Rob West’s Top Energy Themes for 2026 + Your Doomsday Power Backup Plan
This week on the podcast, we welcome back Rob West, founder and CEO of Thunder Said Energy. Founded in 2019, the firm provides research that helps decision-makers identify energy opportunities. Based in Estonia, nine time zones away, Rob is an exceptionally productive energy expert whose work spans a wide range of topics.
We begin by walking through Rob’s Top Ten Themes for Energy in 2026, including the continued steady growth in global oil demand, a waning focus on net zero, EVs, and decarbonization. With that lens, we also discuss Canada’s Pathways carbon capture and storage (CCS) project. Rob then shares his bullish outlook for LNG demand growth, with positive implications for Canada’s aspiration to grow LNG exports. Rob also argues that there is a growing investment case for grid-enhancing technologies to increase the utilization of existing infrastructure and meet rising electricity loads. We also touch on the outlook for copper demand driven by electrification, robotics, and AI data centers, as well as Rob’s expectations for electricity load growth, which are more conservative than some other forecasts.
Finally, Rob and Jackie revisit Jackie’s “doomsday” scenario: what it would actually cost to back up her home during an extended power outage, comparing options such as using stored power from an electric vehicle, a home battery, and a natural gas generator.
Content referenced on this podcast:
- Sign up for Rob’s daily note at his website, https://thundersaidenergy.com/
- Ten Themes for Energy in 2026 from Thunder Said Energy (January 1, 2026)
- Rob’s video: US load growth: unpopular opinions (September 3, 2025)
Please review our disclaimer at: https://www.arcenergyinstitute.com/disclaimer/
Check us out on social media:
X (Twitter): @arcenergyinst
LinkedIn: @ARC Energy Research Institute
Subscribe to ARC Energy Ideas Podcast
Apple Podcasts
Amazon Music
Spotify
Episode 310 transcript
Disclosure:
The information and opinions presented in this ARC Energy Ideas podcast are provided for informational purposes only and are subject to the disclaimer link in the show notes.
Announcer:
This is the ARC Energy Ideas podcast, with Peter Tertzakian and Jackie Forrest, exploring trends that influence the energy business.
Jackie Forrest:
Welcome to the ARC Energy Ideas Podcast. I’m Jackie Forrest.
Peter Tertzakian:
And I’m Peter Tertzakian. Well, Jackie, two years ago today on January 13th, 2024, it was a -30 night, and a very serious situation on the electrical power side occurred. And if there had been an outage, of course, we would’ve all been in dire straits without any power.
Jackie Forrest:
Right. We got that message on our phones that everybody should reduce their use of electricity. And that would’ve had big consequences if we did have rolling brownouts or even a blackout, that would’ve been terrible.
Peter Tertzakian:
Yeah. Yeah. So two years on, we haven’t had the -30 night yet, although the last week in January and the first week in February, typically the coldest nights of the year, so we shall see. But in the intervening two years, we’ve had tremendous growth in electrical power demand, particularly on the AI side across North America, and predicted to grow here quite dramatically. We had a podcast on that late last year. And so, that begs the question, well, what do you do to mitigate personally against the potential for power outages? And I know you’ve got, I’ll call it a doomsday project going, where you want to become much more off-grid and protect yourself against these kind of things. So what are you doing?
Jackie Forrest:
Right. Well, I looked at what it would cost, first of all, to decide if it was worth it. But I do think with the tighter electrical markets, not just in Alberta, across this country, in North America, the fact we’re all interconnected means that even if Alberta at this moment has a bit of spare capacity, some other jurisdiction could cause a shortage that would affect us here, too.
Peter Tertzakian:
Right.
Jackie Forrest:
So I think it’s important to sort of look at what would it cost for my home to be energy secure? And I looked at things like using my EV to back up the house. I’ve got the solar panels. How could I tap into that electricity when it’s available? And even looking at things like power walls or even natural gas generators. So all to say it’s way more complicated than you would think.
Peter Tertzakian:
Well, I know it is, because I actually have a natural gas fired backup generator, and I have a Tesla power wall. And they give comfort and I have solar panels. But nevertheless, you do have to maintain these things and the cost is quite substantial.
Jackie Forrest:
It is. So I think we should talk about this topic as well as many others was one of the smartest people we know in energy. I want to introduce Rob West, founder and CEO of Thunder Said Energy. He’s a repeat guest, so welcome, Rob.
Rob West:
Yeah, thank you very much. I’m delighted to survive doomsday with you guys.
Peter Tertzakian:
I think it’s the third repeat, aren’t you? You’re one of the very few guests that have been on three times in a row.
Rob West:
I think so, yeah.
Peter Tertzakian:
Welcome back.
