
00:17:56.559 --> 00:18:00.160
And they also have a lot of sustainability drivers, most of these large companies do.
00:18:00.319 --> 00:18:06.559
So kind of balancing those out, having those conversations with them, seeing if recycled water could be a thing.
00:18:06.640 --> 00:18:11.440
I will say it is for at least two of the major hyperscalers we have here.
00:18:11.519 --> 00:18:13.279
So conversations are ongoing.
00:18:13.440 --> 00:18:21.200
But yeah, that is definitely a dynamic that I know the data centers themselves look at, and that's part of our conversation on the water side in particular.
00:18:21.279 --> 00:18:24.720
Um, how can we be part of that process?
00:18:25.519 --> 00:18:35.759
Sarah, can you walk us through how water is actually used in a data center, what happens to it, and where does it go after it has been processed through the facility?
00:18:36.319 --> 00:18:43.759
I am not an expert, and as we talked about, there are many different ways that data centers use water for cooling.
00:18:43.920 --> 00:18:59.759
But the typical general model is that water comes in and it either comes in from, you know, uh a source water, you know, surface water, or it comes from a water utility, and then it is used for cooling.
00:18:59.920 --> 00:19:05.200
Sometimes it's a closed loop, sometimes it's an open loop, but a lot of times it's an open loop where there's evaporation.
00:19:05.359 --> 00:19:11.519
Like we talked about, we're seeing a ton of water at a facility level for a hyperscaler coming in.
00:19:11.759 --> 00:19:21.519
And as it goes through that loop, it's goes through many, many times, and the water is a lot of it's evaporated, and so it's becoming more concentrated.
00:19:21.759 --> 00:19:28.720
It also there's some additional chemicals that are being added to it to make sure that the pipes you know aren't corroded.
00:19:28.799 --> 00:19:39.839
Um, and so what ends up coming out, and that would either go to surface water discharge or to a utility, is a very concentrated water.
00:19:39.920 --> 00:19:54.880
So you think about what comes in, and there are um, you know, there's always chemicals and um, you know, additional things in water, even you know, the cleanest, safest water, but all of those um are concentrated and then those additives are concentrated.
00:19:55.359 --> 00:20:01.200
I mean, so that is what comes hopefully to a water reclamation plant.
00:20:01.359 --> 00:20:08.319
Um, one of the things that Ohio was working on earlier this year was a general permit for data centers.
00:20:08.720 --> 00:20:17.039
Right now, it seems like the future is uncertain for the general permit, which I think you know our utility is happy about.
00:20:17.279 --> 00:20:38.799
We would really like to see um regulation that requires individual discharge permits, which would mean that if a data center is going to discharge into a surface water, into a stream or a river, that they would have individual limits that would reflect the water that it's going into.
00:20:39.200 --> 00:20:43.279
You can imagine that this water could also be really warm.
00:20:43.440 --> 00:20:55.359
And so if you are putting a high, you know, a large amount of concentrated water into a stream, you know, you can really change and affect that stream water quality.
00:20:55.680 --> 00:21:02.480
Um, and so, you know, where there's no system to hook up to, we'd like to see individual permits.
00:21:02.640 --> 00:21:10.640
But really, um, what we're advocating for is that this water comes to um water reclamation plants where available.
00:21:10.880 --> 00:21:20.880
We are very invested in making sure that the water in our river that we discharge to stays at a high water quality.
00:21:21.119 --> 00:21:27.759
And that's why we, you know, would want to see the data centers in our area come and we would receive their water.
00:21:27.920 --> 00:21:33.519
We would um treat anything that's organic in the water, we would monitor it, you know.
00:21:33.759 --> 00:21:43.680
After it goes through our system, we have so much water coming through our system, those concentrated contaminants would be diluted through our large flows.
00:21:43.759 --> 00:21:48.000
And it would also change, you know, alleviate that temperature issue.
00:21:48.400 --> 00:21:56.160
Um, and then we could we would just have a better ability to control what's happening in our area of central Ohio.
00:21:56.640 --> 00:21:59.759
Certainly the sewer district and I believe Columbus Water and Power made recommendations.
00:22:00.400 --> 00:22:14.319
To Ohio House and Senate Select Committee data centers recommending either connecting to a treatment works or obtaining an individual permit over a general permit because the general permit just doesn't have those requirements.
