Basically, there was a time (1980's) when fabs in Austin had their own power generation. AMD, which used to have its own fabs, had its own power for them. Then, they (not just AMD but everybody with fabs in Texas) eventually decided it wasn't worth the cost, and started using the regular power grid, because Austin's power had become more reliable. Obviously, as the percentage of solar, wind, and natgas has increased, and the percentage of coal has decreased, this reliability has been thrown into question. Also, the rapid population increase has also just meant a general problem of building new capacity fast enough.
My guess is that they have an eye on it, and it is also possible they will build their own power generation, but more likely they are just getting guarantees of some sort that enough power capacity will be built.
It goes without saying that, just because Fox News is saying it doesn’t mean it’s right. Partisan bickering in either direction isn’t a good way to educate yourself on the issues.
Renewable power throws a wrench in how Texas does grid planning. In every other ISO, there is both a capacity market, which pays providers to commit to making certain generating capacity available, and a generation market, which pays for actual electrical production: https://cpowerenergymanagement.com/why-doesnt-texas-have-a-c.... In Texas, there is just a generation market.
Ordinarily that isn’t a big deal. Generators build adequate capacity so they can be in a position to receive payments for generation at times of peak demand.
Renewables break this down. They undercut traditional generation sources in the summer, but can’t be counted on to be there at times of peak winter demand. So last winter renewables didn’t fail in the sense that nobody was expecting them to generate much power begin with. But the natural gas plants that are irreplaceable for dealing with winter demand are dealing with reduced revenue because renewable sources are underbidding them in peak summer months.
It sounds like a free market working as expected. Maybe something important like this needs regulation to prevent optimization based on only one parameter ($).
Only if they are being told the truth. A functional free market requires symmetrical information (ie. both seller and buyer has the same information about the product). It is just a theoretical thing. But you can get closer to that ideal with regulation and strong buyer protections.
There’s no “free market” for electricity anywhere in the developed world, including Texas. In the US, the grid is centrally planned by several regional transmission organizations. In Texas that’s ERCOT. An artificial market, a bid-auction system, is used to decide the actual price of electricity at any given moment. But in the case of ERCOT they concluded that, because outages prevent generators from making money during periods of peak demand, generators had adequate incentives to invest in reliability without a separate capacity market. And that worked fine for decades!
It’s not a choice of “free market versus regulation.” It’s a heavily regulated market. The question is only about the design of the regulatory scheme.
I don't think they can charge whatever they like. There are some regulations on the max price a generator charge for electricity, which we hit during the last freeze.
I heard about people paying over $10/kWH and having $15k bills for about 5 days of power. I know people who paid well over $1000 for the month and their power was out for much of the cold spell (~50%). I hear they're going to try and lower billing to less than $9000/MWh, but that's not going to encourage building proper infrastructure, and making gouging profitable doesn't encourage any of them to improve reliability. Why bother weatherproofing?
I'm reminded of Enron's "grandma Millie" comments.
The things is there really isn't partisan bickering on both sides of this issue in the first place. There is non-partisan reality where poor planning by non-partisan entity predictable led to service disruption. Even the failure of legislators is itself a non-partisan failure. Ordinary incompetence.
The only partisan bickering is the attempt to incorrectly blame renewables for the lack of capacity when in fact planned downtime, ordinary logistical failures, and failure to winterize are in fact to blame.
Everybody ought to have expected them to need that much power. Not every day and not every winter but everyone ought to have expected another bad winter to come round because they had bad winters in 1957 1960 1973 1985 2015 2017 2021. This includes 3 years out of the last 6.
They weren't prepared because they were short sighted, greedy, and stupid not because solar took so much of the profits and not because insufficient capacity had been built out for lack of such profits. Capacity existed and it sat unused or broke when it was most needed.
For me it was an incredibly laughable shift of blame to renewables. People tried saying wind power wouldn't work in the cold, when Iowa, MN, and the Dakota's make huge amounts of it year round. I'm not a scientist, but pretty sure they get much, much colder
The Texas grid isn't run by stupid hippies who overbuilt solar plants forgetting they don't work in winter. It's run by "libertarians" who think that winterization standards are useless, and that's why nuclear plants, coal plants, and natural gas plants all failed when there was a freeze. This wasn't due to "reduced revenue" it's because the capacity planning didn't require winterization.