Rob West:
Well, you’re not bored of me yet. Not bored of me yet, but there’s still time.
Peter Tertzakian:
No, our audience never tires of you, so welcome back. Hey, Rob, tell us, for those who are first timers with you, tell us about Thunder Said Energy that you founded.
Rob West:
It’s a research firm focused on all the ways we produce, consume, and transform energy. And I’ve been doing this for the last seven years now. The more I learn, the stupider I feel. For the first four to five years of doing this, the momentum was just completely behind net-zero technologies and seemed like it was all anybody wanted to talk about. And now, the theme to rule all themes has switched towards power security, empowering AI and load growth. And so, I hope we get into some really interesting ideas around all of that today.
Jackie Forrest:
For sure. And I will put a link to the Thunder Said Energy site. You can subscribe to their daily newsletter and there’s no cost for that, but if you’d want to get more details, then you can be a subscriber like we are. And we really enjoy your research, Rob. I always feel smarter every week after reading your research.
Well, let’s start with, you just released your top 10 themes for energy in 2026. We just had our 2026 podcast. We want to contrast and compare some of your forecasts. We will put a link to this one as well, this note about the top 10 energy themes, but one of your themes was that net-zero ambitions will continue to soften mostly silently. So explain that and tell us what your thoughts are about Europe, an area that most people in clean energy say, “Well, that’s the area that’s going to continue to have policies that get us investing in clean energy.”
Rob West:
Yeah, it feels like there’s just less interest behind it. And the thing that has really captivated people’s interests is this shiny new toy of what you could do with AI. And so, we’ve seen some examples that I think really illustrate that where, for example, there was a big blue hydrogen project that was going to be built in Teesside in the UK by BP. And last year they pulled the plug and said, “It’s too expensive. We can’t get enough government support to really make this a bankable investment.”
And now that same site is earmarked to be an AI data center. If I look at companies across my coverage, I’ve been doing a lot on robotics and how cool robots could change the world and improve productivity and reshape cost curves. And one company that does robot maintenance now describes stranded assets, not assets that emit CO2, but those that fail to embrace digitization and robotic technologies to keep their costs down.
Late last year, I was looking at the automakers, the OEMs. About a third of them have dialed back their EV aspirations, again, because of the unwinding of incentives and challenges selling those vehicles. But all of them are stepping up their focus on autonomy. And they start talking about net-zero, not to mean a fleet with net-zero emissions, but a fleet that has net-zero collisions and net-zero road fatalities. And so some of these things that I’ve just mentioned there, they really are just seeming to be exciting people in a way that paying 3 trillion dollars a year to decarbonize the world, which wasn’t going pretty well anyway, seemed to excite people less.
Peter Tertzakian:
Based on those comments and my last year’s prediction that net-zero would die, isn’t net-zero de facto dead already? Net-zero carbon emissions? That whole layer on more and more policy in the pursuit of net-zero emissions by 2050, isn’t that movement replaced with electrification and that innovation such as you describe in the pursuit of electrification to drive robotics and AI will necessarily transition us to new energy system circumstances?
Rob West:
I actually think I’m an optimist about this. I don’t think we will get to net-zero. I think we’ll still have about 30 billion tons a year of CO2 by 2050. So it’ll be down 40% from where it is today. It’ll be peaking this decade and going down. But actually, if you kind of flow that through to a climate model, we’ve already had one and a half degrees of warming. We end up with somewhere between two and three. And when people really got worried about climate and decarbonization, there were forecasts out there for like six to eight degrees of warming. So we are going to have huge growth for solar and huge efficiency gains in our energy system, and we might get the big crisis that could have come somewhat averted without having to go all the way to doing the really, really expensive zero means zero kind of scenario.
Peter Tertzakian:
I think we’re saying the same thing in a sense that AI, robotics and electrification are going to rewire and repipe our energy systems over the course of the next 25 years in ways that will reduce emissions and that you don’t need layers and layers of government policy that never worked anyway, or are unlikely to work, that actually innovation and a new set of digital circumstances is going to drive the trends.
Rob West:
Yeah. And it unleashes a lot of really interesting second order effects and third order effects. I mean, one that is on my mind right now is one of my themes for 2026, is the idea that actually, we always used to assume oil demand would grow a million barrels a day every year globally. And then, the net-zero narrative came along and the expectation was, oil demand’s going to peak and it’s going to roll over. And actually, one prediction I have for this year is that we go back to a world where people just as a default start predicting a million barrels a day of growth in oil demand every year from here on out. And it’s linked to, I think, we’ll have lower EV sales this year than last year. It’s linked to maybe we’ll have lower oil prices and that’ll unlock a couple hundred thousand barrels a day rebound effects.