00:22:14.480 --> 00:22:15.119
Is that correct?
00:22:15.680 --> 00:22:31.200
No, a general permit means that anyone, any data center who's eligible can work under this general permit and discharge that concentrated water into the stream, you know, whatever's you know close by their plant.
00:22:31.680 --> 00:22:39.839
But the eligibility was really only tied to how much surface water the data center was taking in.
00:22:40.000 --> 00:22:45.759
So there was really no limit to how big a data center could operate under this general permit.
00:22:45.920 --> 00:22:49.759
Because if you got your water from a water utility, there was no limit.
00:22:49.920 --> 00:22:55.839
The general permit does not consider the stream or the river that you're putting that water back into.
00:22:56.000 --> 00:23:01.680
So it could be, you know, some meandering brook or it could be the Sayoto River.
00:23:01.839 --> 00:23:09.200
And so that's really why we want to make sure that those limits are tied to where that water is going.
00:23:09.440 --> 00:23:18.079
And you can imagine that in some streams, the discharge from a data center could make up most of the water that would be in the stream.
00:23:18.240 --> 00:23:31.920
And the other thing that the data center general permit, one of the limitations or the requirements was that you had to be 500 yards set back from our drinking water intake, from any drinking water intake.
00:23:32.079 --> 00:23:38.000
But there was no limit on clustering or how many data centers could be within that area.
00:23:38.240 --> 00:23:46.400
And so that made us very nervous on the drinking water side that maybe one data center might not affect our source water.
00:23:46.559 --> 00:23:54.160
But if you had a bunch clustered together, and then there's no limit to how many, there's no limit over time either.
00:23:54.400 --> 00:23:56.480
That made us nervous as well.
00:23:57.200 --> 00:24:04.960
And water utilities and water reclamation facilities are really in the business of understanding and protecting their watersheds.
00:24:05.039 --> 00:24:14.960
And so the utilities here on this call have spent billions, literally billions of dollars protecting watersheds from various sorts of discharges.
00:24:15.039 --> 00:24:30.880
And so you can understand why, if there's an approach to a permit that essentially treats all data centers as the same and gives them a lot of flexibility to decide how and where they want to manage discharges, it can be really troubling for the water utilities.
00:24:31.119 --> 00:24:47.359
Like, wait a minute, the public has invested so much, the customers have invested so much in improving water quality, and you just can't, you can't afford to think about the possibility of this somehow slipping as a result of not being really thoughtful in the approach to managing discharges.
00:24:47.440 --> 00:24:49.680
So I think it's a super important issue.
00:24:49.920 --> 00:24:51.359
It's so exciting to have Dr.
00:24:51.440 --> 00:24:58.000
Marston on this call because he's the principal investigator for one of the projects that our subscribers said they most need.
00:24:58.079 --> 00:25:07.839
And that's a project that's focused on understanding the economic considerations and the opportunities for water and wastewater utilities related to data centers.
00:25:08.000 --> 00:25:22.559
What kind of tools can we develop, not like a general tool, but what kind of guidance can we develop for water utilities to help them think through these really specific questions they have to ask, kinds of things that Sarah was just talking about.
00:25:22.880 --> 00:25:33.440
Another project that we have underway that is related to what I mentioned earlier, is a project that was funded specifically by the Southern Nevada Water Authority in Las Vegas.
00:25:33.599 --> 00:25:43.119
So Las Vegas passed an ordinance back in 2021 that put a moratorium on evaporative cooling in the city of Las Vegas.
00:25:43.440 --> 00:25:43.759
Okay.
00:25:44.720 --> 00:25:54.640
They're expected to have their data center footprint increased by tenfold over the next decade or more, but no evaporative cooling.
00:25:54.799 --> 00:25:56.319
So what are we to do?
00:25:56.480 --> 00:26:00.319
Looking at that question, well, what are the other strategies for cooling that are available?
00:26:00.480 --> 00:26:03.039
How might they be used in a city like Las Vegas?
00:26:03.119 --> 00:26:11.680
And then we're doing what we're calling a collaborative water and wastewater utility forum on data centers here in Washington, D.C.
00:26:11.920 --> 00:26:12.480
in August.
00:26:12.640 --> 00:26:22.559
It's going to bring together many dozens of utilities, kind of sharing best practices, sharing the questions they have with experts there in the room to say, well, here's how we did it.