On top of that, when attempting to do rolling blackouts critical gas infrastructure froze and so gas supply plummeted. The gas plants weren't able to generate their capacity because there wasn't enough gas to burn.
So, I wouldn't know, because I haven't watched Fox News in many years. The best low-emotion, technically informed analysis of "Snowpocalypse" is probably this one from "Practical Engineering", made by a San Antonio-based engineer:
The outages were not from any one "main cause", except perhaps "Texas doesn't get cold that often and wasn't ready". But, wind icing up was certainly one of several major causes, along with natural gas pressure dropping so that natgas-powered electrical plants could not keep operating.
The most recent issues for which the Texas electrical grid was in the news for, the near-blackouts earlier this month, were specifically because of unpredictably low wind speeds in west Texas which reduced Texas's expected wind generation from ~20 GW to under 2 GW during the periods of peak demand. For context, peak demand in Texas is around 80 GW. Taking 18+ GW off the table is a huge blow. 18 GW is more than the entire electrical demand of most US states.
Fox news has nothing to do with it. Generation capacity being reduced by nearly 25% due to unpredictably low wind speeds is physics. It's a huge problem being faced, and sticking your head in the sand and crying "fake news" isn't helpful.
https://www.ercot.com/ gives you some idea of what's going on in TX at any given time. One neat thing about summer is that peak demand also tends to be because of AC use which also correlates with sunshine. Works for the summer at least.
It's actually not a meaningful factor AT ALL. Where renewables failed it wasn't because the nature of the technology but because of failure to winterize.
It was, and is, a hugely meaningful factor. Sticking your head in the sand and pretending that it's a republican smear campaign doesn't help.
Texas has more wind power generation than any other state in the US by far. Texas is all aboard the wind train. It's a huge part of the economy. Texas wants wind to win. But that doesn't change reality.
In the last month, Texas has gotten close to electricity demand exceeding supply. A significant factor behind this is that Texas gets nearly 20% of its electricity from wind generation. On your average summer day, wind generates between 15 and 25 GW. However, during the recent heat wave, wind speeds dropped in west Texas (where the bulk of the wind farms are) and wind was only generating less than 2 GW during the hottest part of the day.
Similarly, Texas usually gets about 10 GW of power from solar. However, solar drops off to 0 GW very rapidly around 7 or 8 PM. However, in the summer in Texas, the temperature is still at its peak around 7 PM, so there is still significant demand while solar generation is dropping.
Wind and solar are unlike thermal generation in that we (humans) can choose to burn more oil and create thermal generation when needed. But with solar and wind, we cannot choose to suddenly create more wind or sun. We are at the whims of nature, and until we figure out better solutions for these problems (battery storage, maybe), wind and solar have their disadvantages compared to thermal. Pretending otherwise is not helpful.
> We are at the whims of nature, and until we figure out better solutions for these problems (battery storage, maybe), wind and solar have their disadvantages compared to thermal.
Pumped storage hydro works very well in many places. You can store as much energy as you can store water, and you can bring a lot of generation capacity online fast. Those facilities are pretty cool engineering projects. Texas hasn't got onboard yet AFAIK, but they've got enough land that they could maybe (water's a big factor, too) make it work to offset downtime of their wind and solar.
edit: I think Texas's largely go-it-alone strategy with their power grid might be pretty strategically misguided, but making it work could at least be an interesting problem
> It was, and is, a hugely meaningful factor. Sticking your head in the sand and pretending that it's a republican smear campaign doesn't help.
This could have been a really good comment if you hadn't started it out this way. Wind was one factor, but it did not suffer the largest outages of Texas's various electricity sources - someone in another comment shared a pretty good youtube video by Practical Engineering on this topic. And there really was a republican smear campaign against wind power following the event.
>Wind was one factor, but it did not suffer the largest outages of Texas's various electricity sources - someone in another comment shared a pretty good youtube video by Practical Engineering on this topic. And there really was a republican smear campaign against wind power following the event.
