Turning off electrical grids to prevent wildfire− a complex, technical operation

https://theconversation.com/colorado-is-latest-state-to-try-turning-off-the-electrical-grid-to-prevent-wildfires-a-complex-technical-operation-pioneered-in-california-227639

by rntn • 2 years ago

45 40 2 years ago

theconversation.com

puzzledobserver• 2 years ago

The Practical Engineering Youtube channel has a very interesting video on black starts of an electrical grid [0]. Are there any additional challenges in turning a grid off?

[0] https://www.youtube.com/watch?v=uOSnQM1Zu4w

idiotsecant• 2 years ago

There's a pretty big challenge, which is that you aren't turning off a grid at all -just particular transmission lines. Your typical power flows rely on those lines being up, and without them strange conditions can occur. Its sometimes just as bad to have an unusually lightly loaded line as it is to have an overloaded line. Lightly loaded lines with lots of generation end up messing up line voltages due to capacitive loading. You can end up having to shut down a line or taking other remedial action like shutting down generation because not enough power is flowing through a line.

ckdarby• 2 years ago

I suspect we'll see more openness to this with homes in the future coming with batteries to power for 12-24 hours without the grid.

robocat4• 2 years ago

In my experience 12-24 hours won't help you enough in a city, and is woefully inadequate in rural areas.

There are better ways to spend your money to cover risks than batteries. Maybe if you were getting a battery for other reasons, or as a luxury if you are wealthy enough.

toast05• 2 years ago

I think if you have a big enough time of use differential, and you have significant use during peak pricing, you can justify a local battery system for peak shaving and get backup power 'for free' or vice versa.

12-24 hours would have outlasted any outage I've experienced, until I moved to my current place, but it has a propane fueled whole house generator (and low utility rates with no time of use component; i haven't attempted the math, but I'm pretty sure running from my generator is way more expensive than utility power... OTOH, I've had many utility failures and only one generator failure (12v battery failed due to charger failure))

mjevans• 2 years ago

Nearby site generator? Diesel or Natural Gas, depending on reliability during power outages?

askvictor• 2 years ago

Depends what you need to do with that. Sure, you can't 'live normally' on battery power for too long, but if you prioritise things, your battery can last for a much longer time. Fridge/freezer to prevent spoilage, phone/laptop charger for comms, a couple of lights, a microwave oven. Heating/cooling is tricky as that's a _big_ draw, but there are ways to be efficient with that too.

mschuster913• 2 years ago

Maybe if you were getting a battery for other reasons, or as a luxury if you are wealthy enough.

Batteries are cheap. A 5 kWh battery storage pack can be had for ~1500€ these days, that's enough to ride at least the fridge, freezer and LED lighting for a day or two through a major event. In rural areas, add a small 1-2 kW generator and a transfer switch for another few hundred euros, and now you're completely set even for larger outages.

TechnoTopian• 2 years ago

We already have those batteries: Electric Vehicles. We just need some kind of vehicle-to-grid standardization.

bagels• 2 years ago

What happens when the power goes out when your car is not charged, or if you want to drive somewhere during a power outage?

amluto2• 2 years ago

I would like to know why grid operators turn portions off instead of installing devices to mitigate the fire risk when a fault occurs. “Rapid Earth Fault CLearance” devices (REFCL), also known as ground fault neutralizers, are a newish but commercially available technology that massively reduces current, and thus heating energy, delivered into a ground fault. In testing in Australia, a REFCL-protected line was intentionally dropped into dry plants without setting them on fire.

What gives? I doubt these devices are so expensive that they’re worse than the cost of “public safety power shutdowns” or of the billions of dollars that utilities are spending, and charging customers for, for undergrounding of lines and extra tree work.

mschuster913• 2 years ago

I would like to know why grid operators turn portions off instead of installing devices to mitigate the fire risk when a fault occurs.

The stuff you're mentioning is - at least from a quick Google, I'll admit that - only certified in Australia. Even here in Germany, introducing that onto the grid would take years for certifications. Bureaucracy moves slowly.

Then, you'd need to upgrade a lot of transmission lines including the regional distribution grid (i.e. < 100 kV), a single unprotected line in a danger-prone area is enough to start a deadly fire. And there are a looooot of power lines that haven't had an inspection for decades - the hook that caused the 2018 fire was almost a century old [1]. And so, instead of inspecting, maintaining and upgrading the many thousand km's of national and regional lines, shutting them down during peak fire condition times is cheaper.

IMHO it only makes sense to install REFCL devices as part of a general overhaul of an entire transmission grid area.

[1] https://www.wsj.com/articles/this-old-metal-hook-could-deter...

toast05• 2 years ago

Because they already have the ability to turn portions off. Installing a lot of new equipment means regulatory effort to get a plan approved, and a lot of installation labor.