But the big one is how the age of the internet and AI is increasing mobility. And you can pull together just fascinating data on that. But if you look at population density, people are moving out of really, really dense cities and moving towards suburbs and exurbs and driving more. And I’ve kind of lived that myself moving to a house deep in the forests of Estonia where, like Jackie, I am considering my preparedness for off-grid doomsday scenarios.
Peter Tertzakian:
Well, we can understand that being close to Russia these days.
Jackie Forrest:
I want to come back to though, Rob, Europe, because things like carbon capture storage, industrial decarbonization, green hydrogen, these are things that really do need subsidies. What’s your thoughts on those? And a lot of people say, “Well, Europe still has things like carbon border adjustment coming in this year.”
You’ve got their requirements for e-fuels and that there still will be a market for these things. It may be smaller than what people thought, but what are your thoughts? Do you think in 2026 we’ll still see investments in these areas, just a lot less than you might’ve thought a few years ago?
Rob West:
I kind of agree with Peter that the writing’s on the wall here. This is all going to get unwound. It’s not something that voters actually want to pay for, in my opinion, and it’s kind of pointless. You think about the old prisoner’s dilemma, the classic game theory, two prisoners, if they both collaborate, they get the best possible outcome. But if one of them collaborates and the other one snitches, the one that collaborates gets the worst outcome.
And that’s kind of where we are with high cost decarbonization in Europe. Why are we in Europe, or why are we in Canada, are we in Canada, I guess Australia as well, putting in place policies that make living costs more expensive, make industry more expensive, less competitive, and then all of our competitors [“adversaries”] are not doing that. And you’ve got real problems there. If you’re shouldering the cost and your competitors, whisper, adversaries, are not, and then they’re going to steal your industry and attack you.
And so, you’ve got to really think hard about why are we pushing those costs onto industry and people? And my guess is that you will have policymakers who make a name for themselves by coming out against this. And we’ve already seen it with some parties, for example, in the UK, Germany. You’ve had it in Canada. And actually it’s kind of interesting how some of the support behind those policymakers has been growing.
Peter Tertzakian:
Yeah, I agree. But I also would say further to my prior comments that innovation and economics are going to lead us towards solutions to decarbonize faster than we thought. So it doesn’t mean that a lot of the stuff that was invested in and that society at large of the West was working on or China was working on is finished, but that’s not what we’re saying. There is a certain camp that says, “We’re all going back to oil and gas and that’s the end of the story.”
No, that’s not the end of the story. It’s just that this new set of mega circumstances that necessitate electrification are going to actually create a new set of exciting opportunities to invest in that, by the way, include things that were being nurtured over the course of the last 10 years.
Rob West:
Energy systems aren’t black and white. They’re multiple shades of gray. And I kind of love that point because like I said, I think emissions are going to peak and what’s going to drive the peak is there’s going to be huge growth in solar. That is going to emerge at the bottom of the cost curve, and be the single largest source of growth between now and 2050 as a source of energy supply. And EVs have a great role in some contexts. I became an EV owner last year. I have a plugin. I charge it on a time of use tariff over here. So basically free power at night coming over the border from Finland. And I haven’t been to the gas station in like eight months. It’s so convenient. It just fills itself up every night whenever the power price goes below my set point.
Peter Tertzakian:
You plug your phone in at night, you plug all your electrical peripherals, so why wouldn’t you plug in your car? That’s what I do.
Jackie Forrest:
The only benefit though of pure electric like mine, Rob, is I don’t have any maintenance costs either. You’ve still got to maintain that combustion engine that you barely use. So got to go pure EV.
I want to ask you a question, maybe this is a little unfair, but I’m sure a lot of people in Alberta would be interested in your thoughts on this or even Canada. We are sitting with a big decision in front of us here in Canada. We want to build an oil pipeline. If you haven’t been following the news, we’re getting threats from Venezuela now, the Americans. We want to create new markets for our oil. But there has been a deal put forward between the federal and provincial government here in Alberta that we can build this oil pipeline if we build a 20 billion dollar carbon capture storage project, because one of the things about our heavy oil from the oil sands, it is higher emissions than a lot of other crudes. And what are your thoughts? At this moment in time, are we better to put this 20 billion dollars into the carbon capture storage project or something else in our economy? Assuming the government has to pay for part of it.
Rob West:
Can you be in the carbon capture and storage project and still build the pipeline?
Jackie Forrest:
Well, I mean, I think that should be one scenario we consider at this moment for sure, or delay the carbon capture storage project. But right now, the MOU has it as a requirement for the pipeline to go forward.
Peter Tertzakian:
Well, this is a deeper discussion. As I said, people talk about competitiveness and I say, “Well, what kind of competitiveness are you talking about?”
