One If By Hyperscale, 2 If By Colocation: Texas Prepares For A Data Center 'Invasion'

“The data centers are coming!” proclaimed a June Dallas Morning News editorial cartoon, featuring a Paul Revere-type character to make its point.

Revere is onto something: By 2030, Texas is expected to become the world’s largest data center market, surpassing even Virginia’s vaunted Data Center Alley. As of now, more than 580 data centers are operating or planned to open in Texas.

This boom in the making is fueling excitement about resulting construction jobs as well as concerns about data centers’ electricity usage. But the physical security of these facilities is another issue that Texas should keep in mind, said Liz Lasko, a data center security expert for Ameristar Perimeter Security.

Lasko will speak on a panel concerning physical security best practices at Bisnow's Data Center Investment Conference & Expo: South, to be held Aug. 12-13 in Arlington, Texas. In a conversation with Bisnow, she explained why developers and contractors would be wise to learn from the experiences of earlier data center projects.

Bisnow: How are the experiences of other states relevant to Texas as it ramps up data center construction?

Lasko: A lot of the places that built the fastest and built the first — Silicon Valley, Northern Virginia, Chicago — are now having to retrofit their perimeter security because it wasn't designed to accommodate future growth, and that's an expensive lesson. Some teams in other markets might wish they had specced their security infrastructure differently in 2012. The Texas market should take a look at those experiences and start having conversations about security needs in the design stage and not after the fact. 

Nine years ago, when I started on the owner side, the scale we're building toward today wasn't on anyone's radar — AI and its processing demands simply didn't exist yet as a driver.

Bisnow: Learning from the experiences of other regions is one thing, but Texas being Texas, the state must have some unique construction challenges, right?

Lasko: Yes, there's a different set of site challenges in Texas, often due to its clay soil. That type of earth allows for soil movement that can crack foundations or compromise barrier footing embedment, which impacts crash rating performance if the perimeter fence is not engineered correctly from the start.

Also, you can encounter underground surprises when you start moving dirt on Texas greenfield sites. These can include undocumented utilities, irrigation systems or drainage infrastructure from who knows how long ago. Those things can force midconstruction barrier relocations, resulting in additional time and cost.

Bisnow: What advice do you have for Texas general contractors hoping to avoid or mitigate the issues we’ve been discussing?

Lasko: It starts with the bid process when a GC is reviewing perimeter security specs for the first time. If they're reviewing those specs only after they've won the job, then they're already kind of behind the scope. 

It also should not just be handed over to the civil sub with the assumption that it's a straightforward work item. Perimeter security spans across civil, electrical and security subcontractor scopes simultaneously. The coordination gaps between those three disciplines is where your schedule dies if you don’t figure it out very early on.  

Bisnow: What are some questions about perimeter security that GCs or other stakeholders should ask to avoid problems?

Lasko: It goes back to preconstruction: What are the civil prep requirements for the barrier system before it even ships? Has the electrical rough-in for the gate locations been coordinated, or are we going to stub conduit in the wrong place? What does the geotech report say about the soil-bearing capacity for the bollards? A big one: What happens if the owner's end user requires a totally different crash rating from what was specified? 

Bisnow: The assumption might be that if a problem arises, it can be addressed through value engineering, right?

Lasko: As with anything you buy, whether it is data center-related or a personal item, there's going to be a cheaper version. But the cost conversation needs to happen with the right expertise in the room. The GC is not a security expert and does not know enough about what the hyperscaler or colocator’s basis of design was, or understand the work and time that went into creating a custom security plan, for them to just go with a cheaper solution that is “somewhat equivalent.” “Somewhat” is not going to fully secure your site.

With perimeter security specifically, value engineering doesn't eliminate costs. It really just migrates them because you’ll have to go back later and add certain things. For example, if a fence system isn't designed with built-in conduit pathways, it will require excavation to add cameras or lighting later, when it could have just been taken care of from the get-go. 

Bisnow: What message do you want to leave with the people who are building Texas’ coming wave of data centers?

Lasko: There are a ton of data center security vendors now where there used to be just a few key players. The competition in Texas is getting fierce because everyone wants a piece of the pie, and that underscores the need to vet who you're choosing to have partnerships with because not all of the vendors have been in this market long enough to understand the nuances of building data centers in Texas.

Click here to register for the Aug. 12-13 Texas data center event.

This article was produced in collaboration between Ameristar Perimeter Security and Studio B. Bisnow news staff was not involved in the production of this content.

Studio B is Bisnow’s in-house content and design studio. To learn more about how Studio B can help your team, reach out to studio@bisnow.com.