1 00:00:00,360 --> 00:00:02,719 Speaker 1: This is Tom Rowland's Reese and you're listening to Switched 2 00:00:02,759 --> 00:00:07,240 Speaker 1: on the BNF podcast. Today we're discussing BNF's new Energy Outlook, 3 00:00:07,280 --> 00:00:10,160 Speaker 1: our annual flagship report that represents our long term energy 4 00:00:10,160 --> 00:00:13,000 Speaker 1: and climate scenarios for the transition to a low carbon economy. 5 00:00:13,280 --> 00:00:16,360 Speaker 1: AI data centers have dominated conversations in the clean energy 6 00:00:16,360 --> 00:00:19,360 Speaker 1: sector of late, with forecasts varying wildly over just how 7 00:00:19,400 --> 00:00:22,120 Speaker 1: much extra load there set to place on already strained 8 00:00:22,160 --> 00:00:24,960 Speaker 1: power grids. As with every hot button topic, comes the 9 00:00:25,040 --> 00:00:28,120 Speaker 1: risk of hype versus reality. While data center energy demand 10 00:00:28,160 --> 00:00:31,440 Speaker 1: is unquestionably growing at speed, it currently represents just one 11 00:00:31,480 --> 00:00:34,120 Speaker 1: point four percent of global power demand and only around 12 00:00:34,159 --> 00:00:37,600 Speaker 1: two percent of that figure is actually consumed by AI facilities. 13 00:00:37,960 --> 00:00:40,600 Speaker 1: The demand itself is also uneven, with some regions like 14 00:00:40,640 --> 00:00:43,440 Speaker 1: the US needing significantly large amounts of power than others. 15 00:00:43,520 --> 00:00:45,040 Speaker 1: And when it comes to meeting the needs of these 16 00:00:45,120 --> 00:00:48,120 Speaker 1: energy intensive facilities, can the development of clean power sources 17 00:00:48,159 --> 00:00:50,559 Speaker 1: even be done at the scale and pace required. To 18 00:00:50,640 --> 00:00:53,080 Speaker 1: learn more about the forecast for data center energy demand, 19 00:00:53,120 --> 00:00:55,680 Speaker 1: I'm joined by b and EF's head of Energy Systems Modeling, 20 00:00:55,720 --> 00:00:58,440 Speaker 1: Ian Berrman to discuss findings from the twenty twenty five 21 00:00:58,640 --> 00:01:01,360 Speaker 1: edition of our new Energy Outlook. BNF plans can find 22 00:01:01,360 --> 00:01:04,000 Speaker 1: the full report at BNF go on the Bloomberg terminal 23 00:01:04,080 --> 00:01:06,160 Speaker 1: or on BNF dot com, and if you're not yet 24 00:01:06,160 --> 00:01:08,399 Speaker 1: a client, you can also download the executive summary at 25 00:01:08,440 --> 00:01:10,840 Speaker 1: BNEF dot com. All right, let's get to talking about 26 00:01:10,840 --> 00:01:20,959 Speaker 1: the impact of data centers on this year's NEO with Ian. 27 00:01:23,000 --> 00:01:24,600 Speaker 2: Ian. Welcome to the podcast. 28 00:01:24,840 --> 00:01:26,839 Speaker 3: Thanks Tom. So Ian has. 29 00:01:26,760 --> 00:01:27,880 Speaker 2: Been on this podcast before. 30 00:01:27,959 --> 00:01:31,280 Speaker 1: We had a freewheeling discussion around the power system as 31 00:01:31,280 --> 00:01:35,520 Speaker 1: a whole. He had this crazy metaphor involving multiple people 32 00:01:35,720 --> 00:01:38,280 Speaker 1: on like a bike, I mean, as in millions of 33 00:01:38,280 --> 00:01:40,520 Speaker 1: people all on the same bike like this is not 34 00:01:40,560 --> 00:01:44,200 Speaker 1: a very relatable metaphor. But okay, I got more messages 35 00:01:44,240 --> 00:01:47,800 Speaker 1: on LinkedIn saying what a great episode after that particular 36 00:01:47,840 --> 00:01:50,680 Speaker 1: one than I've ever had before, So I'm sure this's 37 00:01:50,680 --> 00:01:52,000 Speaker 1: gonna be a fascinating discussion. 38 00:01:52,360 --> 00:01:53,440 Speaker 2: I'm going to talk about NEO. 39 00:01:53,960 --> 00:01:57,680 Speaker 1: So Ian tell us about NEO, which is obviously something 40 00:01:57,720 --> 00:01:59,800 Speaker 1: as well that we hype a lot at BNEF, but 41 00:02:00,160 --> 00:02:01,920 Speaker 1: so spell out what any means for those of us 42 00:02:02,000 --> 00:02:03,080 Speaker 1: that are not familiar with it. 43 00:02:03,720 --> 00:02:07,440 Speaker 4: NEO is the New Energy Outlook and it's benef's flagship publication. 44 00:02:07,800 --> 00:02:11,840 Speaker 4: It's our house view on the future of the energy transition. 45 00:02:12,440 --> 00:02:15,280 Speaker 4: We've been publishing it for many, many years now, since 46 00:02:15,320 --> 00:02:17,920 Speaker 4: before I started, so seven or eight years ago, and 47 00:02:18,120 --> 00:02:21,040 Speaker 4: in some ways it's the culmination of everything we do 48 00:02:21,080 --> 00:02:23,760 Speaker 4: because we are looking at the entire energy transition. We're 49 00:02:23,800 --> 00:02:26,160 Speaker 4: bringing in work from all our colleagues and all the 50 00:02:26,240 --> 00:02:29,800 Speaker 4: various different sectors that touch on that into one comprehensive 51 00:02:29,960 --> 00:02:32,280 Speaker 4: and coordinated house view on the. 52 00:02:32,200 --> 00:02:33,519 Speaker 3: Future of energy and the transition. 53 00:02:33,919 --> 00:02:36,840 Speaker 1: Awesome, and so I kind of have this view of 54 00:02:36,880 --> 00:02:40,200 Speaker 1: it as this it's like an orchestra and we're pulling 55 00:02:40,240 --> 00:02:43,720 Speaker 1: together everything that BNFS done in this coordinated way. And 56 00:02:43,800 --> 00:02:46,040 Speaker 1: so where do you sit in this? AREI the composer, 57 00:02:46,160 --> 00:02:48,799 Speaker 1: the conductor? Are you the star violinist? 58 00:02:50,040 --> 00:02:50,480 Speaker 2: How do you read? 59 00:02:50,600 --> 00:02:52,080 Speaker 1: Why is it that I'm talking to you on this 60 00:02:52,120 --> 00:02:54,720 Speaker 1: podcast and not someone else? That's a good question. I'm 61 00:02:55,000 --> 00:02:56,840 Speaker 1: I'm trying to figure out where to go with this analogy. Now, 62 00:02:57,360 --> 00:03:00,000 Speaker 1: I'm definitely not the conductor that would be my boss. 63 00:03:00,400 --> 00:03:02,440 Speaker 1: Do I make the instruments? Possibly? 64 00:03:02,720 --> 00:03:03,959 Speaker 3: Possibly I write the music. 65 00:03:04,280 --> 00:03:07,079 Speaker 4: Maybe that's that's the way so I sometimes will play 66 00:03:07,120 --> 00:03:09,959 Speaker 4: along in an instrument when when and where I'm needed. 67 00:03:10,040 --> 00:03:12,240 Speaker 4: But the Energy Systems modeling team, which I head up, 68 00:03:12,240 --> 00:03:14,720 Speaker 4: our role is to work on the models which go 69 00:03:14,760 --> 00:03:17,120 Speaker 4: into producing this. So I guess, I guess we make 70 00:03:17,160 --> 00:03:18,959 Speaker 4: the instruments that other people play well. 71 00:03:19,240 --> 00:03:21,280 Speaker 1: I feel like I mean, Firstly, obviously the people who 72 00:03:21,320 --> 00:03:24,240 Speaker 1: make the instruments are fundamental to the orchestra, even if 73 00:03:24,240 --> 00:03:26,600 Speaker 1: they don't get the limelight. So maybe I'm glad we're 74 00:03:26,600 --> 00:03:29,080 Speaker 1: having you on this podcast so you can get your flowers. Finally, 75 00:03:29,240 --> 00:03:31,160 Speaker 1: I feel like you might be being a little modest 76 00:03:31,280 --> 00:03:35,040 Speaker 1: because the music that we create with NEO, I think 77 00:03:35,320 --> 00:03:38,440 Speaker 1: very few people understand it in its completeness more than 78 00:03:38,480 --> 00:03:40,080 Speaker 1: you do, which is which is why we have you 79 00:03:40,120 --> 00:03:43,760 Speaker 1: on today. So Neo, it's this outlook to twenty fifty, 80 00:03:43,800 --> 00:03:45,240 Speaker 1: a comprehensive look at everything. 