Because people talk about carbon competitiveness versus what I call business competitiveness. And the answer to your question from a business competitive perspective is no, you can’t build both and be competitive, particularly in light of Venezuela and potential other heavy oils coming to market. So from the perspective of carbon competitiveness, the issue is, and love to get your perspectives, does anybody pay for a decarbonized barrel of oil, an upstream decarbonized barrel of oil? Does anybody pay a premium?
Rob West:
I would think about it in the way I think about an LNG plant. To get an LNG plant built, you have to get about 70% or more of the LNG pre-sold on long-term contracts where a buyer signs an SPA and says, “I am willing to buy this LNG when the plant is built at this pricing formula or spot or plus a margin.”
And you have to have that to take it to the bank. And that has become a staple of project finance for projects that cost around 20 billion dollars for very good reason. And there’s something slightly scary about building something when there isn’t a market demand for what will be in a range of 200 to $300.00 per ton of CO2. And if you can find the buyers who will sign those contracts and say, “We will absolutely take it, and we’ll sign legally binding agreements,” then great, go ahead. You’ve got something that is a business that somebody will pay for.
But I think from projects I’ve been tracking, that’s high on the cost curve, and it’s not a cost that industrial buyers have been rushing to go forward and pay. And I think that might warrant some caution around a technology that cost in that level has a 20 to 40% energy penalty attached to it, has technology risks associated with the degradation rates of A means and the monitoring of reservoirs. This is complex, energy intensive, costly stuff, and do that without buyers at your peril.
Jackie Forrest:
And something to think about that 200 to $300.00 per ton, I actually think who’s going to have to back it is the taxpayer, the government. And if you cared about emissions, there’s certainly cheaper emissions in our economy.
Rob West:
I just think, “How does it move the needle?” In a world of 50 billion tons a year of emissions, does it move the needle at all? That’s, I think, one of the really hard questions. If you knew we were on a road to net-zero and everybody else was doing it, and you had to do it as well because otherwise you’d be barred from global markets, it’s a different story. But we’re in a world with 30 to 50 billion tons a year of emissions and going it alone and making yourself less competitive might not end great.
Peter Tertzakian:
Well, and the other part of this is that, as you know, 15% or thereabouts of emissions is in the upstream. 85% is in the downstream when you actually put the key in the ignition. So not only does it not move the needle, it does not move the needle in the lifecycle of the barrel of oil produced in Canada that goes to the end use. We’re a very small fraction of the emissions in the whole full cycle supply chain.
Rob West:
Yeah. I mean, to some extent, it feels to me like Cinderella is just not allowed to go to the ball. And whatever Cinderella does to try and make herself go to the ball, some new ridiculous requirement about cleaning a carpet will be levied and then she can’t go to the ball.
Jackie Forrest:
Yeah. Well, and another thought I have, too, this was all conceived maybe in a time and era four or five years ago when it seemed like you needed that to differentiate your product. That has changed, reality has changed. The other thing is maybe in 10, 15 years, there is a more cost competitive way of doing this.
Rob West:
I can’t resist, Jackie. You read my mind. I mean, out there in the theoretical state space of 10 to 23 metal organic frameworks that could be created that we haven’t explored yet, there is something out there that with about 120 to 150 kilowatt-hours per ton, which is 95% less than the best CCS DAC projects today, pulls CO2 out of the atmosphere and mineralizes into a limestone construction material. And you could grow artificial islands at sea, in lakes. You could build construction materials out of nothing but thin air and energy, all powered using the leftover, curtailed energy from solar, and you could pull as much CO2 out of the air if you want, and the cost would be sub $100.00 a ton.
And it’s out there. And this brings us to, God, it would be so cool if we could build a multiple Zetaflop chemical analysis AI and start trolling through that state space of metal organic frameworks and [inaudible 00:19:55], solve the problem rather than just throw money and cost and pain at it.
Jackie Forrest:
So we could be investing in a stranded asset because in 10, 15 years with AI, help making your vision a reality, we could have better technologies, too.
Peter Tertzakian:
Well, let’s move on, Rob, because the vision of robotics and AI and electrification depends heavily on other vital commodity for which there are few, if no substitutes, and that’s copper. What are your views on that? The price has been going through the roof lately.
Rob West:
Oh, there’s been so much disruption in the copper market, especially last year. So we had all time highs last year, about $11.00 a kilogram. The thing that has really struck me looking at the market is the marginal cost of getting new copper mines built is somewhere in the range of 12 to 14. So I think that’s always the best bellwether for where long-term prices are going.
Peter Tertzakian:
Where is it today?
