[magnawave]

The result was "an electrical island in the United States," Bill Magness, CEO of ERCOT, said. "That independence has been jealously guarded, I think both by policy makers and the industry."

https://www.statesman.com/story/news/2021/02/16/texas-power-...

Maybe it’s time to rethink that. HaI has an interesting take on something similar: Japan https://www.youtube.com/watch?v=Mo88zA5nq4Q

Running your own grid seems cool Texas style, until you have a regional problem and have no where to turn.

[txlpo78]

This current problems wouldn’t be fixed if Texas wasn’t on its own grid. The Texas grid does have connections to the other grids, and even right now is importing power from both the East and West interconnections.

The problem is that this is a truly regional event and not just isolated to Texas. The entire central US is struggling right now. The SPP (which manages electricity for Oklahoma, Nebraska, Arkansas, and other states) has been struggling with forced blackouts over the last several days as well. They don’t have enough power for their own grid, let alone enough to share with Texas.

If Texas was more interconnected with the SPP, the end result wouldn’t be Texans all having their problems solved. Many Texans would still be without power, but so would many more Oklahomans. The fact that the Texas grid is separate is the only thing keeping OK from having even worse blackouts. Which makes sense, because the entire point of grid isolation is to keep issues localized and not cascade over the entire network. And that’s working to Oklahoma’s benefit right now, but Texas is getting the short end of the stick.

[mike_d]

This is not correct. Because the Texas grid is isolated, the frequency is not synced with the two other major grids and cannot import electricity at any meaningful capacity. (See http://fnetpublic.utk.edu/frequencymap.html)

Frequency conversion is a costly and difficult to scale problem. If Texas was part of the Western grid they could be drawing excess hydroelectric power from the pacific northwest right now for example. Texas also could have contributed to help the California power shortages last year.

Edit: Here is a map of the grid interconnects in Texas with capacity. As of the time of this comment the total importation capacity is less than 1% of demand. https://user-images.githubusercontent.com/22095643/49019079-...

[txlpo78]

Nothing I said in my comment is incorrect. Texas has 5 different connections with the other grids and can import/export through them. But they are irrelevant right now because the other grids do not have enough spare capacity to send to Texas.

>If Texas was part of the Western grid they could be drawing excess hydroelectric power from the pacific northwest right now for example. Texas also could have contributed to help the California power shortages last year.

No. That’s not how the grids work. Just because Oklahoma and Washington are part of the same interconnection, that does not mean that people living in Tulsa can pull power as needed from a dam in Washington, which is why Oklahomans are struggling with power outages today as well. Most power still must be generated locally. Long distance transmission is difficult and inefficient, and often requires converting to DC just like a grid-to-grid connection requires, so you have the same issues as you have when you’re on separate grids.

[reportingsjr]

This does not seem to be true at all. If you look at a power outage map of Texas you can actually see exactly where the ERCOT boundaries are. Everyone else in Texas that's on the other, federal, grids, are not experiencing widespread power outages.

https://poweroutage.us/area/state/texas http://www.ercot.com/content/wcm/landing_pages/89373/ERCOT-I... https://poweroutage.us/area/state/oklahoma

Per your comment about long distance transmission, that doesn't matter in a situation like this. If you're on a large grid you don't necessarily need to transmit power to Oklahoma all of the way from the PNW.

You need the areas surrounding OK to supply excess power to them, then those surrounding areas can get whatever excess they may need from slightly further areas. This needs less and less excess as you go further since every area is over provisioned.

Eventually at some point, yes, the PNW may be supplying excess power to states around them as a result of Oklahoma having outages, but that power isn't going straight from PNW to OK.

[txlpo78]

https://www.kmbc.com/article/southwest-power-pool-again-orde...

Oklahoma has been dealing with rolling blackouts for the past several days. Tell me why this is, since apparently you think Oklahoma is able to magically get power transferred to them all the way from Washington? If WA has the excess capacity, why are Oklahomans still without power?

> Everyone else in Texas that's on the other, federal, grids, are not experiencing widespread power outages.