81 00:03:45,560 --> 00:03:46,800 Speaker 2: And obviously, if we're going to. 82 00:03:46,840 --> 00:03:50,040 Speaker 1: Do that every year and just update our view on 83 00:03:50,080 --> 00:03:53,360 Speaker 1: a twenty five year twenty five year plus forecast, it's 84 00:03:53,440 --> 00:03:55,960 Speaker 1: not you know, you question why we're doing it, because 85 00:03:56,120 --> 00:03:58,000 Speaker 1: not that much can Chette have changed or how we 86 00:03:58,000 --> 00:04:00,800 Speaker 1: see twenty fifty between one year and the next. So 87 00:04:00,920 --> 00:04:03,800 Speaker 1: we always do new things to kind of keep it fresh, 88 00:04:03,840 --> 00:04:06,120 Speaker 1: keep it interesting, have some other insight to say. 89 00:04:06,240 --> 00:04:08,800 Speaker 2: So this year, what would you say the focus was? 90 00:04:09,200 --> 00:04:11,440 Speaker 4: So this year, which is probably not going to surprise 91 00:04:11,480 --> 00:04:14,720 Speaker 4: anyone that's listening, the hot topic has been data centers 92 00:04:14,760 --> 00:04:18,320 Speaker 4: and AI, so we went away. We developed our own 93 00:04:18,320 --> 00:04:22,599 Speaker 4: in house short and long term forecasts for electricity demand 94 00:04:22,600 --> 00:04:25,200 Speaker 4: from data centers and AI. That's a big feature in 95 00:04:25,240 --> 00:04:29,080 Speaker 4: the report. We refreshed our entire base case, which we 96 00:04:29,120 --> 00:04:32,520 Speaker 4: call the economic Transition scenario, which is not exactly a 97 00:04:32,520 --> 00:04:35,560 Speaker 4: business as usual scenario. And it's probably worth unpacking that 98 00:04:35,760 --> 00:04:39,080 Speaker 4: very briefly because I think economic Transition scenario really reflects 99 00:04:39,120 --> 00:04:42,560 Speaker 4: what makes I think NEO unique and BNF's approach to 100 00:04:42,640 --> 00:04:45,560 Speaker 4: producing this type of global analysis unique, And what makes 101 00:04:45,560 --> 00:04:48,080 Speaker 4: the Economic Transition scenario unique is that we're taking an 102 00:04:48,080 --> 00:04:51,240 Speaker 4: economics led approach, and we're doing this informed where we 103 00:04:51,320 --> 00:04:54,800 Speaker 4: can by bottom up modeling, so very detailed and granular modeling. 104 00:04:54,880 --> 00:04:57,240 Speaker 4: And when we say we're putting economics in front, that 105 00:04:57,279 --> 00:05:00,040 Speaker 4: means we're trying to strip out policy, particularly if that 106 00:05:00,040 --> 00:05:03,360 Speaker 4: that's aspirational, any sort of targets which are non binding, 107 00:05:03,400 --> 00:05:05,920 Speaker 4: et cetera. And we're really trying to hone in on 108 00:05:06,000 --> 00:05:09,320 Speaker 4: the technology and economics story, and that this leaves us 109 00:05:09,320 --> 00:05:12,000 Speaker 4: with a story where economics are in the driving seat 110 00:05:12,080 --> 00:05:13,520 Speaker 4: and technology is a story. 111 00:05:13,800 --> 00:05:16,279 Speaker 1: Right, it's almost I mean, this is a fair way 112 00:05:16,320 --> 00:05:19,960 Speaker 1: to describe it. What we're painting there is a picture 113 00:05:20,000 --> 00:05:23,320 Speaker 1: of a world that follows what we would consider to 114 00:05:23,360 --> 00:05:28,600 Speaker 1: be an economically rational ideal without the interference of inconvenient 115 00:05:28,720 --> 00:05:33,479 Speaker 1: human beings who are either pushing for more low carbon 116 00:05:33,520 --> 00:05:38,720 Speaker 1: technologies to reflect the challenges or the crisis should say 117 00:05:38,720 --> 00:05:43,279 Speaker 1: of climate change or other ideologies, or one could even 118 00:05:43,320 --> 00:05:46,080 Speaker 1: say global dysfunctions that can get in a way of 119 00:05:46,160 --> 00:05:47,760 Speaker 1: maybe that pure economic vision. 120 00:05:48,120 --> 00:05:48,800 Speaker 3: Yeah, exactly. 121 00:05:48,839 --> 00:05:54,000 Speaker 4: And policy can push to carbonization faster or slower than 122 00:05:54,080 --> 00:05:57,240 Speaker 4: what the economics would tell you is least cost. And 123 00:05:57,279 --> 00:05:59,680 Speaker 4: as soon as you try and produce like a four 124 00:05:59,760 --> 00:06:03,040 Speaker 4: car if you will, and you're making judgment calls on 125 00:06:03,120 --> 00:06:06,600 Speaker 4: what policies are announced, when, how long they last for, etc. 126 00:06:07,200 --> 00:06:10,680 Speaker 4: You're really putting your analysis on shaky ground. So by 127 00:06:10,800 --> 00:06:13,640 Speaker 4: focusing on that economic store, it's always useful to know 128 00:06:13,680 --> 00:06:15,599 Speaker 4: what the least cost system is, even if you don't 129 00:06:15,640 --> 00:06:17,080 Speaker 4: know what politics are going to be. And I don't 130 00:06:17,120 --> 00:06:18,839 Speaker 4: think many of us know what politics are going to 131 00:06:18,880 --> 00:06:19,960 Speaker 4: be at the moment. 132 00:06:20,279 --> 00:06:24,560 Speaker 1: We do forecast some sort of human I say, brilliant 133 00:06:25,279 --> 00:06:27,240 Speaker 1: if I might, in terms of we look, we do 134 00:06:27,279 --> 00:06:30,360 Speaker 1: take into account cost declines of technology and the sort 135 00:06:30,360 --> 00:06:35,200 Speaker 1: of technological progress we expect to have happen. But we yeah, 136 00:06:35,240 --> 00:06:39,119 Speaker 1: we don't forecast the politics. We don't forecast either people 137 00:06:39,160 --> 00:06:41,960 Speaker 1: being able to miraculously get on the same page. And 138 00:06:42,000 --> 00:06:45,359 Speaker 1: we don't forecast in which particular way people are dividing 139 00:06:45,400 --> 00:06:48,280 Speaker 1: among themselves and tearing each other apart, which is obviously 140 00:06:48,279 --> 00:06:50,600 Speaker 1: impossible to forecast. So it's kind of like a north 141 00:06:50,680 --> 00:06:51,920 Speaker 1: star in a certain sense. 142 00:06:52,279 --> 00:06:54,080 Speaker 4: North starts a good way to put it, and may 143 00:06:54,080 --> 00:06:55,760 Speaker 4: I also don't want to give people the impression that 144 00:06:55,760 --> 00:06:58,400 Speaker 4: it's sort of too cold and rational and is not 145 00:06:58,560 --> 00:07:01,200 Speaker 4: based on the real world necessarily. And I think that 146 00:07:01,320 --> 00:07:03,520 Speaker 4: there's two main reasons why that's not the cases. First, 147 00:07:03,760 --> 00:07:05,680 Speaker 4: I mean, from our sector teams, we do have a 148 00:07:05,760 --> 00:07:08,200 Speaker 4: very good handle on what we think will happen in 149 00:07:08,320 --> 00:07:11,320 Speaker 4: the next at least at least five years for most sectors. 150 00:07:11,600 --> 00:07:14,360 Speaker 4: We've got huge project databases that BNF is very very 151 00:07:14,360 --> 00:07:17,520 Speaker 4: famous for renewable assets and other types of asset classes, 152 00:07:17,560 --> 00:07:20,119 Speaker 4: so we know what's coming and that feeds into the model, 153 00:07:20,120 --> 00:07:22,480 Speaker 4: and it's sort of there's a transition away from what 154 00:07:22,720 --> 00:07:24,880 Speaker 4: is known and fixed in the short term for the 155 00:07:24,880 --> 00:07:26,880 Speaker 4: next few years, and then what the model proposes to 156 00:07:26,920 --> 00:07:29,520 Speaker 4: do after that, and that's sort of a gradual transition 157 00:07:29,600 --> 00:07:30,800 Speaker 4: and handover in the results. 158 00:07:30,920 --> 00:07:31,880 Speaker 3: And then also, not. 159 00:07:32,080 --> 00:07:36,440 Speaker 4: All sectors that we model are as rational, as for example, 160 00:07:36,480 --> 00:07:39,520 Speaker 4: the power system, and so I mean consumer behaviors in 161 00:07:39,920 --> 00:07:42,520 Speaker 4: important part of how we model our ev uptakes. So 162 00:07:42,640 --> 00:07:45,480 Speaker 4: that's just one example of where it's not like completely 163 00:07:45,520 --> 00:07:50,520 Speaker 4: without taking humans into consideration. But the focus is the economics. Ultimately, 164 00:07:50,560 --> 00:07:52,040 Speaker 4: that's the story we're trying to tell. 165 00:07:52,240 --> 00:07:56,000 Speaker 1: And my impression that that's complex enough as it is. 166 00:07:56,360 --> 00:07:59,440 Speaker 1: I think that you said something interesting before the podcast. 