Rob West:
- About $11.00 a kilogram. But I’ve just been increasing my numbers steadily through things that I wouldn’t have modeled a few years ago. I mean, solar, I think this year we’re going to have pretty flat year, so it’s not going to be a huge demand pull, but long run, that’s a big driver. There’s the moving electricity within buildings, within data centers. There’s some upside, but don’t get too carried away. And then within robots and all these things in the industrial AI ecosystem, about 10% of them by mass tends to be copper. So the more of that robomass you deploy, you’ll probably be pulling a couple million tons a year of upside into that market.
Peter Tertzakian:
What about transmission, new transmission lines?
Rob West:
Well, that tends not to be copper. That tends to be aluminum, and it’s the HVDC lines can be copper. So I’ve got a bit in for that. But it’s really once you get to the low voltage and within facilities and within appliances, that copper really accelerates. And then the big one, of course, air conditioning, global demand for air conditioners could quadruple in the next 20 years, and that’s a big part of the market, too.
Jackie Forrest:
Right. And these AI data centers take a lot of copper as well. I think that was one of your recent research notes was on that as well.
Rob West:
Yeah. I mean, that one, actually tread carefully. There are some numbers floating out there that are kind of bogus numbers. NVIDIA put out a big article about how it needs to upgrade to using higher voltage bus bars across all of its data centers. Otherwise, “a gigawatt scale data center would use half a million tons a year of copper,” which would be like one and a half percent the global copper market.
And I thought that was quite interesting. So I tried running a research note on it, and the conclusion of the research note, doing all the maths from first principles, is I’m pretty sure NVIDIA meant half a million pounds, not half a million tons. And so it was actually a typo. And then the typo got picked up by 12 news outlets, four other companies who’ve all repeated it. And it sort of entered the… I get people who go to AI conferences and they come back and say, “Oh, my God, building data centers this year is going to use 50% of all the world’s copper.” And be quite careful there.
Peter Tertzakian:
Only out by a factor of 2000.
Rob West:
Right. Yeah.
Jackie Forrest:
Okay. Bullish on copper, but not maybe that bullish. Okay.
Rob West:
But it’s going to grow like 3.5% a year, but it’s not going to grow 35% this year.
Jackie Forrest:
I want to talk about LNG. You’ve written quite a bit about LNG starting off the year. Of course, there’s this consensus view that there’s going to be a glut because the Qataris are bringing on all this supply, the Americans are bringing on all this supply. And you may not know about it, Rob, but we too in Canada have our first LNG export project with LNG Canada phase one.
Rob West:
You’re in my database.
Jackie Forrest:
Oh, that’s good.
Rob West:
And there are like half a dozen other interesting projects swirling around. It’s not just LNG Canada. There’s a lot of interesting stuff.
Jackie Forrest:
Yes, that’s right. If you’ve been following, our prime minister has an ambition to get 50 million tons by 2030. I mean, the Americans are going to do that in the next couple of years just in terms of incremental growth, but that’s our goal. And I think that is doable with the projects on the books, and then a goal to get to a hundred by 2040. So I just want your thoughts on, are we going to have a glut? And in that respect, is there room for Canada to grow our goals of 50 million tons and then 100 by 2040?
Rob West:
The big driver of this market is the maturation of coal. Remember, the average coal mine in China gets 13 meters deeper every year. The costs of new Chinese coal mines are now coming in at 200 to $300.00 A TPA. That’s like five times what it used to cost a decade ago and what it costs in places like even India and Australia. So there’s real concern about resource depletion of coal in China.
And by 2030, if you have maybe 9 to $10.00 in MCF LNG, you could actually be cheaper than the marginal coals being mined in China on a power generation basis. So if you think about coal consumption globally, 9 billion tons a year, and if you think about the LNG market, we’re talking about something 420 million tons a year. So when you frame it like that, it’s not quite apples to apples because LNG is about twice as energy dense as coal, but that’s how you can end up with models where you have over a billion tons a year of LNG demand by the mid 2040s.
Peter Tertzakian:
It’s also somewhat bullish for coal prices. This is a dynamic in China that happened in the UK circa the 1970s. They had to go deeper and deeper into the mines to get the coal out, and ultimately resulted in the Margaret Thatcher shut down the coal mines, et cetera, because they’re uneconomic amongst other reasons. So sort of replaying that and notwithstanding what you think about coal as an energy source from a sheer investment point, you can see what’s happening with some of the coal stocks. They’ve been good investments.
Rob West:
Well, actually, what drove the coal market last year, China came in about in line, India came in about in line. The US was the one that consumed 10% more coal than I had in my numbers. And that’s this theme of just runaway power demand, question mark, linked to the rise of data centers in AI.
Jackie Forrest:
Hey, well, Peter, you wrote about whale oil as the only energy source that we’ve actually stopped using because we actually ran out of it. Maybe coal will be the next thing, but I think there will always be cool, but it’ll be maybe more and more expensive. It’s interesting. We always talk about limiting things on the demand side by these policies, but actually the thing that will really do it is on the supply side if it gets too expensive.
