1 00:00:00,560 --> 00:00:02,960 Speaker 1: This is Tom Rowland Seriice and you're listening to Switched 2 00:00:03,000 --> 00:00:06,320 Speaker 1: on the B and EF podcast. AI Data centers are 3 00:00:06,320 --> 00:00:09,360 Speaker 1: big news these days. They're also big users of electricity, 4 00:00:09,600 --> 00:00:11,520 Speaker 1: and with a whole fleet of data centers set to 5 00:00:11,520 --> 00:00:14,360 Speaker 1: be rolled out across the US, energy sources need to 6 00:00:14,400 --> 00:00:17,880 Speaker 1: be found to power these behemoths and fast. The US 7 00:00:18,000 --> 00:00:20,640 Speaker 1: does have a bit of an energy superpower. However, it's 8 00:00:20,680 --> 00:00:23,240 Speaker 1: a wash with natural gas, and that gas is cheap, 9 00:00:23,400 --> 00:00:25,840 Speaker 1: and the new Trump administration is throwing its weight behind 10 00:00:25,840 --> 00:00:28,760 Speaker 1: production of the fossil fuel. So might gas be a 11 00:00:28,800 --> 00:00:31,880 Speaker 1: domestic resource for powering the new data center load? BNF 12 00:00:32,000 --> 00:00:34,760 Speaker 1: estimates that the current data center pipeline could add sixty 13 00:00:34,840 --> 00:00:37,320 Speaker 1: nine gigawatts of average power demand to the US grid. 14 00:00:37,520 --> 00:00:40,200 Speaker 1: That's equivalent of about eight point four percent of average 15 00:00:40,240 --> 00:00:42,960 Speaker 1: domestic gas consumption in the US last year. On a 16 00:00:43,000 --> 00:00:45,839 Speaker 1: regional level, though, the story is far more nuanced, with 17 00:00:45,920 --> 00:00:50,240 Speaker 1: data center build up plans disproportionately affecting different electricity grids, 18 00:00:50,440 --> 00:00:52,920 Speaker 1: and given the global shortage of gas turbines, could the 19 00:00:53,000 --> 00:00:56,160 Speaker 1: required plants even be built in the first instance. Today, 20 00:00:56,160 --> 00:00:58,960 Speaker 1: I'm joined by B andF gas market analyst Henry Eaton, 21 00:00:59,080 --> 00:01:01,800 Speaker 1: and we discussed five from his note fueling the Cloud 22 00:01:01,960 --> 00:01:05,160 Speaker 1: data Centers bond with US gas, which BNF clients can 23 00:01:05,200 --> 00:01:07,920 Speaker 1: find a BNF go on the Bloomberg terminal or on 24 00:01:08,000 --> 00:01:10,720 Speaker 1: BNF dot com. All right, let's get to talking about 25 00:01:10,800 --> 00:01:22,760 Speaker 1: US gas and data centers with Henry. Henry, welcome to 26 00:01:22,800 --> 00:01:23,199 Speaker 1: the show. 27 00:01:23,360 --> 00:01:24,160 Speaker 2: Thank you for having me. 28 00:01:24,720 --> 00:01:28,039 Speaker 1: So we're talking about gas and data centers today, and 29 00:01:28,440 --> 00:01:31,720 Speaker 1: we've been doing a lot on data centers either directly 30 00:01:31,840 --> 00:01:35,040 Speaker 1: or indirectly in the podcast this year. But for those 31 00:01:35,080 --> 00:01:38,479 Speaker 1: listeners who haven't listened religiously to everything we talk about, 32 00:01:38,520 --> 00:01:42,360 Speaker 1: could you just maybe frame the scale of the challenge 33 00:01:42,360 --> 00:01:45,920 Speaker 1: slash opportunity as it relates to AI and data centers 34 00:01:46,200 --> 00:01:48,120 Speaker 1: and the US energy system. 35 00:01:48,200 --> 00:01:48,520 Speaker 2: Sure. 36 00:01:48,720 --> 00:01:50,480 Speaker 3: So, I think a lot of people have heard and 37 00:01:50,520 --> 00:01:53,400 Speaker 3: you reading the news a lot about how massive this 38 00:01:53,560 --> 00:01:57,520 Speaker 3: AI opportunity is, both from like a workflow and techside, 39 00:01:57,560 --> 00:02:01,720 Speaker 3: but also the infrastructure needs and the money that's pouring 40 00:02:01,720 --> 00:02:02,680 Speaker 3: into infrastructure. 41 00:02:02,920 --> 00:02:04,200 Speaker 2: The information that. 42 00:02:04,120 --> 00:02:06,000 Speaker 3: We've been able to see or the information that we 43 00:02:06,080 --> 00:02:08,800 Speaker 3: use in our research, shows that there's about one hundred 44 00:02:08,919 --> 00:02:12,359 Speaker 3: and twenty eight gigawatts of data center capacity coming online. 45 00:02:12,480 --> 00:02:14,760 Speaker 3: And I just want to frame that this to really 46 00:02:14,800 --> 00:02:17,400 Speaker 3: show the speed and scale of how much that's growing. 47 00:02:17,560 --> 00:02:20,200 Speaker 3: That number from the time that we put out this research, 48 00:02:20,360 --> 00:02:22,440 Speaker 3: or really at the end of the second quarter this year, 49 00:02:22,480 --> 00:02:24,520 Speaker 3: that number has already grown to one hundred and eighty 50 00:02:24,520 --> 00:02:28,360 Speaker 3: four gigallotts. So the scale and speed that this is growing, 51 00:02:28,440 --> 00:02:31,560 Speaker 3: we almost can't keep up. And I think that really 52 00:02:31,600 --> 00:02:33,600 Speaker 3: just goes to show how much of a. 53 00:02:33,960 --> 00:02:35,240 Speaker 2: Pertinent topic this is. 54 00:02:35,480 --> 00:02:38,720 Speaker 3: You know, we have twenty two gigawotts or so of 55 00:02:38,800 --> 00:02:41,680 Speaker 3: live data center capacity, and so one hundred and eighty 56 00:02:41,720 --> 00:02:45,480 Speaker 3: four in theory is you know, almost almost ten times, right. 57 00:02:45,720 --> 00:02:48,799 Speaker 1: That's a massive number. I mean, the listeners who aren't 58 00:02:49,040 --> 00:02:51,639 Speaker 1: familiar with you know, how big is a gigawatt, and 59 00:02:51,720 --> 00:02:55,560 Speaker 1: a nuclear power plant typically has output between depending on 60 00:02:55,600 --> 00:02:58,239 Speaker 1: its size, like one and two gigawatts. So we're talking 61 00:02:58,360 --> 00:03:01,240 Speaker 1: of the scale of like more than one hundred nuclear 62 00:03:01,240 --> 00:03:05,720 Speaker 1: power plants exactly, which is just an unfathomable amount of 63 00:03:05,840 --> 00:03:08,560 Speaker 1: energy in a system that where demand hasn't really been 64 00:03:08,680 --> 00:03:12,480 Speaker 1: growing that much till now. We also, I mean, and 65 00:03:12,480 --> 00:03:15,080 Speaker 1: we've spoken about this on the podcast before, but again 66 00:03:15,120 --> 00:03:17,600 Speaker 1: it's useful for us to sort of recap just for 67 00:03:18,400 --> 00:03:21,040 Speaker 1: the people who haven't listened to everything that we've done 68 00:03:21,040 --> 00:03:25,880 Speaker 1: on this. Where is this pipeline likely to be realized? 69 00:03:26,360 --> 00:03:27,440 Speaker 2: Right, Yeah, that's a good question. 70 00:03:27,520 --> 00:03:31,400 Speaker 3: So the you know, data centers, you see growth across 71 00:03:31,400 --> 00:03:35,560 Speaker 3: the entire country. The scale of that growth varies quite 72 00:03:35,680 --> 00:03:38,840 Speaker 3: vastly between various regions. In our research, where we split 73 00:03:38,880 --> 00:03:41,600 Speaker 3: the data centers up in our data set into five regions, 74 00:03:41,640 --> 00:03:44,280 Speaker 3: which is essentially the East, which is the eastern seabord 75 00:03:44,320 --> 00:03:48,200 Speaker 3: of the US, South Central which is Louisiana, Texas, Oklahoma, 76 00:03:48,600 --> 00:03:51,680 Speaker 3: the Midwest, the Mountain region, which is essentially the Rockies 77 00:03:52,240 --> 00:03:54,880 Speaker 3: in some of the states in the plains, and then the. 78 00:03:54,800 --> 00:03:56,800 Speaker 2: Pacific Coast as a Pacific region. 