00:26:22.960 --> 00:26:27.759
If I might add something to that, thanks, Peter, for mentioning our project at Virginia Tech.
00:26:28.240 --> 00:26:42.799
If there are utilities and you work with a utility and you want to have your voice heard, we're going to be linking the survey, the initial survey that we'll be conducting to understand your perspectives and to be able to try to meet those needs more effectively.
00:26:43.039 --> 00:26:46.880
So if you'd like to participate in that survey, please complete that by the end of July.
00:26:47.039 --> 00:26:48.960
We'd be happy to follow up with you.
00:26:49.039 --> 00:27:01.359
There's an opportunity for you to indicate that if you would like to be more involved in the project and to have your input not only through the survey, but perhaps through interviews be captured within our reporting, we'd be happy to do that.
00:27:01.759 --> 00:27:02.799
I'm just interested.
00:27:02.880 --> 00:27:07.119
I'm so fascinated by water reuse and purple pipe techniques.
00:27:07.200 --> 00:27:11.119
Um and I'm wondering what's that conversation look like on your end?
00:27:11.279 --> 00:27:15.039
Is there interest in that investment or is it does it seem more difficult?
00:27:15.119 --> 00:27:16.160
So there's pushback.
00:27:16.480 --> 00:27:26.559
Yeah, so unlike Cleveland, Lake Erie, and Cincinnati with the Ohio River, Columbus is in the middle of the state, and we do not have large bodies of water that we can draw water from.
00:27:26.799 --> 00:27:33.119
So uh every drop of water that we can within reason, we we we capture and treat and we hold it for storage.
00:27:33.200 --> 00:27:35.599
And so we have reservoirs, we use groundwater.
00:27:36.000 --> 00:27:40.480
That's one of the reasons we're looking at this, because we do not have a limitless supply of water.
00:27:40.799 --> 00:27:45.519
Ohio is a water-rich state, but we still have to manage our water a little bit different, Central Ohio.
00:27:45.759 --> 00:27:59.440
Probably defining what reuse is just for the general audience, reuse is taking effluent from a water reclamation facility, which in many cases and probably most cases is cleaner than the actual river that the water is going into.
00:27:59.759 --> 00:28:11.279
Uh, what we're evaluating at our Southerly facility is uh at a conservative number
is 40 million gallons per day of water that we can put back into a system, uh purple pipe system.
00:28:11.599 --> 00:28:13.039
Uh thanks collaboration.
00:28:13.119 --> 00:28:14.960
We're very early in this process.
00:28:15.039 --> 00:28:17.759
Um, as far as having the conversations, we know we can do it.
00:28:17.839 --> 00:28:19.200
It's been done in other states.
00:28:19.440 --> 00:28:21.839
We are well in the planning of that purple pipe network.
00:28:22.000 --> 00:28:23.680
It's gonna depend on our partners, though.
00:28:23.839 --> 00:28:27.200
We are not expecting our ratepayers to carry this load.
00:28:27.279 --> 00:28:35.920
Um, that again benefits can benefit us and can benefit the long-term sustainability of uh data centers and other manufacturers.
00:28:36.400 --> 00:28:37.519
That's really interesting.
00:28:37.680 --> 00:28:43.519
I think you know, the idea of uh doing partial or like almost complete treatment, right?
00:28:43.599 --> 00:28:56.960
And then sending that directly to the end user if they don't need potable water makes the most sense rather than putting it back into whatever your receiving water body is just to pull it back out and treat it and then send it again, right?
00:28:57.200 --> 00:29:00.640
So, like sustainability-wise, it just seems like a great opportunity there.
00:29:00.880 --> 00:29:02.799
And then Sarah, I just had another question for you.
00:29:02.960 --> 00:29:12.799
When it comes to, you know, being in a city, there's so many players when it comes to economics, when it comes to development, when it comes to your power and your water.
00:29:12.960 --> 00:29:18.240
How are you guys as a city like really having trying to have those conversations like in a united way?
00:29:18.640 --> 00:29:32.720
It's interesting in central Ohio, like our utility serves, you know, over 20 suburbs, but within our actual cities, we actually don't have the hyperscalers like within the city, because the city, you can imagine, is much more developed.
00:29:32.799 --> 00:29:40.240
Um, in central Ohio, our suburbs have a lot of farmland that is ripe for that development that John is talking about.
