The Mini-Grid Business

Monetizing stranded electricity to reduce demand risk - A way towards 100% renewable mini-grids

• Nico Peterschmidt / INENSUS • Season 1 • Episode 4

Share episode

Send us a text

In this insightful episode, Nico Peterschmidt from INENSUS takes the helm, engaging in a fascinating dialogue with Kyle Hamilton of Nuru Sasu and Erik Hersman from Gridless. Dive into the challenges and innovations of managing surplus electricity in mini-grids, especially in the dynamic landscapes of the Democratic Republic of the Congo.

Kyle sheds light on Nuru Sasu’s pioneering staged implementation approach. While diesel generators serve as temporary bridges between stages of solar PV system expansion for electrifying bustling towns, smaller villages present a unique challenge in harnessing excess power.

Ever wondered about the confluence of cryptocurrency and sustainable energy? Erik delves into Gridless' remarkable success story of deploying bitcoin miners in hydro mini-grids. He explains why hydro mini-grids are easier for bitcoin mining than solar mini-grids. Nico uncovers how INENSUS uses machine learning controllers to overcome the challenge of integrating bitcoin miners into solar mini-grids.

The trio doesn’t shy away from the pressing hurdles: confronting issues like hardware theft, intricate import duties, and the prevailing skepticism around bitcoin. They explore potential alternatives, such as leveraging cloud services with GPUs tailored for AI applications.

By the end of the conversation, one thing becomes irrefutably clear: A future dominated by 100% renewable mini-grids, slashing demand risks, is not only achievable but also likely to set the new gold standard. And with this evolution, the doors to climate finance might just swing wide open for the mini-grid sector. Tune in for an electrifying discussion on the future of sustainable energy!

LinkedIN : 👉 https://lnkd.in/epAXws2S

Twitter: https://x.com/INENSUSgmbh/status/1704059464800739442?s=20

LinkedIn: https://www.linkedin.com/company/inensus-gmbh/mycompany/
Twitter: INENSUS (@INENSUSgmbh) / X (twitter.com)
Visit www.inensus.com for more info.

Speaker 1:

Solar mini-grids have turned from small pilots to an electrification wave. We were there when mini-grid regulation was established, when financial transactions were closed. We saw new technology thrive and companies fail. This is where we tell the stories. This is where we discuss the future the mini-grid business Powered by Enensis.

Speaker 2:

Hello, this is Nico. I am the CEO of Enensis. Today I am welcoming Kyle Hamilton from Nuru Sosu, operating solar battery, diesel hybrid mini-grids in the Democratic Republic of the Congo, and Eric Herzman from Gridless, an expert in decentralized data center operation based in Kenya. We are talking about stranded electricity and demand risk. Kyle is a senior manager of strategy and partnership in Nuru. For six years now he has worked in social and policy research in Congo on land conflicts, on land governance reforms, as well as agribusinesses. Eric is the CEO of Gridless. He has also led various other Africa-focused companies, among them BRCK, providing Wi-Fi technology designed and engineered in Kenya for use throughout the emerging markets. He also founded the iHUB, nairobi's innovation hub for the technology community, bringing together entrepreneurs, hackers, designers and the investment community. Kyle, you operate mini-grids. How many mini-grids of what size do you have?

Speaker 3:

Yeah, thank you, nico, great to be here. Nuru is currently operating four mini-grids ranging from 55 kilowatts, which was our original pilot project in Benin, north Kivu, and now we're currently operating our largest project at 1.3 megawatts, and across those four projects we have a total of 1.69 megawatts of installed capacity, with 2,800 customers being served.

Speaker 2:

Nice, kyle, how do you project electricity demand in mini-grids of these various sizes that you operate before you have electrified the community?

Speaker 3:

Yeah, that's a great question. So I think one of the biggest challenges for operating in these new frontier markets like BRC is there's really a lack of quality data out there. Congo hasn't had a census since 1984, since I was born. So how do we try to understand what is the demand, what is the willingness to pay? Well, we try to use as many data sources as possible. So we conduct market research surveys. We do surveys with a wide range of customers, from households to small, medium enterprises to social institutions, and then we also do data logging, where we connect a data logger to a large business or organization to capture about a week of data to better understand what their current energy demand is and what their future energy needs will be. And so we try to take all of that data and use that to make some assessments. But we also have found that it's even better to draw on existing data that we have from our current foresight.

Speaker 3:

So we try to draw on a wide range of data sources to fill in these gaps and develop some models looking at peak load and making some estimates based on all of that information that we have.