We're talking about different events. The Practical Engineers video is talking about the 2021 winter storm, and in the midst of that storm there were republican talking points about wind failure. That's a different event than the one I am talking about in my comment, which is the general unreliability of renewable energy as witnessed in the current summer where lacking wind generation _really is_ a huge factor in the threatened blackouts, but posters like the one I was responding to are still pretending like any decrying of renewable energy is fake news. Sometimes it is fake news, and sometimes its reality. It's important to recognize the difference.
I'll agree with you that there's greed and incompetence all over Texas's politicians and state management. But it isn't unique to Texas. We have a widespread societal problem with our electrical grid that transcends state lines.
Renweable energy was not the "main cause" because it reliably fails to generate adequate amounts of power at that time of year and so the grid was entirely reliant on other generation (mostly natural gas) to fill in the gap. Basically, it wasn't the fault of renewable energy because everyone knew it was useless anyway.
I have no idea why natural gas was third in your list. Texas's power issues are entirely predicted by natural gas power plants failing (or voluntarily shutting down because of price spikes).
My understanding is that the Texas grid issues are largely due to increased population and lack of a capacity market. Texas has been hitting record high demand again and again in the past few years. Aggregate demand went above ERCOT's worst case projections during the recent heat wave.
It's true, but they are able to buy and sell power from the eastern grid (my understanding is it's less efficient, because it goes to DC and back to AC at the interface, but it does happen normally). They weren't able to buy any power a year and a half ago, because the eastern grid needed theirs for the same winter storm, but presumably during summer heat waves it works better as long as points north aren't all running their air conditioners nonstop as well.
The low natgas line pressure was also related to the fact that points north were using more than usual as well. In fact, one of the responses to the power outage was to stop all natgas flowing from Texas to Mexico. So it's not only the power grid, but also the natgas lines, that connects Texas to points outside.
Given this is just an initial permission with no concrete plan to build they may also be trying to induce a bidding war with another location that is more favorable (but more expensive).
> 2) Texas's power grid has gained a reputation lately for being unreliable.
Having the appearance of something is not the same as actually being that thing. They had one outage during a cold snap, and then warned about high usage, but there has not actually been another outage.
That said, the failure and piss-poor government response to it makes me almost viscerally angry. Texas has promoted itself as an energy leader, from traditional oil and gas production, to large (and growing) production of renewables, we shouldn't be having these failures.
The issue for me isn't the failure of the power grid, extraordinary events happen - its that:
One - it's happened before (twice before).
Two - we should have had a governmental response to adequately winterize the grid.
Three - the deregulated power system in Texas creates perverse incentives to building more capacity in - and power usage in Texas is growing faster than either base load or peaking capacity is.
These are all governmental and regulatory failures, and ones that were both foreseeable and preventable.
They are in fact doing rolling "brownouts" as of a few days ago. I put it in quotes because the friend who told us about it had been without for a couple hours at that point.
This is not being done by ERCOT that manages power distribution in Texas. Your friend had a local power failure, probably caused by a downed power line.
> Here's something that confuses me about building fabs in Texas.
> Texas's power grid has gained a reputation lately for being unreliable.
As stated, there's nothing to be confused about. You'd want to build new fabs in Texas for the same reason you want to buy stocks while they're down rather than while they're up.
The reputation only matters if you're trying to sell your new fab to someone else. If you want to use it yourself, what matters is the reality.
> The reputation only matters if you're trying to sell your new fab to someone else. If you want to use it yourself, what matters is the reality.
Certainly. My point was that (perhaps) the grid in actuality has reliability issues, and the reason I thought that might be true is because of its current reputation.
Also odd since their state government leaders are highly politically motivated and want to make political waves even if their actions are irrational (like on power production, blaming wind turbines instead of recognizing their failure to winterize wind and coal power systems).
I think it's a favorable tax environment. Just remember that these companies claim to be progressive, but if they can save $20 on their taxes, that all goes out the window.
I think the tax environment for companies is favorable, also fewer environmental regulations (and making chips uses some nasty chemicals)
Just wanted to say that taxes in Texas are less favorable than eg California when it comes to middle class citizens - no income tax, but very high property taxes.