00:29:40.720 --> 00:29:45.039
The major siting conversations are not happening with us, those are happening with the suburbs.
00:29:45.279 --> 00:29:55.839
When we talk about like NDAs and these this idea of data transparency, those siting and development conversations are happening outside of us and we don't have that information.
00:29:56.079 --> 00:30:06.079
And so sometimes people look to us, you know, as the utilities, it's like, well, you know, it's you guys, but really um it's a much broader, a broad, much broader conversation.
00:30:06.240 --> 00:30:10.640
The other thing that is happening is that the public is interested.
00:30:10.720 --> 00:30:21.519
Um, and so we're seeing it across Central Ohio, across Ohio, is becoming a topic of conversation in at council meetings or commissioner meetings, you know, things like that.
00:30:21.759 --> 00:30:28.319
And so, you know, of course, that is of interest to the city, um, even though we don't have the hyperscalers within the city.
00:30:28.400 --> 00:30:49.440
So I guess it's continuing to educate the public, keeping them informed with the way that we are approaching this and that we have our ratepayers' best interests in the forefront, our residential ratepayers, but that we also, you know, we are meeting the growth of Central Ohio, not just the growth of data centers.
00:30:49.759 --> 00:30:54.799
Some of our local neighbors, we are seeing bans.
00:30:54.960 --> 00:30:59.440
You know, some of the suburbs are um instituting bans on new data centers.
00:30:59.519 --> 00:31:06.799
And so we're seeing this like at, you know, a city level, we're seeing it at a regional level, we're hearing about it at the state.
00:31:06.960 --> 00:31:08.720
The legislature is really interested in it.
00:31:08.799 --> 00:31:10.880
It's a bipartisan issue.
00:31:11.039 --> 00:31:12.720
Everyone is interested in it.
00:31:12.799 --> 00:31:15.680
Um, and then we're seeing it at the federal level as well.
00:31:16.559 --> 00:31:21.279
Sarah, I was just gonna say it was so interesting what you were talking about in terms of the local dynamics.
00:31:21.359 --> 00:31:21.920
And Dr.
00:31:22.079 --> 00:31:30.240
Marston made reference to northern Virginia, you know, that's like the center of so much of data center development.
00:31:30.319 --> 00:31:34.160
You know, they got 200 data centers and counting in Loudoun County.
00:31:34.400 --> 00:31:56.720
And when we talk about the dynamics and the various moving parts and pieces that drive decisions for where to site, I know Loudoun County reported that back in 2025, they estimated that the tax revenue from data centers was going to be equivalent to 95% of the county's operating budget.
00:31:57.039 --> 00:32:10.559
And when you think about those numbers and what that means actually for taxes for residents, which in some cases are going down because the data centers are coming in, it's like so dynamic.
00:32:10.720 --> 00:32:15.039
All the different moving parts and pieces that are driving how this is playing out.
00:32:15.279 --> 00:32:22.960
And the water utilities are kind of there in the middle of all this, watching all this going on and saying, well, how do we make sure we can meet the need and support the community?
00:32:23.039 --> 00:32:24.400
It's a really interesting time.
00:32:25.680 --> 00:32:26.000
Dr.
00:32:26.160 --> 00:32:27.599
Marston, what's your take on that?
00:32:27.759 --> 00:32:29.839
Just interested, since you're like a local.
00:32:30.319 --> 00:32:36.559
Yeah, so we've had some media coverage with our research related to data centers at Virginia Tech.
00:32:36.799 --> 00:32:41.440
And um I think because of that, it's in the capturing of public attention more broadly.
00:32:41.599 --> 00:32:45.920
I receive a lot of phone calls from concerned citizens in the state
of Virginia.
00:32:46.240 --> 00:32:49.599
People are very concerned with data centers coming into their backyard.
00:32:49.680 --> 00:32:57.599
And so some of that's related to water, some of it's more broadly related to concerns just about AI, some of it is related to noise concerns.
00:32:57.839 --> 00:33:06.720
I think data centers, for whatever reason, have captured the attention of how we use water in ways that other sectors, like agriculture, for instance, quite haven't.
00:33:06.799 --> 00:33:12.960
Uh Sarah earlier talked about this idea of lack of transparency within the industry, data center industry.
