Intro: Welcome to The Energy Shift, a podcast series focusing on the rapidly evolving energy landscape with hosts Luisa Fuentes and James Wright.

James Wright: Good morning, Luisa. Happy Friday.

Luisa Fuentes: Happy Friday. Good morning, James. How’s your week been? How was Infocast?

James Wright: Great. Yeah, yeah, very good. It feels like flu season’s here. We’ve all been kind of increasing hand sanitizer usage. Had a couple of sick kids, which always makes the week challenging. How about you?

Luisa Fuentes: Yeah, I did not go to Infocast this year, so I think I’m better rested than the contingency of people that we sent over, but also feeling, I’m hoping it’s allergies, and I hate having allergies, but it does give me hope that there is spring around the corner.

James Wright: Exactly. So what have we got today?

Luisa Fuentes: So on todays episode we’re going to be discussing the intersection of AI and energy demand and particularly focusing on the implications of the AI boom on our power infrastructure here in the US and as we discussed, transmission. Our guest today is Darrel Koo. He is a Director of Analytics at Orennia. Orennia is a leading provider of energy transition data and analytics software. They cover power markets as well as transition markets.

James Wright: Great, fantastic. Welcome, Darrel. Great to have you on the show. Let’s get going then with finding out a bit about you, if we could. It’d be great to understand just who you are, what Orennia does in this space, and just give a bit of personal background. How did you end up in power analytics?

Darrel Koo: Yeah, absolutely. It’s great to be here. Thanks for having me. So I’ll try to keep it brief, but I started my career initially as an investment research analyst covering the upstream oil and gas space at a boutique research shop here in Calgary. And it was an interesting time because this was kind of near the tail end of the financial crisis, so like 2008, 2009, and sort of the beginnings of the shale revolution. So it was kind of a transition period in a lot of ways, but ended up staying there for over a decade, but by the time I left, that company had evolved from being just a pure research provider, catering mainly to institutional investors like hedge funds, to a more kind of broad data analytics and research platform serving not just the capital market side of things, but also the energy companies themselves. And so we really found that, you know, both cohorts or both groups were really kind of looking for the same thing, which was to distill this massive amount of information and data that we have in the energy sector, whether that’s oil and gas or power, into investable decisions. So that was sort of the earlier part of my career, but quite a few of us from that company are now at Orennia. And so Orennia started about four years ago, and it’s really a continuation of that same idea of providing a kind of all-in-one data and research platform really catered to energy investors and developers, instead of oil and gas Orennia, really focuses on the power and energy transition sectors. So to bring it to digital infrastructure and data centers, to be fully transparent, we as a company and myself kind of came across this space almost by accident. So we had originally designed this platform primarily for power developers and IPPs. So folks building solar wind projects, battery storage projects, infrastructure investors and so on. So really the supply side of the power markets. But as I’m sure a lot of your audience is aware, the recent acceleration of AI and all the associated load growth has really reshaped power markets. And so we were finding that a lot of our existing power developer clients wanted to understand data centers. So they wanted to understand how these developers thought. Where do they site projects? And really they wanted to know if this load growth was real. You know, they don’t want to pursue speculative load growth, if you will. And so, you know, if it is real, how can they participate? And so we were getting a lot of interest from that side of things, but we’re also getting a lot of interest from the load side of the power markets. So specifically data center developers who are really struggling to find power. And so to kind of summarize, we’ve really seen this convergence in workflows between the data center developers and the power developers. We’ve seen them kind of start to coalesce a little bit. And so now as a result, we have hyperscalers and data center developers who using our platform to help them screen markets and as well do kind of really granular site selection kind of workflows, essentially trying to find where is their interconnection capacity.

James Wright: And we’ve certainly been seeing those trends on our side as well, Darrel, exactly the same stuff you’re talking about. So thank you. That was great. Luisa, I want you to kick us off.

