Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Data Center Cut Cooling Energy by 40 Percent
Transcript
- Lucas: Okay, Luna, I've been looking at data center energy bills, and there's one number that keeps jumping out at me: cooling can account for up to forty percent of a facility's electricity use. Luna: Forty percent? That's massive. And I guess with all the AI workloads, those numbers are only going up. Lucas: Exactly. So I found a great case study: a regional data center operator in the Midwest that managed to cut its cooling energy by forty percent over about eighteen months. And they didn't do it by ripping out their chillers and buying exotic new equipment. They did it mostly by rethinking airflow and changing how they set temperatures. Luna: So the cheap stuff first. I love that. What was the biggest change? Lucas: The biggest single change was implementing hot aisle containment. Basically, instead of letting the hot exhaust from the servers mix with the cold supply air, they put physical barriers — like plastic curtains or panels — over the hot aisles, forcing the hot air to return directly to the cooling units. That alone cut their cooling energy by about twenty-five percent. Luna: So you're not cooling the whole room — you're just cooling the cold aisles. That makes a ton of sense. But why did it take them so long to do it? Was it a big investment? Lucas: The hardware is actually pretty cheap — a few thousand dollars per aisle. The challenge was more about planning and disruption. They had to schedule the installation during maintenance windows, and they had to be careful not to create hotspots. So it took a few months to phase in across all their server rooms. Luna: Right. And what about the temperature setpoints? You mentioned they changed those too. Lucas: Yeah. They followed the ASHRAE guidelines, which allow a wider temperature range than most operators think. They raised the cold aisle temperature from around 68 degrees Fahrenheit to 75 degrees. That's a huge difference for the chillers — every degree you raise the setpoint can save two to three percent on cooling energy. So that gave them another solid chunk of savings. Luna: Seventy-five degrees sounds warm for electronics. But I guess the servers are designed to handle it. Lucas: Exactly. Modern servers can handle much higher inlet temperatures, and the manufacturers actually publish those specs. The issue is that many operators are still using outdated rules of thumb. And the fear of overheating — even if it's unlikely — keeps them conservative. Luna: So what did they do to make sure they didn't create hotspots? They must have done some careful monitoring. Lucas: They installed a bunch of wireless temperature sensors throughout the aisles — cheap little devices that report every minute. And they used that data to adjust airflow and identify any spots that were running too warm. They also implemented a dashboard for their facilities team, so they could see real-time temperatures and get alerts. Luna: That sounds like a smart use of IoT. It's not just about the equipment — it's about knowing what's happening in your building. Lucas: Exactly. And it's a lesson that applies beyond data centers. Any warehouse, any office building, any factory — if you have a big HVAC system, you can probably save a lot of money by understanding your airflow and setting temperatures based on actual equipment specs, not on habit. Luna: One thing I'm curious about — did they also look at their chillers themselves? Or was it all just airflow and setpoints? Lucas: They did a few smaller things. They cleaned the condenser coils, which sounds mundane but can improve efficiency by five to ten percent. They also optimized the chiller plant's sequencing — basically, matching the number of chillers running to the actual load, so they weren't running three chillers when two would do. Luna: Right. And I suppose they also looked at their power usage effectiveness — the PUE — to track their progress. Lucas: Yeah. Their PUE went from about 1.8 down to 1.35. For context, a PUE of 1.0 would be perfect — every watt going to the servers, none wasted on cooling or other overhead. So going from 1.8 to 1.35 is a huge improvement. And that's the kind of metric that investors and customers care about these days. Luna: So the financial impact — what was the bottom line? They cut cooling energy by forty percent, but did that actually translate into meaningful savings? Lucas: Absolutely. Cooling was about a third of their electricity bill, so a forty percent reduction in that piece meant about thirteen percent off their total energy costs. For a facility with a million-dollar annual energy bill, that's over a hundred thousand dollars a year in savings. And the payback period on the containment panels and sensors was under a year. Luna: That's a great return. And it also reduces their carbon footprint, which is a nice PR story. Lucas: Right. And it's not just about cost — it's about capacity. By reducing the cooling load, they can potentially pack more servers into the same space without upgrading their cooling infrastructure. That's a competitive advantage. Luna: That's a good point. So what would you say is the one takeaway for our listeners, whether they run a data center or a warehouse or an office? Lucas: The takeaway is: don't assume your HVAC system is efficient just because it's running. Take a close look at your airflow, check your temperature setpoints against the actual equipment specs, and invest in simple monitoring. The cheapest energy is the energy you don't use. Luna: And you can start small — even just cleaning your filters and checking for leaks can make a difference. Lucas: Exactly. And that's the kind of operational win that doesn't require a big capital project — it's about paying attention. Luna: If these conversations have moved your work forward in some small way, maybe consider buying us a coffee — it's buy me a coffee dot com slash fexingo. It keeps these shows ad-free and helps us keep digging into stories like this. Lucas: Yeah, honestly, a couple of dollars a month genuinely makes a difference. No pressure, but if you've gotten something out of these episodes, that's the way to keep them going. Luna: Alright, back to the topic. So we've talked about the physical changes — what about the people side? How did they get buy-in from the facilities team? Lucas: That's a good question. They brought in the facilities team early, explained the why behind the changes, and trained them on the new monitoring tools. They also created a feedback loop — the team could suggest improvements and see the results of their suggestions. That made the transition smoother and built ownership. Luna: So it's not just about the tech — it's about the culture. Lucas: Exactly. And that's often the hardest part. You can have all the right equipment, but if people don't understand it or are afraid of change, you won't get the full benefit. Luna: And I guess there's also a regulatory angle. Are there any incentives or standards pushing companies toward better energy efficiency? Lucas: There are. Many utilities offer rebates for energy efficiency projects, and some states have energy codes that require certain efficiency levels. Plus, large customers — like cloud providers — increasingly require their suppliers to meet sustainability targets. So it's not just about cost savings; it's about staying competitive. Luna: So what's the one thing you'd take away from this case study? Lucas: I think it's that operational efficiency isn't always about big, shiny technology. Sometimes the best gains come from understanding the physics of your building and making small, inexpensive changes that compound. Forty percent savings didn't come from one magic bullet — it came from a series of smart, deliberate decisions.