Yves here. Aside from the shaky state of the electrical grid being of interest to readers in potentially or actually afflicted areas of the US, this story makes clear that loss of power is not just due to the “shit happens” of above-ground power lines meeting ice storms that break tree branches which then sever cables or damage transformers. It describes how cold snaps produce a surge in demand which systems that have suffered from decades of underinvestment cannot handle well. Another part of the problem is underinvestment in maintenance and understocking of transformers in utility inventories.
And the US grid is in decrepit state even as tech bros are rapidly ramping up demand to feed environmentally-destructive and intended-to-be-job-eating AI.
By Andrew Topf, a seasoned writer specializing in natural resource commodities who also has expertise in the oilfield services and heavy construction industries. Originally published at OilPrice
A winter wallop that delivered a blast of Arctic air to two-thirds of the United States last weekend brought with it power failures, partially due to snow, ice and falling trees or branches damaging power lines.
At least 35 deaths were reported, including three brothers between the ages of 6 and 9 who fell through an iced-over pond in Texas, and two people who were run over by snowplows in Massachusetts and Ohio.
The mega-storm reportedly caused blackouts to more than a million customers, especially in Tennessee, Mississippi and Louisiana.
As of Tuesday morning there were still 550,000 outages across the country, according to poweroutage.com, via CBC News, with most of them in the South, where weekend blasts of freezing rain caused tree limbs and power lines to snap, inflicting crippling outages on northern Mississippi and parts of Tennessee.
Officials warned that it could take days for power to be restored.
The largest regional grid, PJM Interconnection, serving 67 million people in the East and Mid-Atlantic, saw electricity prices soar above $3,000 per megawatt hour Saturday morning from earlier levels of less than $200 per MWh (Reuters)
On ISO New England, the grid for six states, spot prices hit nearly $600 per MWh, up sharply from Friday when they were below $100/MWh.
Reuters said that due to constricted gas supplies, regional grid operators asked coal and gas-fired power plants to boost output.
The winter storm was the biggest test for Texas since 2021, when an icy blast nearly caused a catastrophic regional blackout. More than 200 people died when the Electric Reliability Council of Texas lost about half of its generation capacity.
Reuters says Since then, stricter state and federal rules have been implemented to require better winter readiness by utilities and grid operators throughout the country.
Clearly though, it’s not enough. Power, water, sewer, and road/bridge infrastructure have been crumbling across the US for years. In 2025, the American Society of Civil Engineers Infrastructure Report Card graded the US electrical grid with a D+, down from a C- in 2021.
According to Critical Energy Infrastructure Services (CEIS), ASCE’s report cited a shortage of distribution transformers, severe weather events and lack of transmission capacity as some of the reasons for the low grade. The organization predicts the demand for electricity will grow from the 17 GW in 2022 to more than 35 GW in 2030.
It’s an alarm that has been raised for years, but the investment needed to bring everything up to date has not met the increasing demand.
Most power lines are over 30 years old and are designed to handle less electrical load. Almost 70% of transformers are more than 25 years old. The US experiences more outages than most developed countries.
An NBC News review of federal and state records, and internal utility documents, found that in five of the past 11 years, from 2024, parts of the US electrical grid have been hit by blackouts, shut-offs or close calls during cold weather.
CEIS notes the average home in 1970 used under 100 kwh of electricity annually. The service drop for most homes was at 100 amps. In 2025, the average home uses more than 10,000 kwh per year and the standard service drop is 200 amps.
The ASCE report says it would take at least $700 billion to modernize the grid.
On Thursday, Jan. 29, congestion on the PJM grid was expected to soar due to forecasted record winter power demand over the upcoming weekend. This again has led to a sharp rise in spot wholesale electricity prices, which have gone past $1,000 per megawatt this week.
PJM reportedly issued five warnings to utilities about potential power cuts related to transmission line problems. Reuters said Restricted natural gas supplies to power plants have also been a major factor in causing generator outages and boosting spot prices, according to analysts and PJM data.
PJM… forecasts more than 1,400 high-voltage lines would have restrictions on Thursday with disruptions averaging nearly 13 hours….
Before heavy snow and frigid temperatures hit the eastern U.S. last week, congestion on PJM’s transmission network affected only about 60 power lines with constraints lasting an average of about 4 hours, PJM data show…
PJM predicts electricity demand will rise to 148 gigawatts on Friday, setting an all-time record for winter.