I imagine these systems will be put in place over time, but I was just looking at Wikipedia on Reclosers and they said Sensitive Ground Fault protection reduces the risk by 80%; if the risk is high enough, that probably still warrants deenergizing risky lines at times.

amluto2• 2 years ago

These aren’t quite the same thing. Sensitive ground fault protection attempts to shut down the line more quickly in the event of a ground fault. A ground fault neutralizer attempts (and apparently largely succeeds!) at reducing the fault current to near zero during a fault even before anything is shut down.

These technologies can be combined.

samtho• 2 years ago

Electrical grids are massive, like almost unbelievably massive. Ownership is complex in that segments of it are controlled by different parties/companies with some of those segments having shared ownership. There only three grids in the USA and (I think) five total in North American with every device operating on a given grid have the capability to play by the grid’s rules and idiosyncrasies. Grid operators are constantly testing new equipment but it gets tested within sandboxes for years prior to adoption. Even if testing goes well, you still need individual buy-in from equipment operators who are allergic to risk and hate change.

I am not trying to say this is the right or wrong way here, just kind of the reality of operating such a complex network.

bunabhucan3• 2 years ago

We had a fire near Boulder CO that destroyed 1000 homes and had a few fatalities. If there are 100mph winds and dry conditions then it just takes one spark to create an unstoppable fire.

The local utility closed down some (but not all) lines for a recent wind event and everyone is kvetching.

Maybe this isn't obvious but domestic electric service is a marginal business. It's not profitable for every customer. That's why it's a regulated monopoly. In edge and rural areas the lines are not profitable (you can't charge more to joe farmer at the end of hundreds of poles.) New subdivisions with new poles/transformers etc. cost less

Windswept lines and poles out in the middle of nowhere are the places at risk and "make no money." The utility doesn't "pay" for the shutdowns but they can get sued for a fire.

sanderjd6• 2 years ago

The thing that was most unclear with this incident in Colorado was: How did they decide which areas to shut down? It didn't make me feel like the people in charge knew what they were doing, when we left our neighborhood where the power was intentionally turned off (for over a day) to go to our friends' house where it was kept on, despite them having much stronger winds both in the forecast and in actuality. Their house is also in an area with much higher fire risk, surrounded by dry windswept vegetation. It's hard to trust an organization that says "we're turning off your power due to your area's unique risk of fire" when that conclusion seems incomprehensible to anyone aware of the weather and geography of the area. It's very hard not to wonder whether they are ignorant or are not being forthright. Maybe there is some information that would show that they made the right decisions, but if so, they certainly haven't shared it yet!

zmj2• 2 years ago

Agreed this was confusing. It seems like the risk factors that would lead to turning off a neighborhood (non-buried power lines, exposure to wind and available fuels) should have been known and announced in advance.

I'm in Lafayette, and it seemed strange that our power was on when neighborhoods in Louisville and Erie were out.

sanderjd6• 2 years ago

Yep, and pretty much the entirety of Boulder preemptively turned off, but not the areas south along 93 and to the east into Arvada near Rocky Flats, which are the exact super dry and windy areas where the Marshall Fire began.

I actually do think there must be some plausible explanation, but it's hard not to feel sympathetic to the people who are convinced it is a tit for tat by Xcel, targeted at people who have been involved in recent legal confrontations with them. There are good reasons they aren't being given the benefit of the doubt here.

I honestly worry more that truth is they were just winging it without any strong process or data with which to make the decision, and that this is their plan moving forward.

But the good news is that the PUC investigating what happened is the way to get these answers, and is exactly what Governor Polis is pushing for.

michaelrpeskin3• 2 years ago

My power wasn't turned off - I did get the "we might have to turn your power off" warning though.

I live in probably one of the worst areas for wind, and had gusts to 100 that Saturday night - didn't get any sleep!

I had studied the local grid a bit for some other reasons a while ago, here's how I would put the pieces together:

My neighborhood is all buried lines, and everything is buried starting right near our substation. No raised distribution lines to worry about. However, it's still not really distribution lines that are going to cause wildfires, it's the transmission lines that go through the dry forest. Wind taking out a distribution line isn't that big of a deal, and it wouldn't cause much downstream fire risk.

My co-worker who lives about a half mile from me was without power for nearly 2 days. In other power outages, we've compared notes. Even though they live closer to my substation than I do, they're on a different substation.

The transmission line that powers their substation goes right through the mountains. It's really close to lots of dry fuel and the towers are very far apart, so it would be easy to get those lines to swing in the wind. Additionally, the way the winds blow, if their lines did arc and cause a fire, the wind would sweep it right into town.

So you really don't want to look at "why was my neighborhood without power, we didn't get much wind." You want to look at "where does my substation get it's power, and would an arc on that transmission line be catastrophic".