00:33:13.039 --> 00:33:15.839
I'm seeing a shift in that over just over the last several months.
00:33:15.920 --> 00:33:22.240
I'm thinking across social media and kind of media platforms and seeing what these large tech companies are starting to put out before.
00:33:22.400 --> 00:33:26.799
Everything was mired in non-disclosure agreements, um, nothing was made public.
00:33:26.880 --> 00:33:36.960
Um, but now you're starting to see like AWS, you're starting to see Google report their facility, in the case of Google, report their facility level water usage for their data centers.
00:33:37.119 --> 00:33:38.160
I think that's a path forward.
00:33:38.240 --> 00:33:41.359
I've made this case directly to tech companies before.
00:33:41.519 --> 00:33:47.519
I think they have done more harm than good by keeping the this information secret.
00:33:47.599 --> 00:33:52.079
And part of this is because of trade secrets and because of, you know, it's a very competitive space right now.
00:33:52.160 --> 00:33:58.000
And so anything they think that might allow their competitors to have information on their operations, they're they're hesitant to release.
00:33:58.160 --> 00:34:10.159
But the counter to that is you're seeing, you know, these data centers, some of these projects being delayed or potentially canceled in many regions across the country because of public pushback.
00:34:10.239 --> 00:34:18.000
And because and a lot of this is centered, even though energy is a much bigger, in my opinion, much bigger threat than water when it comes to resource utilization.
00:34:18.159 --> 00:34:19.679
People understand water more, right?
00:34:19.760 --> 00:34:24.159
This is this is very fundamental to life, and and people know what a gallon of water is.
00:34:24.239 --> 00:34:26.559
They might not know what a megawatt of energy is.
00:34:26.719 --> 00:34:30.480
And so that visceral reaction is more pronounced when we talk about water.
00:34:30.559 --> 00:34:37.599
And that and I think that's where you're seeing this on the front lines, is people coming to town halls and other types of venues to protest data centers.
00:34:37.679 --> 00:34:40.639
And a lot of that protest is because of water.
00:34:40.719 --> 00:34:43.119
And I think these tech companies start to realize that.
00:34:43.199 --> 00:34:48.079
And the name of the game, especially when we're talking about AI data centers, is speed to operation, right?
00:34:48.320 --> 00:34:51.280
How quickly can they get the data center up and going and operational?
00:34:51.760 --> 00:34:58.559
And I think they are having a turning point where before it was this community engagement was going to be something that would slow them down.
00:34:58.639 --> 00:35:10.800
But I think now they're starting to realize that if they don't do the community engagement, they don't make sure that they're they're taking care of these communities and ensuring they're investing in infrastructure for the long term, they're going to slow themselves down even more, right?
00:35:10.880 --> 00:35:12.960
And that's what you're seeing, I think, happen in real time.
00:35:13.039 --> 00:35:33.360
And so I'm hopeful for the future, looking out you know, the next couple years, is that we'll see these tech companies begin to work with utilities and try to figure out solutions as opposed to kind of imposing their will on these communities and ensuring that they get the resources they need and you know, but not make care of some underlying needs as a community as well.
00:35:33.440 --> 00:35:34.960
And so that's what I'm hopeful for.
00:35:35.119 --> 00:35:36.159
I think you'll see some of that.
00:35:36.239 --> 00:35:41.039
I'm also hopeful for as I think Peter mentioned earlier, this uh technological advancement.
00:35:41.119 --> 00:35:46.000
He noted how this has occurred within the realm of residential water use.
00:35:46.079 --> 00:35:50.960
Um, I think the same thing might happen as well with the the data center industry.
00:35:51.119 --> 00:36:01.280
In fact, just within the last 10 days, there's an announcement by NVIDIA that they are producing new chips that could run uh much hotter temperatures, which means less potential cooling, right?
00:36:01.360 --> 00:36:04.000
And so this could fundamentally shift how water is being used.
00:36:04.079 --> 00:36:06.239
And I think there's a direct parallel.
00:36:06.559 --> 00:36:11.360
You go back 15, 20 years ago, the data center industry was using a ton of energy.
00:36:11.440 --> 00:36:15.599
They're still using a lot of energy, mainly because of their growth, but that energy usage was very inefficient.
