Data Centers Are Quietly Taking over Texas. The Pollution Could Be Catastrophic

Texas will allow big players to do just about anything they want as long as the promises of "future jobs" and "additional tax revenue" are kept, even faintly.

My tenure at one facility ended because of an OSHA investigation (which I do not wish to discuss any further than: just as HR works for their corporate owners, OSHA works for their corporate owners).

Safety codes are written in blood. Faintly. They must often be re-written.

----

My hope is that the current datacenter over-buildout results in a similar disaster to the early-2000s destruction of sound-recording studios by home computers with ProTools (e.g).

For my own futureproofing, I've purchased a 5070Ti for local LLM ponderings.

----

A brother is a statewide whore critter ("elected politician", at the Texas level), and my advice to fellow Texans is to do what I did: leave (as a 5th-gen native, I left over a decade ago; as did our other brother).

----

Different state, same shitty "leadership" :

<https://www.techdirt.com/2026/07/09/muskcorp-tries-to-bribe-...>

We've all seen the documentary Idiocracy (2006) – also check out its predecessor They Live (1988):

<https://www.imdb.com/title/tt0096256/>

Some of the articles claims are valid; the datacenter construction has had disruptive traffic, caused some traffic fatalities, etc. And if my property were right next door, I would also be pretty upset about the dust and noise. But I think this boils down to standard NIMBYism, which is a thorny problem with civilization in general, not specific to data centers or really news--it also sucks when there's loud testing and flyovers from Dyess AFB nearby. The DC has had other impacts, like reduced housing availability and significant increases in average rent, which sucks.

But the core claim of the article is "the pollution could be catastrophic", and I don't think that actually pans out. Solar and wind is abundant and cheap here; Abilene has significantly above-average production of wind especially, and the datacenters being built nearby (there are 3-4 different projects now) are smack-dab in the middle of it.

Yes, the gas generation capacity is permitted to run a year-round. But that's kinda like how you make sure AWS quotas are significantly greater than you'll hit in practice. I'm not a permit/regulatory expert, so I could be wrong here. The reality is: no rational actor would primarily use gas generation. It doesn't make economic sense. Leaning on wind power is significantly more cost-effective. And for short-term power generation gaps, there's 4 hours' worth of on-site battery capacity.

The gas turbines only make sense to run when wind/solar isn't producing for a prolonged period. It seems like the gas turbines would only be the rational choice about 5-15% of the time, based on the past year of ERCOT data (disclaimer that I used an LLM for some of the analysis here).

The DC operator is certainly incentivized to run the datacenter 24/7 to keep the GPUs occupied, so it's worth the cost of paying for a full gas power plant, rather than shutting down during wind/solar low periods. So no, the gas generation doesn't fill the primary role of "backup" power--it's intended to ensure continuous operation. But that doesn't mean that the operator WANTS to run it 24/7 on gas, and they would be incentivized by market forces to reduce gas usage as much as possible, as long as they can stay online.

Would love to understand if I'm off-base on this, or if other folks could double-check that usage estiamte.

I guess I assume if they have the capital required, and a willing seller for land/development, the public's opinion doesn't really matter