Luisa Fuentes: Yeah, sure. Thank you again, Darrel. I love the story of the evolution of the business model that you’re creating at Orennia. So that’s great because it is actually a platform that’s meeting client needs and not just existing, waiting for clients to come. So I love that. To kick things off, I thought we discussed the shift in data center expansion that has taken place over the last few years. I think you’ve kind of started that down that road already. We have gone, least in the world that we sit in, from digital infrastructure being kind of a buzzy topic around late 2022, 2023, to it becoming a major beneficiary of infrastructure investment and ultimately bank financing. Maybe you could set the table for us here with regard to data-centered development and maybe from there, talk to us where you think the natural knock-on effects happen and could be. You’ve already started to talk about the load needs. As an institution, we tend to be laser focused on power load demand, for example, but are there any other areas we should be thinking about?

Darrel Koo: Yeah, absolutely. So, I think maybe we can just set the table a little bit. And I think it’s worth saying that, you know, data centers are not new, even though they’ve increasingly become a lot more mainstream and kind of talked about in, it almost seems unavoidable, whether that’s just the news or kind of energy, participants, like everybody’s talking about data centers, but they’ve actually existed for quite some time. So well over a decade, close to two decades, but the original use for data centers was really for cloud compute. So really that trend was about offloading data and computing processes that used to be done in kind of small server rooms in our offices to these kind of large warehouses known as data centers that are run by cloud providers like AWS. And so that’s really, that was the first part of the data center story, if you will. And I think they’re still very much a big part of the story. I don’t think there’s really accurate numbers on this, but from what we can understand, cloud compute still represents the vast majority of data center load. However, AI is again, spurring a massive amount of capital influx as well as a lot of new players in the space. And so these recent advancements in AI have really caused big tech to deploy just massive amounts of capital and resources to training these AI models. And it’s a very intensive, energy intensive process that requires very specialized hardware. So GPUs, water cooling, et cetera. And so the result of all of that is that there is a significant step change in the power density. So, you know, basically data centers have always existed, but now there’s more being built and also what’s being contained in these data centers is fundamentally changing. And so just to give you some numbers around this, the average nameplate capacity of a data center that’s currently operating in the US is around 20 megawatts compared to the same kind of metric. So the average nameplate capacity of data centers in development. So future data centers, that’s nearly 200 megawatts. So that gap is roughly 10 times in terms of power consumption. Hopefully that gives you kind of a sense of what’s changing here. And as a result of all of that, utilities, regulators, power developers, folks across the energy value chain, they’re struggling to keep up. So what we’re really seeing is this mismatch between all of that investment on the load side of things and the same sort of amount of investment on the infrastructure side of things, the power infrastructure. Those are just fundamentally mismatched in our mind.

James Wright: Thanks for that, Darrel. That’s fascinating. So drilling down a bit into the impact of data center development and low growth that you were just referring to. Could you elaborate for us and for the listeners on some of those trends you’re seeing, particularly around the impacts on greenfield power development? What are you seeing in that space?

