1
00:00:00,000 --> 00:00:05,700
What one thing should future solar mining projects copy from you guys?

2
00:00:06,020 --> 00:00:06,840
Project finance.

3
00:00:07,620 --> 00:00:10,020
One thing that they should never copy.

4
00:00:13,060 --> 00:00:17,220
HashrateUp. Hardware sales, advisory, hosting, and site brokerage.

5
00:00:17,640 --> 00:00:21,680
Find new and used ASIC deals through the website and the Telegram channel below.

6
00:00:22,240 --> 00:00:24,360
Make smarter decisions with HashrateUp.

7
00:00:24,620 --> 00:00:28,160
We are at blockhite 935267.

8
00:00:28,160 --> 00:00:34,380
cash price still an absolute dumpster fire i think when i recorded yesterday the absolute low down

9
00:00:34,380 --> 00:00:44,400
here was 2789 and now we're just above 30 again with that being said um today we're speaking to

10
00:00:44,400 --> 00:00:49,720
spencer ma of sankhar renewables i hope i pronounced that right he can correct me just now

11
00:00:49,720 --> 00:00:56,460
where they recently announced a 20 megawatts solar mining project.

12
00:00:57,620 --> 00:01:02,040
And especially with the current hash price environment as a backdrop,

13
00:01:02,420 --> 00:01:04,940
that is a very interesting thing to talk about.

14
00:01:05,280 --> 00:01:07,800
With all of that being said, Spencer, welcome to the show.

15
00:01:07,920 --> 00:01:09,320
Thank you for your time. How are you doing?

16
00:01:10,980 --> 00:01:12,980
I'm doing well. Thank you for having me.

17
00:01:13,200 --> 00:01:15,560
It's an interesting time to have this conversation.

18
00:01:15,560 --> 00:01:23,700
yeah what makes it interesting from your perspective yeah obviously uh you know when

19
00:01:23,700 --> 00:01:30,300
bitcoin price starts careening down and hash price goes along with it though of course not

20
00:01:30,300 --> 00:01:37,340
not quite as precipitously uh it doesn't make anyone feel good it's not uh confidence inspiring

21
00:01:37,340 --> 00:01:43,340
uh but it also as we as a team internally have been thinking about this of course all week it's

22
00:01:43,340 --> 00:01:45,420
It has everyone's attention.

23
00:01:45,840 --> 00:01:47,680
We also see some silver linings.

24
00:01:48,860 --> 00:01:59,120
So in our case, and I'm sure this is what we're going to get into, we feel like we took a long time to get this current project with Total Energies off the ground.

25
00:01:59,120 --> 00:02:09,700
And over the course of putting the deal together, as much focus as there was on really the energy rate procurement piece, as much if not more went into business model.

26
00:02:09,700 --> 00:02:18,280
And so we've designed a really lean project finance approach towards capitalizing and operating Bitcoin mining assets.

27
00:02:19,280 --> 00:02:24,580
So a lot of our, you know, when times are good, it's easy to mask mistakes that people are making.

28
00:02:24,820 --> 00:02:36,280
But we feel like, you know, being in this down market and we'll see how long it lasts is actually kind of proving or validating some of our some of what we've been selling investors on for a long time,

29
00:02:36,280 --> 00:02:47,640
which is really that our business model and our procurement strategy are designed to be flexible and provide us a limit to the to the downside.

30
00:02:48,680 --> 00:02:56,440
And so even now, as part of part of what we're thinking through is like if the economics of mining Bitcoin were to never get better than they are today,

31
00:02:57,000 --> 00:03:00,060
are we still cash flow positive? Can investors still generate a return?

32
00:03:00,100 --> 00:03:01,880
And the answer, at least right now, is yes.

33
00:03:02,440 --> 00:03:05,840
So, you know, there's reasons for that, that we can explore.

34
00:03:05,840 --> 00:03:08,840
But yeah, that's the silver lining of the moment.

35
00:03:09,020 --> 00:03:17,620
So I guess to answer your question, it's because in both directions, you have the downside of, of course, we all wish we were at hash price 100.

36
00:03:18,400 --> 00:03:29,460
But over the long run, we feel like our views and our thesis and business model will kind of shine even more as we come out of this.

37
00:03:30,200 --> 00:03:34,600
All right. Give people a 30-second rundown, Spencer, of your time in Bitcoin mining.

38
00:03:34,600 --> 00:03:39,480
we talked a bit before you've been around the block as well i just give people some context

39
00:03:39,480 --> 00:03:48,200
since you're a new character on the show sure sure so um my background was as a lawyer for about five

40
00:03:48,200 --> 00:03:56,200
years i left legal practice to get into renewable public policy in uh 2016 2017 working for the city

41
00:03:56,200 --> 00:04:01,300
of New York in 2017 because I'm class of 2017.

42
00:04:02,180 --> 00:04:06,800
So I saw as Bitcoin went to 20,000 at the time, which is a big, exciting thing.

43
00:04:07,180 --> 00:04:11,360
I was working in a pretty interdisciplinary environment with engineers and power markets

44
00:04:11,360 --> 00:04:11,740
people.

45
00:04:12,060 --> 00:04:19,420
And I had learned enough to be dangerous about the problems of power marketing and renewable

46
00:04:19,420 --> 00:04:20,660
energy project finance.

47
00:04:20,660 --> 00:04:35,460
And so I had this basic idea, light bulb moment of, could we use Bitcoin mining as a force multiplier to turn around economics on renewable projects facing congestion, curtailment, negative pricing?

48
00:04:36,360 --> 00:04:41,240
I'm saying all this now as if my phrasing then might have been different because I didn't know.

49
00:04:41,320 --> 00:04:42,180
It was just a basic go.

50
00:04:42,600 --> 00:04:44,500
Renewable energy bad, Bitcoin mining good.

51
00:04:44,740 --> 00:04:47,340
Let's figure out how to use the two together.

52
00:04:47,340 --> 00:04:55,500
I tried in 2018 to pursue co-location with some large renewable assets, particularly wind.

53
00:04:56,760 --> 00:04:59,700
It was just too early in every way.

54
00:05:00,000 --> 00:05:03,540
The idea of institutional grade Bitcoin mining didn't exist.

55
00:05:03,880 --> 00:05:06,480
There wasn't a single publicly traded mining firm at the time.

56
00:05:07,780 --> 00:05:11,460
Credit profile, deal structuring, it was never going to happen.

57
00:05:11,460 --> 00:05:19,520
And so by 2018, back then we had a company called Sangha Systems, which is a bit of a predecessor company.

58
00:05:19,780 --> 00:05:30,400
We wound up developing and operating and then ultimately selling just a pure grid-tied site about two hours west of Chicago in a town called Hennepin, Illinois.

59
00:05:31,320 --> 00:05:37,300
And it was in that time that, you know, the Bitcoin mining industry really blew up.

