by Tim Holt
There has been a lot of talk lately about data centers coming to our little corner of the desert. Meta is building a huge one in El Paso, and Project Jupiter is chugging along in Doña Ana County. Ft. Bliss is building one and no one seems to even talk about it. People, rightly so, have raised questions about electricity, taxes, jobs, development, and especially water. Considering that we live in a desert where the Rio Grande is neither a rio nor grande anymore, water seems like a pretty reasonable thing to worry about.
Our Southwest is experiencing a serious drought, made worse by an exceptionally poor winter snowpack, persistent heat, and a warming climate. We all have heard about how the Colorado River Basin has now endured more than two decades of drought, while Lakes Mead and Powell west of us have fallen to extremely low levels. Closer to home, Elephant Butte Reservoir, a key source of Rio Grande irrigation water for southern New Mexico and the Mesilla Valley, is also extraordinarily low. Low as in roughly 28,000 acre-feet of water remaining in mid-September. Summer monsoon rains provided some local relief but are not nearly enough to restore any major reservoirs.
The result, of course, is increased pressure on both surface water and groundwater, making questions about how agriculture, cities, industry, and new developments use water especially important throughout the Southwest. I get it. I have been here for 65 years. It is hotter than ever, it is drier than ever, and water is at a premium. Just look at places like Ruidoso, where Ski Apache has struggled through increasingly difficult winters with unreliable snowfall for decades now. If they didn’t have manmade snow making machines, they would have folded years ago. Sierra Blanca is gonna have to be renamed pretty soon, seeing as it seems to have less and less “blanca” on it. Sierra Man-made Blanca anyone?
Water is a big number problem. The problem when dealing with big numbers, however, is that they don’t mean much unless you have something to compare them to. Tell me a data center will use 175 million gallons of water a year and my first reaction is, “Holy cow, that’s a lot of water.” And it is. Like 265 Olympic swimming pools of water. And 175 million of almost anything sounds like a lot. I would happily accept 175 million dollars (tax free, of course). I would be less enthusiastic about 175 million mosquitoes.
But big numbers need context. Fortunately, we have a pretty good comparison growing just up the road.
Drive north from El Paso toward Las Cruces on Highway 28 and you will pass through thousands of acres of pecan trees. They are beautiful and green during the summer when everything around them seems to be some variation of brown. We all drive through there to show off to our visiting relatives. “See Grandma, the desert isn’t so bad, right?” They are also an important part of the economy of southern New Mexico, and families have been growing pecans there for generations.
They also use one heck of a lot of water. A big number lot of water.
Doña Ana County has roughly 31,000 bearing acres of pecan orchards. New Mexico State University uses about 50 acre-inches of applied irrigation water per acre in its pecan production estimates. Spread that across all those acres and you end up with something in the neighborhood of 42 billion gallons of irrigation water applied to pecan orchards each year.
That’s billion with a B. Billion.
Before someone from the pecan industry shows up at my house carrying a pitchfork, there is an important qualification. Applying 42 billion gallons doesn’t mean the trees actually consume 42 billion gallons. Some irrigation water moves back into groundwater. Some remains in the soil. Some leaves the field as runoff. A lot of it evaporates. NMSU researchers have a better way of estimating what is actually consumed. It is called evapotranspiration, usually shortened to ET. It combines water evaporating from the land with water that the trees pull from the ground and eventually release into the atmosphere. ET, send water.
Studies of Mesilla Valley pecan orchards have put ET at around 39 to 40 inches of water per year. Across roughly 31,000 bearing acres, that works out to somewhere around 34 billion gallons annually.
Put that number over here on the table for a minute. 34 billion gallons. About 51,500 Olympic swimming pools of water.
Now let’s look at Meta.
El Paso Water says an independent consultant estimated that Meta’s El Paso data center will average about 480,000 gallons of water per day. You may have heard the much scarier number of 2.5 million gallons per day. That number is real, but it is the contractual maximum, not the expected average. The larger amount would be available during unusual conditions such as periods of extreme heat. Yes, I know, we all have been experiencing extreme heat. Thank you, global climate change.
At the expected average, Meta would use about 175 million gallons of water a year. Again, 175 million gallons sounds enormous because it is enormous. But now we have something to compare it with. And importantly, that 480,000-gallon daily estimate isn’t simply Meta telling us how much water it thinks it will use. It comes from an independent consultant’s estimate reported by El Paso Water.
Pecan orchards: about 34 billion gallons in annual evapotranspiration. About 51,500 Olympic swimming pools of water.
Meta: about 175 million gallons in projected annual water demand. About 265 Olympic swimming pools of water.
Now, these aren’t perfectly identical measurements. Pecan ET estimates water actually transferred into the atmosphere through evaporation and transpiration, while Meta’s figure represents its total projected water demand. But that distinction doesn’t erase the enormous difference in scale.