Why is the US power grid so vulnerable to cold snaps? Those who have looked at this problem say the main culprit is natural gas dependency.
According to NBC News, natural gas since 2016 has become the dominant source of power for the US electrical grid, since it’s cheaper than coal and more efficient. In 2024 natural gas provided 43% of the country’s electricity.
But cold-weather events are putting multiple stresses on the gas supply. The stresses threaten 50% or more of US homes that are heated with the fossil fuel.
Here’s how it works: Winter storms create an imbalance between the supply of electricity and the demand for it. When that happens, grid operators are forced to take some electricity demand offline to prevent a larger outage.
Another problem is that many regional electric systems are not connected with other systems, meaning when one system is overwhelmed, other utilities can’t help out.
In extreme cold, says the World Resources Institute, one of the biggest challenges is getting fuel to power plants for electricity generation and to homes for heating…
During winter storms, demand for natural gas spikes as people turn up thermostats, while power plants also need the fuel to generate electricity. This simultaneous surge, known as coincident peak, is a well-known planning challenge. Extreme cold can strain the natural gas system beyond normal operating conditions, making it harder for gas-powered generators to access fuel when they need it the most. Similar stress can happen during extreme heat, when air conditioners drive up electricity demand.
Another issue is lack of proper winterization — preparing equipment for wintry conditions — across different parts of the electricity system. Equipment that isn’t prepared for extreme cold, such as power plants, sensors and natural gas wells and pipelines, can freeze or malfunction. When multiple pieces fail at once, utilities may have no choice but to implement widespread outages to maintain system safety and prevent cascading failures.
The Federal Energy Regulatory Commission (FERC) expects natural gas to increase its contribution to the electrical grid’s power plants by several percentage points in the next decade.
To ensure a reliable source of gas, FERC recommends that utilities invest in winterizing the natural gas system, lawmakers establish tough reliability standards, and gas companies and utilities improve their cooperation, NBC News says.
Other steps that could be taken to improve the electrical grid include: smart grid technology; federal, state and private investment; improving transmission networks; and Distributed Energy Resources (DER), which encourages localized energy production to take strain off the grid. Examples include rooftop solar and the creation of small, independent grids that service facilities such as hospitals and military bases.
28 comments
Comments are closed.
OK, I don’t understand the explanation given in paragraphs after the “According to NBC News” and this particular paragraph:
If a winter storm creates an imbalance my guess it is because some or much of the heating is electrical,not NG fuelled.
Then it is not the former but gas supply constraints? NG supply constraints pushed homes/buildings equipped with NG boilers to use electrical means for heating (instead or besides NG), and then this caused electric imbalances? I found the explanation messy.
On the positive side it seems that the high voltage lines where able to stand the mechanical forces created by ice and wind in such severe conditions which means good engineering calculations were made in the past taking in account the possibility of such adverse events. Quite a pain in the neck those calculations!
If you have a gas fired or oil fired boiler (or furnace), it requires electricity to run. No electricity, no heat.
Well, yes but that doesn’t contribute significantly to power peaks, it is more or less constant and only adds a few milliampere no matter how cold it is outside..
I have a high efficiency gas furnace. It has two electric motors. One motor for circulating the hot air through the house, the other motor sucks air in from the outside for burning gas and then exhausting the combustion products to the outside. This is equivalent to running a couple of room air conditioners all day.
Those furnaces do not require a lot of electricity. They could be kept in operation with a generator or batteries.
It seems the transition from NG and other direct home energy sources to electricity for home heating in the US is going quite well. Its a very welcome trend, especially if its paired with domestic solar panels.
The issue of course with extreme weather and infrastructure is that all infrastructure is designed within certain local/regional tolerances depending on historical patterns – when those patterns change rapidly, thats when you run into deep trouble, and this is what we are seeing worldwide.
Of course, there are also different ‘types’ of extreme weather events with different impacts. We’ve had appalling weather the past 2 weeks in Ireland with lots of rain and flooding, but little impact on infrastructure – if anything, the storminess has led to consistent high levels of wind energy and reduced gas usage. If the predicted cold snap arrives later in the month it could be a very different story (unless it comes with plenty of wind too).
Such increase on electricity usage for heating from 35% to 43% might indeed explain, at least in part, the imbalances seen during the arctic blast. And probably such transition might have been steeper in some Southern states.