79 00:03:57,000 --> 00:04:00,200 Speaker 3: To elaborate more on where we see that growth. The 80 00:04:00,240 --> 00:04:04,480 Speaker 3: Eastern region, I think the Eastern Seaboard encompasses Virginia and 81 00:04:04,840 --> 00:04:08,400 Speaker 3: all the states that border the Atlantic. That region houses 82 00:04:08,880 --> 00:04:12,400 Speaker 3: something a little under fifty percent of all data center 83 00:04:12,480 --> 00:04:15,920 Speaker 3: capacity that's live today and in that US in the US, Yeah, 84 00:04:16,040 --> 00:04:19,120 Speaker 3: and that's going to grow to about sixty percent given 85 00:04:19,160 --> 00:04:21,600 Speaker 3: the pipeline that we can see. And so that is 86 00:04:21,640 --> 00:04:25,240 Speaker 3: a region that really really has already dominates the space 87 00:04:25,320 --> 00:04:27,479 Speaker 3: and will as from what we can see will continue 88 00:04:27,520 --> 00:04:29,320 Speaker 3: to do so going into the future. 89 00:04:29,560 --> 00:04:30,800 Speaker 2: There are a couple other like. 90 00:04:31,200 --> 00:04:35,360 Speaker 3: Very interesting dynamics going on where the Texas, Louisiana, Oklahoma 91 00:04:35,480 --> 00:04:39,359 Speaker 3: area right now, excluding the bitcoin mining data centers, is 92 00:04:39,360 --> 00:04:42,960 Speaker 3: the smallest market for live capacity live data center capacity 93 00:04:42,960 --> 00:04:46,039 Speaker 3: in the US. However it's the quickest growing. It's growing 94 00:04:46,080 --> 00:04:48,719 Speaker 3: the quickest, so that's going to become the second largest 95 00:04:48,880 --> 00:04:52,560 Speaker 3: market for data centers. And I think the last interesting 96 00:04:52,800 --> 00:04:57,400 Speaker 3: thing here is the Pacific region, which is Washington, Oregon, California, 97 00:04:57,600 --> 00:05:01,800 Speaker 3: that was the second largest market live today but will 98 00:05:01,800 --> 00:05:05,240 Speaker 3: actually become the smallest market in the future. And so 99 00:05:05,520 --> 00:05:08,040 Speaker 3: what I like to call is it's a mature market. 100 00:05:08,120 --> 00:05:10,600 Speaker 3: It is going to see growth. It's going to essentially double, 101 00:05:10,760 --> 00:05:12,800 Speaker 3: but the other regions in the US are basically going 102 00:05:12,839 --> 00:05:15,440 Speaker 3: to four or five x at the very least, whereas 103 00:05:15,560 --> 00:05:16,160 Speaker 3: the West. 104 00:05:15,920 --> 00:05:17,159 Speaker 2: Coast is really just going to double one. 105 00:05:17,200 --> 00:05:18,920 Speaker 1: I think there are a couple of challenges there that's 106 00:05:18,960 --> 00:05:20,840 Speaker 1: so interesting. I mean, I think just because you know, 107 00:05:20,960 --> 00:05:22,720 Speaker 1: we all think of the West Coast and we think 108 00:05:22,760 --> 00:05:24,640 Speaker 1: of tech, and that's where it's going on, and you 109 00:05:24,680 --> 00:05:26,800 Speaker 1: know that might still be where a lot of the 110 00:05:27,440 --> 00:05:31,560 Speaker 1: I suppose the human intellect will be behind those industries, 111 00:05:31,600 --> 00:05:36,200 Speaker 1: but maybe not the computer hardware. I think that I 112 00:05:36,240 --> 00:05:38,279 Speaker 1: think it's just worth pointing out here as well, particularly 113 00:05:38,320 --> 00:05:40,680 Speaker 1: to people who are outside of the US. You know, 114 00:05:40,680 --> 00:05:42,520 Speaker 1: if this what we're talking about sounds like a very 115 00:05:42,560 --> 00:05:46,600 Speaker 1: local US thing. I think today about fifty percent of 116 00:05:46,720 --> 00:05:50,680 Speaker 1: global data center capacities in the US, and I think 117 00:05:50,680 --> 00:05:54,080 Speaker 1: that ratio stays roughly the same in our projections. So 118 00:05:54,240 --> 00:05:56,159 Speaker 1: you know, when you're talking about maybe half of it 119 00:05:56,200 --> 00:05:59,240 Speaker 1: being on the Eastern Seaboard in the US, that then 120 00:05:59,320 --> 00:06:01,120 Speaker 1: means that's a core sort of the data centers in 121 00:06:01,160 --> 00:06:04,280 Speaker 1: the world. So this is not just something that is 122 00:06:04,640 --> 00:06:07,560 Speaker 1: a interesting question from a US perspective. This is a 123 00:06:07,680 --> 00:06:11,440 Speaker 1: kind of a thing that is a global trend happening 124 00:06:11,720 --> 00:06:14,280 Speaker 1: in the US. That's how I would describe it. Kind 125 00:06:14,279 --> 00:06:16,640 Speaker 1: of last question before we kind of dive more specifically 126 00:06:16,680 --> 00:06:20,960 Speaker 1: into your research and your specialization. But another framing is 127 00:06:20,960 --> 00:06:23,359 Speaker 1: is how big are these data centers that we're talking about. 128 00:06:23,600 --> 00:06:26,080 Speaker 2: Yeah, that's a great question to start off with some numbers. 129 00:06:26,120 --> 00:06:29,159 Speaker 3: I think the average size of a data center today 130 00:06:29,640 --> 00:06:32,920 Speaker 3: is just under twenty megawatts. And you know, we started 131 00:06:32,920 --> 00:06:35,360 Speaker 3: the conversation off by talking about how the pipeline is 132 00:06:35,520 --> 00:06:38,680 Speaker 3: one hundred and eighty four gigawatts, right, The average size 133 00:06:38,680 --> 00:06:42,000 Speaker 3: of these data centers, particularly in various stages, the ones 134 00:06:42,000 --> 00:06:44,400 Speaker 3: that are announced but not built yet, so the early 135 00:06:44,400 --> 00:06:46,919 Speaker 3: stage data centers, they grow to about over two hundred 136 00:06:46,960 --> 00:06:49,920 Speaker 3: and twenty or something megawatts. So not only is our 137 00:06:50,080 --> 00:06:53,400 Speaker 3: pipeline almost going to ten x, but the size or 138 00:06:53,440 --> 00:06:56,039 Speaker 3: the average size of our data center is going to 139 00:06:56,680 --> 00:06:59,120 Speaker 3: really increase about the same it's going to They're they're 140 00:06:59,160 --> 00:07:01,039 Speaker 3: going to mass for me, I. 141 00:07:01,000 --> 00:07:03,640 Speaker 1: Mean a two hundred and twenty men data center, that's 142 00:07:03,680 --> 00:07:06,680 Speaker 1: about a very normal size for like a gas power plant. 143 00:07:06,800 --> 00:07:07,000 Speaker 2: Yeah. 144 00:07:07,279 --> 00:07:11,800 Speaker 1: Yeah, So I mean that segues nicely into your research 145 00:07:11,840 --> 00:07:14,760 Speaker 1: because you're from our US Gas team. Our US Power 146 00:07:14,800 --> 00:07:17,880 Speaker 1: team has already done a bunch of research on data 147 00:07:17,920 --> 00:07:20,920 Speaker 1: centers and the implications on the grid, and we had 148 00:07:21,120 --> 00:07:24,720 Speaker 1: Helen co and Natalie Leamandebrata from that team on the 149 00:07:24,760 --> 00:07:27,880 Speaker 1: podcast earlier this year. You're looking at specifically from a 150 00:07:27,920 --> 00:07:31,480 Speaker 1: gas angle, So as a gas analyst, what's interesting about this? 151 00:07:32,120 --> 00:07:35,120 Speaker 3: Looking at the whole data center boom from a gas perspective, 152 00:07:35,360 --> 00:07:38,440 Speaker 3: I'd say there are two things. One, the developers and 153 00:07:38,520 --> 00:07:43,240 Speaker 3: operators are looking to gas to power their data centers, 154 00:07:43,320 --> 00:07:47,320 Speaker 3: and that's for reliability reasons, base load reasons, right, the 155 00:07:47,360 --> 00:07:51,360 Speaker 3: dispatchable characteristics of gas that make it favorable to power. 156 00:07:51,120 --> 00:07:52,040 Speaker 2: These data centers. 157 00:07:52,240 --> 00:07:56,080 Speaker 3: So there's interest in natural gas as a topic from 158 00:07:56,160 --> 00:08:00,200 Speaker 3: that side. There's also a lot of interest from the 159 00:08:00,280 --> 00:08:05,400 Speaker 3: natural gas market participants, the midstream developers, the power utilities, 160 00:08:05,760 --> 00:08:08,360 Speaker 3: the people who are trading gas every day, who want 161 00:08:08,400 --> 00:08:12,400 Speaker 3: to know how much impact is this revolutionary technology going 162 00:08:12,480 --> 00:08:15,600 Speaker 3: to have on my business and where are the opportunities 163 00:08:15,760 --> 00:08:19,680 Speaker 3: and so naturally, the amount of power that these data 164 00:08:19,720 --> 00:08:22,440 Speaker 3: centers require is also going to call quite a bit 165 00:08:22,520 --> 00:08:24,920 Speaker 3: on natural gas, not just because it calls on the 166 00:08:24,920 --> 00:08:27,960 Speaker 3: power system in general, but also because gas. 167 00:08:27,880 --> 00:08:30,239 Speaker 2: Is favorable for empowering these data centers. 