Darrel Koo: Just to circle back a little bit on the last point. So, the result of that mismatch that I spoke about is really that low growth is straining against the grid’s ability to serve it reliably. Right. And, you know, also there’s a very kind of fundamental difference between how big tech operates and utilities operate. Right. I think they work on very different timescales and have very different motivations. So. I think that’s an interesting sort of overarching theme that I think is worth noting. But just to kind of dig into the power side of things or the greenfield power development side of things, I think it’s worth mentioning that the strategies of these big tech companies have really evolved as it relates to energy and environmental strategies. If we rewind a few years ago, so actually not that long ago, how big tech has really approached energy and approached carbon emissions is that they were really doing this annual matching exercise. So they looked at how much power they were consuming on an annual basis, largely from data centers. And they essentially matched that using RECs, renewable energy credits, and largely unbundled RECs. Not necessarily buying the power itself, but just the certificates. that was sort of the earliest version of energy procurement, if you will, from Big Tech. But over time, as you can imagine, there’s been kind of increasing rigor towards this. So, the community pushback started to grow. I think even they themselves realize that, you know, this kind of approach is not necessarily the best way of dealing with carbon emissions. And so now we’re seeing from folks like Google and Microsoft that they’re striving for 24-7 carbon-free energy with locational and hourly matching. So not just matching on an annual basis, but on an hourly basis and even on a grid-by-grid basis. So as you can imagine, this is a very difficult equation to solve. The hyperscalers as a result have been sort of the largest purchasers of renewable energy. largely through these virtual PPAs and based on our kind of public, our data set of public transactions, hyperscalers and data center developers are responsible for procuring at least 30 gigawatts of wind and solar capacity in the U.S. So think that’s actually quite commendable. That’s a massive amount of investment that they’ve poured into this space. Yeah. But as you can kind of imagine, and I think a lot of folks again, in the renewable energy space understand this, but traditional renewable power generation technologies like solar, wind and storage, they have limitations as it relates to achieving these 24-7 carbon-free energy goals. Namely, the sun doesn’t shine all the time, the wind doesn’t blow 24-7. And when you match that up with how data centers run, so the data centers are continuously run, especially these kind of cloud compute critical IT infrastructure facilities. So they have very high load factors. Again, making this a very hard equation to solve. And on top of that available interconnection capacity is very difficult to find, especially in these primary data center markets. So, a lot of variables here, but the result of all of that is that, you know, the, result of all these constraints and all these kinds of motivations is that there’s new kind of solutions being proposed. So there’s novel kind of rate making structures that are being proposed like clean transition tariffs, which, know, if we want, we can go into that a little bit. But the idea there is where, you know, an off taker like Google, for example, will pay a premium for electricity from, let’s say like a next gen geothermal project. And that way they can again sort of make progress towards their 24-7 covered free energy goals and at the same time not kind of have external impacts on the rate base. So that’s a regulatory solution being proposed.

James Wright: A lot of moving parts, as you said. If I can maybe summarize what you were saying, Daryl, is it fair to kind of say then, particularly when you think about the hyperscalers, they want it clean, but they need it firm.

Darrel Koo: That’s right. I think that’s sort of the holy grail is if there were some sort of technology that could provide them with carbon-free energy on a 24-7 basis, that would be perfect. And again, I think that’s kind of why you’re seeing this cohort move towards nuclear power. But obviously, those are few and far between those projects. And so they’re exploring. small modular reactors, next-gen geothermal. Again, so these are like on paper very good, but I think what they lack is that speed and that scale because they’re really trying to build these things very quickly.

James Wright: Which is a perfect segue to what I was going to ask you next, which is, as you said, given some of those technologies, practically speaking many, many years away still from commercial deployment, I should say, are you then like near a term now, are you seeing that that thirst for electrons impacting both renewables demand and this resurgence of CCGTs in equal proportion?

Darrel Koo: In terms of the proportion, I think that’s hard to pin down necessarily, but I think the demand for renewables will likely continue. So again, through these kind of renewable virtual PPAs, I think Big Tech will continue to procure those again, because they have these kind of 24-7 carbon-free energy goals. And even though they’re not perfect, they do help. And we’re also seeing examples of co-locating renewable generation with data centers. So we recently saw an announcement of a partnership between Intersect and Google, which is really to build data center campuses in these industrial parks where you have high wind and solar resource potential. And you can kind of pair that with this data center load. But even in those cases, they have to be supplemented with grid power because of that intermittency problem that we talked about, and or on-site gas generation. We’re not sure exactly what that looks like. I’m not sure there’s going to be a one size fits all solution. But I think a pure renewable solution seems a little impractical. Massively over-sizing your generation relative to the data center capacity seems a little inefficient. And so I think grid power is going to have to come into play and potentially gas as well. And we are seeing growing representation of gas in the US interconnection cues. And we’re hearing about kind of delayed thermal retirements as well in response to these reliability needs.

James Wright: I’m just curious real quick as well, just because you made me think of it when you were talking about some of those challenges. What about behind the meter best battery storage? Is that featuring? You’ve seen that coming through to the formal.