I've driven through Texas, and there is this 300 mile+ stretch, in the North, that is hell on earth. Pure putrid smog, the skies darkened, the air thick, dank, cloying from refineries or something.
So yeah, great place to build something, and save cash on tech to cleanup your waste. Yay Texas!
I swear people must die 10 years sooner in that stretch.
There are worse places just in Texas - there's a chemical refinery locus to the east side of Houston near the port and bigger one near the border with Louisiana with so many flares at night it reminds one of Mad Max a bit, though there are various plants scattered along the gulf coast as well, were a half dozen that I knew of in the mostly rural county I grew up in to the south of Houston.
Flare gas recovery systems do burn the excess gas to power the refineries. I’d guess the generated heat is used directly without conversion to electricity.
That’s absolutely appalling, whether or not it shaves a full ten years off people. Reminds me of the one time I visited Ghengzao in China. The pollution was so bad it was cloying.
CA property taxes are highly regressive. Old boomers get to pay sub 0.1% tax rates because of Prop 13 while new, young, and poorer buyers get to pay tax on inflated property valuations.
When it comes to increasing class mobility high property tax >>> high income tax.
I don't live in CA, but forcing low or fixed income people out their homes just because their property increased in value which is often outside their control is not good.
From my understanding part of the problem also is even landlords and businesses benefit from prop 13 and same with people that own multiple properties. Honestly, I think something like Idaho's Homeowners Exemption is better. The owner has to live in it to qualify.
https://tax.idaho.gov/i-1051.cfm
Also states like Nevda have done some other interesting things like letting family farms get a Limited Allodial Title for a period in time before stopping. I just think prop 13 was not planned out very well and applies to too many entities, but the idea of not wanting to force people out their homes is certainly a problem other states worry about.
They can take a line of credit on their home and pay for the taxes, then when they die their home is not passed on to their children for free (and inheriting the low property taxes!). At CA property tax rates, people would not be forced out of their home in their lifetime.
Unproductive assets like real estate should not be enablers of generational wealth.
That's kinda how the Idaho Property tax deferral program works. Prop 13 from my understanding applies to businesses, landlords running apartments, people with multiple properties ect... A tax reduction that only applies if the owner is living it seems fine in my opinion. Most other states property tax relief programs require that property to be the owners primary residence.
Property taxes are weird as they are basically one of the few taxes that continue to happen even without any transactions. Almost all other taxes happen when there is a transaction. For example sales tax, when you buy something; Income tax when you get paid; Tariffs when something is imported ect... Property taxes can also can increase without you doing anything so if your income is fixed or does not increase fast enough you could lose your house due to a tax lien.
If your worried about generational wealth that's why we have inheritance and estate taxes. The federal estate tax can get as high as 40%. However, a home is a productive asset if people are living in it. It may not be as productive as some other uses of property like running a business from it, farming the land ect...
> Property taxes can also can increase without you doing anything so if your income is fixed or does not increase fast enough you could lose your house due to a tax lien
Like I said, you won't lose your house within your lifetime at current property tax rates. I see no reason why we should guarantee owners' houses for any longer than that instead of freeing up those houses on the market for new owners.
Tax deferral is a great idea, tax relief not so much.
that's a fair criticism a lot of the time, but I don't think Samsung ever makes any real claims of being progressive. not too many woke companiese have a division that builds tanks.
There is plenty of politics in Texas, for sure, but it wasn't coal power systems that failed in the winter, it was wind, natgas, and one nuclear power plant.
Iirc wind performed more or less as forecast (admittedly not great, but as expected) during Feb 2021 and nearly all of the shortfall came down to natural gas plants and infra failing.
I'm all in favor of wind power in Texas, and want more of it, but it did do worse than expected in Feb 2021. It's just a problem to be fixed, not a reason to drop wind power (west Texas has a lot of wind), but it did actually happen.
Yes--a lot of wind was offline. But all of the expected wind capacity could have been met and the grid would have still needed to shed quite a bit of load. In fact, all of the expected wind, solar, coal, and nuclear could have been online and the grid would have still had to shed a meaningful amount.
The scale of the natural gas outages was larger than the entire load-shed, afaict.
