Hacker News Reader

Comment by piltdownman | Hacker News Reader

Texas is the only state in the lower 48 that has no major connections to neighboring power grids. That means growing energy demand in Texas must be met by new power generation in Texas.

Texas has not improved energy efficiency standards since the 2021 blackout, and have resisted all attempts at increasing the governance of the gas generation.

The Electric Reliability Council of Texas' "Capacity, Demand and Reserves report" even details a scenario in which massive energy demand growth in the state surpasses available supply in 2026.

https://www.ercot.com/files/docs/2025/02/12/CapacityDemandan...

Now add data centre demand in a climate where passive-cooling isn't viable and inter-state redundancy is non-existent.

View on HN

https://app.electricitymaps.com/map/zone/US-TEX-ERCO/live/fi...

Despite all the yapping, there's a crap ton of (ever growing) solar and wind and battery storage and what not in Texas. And ERCOT does have a power link with the neighboring SPP.

https://www.spp.org/documents/71831/ercot-spp%20coordination...

No one wants the grid to collapse after all (except you I guess, for whatever reason).

https://en.wikipedia.org/wiki/Texas_Interconnection

> The Texas Interconnection is tied to the Eastern Interconnection with a 220 MW DC tie near Oklaunion, and a 600 MW DC tie near Monticello, and is tied to NERC (North American Electric Reliability Corporation) systems in Mexico with a 300 MW DC tie near McAllen, and a 100 MW VFT tie near Laredo.[29] There is one AC tie switch in Dayton, Texas

Including in-flight projects, that totals several GW. Texas demand today is pretty low for noon - clouds and rain in much of the state, and we're at 70 GW.

Building interconnects costs money. What's the ROI on building more?

Positive ROI becomes present if Texas has large amounts of nondispatchable power and the neighboring areas do not have the same. So if Texas builds more wind and solar than they can consume themselves, then it makes sense to invest in more interconnection capacity. But that's predicted on excessive production first.

Finally, let's just say they built 70 GW of interconnection, because you seen to think that number is important. Do you think Louisiana and Arkansas have that much generation and transmission capacity themselves? Let alone EXCESS?

Washington and Oregon are on the same grid, with about peak generation of 31 GW and 18 GW, yet only about 7 GW of transfer capacity exists, and it's only that high because of sales to California because of seasonal hydroelectric oversupply, an oversupply which doesn't exist in Texas, Arkansas, or Louisiana.

Significantly higher interconnection capacity does not yet make financial sense. They're right to not yet waste money on them.

I haven't been arguing for or against anything, and I don't think the original thread creator was, either.

The comment from @piltdownman merely clarified that Texas is for the most part an isolated grid. That if demand matches predictions from our state regulator, Texas needs to continue massively increasing energy supply.

Texas is the second-biggest state. Where are they going to make major connections to -- the Great Prairie? There's a whole New Mexico-worth of sparse population between Dallas/Austin/San Antonio before you get to New Mexico itself, which you would then need to cross halfway before you hit a major population center.

Enron fiasco put a local power company here in WA in insurmountable debt because they couldn't ship power to California because the lines were already overloaded. If you build a major new power-consuming plant in Washington, you'll need to get power from someplace closer than half the width of Texas (and only even that far because historically we had coal power plants in Montana, so there's existing long-distance transmission).

I'm not saying they haven't made mistakes, but saying a place had "no major connections" is both wrong and ignores why. El Paso has "major" connections to New Mexico. It shouldn't be surprising Dallas doesn't have "major" connections to New Mexico, just like Denver CO doesn't to Portland OR.

>> El Paso has "major" connections to New Mexico.

El Paso is not part of the Texas grid (ERCOT):

https://www.epa.gov/green-power-markets/us-grid-regions

Neither are the panhandle nor the northeastern portion.

GGP didn't say ERCOT, they said "Texas". Which further demonstrates how ignorant it is to say there are "no major connections".

Here's some reading:

https://en.wikipedia.org/wiki/Texas_Interconnection

Most of Texas is on this grid, not connected to any other states. The total amount of power that can flow between the Texas Interconnect and the rest of the US grid is a small fraction of the Texas grid's demand. That was the point being made by the original commenter.

That's a stupid point to be trying to make and irrelevant.

> Texas is the second-biggest state. Where are they going to make major connections to ...

Anywhere else in the transmission network that has excess generation capacity. If the biggest state benefits from interconnection, why wouldn't the second biggest benefit too?

Daily demand is cyclic, even with no additional capacity added, being connected to grids in other timezones that have different demand peak-times is a net win.

>in a climate where passive-cooling isn't viable

Chips can also be water cooled, and Texas borders an endless supply of water in the Gulf of America.

Are you being serious or sarcastic?

Completely serious. Sea water cooling for data centers is already in use here in Helsinki. And water temperature in the Gulf of America doesn't seem to go above 30 C even in the summer.

https://news.ftcpublications.com/core/coastal-cities-test-se...

https://www.aoml.noaa.gov/phod/regsatprod/gom/sst_map.php

Sea water would have to be piped to places like Dallas, whereas much of Finland's people/data centers are closer to the shoreline. A solution that works in certain places does not generally work everywhere.

There's no need to build them in Dallas. Just keep them near the shoreline. The only reason you have so many data centers in North Jersey is so high frequency traders can make their calls 0.02ms before their competitors on the NYSE. For an average user, the 1ms ping between Dallas and Corpus Christi is not a problem.

They are many already planned/under construction/built/expanding, there are also wildfires and hurricanes to consider.

Its easy to offer hypothetical solutions for a far when details are amis. It's another thing to deal with local politics

We have a home in Houston. Over by the big gates down Sunset.

I have no particular objection to building datacenters there.

All that said, I think you're being real hand-wave-y here. The climate in Houston is materially different than it is in Dallas. So there's likely a reason these companies have been locating their enormously expensive assets where they have been locating them.

Houston floods. Often.

Houston gets hit by Hurricanes.

Houston is hot and humid with relatively high ground water tables.

Don't even get me started on the power and the durability issues arising from the use of sea water.

There's a lot to work out there. You put what you can afford to lose in Houston, but you don't bet the farm on coastal Texas. You just can't do that if you want resilient infrastructure.

Building closer to shore can bring other problems like a corrosive salt filled atmosphere and higher hurricane and flood dangers.

Im not saying it can't be done, but building near the ocean has its own set of drawbacks and costs.

Reply on news.ycombinator.com