Oh, yes! In Spain we are having quite an “Irish” weather too! (if you don’t mind) Windy and quite wet since the end of January. With wind energy reaching nearly 50% of generation for several days. Only one day, because an excess of wind, (many wind turbines had to stop) we had some risk of blackout. Portugal has had some trouble. The region around Porto (Oporto) must be seeing record rains.
My approach to solar-updating an older (as in can’t be insulated retrospectively to a high standard) USA house is to use the gas if you have it. The batteries can run the 120V electronics and electrics (1/3hp air handler motors), fridge etc because they are low power. Gas heat is useful in winter because it’s needed at night when the solar isn’t generating. AC use is matched to solar generation much better. Also, it’s very rare for gas to get cut off, so the battery storage works well during extended power cuts, when the tradeoff is mostly how much cooling to do during hot times.
A word of warning about heat pumps: they need a supplemental heat source for regions that go much below freezing, and they work best maintaining a set point temperature. They are unhappy to bring a house up quickly in the morning if you set a lower temperature overnight.
Note that in the US, wood pole construction for transmission is very common for “load serving” network voltages under 200 kv (perhaps the majority) and it is also present for 230 and 345 kv “bulk power” networks. It is very susceptible ice failure. More recent construction is steel Poles.
Ice on transmission lines can be disastrous. In January 1998 a multi-day ice storm in Eastern Ontario, Southern Quebec, the Maritimes, Maine and Northern New York State wiped out all transmission, including thousands of steel lattice towers. There was asymmetric ice of 10cm+. The recent band of ice storms in the middle Southern states has been experienced in the past.
It is nearly impossible to protect transmission lines from such ice events, as it also is for tornados. When these events occur, the network collapses and restoration is stepped and lines operate radially. Full network restoration takes months. Thus, some generation sources may be unavailable while others are stretched to the maximum.
Maybe the day will come when for certain regions, that transmission lines will be laid in tubes underground. Yeah, it is more expensive to do it this way but it may be the only way to go and may even be cheaper long term.
This sub-thread is way off:
1. “Note that in the US, wood pole construction for transmission is very common for “load serving” network voltages under 200 kv (perhaps the majority) and it is also present for 230 and 345 kv “bulk power” networks. It is very susceptible ice failure. More recent construction is steel Poles.” Wood is not susceptible to ice failure (or can be made to not be, anyway). Northern Canada, Norway, Sweden, Russia, Finland all use wood poles. Wood poles *are* susceptible to things like woodpeckers, though, and they do not last as long as steel.
2. “Maybe the day will come when for certain regions, that transmission lines will be laid in tubes underground. Yeah, it is more expensive to do it this way but it may be the only way to go and may even be cheaper long term.” Not even remotely close to the mark. First, “transmission lines laid in tubes underground” = cables. These are obscenely expensive and very difficult to maintain. They exist for dc links (and are still obscenely expensive) because ac produces “parasitic capacitance” against the wall of the tube (basically, it siphons off the power that is transmitted. This also limits the practical cable length.
I believe that a couple of decades ago the idiots in Denmark’s parliament passed a law mandating something similar, once again demanding that the laws of physics be bent to shape their political goals. I don’t think I bothered waiting to see how it turned out.
“Ice storms are extremely destructive to powerlines and it’s a known hazard. The last few years utilities have been starting to use drones to knock ice from lines and even hanging trees.”
Just in case you want a simplistic overview of the details: The ice forms on a T-line and basically adds to the weight of the line. In high winds, this produces “galloping”, where the lines start swinging back and forth. The momentum from this is often enough to shake the towers on each ends.
Using drones to knock ice off the lines is a little silly–ice melting (connecting a mobile transformer to the line and then using current to heat the line to melt the ice) makes a lot more sense, I think, unless I am really missing something. That’s how the northern countries typically deal with ice buildup, anyway
I had 50% of the US grid as customers in my annual reliability survey which ran until 2018. Wood construction is FAR more susceptible to structural failures for any cause than steel, particularly the cross arms. Much of the areas recently affected in the middle Southern US, wood construction predominates and much of it is in excess of 50 years old. US infrastructure ain’t like Sweden, Finland or Norway.
Ice storms are extremely destructive to powerlines and it’s a known hazard. The last few years utilities have been starting to use drones to knock ice from lines and even hanging trees. Tree pruning or lack there of is an issue that isn’t about powerlines. Some utilities do a great job others not. Often it’s local people who protest about cutting trees putting the utility in a tough spot.