167 00:07:59,640 --> 00:08:02,840 Speaker 1: That is the number one kind of piece of feedback 168 00:08:02,880 --> 00:08:05,240 Speaker 1: you get from clients about this is that they don't 169 00:08:05,280 --> 00:08:07,440 Speaker 1: like a particular kind of charts. Tell me what kind 170 00:08:07,440 --> 00:08:08,280 Speaker 1: of chart don't they like? 171 00:08:08,400 --> 00:08:10,840 Speaker 4: Yeah, So I've been involved in here for a while now, 172 00:08:10,880 --> 00:08:14,200 Speaker 4: and I mean, sometimes, well maybe this is feels like 173 00:08:14,240 --> 00:08:16,520 Speaker 4: a couple of years ago now, but sometimes there's not 174 00:08:16,640 --> 00:08:19,760 Speaker 4: big disturbances in markets, and things tend to tick along, 175 00:08:19,800 --> 00:08:21,680 Speaker 4: and I feel like people sort of get used to 176 00:08:21,800 --> 00:08:24,320 Speaker 4: charts and they can go up, they can go down, 177 00:08:24,360 --> 00:08:26,320 Speaker 4: but they tend to be sort of like quite constant. 178 00:08:26,360 --> 00:08:28,960 Speaker 1: They show a direction, they show a direction, which direction 179 00:08:29,120 --> 00:08:32,160 Speaker 1: plays itself out in a way that you're like, okay, cool, 180 00:08:32,280 --> 00:08:33,160 Speaker 1: so that's the trend. 181 00:08:33,520 --> 00:08:34,200 Speaker 3: And yeah. 182 00:08:34,520 --> 00:08:36,400 Speaker 4: I mean one of the charts that I've got the 183 00:08:36,520 --> 00:08:38,640 Speaker 4: most questions about, or the general type of charts, is 184 00:08:38,679 --> 00:08:40,640 Speaker 4: where that doesn't happen, where you'll have something that goes 185 00:08:40,720 --> 00:08:43,520 Speaker 4: up and then it goes down and people immediately latch 186 00:08:43,559 --> 00:08:45,240 Speaker 4: onto that and they're like, what's happening here? And I mean, 187 00:08:45,280 --> 00:08:47,160 Speaker 4: I think that's part of what makes us a unique 188 00:08:47,160 --> 00:08:48,880 Speaker 4: cause we try to do that bottom up modeling, and 189 00:08:48,920 --> 00:08:50,760 Speaker 4: I think when you do that, you can see these 190 00:08:50,800 --> 00:08:54,240 Speaker 4: sort of weird changes of directions occasionally in your results. 191 00:08:54,240 --> 00:08:56,360 Speaker 4: I think if you're doing a trend based analysis more 192 00:08:56,360 --> 00:08:58,040 Speaker 4: top down, you wouldn't see that at all. 193 00:08:58,040 --> 00:08:58,960 Speaker 3: You would miss that. 194 00:09:00,040 --> 00:09:02,520 Speaker 4: I'm always amazed by client's ability to hone in and 195 00:09:02,600 --> 00:09:05,839 Speaker 4: identify those charts that change tackicy. It has come at 196 00:09:05,880 --> 00:09:07,160 Speaker 4: me with questions. 197 00:09:06,880 --> 00:09:09,560 Speaker 1: If you're been a f client, presumably by the fact 198 00:09:09,600 --> 00:09:11,920 Speaker 1: that you even find our content useful. You do not 199 00:09:12,080 --> 00:09:16,199 Speaker 1: fear change, but you might still fear change of change. 200 00:09:16,679 --> 00:09:19,280 Speaker 1: And that's what those charts are, is like the change. 201 00:09:19,320 --> 00:09:22,400 Speaker 1: We've got some change, and then the change changes, and 202 00:09:22,720 --> 00:09:23,800 Speaker 1: that's a little bit weird. 203 00:09:24,080 --> 00:09:26,240 Speaker 4: Yeah, I mean, it's probably usual to talk about a 204 00:09:26,240 --> 00:09:28,079 Speaker 4: little bit. I mean, the main culport of this has 205 00:09:28,120 --> 00:09:31,240 Speaker 4: been when you zoom out, our results are basically saying, 206 00:09:31,240 --> 00:09:34,200 Speaker 4: if you look at US gas demand, we see a 207 00:09:34,280 --> 00:09:37,719 Speaker 4: decline until twenty thirty and then growth again, and sort 208 00:09:37,760 --> 00:09:42,000 Speaker 4: of people immediately what's behind that, And it's not a 209 00:09:42,040 --> 00:09:44,960 Speaker 4: super complex idea to unpack. Essentially, what we have today 210 00:09:45,160 --> 00:09:48,880 Speaker 4: are many parts of the US where we're still building renewables, 211 00:09:48,880 --> 00:09:51,600 Speaker 4: and renewables are still in the money, and that expansion 212 00:09:51,640 --> 00:09:55,640 Speaker 4: of renewables continues, and so there's some displacement of natural 213 00:09:55,679 --> 00:09:58,480 Speaker 4: gas demand in the power system towards twenty thirty. But 214 00:09:58,520 --> 00:10:01,720 Speaker 4: then around twenty thirty at least when we've modeled this 215 00:10:01,760 --> 00:10:03,160 Speaker 4: to the last year or two, I mean, the whole 216 00:10:03,200 --> 00:10:05,720 Speaker 4: system is moving so equally. Reinpoints the wrong sort of 217 00:10:06,040 --> 00:10:08,880 Speaker 4: terminology here, but we sort of reach a penetration level 218 00:10:08,920 --> 00:10:11,840 Speaker 4: of renewables where the system struggles to move past that 219 00:10:12,000 --> 00:10:15,080 Speaker 4: and so post twenty thirty, what happens. We still build renewables, 220 00:10:15,280 --> 00:10:17,800 Speaker 4: but what you see is the whole system is getting bigger, 221 00:10:17,880 --> 00:10:20,920 Speaker 4: and renewables role in that system grows more or less 222 00:10:20,920 --> 00:10:23,640 Speaker 4: in proportion with that system wide growth, which means the 223 00:10:23,679 --> 00:10:26,040 Speaker 4: other parts of the system also grow in proportion there 224 00:10:26,040 --> 00:10:29,560 Speaker 4: and you see a return to gas demand growth after 225 00:10:29,600 --> 00:10:30,280 Speaker 4: twenty thirty. 226 00:10:30,480 --> 00:10:32,200 Speaker 1: I would go so far as to say that this 227 00:10:32,360 --> 00:10:34,640 Speaker 1: is part of the real value add I mean the 228 00:10:34,679 --> 00:10:36,400 Speaker 1: fact that you have these and clients hone in on 229 00:10:36,440 --> 00:10:38,280 Speaker 1: them and ask questions. And maybe it's not because they 230 00:10:38,280 --> 00:10:40,720 Speaker 1: fear the change that's happened to change. It might just 231 00:10:40,760 --> 00:10:43,640 Speaker 1: be that they've actually identified this is where the value 232 00:10:43,679 --> 00:10:46,720 Speaker 1: really is. Is Some of these things kind of reminds 233 00:10:46,760 --> 00:10:49,800 Speaker 1: me back in my early days as an analyst. My team, 234 00:10:50,200 --> 00:10:53,760 Speaker 1: as an April Full we wrote a pretend research note 235 00:10:53,800 --> 00:10:55,840 Speaker 1: and sent it to the editors. 