60
00:05:37,680 --> 00:05:41,400
You know, you started having these huge conferences, institutional capital flowing into the space.

61
00:05:41,460 --> 00:05:46,740
But we were then and still are a pretty lean team.

62
00:05:46,920 --> 00:05:50,300
So it's not like we could expand beyond the one site.

63
00:05:50,380 --> 00:05:52,620
It was a large site in Hennepin, 80 megawatts.

64
00:05:53,380 --> 00:06:02,040
And so when we sold in 2022, we were excited to kind of take a lot of what we were learning and put it in practice.

65
00:06:02,040 --> 00:06:16,280
The big one really, like the sort of macro thing that we observed was our ability to develop institutional capacity, logic, financial modeling, software, hardware, you name it.

66
00:06:16,620 --> 00:06:26,600
That would say, okay, any given five minute energy trading interval we can develop if we know hash price and we know the efficiency of our ASICs, we have a strike price, right?

67
00:06:26,600 --> 00:06:28,920
This is now well understood back then.

68
00:06:29,020 --> 00:06:30,860
We were kind of new to it.

69
00:06:32,040 --> 00:06:41,280
We felt like we were at the vanguard. And so we said, well, could we put together renewable projects that take advantage of this insight?

70
00:06:42,080 --> 00:06:57,940
And so, but we couldn't do it until we sold. So we sold in July 2022 and then came back full force with a specific and exclusive focus on originating sites through distressed renewables.

71
00:06:57,940 --> 00:07:19,520
And then from late or so, let's say from 23, late early 2023 to late 2024, it was a lot of trying on different business model ideas that went from a core focus for a while on convincing large renewable independent power producers that they should be the Bitcoin miner and we would build it and operate it for them.

72
00:07:19,520 --> 00:07:26,440
That actually did have some traction, but ultimately the crypto went through then made it a bit difficult.

73
00:07:27,200 --> 00:07:42,620
And ultimately our business model settled on what it is today, which is we use our company, Sanger Renewables or Sanger Development, as a development platform, similar to like real estate finance or renewable energy project finance,

74
00:07:42,620 --> 00:07:46,360
whereby we use our development company

75
00:07:46,360 --> 00:07:49,920
to originate discrete project level financings

76
00:07:49,920 --> 00:07:53,760
where we then, once the project matures to a certain point

77
00:07:53,760 --> 00:07:55,420
and it's ready to be capitalized,

78
00:07:55,540 --> 00:07:58,980
we take an investment directly at the project level.

79
00:07:59,780 --> 00:08:02,520
So the balance sheet, the P&L, the assets,

80
00:08:02,640 --> 00:08:05,520
all of it is held at the project level.

81
00:08:06,080 --> 00:08:08,720
And our company, this has allowed us to be really efficient.

82
00:08:09,200 --> 00:08:11,500
We take some fees that really are a function

83
00:08:11,500 --> 00:08:18,100
of the profitability of the underlying cash flows. And that has forced us to become extremely

84
00:08:18,100 --> 00:08:23,000
disciplined around making sure that we're always cash flow positive, right? We're not playing

85
00:08:23,000 --> 00:08:29,300
a stock market, stock price appreciation game. We're really playing an IRR game.

86
00:08:30,420 --> 00:08:36,520
And so, you know, especially in times like this, you know, you have to be able to think through

87
00:08:36,520 --> 00:08:42,860
margin per megawatt hour. And then for us, as we grow into a portfolio of sites, so we think about

88
00:08:42,860 --> 00:08:48,260
margin per megawatt hour and all the adaptive strategies necessary to maximize that. And then

89
00:08:48,260 --> 00:08:53,980
as a company, as an enterprise, as a platform, how many megawatts under management can we get?

90
00:08:53,980 --> 00:09:01,240
So we're leveraging the strong momentum coming out of this 20 megawatt project with Potala

91
00:09:01,240 --> 00:09:06,580
energies into conversations with essentially like all of the large renewable power producers

92
00:09:06,580 --> 00:09:12,060
globally, not even just in the US, a massive portfolio of megawatts under management,

93
00:09:12,600 --> 00:09:17,600
all applying the same logic of being extremely disciplined on the cash flows.

94
00:09:17,920 --> 00:09:23,740
So we're excited to share that Sanger Renewables 20 megawatts facility in West Texas is now

95
00:09:23,740 --> 00:09:29,240
fully energized, developed behind the meter at a solar site and supported by energy partners,

96
00:09:29,240 --> 00:09:32,300
including Total Energies and Lynx Energy Partners.

97
00:09:32,820 --> 00:09:35,400
This project demonstrates a practical model

98
00:09:35,400 --> 00:09:39,660
for monetizing renewable generation using Bitcoin mining.

99
00:09:40,040 --> 00:09:43,160
Can you tell me more about these two entities?

100
00:09:43,420 --> 00:09:45,060
You mainly mentioned Total Energies

101
00:09:45,060 --> 00:09:48,340
and what kind of their problem is today

102
00:09:48,340 --> 00:09:49,600
and how you're solving it?

103
00:09:51,260 --> 00:09:52,000
Sure, sure, sure.

104
00:09:52,120 --> 00:09:55,160
I'll start with, we say Total Energies

105
00:09:55,160 --> 00:09:56,460
because it's a French company.

106
00:09:56,460 --> 00:10:06,660
so Total Energies is a top 10 energy conglomerate globally they were traditionally an oil and gas

107
00:10:06,660 --> 00:10:13,580
company and then over the last you know let's say 10 to 20 years especially in the U.S. have

108
00:10:13,580 --> 00:10:20,140
really gotten into the power business so with they own a fairly large portfolio of renewable

109
00:10:20,140 --> 00:10:28,180
generation of physical power plants in the U.S. and around the world. And then they also serve

110
00:10:28,180 --> 00:10:35,520
that portfolio in their capacity as traders and retail energy market participants.

111
00:10:35,520 --> 00:10:44,140
So they have sort of every part of the energy business conglomerated into their overall

112
00:10:44,140 --> 00:10:51,380
parent company. They, in this case, at the asset called Oberon that we're talking about,

113
00:10:51,940 --> 00:10:59,600
they own 150 megawatt operating solar facility. That facility, given its location and sort of the

114
00:10:59,600 --> 00:11:23,466
dynamics of the West Texas power market it a strong asset has a lot of solar production But as is the case with a lot of these assets a fair amount of the time their nodal pricing which is where you know they able to mark to market their energy just trades negative or at just a very low price

115
00:11:23,466 --> 00:11:38,946
So when we started conversations with them, the idea was, could we use both the co-location at the physical asset behind the meter, literally physically behind the meter on a parcel of land that's directly adjacent to the solar facility?

116
00:11:38,946 --> 00:11:55,546
And could we use their teams, retail energy trading and other energy service, you know, capabilities to put the appropriate regulatory, commercial and engineering sort of wrap around this project.