Speaker 2:

According to my experience, what drives demand in a mini-grid is the level of entrepreneurship available locally. In a large town or city, there are many entrepreneurs, and some of them will most likely be successful. However, in a small rural village, just one excellent entrepreneur can change the trajectory of the village economy completely. This one entrepreneur may not even live in the village before electrification, but may be attracted by the new opportunities arising from electricity supply. So how do you project the development of electricity demand in a village if you do not know where there will be successful entrepreneurs in the village down the line?

Speaker 3:

Great question, nico. We have two different kind of projects. We have kind of our core project pipeline, and those projects are really centered in urban DRC. So the Democratic Republic of Congo is one of the most exciting energy markets in the world. There are more than 16 cities with populations of 300,000 or more, and those markets already have significant unmet demand, and so even the cities that say 50% energy access. Well, those cities often have very underperforming energy services and so they're not able to meet existing demand, even among the customers who are technically being served with energy. And so we try to go into these communities where there's a lot of unmet demand, where there is insufficient access to energy access right now, and to provide energy services to these communities. We look at communities with densely populated areas where there's high concentrations of commercial industrial offtakers, where there's high concentrations of productive users, small medium enterprises, and we try to sell about 50% of our total output to these commercial industrial offtakers, which then enables us to provide energy access at scale to households, to small medium enterprises and to social institutions.

Speaker 3:

So that's kind of our core focus displacing diesel, providing energy access to densely populated urban communities.

Speaker 3:

Now when we're talking about rural yes, rural is definitely more challenging we still also look to sell about 50% of our energy output to commercial industrial offtakers.

Speaker 3:

Possible that can be a telecom communications tower, for example, that can be a hospital in the area, it can be a small agribusiness operation. But it is more difficult and right now we're kind of seeing in rural communities we need a significant amount of grant funding honestly to serve those communities. So, for example, we have two very, very remote rural mini grids in communities adjacent to Gramba National Park in order to get those sites up and running, to kind of provide alternative livelihoods that create the infrastructure for ecotourism in the future but also just provide new alternative livelihoods. Now We've had to mobilize quite a bit of grant capital to do that, whereas for our large urban projects we're able to bring in a wider range of blended finance to make those projects happen, from venture capital to development finance institutions et cetera, to get those projects built. They're much more bankable because we have significant unmet demand with a wide range of customers, from commercial, industrial to household.

Speaker 2:

Yeah, we've heard in other episodes in this podcast already from other mini grid companies that many of the mini grid companies are currently looking at the larger sites. Then you already said that DRC has many of these larger sites and they are still available and unelectrified. There's lots of opportunity out there, but we also shouldn't forget the smaller sites, because that is where the long-term play will actually be, as soon as the larger sites are all covered. Coming back to this design question, how do you design your system following the demand projection for your larger sites and how you design your system following your projections for your smaller sites?

Speaker 3:

Yeah, thank you, nico. For our larger sites we really try to hit 100% utilization within 12 to 18 months after achieving commercial operation date. At commercial operation date we're really trying to get as close to that 100% as possible. We try to design our sites to meet demand as much demand as we can, but we also try to undersize slightly so that we hit that 100% utilization rate, which then enables us to go back to investors and make the case to expand our operations there. For smaller sites, it is definitely more of a challenge, I will say at this point, given the ongoing demand in urban Congo, with so many 16-plus cities still needing to be electrified, I think right now we're really having to mobilize grant capital to subsidize the risk in order to make it viable.

Speaker 2:

All right, let's talk about what we call stranded electricity. Can you explain to our audience what stranded electricity in mini grids is and where it comes from?

Speaker 3:

Yeah, so stranded electricity is the electricity that we could generate and sell to customers, but we're not able to because there isn't enough customers to consume that electricity. With our big sites we plan to achieve 100% utilization and that means have zero kilowatt hours that we're not selling, zero percent stranded electricity that we're not selling. 12 months after commercial operation date, during that ramp up period, there are going to be kilowatt hours we're not able to sell. These are kilowatt hours that we have the capacity to generate but we're not able to sell to our customers because they're not connected yet. The demand isn't there yet. Whatever the case may be, that's the stranded amount of electricity.

Speaker 2:

I see, and then you probably take the risk of not being able to supply all the demand right, if you undersize to a certain extent.

Speaker 3:

Yeah. So what we'd rather do is undersize, hit 100% utilization and then go back to our investors and say, hey, look what we've achieved. This is really exciting. There's a lot more demand. We're seeing that there are still thousands of more households to connect. We're seeing that there are more businesses emerging. We need to connect and we can build that case to expand. So we usually design our sites, our site locations, to allow for expansion as well.