There are companies now producing such drones and it is and will have a huge effect into the future to minimize ice and really it’s just freezing rain that is the issue.
Generally freezing rain is highly localized and expected by forcast so the drones can be deployed ahead of time.
Tornados while doing total devastation in their path are relatively small in area and only underground services can protect against that.
“CEIS notes the average home in 1970 used under 100 kwh of electricity annually. The service drop for most homes was at 100 amps. In 2025, the average home uses more than 10,000 kwh per year and the standard service drop is 200 amps.”
This is the main issue. Terran humans are using electricity at a far higher rate than in the past, per person. We should ask; what are the basic sources of that extra demand? [What are the overall uses today versus fifty-five years ago? Is this “average” usage a true average of total electricity use divided by population? Does it include items of peripheral utility to the “average” person as well as ‘basic’ uses, such as home use? This is a case where the definition of terms is definitive.]
The second item which will have major impacts on the “average” utility ‘customer’ is the “official” policies of the electricity suppliers concerning the relative importance of “customer” demands. Roughly speaking, who will have first use rights? Will home heating use be sacrificed upon the altar of corporate and governmental electronic surveillance regime use?
I do hope that data farm plants be supplied with robust back-up power generation capacity on site. I can easily see enraged mobs destroying electric power connections for regional data plants as ‘revenge’ for the local deaths of people due to electricity cut offs done in favour of these “industrial” users.
We live in interesting times.
100 kWh of electricity annually per household isn’t a credible number. That’s the equivalent of running one 60W lightbulb for 4.6 hours per day. I’m pretty sure it’s a typo.
According to the US Energy Information Administration (at https://www.eia.gov/todayinenergy/detail.php?id=49036), people were consuming about 2,300 kWh of electricity per person annually back in 1970. It’s risen to about 4,500 kWh (again, per person) of electricity today. So it’s basically risen by a factor of 2. Not a factor of 100.
As for causes, the EIA also has the following article on why electricity consumption has increased over the years: https://www.eia.gov/analysis/studies/buildings/households/pdf/drivers_hhec.pdf. The main drivers include larger houses, more air-conditioning, more clothes dryers, oil- and gas-fired furnaces being replaced with heat pumps, and increased use of personal electronics.
We do have lower power light bulbs now and a effort was made in more enlightened times to greatly increase the efficiency of appliances like refrigerators and to add more insulation to houses. This was at a time when the country was concerned about both pollution and depending on foreign oil.
Now Trump probably favors taking the insulation out of houses so his energy buddies can sell more of it.
Locally Duke Energy asked customers to conserve as much as possible during the recent (with more to come later) freeze. Our power stayed on, but the water company seems to be busy around town with breaks in aging water pipes. I had a yard faucet leak myself but was able to get it fixed in a couple of hours after a trip to Home Depot. The run of very cold days we’ve been having is very unusual–perhaps not since the 1980s. Our recent climate refugees may now be considering Fiji.
If you look at the national electrical usage by year, it’s almost flat from 2000-2022. And extremely flat from 2007 to 2020. The link below has a graph in it that shows it.
A large part is due to efficiency standards
It’s one of the reasons that the utilities are trying to catch up to the growth being driven by all sorts of things of new loads also with new environmental needs like hotter summers: AC, EV, heat pumps, data centers, IOT etc. I mean a whole generation of utility design engineers went by without much to do.
https://www.eia.gov/energyexplained/electricity/use-of-electricity.php
I wonder how many of these articles are AI written with such glaring errors like the 100kwh year.
Actually, it’s flat due to de-industrialization (and efficiency improvements) – see China, which also has improved efficiency, as the US population doubled. Read down in the eia link and see industry is the LARGEST user of electricity.
Of course, an article cannot be all things, but any article should note the rapidly growing delivery timelines for utility components like transformers* – in part due to said de-industrialization and rare earth embargoes to the US.
If (and it’s an unlikely if) the US were to try to upgrade its system, the 750 billion would almost certainly rise into the many trillions and take decades due to shortages and a lack of western manufacturing capacity, mining, etc. – copper is acting more and more like a precious metal, for example.**
* https://kdwalmsley.substack.com/p/another-casualty-of-the-trade-wars?utm_source=publication-search
** https://kdwalmsley.substack.com/p/criminal-gangs-are-ripping-copper?utm_source=publication-search
“it takes an average of 29 years to open a mine in the US.”