117
00:11:55,906 --> 00:12:05,146
And it was new for everyone. You know, there's a few people in the market doing something similar, but it's, you know, you can't point to a large volume of projects like this.

118
00:12:05,146 --> 00:12:19,186
So it was a rather intensive collaboration where, you know, every day between the legal teams, regulatory teams and, you know, the analysts and traders, we had to think through a lot of dimensions to what does this really look like?

119
00:12:19,186 --> 00:12:25,666
And it took us about 18 months from the time that we, I think, signed a term sheet till we actually went live.

120
00:12:26,066 --> 00:12:44,366
And all the way, we were just kind of chugging along on our side to ensure that we were capitalizing the project the right way and setting up sort of an LLC structure that allowed us to get the project off the ground and, you know, very, very effectively with a low cost of capital.

121
00:12:44,366 --> 00:12:51,666
So Lynx Genco, Lynx Energy Partners are actually a group that we co-develop with.

122
00:12:51,906 --> 00:12:53,646
So they're on our side of the equation.

123
00:12:54,326 --> 00:12:55,786
I can't say enough about them.

124
00:12:55,786 --> 00:13:07,386
They effectively, Lynx and Sangha are in many ways with respect to these projects, at least fully partnered, effectively one in the same team.

125
00:13:07,386 --> 00:13:19,206
So Lynx is a group of about six partners and some analysts that have many years of experience in, let's just say, energy markets.

126
00:13:19,866 --> 00:13:24,586
And so they helped us really on the fine grain aspects of the commercial negotiations with Total.

127
00:13:25,366 --> 00:13:33,806
And we are working every day with Lynx right now to have a similar conversation to all the rest of the large renewable IPPs out there.

128
00:13:33,806 --> 00:13:47,546
Okay, so to break it down a bit more simply, what you're saying effectively is, you know, you're giving Total Energies a hedge on how much they can monetize their kilowatt hours for.

129
00:13:48,126 --> 00:13:56,466
And whenever market conditions are, well, favorable to mine Bitcoin instead of selling it to the grid, they have an option to do that.

130
00:13:56,686 --> 00:13:58,886
Would you agree with that sort of summary?

131
00:13:58,886 --> 00:14:02,266
Buying Bitcoin miners should be simple, safe and fast.

132
00:14:02,846 --> 00:14:05,406
AltaEar technology delivers exactly that.

133
00:14:05,966 --> 00:14:08,586
Based in the US and shipping directly from St. Louis,

134
00:14:08,966 --> 00:14:13,346
AltaEar eliminates all the shady messaging groups, customs and tariffs issues,

135
00:14:13,566 --> 00:14:16,306
and makes sure Bitcoin miners get to you on time.

136
00:14:16,566 --> 00:14:19,926
AltaEar stocks leading commercial ASIC miners, efficient home units,

137
00:14:19,926 --> 00:14:24,486
and all the critical parts you need, backed by real inventory, real support,

138
00:14:24,666 --> 00:14:28,506
and a team of people that you can rely on and actually mines Bitcoin.

139
00:14:29,146 --> 00:14:31,846
Thank you so much, AltaEar technology, for supporting the show.

140
00:14:31,846 --> 00:14:34,386
And without further ado, let's get back to it.

141
00:14:34,626 --> 00:14:40,046
There are parts of the deal that you could say look like a hedge in nature.

142
00:14:40,726 --> 00:14:44,886
I think the core of the deal, maybe I wouldn't use the word hedge, I'd use the word floor.

143
00:14:45,566 --> 00:14:50,006
That is the word that we actually use in the actual documents.

144
00:14:51,026 --> 00:14:55,046
So yeah, I mean, I have to break it down without getting too in the weeds about it.

145
00:14:55,186 --> 00:14:59,186
You could think of it as there's three buckets of how we buy energy.

146
00:14:59,826 --> 00:15:04,366
The first is from a floor down to negative infinity.

147
00:15:04,366 --> 00:15:09,286
So we just index what we buy the energy for to the real-time energy market.

148
00:15:09,886 --> 00:15:12,006
And we put a floor in place.

149
00:15:12,206 --> 00:15:23,406
I won't say what that is, but there's a floor in place so that anytime there's a 15-minute energy settlement interval that settles at the floor or below, it could get negative infinity.

150
00:15:23,786 --> 00:15:28,306
Of course, it never goes to negative infinity, but negative 20, negative 30, which happens.

151
00:15:28,306 --> 00:15:40,466
We're buying it up to that floor price. So the core value proposition to Total was really the uplift they get by eliminating the exposure below the floor.

152
00:15:40,926 --> 00:15:47,946
Between the floor and our break even strike price in dollars per megawatt hour, we're just buying at the market.

153
00:15:48,706 --> 00:15:53,246
So it's the same kind of exposure they have when facing the ERCOP merchant market.

154
00:15:53,246 --> 00:16:11,506
And then when our break-even, which at the moment is much lower than it was a few months ago, when our break-even is surpassed by the real-time price, we're simply turning off the load and Total is able to push those electrons into the open market.

155
00:16:11,506 --> 00:16:29,826
Right. So if we look at you guys basically being that additional tool, give viewers a bit of context, you know, over the last year, over the last couple of months, how much uptime did you guys see? How much was that tool actually utilized? And what machine choice goes along with that?

156
00:16:29,826 --> 00:16:52,686
Yeah, so we spend an inordinate amount of time mapping our expectations for uptime, which is, of course, a function of the way we think about hash price and then how hash price equates to a strike price and how does that strike price actually play out in the real-time energy market.

157
00:16:52,686 --> 00:16:59,966
So we went live in late November around Thanksgiving time.

158
00:17:01,406 --> 00:17:09,906
And in that time, what we've observed is that, of course, this is a period of time in which our strike price is considerably low.

159
00:17:09,906 --> 00:17:25,786
So with hash price, given the machines that we're using, which average out to about roughly call it an 18 and a half tools per tera hash, where we're looking at in today's terms, we're looking at a strike price of around $70 a megawatt hour.

160
00:17:26,246 --> 00:17:32,286
If I can remember all the way back to early December, that was probably closer to 85 or 90.

161
00:17:32,446 --> 00:17:36,926
And if I think back to last summer, we were probably at like 120.

162
00:17:36,926 --> 00:17:42,206
um so really price quickly spencer explain the strike price to the viewers that don't understand

163
00:17:42,206 --> 00:17:48,786
sure sure sure it's it's it's a fairly simple thing if you have your hash price which is the

164
00:17:48,786 --> 00:17:54,886
revenue you know the commodity value of your revenue in it's just the unit adjustment so you

165
00:17:54,886 --> 00:18:04,486
say that's in um you know uh dollars per terahash per day if i know that and i know the number of

166
00:18:04,486 --> 00:18:12,046
units of energy that go into that equivalent amount of compute, and I take my hash price and

167
00:18:12,046 --> 00:18:18,526
I divide it into the number of megawatt hours I need to use to generate that amount of hash price,

168
00:18:19,126 --> 00:18:24,806
then now I have a dollar per megawatt hour strike price. And now that's what allows us to perform

169
00:18:24,806 --> 00:18:32,326
an arbitrage between revenue and cost. And so what we then do is we're just looking at actually

170
00:18:32,326 --> 00:18:37,386
every five minute interval, there's price discovery in the ERCOT merchant market.