Peter Tertzakian:
I think the broad theme though is the pull on all primary sources to produce electricity, including nuclear coal, everything, because the generation of electricity is no longer just say wind and solar and batteries, which is what people were thinking. It’s now going not only into natural gas, we know that, but also coal and actually even oil fired power generation. I don’t know if you followed the demand for reciprocating or combustion engines, because you can’t get enough gas turbines. So gas fired reciprocating engines and now even oil fired reciprocating engines to produce electrical power, which ultimately means a pull on all of these commodities.
Rob West:
Energy is energy is energy. It’s all fungible.
Jackie Forrest:
Okay. Well, let’s talk about power grids, because generation is just one piece of it. And Peter’s described our current situation for electricity as a barbell transition where we’re building out the generation on one side and demand on the other, but no one’s investing in the grid. And in 2026, one of your predictions is this is the year we’ll tackle that with new technology that increases the use of our grid. And I thought that was a great prediction, although we’ve been predicting that for a while. It seems so obvious, but there’s all these non-technology barriers like regulatory requirements or the incentives of the utilities to actually do that when they only get paid for putting in hard equipment. So why do you think 2026 is the year that those barriers go away?
Rob West:
Because we need to. I think we need to in a way that we didn’t. We had 15 years in the US where power demand went sideways and every year the ISOs came out and they projected 2 to 3% a year growth. And every year, growth disappointed and we just didn’t have tightness in our grids the way we do now and now it is a problem. And so, everybody is attuned to that problem. You’d have to be living under a huge rock to have missed that people think power grids are a little tight right now.
And so, that is what sparks people to start looking for solutions. And I think just to spell out some of the things that I think are going to happen here, I think time of use tariffs are fantastic, and they’re going to be rolled out more broadly. I think data centers are going to realize they can use way less energy and way more flexibly. And the old adage that no, we must have 100% uptime or 99.99999, actually they’re going to realize it’s better to get a grid connection with 95% availability in the next couple of years than 99.99999 in 12 years time.
And so, I’ve already seen patents and business models focused on that and being able to allocate which AI loads need to happen right away and which ones can be deferred until power is available. And so, I think just a lot of this stuff is coming down the pipe, and flexibility is going to create a lot of value and opportunity.
Jackie Forrest:
I do want to put a link to your video that you did before Christmas on your view that these crazy high load growth expectations are going to disappoint people. They’re not going to be quite as high. There’s people with like 3% to 5% annual growth for electricity, for instance, in North America and your arguments for why that won’t happen. So we will put a link to it, but just quickly, why are you skeptical of these 3 to 5%? Because hey, if you look at all these AI data centers, I think even you expected 150 gigawatts of AI data centers to be built globally when we talked to you a year and a half ago. It seems like we’re going to have just this explosive load growth.
Rob West:
That’s actually not that much though. That adds about 1% a year, the AI ecosystem, everything in it. And on the other side of that equation, you have increasing chances to improve efficiency, move those loads around, and thus trim the amount of new capacity that needs to be added.
Even the AI numbers that I said there, there’s a lot of uncertainty in them. I’ve had in my numbers that the energy efficiency of AI would improve by about 30% a year. Last year, Google improved it not by 30%, but by 30 times. And looking across the old ideas about how we would get better models in AI, the concept was brute force scaling. If you just put a thousand times more data and more model through the model, that was how you were meant to be getting these improvements in performance. But now, you’re getting a lot of the improvements in performance through better learning algorithms, software, mixture of experts, all this other cool, nerdy stuff.
And then the final part, Jackie, is a lot of the growth that was meant to come was meant to be driven by decarbonization and EVs. And I’ve really cut my numbers there, because in the US they expired the $7,500.00 EV incentive at the end of September and the numbers 4Q, we were down 40% in EV sales in the US. So with fewer EVs getting sold and fewer heat pumps and less electrification of combustion heat, I would be downgrading my numbers by about a percent per year just based on that kind of policy-driven momentum.
Peter Tertzakian:
Which leads us back to Jackie’s doomsday project.
Jackie Forrest:
Okay. Yeah.
Peter Tertzakian:
So do you want to talk as we conclude our podcast about how to analyze that one? Because I can tell you with the investments that I’ve made, and by the way, I’ve bought a heat pump, too. I don’t know if you knew that. The payback on most of these other things like power walls and what have you, is probably never. So I’ve just done it, because I’m an energy nerd and I like to figure out how these sorts of technologies work.