As an example, consider the replacement of the Francis Scott Key Bridge in Baltimore – costs have already skyrocketed from an estimated $400 million to $5 billion, while completion is pushing back by years.
Thanks for checking this for all of us. 100x jumped out at me, too.
I really hope I never see a mob attack a sub-station or high-voltage line, as it would be a stupendously dangerous thing to do. Voltages on the high-voltage side are so high that they can jump through the air and electrocute a person without any physical contact required. Available short-circuit currents on medium-voltage side are so high that the resulting arc-flash from a short-circuit event can inflict whole-body third-degree burns from yards away. And if the high-voltage side is shorted out instead, you might black out the whole city instead just the “industrial” user being attacked. Mind-blowingly foolish.
But fortunately, even with their excessive political influence, I don’t think data center owners will ever manage to rig things so that ordinary households are disconnected before they are. I can’t think of faster way for a politician to lose an election than to have voters learn that AI algorithms kept running while people literally froze to death.
Heck, I’ve seen a variant of that decision process in my own experience. About 15 years ago, the substation that feeds the factory where I work (manufacturing power generation equipment) was damaged when a truck driver had a heart attack and crashed into it. A large transformer was taken off-line, and the second big transformer that was still energized now had to feed both my factory and a nearby hospital in parallel. It started to overheat.
Our plant manager got a phone call from the utility, giving us 10 minutes warning that our power was going to get cut off. So we ran around frantically, turning off computers and servers and getting manufacturing equipment into a safe state. And then they cut us off. Nobody complained. We all understood. It was more important to keep the hospital up than our factory. Lives depended on it.
Thanks for the report from the trenches as it were. As my comment showed, most people have no idea how dangerous electricity systems can be. (I worked in commercial construction, and I didn’t know the physics of high level power installations.)
As for your comment about the “consciences” of elite class managers, I beg to differ. I have observed truly callous and ruthless people in positions of power. They wouldn’t really care about that hospital if the factory to be shut down was of any importance to their goals. You echoed the ethos of on the ground builders and makers. The “real” world is intimately bound up with that cohort’s actions. Many managers are disconnected from that “real” world. The public relations aspects of such a dire decision would have greater weight in their calculations than any actual effects upon the public.
I must mention that I am speaking from an extremely cynical outlook. Right or wrong, such a viewpoint should not be dismissed out of hand. Too many preventable disasters loom in the shadows of history for that to happen.
Stay safe. (I will tell the mob to be careful where they point those pitchforks.)
Oh, I have no doubt that there are “truly callous and ruthless people in positions of power” who don’t care about anything but their own goals, consequences be damned.
But fortunately, they’re not in control of the actual switches. Shutting power off to a data center or factory or hospital often requires sending a human being down to the sub-station to turn off a giant switch. And because these switches handle such large amounts of power and can fail catastrophically (emitting giant blasts of UV radiation and fireballs in the process), you have to put on an arc-flash protection suit before you even approach the switch.
And the number of “elite class managers” who are willing to go through this level of hassle and risk is essentially zero. [Heck, they probably don’t even have the keys necessary to access the sub-station innards.] So the guy actually flipping the switch is some non-managerial grunt who’s been through a lot of training and understands the consequences of what he’s doing. And because he has physical control of the switch, he has ultimate veto power. He can refuse an order that will kill people.
“And because he has physical control of the switch, he has ultimate veto power. He can refuse an order that will kill people.”
And therein lies the real power of guilds and unions.
It is sad that “real” unions are not thick on the ground in America.
Stay safe.
Transmission companies deal with risk (both personal and societal) on a level that’s hard for regular worker bees to understand, precisely because the consequences of failure or mistakes can be so catastrophic.
I did some work for one once, and safety consciousness pervaded everything they did. They were a typical large corporate and some of their values were definitely more aspirational than real, but THAT one they lived and breathed. They were serious enough about it that those of us in regular low-risk desk jobs were expected to comply with the same general standards, not because it was necessary – they knew perfectly well it wasn’t – but simply so that nobody would ever have to think about whether they applied. They applied to EVERYONE. End of story. One more source of potentially fatal misunderstandings eliminated. If that meant a bit more inefficiency in some parts of the business, it was a small price to pay.