Completely wrong. Eastern Texas (eg Orange), which is under MISO, and is dealing with blackouts. And parts of the Texas panhandle like Lubbock, which is also not part of the Texas grid, is also struggling with power outages.

[jeffbee]

I'm sure you can see that a rolling outage affecting 200k people for 4 hours is quite different than an outage affecting four million customers for 3 days.

Check out the map. It's pretty clear that what you said is wrong. ERCOT territory is all broken, panhandle, east Texas, and El Paso area are not having problems. https://poweroutage.us/area/state/texas

[Element_]

I am not sure it's that simple, I read some where else yesterday that at least one of the non-ERCOT grids had paid to winterize their local power plants after the last ice storm so their plants have been operational throughout this storm and as a result had no outages. I have no idea how accurate that is though, I don't know anything about the electric grid...

[briffle]

There is a 3.6GW DC line that goes from about an hour East of Portland, Oregon, down to LA. Its 2 wires. Texas doesn't have any interconnections with the west. But even if they did, 3GW would not be nearly enough to solve their problem could could replace many natural gas plants that are currently down.

[blake1]

AC transmission absolutely does work over large distances. It’s just not a point-to-point system.

Imagine four cities in a row, all connected with AC. City A generates extra power, which gets sucked up by city B, whose power goes to the next city down the line, to city Z.

Sure, it’s not actually that simple, but when was the last time NY literally had no power? They benefit from being highly connected.

TX is paying for being isolated.

Their handful of DC interconnects do not have the capacity to power their mini grid. They’re short 35GW of generation, and I assume the DC ties are at capacity.

[txlpo78]

> when was the last time NY literally had no power? They benefit from being highly connected.

Do you not remember the blackouts of 2003? Multiple entire states went dark for hours, and the “highly connectedness” was a huge part of the problem. The only reason it wasn’t even worse is specifically because grid isolation stopped it from propagating further, just like what’s happening here.

[blake1]

Just look at the first image on the Wikipedia page, showing the extent of the blackout. This is far smaller than the footprint of the eastern interconnect. The control software at the time made some simplifying approximations which left the grid vulnerable to problems cascading between operators. I do not think they have quite the same problem anymore.

https://en.m.wikipedia.org/wiki/Northeast_blackout_of_2003

[pixl97]

There is another problem here. Texas generates as much power as the second and third states on the generation list. The outage Texas had would have sagged most of the US.

Separate grids are great for other kinds of emergencies, if we get a big solar flare then the splits save each grid.

We need a east to west DC long distance interconnect to haul power across the country.

[mike_d]

> that does not mean that people living in Tulsa can pull power as needed from a dam in Washington

Not directly, but by way of demand shifting, effectively yes. Northern California is fed by Washington, SoCal by NorCal generation, etc. until you get excess capacity closer to the demand sink.

[mjevans]

You do still need generation closer to Texas that works, but as a whole the grid can balance the generation and output a bit depending on the gradient between the sources.

Imagine a 'bouncy castle' with several input fans. Texas is like an entry ramp that isn't hooked up to either of two big banks of fans and sinks next to it. If it were just ganged in with one of those other two groups even though Texas is having a bad time the other blowers could compensate in aggregate.

[oasisbob]

A good example of this is the PNW interties which move (primarily) hydro power to California from Washington and Oregon.

One system is DC. The other is AC. They both do primarily the same thing through elaborate systems.

Washington can send electricity south via AC - doesn't really mean CA and WA are functioning on the same grid.

[aardvarkr]

FYI the Texas grid runs on a different frequency than the other two grids and as a result incurs MASSIVE efficiency penalties for that hubris. If I recall correctly it has to be converted from AC to DC then back to Texas’ AC.

[O5vYtytb]

Technically they run on at different phase, but the same frequency (60hz)

[quickthrowman]

You recall correctly, and interestingly, the same method is used for variable-frequency drive motor controllers

[txlpo78]

Yes, that’s correct. Texas has 5 ties to the western/eastern US interconnections as well as with the Mexico grid. But none of that matters right now because those grids don’t have excess capacity to send to Texas anyway.