168 00:08:30,440 --> 00:08:33,880 Speaker 1: I question I have is it makes sense that gas 169 00:08:33,920 --> 00:08:37,920 Speaker 1: is well positioned to meet this demand, and the US 170 00:08:38,120 --> 00:08:39,959 Speaker 1: produces a lot of natural gas. Is some of the 171 00:08:40,000 --> 00:08:42,920 Speaker 1: cheapest natural gas in the world. So I can see 172 00:08:42,920 --> 00:08:46,280 Speaker 1: that that is putting a lot of advantages in the 173 00:08:46,400 --> 00:08:49,920 Speaker 1: US's hands when it's thinking about its data center development strategy. 174 00:08:50,080 --> 00:08:53,680 Speaker 1: And so obviously, if these data centers do depend someone 175 00:08:53,720 --> 00:08:55,960 Speaker 1: on gas, that will mean more gas demand to what 176 00:08:56,040 --> 00:08:58,200 Speaker 1: extent is this because I know your team looks at 177 00:08:58,240 --> 00:09:01,280 Speaker 1: the gas balances. You know how much is coming in 178 00:09:01,320 --> 00:09:03,280 Speaker 1: and out of storage, how much is coming out of 179 00:09:03,280 --> 00:09:05,960 Speaker 1: the ground, how much is being consumed. So obviously this 180 00:09:06,040 --> 00:09:08,000 Speaker 1: is going to affect that. But how much of it 181 00:09:08,040 --> 00:09:11,040 Speaker 1: is going to affect the infrastructure, you know, the pipelines, 182 00:09:11,280 --> 00:09:14,040 Speaker 1: the ability to deliver I suppose what I'm really saying 183 00:09:14,120 --> 00:09:15,920 Speaker 1: is how much of this is just a question. Yeah, 184 00:09:15,920 --> 00:09:19,880 Speaker 1: there'll be more gas consumed versus than needing to be 185 00:09:20,480 --> 00:09:23,840 Speaker 1: investment in the infrastructure around gas because of this. 186 00:09:24,440 --> 00:09:27,440 Speaker 3: Yeah, that's a That's an interesting question and a tough 187 00:09:27,440 --> 00:09:29,280 Speaker 3: one to give a direct answer to. 188 00:09:29,440 --> 00:09:31,920 Speaker 2: I'd say, without a doubt, there. 189 00:09:31,840 --> 00:09:34,920 Speaker 3: Is no question that it's going to increase natural gas demand. 190 00:09:34,920 --> 00:09:38,080 Speaker 3: I think that is a foregone conclusion and is known 191 00:09:38,160 --> 00:09:40,280 Speaker 3: by people that are that are looking at it and 192 00:09:40,400 --> 00:09:42,880 Speaker 3: are that have interested in it. The infrastructure side of 193 00:09:42,920 --> 00:09:47,200 Speaker 3: things is a more interesting one. I would say, right now, 194 00:09:47,440 --> 00:09:50,480 Speaker 3: the US has enough infrastructure. We certainly have the gas 195 00:09:50,520 --> 00:09:52,520 Speaker 3: production to get the gas out of the ground, right, 196 00:09:52,600 --> 00:09:54,440 Speaker 3: So it's a matter of can we get it to 197 00:09:54,720 --> 00:09:57,360 Speaker 3: where the data centers are and do the data centers 198 00:09:57,440 --> 00:10:00,319 Speaker 3: have for example, a turbine that can turn that natural 199 00:10:00,320 --> 00:10:01,520 Speaker 3: gas into electricity. 200 00:10:01,679 --> 00:10:03,840 Speaker 2: I think what you'll begin to see. 201 00:10:03,920 --> 00:10:06,120 Speaker 3: And part of the reason why I brought the difference 202 00:10:06,200 --> 00:10:08,240 Speaker 3: between the pipeline that we looked at the time we 203 00:10:08,240 --> 00:10:10,559 Speaker 3: put out this research in the pipeline now has grown 204 00:10:10,559 --> 00:10:13,720 Speaker 3: another fifty percent, is to show that I think there 205 00:10:13,800 --> 00:10:18,120 Speaker 3: will need to be infrastructure upgrades no matter what, especially 206 00:10:18,120 --> 00:10:20,520 Speaker 3: if all of this data center demand does come to 207 00:10:20,679 --> 00:10:23,520 Speaker 3: fruition and it's not just companies announcing that they're going 208 00:10:23,559 --> 00:10:24,959 Speaker 3: to build them and then they don't get built. 209 00:10:25,120 --> 00:10:26,360 Speaker 1: So which is which is possible? 210 00:10:26,360 --> 00:10:27,800 Speaker 2: Which is one hundred percent possible? 211 00:10:28,679 --> 00:10:31,280 Speaker 1: That we've seen pipelines before for different things, and yeah, 212 00:10:31,280 --> 00:10:33,000 Speaker 1: we don't always get one hundred percent of what's in 213 00:10:33,040 --> 00:10:33,640 Speaker 1: the pipeline. 214 00:10:33,720 --> 00:10:38,079 Speaker 3: And I think the interconnection que problem exacerbates that problem 215 00:10:38,160 --> 00:10:40,560 Speaker 3: as well, because really the number one priority for these 216 00:10:40,600 --> 00:10:43,440 Speaker 3: developers to get online as quickly as they can, right, 217 00:10:43,520 --> 00:10:46,240 Speaker 3: So they're going to submit a data center to every 218 00:10:46,280 --> 00:10:49,800 Speaker 3: interconnection possibility that they can just on the off chance 219 00:10:49,840 --> 00:10:51,679 Speaker 3: that they get it. It gives them an opportunity to 220 00:10:51,679 --> 00:10:53,880 Speaker 3: build that. So there might be the amount of capacity 221 00:10:53,880 --> 00:10:54,840 Speaker 3: that we see could be inflated. 222 00:10:54,840 --> 00:10:56,360 Speaker 2: It could not be right, right, right right. 223 00:10:56,400 --> 00:10:58,480 Speaker 1: I think, so we know there's going to be more 224 00:10:58,720 --> 00:11:01,160 Speaker 1: you know, how much is a question, and in a way, 225 00:11:01,240 --> 00:11:02,800 Speaker 1: some of the constraints on the system, and you mentioned 226 00:11:02,800 --> 00:11:05,240 Speaker 1: the interconnection queue, but you're also alluding to the need 227 00:11:05,320 --> 00:11:08,400 Speaker 1: to invest in more gas infrastructure, and if that doesn't happen, 228 00:11:08,520 --> 00:11:11,640 Speaker 1: then some of this pipeline might not materialize. But we'd 229 00:11:11,640 --> 00:11:14,079 Speaker 1: be wrong to say like, oh, well, the infrastructure was 230 00:11:14,120 --> 00:11:16,800 Speaker 1: fine because they never built as much as they were 231 00:11:16,800 --> 00:11:18,640 Speaker 1: going to. But you will say, well, that was because 232 00:11:18,679 --> 00:11:21,520 Speaker 1: the infrastructure was exactly Yeah, given what we've just said, 233 00:11:21,679 --> 00:11:24,199 Speaker 1: maybe this is a dumb question. Expecting to be able 234 00:11:24,280 --> 00:11:26,560 Speaker 1: to answer this. It's hard to say exactly how much 235 00:11:26,640 --> 00:11:28,640 Speaker 1: is going to get built. But you mentioned that they'll 236 00:11:28,679 --> 00:11:31,960 Speaker 1: be more gas consumed in the US because of this. 237 00:11:32,400 --> 00:11:34,840 Speaker 1: A question I have is, do you mean there'll be 238 00:11:34,880 --> 00:11:38,240 Speaker 1: more gas consumed compared to if there hadn't been the 239 00:11:38,280 --> 00:11:41,680 Speaker 1: data centers, which seems fairly uncontroversial because it's an additional 240 00:11:41,720 --> 00:11:43,600 Speaker 1: source of demand, or do you mean there'll be more 241 00:11:43,640 --> 00:11:46,280 Speaker 1: gas full stop? Because if we look out far enough 242 00:11:46,280 --> 00:11:48,520 Speaker 1: from the future, currently you know gas is displacing coal, 243 00:11:48,960 --> 00:11:52,160 Speaker 1: increases gas demand. Let's leave exports to the side, because 244 00:11:52,200 --> 00:11:54,520 Speaker 1: that is an increasing source of demand for gas. But 245 00:11:54,640 --> 00:11:57,480 Speaker 1: so we're