Darrel Koo: Yeah, we are. so the behind the meter sort of concept is increasing when we talked about, again, in the past, it was certainly not the preferred solution because it is relatively expensive. And so I think most folks on the data center side prefer a grid connection. And I think that still is going to be the end game for a lot of these projects. But again, if the data center is serious, if you will, right, like if the off-taker is very serious for having this kind of facility come online and that grid power isn’t available, then I think there’s really no other path other than some behind the meter solution. And exactly what that looks like, what that configuration looks like, I think is still up in the air, but I think, yeah, think BESS will likely play a role as well.

James Wright: Excellent. Thank you.

Luisa Fuentes: Following up on the topic that James just raised, you know, it is a challenge to have non-interruptible power and also have it be green, but challenges often become opportunities. And so the way we’re starting to think about it, and I think you raised SMRs and geothermal previously, is that this drive for kind of green electrons and non-interruptible power could result in stimulus for some of these newer technologies. Maybe just getting your thoughts on things like hydrogen and turbines, carbon capture, et cetera. Where are you seeing the potential opportunities here in the near term and maybe the medium term?

Darrel Koo: It is very exciting for some of these next gen generation technologies because you have this buyer who is very motivated to get carbon free energy, right? That’s baseload in nature or dispatchable in nature and they have the purchasing power, if you will, right? So I think it’s a very perfect kind of environment to accelerate some of these more nascent technologies that we already kind of mentioned like SMRs, carbon capture plus gas and advanced geothermal. And having said that, I think all these hyperscaler companies, have, yes, carbon free energy goals, but they also have strategic goals. So they’re competing with each other. And also there is this national security imperative as well. So I think there’s this very powerful drive as well to get this capacity online as quickly as possible. So in the near term, I think we are ultimately going to see a boost in gas generation. And we’ve seen this already through the Meta and Entergy partnership in Louisiana and the associated gas generation that came along with that. So I think that’s a perfect kind of example. Although I would say often those kind of announcements are paired with a pathway to future CCS. So I think Big Tech does still want, you know, an ultimate sort of real option, if you will, to add CCS to thermal projects. But I think, all of these technologies, whether it’s gas and CCS, SMRs or next-gen geothermal, they’re all attracting a ton of interest from Big Tech because they have those very desirable qualities of high reliability and carbon free qualities or close to carbon free. But again, in a lot of cases, they don’t check the speed and scale boxes. So I think they are making these investments, but they’re not going to fill the void of the near term, I would say.

Luisa Fuentes: As someone who spends a lot of time looking at kind next generation technologies, what I like about the potential here is the fact that there is the demand already. And so in my mind, that tends to drive things to a logical conclusion where if it’s economic, it will it will happen as opposed to coming up with technologies that are interesting and very low carbon, but nobody’s really asking for. Let’s go a bit deeper now, Darrel. Maybe you could break down what the key regions have been so far for data center development. Obviously, Virginia and ERCOT come to mind. And maybe you can address for us why they have been such hot markets. What are some of the key requirements for data center siting? What are some of the secondary regions that you feel could be developed with regard to data center? I’m thinking of Wyoming, for example.

Darrel Koo: Some of the preferred areas for data centers was where land was relatively cheap, favorable tax and regulatory conditions were there, and also access to water for cooling, since that’s a massive kind of resource that’s being used by these facilities as well. And as well, data centers tend to avoid areas with natural hazard risks, floodplains as well, for obvious reasons. those were sort of the initial kind of key characteristics, I would say. And the primary markets, again, that have cropped up as a result of that is, for example, Virginia, that’s the number one spot is the epicenter. It’s really the ideal location in terms of national and international fiber connectivity. It’s close to population centers. And historically, there was a lot of available power and land. However, that is very much not the case anymore. So Virginia is now very much constrained. And even though it’s constrained, there’s still around 40 gigawatts of proposed data center capacity from announced projects. So it’s not really slowing down, I guess is another way to put it. And meanwhile, Texas and Arizona are sort of the runner-ups in terms of data center capacity. And Texas is interesting because I think it has some unique qualities because it’s attractive in the sense that it has a relatively fast interconnection process compared to some of these other markets. And it’s also unique in the sense that there’s a lot of stranded energy assets, like for example, excess natural gas that’s being produced in the Permian Basin in West Texas that has made it particularly attractive to cryptocurrency miners who are looking to capture a bit of an arbitrage profit, if you will. having said all that, that’s sort of the primary market and historical sort of look. Now going forward, these emerging kind of AI workloads where you’re again, training these models, this is a less latency sensitive process as we understand it, which means that additional markets are potentially viable, right? And so we’re starting to see some of these AI data centers crop up in areas like Pennsylvania, Ohio, Indiana, Louisiana, Wyoming, you already mentioned. So these are areas that historically have not really been the most obvious targets for data center development. But again, because of this environment of constrained primary markets and just the need for interconnection capacity and AI being a little bit more regionally flexible if you will some of these other markets have cropped up.