00:36:15.760 --> 00:36:28.320
And so there were a lot of projections uh around 2010, 2012 that data centers um were going to be consuming 20 to 30 percent of global energy, of global energy demand, right?
00:36:28.559 --> 00:36:31.039
Um, and that number seemed outrageous.
00:36:31.119 --> 00:36:41.039
You kind of look into a little bit more, and what was happening is people were projecting out uh the similar levels of efficiency going forward with this uh increase in the number of data centers.
00:36:41.119 --> 00:36:51.199
But what actually happened is you would we saw between 2010 and 2018 a 550% increase in compute, but only a 6% increase in electricity demand.
00:36:51.360 --> 00:36:56.000
And that's because we are realized
these economies of scale, there is a technological advancement.
00:36:56.320 --> 00:37:02.960
And so I think if if we bring smart people together, people are willing to work towards solutions to reduce water demand.
00:37:03.039 --> 00:37:04.639
I think there's a path forward there.
00:37:04.800 --> 00:37:13.280
But it takes it takes a collective effort, um, and and hopefully um we've had some some great utilities on this panel that are part of that solution.
00:37:14.079 --> 00:37:18.960
Everybody's involved in this conversation, and we're seeing it at the Ohio State House.
00:37:19.039 --> 00:37:29.599
Um HB646 just barely didn't pass um right before break, and it could pass later this year, but it really is.
00:37:29.679 --> 00:37:39.760
It's a bipartisan bill that really moves forward a lot of these issues that the public, that utilities, that everyone has been interested in.
00:37:40.000 --> 00:37:51.280
It would require reporting of energy, reporting of water usage, um, requiring best practices for water conservation and efficiency.
00:37:51.679 --> 00:37:56.239
Um it would require wastewater monitoring, wastewater quality monitoring.
00:37:56.320 --> 00:38:04.880
It would require Ohio EPA and the Ohio Natural UmDNR to both develop regulations.
00:38:05.280 --> 00:38:09.119
I really was really excited about this to see everyone come together.
00:38:09.280 --> 00:38:12.000
You know, we don't see that very often in Ohio.
00:38:12.079 --> 00:38:15.440
You know, we don't see it very often on a federal level either.
00:38:15.599 --> 00:38:21.039
But it's great to see that everyone is interested in saying, let's demystify this.
00:38:21.199 --> 00:38:28.559
You know, we need the data centers, but let's do it in a responsible way where we're creating the regulation so that we can be successful.
00:38:29.039 --> 00:38:31.039
I think data centers are going to continue to grow.
00:38:31.199 --> 00:38:34.159
I think they're gonna continue to grow at a remarkable pace.
00:38:34.239 --> 00:38:38.400
And there's gonna be a lot of challenges for us to address going forward.
00:38:38.559 --> 00:38:40.079
But I'm pretty hopeful about this.
00:38:40.239 --> 00:38:46.400
I think we're starting to see things open up in a really positive way, and hopefully we can you know build on that going forward.
00:38:47.599 --> 00:38:48.719
I'll jump in next here.
00:38:48.800 --> 00:39:06.719
So where I see it, I think Columbus um within the next five, 10 years, we are gonna have a purple pipe network uh for uh a prime location for data centers in particular on a water side to go, uh, which again should help with their calculations on the water versus power that they that they do to make those decisions.
00:39:07.199 --> 00:39:10.480
Um so I think we'll see that in Columbus.
00:39:10.639 --> 00:39:24.400
And I I guess for the the general the AI for general public and just data centers in general, I do think there's a little merit to the national security perspective of this, having these located in the US versus other countries.
00:39:24.559 --> 00:39:31.519
So I think that's an important piece that uh I think sometimes folks don't uh they overlook maybe.
00:39:31.920 --> 00:39:42.719
I think one thing that um Peter, you talked or you talked about was the tax incentives and how um Virginia was where it was really making such a positive difference.
00:39:42.960 --> 00:39:44.559
The story's different in Ohio.
00:39:44.719 --> 00:39:55.760
You know, just a month ago, the story broke that um Ohio has missed over a billion dollars in tax revenue because of tax incentives.
00:39:56.000 --> 00:40:02.480
It seemed like in the articles that people didn't even know this, um, that, you know, this was unrealized.
00:40:02.719 --> 00:40:10.800
Um, right now there's a um temporary ban on future tax incentives, um, which um I'm not sure what will happen with that.