236 00:10:56,360 --> 00:10:57,840 Speaker 2: I mean, this is how nerdy we were. 237 00:10:58,000 --> 00:11:01,280 Speaker 1: And in it we had a forecast, we had the methodology, 238 00:11:01,280 --> 00:11:04,960 Speaker 1: and we had developed a methodology called the IACAGR, which 239 00:11:05,000 --> 00:11:09,000 Speaker 1: stands for Indiscriminate application of a compound annual growth rate. 240 00:11:09,360 --> 00:11:13,360 Speaker 1: So obviously just making fun of like how you can 241 00:11:13,440 --> 00:11:15,680 Speaker 1: make something look with the sort of like the laziest 242 00:11:15,720 --> 00:11:18,240 Speaker 1: possible methodology, and I think this just really emphasizes that 243 00:11:18,280 --> 00:11:21,320 Speaker 1: you definitely haven't done the indiscriminate application of a compound 244 00:11:21,320 --> 00:11:24,400 Speaker 1: annual growth rate. There's a lot of complexity behind what 245 00:11:24,440 --> 00:11:27,160 Speaker 1: you've done, and so even with some of the policy 246 00:11:27,400 --> 00:11:30,240 Speaker 1: interventions going one way or other stripped out, there's still 247 00:11:30,280 --> 00:11:33,840 Speaker 1: a lot of the zigzagging that happens within this as 248 00:11:33,880 --> 00:11:36,800 Speaker 1: some of these different dynamics play out. So speaking of 249 00:11:37,000 --> 00:11:39,480 Speaker 1: different dynamics that up playing out, let's get on to 250 00:11:39,640 --> 00:11:43,960 Speaker 1: one of the big additions of this report, which was 251 00:11:44,160 --> 00:11:49,800 Speaker 1: bringing in an analysis of the impact that data centers, 252 00:11:49,800 --> 00:11:53,679 Speaker 1: principally driven by power demand for AI, how they could 253 00:11:53,920 --> 00:11:57,640 Speaker 1: impact the energy transition. So for full context, we did 254 00:11:57,640 --> 00:12:02,000 Speaker 1: a podcast with our colleagues Natalie Lamandebrata and Helen co 255 00:12:02,080 --> 00:12:04,959 Speaker 1: and they were talking about specifically maybe the nearer term 256 00:12:05,040 --> 00:12:07,040 Speaker 1: view and particularly to the US, although I think a 257 00:12:07,040 --> 00:12:10,040 Speaker 1: lot of those themes apply more broadly, and so that 258 00:12:10,160 --> 00:12:12,559 Speaker 1: was part of obviously your analysis, but you also looked 259 00:12:12,559 --> 00:12:15,199 Speaker 1: in the longer term as well. You extended this out 260 00:12:15,240 --> 00:12:17,760 Speaker 1: to twenty fifty and obviously the team as well took 261 00:12:17,800 --> 00:12:20,400 Speaker 1: this global. So, yeah, what were some of the things 262 00:12:20,520 --> 00:12:23,199 Speaker 1: that you found doing this analysis? 263 00:12:23,800 --> 00:12:26,440 Speaker 4: So, I mean, first off, I'm definitely standing on the 264 00:12:26,520 --> 00:12:28,360 Speaker 4: shoulders of giants when it comes to the work that 265 00:12:28,440 --> 00:12:30,360 Speaker 4: Natalie and Helen have done here. This is a couple 266 00:12:30,400 --> 00:12:33,400 Speaker 4: of couple of interesting dynamics here is one, because of 267 00:12:33,520 --> 00:12:37,640 Speaker 4: latency concerns, that forward looking view of where where is 268 00:12:37,720 --> 00:12:41,240 Speaker 4: data center demand coming from? It's a lot more globally 269 00:12:41,280 --> 00:12:44,640 Speaker 4: distributed than you might otherwise imagine. I mean that's not 270 00:12:44,679 --> 00:12:47,160 Speaker 4: to downplay the role that the US plays, Like the 271 00:12:47,240 --> 00:12:50,319 Speaker 4: US is still huge, say China's massive. 272 00:12:50,120 --> 00:12:53,680 Speaker 1: Approaching fifty percent of data centers today are in the US. 273 00:12:53,840 --> 00:12:56,280 Speaker 4: I think, well, yeah, no, that's that's more or less 274 00:12:56,280 --> 00:12:58,720 Speaker 4: on the money. And yeah, and so we see like 275 00:12:58,760 --> 00:13:02,040 Speaker 4: a large geographic spread, at least towards the longer term. 276 00:13:02,040 --> 00:13:03,480 Speaker 4: And there's sort of a dynamic here where I mean, 277 00:13:03,520 --> 00:13:06,080 Speaker 4: the US is definitely sort of leading the pack here, 278 00:13:06,120 --> 00:13:08,640 Speaker 4: but other economies who might be a bit later that 279 00:13:08,720 --> 00:13:13,040 Speaker 4: party will eventually catch up in their demand for general 280 00:13:13,120 --> 00:13:15,600 Speaker 4: data and AI specific use cases will grow. 281 00:13:15,400 --> 00:13:16,160 Speaker 3: Over time too. 282 00:13:16,640 --> 00:13:19,000 Speaker 4: There's also I mean there's a seg here as well, 283 00:13:19,040 --> 00:13:23,120 Speaker 4: because that story of renewed gas demand growth in the 284 00:13:23,200 --> 00:13:27,280 Speaker 4: US post twenty thirty is partly responsible from data centers 285 00:13:27,320 --> 00:13:30,320 Speaker 4: as well. There's definitely a long term growth there data 286 00:13:30,360 --> 00:13:33,839 Speaker 4: centers and interestingly electric vehicle demand as well. So that 287 00:13:34,240 --> 00:13:36,600 Speaker 4: was another sort of high level result that popped out 288 00:13:36,600 --> 00:13:40,840 Speaker 4: which was quite interesting, is that evs in most places 289 00:13:41,080 --> 00:13:44,640 Speaker 4: by twenty thirty is still much more significant demand than 290 00:13:44,840 --> 00:13:47,400 Speaker 4: data centers may potentially not much more significant. Other places 291 00:13:47,400 --> 00:13:48,920 Speaker 4: are a bit closer, and I mean US is one 292 00:13:48,920 --> 00:13:51,199 Speaker 4: of the few exceptions where we think that data center 293 00:13:51,240 --> 00:13:54,400 Speaker 4: demand will outpace demand from EV's in the short term, 294 00:13:54,400 --> 00:13:56,800 Speaker 4: but eventually I think EV's will catch up by about 295 00:13:56,840 --> 00:13:57,439 Speaker 4: twenty thirty. 296 00:13:57,720 --> 00:13:59,439 Speaker 2: So where else would we see that trend. 297 00:13:59,800 --> 00:14:00,560 Speaker 3: That's a good question. 298 00:14:00,600 --> 00:14:03,240 Speaker 4: I think we're probably if you looked at markets where 299 00:14:03,440 --> 00:14:06,560 Speaker 4: EV growth is a bit behind the curve, where you 300 00:14:06,559 --> 00:14:08,880 Speaker 4: still have strong demand from data centers. So I think 301 00:14:09,040 --> 00:14:10,920 Speaker 4: I haven't seen these charts, but my guesses would be 302 00:14:10,960 --> 00:14:14,640 Speaker 4: potentially places like Australia or Malaysia where there's a reasonable 303 00:14:14,640 --> 00:14:17,240 Speaker 4: amount of data centers going in but they're not necessarily 304 00:14:17,280 --> 00:14:18,760 Speaker 4: particularly strong EVA markets. 305 00:14:20,160 --> 00:14:21,000 Speaker 2: It's really interesting. 306 00:14:21,000 --> 00:14:25,000 Speaker 1: I mean, actually we just recorded a podcast with Colin Mcherrika, 307 00:14:25,040 --> 00:14:27,880 Speaker 1: who heads up our advanced Transport analysis, and we were 308 00:14:27,880 --> 00:14:31,440 Speaker 1: talking about EVO, which is the transport counterpart to NEO, 309 00:14:31,520 --> 00:14:33,240 Speaker 1: So that's the electric vehicle out look, and we had 310 00:14:33,240 --> 00:14:35,200 Speaker 1: a long discussion about how the US is going to 311 00:14:35,240 --> 00:14:37,680 Speaker 1: be moving slower than other markets on electric vehicles. But 312 00:14:37,880 --> 00:14:39,960 Speaker 1: you know, it's interesting there's this pattern that some of 313 00:14:39,960 --> 00:14:42,040 Speaker 1: those markets they might be getting some data centers, so 314 00:14:42,080 --> 00:14:44,920 Speaker 1: there's going to be demand growth either way. I remember 315 00:14:45,160 --> 00:14:50,200 Speaker 1: a really interesting point that you made in a meeting 316 00:14:50,440 --> 00:14:53,200 Speaker 1: when we were just talking about our data center analysis, 317 00:14:53,280 --> 00:14:56,000 Speaker 1: because you know, there's this question that is sort of 318 00:14:56,040 --> 00:14:58,760 Speaker 1: everyone wants the answer to all this data center demand, 319 00:14:58,760 --> 00:15:01,240 Speaker 1: all this electric vehicle to Is it going to mean 320 00:15:01,240 --> 00:15:03,240 Speaker 1: more renewables and more gas? Is it going to mean 321 00:15:03,280 --> 00:15:06,600 Speaker 1: more emissions? And I remember you explaining this really nuanced point, 322 00:15:06,680 --> 00:15:10,000 Speaker 1: which is that you can create a model that has 323 00:15:10,400 --> 00:15:13,160 Speaker 1: all of those extra things layered in and you can 324 00:15:13,200 --> 00:15:16,200 Speaker 1: see the difference to the power generation mix. But it's 325 00:15:16,280 --> 00:15:19,600 Speaker 1: not necessarily correct to just say that the difference to 326 00:15:19,640 --> 00:15:22,680 Speaker 1: the mix is what is attributing that to those data 327 00:15:22,720 --> 00:15:25,880 Speaker 1: centers or electric vehicles? Can you just explain that? Am 328 00:15:25,920 --> 00:15:26,960 Speaker 1: I remembering this right? 