just looking at us domestic consumption because the 246 00:11:57,480 --> 00:12:00,000 Speaker 1: other thing that's happening is renewables are getting built, which 247 00:12:00,000 --> 00:12:03,760 Speaker 1: which is displacing coal, and eventually we'll start displacing gas. 248 00:12:03,840 --> 00:12:07,240 Speaker 1: So in the long run that downward pressure on gas demand, 249 00:12:07,320 --> 00:12:10,320 Speaker 1: I suppose I'm saying, do we think that data centers 250 00:12:10,320 --> 00:12:13,040 Speaker 1: more than offset that to the point where domestic gas 251 00:12:13,080 --> 00:12:15,080 Speaker 1: demand is growing, It just means that it's going to 252 00:12:15,120 --> 00:12:16,320 Speaker 1: be more than would have been. 253 00:12:16,600 --> 00:12:19,120 Speaker 2: That's an interesting question. I don't have the perfect answer 254 00:12:19,120 --> 00:12:19,360 Speaker 2: for you. 255 00:12:19,440 --> 00:12:22,080 Speaker 3: I think that's going to come down to whether we 256 00:12:22,200 --> 00:12:26,199 Speaker 3: build out the power infrastructure to connect these data centers 257 00:12:26,240 --> 00:12:28,240 Speaker 3: to the grid. So I think if they're connected to 258 00:12:28,280 --> 00:12:31,760 Speaker 3: the grid, I think that natural progression that you explain, 259 00:12:31,840 --> 00:12:35,640 Speaker 3: where gas replaces coal and then renewables and batteries start 260 00:12:35,679 --> 00:12:38,120 Speaker 3: to eat into that gas share, I'd say that you'd 261 00:12:38,160 --> 00:12:42,360 Speaker 3: see gas's share of the fuel mix begin to fall 262 00:12:42,640 --> 00:12:45,160 Speaker 3: far out into the future. Right now, what we see 263 00:12:45,400 --> 00:12:47,760 Speaker 3: is I think in twenty twenty four, gas represented about 264 00:12:47,800 --> 00:12:48,719 Speaker 3: forty three percent of. 265 00:12:48,640 --> 00:12:50,439 Speaker 2: The fuel mix in the power sector. 266 00:12:50,440 --> 00:12:53,240 Speaker 3: In the power sector, right and that has grown in 267 00:12:53,240 --> 00:12:55,840 Speaker 3: the past couple of years. So renewables are eating part 268 00:12:55,840 --> 00:12:59,240 Speaker 3: of that share, but gas is gaining more from replacing coal. 269 00:13:00,120 --> 00:13:02,840 Speaker 1: Without this data center demand. Where gas is growing. 270 00:13:02,600 --> 00:13:05,600 Speaker 3: Gas is growing either way, right, and so that's an 271 00:13:05,640 --> 00:13:08,760 Speaker 3: important thing to note. I'd also say, like back on 272 00:13:08,800 --> 00:13:12,440 Speaker 3: your infrastructure question, if data centers can't connect to the 273 00:13:12,440 --> 00:13:15,760 Speaker 3: grid and they decide to go for gas generation, they're 274 00:13:15,760 --> 00:13:18,600 Speaker 3: not disconnecting from the grid, They're just connecting to a 275 00:13:18,720 --> 00:13:20,559 Speaker 3: different grid, which is the gas grid. 276 00:13:21,520 --> 00:13:25,319 Speaker 1: So you're now talking not about data centers creating more 277 00:13:25,360 --> 00:13:27,560 Speaker 1: demand for power from the grid and some of that 278 00:13:27,640 --> 00:13:29,880 Speaker 1: will be met by gas plants connected to the grid. 279 00:13:30,040 --> 00:13:34,600 Speaker 1: You're talking about data centers having their own gas power supplies. 280 00:13:34,720 --> 00:13:35,640 Speaker 1: Is that what you're getting at? 281 00:13:35,760 --> 00:13:38,920 Speaker 3: That is what I'm getting at, if that's feasible for 282 00:13:39,000 --> 00:13:41,400 Speaker 3: these developers. And I don't want to get into the 283 00:13:41,440 --> 00:13:44,839 Speaker 3: question yet of the whole turbine unavailability, because that's a 284 00:13:44,880 --> 00:13:46,720 Speaker 3: whole separate story, right, So if they let's say that. 285 00:13:46,920 --> 00:13:49,000 Speaker 1: Some listeners don't know about that yet, but there's a 286 00:13:49,000 --> 00:13:51,000 Speaker 1: whole other story about turbine availability. 287 00:13:51,120 --> 00:13:54,400 Speaker 3: Yes, I'd say that because speed is so important and 288 00:13:54,480 --> 00:13:57,760 Speaker 3: interconnection times in certain regions are so long, that there 289 00:13:57,960 --> 00:14:02,240 Speaker 3: is a question that these operator and developers are asking themselves, 290 00:14:02,240 --> 00:14:04,440 Speaker 3: which is, can we use gas to get online quicker? 291 00:14:04,480 --> 00:14:06,680 Speaker 3: And it may not be like a permanent solution, Maybe 292 00:14:06,679 --> 00:14:08,680 Speaker 3: it's just a temporary solution, but it is a question 293 00:14:08,720 --> 00:14:10,040 Speaker 3: that they're asking themselves for sure. 294 00:14:10,120 --> 00:14:12,840 Speaker 1: I mean it's a really interesting thought though, because you know, 295 00:14:12,920 --> 00:14:17,520 Speaker 1: we have these two parallel energy delivery systems in the US. 296 00:14:17,600 --> 00:14:19,600 Speaker 1: You could say, which is the power grid and the 297 00:14:19,640 --> 00:14:22,120 Speaker 1: gas network? And I know they're not the only two 298 00:14:22,320 --> 00:14:24,880 Speaker 1: energy delivery systems, but let's just talk about these two things. 299 00:14:24,960 --> 00:14:28,000 Speaker 1: They interact with each other, but they are also independent 300 00:14:28,040 --> 00:14:30,400 Speaker 1: from each other to some extent. The picture you're painting 301 00:14:30,520 --> 00:14:34,360 Speaker 1: is if that first energy delivery mechanism, the grid, is 302 00:14:34,920 --> 00:14:37,880 Speaker 1: too much of a constraint, then data centers might go 303 00:14:38,000 --> 00:14:41,400 Speaker 1: straight to the second one and kind of rely on that. 304 00:14:41,440 --> 00:14:44,160 Speaker 1: So then that leads me to the question, how does 305 00:14:44,200 --> 00:14:47,400 Speaker 1: that influence the location of these data centers? Because I 306 00:14:47,400 --> 00:14:51,280 Speaker 1: imagine that this the gas grid has more capacity in 307 00:14:51,320 --> 00:14:52,520 Speaker 1: some parts than others. 308 00:14:52,680 --> 00:14:53,800 Speaker 2: Yeah, that's absolutely correct. 309 00:14:53,920 --> 00:14:58,520 Speaker 3: We've seen things get more constrained in the gas grid, generally, 310 00:14:58,680 --> 00:15:01,480 Speaker 3: specifically coming out of some the large production regions as 311 00:15:01,760 --> 00:15:02,920 Speaker 3: those producers try to meet. 312 00:15:02,920 --> 00:15:04,160 Speaker 2: Look a f natural gas to man. 313 00:15:04,360 --> 00:15:08,480 Speaker 3: You'd see in certain regions, for example, in northern Virginia, 314 00:15:08,560 --> 00:15:12,720 Speaker 3: or data center ally, as it's often coined, where a vast, 315 00:15:12,800 --> 00:15:15,880 Speaker 3: vast majority of capacity is going. I think I said 316 00:15:16,000 --> 00:15:18,640 Speaker 3: half of future capacity is going to the east. About 317 00:15:18,760 --> 00:15:21,200 Speaker 3: half of that eastern capacity is going to Virginia. 318 00:15:21,400 --> 00:15:22,600 Speaker 1: So key, lucky Virginia. 319 00:15:22,680 --> 00:15:24,760 Speaker 3: If you do another half a half, right, let's call 320 00:15:24,800 --> 00:15:27,880 Speaker 3: it a half of us capacity. Another half is in Virginia, 321 00:15:28,040 --> 00:15:29,880 Speaker 3: so let's call it a quarter of us. And then 322 00:15:29,960 --> 00:15:32,480 Speaker 3: I don't know, an eighth of the world capacity, eighth. 323 00:15:32,280 --> 00:15:35,080 Speaker 1: Of eighth of the world date cent capacity is in 324 00:15:35,160 --> 00:15:36,200 Speaker 1: one little stay. 325 00:15:36,000 --> 00:15:36,640 Speaker 2: In one state. 