Rob West:
The thing is, Peter, what’s the payback on having an iPhone? People don’t have it because they do a payback announcement. They have it because they think it’s cool. And I kind of feel the same way about a lot of the stuff in this home resiliency, home automation. One of my favorite things in the house that we’ve just built is after dark, all of the lights just go on motion sensors. So there’s that stat. 50% of the time the light is on, nobody’s in the room, not in our house. And I can get up in the morning and go downstairs, make my pot of green tea and come back up to my desk and I touch zero lights. The stairs just know that it’s 5:45 AM. There’s motion, turn those lights on at 10% brightness, and I love it. The payback, terrible, but it’s cool.
Peter Tertzakian:
Let’s get to the economics of this.
Jackie Forrest:
Okay. Let’s get to this thing.
Peter Tertzakian:
Because the capital cost of doing what you’re proposing to do is largely out of reach of the average person.
Jackie Forrest:
Yeah. Well, and special thanks to my husband who got all these cost estimates because that one itself was a bit of work. So we looked at the power wall, not the Tesla one. We found a lower priced option called Franklin. We looked at a natural gas generator, and now I would need a new EV unfortunately to do this, but you can tap into the batteries of your EV if you get the right EV. Unfortunately, in North America, there’s only really two cars that support that in Canada at least right now, neither of which I would like to drive, the F-150, which is now not even being made, the Lightning one. But the Cyber Truck for Tesla is the only other vehicle that will support that at the kind of flow rate I need. So I looked at all those different options, and I actually had a bias to the electric car charging my home because I thought that would be really cool. Because we have this huge battery that’s not being utilized and those batteries could last your home like seven days in a power outage. And here it’s just sitting in the garage if we had a power outage and I can’t tap into it.
So that was my bias going in. But it turns out that it’s just really expensive for the EV option, because not only do I have to buy this car, which the F-150s are pretty cheap right now, because they haven’t been a success. So you can get a massive battery for like 50 or 60,000 Canadian. But what I found is I probably need to add 15,000 Canadian just to tap into the power of that battery, because I need a box that’s going to take the direct current to alternating current. Rob actually taught me about this, too. I need to create a cycle in my home, an alternating current cycle, because I’m not getting that off the grid in a power outage. And so, all that together is at least 15 grand. So I’m looking at like 95 grand probably to buy this EV if it’s a Tesla one, or if I get a cheap Ford, at least still 80 grand.
Peter Tertzakian:
And by the way, you have to still drive in a black-out. You’re going to have to drive around to buy rancid groceries from unrefrigerated shelves in this grocery store.
Rob West:
It’s a distressed purchase.
Jackie Forrest:
I’ve got that combustion engine vehicle back up in that scenario. But yeah, the electric car thing just really didn’t work because of that cost.
Rob West:
Well, to make you feel a bit better, Jackie, the F150 Lightning, the energy consumption is like twice like a regular Tesla Model S. It’s like 0.35, 0.4.
Jackie Forrest:
For going down the road for each kilometer you travel.
Peter Tertzakian:
It’s heavy, right? It’s so heavy.
Rob West:
Yeah.
Jackie Forrest:
But I mean, you can buy… Some of them are going for like 50 or 60,000 Canadian and they have like 125 kilowatt battery. They’re cheap just for the battery. The problem is the logistics of using that power in your home, it costs you a lot. And then there’s the complication of… I had this other great idea that while the sun was shining in this outage the next day, I could be charging my F-150, so I could be driving around a bit more. But that in itself had a whole series of additional costs, because I can’t get the power off my solar panels unless I have power to the inverter. Anyway, that was going to cost another 10 grand.
Peter Tertzakian:
Wow, especially in winter where there’s 16 hours of darkness here.
Jackie Forrest:
Yeah. Okay. So the other option was the power wall, which was about $23,000.00 for our home. The problem is I only get about a day of backup. So if I look at it cost per day of backup, it’s like 15 grand for every day that I have backup for each power outage. So I just felt like that is a super expensive option, even though it was kind of a nice unit. It had all that inverter problem solved within the box, but it’s just so expensive for the amount of backup.
Because remember, with the electric car, I was getting maybe 7 to 10 days of backup. Now I’m getting like 1. And the natural gas generator was the other option, which I wasn’t a real fan of, because I want the cool technology and the low emissions and everything. But that one was about 17,000 and I could just back up my house forever in a power outage. Because I did some work. It seems like even in a serious power outage, we’d have natural gas flowing, because a lot of those natural gas compressors have backup systems in themselves.
