Data center debate needs more transparency
As data centers continue to come online across North Texas—and resident opposition grows—the Fort Worth Report invited Mohammed Islam, an assistant professor of computer science and engineering at The University of Texas at Arlington, to join their Informed Discussion series for a panel discussion on the industry’s impact on Fort Worth and Tarrant County.
During the panel, Dr. Islam provided his perspective on misinformation about data centers, how closed-loop cooling technology works and where he believes the industry still falls short on transparency.
Q: What's the biggest piece of misinformation you've heard in the data center debate?
Islam: There's been misinformation on both sides. We shouldn't say there's nothing to worry about; there is real growth in water and energy consumption, though it's sometimes exaggerated on the negative side. But I've also seen misinformation from the tech industry. For example, there's a lot of talk about new data centers that use very little water, but they don't tell you that water use is concentrated in just one or two months out of the year. So if you spread it across the year, it looks like the center uses 10 times less water than an older facility, but during the summer months it's putting the same amount of pressure on the system. Companies are also asking for infrastructure sized for very large gallons-per-day allowances, which puts stress on water infrastructure. So the numbers are being cherry-picked on that side, while they're exaggerated on the other. Saying data centers alone are going to create a drought isn't true either. What I really want to see is more transparency from the data center industry about exactly what's happening.
Q: How does closed-loop cooling technology work, and how much water does it actually save?:
Islam: Every megawatt or gigawatt of energy a data center uses eventually becomes heat that has to be removed, and water is one of the best conductors for carrying that heat away. In a closed-loop system, water carries heat from the servers inside the data center out to a cooling facility, gets cooled down, and comes back in, running in a loop, which is why it's called closed loop. Ideally, you'd get to zero water use with fully dry cooling, but even closed-loop systems have some water footprint because you can't run the same water forever; it has to be replaced periodically. Dry cooling is close to zero water use; closed loop is almost zero. The downside is you can't run these systems year-round in a place like Texas. Older data centers historically used evaporative cooling, which is actually more efficient at cooling than dry systems, but it consumes a lot of water. Moving to closed-loop systems trades some water savings for a slightly larger energy footprint. These systems are still fairly new; newer data centers are increasingly using them, but not all facilities have made the switch. There's also research happening right now, including at UTA, on getting heat away from the chips more efficiently, so I'm hopeful this area keeps improving.
Q: Is it expensive for existing data centers to adopt this technology?
Islam: Yes, retrofitting an existing data center is expensive because you may have to shut it down or replace major systems. My own opinion is that the money data centers spend on infrastructure, relative to the revenue they generate, isn't comparable to a lot of other water-using industries. Data centers can afford to spend more on this than most other industries can. If you asked a lower-margin water-using industry to go to zero-water or dry cooling, they'd likely go out of business; data centers won't. I don't have an exact percentage on the water savings, but you're looking at something like 10 to 20 times less water use than a traditional facility, at minimum.
Related: Nalley: ‘We delivered southern hospitality’
Q: Are there things about data centers people should justifiably be concerned about?
Islam: I do think there's a lot of information out there, but when academics like us try to dig in, we often can't find what we need. I don't want transparency dictated by the industry itself; I want it dictated by policymakers. Companies tend to show us their best examples, but the overall numbers don't always match. We see very advanced technology in some data centers, yet Google's own water consumption numbers have grown a lot in recent years, so they're showing us the poster children while a lot of other facilities are still doing things the old way. My main concern is around the nuances. Take the moratorium on new grid connections, that's actually sending an odd message, because now some developers may think that if a data center isn't connected to the grid, they can do whatever they want: use worse water technology, burn a lot of fossil fuel, since they're not technically breaking any grid-connection rule. We shouldn't have policy that only looks at one side of the problem.
Q: How do we objectively determine what makes a "good" operator, and how should that factor into city or state regulation?
Islam: I wouldn't say any operator in the queue is simply "good" or "bad," some may just be better than others, and I want more transparency from all of them than we currently get. One thing worth noting: data centers can actually become an asset to the power grid, because they're large, highly controllable loads that can adjust demand quickly. Right now they're not being asked to use that capability for the grid's benefit, because it doesn't make financial sense for them to slow down their own operations to help. But I think we should be looking at this from the community's perspective first, what protections do we need if a large amount of new demand is coming in, rather than deciding upfront which companies are "better" and just letting them build. For example, right now some data centers are being encouraged to build their own gas power plants next to their facilities because of the moratorium, and that's not a good outcome; burning gas means more carbon emissions, and that's a decision with consequences lasting far longer than the life of the project. If new power generation is going in next to a data center, I'd rather see renewables like solar, not just more gas.
Q: What can concerned citizens do to better understand this issue?
Islam: A lot of the information out there, from the companies building these data centers, comes with an agenda behind it; it's not necessarily the full picture. People should look at a range of sources and also talk to groups without a stake in the outcome, like local newsrooms. I think there's a role for academics here, too. We should be working to create more trustworthy, nonpartisan sources of information on this.
About The University of Texas at Arlington (UTA)
The University of Texas at Arlington is a growing public research university in the heart of Dallas-Fort Worth. With a student body of over 42,700, UTA is the second-largest institution in the University of Texas System, offering more than 180 undergraduate and graduate degree programs. Recognized as a Carnegie R-1 university, UTA stands among the nation’s top 5% of institutions for research activity. UTA and its 300,000 alumni generate an annual economic impact of $28.8 billion for the state. The University has received the Innovation and Economic Prosperity designation from the Association of Public and Land Grant Universities and has earned recognition for its focus on student access and success, considered key drivers to economic growth and social progress for North Texas and beyond.