326 00:15:36,920 --> 00:15:39,560 Speaker 3: So I mean pipelines there like we see that if 327 00:15:39,760 --> 00:15:44,040 Speaker 3: let's say all future just for just theoretic all future 328 00:15:44,240 --> 00:15:46,880 Speaker 3: data center capacity says, Okay, the wait time's too long. 329 00:15:47,040 --> 00:15:48,920 Speaker 3: We can get a gas turbine and we're going to 330 00:15:48,960 --> 00:15:49,920 Speaker 3: connect to the gas grid. 331 00:15:50,120 --> 00:15:51,040 Speaker 2: I don't know if that's. 332 00:15:50,920 --> 00:15:54,320 Speaker 3: A feasible solution for them, because pipelines there, they already 333 00:15:54,360 --> 00:15:57,560 Speaker 3: fill up during times of seasonally high demand. Your your 334 00:15:57,600 --> 00:16:00,200 Speaker 3: cold peaks in the winter, you're super hot peaks in them. 335 00:16:00,480 --> 00:16:03,760 Speaker 3: So those pipelines are full and the gas flows generally 336 00:16:03,800 --> 00:16:06,960 Speaker 3: north to south through Virginia through through you know, three 337 00:16:07,240 --> 00:16:11,720 Speaker 3: major pipeline systems there. Building more infrastructure there, particularly in 338 00:16:11,800 --> 00:16:14,760 Speaker 3: east in these developed states where land is. You know, 339 00:16:14,920 --> 00:16:16,920 Speaker 3: once you get into the Washington, DC area, you can 340 00:16:16,960 --> 00:16:18,880 Speaker 3: imagine it's probably not so easy to put a pipeline 341 00:16:18,880 --> 00:16:22,640 Speaker 3: through it. That's an extremely extremely difficult feat to achieve 342 00:16:22,680 --> 00:16:24,800 Speaker 3: and takes years and years and years. So this data 343 00:16:24,840 --> 00:16:28,239 Speaker 3: center capacity, they're all trying to come online much quicker 344 00:16:28,240 --> 00:16:30,480 Speaker 3: than you can build the infrastructure if you can get 345 00:16:30,480 --> 00:16:34,600 Speaker 3: approved for it, right, So there are constraints there for sure, 346 00:16:34,960 --> 00:16:39,520 Speaker 3: especially if these operators decide to go for like temporary 347 00:16:39,560 --> 00:16:42,000 Speaker 3: gas solutions. You know, you can bunch up a small 348 00:16:42,080 --> 00:16:45,400 Speaker 3: bunch of small gas generators, I think enough. Helen and 349 00:16:45,480 --> 00:16:49,560 Speaker 3: Natalie's research that you mentioned earlier estimated and through surveys, 350 00:16:49,720 --> 00:16:52,800 Speaker 3: found that the interconnection time is seven years, right, Virginia. 351 00:16:52,920 --> 00:16:55,520 Speaker 3: That's a long time to wait for a competitive industry 352 00:16:55,600 --> 00:16:57,920 Speaker 3: like AI. If you want to come online as quickly 353 00:16:57,920 --> 00:17:00,440 Speaker 3: as possible, are you willing to wait seven years? 354 00:17:00,800 --> 00:17:01,000 Speaker 4: Yeah? 355 00:17:01,040 --> 00:17:04,440 Speaker 1: I mean it's not just a competitive industry between different companies, 356 00:17:04,480 --> 00:17:08,200 Speaker 1: but it's competitive between different countries. You know, in the US, 357 00:17:08,280 --> 00:17:13,239 Speaker 1: AI is seen as a strategic industry, you know, with 358 00:17:13,400 --> 00:17:17,920 Speaker 1: maybe even national security. And the biggest competitor is China. 359 00:17:18,600 --> 00:17:21,800 Speaker 1: And China's grid is it growing and expanding has been 360 00:17:21,840 --> 00:17:24,480 Speaker 1: for decades, So I mean very naively, and I'm sure 361 00:17:24,480 --> 00:17:26,720 Speaker 1: someone on our China team will say this as an oversimplification, 362 00:17:26,800 --> 00:17:28,480 Speaker 1: but you know, I was a little bit of extra 363 00:17:28,520 --> 00:17:31,560 Speaker 1: demand when you're all already growing x percent a year 364 00:17:31,680 --> 00:17:34,440 Speaker 1: compared to the US, where both for power and gas, 365 00:17:34,600 --> 00:17:37,199 Speaker 1: is maybe like you're more at a standing start on 366 00:17:37,320 --> 00:17:37,800 Speaker 1: this stuff. 367 00:17:37,880 --> 00:17:38,920 Speaker 2: Yeah, it's so interesting. 368 00:17:38,960 --> 00:17:40,720 Speaker 1: I mean, one of the things that in Helen and 369 00:17:40,800 --> 00:17:45,560 Speaker 1: Natali's research they've spoken about is that energy ideal locations 370 00:17:45,640 --> 00:17:48,679 Speaker 1: versus the locations that are ideal from everything else that 371 00:17:48,800 --> 00:17:52,440 Speaker 1: data center developers might care about. So connection to the well, 372 00:17:52,440 --> 00:17:54,800 Speaker 1: obviously everywhere can connect to the internet pretty much, but 373 00:17:54,840 --> 00:17:57,080 Speaker 1: you know, where are you going to get the bandwidth? 374 00:17:57,240 --> 00:17:59,919 Speaker 1: Where is the local skills? You know, there's a kind 375 00:17:59,920 --> 00:18:02,640 Speaker 1: of clustering effect, and that's why so many in Northern 376 00:18:02,680 --> 00:18:05,160 Speaker 1: Virginia now from a power sector perspective, if you didn't 377 00:18:05,200 --> 00:18:07,280 Speaker 1: care about any of that stuff, there are regions of 378 00:18:07,320 --> 00:18:10,000 Speaker 1: the country where the wind resources are so great, power 379 00:18:10,040 --> 00:18:13,199 Speaker 1: prices get solow. There's a bunch of electricity that's just 380 00:18:13,280 --> 00:18:16,080 Speaker 1: waiting to be kind of mopped up by a data 381 00:18:16,080 --> 00:18:17,960 Speaker 1: center if someone would just build one there, And there's 382 00:18:17,960 --> 00:18:20,480 Speaker 1: the reasons why they don't, but they might do one 383 00:18:20,560 --> 00:18:23,760 Speaker 1: day when these areas like Virginia just become saturated. So 384 00:18:23,760 --> 00:18:27,320 Speaker 1: I'm curious to know what is the gas equivalent of 385 00:18:27,560 --> 00:18:29,640 Speaker 1: the west of the Rocky Mountains, Because if you're saying 386 00:18:29,680 --> 00:18:32,960 Speaker 1: that the kind of the gas delivery grid is constrained 387 00:18:32,960 --> 00:18:35,520 Speaker 1: in many ways, then there must be regions where you 388 00:18:35,600 --> 00:18:38,399 Speaker 1: have cheap gas that can't get out. They're just waiting 389 00:18:38,440 --> 00:18:40,679 Speaker 1: for someone to eat it up. So, yeah, where is 390 00:18:40,720 --> 00:18:41,359 Speaker 1: that in the US? 391 00:18:41,480 --> 00:18:42,120 Speaker 2: Yeah? 392 00:18:42,160 --> 00:18:44,879 Speaker 3: Well, for the gas, that's the Permeuan Basin, which is 393 00:18:44,920 --> 00:18:48,840 Speaker 3: an oil production basin in West Texas and New Mexico. Essentially, 394 00:18:49,080 --> 00:18:53,200 Speaker 3: the drilling economics are favorable to drill oil and gas 395 00:18:53,240 --> 00:18:55,560 Speaker 3: is a byproduct of that, and so they actually have 396 00:18:55,680 --> 00:18:58,400 Speaker 3: all this excess gas that oftentimes they can't even get 397 00:18:58,440 --> 00:19:00,359 Speaker 3: out of the premium basin in the They have to 398 00:19:00,400 --> 00:19:03,240 Speaker 3: flare it and sometimes to sell it it's negative prices, 399 00:19:03,280 --> 00:19:05,199 Speaker 3: so they're paying someone to take it away from them. 400 00:19:05,280 --> 00:19:05,399 Speaker 2: Right. 401 00:19:05,480 --> 00:19:08,800 Speaker 3: Wow, So if you could get paid to consume energy 402 00:19:09,080 --> 00:19:11,280 Speaker 3: to power your data center, that's a pretty good deal. 403 00:19:11,600 --> 00:19:14,800 Speaker 2: Like you said, though, there are a lot more factors. 404 00:19:14,240 --> 00:19:17,439 Speaker 3: That these developers need to consider before they go to 405 00:19:17,520 --> 00:19:21,800 Speaker 3: that route. I'd say regardless, Texas is a pretty interesting market, 406 00:19:22,080 --> 00:19:25,240 Speaker 3: probably pretty favorable for new builds right now. I think 407 00:19:25,240 --> 00:19:27,119 Speaker 3: that's why back to one of my first points was 408 00:19:27,119 --> 00:19:30,520 Speaker 3: that Texas and Louisiana is essentially the fastest growing market 409 00:19:30,520 --> 00:19:33,000 Speaker 3: in the US because you may not have negative gas 410 00:19:33,040 --> 00:19:35,440 Speaker 3: prices if you're near Dallas, but they're going to be 411 00:19:35,480 --> 00:19:39,040 Speaker 3: pretty cheap. Right Building new infrastructure there as much easier. 