I toured South Texas Project nuclear plant years ago and remember being shocked that their turbine deck was open-air. "We don't really get inclement weather here", I was told. Unfortunately, unit 1 is down bc a feedwater pressure sensing line failed due to cold. Unit 2 at 100%

They gave me a nice hat though, which is by far the best swag I've gotten for just going on a plant tour

Even at 19, I had the sense that having an open-air turbine deck = wider potential for failures, even if bad weather is rare. After pushing the engineer, I was again told "these aren't safety-critical systems"... ok but if it fails, the reactor stops producing power (see: now)

Yeah, a sensor failure of a feedwater system is not a reactor safety issue. The reactor is fine. The issue is, we rely on reactors to produce electrical power, even (especially) in bad weather. Secondary systems are still critical in that regard

Here's an overhead view, you can see the turbine and generator on the roof. The one thing I have to admit, their turbines are a really pretty blue color, which you can sort of tell even here.

Not hating too much on the muted pastel yellows and greens that I've seen in many other plants (images not mine).... but I admit I like the bold blue better

Importantly, Comanche Peak, the only other nuclear power station in Texas (which I have never been to) also appears to have an open-air turbine deck. Hopefully STP unit 2 and Comanche Peak (2 units) are both taking steps to prevent any more reactors tripping (going offline)

I don't want to give the sense that STP and other reactors with open turbine decks have never considered that equipment could freeze. Cold weather conditions are part of the severe weather conditions that nuclear plants must plan and prepare for.

I don't have access to STP procedures, but in a document submitted to the NRC, training guidelines say that personnel should be trained to conduct system walkdowns during freezing conditions, where sensing lines are specifically noted as a system to be monitored

Reiterating that I don't know the whole story here, besides what has been reported by the NRC and in media statements such as this one reported by @Atomicrod
https://atomicinsights.com/south-texas-project-unit-1-tripped-at-0537-on-feb-15-2021/

Given what has been reported, it's reasonable to think that more systems should be freeze protected going forward, and/or systems with the potential to fail due to cold should be identified & temporarily insulated faster in emergency conditions. But this was not an unknown risk

Well this is an interesting and entirely-different cold weather failure mode! https://twitter.com/TXAg2014/status/1361888243348824065

Hearing from a former senior reactor operator at STP that they're beginning the process of coming back online https://twitter.com/rickrocket1/status/1362043771853164546?s=19

https://twitter.com/nuclearkatie/status/1362085438823878656