James Wright: For folks like me who’s still kind of barely off a fax machine, when you talk about latency and sensitive, you just mean the physical distance of the hyperscaler from the customer isn’t so sensitive. Is that the kind of point there?

Darrel Koo: Yeah, the point being, if you’re streaming Netflix or YouTube or something like that, the closer you are, the end user is to where that compute is happening or where the servers are, the faster the performance is going to be, which is good. It’s good from the client standpoint, the customer standpoint. Now, going forward, some of these kind AI workloads like training, chat GPT. A lot of that processing does not require back and forth with end users. So that’s kind of known as the training mode, if you will. And so that process can be further away, but you raise an interesting point because ultimately these AI models will have to be used and will have to be deployed, right? And they call that inferencing. So essentially it’s, in my mind, that’s going to be very similar to cloud compute, right? You want that to be fast, right? If you’re querying chat GPT, you don’t really want slow performance there. So I think you still will need to be close to users at the end of the day.

James Wright: That’s helpful. Thank you. Let’s switch gears a bit. So one topic that Luisa and I were talking about a couple of weeks ago was the state of transmission here in the US. So with that, with your data center hat on, as we’ve just been discussing, do you have concerns, Darrel, that the current grid can keep up with the low demands that these data centers are putting on the grid right now? I mean, it feels like just from our seat, there’s still a real disconnect with the billions of dollars needing to be and frankly being spent on this sector at the moment. But the dollars in the ground on the basic wires, right, the transmission network is really still lacking. So do you agree and what’s the impact you’re seeing? What’s the data telling you about?

Darrel Koo: I completely agree. Soagain, as we kind of mentioned earlier, there’s been this mismatch in pace, right? In terms of the investment into data centers, GPUs, et cetera. That’s just massively outpaced investment into transmission or the wires, as you say, as well as the power generation side of things. So the result is that finding interconnection capacity is the number one pain point for data centers. And that’s the number one ball neck. And I think that will continue to be the case for the near future, barring any sort of widespread reform of the power and transmission sectors. I think the grid will continue to be a constraint and that will really kind of govern the pace of data center load growth. And some might say that’s a good thing. Some might say, you know, perhaps we need to be a little bit more cautious in terms of building out infrastructure that can’t be undone. I think, that’s an important thing to note. And as well, what this will really do is I think, sort out the real projects from the speculative ones. So because there’s so much, again, hype and capital flowing into the space, there’s a lot of folks who are just trying to participate in this trend, but don’t even have experience building a data center, quite frankly. And so I think there’s going to be a lot of separation from true projects from less real projects.

Luisa Fuentes: Moving on to kind of the macro environment. I think we can safely say there are a lot of sectors currently feeling uncertainty given the stances of the new administration, some of the executive orders that have been issued, et cetera. However, the potential on AI and data centers specifically seem to be hugely positive. Love to hear your thoughts on this and how you see the market reacting.