171
00:18:38,046 --> 00:18:45,366
So we have various software and APIs and firmware that allow us to take a feed of the real-time

172
00:18:45,366 --> 00:18:53,366
strike price and a feed of the real-time cost. And we have an overlay and an algorithmic approach

173
00:18:53,366 --> 00:19:02,086
to allow us to turn off when we're functionally underwater on a dollar per megawatt hour,

174
00:19:02,086 --> 00:19:08,026
over dollar per megawatt hour basis. And so actually a conversation I have more and more

175
00:19:08,026 --> 00:19:14,066
with the energy companies, because it's vital to thinking through how do we actually put together

176
00:19:14,066 --> 00:19:20,186
the energy procurement, is that we try to explain that we're not really, sometimes as a shorthand

177
00:19:20,186 --> 00:19:27,526
in the past, I would say something to the effect of, we're really looking for access to a site

178
00:19:27,526 --> 00:19:36,446
that allows us to get to 95% uptime at $35 a megawatt hour all in or less. And those are both

179
00:19:36,446 --> 00:19:42,626
annualized figures. It turns out more recently, I realized that while shorthand is good because it

180
00:19:42,626 --> 00:19:47,426
allows you to communicate efficiently, it also kind of led to certain misunderstandings. And so

181
00:19:47,426 --> 00:19:53,046
the way I would describe it now, it's a bit more nuanced, is that what we're really solving for

182
00:19:53,046 --> 00:20:00,106
is that we need to see in our underwriting, which is a function of we look at pricing,

183
00:20:00,106 --> 00:20:06,366
both historically and on a forward-looking basis, and levels of actual production from the asset

184
00:20:06,366 --> 00:20:13,466
that would allow us to convince ourselves, underwrite for ourselves, that there's a sufficient

185
00:20:13,466 --> 00:20:21,646
volume and magnitude, or frequency, I should say, frequency and magnitude of profitable megawatt

186
00:20:21,646 --> 00:20:29,366
hours. And from there, we're able to back into whether there's a risk-adjusted, attractive IRR

187
00:20:29,366 --> 00:20:36,406
that we can bring capital into a deal. And so you had asked before which machines we're using.

188
00:20:36,406 --> 00:20:42,446
We actually, we're using a blend. We're using hydro cooling all around because when you're

189
00:20:43,166 --> 00:20:50,446
arbitraging, meaning curtailing and tuning the firmware as frequently as we are, we needed a

190
00:20:50,446 --> 00:20:55,166
robust infrastructure. Otherwise, you know, in an air-cooled environment, it's very physically

191
00:20:55,166 --> 00:20:59,326
taxing to the machines to be going out and off so much. So we use hydrocooling.

192
00:20:59,326 --> 00:21:03,226
Sorry, Spencer, to interrupt you there. Sorry. This is very important, right? Because everybody

193
00:21:03,226 --> 00:21:08,806
thinks, oh, Bitcoin is still super flexible, right? But these machines, you have to keep in

194
00:21:08,806 --> 00:21:13,966
mind, they're designed when they come out to run 100% of the time. They're not designed to be

195
00:21:13,966 --> 00:21:17,206
constantly thermal cycling. So that's just important to note.

196
00:21:17,206 --> 00:21:23,586
but that's right and so and and there's no perfect solution to it um but what what we tried to do

197
00:21:23,586 --> 00:21:31,586
and we were willing we thought that it made sense um in the overall pro forma for how we

198
00:21:31,586 --> 00:21:37,466
capitalize the project it was worthwhile to us to spend a little more on the capex

199
00:21:37,466 --> 00:21:46,326
to get hydro cooling machines and hydro cooling infrastructure so that uh we could have more

200
00:21:46,326 --> 00:21:53,766
reliable performance. So it allows us to overclock, it allows us to underclock, and allows us to tune

201
00:21:53,766 --> 00:21:58,166
the machines much more rapidly to put the hydro cooling in. Even then, of course, nothing's

202
00:21:58,166 --> 00:22:03,046
perfect, right? These are physical systems, you run into issues, sometimes it's the issues you

203
00:22:03,046 --> 00:22:11,306
knew about, sometimes it's some things take you by surprise. And so, yeah, we went with GreenHash

204
00:22:11,306 --> 00:22:12,986
are the containers we're using,

205
00:22:13,166 --> 00:22:13,626
which is, you know,

206
00:22:13,666 --> 00:22:29,812
hydro infrastructure And then about two of the machines that we using on are Watts Miner their hydro series the M63S And then more recently we brought the last third of the capacity on and we very pleased with this relationship

207
00:22:29,992 --> 00:22:33,532
We see a long-term future in the relationship with actually Auradyne.

208
00:22:34,132 --> 00:22:42,432
So Auradyne has some machines that are doing things a bit differently than the rest of the ASIC manufacturers.

209
00:22:42,432 --> 00:22:45,712
in particular what really got us excited about them.

210
00:22:45,832 --> 00:22:48,172
First of all, their team has been amazing to work with.

211
00:22:48,272 --> 00:22:50,992
We're getting the attention of the senior team

212
00:22:50,992 --> 00:22:52,732
and any questions we have,

213
00:22:52,832 --> 00:22:55,372
any requests for additional features,

214
00:22:55,452 --> 00:22:57,352
they've been very quick to respond.

215
00:22:58,072 --> 00:23:00,532
And the key one is that they've enabled us

216
00:23:00,532 --> 00:23:03,312
to rapidly tune.

217
00:23:03,492 --> 00:23:06,312
So both on the up throttle and the down throttle

218
00:23:06,312 --> 00:23:07,772
on ASIC performance,

219
00:23:08,112 --> 00:23:11,452
we're seeing them come in at sub minute,

220
00:23:11,452 --> 00:23:20,192
even sometimes sub 30 seconds, which is actually critical given that what we're trying to capture is the spread on energy, which trades pretty rapidly.

221
00:23:20,472 --> 00:23:32,652
So that has actually significantly boosted our margin is the ability to really perform almost to the standard of like a battery system, you know, the ability to throttle up and down.

222
00:23:33,312 --> 00:23:40,652
So briefly talk to me again, just very briefly, because you already said it's just critical to the next question on uptime.

223
00:23:40,652 --> 00:23:42,792
What uptime are you targeting?

224
00:23:44,072 --> 00:23:50,012
Yeah, we don't really target a specific uptime.

