Water reservoir and bridge infrastructure
Long-range water planning matters even when short-term reservoir conditions improve.

What Is Happening

Data center growth is accelerating, and that changes the local water conversation even before a specific Temple project appears.

Lawrence Berkeley National Laboratory's June 2026 United States Data Center Energy Usage Report: 2025 Update estimates that U.S. data centers could account for 11.8% of total U.S. electricity use by 2030, with a scenario range of 9.5% to 15.3%. That does not tell Temple how much water a future project would use, but it does explain why more communities are being asked to host facilities that run continuously and need dependable cooling.

The water side is more nuanced than many headlines suggest. A June 2025 Lawrence Berkeley National Laboratory review found that workload-level data center water use can vary by more than 10,000-fold depending on factors such as server efficiency, grid water intensity, cooling-system type, climate zone, and utilization. In other words: "data center" is not a precise water category by itself.

The cooling choice matters because some designs are explicitly water intensive. The U.S. Department of Energy's Federal Energy Management Program says cooling towers reject heat through evaporation and, by design, use significant amounts of water. DOE also notes that operators can lower freshwater demand through measures such as higher cycles of concentration, reverse-osmosis reuse of blowdown, and some direct liquid cooling configurations.

Temple's water reality is not hypothetical. The Brazos River Authority says Temple has contracted with the authority for 30,453 acre-feet of water per year and draws supply from the Leon River downstream from Lake Belton. The same BRA customer profile says the authority has provided wastewater treatment services for Temple since 1975.

Bell County planning already assumes tighter water conditions over time. In the Texas Water Development Board's Bell County water-planning summary, Temple is listed with a projected 2070 water need of 17,103 acre-feet per year under drought-of-record planning assumptions if recommended strategies are not implemented. The same summary shows Temple's recommended strategies as aquifer storage and recovery, conjunctive use, municipal conservation, and other surface water. Table A2 in that summary indicates Temple's projected need rises to 47% of demand by 2070 under those planning assumptions.

That does not mean Temple is about to run out of water tomorrow. The same TWDB Bell County summary says that between January 2016 and November 2023, no public water systems in Bell County reported having approximately 180 days or less of water supply remaining. But it does mean long-range water scarcity is already part of the official planning frame.

Current conditions also show why the debate cannot be reduced to one lake-level screenshot. The BRA's drought notices page shows Lake Belton was included in a Stage 1 drought watch initiation on March 30, 2026. Yet the BRA homepage listed Belton Lake at 594.34 feet on August 9, 2026, slightly above the reservoir's 594-foot conservation-pool elevation listed by TWDB. Short-term conditions can improve while long-term drought planning pressure still gets worse.

Regional utilities are planning accordingly. On the BRA's current projects page, the Belhouse Drought Preparedness project is described as a raw-water transfer system between Belton Lake and Stillhouse Hollow Lake, with an estimated request for bids in fiscal year 2030 Q1. That is not a data center project. It is evidence that Central Texas water providers are already spending time and money on drought resilience before layering in more large-load growth.

Why It Is Happening

The real argument is not about whether computers "drink water." The real argument is about which kind of facility a city is being asked to host, under what cooling design, and on top of which water system.

A lot of public debate treats data centers as a single environmental category. They are not.

Some facilities rely more heavily on evaporative cooling. Some lean harder on air cooling, hybrid systems, or direct liquid cooling setups that reduce site water use but may raise other costs or energy tradeoffs. Some can reuse non-potable water streams more effectively than others. Some are designed around peak summer performance in hot climates. Some are not.

That is why Temple should be skeptical of both easy narratives.

The first bad narrative is: data centers barely use water now, so there is nothing to discuss.

That is too simplistic. DOE is explicit that cooling-tower-based systems are water intensive by design, and Central Texas is exactly the kind of hot-weather region where cooling design matters. A city that never asks what water source is being used, what the peak-day draw looks like in August, or how the facility behaves under drought restrictions is not doing diligence. It is outsourcing diligence to the developer.

The second bad narrative is: every data center will automatically wreck the local water system.

That is also too simplistic. The LBNL review makes clear that water use varies dramatically across workloads and facility designs. Some operators can cut freshwater demand materially through reuse, filtration, and alternative cooling configurations. The question is not whether the industry can say the words "closed loop." The question is whether the specific project in front of Temple would commit to measurable limits and enforceable fallback conditions.

For Temple, the right framework is boring on purpose:

  1. What water source would the facility use in normal operations: potable, raw, reclaimed, or some combination?
  2. What is the maximum daily draw during hot-weather peak conditions, not just the annual average?
  3. What cooling architecture is actually proposed, and how does it perform during drought and extreme heat?
  4. What reuse systems are included from day one rather than promised later?
  5. Where does the facility sit in the drought-curtailment order if Lake Belton conditions deteriorate?
  6. What public infrastructure upgrades would be required, and who pays for them?
  7. What reporting would residents receive after the ribbon cutting, not just before it?

Those are infrastructure questions, not ideological ones.

Temple's own planning context makes them more urgent. If TWDB planning already shows large future drought-of-record needs for Temple without additional strategies, then any major new water-dependent industrial user should be evaluated against that future, not against a single wet month.

This is why water debates around data centers often feel muddled. People are arguing about symbols when they should be arguing about operating conditions.

What Might Happen Next

My forecast is that Texas communities will stop asking whether they are "for" or "against" data centers and start asking which water terms they can force into the deal.

That is the mature version of this debate.

For Temple specifically, I expect three things.

First, any serious future digital-infrastructure proposal will face sharper public scrutiny on water than earlier industrial recruitment pitches did. People now understand that power, water, and heat rejection are not side details. They are the project.

Second, utilities and developers will get better at presenting water-saving language. That will make public review more difficult, not easier. "Efficient cooling," "closed loop," and "reuse-ready" can mean very different things depending on the actual design. Temple will need operating thresholds, not marketing adjectives.

Third, reclaimed water and stricter drought-trigger rules will become more central to credible project design in Central Texas. The projects that survive public skepticism will be the ones that can show how freshwater demand is reduced, how peak-day demand is managed, and how the city keeps leverage when drought conditions tighten.

If Temple ever gets a large data-center proposal, the city should not begin with "How many jobs?" It should begin with "Show the water math."

If the developer cannot document the cooling system, source mix, drought behavior, reuse plan, and monitoring commitments in plain language, the city should assume the risk is being shifted onto the public.

If the developer can document those things and accept meaningful conditions, then Temple can have a more serious conversation about whether the trade is worth it.

That is the real water debate.

It is not whether technology uses resources.

It is whether Temple is disciplined enough to price the risk before it approves the growth.

Sources

Image credits: Temple and infrastructure images are tracked in the Pheorix image credit files.