Darrel Koo: Yeah, it seems that the new US administration is broadly supportive of AI and the kind of underlying data center trends. So it seems like they’re pointed in the same direction in terms of, you know, what big tech wants and, you know, the interests of national security. So we’ve seen, you know, the Stargate announcement, obviously. So, you know, the US government seems to be sort promoting or facilitating capital flows into digital infrastructure, or at least in the US. However, I would say capital does not really appear to be the constraint here. So I don’t think that more capital necessarily needs to be spent on building data centers. Rather, the constraint is on the power infrastructure side of things. It’s the wires, it’s the supply chain, and it’s not clear me how the change in administration can really unblock these. But having said that, the risk of further regulatory barriers seems low. I think the markets are pricing in a world where AI and data center development is relatively unregulated, at least at the federal level. However, it’s, I think, important to note that there’s growing pushback on the regional or local side of things. A lot of these regional utility regulators, consumer advocacy groups, the communities themselves are starting to push back, particularly in these really saturated markets where initially, you know, it was a good thing having these data centers come in and, you know, add to the, to the tax base. But over time, there’s been, there have been some of these negative externalities, right? Like, you know, the fact that they’re you know, impacting the grid massively and potentially impacting power prices for other repairs, right? So I think these are the sorts of things that might again, modulate or govern data center growth, I would say.

Luisa Fuentes: Interesting. And other things we think about as we look at this very hot space and how it might be tempered is the potential impact for there to be an overbuilt situation with regard to power supporting data centers, particularly as we see new language models, the DeepSeek example obviously being one, but it is a reminder that we’re still in a very young sector that’s impacting all of this hockey stick load growth predictions. How does the sector develop where it could end up in this overbuilt situation? How are developers on the power side thinking about this as they kind of take individual FID decisions?

Darrel Koo: Yeah, it’s a great question. I think this really illustrates how difficult it is to forecast low growth from AI and data centers. So I think the utilities have a really hard job. And I think, announcements like DeepSeek just show that, you know, compared to other forms of low growth, like you know, manufacturing or industrial, it’s really hard to know, you know, how data center power consumption will change, right? Because you have so many moving parts, have efficiencies coming out on the software side, as you mentioned, but as well on the hardware side. So GPUs are getting more advanced and becoming more efficient. So again, it’s going to be very difficult to exactly pin down where this ultimately leads, but we would expect near term capacity, Data center capacity that’s getting built out to be relatively unaffected by these things. Until we start to see project cancellations or rate of change shift in terms of new announcements, I think that’s the thing to watch for. But to your point about overbuilding, think that’s definitely a concern. Again, going back to the regulatory side of things, I think that’s something that the public utility commissions are very concerned about, right? So they’re starting to institute some of these protective mechanisms like large low tariffs that are designed to sort of protect against this overbuild situation where the load doesn’t show up and you have built all this generation capacity, all of this transmission capacity and it’s not being used. Who pays for that, right? And I think that’s really the point of some of these large load tariffs and regulations that are designed to sort of shift the risk and make sure that the big tech kind of players don’t overburden the rest of the repairs.

Luisa Fuentes: Makes a lot of sense. Thanks, Darrel.

James Wright: So finally, Darrel, as we bring this to a wrap, Orennia published a thought piece recently on the role of power market analytics in data center development. Could you give us a quick preview of that and what the key findings were?

Darrel Koo: Yeah, try to summarize it relatively quickly, but we recently published a report on data center siting workflows, kind of leveraging our data sets. So leveraging our data set of known data centers across the U S and Canada, we’ve defined various data center markets. We then layered in our transmission analytics and power markets, forecasts. And the result is that we were able to sort of, quantify risk in terms of power and transmission between these primary markets as well as the secondary market. So, it’s going to be hard to kind of summarize all of that analysis, but essentially the goal of that was to help data center developers screen regions at a macro level. But obviously after doing that, developers have to screen for bus level interconnection capacity, fiber, gas availability regional generation build out from the connection queues. So really, we’re just trying to frame this data set problem in the context of power and transmission. That’s a bit of a summary of what we did there.

James Wright: Sounds great, look forward to reading it. Thank you. All right. So as we draw to a close, Darrel, this is the Energy Shift podcast. So we’d like to finish off with something that shifted all of our weeks. Luisa, why don’t you start us off today?

Luisa Fuentes: I’m focusing today on adaptability and AI itself, which I’ve been very resistant to using. However, I have two kids who are going to be 14 and 12 in May, and I think this is going to be a regular part of all of our lives. So much like we adapted and adapted to cell phones, laptops, iPads, et cetera, I am going to download ChatGPT and I’m going to start getting up to speed. I’m familiar with the tools of the future. Evolve or die.