412 00:19:39,080 --> 00:19:40,600 Speaker 2: They've been doing it for years. They still do it. 413 00:19:40,640 --> 00:19:43,640 Speaker 3: We build a lot of infrastructure, particularly to meet new 414 00:19:43,760 --> 00:19:48,320 Speaker 3: which look finnatural Estiman and so there's more pipeline capacity 415 00:19:48,359 --> 00:19:50,400 Speaker 3: being built to carry that gas out of the Permian, 416 00:19:50,600 --> 00:19:53,480 Speaker 3: and it has Internet exchange points. So it's like maybe 417 00:19:53,480 --> 00:19:56,000 Speaker 3: you don't build directly in the Permian, but you build 418 00:19:56,000 --> 00:19:57,040 Speaker 3: trice even if. 419 00:19:56,960 --> 00:19:58,760 Speaker 2: You build close to it. I think you get some 420 00:19:58,800 --> 00:20:00,320 Speaker 2: of those benefit fits. 421 00:20:00,800 --> 00:20:04,280 Speaker 1: Anyways, so in a world where gas is the defining 422 00:20:04,359 --> 00:20:06,400 Speaker 1: fuel for data censers, which we you know, we may 423 00:20:06,440 --> 00:20:10,160 Speaker 1: well be entering into that kind of tips the scales 424 00:20:10,160 --> 00:20:11,840 Speaker 1: in favor of Texas being where a lot of this 425 00:20:11,840 --> 00:20:12,600 Speaker 1: stuff gets built. 426 00:20:12,760 --> 00:20:13,439 Speaker 2: Yeah, I think so. 427 00:20:13,600 --> 00:20:18,040 Speaker 3: I think for flexibility and look, I mean interconnection times 428 00:20:18,040 --> 00:20:19,919 Speaker 3: and Texas are still quite long. But if you can 429 00:20:19,920 --> 00:20:21,680 Speaker 3: get your hands on a gas turbine, it's a pretty 430 00:20:21,680 --> 00:20:22,680 Speaker 3: favorable place today. 431 00:20:22,880 --> 00:20:24,479 Speaker 1: If you can get your hands on a gas turbine. 432 00:20:24,560 --> 00:20:27,960 Speaker 1: Let's unpack that a little bit. Why is that an if? 433 00:20:28,480 --> 00:20:32,720 Speaker 3: Well, it's an if because the suppliers, the builders of 434 00:20:32,760 --> 00:20:37,520 Speaker 3: these gas turbines have backlog order sheets for years and 435 00:20:37,800 --> 00:20:40,680 Speaker 3: I'm not an expert on this topic at all. Our 436 00:20:40,680 --> 00:20:44,240 Speaker 3: team is doing research with another BNF supply chain team 437 00:20:44,440 --> 00:20:46,679 Speaker 3: to put out more in depth research on. 438 00:20:47,280 --> 00:20:48,840 Speaker 2: What's going on and what's going on there. 439 00:20:49,160 --> 00:20:52,040 Speaker 3: It was when we were creating this current research that's 440 00:20:52,080 --> 00:20:54,760 Speaker 3: trying to estimate guests demand and power demand and where 441 00:20:54,760 --> 00:20:57,840 Speaker 3: they're going in location with infrastructure, that we were like, well, 442 00:20:57,840 --> 00:20:59,120 Speaker 3: maybe they can't even get them. 443 00:20:59,240 --> 00:21:02,080 Speaker 1: It's so interesting because there's all these different ways that 444 00:21:02,320 --> 00:21:05,359 Speaker 1: data centers in the future could be powered, and you know, 445 00:21:05,400 --> 00:21:07,840 Speaker 1: and it could be that some things meet the demand 446 00:21:07,880 --> 00:21:09,359 Speaker 1: in the near term and other things meet them in 447 00:21:09,400 --> 00:21:12,920 Speaker 1: the long term, But it seems like every possibility has 448 00:21:13,160 --> 00:21:16,240 Speaker 1: it's like an obstacle course between say, you know, renewables 449 00:21:16,280 --> 00:21:18,399 Speaker 1: and gas. So you have two people racing on an 450 00:21:18,400 --> 00:21:22,480 Speaker 1: obstacle course, but they face different obstacles. So renewables, it's like, 451 00:21:22,640 --> 00:21:26,000 Speaker 1: grid constraints are a bigger issue. There's the intermittency that's 452 00:21:26,359 --> 00:21:28,920 Speaker 1: that's a challenge that needs to be resolved. With gas, 453 00:21:29,040 --> 00:21:30,960 Speaker 1: you don't have you know, if you're going on site, 454 00:21:30,960 --> 00:21:33,720 Speaker 1: you don't have grid constraint issues. There's can the gas 455 00:21:33,760 --> 00:21:36,159 Speaker 1: network support it? You know, there's a location, right, I 456 00:21:36,160 --> 00:21:38,359 Speaker 1: guess that replies renewables as well. But then there's like 457 00:21:38,400 --> 00:21:40,640 Speaker 1: can you get the turbine? Yeah, So it's like there's 458 00:21:40,680 --> 00:21:45,080 Speaker 1: this race between these options, and there's unresolved issues in 459 00:21:45,119 --> 00:21:49,440 Speaker 1: both pathways, and maybe which issues get resolved quickest might 460 00:21:49,560 --> 00:21:51,280 Speaker 1: determine what the outcome is. 461 00:21:51,720 --> 00:21:53,680 Speaker 2: I think that's exactly what's going to determine what the 462 00:21:53,720 --> 00:21:54,240 Speaker 2: outcome is. 463 00:21:54,440 --> 00:21:57,000 Speaker 3: What I tell people if I'm talking to friends or 464 00:21:57,040 --> 00:21:59,320 Speaker 3: family about work and they ask me about it, I 465 00:21:59,520 --> 00:22:02,639 Speaker 3: always say, we can only build data centers and therefore 466 00:22:02,680 --> 00:22:05,680 Speaker 3: develop AI as quickly as we can power these things. Yeah, 467 00:22:05,800 --> 00:22:09,840 Speaker 3: and so it's very much a power constrained industry. 468 00:22:09,520 --> 00:22:11,760 Speaker 2: And really in a time where you don't want. 469 00:22:11,640 --> 00:22:14,000 Speaker 3: It to be constrained, especially if we go back to 470 00:22:14,359 --> 00:22:18,480 Speaker 3: it's a competition between companies, but it's also competition between countries. 471 00:22:18,160 --> 00:22:20,360 Speaker 2: And all of a sudden, you're years and years old. 472 00:22:20,359 --> 00:22:24,040 Speaker 3: Infrastructure is what's preventing you from being able to get 473 00:22:24,080 --> 00:22:24,560 Speaker 3: a leg up. 474 00:22:24,600 --> 00:22:28,520 Speaker 1: Maybe it's an interesting point. I mean, in our world 475 00:22:28,560 --> 00:22:31,320 Speaker 1: as benf analysts, so many of the questions have changed 476 00:22:31,480 --> 00:22:33,679 Speaker 1: in the last few years, from you know, what's going 477 00:22:33,760 --> 00:22:37,679 Speaker 1: to decarbonize the energy system to now, what's going to 478 00:22:37,680 --> 00:22:40,800 Speaker 1: meet all this new demand? And the cliched answer which 479 00:22:40,800 --> 00:22:44,280 Speaker 1: I always found frustrating when people would talk about decombonization 480 00:22:44,359 --> 00:22:46,879 Speaker 1: and say, oh, we're in favor of all of the above, 481 00:22:47,000 --> 00:22:49,800 Speaker 1: which for me, if someone said that a conference, I 482 00:22:49,840 --> 00:22:51,879 Speaker 1: was like, that's code for you don't have an answer 483 00:22:51,880 --> 00:22:54,560 Speaker 1: to the question, because there's surely more you can say 484 00:22:54,560 --> 00:22:58,120 Speaker 1: than all of the above. But I find myself thinking about, 485 00:22:58,200 --> 00:23:00,399 Speaker 1: you know, how we power data centers and you know 486 00:23:00,480 --> 00:23:03,000 Speaker 1: this pressing need to meet demand is maybe it's all 487 00:23:03,040 --> 00:23:06,240 Speaker 1: of the above. But it's interesting that the US government 488 00:23:06,320 --> 00:23:10,359 Speaker 1: currently is dialing back support for certain types of energy 489 00:23:10,640 --> 00:23:14,240 Speaker 1: without necessarily turning the dial up for others as much 490 00:23:14,280 --> 