329 00:15:27,080 --> 00:15:27,800 Speaker 2: You are? 330 00:15:27,840 --> 00:15:29,560 Speaker 4: And I mean there's a there's a chart in NEO 331 00:15:29,600 --> 00:15:32,480 Speaker 4: which I think illustrates this point quite well. So we 332 00:15:32,520 --> 00:15:35,640 Speaker 4: with our economic transition scenario, we modeled in an additional 333 00:15:35,680 --> 00:15:38,840 Speaker 4: scenario sort of behind the scenes, where we stripped out 334 00:15:38,960 --> 00:15:42,080 Speaker 4: demand from data centers, and so we solved these two 335 00:15:42,360 --> 00:15:45,000 Speaker 4: nearly identical scenarios. The only difference is that missing data 336 00:15:45,000 --> 00:15:47,080 Speaker 4: center demand. And you can then look at those two 337 00:15:47,120 --> 00:15:49,760 Speaker 4: sets of results and compare them and start to infer 338 00:15:50,040 --> 00:15:52,400 Speaker 4: what the effects are of that data center demand. And 339 00:15:52,840 --> 00:15:54,480 Speaker 4: this goes into a chart which we have in the 340 00:15:54,520 --> 00:15:56,840 Speaker 4: report where you see that potentially out to twenty thirty, 341 00:15:56,840 --> 00:16:00,960 Speaker 4: about two thirds of the additional demand and for data 342 00:16:01,000 --> 00:16:03,680 Speaker 4: centers is met by fossil fuels. And you can't stop 343 00:16:03,680 --> 00:16:05,760 Speaker 4: there though, because that's not the actual message. That's what 344 00:16:05,800 --> 00:16:07,880 Speaker 4: the chart shows. But you sort of have to unpack 345 00:16:08,040 --> 00:16:10,440 Speaker 4: what that sensitivity means, and I think it goes back 346 00:16:10,440 --> 00:16:13,600 Speaker 4: to this concept of additionality. So, first off, we're not 347 00:16:13,640 --> 00:16:15,920 Speaker 4: saying that two thirds of the electrons that are going 348 00:16:15,960 --> 00:16:17,800 Speaker 4: to go into data centers are going to come from 349 00:16:17,920 --> 00:16:19,240 Speaker 4: fossil fueled plants at all. 350 00:16:19,280 --> 00:16:20,200 Speaker 3: That's not what we're saying. 351 00:16:20,320 --> 00:16:22,040 Speaker 4: I think we're probably saying the opposite I think we're 352 00:16:22,080 --> 00:16:25,200 Speaker 4: looking at the sort of corporate ppa activity and there's 353 00:16:25,240 --> 00:16:27,680 Speaker 4: like huge demand there. So like these data centers at 354 00:16:27,760 --> 00:16:30,000 Speaker 4: least normally are going to be powered more often than 355 00:16:30,000 --> 00:16:33,680 Speaker 4: not by renewables, and that's not just a US trend. 356 00:16:33,640 --> 00:16:36,000 Speaker 1: And those I mean, the interesting thing about it as 357 00:16:36,040 --> 00:16:38,440 Speaker 1: a market, and this is very different from electric vehicles, 358 00:16:38,520 --> 00:16:41,360 Speaker 1: is it's so concentrated with a small number of very 359 00:16:41,480 --> 00:16:45,560 Speaker 1: huge organizations that have considerable power in the market. Yeah, 360 00:16:45,600 --> 00:16:49,080 Speaker 1: and a lot of those organizations are investing heavily in 361 00:16:49,720 --> 00:16:51,240 Speaker 1: renewables supply, so. 362 00:16:51,760 --> 00:16:54,920 Speaker 4: And nuclear and other things as well. And I mean 363 00:16:54,920 --> 00:16:57,640 Speaker 4: we also hear stories of them building on site gas 364 00:16:57,680 --> 00:16:59,880 Speaker 4: generation as well. So there's a lot of things happening. 365 00:16:59,880 --> 00:17:02,120 Speaker 4: But I think this concept of coming back to this 366 00:17:02,160 --> 00:17:04,960 Speaker 4: concept of additionality. So we're not saying that two thirds 367 00:17:04,960 --> 00:17:06,760 Speaker 4: of the power that goes into these things is coming 368 00:17:06,760 --> 00:17:10,560 Speaker 4: from fossil fuels. What our analysis shows is that when 369 00:17:10,640 --> 00:17:13,959 Speaker 4: you add this additional demand to the system, can we 370 00:17:14,119 --> 00:17:17,280 Speaker 4: meaningfully increase the rate at which we build renewables in 371 00:17:17,320 --> 00:17:20,600 Speaker 4: the next five to ten years And our analysis tends 372 00:17:20,640 --> 00:17:22,960 Speaker 4: to indicate that we can't. And so this is the 373 00:17:23,000 --> 00:17:26,400 Speaker 4: important idea of additionality it we're going to build these 374 00:17:26,440 --> 00:17:29,960 Speaker 4: data centers and nominally they're going to be powered by renewables, 375 00:17:29,960 --> 00:17:32,959 Speaker 4: But that doesn't mean extra renewables built in the world. 376 00:17:33,119 --> 00:17:35,760 Speaker 4: It probably just means renewables that were already going to 377 00:17:35,800 --> 00:17:38,800 Speaker 4: be built somewhere else now being built and attributed to 378 00:17:38,920 --> 00:17:40,959 Speaker 4: data centers. Does that means the rest of the system 379 00:17:41,080 --> 00:17:43,040 Speaker 4: becomes a little bit more more fossil heavy. And I 380 00:17:43,080 --> 00:17:45,919 Speaker 4: think that's the way to understand the chart. That concept 381 00:17:45,960 --> 00:17:48,960 Speaker 4: of additionality is that we're not building many more renewables 382 00:17:49,000 --> 00:17:51,120 Speaker 4: because of this additional demand, and the reason being we're 383 00:17:51,119 --> 00:17:53,520 Speaker 4: sort of near or at the limits of what we 384 00:17:53,560 --> 00:17:56,680 Speaker 4: think renewable supply chains can manage in the next five 385 00:17:56,720 --> 00:17:58,480 Speaker 4: to ten years and what the grid could manage in 386 00:17:58,520 --> 00:17:59,720 Speaker 4: the next five to ten years. 387 00:18:00,040 --> 00:18:02,840 Speaker 1: I mean, it's really interesting because I think where we 388 00:18:02,880 --> 00:18:07,080 Speaker 1: get to here is on that boundary between our economically 389 00:18:07,160 --> 00:18:11,800 Speaker 1: rational model and something that reflects the what happens in 390 00:18:11,840 --> 00:18:14,800 Speaker 1: the fog of war. And I could make the argument 391 00:18:14,920 --> 00:18:16,919 Speaker 1: that when you say all those renewables would have been 392 00:18:16,920 --> 00:18:19,480 Speaker 1: built anyway, but now they're going to be built with 393 00:18:19,520 --> 00:18:21,560 Speaker 1: a PPA for a data center, you could make the 394 00:18:21,640 --> 00:18:24,160 Speaker 1: argument actually they weren't going to get built anyway, even 395 00:18:24,200 --> 00:18:26,800 Speaker 1: though it was the economically rational thing from a system 396 00:18:26,840 --> 00:18:29,679 Speaker 1: point of view, because might just be the market design 397 00:18:29,760 --> 00:18:31,879 Speaker 1: didn't favor that. So I and we also have to 398 00:18:31,960 --> 00:18:34,720 Speaker 1: keep that in mind, is that sometimes these things can 399 00:18:34,840 --> 00:18:38,560 Speaker 1: act as of forcing mechanism to make maybe the outcome 400 00:18:38,560 --> 00:18:42,080 Speaker 1: that should have happened anyway on an economic basis happen. 