225
00:23:50,412 --> 00:23:51,632
What do you expect?

226
00:23:52,312 --> 00:23:52,832
Better word.

227
00:23:53,132 --> 00:23:56,832
Yeah, well, I can't answer that simply

228
00:23:56,832 --> 00:24:01,912
because it's fundamental to our design and our business model.

229
00:24:02,852 --> 00:24:05,432
So I'll try to do it as briefly as I can,

230
00:24:05,512 --> 00:24:09,212
but maybe we can go into this on a longer question format.

231
00:24:09,212 --> 00:24:29,032
But essentially what we do is we've developed proprietary hash price forward curves, which are actually very similar to the way energy analysts would look at L&P forward curves in the energy markets, which you can look at a month out all the way to 10 years out.

232
00:24:29,032 --> 00:24:39,312
And then you can extract some meaningful position about where you expect hash price to settle in any given hour going out 10 years.

233
00:24:39,612 --> 00:24:53,392
What we're really looking for is to see that if we know that hash price or we can project that hash price in defensible, conservative ways will equate to an energy denominated strike price.

234
00:24:53,392 --> 00:25:03,612
What we really just want to see is that that, again, is of a sufficient frequency and magnitude to generate an attractive return to our investors.

235
00:25:03,872 --> 00:25:05,192
And these are project level investors.

236
00:25:05,992 --> 00:25:19,412
So the whole thing in many ways is simplified to it costs us X hundreds of thousands or maybe into the millions, low, low single million per megawatt to physically build the facility.

237
00:25:19,412 --> 00:25:24,752
And then we need to amortize that into all of the megawatt hours in which we would run.

238
00:25:25,312 --> 00:25:34,412
And how often we would run and the margin within any single hour is a function of the energy forward curve overlaid against our own hash price forward curve.

239
00:25:35,072 --> 00:25:37,412
And so the answer is it depends.

240
00:25:37,932 --> 00:25:48,372
But to give you a range, to not completely wiggle out of the question, we're looking at certain assets right now that we can run exclusively behind the meter where we never import from the grid.

241
00:25:48,372 --> 00:25:53,692
generally you can only do that with wind assets also hydro assets but hydro is kind of its own

242
00:25:53,692 --> 00:25:58,612
thing so on wind assets for example we were we're close to getting to a term sheet on

243
00:25:58,612 --> 00:26:05,312
a 200 megawatt wind facility where we offered to do 50 megawatts of mining and what we found is

244
00:26:05,312 --> 00:26:12,352
that by downsizing the ratio i guess it increases the ratio but downsizing the mine relative so 50

245
00:26:12,352 --> 00:26:20,012
megawatts against 200 megawatts of wind, we can sustain about somewhere between 70 and 75% uptime,

246
00:26:20,012 --> 00:26:27,412
which generally would be too low for us. But the reason why in this case, it's acceptable is

247
00:26:27,412 --> 00:26:33,272
because for those hours, for those megawatt hours, we're looking at like a $25 megawatt hour rate.

248
00:26:33,872 --> 00:26:39,772
So the lower the rate, the more tolerance for a suppressed uptime. And then the last piece of

249
00:26:39,772 --> 00:26:44,932
the equation that we spend just a tremendous amount of time modeling and thinking through is,

250
00:26:45,512 --> 00:26:51,752
well, what's the combination of infrastructure and CapEx and development costs at which we feel like

251
00:26:51,752 --> 00:26:59,312
that would generate an attractive return? So we kind of go into each project and we look at it

252
00:26:59,312 --> 00:27:04,912
on a bespoke basis, whether we would go air-cooled or hydro-cooled, right? And we take all the

253
00:27:04,912 --> 00:27:09,292
physical considerations in as well. And we're always talking to vendors to understand where

254
00:27:09,292 --> 00:27:15,292
the latest technology is, which will actually perform the best. Can it perform for that specific

255
00:27:15,292 --> 00:27:20,892
use case based on that specific energy procurement strategy? And at the other end of the spectrum is

256
00:27:20,892 --> 00:27:26,492
what we have with Total, which of course, right now with hash price where it is, we're running at

257
00:27:26,492 --> 00:27:32,012
less than 95% of the time. We're running at even right now, something like 85% of the time.

258
00:27:32,012 --> 00:27:35,972
but we think over a long

259
00:27:35,972 --> 00:27:37,312
over a 10 year horizon

260
00:27:37,312 --> 00:27:38,612
which is what we underwrite for

261
00:27:38,612 --> 00:27:41,312
we can achieve about a 90% uptime

262
00:27:41,312 --> 00:27:42,352
with an all-in rate

263
00:27:42,352 --> 00:27:45,132
which includes the energy commodity charges

264
00:27:45,132 --> 00:27:47,452
what I'll call the wires charges

265
00:27:47,452 --> 00:27:48,712
so payments to the utility

266
00:27:48,712 --> 00:27:50,632
and taxes

267
00:27:50,632 --> 00:27:53,492
we think we can come to about a 90% uptime

268
00:27:53,492 --> 00:27:56,512
at about a 28 to $33

269
00:27:56,512 --> 00:27:57,752
megawatt hour figure

270
00:27:57,752 --> 00:27:59,452
fantastic

271
00:27:59,452 --> 00:28:00,332
that's amazing

272
00:28:00,332 --> 00:28:16,072
So the reason why I asked that question is, first of all, I really like that approach, which we see on more solar sites where you take, you know, higher efficiency machines, they turn on first, they respectively have a higher uptime.

273
00:28:16,192 --> 00:28:26,232
And then you have sort of lower efficiency machines that turn on later and have less uptime hours, but still get the ROI done because they cost less initial investment.

274
00:28:26,232 --> 00:28:35,692
So the question I have for you is, why not say to Total, look, we're only going to buy behind the meter.

275
00:28:36,252 --> 00:28:38,232
Our uptime is probably going to be less than 50%.

276
00:28:38,812 --> 00:28:50,272
And then we're buying used machines that are dirt cheap on the market, basically scrap value paperweights to make those ROIs insane, basically.

277
00:28:50,552 --> 00:28:51,652
Why not do that?

278
00:28:51,652 --> 00:29:02,012
yeah i mean i i think that actually if our underwriting uh believed that that was true

279
00:29:02,012 --> 00:29:06,092
we probably would consider it i just i think we're dubious about whether the math actually

280
00:29:06,092 --> 00:29:11,972
gets there some and it has it has to do with two things one is that you're still spending

281
00:29:11,972 --> 00:29:19,132
no matter how lo-fi them you can you know you can talk about s9s or something um even with s9s

282
00:29:19,132 --> 00:29:24,672
you're still having to sink the same amount of marginal CapEx into your infrastructure for your

283
00:29:24,672 --> 00:29:29,392
electrical infrastructure. There's development costs, right? So it's not the amount we spend on

284
00:29:29,392 --> 00:29:34,592
lawyers and engineers is not trivial. The amount we spend on civil, it all has to be done up to code.