At least in my immediate area, the substations that were shut down were exactly the ones I would have shut down just by looking at a map and saying "if it's windy and dry here, do I want an energized line?".

I don't know about the decisions in the wider Front Range area, but I would be willing to bet that 1) it's a pain to de- (and re-) energize lines, so they don't do it without weighing the risk, and 2) much of our transmission lines out here are in the mountains to the west, which is terrible when you have westerly winds that will blow the fire into town, so they have a high risk weighting.

sanderjd6• 2 years ago

Great comment!

I'll look this up myself too, but do you have things I could read on the relative risk level of transmission vs distribution lines? They seem risky in different ways to me.

At least in the case of what (possibly) started the Marshall Fire, I thought that seemed like a distribution line that was knocked down, not a transmission line that arced. But that could be wrong (and of course it's disputed that it was a cause of the fire at all).

At least in my immediate area, the substations that were shut down were exactly the ones I would have shut down

First of all, how did you determine which ones were shut down? A big part of the problem, in my view, is that information of the form "this is what we're shutting down, and why we think it's high risk" did not seem to be forthcoming anywhere. To put it more broadly and with a finer point to it: there was no communication and nobody, including, apparently, the municipal emergency services organizations, had any idea what was going on, or why.

But secondly, from cobbling together information from their website, gossip, social media, and driving around that day, my impression was the opposite of yours, for my surrounding area. My impression was that the pattern was extremely mystifying.

But sure, maybe the communication issue is the first order problem to solve. I certainly hope you're right that they based the decisions on a good risk analysis. But if so, they didn't tell people what that was. And lacking information on that, few will be willing to give them the benefit of the doubt that you're giving them on this, even if they deserve it.

And I don't just mean communication in the period of time just before during and after the storm. Did I ignore an email from them in the past few months announcing that they are planning on incorporating emergency preemptive shutoffs during this fire season? I don't think so! (It's also weird to me that they seemingly didn't need to make the PUC aware of this ahead of time, with an announcement like that being required as part of approving that plan; it isn't only Xcel that I think messed up here.)

It's easy for monopoly service providers to come off as arrogant and aloof, and lose peoples' trust!

bunabhucan3• 2 years ago

Your assumption is that the outage is about the houses that are dark and not some line of silent poles/towers stretching beyond where the people are. The N Boulder outage was probably for rural lines N of Boulder rather than any neighborhood risk.

I work in GIS for utilities and know where about a mile of my feeder comes from but I can't figure out where exactly it goes after that.

Somebody probably overlaid a few maps (weather forecast, estimated fire risk, electric lines) and picked feeders based on that.

sanderjd6• 2 years ago

That wasn't my assumption, for what it's worth. I understand that it's about the lines not the end points. It's a good point though, that the lines they worried about the risk of might be nowhere nearby.

I suspect you're right about what the process was. I think a more succinct version of my view is this: I think we shouldn't need to speculate about this, we should have the details of this process available to us. I understand that info about specific lines might be sensitive, but we should know what the process is for this.

ThinkBeat• 2 years ago

There is a bad snowstorm coming in folks. Snow, high winds and temperatures best thought about in Kelvins. Just joking with ya-all. Looks like we will be snowed in for a week or so. To keep ya-all safe, the utility will be turning off the grid. Remember to stay warm.

zmj2• 2 years ago

That's actually a real concern for using this strategy in Colorado in the fall. The Marshall Fire (mentioned in the article) was followed by snow and temperatures well below freezing. Sadly, this probably killed some pets that escaped the fire.

bunabhucan3• 2 years ago

Something that might not be obvious: the shape of feeders can change based on switching operations. New PV on the big expensive house next to open space can create problems (you now have a generator right at the end of the grid) and new EVs can be roughly as much demand as a new house. Sometimes the fix is to change what feeds what, especially in a dense city. The utility obviously knows what feeder every meter is on but may not have anticipated (or paid for!) bulk calling the exact customers based on that. CA had rolling outages (caused by Enron) and implemented a system for notification but it was work, took years to design and put in place.

The systems for managing switching and restoring power are quick/live/critical whereas the systems for building / maintaining are more slow / archival / non-critical. Both are typically legacy systems mired in an interconnected web with other legacy systems and very, very slow to change. Anything that looks like a bulk customer / phone list api is even slower.

evilotto• 2 years ago

Why is it that the costs "trickle down to the electricity consumers" while utilities like PG&E report record profits? Are profits more important than safety, or customers?

kkfx• 2 years ago

Ehm... Sorry, no. To prevent wildfires from the grid you need to make most cable underground, where you can't keep them in a clear area, large enough to avoid lighting up a fire if a pole drop.

Turning off electrical grids to prevent wildfire− a complex, technical operation

Related Stories