Speaker 2:

Kai, what is your renewable energy fraction? What percentage of energy is coming from solar, pv or from other renewable sources in your systems? What fraction is coming from diesel generation? Let me explain this quickly where this question is coming from. You said that you have 100% utilization rate in your larger systems and you have rainy seasons. You've got off rainy seasons. You have demand development over time, over years. There will probably also be seasonal patterns in consumption and all of that somehow needs to be buffered by the diesel generators. If you don't have any surplus electricity, any stranded electricity, then you probably have a relatively high portion of diesel electricity that is being injected into the distribution network, isn't it?

Speaker 3:

Yes, so we have a commitment to a 75% renewable energy fraction over the lifetime of our sites. And as they start and as we're ramping up and adding new connections and getting closer to that 100% utilization, we're more or less 100% solar or renewable sources at that point. But as we get closer to that 100%, you see the increase of diesel. And so for right now, in our flagship project, currently the renewable energy fraction over the project lifetime is still over 75% renewable, but it's actually below 70% right now because we've had to draw on more diesel.

Speaker 3:

That said, we're inactive conversations with a hydro business nearby to do an interconnection to displace all of our diesel from our GOMA site using hydro. So hopefully it will be 100% renewable again. But yes, that is definitely a challenge. And so, thinking about, kind of over the long term, what does that look like? Well, we're planning to add a second 3.7 megawatt site, building on our 1.3 megawatts in GOMA and do an interconnection between those two sites to kind of, again over the long term, ensure that we have that 75% renewable energy fraction achieved.

Speaker 2:

Can you tell me how stranded electricity is related to this other term that well drives the mini-grid sector, which is demand risk, and how dangerous demand risk is for a mini-grid company?

Speaker 3:

Yeah, demand risk is significant, because we need to have the demand in order to achieve 100% utilization as quickly as possible, because that ensures that your project is bankable and it means that you're able to generate the revenues you need to maintain and operate your business, but also allow for expansion efforts and replacement costs. And so, yeah, managing demand is a significant challenge.

Speaker 2:

All right, eric, I know that you have got a solution to monetize stranded electricity. What does that look like?

Speaker 5:

Hey guys, yeah, so thanks a lot for having me on, nico, and actually it's fascinating listening to what Kyle was talking about with Nuru, because I've known about Nuru since I think about 2015 or 2016. To have somebody with that much experience in mini-grid is great, because I'm new to the energy sector. I'm very much more experienced in the data and the connectivity side of the internet, so, anyway, yeah, so what we do at Gridless is really we work with these small energy producers that often run their own mini-grids in these rural communities, and what we do is we look for the ones who have stranded or wasted energy and we soak it up with a small data center put on site. Now, our data centers are a little different in that they Bitcoin mining data centers, which means that the computers are very specialized for just doing Bitcoin mining. What does that mean? There's a network of computers around the world that verify every new transaction on the network, and it's also where a new Bitcoin is generated, and so that's how new Bitcoin enters the world, by paying it out to us. So that's how we earn. Every 10 minutes, a new block is formed and is released as an award to Bitcoin miners around the world. That's how we monetize our business. That's our incentive for existing.

Speaker 5:

Now, bitcoin mining takes up a fair amount of energy. Each computer takes somewhere around three kilowatts, so that means it's fairly heavy. But that's exactly what we need in Africa, where we have small energy producers who need anchor tenants. We work with them to figure out if it's a new site, how much excess energy they'll have. Oftentimes in the beginning there's not going to have 100% usage.

Speaker 5:

I think Neurome might be a little unique in that, compared to a lot of the other organizations we speak with, who often have only 25 to 40% usage in the beginning, and then we come along and soak up the rest of that. Now that usage for them does include you could be a tea factory or a mobile phone tower as well as the community usage. But then we come in and take up the excess and then, as that community has surges in need throughout the day and evening cycles, we reduce the amount of computers that we have turned on so we can turn them on and off within seconds, and that allows us to be very responsive to their needs on their network. But yeah, that's what we do with Bitcoin miners and renewable energy mini grids in Africa.

Speaker 2:

Good Nice concept Stranded electricity does not only occur in solar mini grids, but especially also in hydro mini grids. Can you explain what the big advantage of hydro over solar is regarding Bitcoin mining?

Speaker 5:

Yeah, so you know. I mean, it's pretty simple, Everybody gets it. There's only so many hours of sun in a day, so you get eight to 10 hours of decent sunshine for solar and the rest. If you want to keep mining to be more efficient, you need to be able to have it 24 seven. The only way to do that is with battery storage, and battery storage right now is fairly expensive, so it makes it just more complicated. To get a return on your investment Takes longer. So hydro is better for us in that it might be a little bit cyclical and rainy and dry seasons in Africa, but it is still fairly well understood as what the cycles will be and with that certainty it brings us a better return on investment over time and we can fine tune our data centers to the needs of those rainy and dry seasons throughout the year.