285
00:29:34,592 --> 00:29:41,112
So there's kind of like a minimum bogey just to get onto the site. And you're going to absorb that

286
00:29:41,112 --> 00:29:49,032
no matter what into your marginal cost on your CapEx, like your cost per megawatt.

287
00:29:49,132 --> 00:30:04,312
So you're already kind of you're already in a certain amount. And then the other thing is that, you know, I mean, on solar, it's quite difficult. Like the most you're ever going to do is the capacity factor of the solar itself, which means you're capped at.

288
00:30:04,312 --> 00:30:11,272
it's really more like, again, you could downsize the mine relative to the size of the solar,

289
00:30:11,792 --> 00:30:19,472
and maybe you could get up to like a 40% runtime. But now you're still going to have effectively a

290
00:30:19,472 --> 00:30:25,552
cost on your nodal, if you're going to index that to the nodal cost, the nodal real time,

291
00:30:26,112 --> 00:30:31,452
with where energy markets are right now, it's not like you're going to do that and get a one

292
00:30:31,452 --> 00:30:36,032
cent rate or $10 megawatt hour rate. You're still going to get around, might be a little bit better

293
00:30:36,032 --> 00:30:44,792
than the same 28 to 33 I just gave you, but maybe you're going to get $3 off, maybe it comes to 25

294
00:30:44,792 --> 00:30:52,992
to 28. So now actually that lower strike price, because you have the NS9, I mean, S9, I haven't

295
00:30:52,992 --> 00:30:58,812
checked in a while, but if I had to guess, NS9 is probably generating something like $20 a megawatt

296
00:30:58,812 --> 00:30:59,732
hour gross.

297
00:31:01,732 --> 00:31:02,632
Maybe less.

298
00:31:02,812 --> 00:31:03,512
I don't even know.

299
00:31:05,652 --> 00:31:06,172
Maybe.

300
00:31:06,732 --> 00:31:07,252
Mental math.

301
00:31:08,292 --> 00:31:08,672
I don't know.

302
00:31:09,592 --> 00:31:10,852
It doesn't matter.

303
00:31:11,092 --> 00:31:11,992
We get the point.

304
00:31:12,712 --> 00:31:14,132
I was thinking of an S19.

305
00:31:14,332 --> 00:31:16,452
An S9 is probably $7.

306
00:31:16,452 --> 00:31:17,192
I can tell you

307
00:31:17,192 --> 00:31:20,412
this

308
00:31:20,412 --> 00:31:21,852
beauty right here.

309
00:31:22,652 --> 00:31:23,572
This M30S.

310
00:31:24,352 --> 00:31:26,412
Do you want to guess how much per kilowatt hour

311
00:31:26,412 --> 00:31:27,152
it makes today?

312
00:31:27,152 --> 00:31:32,492
I'm trying to remember the Jules per terahash on the M30.

313
00:31:32,672 --> 00:31:33,352
Was it around?

314
00:31:33,592 --> 00:31:34,352
It was like 38.

315
00:31:35,452 --> 00:31:35,792
Yeah.

316
00:31:35,972 --> 00:31:36,192
Yeah.

317
00:31:36,372 --> 00:31:39,052
So, yeah, it's probably about 15 bucks.

318
00:31:40,232 --> 00:31:40,372
Yeah.

319
00:31:40,832 --> 00:31:43,272
No, it's actually 33 per megawatt.

320
00:31:43,292 --> 00:31:43,612
Okay.

321
00:31:44,152 --> 00:31:44,492
Okay.

322
00:31:45,712 --> 00:31:46,632
So, it's not too terrible.

323
00:31:47,692 --> 00:31:53,552
But if my cost for the hours I run is already there, there's actually no margin in it.

324
00:31:54,252 --> 00:31:54,692
No, I hear you.

325
00:31:54,692 --> 00:32:00,912
I mean, what you're saying is, look, in the US, we have fixed cost irrespective of what machine we choose.

326
00:32:01,472 --> 00:32:05,032
Yes, the ASICs make up the largest chunk of the CapEx overall.

327
00:32:05,592 --> 00:32:11,852
But the part that's not variable, according to the machine, already makes the project, gets it to a certain base price.

328
00:32:12,092 --> 00:32:15,272
And if we want to clear that, it doesn't make sense to use the old machines.

329
00:32:15,272 --> 00:32:23,372
And then it makes more sense to buy power from the grid, get to 85, 90% uptime and just go that route.

330
00:32:23,372 --> 00:32:25,252
I hear that, but it just means it's interesting.

331
00:32:25,352 --> 00:32:27,632
It's an interesting thought because it means, you know,

332
00:32:27,692 --> 00:32:29,272
countries like South Africa where I live,

333
00:32:29,732 --> 00:32:33,592
where maybe the codes are less expensive to get to,

334
00:32:34,112 --> 00:32:36,892
where certainly labor fees and lawyer fees are lower,

335
00:32:37,212 --> 00:32:40,472
they actually have an advantage now because that fixed portion,

336
00:32:40,632 --> 00:32:43,932
that fixed cost portion is way lower, right?

337
00:32:44,092 --> 00:32:49,752
And I mean, you know, we would never install like AH3880 machines

338
00:32:49,752 --> 00:32:53,252
from Auradyne at, I don't know, I don't know what you bought them at.

339
00:32:53,372 --> 00:32:56,372
I remember four weeks ago, there were like 10,000 a pop.

340
00:32:58,032 --> 00:33:01,132
And installed it here on a solar site in South Africa

341
00:33:01,132 --> 00:33:02,652
where we have 30% uptime.

342
00:33:02,912 --> 00:33:05,152
You're never going to see that because you can't make the money back.

343
00:33:05,452 --> 00:33:08,492
That was just important for me to convey.

344
00:33:09,772 --> 00:33:13,012
BitX, solomining.de got it.

345
00:33:13,152 --> 00:33:14,952
Premium made in Germany.

346
00:33:15,132 --> 00:33:15,912
You want to go fancy?

347
00:33:16,392 --> 00:33:19,952
Go for NerdX or NerdQX++ for a lot of power.

348
00:33:20,352 --> 00:33:21,012
Worldwide shipping?

349
00:33:21,512 --> 00:33:23,632
Check out solomining.de today.

350
00:33:23,972 --> 00:33:24,312
Yeah, yeah.

351
00:33:24,412 --> 00:33:34,118
I mean I be curious to look at the numbers on that I highly skeptical that you can really make it work anywhere with solar specifically because it just the nature of solar production

352
00:33:35,118 --> 00:33:41,778
Solar has benefits that wind does not, in particular that it more predictably is producing,

353
00:33:41,978 --> 00:33:46,258
at least in US markets where you have 4CP, 5CP, 12CP.