So for 17,000, that’s actually just the price pretty much for me to get my electric car working with the house. I think there’s just no way to beat that, especially because it gives you a high amount of capacity, too. That’s the other thing. The actual maximum flow from some of these battery options would mean that I probably wouldn’t be able to meet all my requirements at peak, but with this natural gas generator, it’s putting out so much electricity that we wouldn’t have to change our lifestyle at all. We’d just operate like we did before the power outage, where with the electric options, we’d be all trying to conserve to make our battery last longer. So anyway, Rob, you’ve looked at this yourself. What’s your conclusion? Are you backing up the house with the natural gas or some of these electric options?
Rob West:
I’m backing up my house with wood, because in the dead of winter, I just need a day of iPad to sit and read before the power comes on and just to kind of stay vaguely warm. So we have a fireplace and that will suffice.
The analysis you did, I think is exactly right. I think the conclusion is you should try and buy Peter’s Generac unit off of him. That will be your most economical way to have long-term power in an outage. And I guess the only other thing that strikes me is actually grids are so amazingly reliable. It’s so rare that you have a protracted outage, at least in my part of the world. But there’s that nice Estonian saying that, On a rainy day, some people get wet and other people feel the rain. And I kind of think the day we have a power outage, yeah, I’ll have a great reading day, and we’ll eat the ice cream out of the freezer and have fun with it. And maybe that’s the right answer for my long-term anxiety.
Peter Tertzakian:
Yeah. Well, I’ll show my age again. When I grew up, power outages, especially during the winter, were, I’ll just call it routine, but they would happen maybe once a month. And my mother just had candles in the drawer and we just go get the matches and the candles and put the candles on. And it was like, “Okay, we’ll just wait this out.”
Rob West:
We had one the first year we moved to Estonia when we were actually staying at my in-laws house. And there was a power cut, and just magically my in-laws just appeared with candles ready to go. And I was like, “Whoa, how do you know how to do this?”
And they sort of looked at me very dryly and said, “We lived through Soviet times. This is not our first power cut.”
Jackie Forrest:
Well, Peter, you’ve had your natural gas generator at your home for how long and how many times have you used it?
Peter Tertzakian:
12 years. In the first few years, the grid was less predictable out in the country, so it would kick in in the winter two, three times on occasion for several hours. But as time went on and the city of Calgary expanded and the grid got better, I would say now maybe once a winter it might kick in. And actually, it kicks in more in the summer when there’s lightning strikes.
Jackie Forrest:
Now I would just say the past is not always what the future is.
Peter Tertzakian:
Exactly.
Jackie Forrest:
And when you look at these like… Well, Rob’s maybe a little less excited about load growth in North America, but when you look at some of these predictions and you look at how long it takes us to build generation, I like your optimism in terms of all these new technologies unlocking all this unutilized grid capacity. But history has shown that utilities are very slow with this stuff. They may need some outages to get them thinking in a new paradigm. And so I think the future, we may need it more. Maybe my backup plan will be just to come over to your house in my electric car while I still have power in my car.
Peter Tertzakian:
Right. But here’s what I do recommend. Here’s what I do recommend, and it’s a far cheaper option. And the thing that you really should be backing up is your internet. And you can do it for a couple hundred bucks. But if you get a good industrial battery backup for your network within your house, probably cost you 750 to $1000.00, and it will supply your router with power.
Jackie Forrest:
In an outage, right.
Peter Tertzakian:
In an outage for several, several hours and your computers.
Jackie Forrest:
Right. If there’s a major power outage, your internet may still work if you had this backup power.
Peter Tertzakian:
Yeah.
Jackie Forrest:
Do you have that?
Peter Tertzakian:
I do.
Jackie Forrest:
See, I’m planning for the doomsday scenario. Peter’s already got it in place.
Peter Tertzakian:
I do.
Jackie Forrest:
Rob, do you have this backup power for your internet?
Rob West:
No, I don’t. But I do have a smart router that will tell me when it’s disconnected and dead. So it’s not particularly helpful to know that. I know because my internet isn’t working, but it will tell me, “No internet, dead.”
Jackie Forrest:
Then you’ll know the reason. So, good.
Rob West:
Yep, exactly.
Jackie Forrest:
Well, Rob, as always, it’s been great to chat with you and learn about your work. Certainly thinking ahead about 2026, and we really appreciated the conversation today.
Rob West:
I hope we make it through the whole year with none of us requiring any doomsday solutions.
Peter Tertzakian:
Yes.
Rob West:
But it’s good to know we are increasingly prepared as a result of this conversation. So, thank you.
Peter Tertzakian:
Absolutely. And I am going to go back and read my, what is it, 17th century Cinderella story to remind myself what you were talking about with Cinderella’s ball, because I’ve completely forgotten that story.
Jackie Forrest:
Thank you, Rob, and thanks to our listeners. If you enjoyed this podcast, please rate us on the app that you listen to and tell someone else about us.
Announcer:
For more ideas and insights, visit ArcEnergyInstitute.com.