[teraflop]

It's not really a matter of whether the grids have capacity; the ties themselves can only handle a limited amount of power.

As per ERCOT's status page, both of the high-voltage DC ties between Texas are currently operating at >99% of their rated capacity, and they have been every time I've checked since yesterday. They're not being limited by the availability of power from the other side.

[txlpo78]

You’re missing the point. Even if the ties had more capacity, the supply of power on the other side of the ties is not there. It’s a two-pronged issue, and you won’t solve the problem by only focusing on one of the prongs.

[blake1]

This is incorrect. MISO, the system to the north and east of TX, has capacity. The DC ties cannot handle it. You can see this by checking the price signals on their page. Right now, the TX hub is about $1,000 but the MS hub is about $60.

https://api.misoenergy.org/MISORTWD/lmpcontourmap.html

[margalabargala]

It seems highly unlikely that exactly 100% of the capacity of the interconnects is, coincidentally, precisely equal to the amount of excess power available to be fed into the interconnects at the moment. Do you have anything to back up this extraordinary claim?

[ineedasername]

If you browse a few of the ISO pages for other states/regions you'll see a bunch that have excess capacity above their projected peak for the day.

Here's one: https://www.iso-ne.com/

Here's another: http://www.caiso.com/TodaysOutlook/Pages/default.aspx

But there's a limited capacity for Texas to bring in power over just 5 connections, combined with their choice that makes conversion to something compatible with the Texas grid much less efficient.

By isolating & not focusing on compatibility they have made it very difficult to have more robust redundancy in their grid.

[s0rce]

Do the other grids publish their generation/demand statistics live? I remember being able to check the CalISO page during the rolling blackouts in California during the summer.

[stefan_]

They have absolutely 0MW capacity that they would be willing to put online to earn millions an hour? Seems unlikely.

[7thaccount]

I tried to say the same above. When the entire south is low on capacity, being fully AC interconnected probably doesn't help much even if the DC ties had more capacity.

[ineedasername]

A few limited connections are not sufficient to fully take advantage of all areas of the country where there may be extra capacity. It wouldn't solve the problem, but would have helped. Oklahoma for example has been able to stop rolling blackouts. Texas is still deep in this.

[austinheap]

This is a two day old account repeating exactly one trope FWIW.

[ineedasername]

Yes, basically Texas is great, this situation was unavoidable, nobody did anything wrong, and people should just suck it up.

[pkulak]

How hard can it be to have your own grid, but also have inputs on the edges to draw from? Or would that no longer be independent?

[teraflop]

Texas does have connections to other grids. But the problem is that since the grid frequencies aren't synchronized, you can't just plug one into the other.

Transferring energy between grids requires either converting it from AC to high-voltage DC and back using solid-state electronics, or converting it via mechanical energy using a variable-frequency transformer. With either approach, you need bulky and expensive equipment in proportion to how much power you want to handle. These connections are designed to smooth out (and profit from) short-term capacity fluctuations, not to power the entire state.

Currently, the Texas grid has two DC ties operating at full capacity and drawing about 800MW from the Southwest Power Pool. But that's a drop in the bucket compared to the ~45GW of current demand, or the estimated 70-80GW of demand that would be likely if it weren't for the outages.

http://www.ercot.com/content/cdr/html/real_time_system_condi...

[pkulak]

Thank you! I didn't think about the AC synchronization.

[theshrike79]

Electricity networks are a really fascinating hole to dive in. I did software and firmware for Smart Metering solutions for a half decade and know more about that stuff than I'll ever need.

[martinald]

TX grid is connected via HVDC, but keep in mind that even if it was "fully" connected there is no hope in hell it would be much better. There wouldn't be enough transmission to transfer 40GW+ from the East or West coast grids to TX. It's an enormous amount of power to go offline. I don't think California for instance has more than 10GW of transmission north to south.

Basically, no amount of grid infrastructure can really help you much when you lose 50%+ of generation capacity on your highest demand days in history.