401 00:18:42,280 --> 00:18:44,439 Speaker 1: I think you're entirely right, and that's I was careful 402 00:18:44,480 --> 00:18:47,560 Speaker 1: my language. I didn't I used infer because the results 403 00:18:47,560 --> 00:18:49,560 Speaker 1: don't tell you something's definitely going to happen. It just 404 00:18:49,560 --> 00:18:51,880 Speaker 1: gives you a hint at what might be what might 405 00:18:51,920 --> 00:18:54,439 Speaker 1: be the reasons. And I mean I've also made a 406 00:18:54,480 --> 00:18:57,080 Speaker 1: similar argument before I said, Look, before we started talking 407 00:18:57,080 --> 00:19:00,119 Speaker 1: about data centers an AI, we already were running to 408 00:19:00,240 --> 00:19:02,760 Speaker 1: huge challenges in trying to connect the renewables that we 409 00:19:02,760 --> 00:19:04,720 Speaker 1: were already going to build over the next five to 410 00:19:04,760 --> 00:19:07,760 Speaker 1: ten years. We were running into connection Q issues and 411 00:19:08,160 --> 00:19:10,520 Speaker 1: some supply chain issues. And this is before we started 412 00:19:10,560 --> 00:19:13,280 Speaker 1: worrying about tariffs and whatnot. So just connecting the amount 413 00:19:13,280 --> 00:19:16,080 Speaker 1: of renewables that we thought was economically irrational in our 414 00:19:16,200 --> 00:19:18,000 Speaker 1: modeling over the next few years was already a challenge, 415 00:19:18,040 --> 00:19:20,960 Speaker 1: and we were already constraining the model there. And to 416 00:19:21,000 --> 00:19:22,800 Speaker 1: flip what I just said on his head, like, the 417 00:19:22,840 --> 00:19:26,879 Speaker 1: amount of resources that the Amazons, the Googles, the Microsoft, 418 00:19:26,960 --> 00:19:29,679 Speaker 1: and the Metas have in the world are enormous, so 419 00:19:29,880 --> 00:19:34,120 Speaker 1: they can actually apply the capital and political influence required 420 00:19:34,160 --> 00:19:36,679 Speaker 1: to in many cases solve some of these issues that 421 00:19:36,760 --> 00:19:39,040 Speaker 1: might have persisted for longer otherwise. So we've gone from 422 00:19:39,040 --> 00:19:42,040 Speaker 1: a situation where we didn't necessarily know how we were 423 00:19:42,040 --> 00:19:44,720 Speaker 1: going to sort of muster the support we needed in 424 00:19:45,119 --> 00:19:47,080 Speaker 1: these various areas to build this amount of renewables, and 425 00:19:47,080 --> 00:19:49,040 Speaker 1: all of a sudden, we've got these huge players like 426 00:19:49,119 --> 00:19:51,240 Speaker 1: pushing all the right buttons to try and get these 427 00:19:51,240 --> 00:19:53,880 Speaker 1: things built as quickly as possible. If we sat here 428 00:19:53,880 --> 00:19:56,920 Speaker 1: in a year's time, I'm assuming that in the NEO 429 00:19:57,080 --> 00:20:01,320 Speaker 1: twenty twenty six we will also have inc data centers, 430 00:20:01,440 --> 00:20:03,560 Speaker 1: because we can't not include them unless in the next 431 00:20:03,680 --> 00:20:06,600 Speaker 1: year or so everyone decides that actually, AI was such 432 00:20:06,600 --> 00:20:07,840 Speaker 1: a terrible idea. 433 00:20:07,720 --> 00:20:10,280 Speaker 2: And you know, I highly doubt that's going to happen. 434 00:20:10,480 --> 00:20:11,960 Speaker 1: And one of the things as you were speaking that 435 00:20:12,000 --> 00:20:14,920 Speaker 1: I was thinking about was this idea of on site 436 00:20:14,960 --> 00:20:18,399 Speaker 1: generation and that's been discussed a lot as a potential solution. 437 00:20:18,440 --> 00:20:20,440 Speaker 1: I suppose I've got this question all in the wrong order. 438 00:20:20,480 --> 00:20:22,600 Speaker 1: I should have said, did we model on site generation 439 00:20:23,000 --> 00:20:25,119 Speaker 1: this year? And do you think we will model it 440 00:20:25,160 --> 00:20:25,600 Speaker 1: next year? 441 00:20:25,840 --> 00:20:26,600 Speaker 3: So it's a good question. 442 00:20:26,680 --> 00:20:30,120 Speaker 4: So the short answer is no, So all our demand 443 00:20:30,400 --> 00:20:32,560 Speaker 4: is grid connected. That doesn't mean to say there aren't 444 00:20:32,600 --> 00:20:34,720 Speaker 4: constraints that we can apply to the grid in the model, 445 00:20:34,760 --> 00:20:37,200 Speaker 4: but all demand is grid connected. I don't think it's 446 00:20:37,200 --> 00:20:39,720 Speaker 4: a terrible assumption for data centers and AI. So I mean, 447 00:20:39,840 --> 00:20:43,760 Speaker 4: we've talked before about this five nines reliability constraints. So 448 00:20:43,800 --> 00:20:46,120 Speaker 4: we need ninety nine point nine nine nine percent uptime 449 00:20:46,320 --> 00:20:49,639 Speaker 4: on many of these facilities, which are highly critical parts 450 00:20:49,720 --> 00:20:52,280 Speaker 4: of the infrastructure that powers the modern world. And I mean, 451 00:20:52,320 --> 00:20:53,920 Speaker 4: when you do the math, ninety nine point nine o 452 00:20:54,040 --> 00:20:56,200 Speaker 4: nine of all the hours in the year of cross 453 00:20:56,200 --> 00:20:58,520 Speaker 4: the year gives you five minutes of downtime. And so 454 00:20:58,600 --> 00:21:03,080 Speaker 4: it's very hard to design a system that's off grid 455 00:21:03,200 --> 00:21:06,080 Speaker 4: that only gives you five minutes of downtime with a 456 00:21:06,119 --> 00:21:08,000 Speaker 4: high chance of likelihood. 457 00:21:08,160 --> 00:21:11,919 Speaker 1: I mean, I suppose there's being completely off the grid, 458 00:21:12,080 --> 00:21:14,560 Speaker 1: which might create those issues. But oh sorry, we're looking 459 00:21:14,560 --> 00:21:16,400 Speaker 1: at this from the site. Yeah, two separate things here. 460 00:21:16,440 --> 00:21:19,800 Speaker 4: So when it comes to on site generation, so that 461 00:21:20,000 --> 00:21:22,639 Speaker 4: in theory is something that our model proposers, so we 462 00:21:22,680 --> 00:21:25,119 Speaker 4: don't make a judgment call of where that generation has 463 00:21:25,160 --> 00:21:27,400 Speaker 4: to sit in the grid. So some of the generation 464 00:21:27,480 --> 00:21:30,800 Speaker 4: that our model proposers could be on site generation for 465 00:21:30,840 --> 00:21:31,680 Speaker 4: those facilities. 466 00:21:31,960 --> 00:21:35,119 Speaker 1: Okay, So actually, in a way you do factor it 467 00:21:35,160 --> 00:21:39,080 Speaker 1: in kind of implicitly, but we don't make it explicit exactly. 468 00:21:39,160 --> 00:21:41,960 Speaker 4: So everything's grid connected, but we don't make a judgment 469 00:21:41,960 --> 00:21:44,679 Speaker 4: call on where things have to sit in the grid 470 00:21:44,720 --> 00:21:46,080 Speaker 4: for the whole thing to make sense. 471 00:21:45,840 --> 00:21:49,240 Speaker 1: And whether it's one four hundred megawatt gas plant or 472 00:21:49,280 --> 00:21:51,840 Speaker 1: four hundred one megawat gas plants exactly. 473 00:21:51,920 --> 00:21:54,399 Speaker 4: I mean, I think for the reliability reasons we mentioned, 474 00:21:54,440 --> 00:21:57,760 Speaker 4: it might be more likely to build more smaller gas 475 00:21:57,760 --> 00:22:00,000 Speaker 4: plants for one of these things rather than one line 476 00:22:00,240 --> 00:22:02,879 Speaker 4: one which is harder to maintain it that uptime level. 477 00:22:03,119 --> 00:22:05,440 Speaker 1: So it sounds like in a way you have got 478 00:22:05,440 --> 00:22:07,679 Speaker 1: that covered. And obviously we're making no promises here, but 479 00:22:07,720 --> 00:22:09,639 Speaker 1: what do you think might be some of the things 480 00:22:09,680 --> 00:22:12,080 Speaker 1: that we add in next years NEO. 481 00:22:12,160 --> 00:22:13,800 Speaker 3: Then when it comes to data centers. 482 00:22:13,800 --> 00:22:15,439 Speaker 4: I mean, I think we just we'll have a year's 483 00:22:15,480 --> 00:22:17,680 Speaker 4: more data behind us, and we'll have a better view 484 00:22:17,680 --> 00:22:20,040 Speaker 4: on what the short and then the long term looks like. 485 00:22:20,280 --> 00:22:22,760 Speaker 4: And one of my favorite data points this year is 486 00:22:22,800 --> 00:22:25,240 Speaker 4: if you look at the data center demand that we 487 00:22:25,280 --> 00:22:27,679 Speaker 4: put in the model, the fraction of that which is 488 00:22:27,840 --> 00:22:31,679 Speaker 4: AI facilities at the moment, it's about two percent of 489 00:22:31,720 --> 00:22:34,920 Speaker 4: old data center capacity is AI specific and the rest 490 00:22:35,000 --> 00:22:37,040 Speaker 4: is just our general run of them, will data centers, 491 00:22:37,200 --> 00:22:40,240 Speaker 4: backbone of the Internet, the cloud, et cetera. So that's 492 00:22:40,280 --> 00:22:42,480 Speaker 4: just two percent of the total. And then data centers 493 00:22:42,520 --> 00:22:46,000 Speaker 4: as an entire demand class are only about one point 494 00:22:46,160 --> 00:22:49,840 Speaker 4: four percent of the entire global power demand. 