Speaker 2:

Eric, you have recently signed a contract with Virunga on the deployment of Bitcoin mining in their Zengar Mina hydro mini grid in Zambia. What does that project look like?

Speaker 5:

Yeah, so Zengar Mina is very interesting. It's a new site for us. It'll be live next month. It's a 700 kilowatt site and they have stranded power, depending on the rainy or dry season, between about 300 and 500 kilowatts, and so we're going to be deploying a 20 foot container that holds about 160 Bitcoin computers, the data center, and that can soak up that excess energy and make it useful. We also have sites in Malawi, a 200 kilowatt site and another 500 kilowatt site in Kenya, and we'll be expanding to a few more throughout the remainder of this year.

Speaker 5:

The idea is that each of these sites that we go into, some of them have been around for years and just have not had an answer to their anchor tendency issue, which is how do they find an industrial offtaker that can actually be kind of geographically agnostic, as most of these places are way off in the middle of nowhere and larger businesses don't have a reason to be out in that area, and so we can be there. We can bring our own connectivity you do need good connectivity to do Bitcoin mining and we can set up in the middle of nowhere. It works just as well there as it does sitting in the city or in the outskirts of a city in the US or Europe. What's important to understand about Bitcoin mining as well and this is part of the basis for our existence is that we believe that Bitcoin mining needs to be decentralized further, and that means that while Bitcoin mining is very heavy in the US, russia, china, iceland, different parts of Asia and Europe, it does need to distribute itself further throughout Africa and Asia and Latin America.

Speaker 5:

And we chanced upon the fact that mini grids and small power producers have a big need here.

Speaker 5:

That need is built on the backs of an infrastructure that doesn't have a really good mini grid or any grid interconnect and that allows for us to walk in and really supply something that's a win-win for us and the power producer.

Speaker 5:

They literally are stranded without any other buyer, so it's very hard for them to be financially sustainable. When we come in, we're saying, hey, listen, we can have a very creative and flexible agreement where we don't have to be very strict on the amount of power, we get the uptime on it and they don't have to be very strict on the usage of it, and that allows us to both work very well together. We structure these deals as revenue sharing agreements so that we're both incentivized on the same thing and that has worked really well so far. They make a percentage of the Bitcoin that we generate and then they decide as a company how they want to liquidate that Bitcoin whether they want to hold onto it in their treasury or whether they want to liquidate it immediately into some other type of currency and utilize it to pay their operations.

Speaker 2:

Most of our listeners will probably be most interested in learning how many US dollar cents per kilowatt hour they can expect from a collaboration with you, if they work with you. Of course, I know this depends on the Bitcoin exchange rate and on the difficulty adjustment in the Bitcoin network and many things, but can you give them an idea of what we're looking at just roughly?

Speaker 5:

It's really hard to pin down for the reasons you just stated. The Bitcoin price is dynamic, so is the difficulty adjustment on a two-week rolling average. What I would say is that in a downtime you'll make a few cents US cents per kilowatt, but in a good time you'll make many more cents per kilowatt than you would selling to an expensive customer. So it varies quite a bit, but any of it any sense is better than what they're making today, which is zero because it's wasted energy.

Speaker 2:

True, true, and that's why we're having this podcast session today. Most of the mini-grid companies that are expanding in sub-Saharan Africa these days are solar mini-grid operators and they're running solar battery mini-grids. Kyle, you mentioned that you're running solar battery, diesel mini-grids. There are some of these companies out there, but many of them are being pushed towards 100% renewable energy and this means that they have, especially during dry seasons, especially during midday or in the afternoon, they have significant surplus electricity, excess electricity, stranded electricity available for things like Bitcoin mining. Now you already mentioned, eric, that, due to the short availability of sun during a 24-hour day, it is relatively difficult to amortize Bitcoin miners in a short period of time, because they just run for a few hours a day. As soon as you bring in batteries to store electricity, things become a little bit too expensive. Now we at the Nensos, we had a look at this some years ago and developed a controller which allows miners to run in an optimal manner using the battery capacity in a solar battery mini-grid, making use of that available electricity. What this algorithm does is it projects the consumption of electricity over the coming 48 hours using a machine learning algorithm, basically looking at the history of consumption, long-term history, finding patterns and projecting these patterns into the future to project the consumption of the coming 48 hours. In parallel, it retrieves weather projections from the internet and optimizes these weather projections again with machine learning, for the coming 48 hours. It says well, if the weather projection has said this, in reality, what happened was that this experience it deploys to the projection for the coming 48 hours. These two time series go into a deterministic optimization tool, a mathematical optimization tool, which then tells the miners please switch on now, because I know that tomorrow morning the sun will shine and the demand will be low. Please reduce the state of charge of your batteries during nighttime, using the battery electricity for mining, because then I have enough battery capacity available the next day to fill the battery. Use all the solar electricity, keep the miners running, supply electricity to the community and keep everybody happy and not lose any electricity. It basically optimizes battery usage.