00:40:10.960 --> 00:40:14.639
But what I'm hoping for is that that becomes part of that conversation.
00:40:14.719 --> 00:40:29.360
And we say, no, you know, these are billion-dollar companies, our community should be receiving this money, um, and we shouldn't be continuing this very, very heavy tax incentives that have been um historical in Ohio.
00:40:29.519 --> 00:40:34.800
And that also helps, you know, this discussion of whether the public accepts the data centers.
00:40:34.960 --> 00:40:54.880
And then what I would want people to know is that utilities like Columbus Water and Power, we are very focused and very intentional on making sure that our residential ratepayers can't have uh affordable bills, that they can afford the water that they need.
00:40:55.039 --> 00:40:57.519
Our drinking water is an essential service.
00:40:57.679 --> 00:41:05.599
And so as we have this discussion, as Sean mentioned, you know, we're asking these tech companies to come to the table.
00:41:05.760 --> 00:41:14.480
They are benefiting from being in central Ohio, they're benefiting from being in Ohio, and we want to make sure that they're paying their fair share.
00:41:14.639 --> 00:41:37.519
And that really gets to that question of long-term environmental protection, long-term water stability is making sure that everyone's paying their fair share and that we are protecting as we go, that we're not waiting, you know, 10 years from now and looking back and saying, oh, wait a minute, I guess we shouldn't have had, you know, six data centers on one part of a stream.
00:41:37.920 --> 00:41:46.559
I think there's also, you know, going to be uh chances for water and wastewater utilities to start to learn from each other.
00:41:46.639 --> 00:41:48.880
And so a lot of this has happened kind of ad hoc.
00:41:49.039 --> 00:41:53.440
Uh, each utility kind of feeling like they're on their own when it comes to finding solutions.
00:41:53.679 --> 00:42:27.119
I'm hopeful through uh our research with the Water Research Foundation, as well as other folks that are doing research in this space, we can help uh utilities come together to share information, to build out best practices so that they can serve these data centers without causing harm or threat to their existing customer base, and also have the tools that they need to make informed decisions with respect to infrastructure build out, with respect to financing, uh and contracts, and probably with the most level uncertainty, demand forecasting, right?
00:42:27.199 --> 00:42:34.000
So there's that we talked about this earlier in the podcast about how much uncertainty there is and how this is going to play out in the next few years.
00:42:34.159 --> 00:42:41.039
Hopefully, by sharing information, learning from each other, we can get better project projections, which ultimately means better service.
00:42:43.440 --> 00:42:45.199
Thank you everybody for those reflections.
00:42:45.280 --> 00:42:45.840
That was great.
00:42:46.079 --> 00:42:47.920
Mike, I'm gonna swing it around to you.
00:42:48.079 --> 00:42:48.960
What are you thinking?
00:42:49.360 --> 00:42:50.719
It's been a whirlwind.
00:42:53.519 --> 00:42:54.559
Uh that I didn't know.
00:42:54.639 --> 00:43:07.679
I think there has been a certain level of anxiety and uncertainty, uh, as well as just wondering about the economic prospects as we see more data centers here in Northeast Ohio, and this has really shed some light on it.
00:43:07.920 --> 00:43:18.719
You know, I'm curious how things move forward with increased transparency, requirements for accountability on these huge investments and use of public resources.
00:43:19.039 --> 00:43:22.480
So I appreciate all the the input that we've gotten from our guests.
00:43:23.360 --> 00:43:35.440
I think for me, there's like a lot spinning around in my mind about like where capitalism meets government function, meets science, and how are all of these things going to kind of coalesce at the end of the day?
00:43:35.599 --> 00:43:37.360
Where is regulation gonna come into play?
00:43:37.519 --> 00:43:41.039
Where is technology gonna be changing all of those factors?
00:43:41.280 --> 00:43:52.159
And how do we react as a public and as a community to like make sure that we are protecting our clean water resources, that we're protecting our energy sources to the best that we can?
00:43:52.320 --> 00:44:00.320
So I think my takeaway is intersectionality, like everything is so closely knit together here when it comes to data centers and their impacts.
00:44:00.800 --> 00:44:04.159
Thank you everyone so much for joining us today.
00:44:04.400 --> 00:44:09.440
We're happy to have you on Clean Waterworks, our podcast here at the Zero District about clean water.