495 00:22:50,040 --> 00:22:52,480 Speaker 1: And so so right now this is like a small 496 00:22:52,560 --> 00:22:54,120 Speaker 1: fraction of a small fraction. 497 00:22:53,840 --> 00:22:54,840 Speaker 3: A small fractions for it. 498 00:22:54,880 --> 00:22:57,520 Speaker 4: And I love an analogy, Tom, so indulge me, Andre 499 00:22:57,560 --> 00:22:57,800 Speaker 4: we go. 500 00:22:57,880 --> 00:22:58,960 Speaker 2: We haven't had enough to date. 501 00:23:00,520 --> 00:23:02,120 Speaker 4: So the way I like to think about it, it's, 502 00:23:02,200 --> 00:23:05,680 Speaker 4: right right now, this demand from AIS just a very 503 00:23:05,800 --> 00:23:08,480 Speaker 4: very small ripple on the horizon we're looking out to 504 00:23:08,480 --> 00:23:10,560 Speaker 4: see and we can see this very very small ripple, 505 00:23:10,720 --> 00:23:13,440 Speaker 4: and we really don't know in five or ten years 506 00:23:13,440 --> 00:23:15,600 Speaker 4: time whether this is just going to be a small 507 00:23:15,640 --> 00:23:18,040 Speaker 4: wave that sort of peters out or a tsunami. 508 00:23:17,520 --> 00:23:19,080 Speaker 3: That washes the power system away. 509 00:23:19,240 --> 00:23:21,560 Speaker 4: We really don't know at the moment, but there are 510 00:23:21,600 --> 00:23:25,360 Speaker 4: some clues because, to continue this analogy, the coastline's not flat. 511 00:23:25,400 --> 00:23:28,600 Speaker 4: There are parts of this coastline which extend file into 512 00:23:28,600 --> 00:23:30,600 Speaker 4: the sea, and we can look at these places today 513 00:23:30,600 --> 00:23:32,640 Speaker 4: and see the challenges they're facing. So, I mean, it's 514 00:23:32,720 --> 00:23:35,160 Speaker 4: not a surprise we're going to talk about PGM now. 515 00:23:35,359 --> 00:23:37,359 Speaker 4: Basically the part of the US, one of the grids 516 00:23:37,359 --> 00:23:40,040 Speaker 4: in the US where we see a hyper concentration of 517 00:23:40,200 --> 00:23:40,960 Speaker 4: data centers. 518 00:23:41,440 --> 00:23:44,160 Speaker 1: Yeah, that's like forty percent of US state centers roughly, 519 00:23:44,200 --> 00:23:45,240 Speaker 1: I think are in PJM. 520 00:23:45,440 --> 00:23:48,119 Speaker 4: Yeah, it's it's it's absolutely massive, And so you can 521 00:23:48,160 --> 00:23:50,679 Speaker 4: look to those regions now and you can see the 522 00:23:50,760 --> 00:23:53,159 Speaker 4: challenges they're facing and the struggles they're going through, and 523 00:23:53,200 --> 00:23:55,320 Speaker 4: you can sort of think about the implications for the 524 00:23:55,359 --> 00:23:57,879 Speaker 4: wider system further down the path. 525 00:23:59,440 --> 00:24:04,639 Speaker 1: So I'm I'm running with your rather doomsday like analogy. Here, 526 00:24:04,840 --> 00:24:07,760 Speaker 1: You're on the beach, there's a little ripple. Some people 527 00:24:07,920 --> 00:24:12,439 Speaker 1: get up and start running, and other people are like, really, 528 00:24:12,680 --> 00:24:13,720 Speaker 1: you just need to chill. 529 00:24:13,880 --> 00:24:16,360 Speaker 2: There's nothing going on here? Where are you? 530 00:24:16,880 --> 00:24:20,359 Speaker 1: Are you grabbing your towel and making it for the mountains, 531 00:24:20,560 --> 00:24:23,360 Speaker 1: or are you just a bit calmer waiting to see 532 00:24:23,359 --> 00:24:25,400 Speaker 1: what will happen, and whatever happens, you're like, I'm pretty 533 00:24:25,400 --> 00:24:26,560 Speaker 1: confident we can handle this. 534 00:24:26,920 --> 00:24:30,280 Speaker 4: So if you look at our forecast compared to other 535 00:24:30,359 --> 00:24:33,960 Speaker 4: research houses, we're more likely to be the ones staying 536 00:24:33,960 --> 00:24:37,520 Speaker 4: on the beach ordering another dakery. So we tend to be. 537 00:24:37,880 --> 00:24:39,680 Speaker 1: Or towards while the dakery is are cheap because no 538 00:24:39,720 --> 00:24:41,320 Speaker 1: one else is around to buy them as well. 539 00:24:41,440 --> 00:24:44,199 Speaker 4: Exactly, So we tend to be the more towards the 540 00:24:44,240 --> 00:24:47,960 Speaker 4: more conservative end of the demand forecast that you see 541 00:24:48,040 --> 00:24:50,080 Speaker 4: for data centers and AI out there. 542 00:24:50,119 --> 00:24:52,520 Speaker 1: But put that in numbers because like our twenty fifty 543 00:24:53,000 --> 00:24:54,639 Speaker 1: I think we still say data sense will be a 544 00:24:54,680 --> 00:24:56,680 Speaker 1: pretty significant percentage. 545 00:24:56,240 --> 00:24:58,520 Speaker 2: Of demand in twenty fifty of power demand. 546 00:24:58,920 --> 00:25:02,240 Speaker 4: Yeah, we're we're going from about one point four percent 547 00:25:02,280 --> 00:25:06,560 Speaker 4: of global power demand today by twenty thirty that's about 548 00:25:06,640 --> 00:25:09,440 Speaker 4: three percent, and by twenty thirty five that's four point 549 00:25:09,440 --> 00:25:11,760 Speaker 4: five percent. If we keep going out to twenty fifty, 550 00:25:11,760 --> 00:25:14,439 Speaker 4: it's almost nine percent of all electricity demand. 551 00:25:14,560 --> 00:25:16,879 Speaker 1: And to be clear, I mean we're seeing we're assuming 552 00:25:16,880 --> 00:25:20,960 Speaker 1: a certain amount of electrification generally in kind of all economies, 553 00:25:21,040 --> 00:25:24,680 Speaker 1: So a bigger percentage of a bigger volume, right, Yeah. 554 00:25:24,520 --> 00:25:27,040 Speaker 4: So the system is getting huge and yeah, and I 555 00:25:27,040 --> 00:25:29,119 Speaker 4: think another thing is that, like originally I said, like 556 00:25:29,160 --> 00:25:31,240 Speaker 4: some of this demand, there's more, maybe more of a 557 00:25:31,320 --> 00:25:33,879 Speaker 4: spread across geographies than you might think. It's not just 558 00:25:33,920 --> 00:25:36,400 Speaker 4: the US only story, but I mean the US definitely 559 00:25:36,440 --> 00:25:38,200 Speaker 4: is at the pointy end of the wedge here, and 560 00:25:38,400 --> 00:25:41,480 Speaker 4: particularly that that PGM region we mentioned. So, I mean 561 00:25:41,520 --> 00:25:44,280 Speaker 4: those figures might be in twenty fifty, like eight point 562 00:25:44,400 --> 00:25:47,480 Speaker 4: seven percent worldwide, but it's it's closer to about twenty 563 00:25:47,480 --> 00:25:49,440 Speaker 4: five percent in PGM at that point. 564 00:25:49,560 --> 00:25:49,879 Speaker 2: Wow. 565 00:25:50,040 --> 00:25:52,880 Speaker 1: Okay, it's really interesting. So it might be the own 566 00:25:53,000 --> 00:25:56,760 Speaker 1: s beaches have another Dakari and another one. Pick up 567 00:25:56,760 --> 00:25:59,480 Speaker 1: your towel and start getting ready for change. I think 568 00:25:59,680 --> 00:26:03,040 Speaker 1: this wave is not going to hit all beaches equally. 569 00:26:03,320 --> 00:26:04,000 Speaker 3: Yeah, in case it. 570 00:26:04,000 --> 00:26:06,560 Speaker 1: Lost a thread of this metaphor, I mean that some 571 00:26:06,840 --> 00:26:09,440 Speaker 1: regions and some markets will be affected by this more 572 00:26:09,480 --> 00:26:10,080 Speaker 1: than others. 