Prove the Water Before You Close: What a Texas Data Center Site Can Really Supply

A data center site is only as good as the water it can actually permit, so prove the supply before you close on the land. On this episode of the MES Live Show, Mike Groel walked through how a Texas developer finds out what is really behind a utility's friendly answer, and what a well log does and does not tell you. The short version: start with your real cooling demand, check the permitted rights, and get hard evidence before capital is committed.

How much water does your data center actually need?

Start with the demand your real cooling concept creates, then compare it against a source that has permitted water behind it. Take the gallons per day the design needs and work down the list of sources against that number, one source at a time. It is simple arithmetic.

Do not put in a low demand early to make the project look smaller. If you have expansion plans, lock up and entitle the land for them now, not after the first building.

A phased or warehouse project that can sit a few years before it is built is a different case. That timing is a choice you can make.

This is the kind of question we work through with developers and builders before a site is committed.

What can you learn about a utility before you ask it for water?

A lot, from public records, and doing that work first keeps your demand out of a public meeting you are not ready for.

The state's public water system records show what a system has actually permitted, from its pumps and groundwater storage to its connection count. Read those against each other and the shortfalls show up quickly.

The Texas Water Board also publishes planned water projects for an area, with funding and dates on them. Layer that on top of the system records as another data point.

Without desktop research, you have to publicly disclose your demand in a feasibility study with the city, water district or utility. Those meetings get contentious, and people show up.

One developer asked a small utility for 1,000 LUEs, living unit equivalencies. That system served 1,600 total, and its permitted water rights covered only about 100 more.

The same records reported a 45% water loss, which the district manager disputed. Mike's lesson from that call: ask in a way that does not suggest the manager is running a poor system. Frame it as finding water the district may not know it has.

Where can water come from when the utility line is not enough?

Utility, groundwater, reuse and rainwater capture, in whatever combination holds up for your site. These are options, not a sequence.

On the utility side, the feasibility study compares its permitted rights and contracts against your full buildout, not against a first phase.

In a groundwater conservation district, an exempt well is usually tied to acreage, and the right is proportional to the land. A campus that buys 5,000 acres and builds on a couple hundred can supply itself.

Domestic wastewater on site is an easy permit, but a data center does not have many people, so the volume is small. The bigger play is a wastewater collection main running past the site, where untreated wastewater can be pulled for cooling.

Remember that a closed loop still needs water to fill it, one to two million gallons on some projects, and it loses about 10% to heat and evaporation.

Rooftop capture drains into a tank and gets treated for cooling reuse. Sizing that tank is an iterative calculation against rain events and how much demand it can supplement.

Why is a well log not a dry-year test?

Because a log records one day of drilling, and a wet year can make a marginal well look strong.

Start with a radius map of the drilling logs the Texas Water Board keeps. That shows likely yield and depth, and it supports a yes or no on a well before anyone spends money.

Add the geology, both surface and aquifer, to see where the water can actually come from.

Then ask when the test was done. Mike has seen a log with strong production while the wells around it did poorly, drilled in a wet season with a high water table influenced by surface water.

Your campus has to run through the dry year, not the year the log was written.

What does water due diligence cost, and what does skipping it risk?

Thousands for an engineer and a test well, against utility asks that can reach the tens of millions.

Checking costs thousands: an engineer to read the records and the logs, and possibly a test well on your own ground.

Some utility agreements land in the $20 million range for 500 LUEs. That is the number to weigh a test well against.

The health of the system matters too. A utility that cannot refill after a main break stops your cooling, and that is a large investment to put at that risk.

Electrical connections get the emphasis at most data centers. Water is the give and take that follows, so get good evidence before full buildout.

We put that case in writing before a site is committed, which is the same ground water certainty before site commitment covers.

Watch the full episode

The full episode is on YouTube: https://www.youtube.com/watch?v=Vyya4bkp_UA (opens in a new tab)

Frequently asked questions

What is an LUE?

A living unit equivalency, the unit many Texas utilities use to measure demand. A data center's water request is often expressed in LUEs, which makes it easy to compare against what a system has permitted.

Where do I find public records on a utility's water supply?

The TCEQ public water system records show permitted pumps, groundwater storage, connection counts and reported water loss. The Texas Water Board publishes planned water projects for an area with funding and dates.

I'm a developer with a Texas data center site under contract, and the utility says it can serve us. What should we ask before we close?

Ask for the permitted water rights and the contracts behind that answer, and compare them to your full buildout rather than your first phase. Also ask about the system's reported water loss and overall health.

I'm scoping a data center in a Texas groundwater conservation district, and the well log next door looks strong. Can we count on it?

Not on its own. Check when that test was drilled and whether the year was wet or dry, then look at the surrounding logs and the geology. If the yields still look possible, a test well on your site is worth the money.

Does a closed loop cooling system still need water?

Yes. It takes one to two million gallons to fill on some projects, and it loses roughly 10% to heat and evaporation. Reuse and rainwater capture can offset part of that.

Is rainwater capture enough for a data center?

Treat it as a supplement. Rooftop capture into a tank and treated for cooling is sized by an iterative calculation of tank size, rain events and how much demand it can offset.

Related Resources

• Data Center Water Supply in Texas: How to Build a Water Plan That Holds Up
• Texas Data Center Water Supply: What Developers Are Getting Wrong Before They Break Ground
• Can Treated Wastewater Cool a Texas Data Center? How to Cut Potable Demand and Plan the Blowdown
• Data Center Water Supply in Texas: How to Find More Water Before You Rely on Electricity

Is the water behind your utility's answer real?

We read the permitted rights, the contracts and the well logs before you commit, and tell you plainly what holds and what needs a test.

We specialize in:

• Water engineering and supply feasibility
• Wastewater engineering and reuse planning
• Site civil engineering for land development
• Water due diligence before site commitment
• Utility coordination on permitted rights
• Construction administration

Modern Engineering Solutions, McKinney, Texas. Contact: (214) 833-6748 or mod-eng.com