My understanding is that this is incorrect, and it was coal power plants that weren’t properly winterized that failed. Of course you had a lot of people claiming the opposite, without evidence.
EDIT: Okay there was some loss in capacity in wind power, but the greatest loss in capacity was natural gas plants, which lost 15GW of capacity. I confused natural gas with coal as I don’t consider either to be climate friendly. Wind lost about 3GW of capacity.
By the way, natgas is actually much friendlier (well, less hostile) to the environment than coal, if you're looking at carbon/kWh or all of the other things released into the atmosphere when you burn coal.
However, you can much more easily make a giant mound of coal next to your power plant, than store up enough natgas (LNG is much more expensive to store). So, the power plants required natgas to show up in the pipeline, and when the temperature dropped and everybody cranked up the heat, not only in Texas, the natgas pressure in the pipelines dropped. Oil (liquid) and coal (solid) are just fundamentally easier to stockpile than natgas or wind (gas) or solar (light).
A big freakin' battery would help with this, but that isn't a cost-effective option (yet?).
There is no way in hell you can run a modern fab using on-site power anymore. Maybe in the 80s/90s, but now you got things like EUV light sources that require megawatt-class power delivery all on their own. Just stepping back 1 generation to DUV and you are "only" talking ~100kW per machine, which is still insane.
What would probably happen is a special grid arrangement. If you are going to build eleven of these things, entire power plants, switchyards and transmission lines will need to be constructed ahead of time. They will likely need to run isolated from the broader grid in some manner. I cannot imagine Samsung executive management agreeing to this many factories (even with subsidies) unless some guarantee could be made WRT power delivery.
Sure. My point is that you might think they have some sort of fancy backup system or to be totally self-contained because otherwise the alternative is that when the grid has problems, they shut down, and surely they wouldn't do that?! But turns out, they do, and that is not merely a hypothetical and happened quite recently. So, the answer to 'what alternative do they use' may just be 'nothing'.
Proposed starts of operations are 2034-2042. I'd expect a fair bit of on-site generation/storage, but I also would hope ERCOT gets its act together by then (or is replaced with a multi-state system).
Fabs have all sorts of power systems because even the slightest blip or glitch in power could mean millions in losses
I know a decade or two ago IBM was using superconducting magnets for power smoothing to handle sub-millisecond events. Then you have battery and flywheel systems for bigger drops that are designed to carry load until generators can start and accept load (usually a very short period of time if the generator's oil and coolant are on heaters.)
Texas is very good at water planning. Solar does make sense if you have lots of land. How much skilled labor actually works at a fab? Attracting skilled labor to work far enough out, that that amount of land is affordable could be a problem. If the weather, and power situation remain. It might be hard in any case.
Depends where in Texas. Houston area is more similar to the Deep South in weather, no shortages of water, in fact dealing with floods is a huge part of their urban planning challenge.
> Probably. I would think that would be common in that industry no matter the location
In particular, I was thinking about the power disruptions in Taiwan earlier this year [0].
IIRC, there was concern that a power drop could ruin some/all chips currently in the fab pipeline, and might also require several days to re-initialize the pipeline's equipment.
With the recent Supreme Court's decision limiting the EPA's power (unless Congress acts to delegate the specific power that the court says the EPA lacked), soon it should be possible for the Texas grid to become more reliably by increasing the us of natural gas and decreasing the relative use of wind and solar.
Out of curiosity, what makes you think that the EPA is responsible for Texas having such a high rate of renewable energy? This seems slightly surprising, because there are large portions of the US grid without as many renewables. Could you point me to an article explaining your theory in more detail, with evidence?
One alternate hypothesis might be that renewables are cheap and profitable, with low capital requirements. In a place like Texas, with lots of sun and wind, it might be tempting to build more renewables than the grid could support. Combined with weak regulation, this would risk destabilizing the grid.
In a complex system like an electrical grid, failures can occur for all sorts of complicated reasons. I'd be interested in understanding what happened here, based on evidence.
1) Power disruptions in fabs are very expensive.
2) Texas's power grid has gained a reputation lately for being unreliable.
If those things are true, how do they plan to deal with that? Heavy investment in onsite power generation?