00:23:16,400 Speaker 1: as they might like you to think, or like might 491 00:23:16,520 --> 00:23:19,679 Speaker 1: like the fossil fuel industry to think. So the foot 492 00:23:19,760 --> 00:23:23,720 Speaker 1: is not necessarily on the accelerator for the energy system 493 00:23:23,760 --> 00:23:28,080 Speaker 1: as much as maybe one might expect, given how much 494 00:23:28,200 --> 00:23:31,360 Speaker 1: central importance this as as a topic. Yeah, I want 495 00:23:31,359 --> 00:23:35,560 Speaker 1: to talk about one other kind of strategic topic relating 496 00:23:35,560 --> 00:23:39,720 Speaker 1: to gas, which is liquefied natural gas energy. You already 497 00:23:39,880 --> 00:23:42,760 Speaker 1: alluded to it a little bit, but that is something 498 00:23:42,800 --> 00:23:45,280 Speaker 1: that the current administration, and to be honest with you, 499 00:23:45,359 --> 00:23:48,480 Speaker 1: I mean, the previous administration kind of flirted and chopped 500 00:23:48,480 --> 00:23:50,400 Speaker 1: and changed on where it stood on energy a little bit. 501 00:23:50,520 --> 00:23:53,280 Speaker 1: But certainly at the moment in government has a consensus 502 00:23:53,320 --> 00:23:55,679 Speaker 1: that that you know, the US destiny is to be 503 00:23:55,680 --> 00:23:58,639 Speaker 1: an exporter of LNG and that will be not just 504 00:23:58,960 --> 00:24:01,560 Speaker 1: lucrative for the kind of but it will also serve 505 00:24:01,640 --> 00:24:04,679 Speaker 1: the country's strategic interests. And so that has kind of 506 00:24:04,720 --> 00:24:07,600 Speaker 1: been a source of growth for US gas demand as 507 00:24:07,680 --> 00:24:11,800 Speaker 1: the export terminals have developed. So how much does this 508 00:24:12,160 --> 00:24:16,359 Speaker 1: potential growth in demand from data centers compare to the 509 00:24:16,480 --> 00:24:18,800 Speaker 1: potential growth in gas demand from LNG. 510 00:24:19,480 --> 00:24:20,840 Speaker 2: Yeah, that's a good question. 511 00:24:21,040 --> 00:24:23,800 Speaker 3: I'll start with staying the stage for how big LNG 512 00:24:24,040 --> 00:24:27,879 Speaker 3: is in the US right now. As of September twenty 513 00:24:27,880 --> 00:24:32,480 Speaker 3: twenty five, LNG exports are about sixteen billion cubic feet 514 00:24:32,480 --> 00:24:36,679 Speaker 3: per day. To put that into perspective, it's around I 515 00:24:36,720 --> 00:24:42,359 Speaker 3: think sixteen percent of average domestic gas production or just average. 516 00:24:41,960 --> 00:24:43,479 Speaker 2: Demand in the US. 517 00:24:43,640 --> 00:24:44,320 Speaker 1: And it's growing. 518 00:24:44,560 --> 00:24:47,359 Speaker 3: Our estimates is essentially going to double by twenty thirty, 519 00:24:47,480 --> 00:24:49,160 Speaker 3: that's four or five years from now. 520 00:24:49,359 --> 00:24:50,680 Speaker 1: For wow, double twenty thirty. 521 00:24:51,280 --> 00:24:54,040 Speaker 3: For the next five years, it's going to double to thirty, 522 00:24:54,240 --> 00:24:56,240 Speaker 3: which is going to make it the second largest and 523 00:24:56,480 --> 00:24:59,240 Speaker 3: that's you know, thirty bcfd at the peak in the 524 00:24:59,240 --> 00:25:01,640 Speaker 3: winter in twenty thirty is still going. 525 00:25:01,560 --> 00:25:02,919 Speaker 2: To be up there right right. 526 00:25:03,280 --> 00:25:06,800 Speaker 3: It's going to become our second largest demand source in 527 00:25:06,880 --> 00:25:09,320 Speaker 3: the US. And just to remind you, I mean, even 528 00:25:09,400 --> 00:25:11,320 Speaker 3: like a little more than ten years ago, we were 529 00:25:11,359 --> 00:25:15,560 Speaker 3: importing energy right right, and now we're the world's largest 530 00:25:15,560 --> 00:25:18,320 Speaker 3: exporter of LERG. So's it's going to double, right from 531 00:25:18,400 --> 00:25:20,960 Speaker 3: fifteen sixteen bcfd do around thirty by twenty thirty. We 532 00:25:21,000 --> 00:25:23,560 Speaker 3: have all sorts of new energy terminals coming online, and 533 00:25:23,600 --> 00:25:26,560 Speaker 3: they've been planning in the works no matter what administration 534 00:25:26,640 --> 00:25:28,639 Speaker 3: you look at, and there is a push to use 535 00:25:28,680 --> 00:25:31,159 Speaker 3: it as a bargaining chip for trade deals and whatnot. 536 00:25:31,240 --> 00:25:34,000 Speaker 3: So that's our estimate for look a financial gas demand. 537 00:25:34,119 --> 00:25:37,040 Speaker 3: It is the greatest demand driver in the US for 538 00:25:37,119 --> 00:25:39,800 Speaker 3: the US gas markets and really puts the most pressure 539 00:25:39,960 --> 00:25:43,880 Speaker 3: for new production as well. If we look at data centers, 540 00:25:44,119 --> 00:25:46,760 Speaker 3: our estimate based on the pipeline that we could see 541 00:25:47,040 --> 00:25:51,160 Speaker 3: two quarters ago is almost seven billion cubic feed per day. 542 00:25:51,440 --> 00:25:57,200 Speaker 3: So we're talking about LNG growing fifteen bcfd doing cubic 543 00:25:57,200 --> 00:25:59,600 Speaker 3: feed per day in the next five years. And I'm 544 00:25:59,600 --> 00:26:02,359 Speaker 3: saying data centers, as big as they are, are going 545 00:26:02,400 --> 00:26:05,359 Speaker 3: to be seven billion cubic for you per day, and 546 00:26:05,400 --> 00:26:06,640 Speaker 3: that's probably later than the. 547 00:26:06,560 --> 00:26:08,360 Speaker 2: Next five years. It takes a while to build. 548 00:26:08,200 --> 00:26:10,879 Speaker 1: These things, so it's not as much, but it is 549 00:26:10,960 --> 00:26:12,840 Speaker 1: in the same ballpark. It's not a rounding era. 550 00:26:13,040 --> 00:26:16,680 Speaker 3: It's not insignificant. Seven bcfd is not insignificant at all. 551 00:26:17,640 --> 00:26:19,679 Speaker 3: That's a number that people have to pay attention to. 552 00:26:19,800 --> 00:26:22,240 Speaker 3: And again, the pipeline's growing, right. I said it was 553 00:26:22,240 --> 00:26:23,800 Speaker 3: one hundred and twenty eight when I looked at it, 554 00:26:23,840 --> 00:26:25,080 Speaker 3: and now it's one hundred and eighty four. 555 00:26:25,320 --> 00:26:27,960 Speaker 1: So it's interesting because as I reflect on this, you know, 556 00:26:27,960 --> 00:26:30,440 Speaker 1: I mentioned earlier that the US is at a standing 557 00:26:30,520 --> 00:26:33,240 Speaker 1: start when it comes to sort of energy growth compared 558 00:26:33,240 --> 00:26:36,520 Speaker 1: to China, which has been growing for a long while, 559 00:26:36,560 --> 00:26:38,360 Speaker 1: and so you know, naively you might say, well, what's 560 00:26:38,400 --> 00:26:40,040 Speaker 1: a little bit more growth, But when you look at 561 00:26:40,040 --> 00:26:43,480 Speaker 1: it from that perspective of gas and LNG growth, actually 562 00:26:43,480 --> 00:26:45,400 Speaker 1: the US is not at a standing start. There's this 563 00:26:45,680 --> 00:26:49,120 Speaker 1: huge trend of growth already happening, and then maybe there's 564 00:26:49,119 --> 00:26:52,080 Speaker 1: something to be said for piggybacking off of that. I mean, 565 00:26:52,119 --> 00:26:56,640 Speaker 1: do you see a scenario where these pipelines and infrastructure 566 00:26:56,640 --> 00:26:59,439 Speaker 1: that is being built to take gas from where it's 567 00:26:59,440 --> 00:27:02,359 Speaker 1: being originated US to bring it to LNG export terminals 568 00:27:02,640 --> 00:27:05,560 Speaker 1: that might be where also there's going to be a 569 00:27:05,560 --> 00:27:08,960 Speaker 1: new kind of clustering of data centers, because yeah, it's 570 00:27:08,960 --> 00:27:10,760 Speaker 1: just taking it, you know, taking advantage of where the 571 00:27:10,800 --> 00:27:11,480 Speaker 1: highways are. 572 00:27:11,760 --> 00:27:13,480 Speaker 3: Yeah. One of the things I miss when talking about 573 00:27:13,560 --> 00:27:16,159 Speaker 3: LNG for those who maybe aren't aware, is all of 574 00:27:16,200 --> 00:27:18,960 Speaker 3: that demand. All these export terminals are on the Gulf coast, right, 575 00:27:19,000 --> 00:27:21,800 Speaker 3: They're in Louisiana and they're in Texas. So to get 576 00:27:21,800 --> 00:27:24,560 Speaker 3: that additional fifteen billion tw week feet per day of 577 00:27:24,600 --> 00:27:26,080 Speaker 3: gas that they're going to need by twenty thirty. We 578 00:27:26,119 --> 00:27:28,399 Speaker 3: need to build a good amount of pipeline capacity. And 579 00:27:28,400 --> 00:27:31,280 Speaker 3: these are pipelines that have been planned for many years 580 00:27:31,280 --> 00:27:33,520 Speaker 3: and are coming online now and we'll continue to come 581 00:27:33,560 --> 00:27:35,560 Speaker 3: online incrementally through the next five years. 