573 00:26:10,920 --> 00:26:13,920 Speaker 3: Just to be clear, Yeah, no, I think that's a 574 00:26:13,920 --> 00:26:14,560 Speaker 3: good way to put it. 575 00:26:14,600 --> 00:26:18,120 Speaker 4: I think it's not panic stations everywhere. It's also it's 576 00:26:18,119 --> 00:26:20,479 Speaker 4: not also not an insignificant challenge, Like if we are 577 00:26:20,560 --> 00:26:23,160 Speaker 4: going to keep building data centers, particularly at the pace 578 00:26:23,200 --> 00:26:24,920 Speaker 4: that we currently want to build them, there are huge 579 00:26:25,000 --> 00:26:27,160 Speaker 4: challenges in both in the short term and the long 580 00:26:27,240 --> 00:26:29,399 Speaker 4: term to become a reality. And I think part of 581 00:26:29,400 --> 00:26:31,160 Speaker 4: the challenge of the modeling if you sort of zoom 582 00:26:31,200 --> 00:26:32,960 Speaker 4: out a little bit, and this is maybe more a 583 00:26:33,000 --> 00:26:35,600 Speaker 4: story for sort of the more developed economies, but we've 584 00:26:35,600 --> 00:26:37,080 Speaker 4: been doing neo for a while now, and if you 585 00:26:37,119 --> 00:26:39,840 Speaker 4: look back at electricity demand over the last like ten 586 00:26:39,960 --> 00:26:42,800 Speaker 4: or so years, I think we basically got out of 587 00:26:42,800 --> 00:26:45,680 Speaker 4: the habit of having new demand. We were quite comfortable 588 00:26:45,680 --> 00:26:48,480 Speaker 4: with demand being quite flat and they're not being demand growth, 589 00:26:48,560 --> 00:26:50,280 Speaker 4: and we just sort of were quite happy to sit 590 00:26:50,320 --> 00:26:52,720 Speaker 4: there and twiddle our thumbs. And if in many regions 591 00:26:52,760 --> 00:26:57,240 Speaker 4: in particular because of basically energy efficiency gains, partly because 592 00:26:57,280 --> 00:27:00,959 Speaker 4: of like the shift from industrial to service economies, you 593 00:27:01,000 --> 00:27:03,480 Speaker 4: even saw falling demand. And so I mean there's a 594 00:27:03,520 --> 00:27:06,639 Speaker 4: regional variation here, but like basically, if you ignore evs 595 00:27:06,640 --> 00:27:08,959 Speaker 4: and data centers, a lot of places, the more developed 596 00:27:08,960 --> 00:27:11,200 Speaker 4: parts of the world would see falling demand over the 597 00:27:11,280 --> 00:27:13,359 Speaker 4: next ten or twenty years, and then you're add in 598 00:27:13,440 --> 00:27:15,960 Speaker 4: evs and it sort of evens out to a bit 599 00:27:16,000 --> 00:27:17,760 Speaker 4: more stable, and then when you ad data centers on 600 00:27:17,760 --> 00:27:20,040 Speaker 4: top of that, you're getting growth pretty much everywhere. 601 00:27:20,119 --> 00:27:22,760 Speaker 1: Now it's so interesting, And you know, I head up 602 00:27:22,760 --> 00:27:25,800 Speaker 1: our global power market analysis where we focus a lot 603 00:27:25,840 --> 00:27:28,720 Speaker 1: on the outlook for prices, and I think that and 604 00:27:28,800 --> 00:27:31,000 Speaker 1: this was before I was heading this up, but before 605 00:27:31,119 --> 00:27:33,760 Speaker 1: this wave of demand growth came into town. The themes 606 00:27:33,800 --> 00:27:36,159 Speaker 1: that we were talking about was like price cannibalization. How 607 00:27:36,200 --> 00:27:38,840 Speaker 1: much is price going to get cannibalized by these renewables, 608 00:27:38,960 --> 00:27:41,600 Speaker 1: And that is still an important topic, but that is 609 00:27:41,680 --> 00:27:44,679 Speaker 1: also predicated on this idea of kind of flat or 610 00:27:44,680 --> 00:27:47,560 Speaker 1: declining demand. Now you've got demand growth, It's like, yeah, 611 00:27:47,760 --> 00:27:49,960 Speaker 1: price cannibalization is still a factor, but it might be 612 00:27:50,000 --> 00:27:53,639 Speaker 1: offset by other factors as well, including the growth of demand. 613 00:27:53,880 --> 00:27:58,880 Speaker 1: So suddenly it's a lot more interesting and creates opportunities. 614 00:27:58,920 --> 00:28:00,920 Speaker 1: I think this is the key point is a lot 615 00:28:00,960 --> 00:28:05,639 Speaker 1: of new power capacity will be renewable, and it maybe 616 00:28:05,960 --> 00:28:08,919 Speaker 1: provides a more optimistic outlook for the economics of that 617 00:28:08,960 --> 00:28:11,840 Speaker 1: power capacity as well. So I would say that even 618 00:28:11,880 --> 00:28:14,560 Speaker 1: if we're saying in our economic modeling that there might 619 00:28:14,600 --> 00:28:18,399 Speaker 1: be more gas consumed, which from an environmental point of 620 00:28:18,480 --> 00:28:20,800 Speaker 1: view is not good. Of course, people in the renewable 621 00:28:20,920 --> 00:28:24,359 Speaker 1: energy industry should be looking at this as and I 622 00:28:24,400 --> 00:28:27,240 Speaker 1: think they are looking at this as an opportunity for 623 00:28:27,320 --> 00:28:31,119 Speaker 1: them to make their mark. And how it actually plays 624 00:28:31,160 --> 00:28:35,720 Speaker 1: out in reality, whether it's majority of gas generation meeting 625 00:28:35,760 --> 00:28:38,000 Speaker 1: that demand or renewables, it kind of depends on how 626 00:28:38,040 --> 00:28:41,360 Speaker 1: people play their cards and strategize and make the right moves. 627 00:28:41,400 --> 00:28:43,240 Speaker 1: And so it's going to be really interesting to see 628 00:28:43,280 --> 00:28:46,040 Speaker 1: all of that. But the point is in the power sector, 629 00:28:46,120 --> 00:28:48,400 Speaker 1: suddenly it's not just about keeping the lights on. 630 00:28:48,520 --> 00:28:49,800 Speaker 2: There's opportunity as well. 631 00:28:50,040 --> 00:28:52,640 Speaker 4: Yeah, and I mean it is, it's just more dynamic. 632 00:28:52,720 --> 00:28:55,040 Speaker 4: Is more exciting when you when you have to accommodate 633 00:28:55,160 --> 00:28:57,760 Speaker 4: new growth, and I mean first level, it makes my 634 00:28:57,840 --> 00:29:01,000 Speaker 4: job more interesting. It's you've got to sort of that 635 00:29:01,000 --> 00:29:03,920 Speaker 4: double whemmy challenge of like what are the new things 636 00:29:03,920 --> 00:29:07,360 Speaker 4: we build and where do we canniblize and potentially replace 637 00:29:07,440 --> 00:29:08,520 Speaker 4: or retire the old assets. 638 00:29:08,760 --> 00:29:12,560 Speaker 1: Well, for every new dynamic, that is another kink in 639 00:29:12,600 --> 00:29:16,440 Speaker 1: one of your charts. That is more the change is changing, 640 00:29:16,800 --> 00:29:19,760 Speaker 1: and that is more questions from clients asking for a 641 00:29:19,800 --> 00:29:23,760 Speaker 1: more detailed explanation of what is going on. So Ian, 642 00:29:23,880 --> 00:29:25,480 Speaker 1: thank you very much for joining today. 643 00:29:25,760 --> 00:29:27,400 Speaker 3: Nice Thanks Tom, See you on the beach. 644 00:29:27,600 --> 00:29:38,840 Speaker 2: See on the beach. Today's episode of Switched On was 645 00:29:38,880 --> 00:29:42,760 Speaker 2: produced by Cam Gray with production assistants from Kamala Shelling. 646 00:29:43,000 --> 00:29:45,920 Speaker 1: Bloomberg ne EF is a service provided by Bloomberg Finance 647 00:29:46,000 --> 00:29:49,080 Speaker 1: LP and its affiliates. This recording does not constitute, nor 648 00:29:49,080 --> 00:29:52,360 Speaker 1: should it be construed, as investment ad vice, investment recommendations, 649 00:29:52,480 --> 00:29:55,440 Speaker 1: or a recommendation as to an investment or other strategy. 650 00:29:55,440 --> 00:29:58,880 Speaker 4: Bloomberg Annia should not be considered as information sufficient upon 651 00:29:58,920 --> 00:30:00,640 Speaker 4: which to base an invest decision. 652 00:30:00,720 --> 00:30:03,720 Speaker 1: Neither Bloomberg Finance LP nor any of its affiliates makes 653 00:30:03,720 --> 00:30:07,480 Speaker 1: any representation or warranty as to the accuracy or completeness 654 00:30:07,480 --> 00:30:09,880 Speaker 1: of the information contained in this recording, and any 655 00:30:09,920 --> 00:30:13,200 Speaker 3: Liability as a result of this recording is expressly disclaimed.