The annual evapotranspiration from Doña Ana County’s pecan orchards is roughly 190 times Meta’s projected annual water demand. Here’s another way to look at it. About 160 acres of mature pecan trees have roughly the same annual evapotranspiration as the projected water demand of the entire Meta data center.
Doña Ana County has around 31,000 pecan-bearing acres.
Then there is Project Jupiter, which makes the comparison even stranger. According to Oracle, the four data-center buildings will need about 10 million gallons to initially fill their cooling systems. After that, the water-based cooling liquid circulates through a closed system rather than constantly being dumped and replaced.
According to Oracle’s current projections, the cooling system should need no more than about 1,000 gallons of makeup water per year. Its fuel-cell system adds about 167,000 gallons annually. There will also be ordinary water use for employees, bathrooms, kitchens and all the other things that happen when human beings occupy a building.
When you spread the initial fill across 15 years and include projected workplace water use, Project Jupiter comes out somewhere around 8 million gallons per year.
That is roughly the annual evapotranspiration of eight acres of mature pecan trees. Eight acres.
Now, none of this means data centers are wonderful. Far from it. Nor should you assume that I am “pro-data center.” I am, however, pro “let’s-compare-big-water-users-to-big-water-users.” It also doesn’t mean Meta should get whatever it wants. It doesn’t mean Project Jupiter deserves a free pass. Data centers consume enormous amounts of electricity. There are legitimate questions about tax incentives, power generation, transmission lines, infrastructure, noise, backup generators, environmental effects, and how many permanent jobs these gigantic facilities actually create. Ask those questions. Ask lots of them.
If water is going to be one of our major objections to these projects, shouldn’t we at least put the water numbers into perspective?
Which brings me to something called “selective indignation.”
Selective indignation, a term I learned as a child listening to the daily Paul Harvey News (12PM everyday on KHEY 690AM), happens when we become furious about one thing while barely noticing something very similar happening right next to it. It doesn’t necessarily mean our original concern is wrong. It means we have decided that one version of the problem deserves our attention while another version doesn’t.
Humans seem particularly good at doing this with things that are new.
A data center is new. It is huge. It has Meta or Oracle, or the US Army attached to it. There are politicians, corporate spokespeople, tax agreements, construction crews, environmental groups, public meetings and newspaper stories. Put “data center” and “millions of gallons of water” in the same headline and people understandably start paying attention.
Pecan orchards are different. They have been here forever, or at least they seem that way. They are green and beautiful, and they make the drive to Mesilla look like something other than the Chihuahuan Desert. Nobody drives past a pecan orchard and thinks, “There goes an industrial-scale water consumer.”
Nobody holds a press conference when somebody opens an irrigation gate for hundreds of thousands of gallons of water to flood an orchard. When something has been using water for 50 or 75 years, we tend to think of that use as normal. When something new comes along and asks for water, suddenly every gallon gets counted.
Some of that makes perfect sense. New development should have to explain itself. Agriculture supports families, workers, processors and businesses, and represents generations of investment. I am certainly not suggesting we bulldoze the pecan orchards and replace them with server farms. Besides everything else, that would make for a particularly ugly drive to Mesilla for a combo plate at La Posta. I am suggesting that we ought to be consistent.
If 175 million gallons of annual water demand from Meta worries us, shouldn’t 34 billion gallons of annual pecan evapotranspiration at least be part of the same regional water discussion? If Project Jupiter’s water use concerns us, isn’t it useful to know that a relatively small patch of mature pecan trees can use a comparable amount?
Those questions aren’t anti-pecan. They aren’t pro-data center. They are questions about how we decide which water use deserves our attention. And maybe that is the more interesting issue.
Living in a desert means we probably don’t have the luxury of deciding which gallons count based on whether we like what they are being used for. We need to know how much agriculture uses, how much industry uses, how much cities use, how much we pump from aquifers, how much returns to the system and how much is actually consumed.
We should absolutely ask Meta where its water will come from. We should ask Project Jupiter the same thing, as well as Ft. Bliss. We should demand good numbers, conservation plans and honest answers about what happens during drought. But perhaps we should ask those questions of everybody.
It is easy to become angry about a giant new building filled with computers. It is harder to look at something beautiful, familiar and economically important and ask the same uncomfortable questions. It doesn’t mean the beautiful, familiar thing is bad. It doesn’t mean the new thing is good. It simply means that outrage isn’t a particularly useful unit of measurement. Gallons are.
And if we’re going to worry about water in the desert, maybe we should start by counting all of it.
Sources
USDA National Agricultural Statistics Service — 2025 New Mexico Pecan County Estimates
New Mexico State University — Sample Pecan Enterprise Budgets to Produce Pecan, 2024
New Mexico Water Resources Research Institute — Remote Sensing of Evapotranspiration in Pecan Orchards
El Paso Water — Meta and Water Q&A
Oracle — Project Jupiter’s Water Use: What Southern New Mexico Residents Should Know
Oracle — Project Jupiter Water Facts
Read: Read More