James Wright: Well said. Our kids are the same age, so I feel your pain there. Darrel, what about you?

Darrel Koo: Yeah, I think mine is pretty similar to Luisa’s and kind of keeps with the theme. So I recently read an article on Wired Magazine about OpenAI’s deep research agent. So admittedly, I was pretty skeptical about some of these kind of lofty expectations put out there about how AI is going to completely revolutionize all of our lives and every aspect of society. But it does seem like these AI tools are getting lot better. over time and you know I myself I do use chat GPT from from time to time as well I find it pretty helpful at certain things but this article talks about how you know compared to you know the models that are currently being used that just generate text some of these new research agents are able to provide more or less like a full-on research report with charts with citations it seems quite interesting I haven’t used it myself but It does seem like maybe even some of our white collar jobs might be at risk. So interesting times.

James Wright: That’s a depressing thought for a Friday. Okay, well, I’ll try and bring it back up again to finish. So look, as Luisa said at the top, I’m fresh back off the plane from Infocast Solar and Wind, the investment summit down in Phoenix. Firstly, it shifted my body. I mean, wow, when you live in Chicago, Phoenix in March is just delightful. But on a more business level, I was really impressed with just the combination of optimism and stoicism across the industry. The wind and solar industries, as we all know, have been through some pretty significant headwinds before this period we’re in feels again like one of those, but the general feeling I already got was one of kind of keep calm and carry on across everyone we met. So that was kind of refreshing. And now I’m back in Chicago where I was greeted by snow and hail storms on landing. So back to reality. So that was fantastic. Thank you, Darrel. Really appreciate it. Some really good insights there, some great statistics, and we’re going to keep really close with that research that you guys are pumping out because it’s trends that we’re seeing on the ground day to day in our business that you’ve been talking about. So thank you again.

Darrel Koo: Thanks James. Thanks, Luisa. Appreciate it.

Outro: Please join us next time on The Energy Shift as we continue to tackle some of the hottest topics in the US energy transition landscape, providing fresh insights and viewpoints to help you shift your perspective.

Disclaimer: The materials disclosed on this podcast are for informational purposes only and subject to our Code of Conduct as well as applicable IIROC and FINRA rules. The information and data contained herein has been obtained or derived from sources believed to be reliable, without independent verification by CIBC Capital Markets and, to the extent that such information and data is based on sources outside CIBC Capital Markets, we do not represent or warrant that any such information or data is accurate, adequate or complete. Notwithstanding anything to the contrary herein, CIBC World Markets Corp (and/or any affiliate thereof) shall not assume any responsibility or liability of any nature in connection with any of the contents of this communication. This communication is tailored for a particular audience and accordingly, this message is intended for such specific audience only. Any dissemination, re-distribution or other use of this message or the market commentary contained herein by any recipient is unauthorized. This communication should not be construed as a research report. The services, securities and investments discussed in this report may not be available to, nor suitable for, all investors. Nothing in this communication constitutes a recommendation, offer or solicitation to buy or sell any specific investments discussed herein. Speakers on this podcast do not have any actual, implied or apparent authority to act on behalf of any issuer mentioned in this podcast. The commentary and opinions expressed herein are solely those of the individual speaker(s), except where the author expressly states them to be the opinions of CIBC World Markets Corp. The speaker(s) may provide short-term trading views or ideas on issuers, securities, commodities, currencies or other financial instruments but investors should not expect continuing analysis, views or discussion relating to those instruments discussed herein. Any information provided herein is not intended to represent an adequate basis for investors to make an informed investment decision and is subject to change without notice. CIBC Capital Markets is a trademark brand name under which Canadian Imperial Bank of Commerce (“CIBC”), its subsidiaries and affiliates provide products and services to our customers around the world. For more information about these legal entities, as well as the products and services offered by CIBC Capital Markets, please visit www.cibccm.com.

Knock, knock. Who’s there? It’s AI

For the full catalog, visit