[JamesBarney]

If your gap is 40GW, and you could get another 8GW. That's enough to give everyone an extra hour of electricity every 5 which would be game changing. The different between a 33 and 45 degree house is enormous.

[s0rce]

Seems like CA might has 15-20GW of spare NatGas generation capacity. Looking back at the summer peak (9/6/20) I see 25GW generated at the peak, while we are currently only needing to generate about 8-9GW from NatGas (looking over night when solar drops off).

[snowwindwaves]

The parent said there is not enough transmission capacity not generation capacity. Generation needs to be near loads. If not then transmission lines are required to bring current from generators to loads. Transmission lines have fixed capacities, like your internet connection can only transmit so much data/s, they can only transmit so much power.

[s0rce]

I thought they were saying there wasn't enough transmission capacity but it wouldn't matter anyways since the rest of the grid didn't have any to spare. Maybe I misinterpreted or responded to the wrong comment.

[8note]

It's likely that the transmission lines are less of a bottle neck than the frequency conversion. Power lines are pretty darned efficient, and given the cold weather, theyd be more efficient than usual too

[snowwindwaves]

Electrical engineers wouldn't build a power line to Texas with a capacity of 5000 MW and then put intertie hardware such as phase shifters or ac-dc-ac converter with capacity of 1000 MW on the end of it. So it is likely that the transmission capacity and intertie capacity are exactly the same!

[pixl97]

They make long distance DC innerconnects, while expensive they are resistant to particular types of problems and can connect to multiple different grids at the same time.

[bsder]

That actually isn't true:

https://3dfs.com/articles/wasted-electricity-vs-lost-electri...

Almost 62% of electricity is lost in the grid. AC is the primary reason (matching, vibration, I2R, etc.).

HVDC is WAY better at transmitting power over any appreciable distance.

[autoditype]

From the articles I've read they do:

> Even today, ERCOT is also not completely isolated from other grids — as was evident when the state imported some power from Mexico during the rolling blackouts of 2011. ERCOT has three ties to Mexico and — as an outcome of the "Midnight Connection" battle — it also has two ties to the eastern U.S. grid, though they do not trigger federal regulation for ERCOT. All can move power commercially as well as be used in emergencies, according to ERCOT spokeswoman Dottie Roark. A possible sixth interconnection project, in Rusk County, is being studied, and another ambitious proposal, called Tres Amigas, would link the three big U.S. grids together in New Mexico, though Texas' top utility regulator has shown little enthusiasm for participating.

https://www.kvue.com/article/weather/texplainer-why-does-tex...

[gitgreen]

The latter. They do not want federal regulation.

For the pedantic TX does have connections to other grids but the capacity is so low that it wouldn't have mitigated this event and is under whatever threshold is set for federal oversight.

[afavour]

It would be independent from a practical point of view but from a regulatory one it would become subject to federal regulations, which is the real reason Texas keeps theirs separate.

[hangonhn]

The Tres Amigas Super Station is, ironically, being built in TX. If it was done, it could help the situation a lot. 30 GW of capacity!

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

[reportingsjr]

Nit: IF this gets built (big if), it will be in New Mexico, not Texas.

It's also very unlikely that it would be near 30GW of transmission capacity.

[wnevets]

Regulations probably.

[cat199]

> Running your own grid seems cool Texas style, until you have a regional problem and have no where to turn.

This cuts both ways, and most of the rest country is often a benefactor - many US/NA companies (and I'm assuming govt/military) have a 3rd DR location on the TX grid precisely because it provides an additional point of redundancy. TX also wasn't impacted by CA mismanagement of it's grid, for example..

[booi]

Right, TX is just impacted by its own mismanagement of its grid.

Most companies that have and need DR put them in locations with independent power, not based on which grid they are in.

[unethical_ban]

I echo the sentiment that federation (in most things) isn't itself a bad thing. It just happens that in the case of electrical reliability, ERCOT has been the root of these issues, given their warnings from years ago.

[throwaway667555]

No, we shouldn't rethink independence. We should build back better.