354
00:33:46,918 --> 00:33:49,998
Generally, what you're going to find is that those peak intervals are being determined

355
00:33:49,998 --> 00:33:53,118
during periods of time where you can run on the solar asset.

356
00:33:53,118 --> 00:33:59,638
So just to speak to, I'm kind of naturally leading to a question that comes up a lot.

357
00:33:59,738 --> 00:34:03,098
I'll give an answer, even though you didn't ask it, which is like, what are the benefits

358
00:34:03,098 --> 00:34:06,498
of co-location with solar as opposed to just like a grid site?

359
00:34:07,298 --> 00:34:08,678
And it's a few things.

360
00:34:10,098 --> 00:34:13,898
And at various times, any one of these might feel like the more important one.

361
00:34:13,979 --> 00:34:21,398
So one is just what we're specifically solving for and filtering for are sites that present

362
00:34:21,398 --> 00:34:31,678
a negative node to load zone basis so that on an annualized, if you annualize it, the node should

363
00:34:31,678 --> 00:34:37,798
trade at anywhere between $10 and $15 a megawatt hour cheaper than the load zone. And you can,

364
00:34:38,058 --> 00:34:44,658
as a pure load serving substation at grid-tied site, the only way you can only buy at the load

365
00:34:44,658 --> 00:34:48,778
zone or the hub, you can't get access to normal pricing unless you're co-located with a generator.

366
00:34:48,778 --> 00:34:54,198
So that's 0.1, which is huge, $10 to $15 a megawatt hour is a gigantic advantage.

367
00:34:55,078 --> 00:34:56,838
Of course, that's on a projected basis.

368
00:34:56,998 --> 00:34:58,178
It's not every node.

369
00:34:58,318 --> 00:35:00,178
We get really into the weeds there.

370
00:35:00,498 --> 00:35:04,158
The other is that the shape of a solar production is exactly what I just said.

371
00:35:04,479 --> 00:35:14,919
When the rest of the grid in ERCOT that's running just on a grid connection is curtailing in the peak of the summer because they're trying to avoid the 4CP tag,

372
00:35:14,919 --> 00:35:21,898
We can actually avoid the 4CP tag because our net metering is still showing zero to ERCOT, but we're running.

373
00:35:22,018 --> 00:35:23,658
We're just running on the solar resource.

374
00:35:24,479 --> 00:35:25,758
So that's huge.

375
00:35:25,838 --> 00:35:30,038
It gives us a gigantic leg up on mitigating our wire charges.

376
00:35:30,959 --> 00:35:40,778
And then the third is the advantages from a speed to market and cost timeline in that we don't have to make any kind of major infrastructure improvements.

377
00:35:40,778 --> 00:35:48,178
Of course, there's still some work we have to do to go from the existing substation, all the medium to low voltage work and all the civil and what have you.

378
00:35:48,318 --> 00:35:59,658
But being able to piggyback on an IPP's existing physical and regulatory interconnection presents tremendous advantages in access.

379
00:36:00,479 --> 00:36:07,178
Of course, you have to be a sophisticated, reputable player for them to be willing to allow you to come onto their site.

380
00:36:07,178 --> 00:36:20,778
So that's one of the things that's exciting for us now is Total Energies is a very reputable, significant company. So the fact that we were able to get a deal done with a company of that stature speaks volumes to our credibility.

381
00:36:20,778 --> 00:36:29,258
to close out um before we head into the rapid fire um question spencer explain to me how you

382
00:36:29,258 --> 00:36:34,518
guys handled you know you said you're working a year on this deal first of all like what did

383
00:36:34,518 --> 00:36:39,698
these initial conversations look like for a big global player like this getting into sort of

384
00:36:39,698 --> 00:36:46,158
big combining and being affiliated number one and number two how do you navigate modeling when the

385
00:36:46,158 --> 00:36:53,959
market over the course of the year changes completely yeah so i think we've seen we've

386
00:36:53,959 --> 00:37:00,358
been working on this i think our first conversations with total uh were taking shape in

387
00:37:00,358 --> 00:37:06,778
november 2023 so it's actually been a little over two years we went live almost exactly two years

388
00:37:06,778 --> 00:37:13,979
after initial conversations and i think they were just sincerely curious um about like what is this

389
00:37:13,979 --> 00:37:21,558
help us understand it. By then there were already some nascent examples of project formats that had

390
00:37:21,558 --> 00:37:27,338
worked. So Luna had been working on a couple of deals that I think are public. There was

391
00:37:27,338 --> 00:37:32,998
an Xterra deal with, I think it was initially maybe with Compute North and then it changed

392
00:37:32,998 --> 00:37:37,678
to US Bitcoin. And now I think it's a marathon. So there were a few of these behind the meter

393
00:37:37,678 --> 00:37:43,898
deals, but not many. And so we were lucky to have some personal relationships with the right

394
00:37:43,898 --> 00:37:48,298
people at Tautologist, friends from college, that kind of thing. And they were able to at least get

395
00:37:48,298 --> 00:37:54,198
us access to the data. And I would say the first six months were just education. It was just,

396
00:37:54,558 --> 00:38:00,459
you know, it was education both about how Bitcoin mining works, what our economics look like,

397
00:38:00,538 --> 00:38:04,678
what we could withstand, what we couldn't withstand. And just as much was an education

398
00:38:04,678 --> 00:38:10,398
on who we are and why we should be trusted to operate a critical infrastructure at an asset

399
00:38:10,398 --> 00:38:14,638
that they own. And so we just worked our way through and little by little, I think it was

400
00:38:14,638 --> 00:38:19,618
just a learning process. I don't know that I could distill it to any one thing. I will tell you that

401
00:38:19,618 --> 00:38:28,078
we were the agents of momentum because we had to be, we were really hungry for the deal.

402
00:38:28,618 --> 00:38:31,698
Of course, Sotala is a gigantic company that has a lot of initiatives going on.

403
00:38:32,479 --> 00:38:37,058
And as time went on, as they saw that we were able to hang, we have a really competent legal

404
00:38:37,058 --> 00:38:43,678
team and Hush Blackwell has been a great firm for us. And we, you know, I think time and time again,

405
00:38:43,678 --> 00:38:49,238
just given how new this stuff is, we kind of had to get out in front and say, guys, we think it's

406
00:38:49,238 --> 00:38:54,558
this. Tell us if you think differently. But, you know, we think whether it's a regulatory question,

407
00:38:54,618 --> 00:39:01,278
a pricing question, the answer should be should be this. And we work collaboratively. And over time,

408
00:39:01,278 --> 00:39:07,578
I think Total felt the more kind of the American team, the U.S. Houston-based team that we were

409
00:39:07,578 --> 00:39:13,738
dealing with that are, call it, like more sort of mid-level and, you know, various divisions

410
00:39:13,738 --> 00:39:19,218
were willing to introduce us to the teams in France and in Geneva that are a bit more at the