You would think that mathematics would read its ugly head as far as the vulnerability of the US grid is concerned. So much of the grid is using really old equipment, some of it a century or more older. At the same time the population is increasing as is power usage with always-on electronics so pervasive. Then on top of all this you have a rapid buildup of data centers who are conveniently not expected to kick in money to meet their new power demands. The maths doesn’t work. Sooner or later somethings gotta give as ramped up demand meets a more and more fragile electrical grid. I can see a time coming soon when average Americans will have to routinely take steps to insulate themselves from when their local grid fails and they have no power. It would be worse if it was seen that power was still being routed to their local data centers to keep them running. That would be adding insult to injury.
“You would think that mathematics would read its ugly head as far as the vulnerability of the US grid is concerned. So much of the grid is using really old equipment, some of it a century or more older”
Yes, but really the biggest problem by far is that zoning/permitting to build transmission lines is extraordinarily difficult, time consuming, and controversial. This forces utilities to basically overload existing lines and to prevents redundant paths, which means that a fault gets amplified **across the entire system**, not just in a local area (in technical terms, it increases the short circuit faults within the entire system). It also means that when a line/transformer/whatever goes out, there is no “easy” way to shift the electricity–the choice is to either overload lines that remain functioning (putting them under more stress) or permitting outages.
And by the way, the US is not ***by far*** the worst off regarding the state of its electric grid. The entire West has basically entered into a mutual suicide pact with stupid policies that ignore reality for short-term considerations, and then each state outdoes each other trying to come up with a more extreme implementation of these stupid policies. Seen in this light, silly things like Europe sanctioning its primary source of energy while Germany televises the destruction of its nuclear generators are, while shocking, unsurprising–they are merely a continuation of mentally retarded, dysfunctional decision-makers and enablers.
And pro-tip: if you think the electric grid is in crappy shape, go investigate the state of the water system. You will be stunned.
” another problem is that many regional system are not connected” This is incorrect unless the author is talking about the 3 main grids in the US. Eastern, Western, and texas which don’t interconnect.
It’s actually one of the reasons that say PJM has price spikes in that they are buying power from outside of their usual suppliers and so get charged a premium. But the lines are all interconnected.
Transformers age. Yes many are old, but as long as they don’t get hot they last an extremely long time. There isn’t anything going on in them. Heat is what destroys the insulation which will cause them to fail. It’s why transformers are so oversized and have reduced loads on them, to make them last a very very long time.
30 yr old powerlines; If they mean the actual lines, sure its just aluminum wire which doesn’t really wear out and some probably are 30+ years old. The actual maintenance is in things like insulators on the poles, replacing poles etc. Many steel towers are 70+ years old, not an issue, the golden gate bridge is 90 years old. It’s the wear parts, insulators, etc that move and need to be replaced which they usually are. PG&E lied about the maintenance which caused a failure of a 50+ year old part that caused the paradise fire. Utilities do preventative maintenance all the time.
A lot of the issues with power lines has to do with trees, as they grow they need to be trimmed or cut back which a lot of people don’t like. A coop in the sierra’s has a perfect record of no fires while others in its area have many becuause they actually do a great job of tree trimming.
The local utility was upgrading the main power lines in my area a few months ago, above ground. Many people wanted buried. The utility provided numbers that and its a pretty easy place to do it in, that it would cost about 5-10X more and its more expensive to work on and repair. In an above ground you can find the issue pretty easily and easy to fix, in buried, you have to locate the issue, then dig it up etc, which they explained was a lot slower and more expensive. Other places might have different numbers.
On a personal note, I really really don’t like articles that start with fear based numbers, ie ” at least 35 deaths associated with the storm” When the actual numbers if you look into it will be less because people are not at work, or driving etc so less accidents and deaths. And pretty sure people have been dying in storms long before AI and data centers.
“This is incorrect unless the author is talking about the 3 main grids in the US. Eastern, Western, and texas which don’t interconnect.”
Actually, they do interconnect. The connections can be severed, though, if the operators choose to do so.
Your comments about transformer and t-line aging are spot on. A transformer can basically last forever as long as the oil gets processed from time to time and the thermal insulation doesn’t break down, and the insulation only breaks down due to “over” heating. The regulatory process, though, makes it difficult to order a large transformer that lasts “forever”, even if the marginal cost is negligible compared to the overall project costs (as in, under 1% of costs).