582 00:27:35,600 --> 00:27:37,320 Speaker 2: So we're building quite a good amount. 583 00:27:37,080 --> 00:27:40,920 Speaker 3: Of pipeline capacity in Texas, in Louisiana to get gas 584 00:27:41,000 --> 00:27:44,720 Speaker 3: out of the Permium basin, which is favorable gas prices, 585 00:27:44,880 --> 00:27:49,200 Speaker 3: to bring those to LNG terminals, the Hainesville production region, 586 00:27:49,240 --> 00:27:52,840 Speaker 3: which is in eastern Texas, and Louisiana, which is more 587 00:27:52,840 --> 00:27:55,280 Speaker 3: expensive gas that also feeds into. 588 00:27:55,000 --> 00:27:56,080 Speaker 2: A lot of those terminals there. 589 00:27:56,080 --> 00:27:59,280 Speaker 3: So we're building you know, essentially maybe just under but 590 00:27:59,440 --> 00:28:03,479 Speaker 3: fifteen b cfd of pipeline capacity as well. The supply 591 00:28:03,600 --> 00:28:05,720 Speaker 3: has to get to the term end somehow, right, So 592 00:28:05,800 --> 00:28:09,000 Speaker 3: I think it goes back to my point of Texas 593 00:28:09,000 --> 00:28:11,560 Speaker 3: being a favorable place moving forward. It's going to say, 594 00:28:11,560 --> 00:28:14,600 Speaker 3: all right, lead to Texas, lead to Texas in the 595 00:28:14,640 --> 00:28:15,600 Speaker 3: gas industry and. 596 00:28:16,160 --> 00:28:19,399 Speaker 1: Maybe in the gas powered data center yeah field as well. 597 00:28:19,400 --> 00:28:21,800 Speaker 1: I mean I realized, well, maybe you should say all 598 00:28:21,840 --> 00:28:25,200 Speaker 1: pipelines lead to Texas. Yeah, so even you know, because 599 00:28:25,240 --> 00:28:27,440 Speaker 1: I don't want to oversimplify this and say that if 600 00:28:27,520 --> 00:28:30,439 Speaker 1: gas wins out as the thing that's going to fuel 601 00:28:30,480 --> 00:28:32,879 Speaker 1: this data center growth, then they're all going to be 602 00:28:32,880 --> 00:28:34,960 Speaker 1: in Texas because we talked about how there's you know, 603 00:28:35,040 --> 00:28:37,640 Speaker 1: growth potential on the on the East coast, et cetera. 604 00:28:37,720 --> 00:28:40,600 Speaker 1: But it does seem that there's a real opportunity there. 605 00:28:41,040 --> 00:28:42,960 Speaker 2: I think there's a real opportunity there. 606 00:28:43,160 --> 00:28:44,959 Speaker 3: They're not all going to go there because again, if 607 00:28:45,080 --> 00:28:46,800 Speaker 3: they all go there, you're going to have the same thing. 608 00:28:46,680 --> 00:28:49,120 Speaker 2: That's you're going to have, right, they just have to 609 00:28:49,120 --> 00:28:49,600 Speaker 2: spread out. 610 00:28:49,760 --> 00:28:50,000 Speaker 1: Really. 611 00:28:50,320 --> 00:28:53,720 Speaker 3: I think for a long time Texas was underdeveloped. It 612 00:28:53,840 --> 00:28:56,920 Speaker 3: was the smallest region. Right now for a live capacity, again, 613 00:28:56,920 --> 00:28:58,000 Speaker 3: it's going to be the second largest. 614 00:28:58,040 --> 00:29:00,400 Speaker 1: So it's kind of like there was this way of 615 00:29:00,560 --> 00:29:05,960 Speaker 1: bitcoin mining facilities and maybe that was the foreshadowing. Yeah, 616 00:29:06,840 --> 00:29:09,200 Speaker 1: you know, bitcoin mining is a bit simpler than AI 617 00:29:09,280 --> 00:29:11,600 Speaker 1: data centers. You don't quite have many constraints. 618 00:29:11,840 --> 00:29:14,680 Speaker 3: Yeah, and bitcoin mining they have the luxury of not 619 00:29:14,840 --> 00:29:18,680 Speaker 3: needing or necessarily caring as much about for example, wait 620 00:29:18,760 --> 00:29:22,400 Speaker 3: and see it serve your application on your phone or 621 00:29:22,480 --> 00:29:25,840 Speaker 3: someone's computer. Right, They can go to the Permium basins 622 00:29:25,880 --> 00:29:28,320 Speaker 3: actually are the areas with lots of solar and wind 623 00:29:28,360 --> 00:29:31,080 Speaker 3: generation in West Texas and say we'll just take that 624 00:29:31,200 --> 00:29:34,280 Speaker 3: power to mind bitcoin. So there are more considerations for 625 00:29:34,720 --> 00:29:37,600 Speaker 3: someone trying to develop, for example, a large language model. 626 00:29:38,320 --> 00:29:41,120 Speaker 1: This has been really interesting conversation and this picture is 627 00:29:41,120 --> 00:29:43,520 Speaker 1: growing in my mind of this obstacle course being run 628 00:29:43,560 --> 00:29:46,040 Speaker 1: in parallel, and you know, one of the tracks on 629 00:29:46,080 --> 00:29:49,760 Speaker 1: that course is the gas pathway, and it's really fascinating 630 00:29:49,840 --> 00:29:51,959 Speaker 1: to see, you know, what the future could look like 631 00:29:52,040 --> 00:29:55,240 Speaker 1: and how that changes the dynamics. And we are internally, 632 00:29:55,280 --> 00:29:56,960 Speaker 1: we look at all of this collectively. You know, it's 633 00:29:56,960 --> 00:29:59,240 Speaker 1: not just Henry puts his head in the sand and 634 00:29:59,280 --> 00:30:01,440 Speaker 1: thinks about gas and our power team just thinks about 635 00:30:01,480 --> 00:30:03,200 Speaker 1: renewable as we're looking at it all together. But I 636 00:30:03,240 --> 00:30:08,000 Speaker 1: think there's some really interesting competitive dynamics and different hurdles 637 00:30:08,000 --> 00:30:09,800 Speaker 1: they face that we're going to have to be keeping 638 00:30:09,800 --> 00:30:13,080 Speaker 1: our eye on. And so there's been a really illuminating conversation. Henry, 639 00:30:13,160 --> 00:30:15,080 Speaker 1: So thank you so much for joining today, Thank you 640 00:30:15,160 --> 00:30:15,600 Speaker 1: for having me. 641 00:30:24,680 --> 00:30:27,800 Speaker 4: Today's episode of Switched On was produced by Cam Gray 642 00:30:28,000 --> 00:30:31,720 Speaker 4: with production assistants from Kamala shelling. Bloomberg NIF is a 643 00:30:31,760 --> 00:30:34,840 Speaker 4: service provided by Bloomberg Finance LP and its affiliates. This 644 00:30:34,960 --> 00:30:37,640 Speaker 4: recording does not constitute, nor should it be construed, as 645 00:30:37,680 --> 00:30:41,640 Speaker 4: investment advice, investment recommendations, or a recommendation as to an 646 00:30:41,640 --> 00:30:44,840 Speaker 4: investment or other strategy. Bloomberg ANIFF should not be considered 647 00:30:44,880 --> 00:30:48,160 Speaker 4: as information sufficient upon which to base an investment decision. 648 00:30:48,280 --> 00:30:51,240 Speaker 4: Neither Bloomberg Finance Lp Nor any of its affiliates makes 649 00:30:51,280 --> 00:30:55,000 Speaker 4: any representation or warranty as to the accuracy or completeness 650 00:30:55,000 --> 00:30:58,000 Speaker 4: of the information contained in this recording, and any liability 651 00:30:58,040 --> 00:30:59,960 Speaker 4: as a result of this recording is expressed. 652 00:31:00,080 --> 00:31:00,680 Speaker 2: Leek is quiet 653 00:31:09,400 --> 00:31:09,680 Speaker 1: MHM