411
00:39:19,218 --> 00:39:25,979
sort of close to the executive level. And I just tried at every step of the way to be credible,

412
00:39:25,979 --> 00:39:34,638
professional, serious, right? I spend a lot of my time thinking about how not to be perceived as

413
00:39:34,638 --> 00:39:40,058
sort of a cowboy who's just trying to say that, you know, we're always going to go up and to the

414
00:39:40,058 --> 00:39:48,138
right, which was like a lot of what was happening from 2019 to 2023, call it. And so we just try to

415
00:39:48,138 --> 00:39:55,318
be a professional developer. It so happens that we deal with a new commodity asset class called

416
00:39:55,318 --> 00:40:03,398
hash price. But we think there should be some sort of first mover advantage in so doing. And that

417
00:40:03,398 --> 00:40:07,018
kind of feeds into an answer to your second question, which is like, how do we think about

418
00:40:07,018 --> 00:40:12,298
a changing set of economics? I would say the answer to that question is probably the place

419
00:40:12,298 --> 00:40:20,078
where if we have a secret sauce, it lies in the answer to that question. And I can say for now,

420
00:40:20,158 --> 00:40:24,338
I mean, we could do a whole other hour on just breaking down, probably a few hours breaking down

421
00:40:24,338 --> 00:40:33,698
exactly how we get there. But I think the insight is that we, so as a meaningful anecdotal point,

422
00:40:34,398 --> 00:40:39,598
we're the first people we're aware of, certainly at any scale, and maybe literally the only people

423
00:40:39,598 --> 00:40:47,298
to have convinced a traditional bank to give us long-term, whole project-backed,

424
00:40:47,298 --> 00:40:52,398
low interest rate finance. And the way that we, and we, so from the beginning, we wrapped

425
00:40:52,398 --> 00:40:58,499
our whole way of presenting the pro forma and the project model around the expectation of what would

426
00:40:58,499 --> 00:41:06,419
a bank ask. And so you can't go in and present like a linear assumption on hash price to a bank

427
00:41:06,419 --> 00:41:11,558
that the first thing they're going to do is say, this is not linear. So we thought through, I'll

428
00:41:11,558 --> 00:41:17,198
just say for now, how to put a shape to a hash price forward curve, which is a distribution over

429
00:41:17,198 --> 00:41:22,778
time that adequately represents the cyclicality and volatility of the underlying cash flows.

430
00:41:23,578 --> 00:41:27,499
And then we also, that gives it kind of a relative distribution and a shape.

431
00:41:27,698 --> 00:41:32,698
And then we're very serious about how conservative we are in saying that, yes, there's that shape

432
00:41:32,698 --> 00:41:37,499
that plays out over time and across cycles, but the general trend line as you move out

433
00:41:37,499 --> 00:41:39,098
in time is margin compression.

434
00:41:39,878 --> 00:41:44,798
And so we said, as long as we can, within those parameters, still show that we can cover

435
00:41:44,798 --> 00:41:53,058
our debt service and meet all of our covenants, then we've effectively financially engineered

436
00:41:53,058 --> 00:41:56,778
all the exotic sexiness of it being Bitcoin out of the picture.

437
00:41:57,439 --> 00:42:03,338
And this is just energy denominated cash flows against energy denominated revenue against

438
00:42:03,338 --> 00:42:04,718
energy denominated cost.

439
00:42:05,718 --> 00:42:10,758
And as long as you're smart about how you put like loan covenants together and you have

440
00:42:10,758 --> 00:42:17,718
the right approach towards working capital and reserves and debt service reserves,

441
00:42:18,298 --> 00:42:24,439
then we're never going to be exactly accurate in any given hour, quarter, or even year.

442
00:42:24,698 --> 00:42:31,578
But in the aggregate over a 10-year term loan horizon, we should be directionally accurate.

443
00:42:33,018 --> 00:42:39,218
Love that. Spencer, thank you so much for all this time and the level of transparency you have

444
00:42:39,218 --> 00:42:41,198
given us. I could probably speak to you for another

445
00:42:41,198 --> 00:42:41,558
hour.

446
00:42:42,959 --> 00:42:45,058
Quick rapid fire round. Quick answers

447
00:42:45,058 --> 00:42:47,078
only, as difficult as it might seem.

448
00:42:47,618 --> 00:42:49,058
One word to describe the

449
00:42:49,058 --> 00:42:50,718
project now that it's live.

450
00:42:53,398 --> 00:42:53,878
Complicated.

451
00:42:55,258 --> 00:42:57,018
Biggest assumption that turned

452
00:42:57,018 --> 00:42:58,138
out to be wrong.

453
00:43:02,878 --> 00:43:04,318
That we could move

454
00:43:04,318 --> 00:43:07,158
the utility faster

455
00:43:07,158 --> 00:43:07,838
than we could.

456
00:43:09,218 --> 00:43:14,999
What one thing should future solar mining projects copy from you guys?

457
00:43:15,798 --> 00:43:17,038
Project finance.

458
00:43:17,798 --> 00:43:20,218
One thing that they should never copy.

459
00:43:25,858 --> 00:43:32,979
Apologies to Total if they ever listen to me on this and they know I say it with love, but letting the trading team lead the way.

460
00:43:34,558 --> 00:43:35,178
All right.

461
00:43:35,218 --> 00:43:35,738
I love that.

462
00:43:35,738 --> 00:43:39,738
Spencer, give people a handoff and where they can find more about you.

463
00:43:40,618 --> 00:43:46,979
Sure. Just sangarenewables.com or on my LinkedIn, Spencer Marr.

464
00:43:47,419 --> 00:43:49,499
I don't really do Twitter or anything.

465
00:43:49,919 --> 00:43:53,058
So yeah, our website or LinkedIn.

466
00:43:53,538 --> 00:43:54,338
Amazing. Cool.

467
00:43:54,658 --> 00:43:55,999
All right, guys, I hope you learned something.

468
00:43:56,118 --> 00:43:56,919
I know I have.

469
00:43:57,058 --> 00:43:59,358
And whenever that happens, I know it was a quality episode.

470
00:43:59,718 --> 00:44:02,378
Thank you so much, Spencer, for the education and the exchange.

471
00:44:02,439 --> 00:44:02,959
I really loved it.

472
00:44:02,959 --> 00:44:06,858
you have to come back at some point because i think i have a bunch more questions for you

473
00:44:06,858 --> 00:44:11,078
with that being said you guys know how this works leave a like comment subscription

474
00:44:11,078 --> 00:44:16,538
you know i have to say this the subscriptions are important why because it shows the sponsor that

475
00:44:16,538 --> 00:44:23,018
this is valued content with that being said i hope you survive hash price um follow spencer's model

476
00:44:23,018 --> 00:44:28,518
be very conservative about how you model you can only be surprised to the upside if it works at 25

477
00:44:28,518 --> 00:44:33,338
it'll work at 45 that being said all the best cheers see you next time bye