Speaker 2:

Then we have a specific view on the latest technology of batteries, which is the lithium ion LFP battery technology. These type of batteries usually come with 6,000 cycles and a calendar lifetime of some 15 years or so, as the cycling is not that much usually and the calendar lifetime determines the end of life of these types of batteries in the solar mini grid. This means that you have free cycles available. Now these free cycles that don't cost you anything are basically free of charge. After all, you have stranded electricity which comes at 0 cents per kilowatt hour because it's stranded otherwise it wouldn't be used. If you use these extra cycles which you have bought already but you're not using, you have some free cycles which also come at 0 US dollar cents per kilowatt hour.

Speaker 2:

Now, if you deploy this algorithm, we believe that you can run Bitcoin miners in a very efficient way, even in solar mini grids, because during at least dry seasons, you can run many of the miners 24-7. We're doing that in Uganda with some miners in a small mini grid of some 60 kilowatt peak and 140 kilowatt hour battery storage LFP force close to one and a half years now. This is running quite well. We believe that we will soon be ready to then also offer this algorithm to other mini grid companies. Do you think, eric, that this is something that you and your company may be interested in, and under which conditions? Would you consider, then, to also look into solar mini grids?

Speaker 5:

We've been very curious about how to do this with solar since the beginning. It's something that we really do think is an attractive business if we can figure out how to unlock the ROI. What it all boils down to is what is the return on investment? It's just lower hanging fruit to be in the hydro world than it does in the solar. For a long time we've been thinking about hybrid wind and solar. Maybe that would be a good solution. But yeah, listen, if you can make the batteries run more efficiently and you can plot out the trajectory of that battery life and the uptime of the miners on a site, then you can run a model that you know what efficiency of mine are to run on there. The efficiency of miners is important. These computers come with different efficiency levels on their energy usage and their tarash, which is their compute cycles. In a world where you can know what the cost of energy is and you can project it and you can project the uptime of it, and you can determine to use a certain efficiency level of a Bitcoin mining computer because you know it costs this much less than the highly efficient ones, I think you could probably make that work. Of course, it will struggle a little bit compared to some of the other types of energy, like hydro. It doesn't mean it's a bad use of it either. It means it could be used.

Speaker 5:

I think there's actually two ways to think about it. One is you've already have sunk costs into batteries and you have some efficient use, like your system, to manage and run them. Then buying less expensive, lower efficiency miners that you know you're not going to need to get as high of a return on investment can work. The other way to think about it, too, is taking older stock Bitcoin miners that have already basically returned their ROI and are no longer going to be needed by larger companies who are buying the next most efficient machine. You can cycle those in. Now the ROI is even much less and the requirements to run them are less as well. Nobody's looking as deeply at CAPEX as they would if it was a very expensive computer that you're buying, however okay.

Speaker 5:

So that's meant me thinking about Bitcoin miners on the solar space. I think there's another really attractive thing to look at, too, which is if you're already on solar and you don't have to spend as much of your CAPEX on batteries, you could look at the usage of, like an AI, machine learning, rendering, farm data center. The reason why is that GPUs just don't use as much energy as the Bitcoin mining A6 do. You could also run those, and it comes with a different model and a different problem, which is that you don't know what the cost or the price of your compute cycle is going to be to the market. You have to build the sales and marketing team maybe to that where you don't have to do that for Bitcoin mining, but I think there is something there that's worth looking into and scratching an itch on, especially in smaller 50 to 100 kilowatt solar sites.

Speaker 3:

Eric, I think for us at Neuro we signed power purchase agreements with off takers and working with a data miner or a data center that would fall into that category. And you asked about demand risk earlier, nico. I think the stronger, the more solid the business. So business that has more years of operations, very clear loads and demand and that's a company that's not going anywhere, that's like very attractive. So I do think data miners, data centers, are very attractive businesses for us to work with. I'm just curious, when you signed a power purchase agreement, how many years you commit to typically what feels good to you in those power purchase agreements. What does that look like for you? And then I have another follow-up question after that.

Speaker 5:

Yeah, so, kyle, that's a good question, and this is why I'm glad we're talking Really. I mean, we feel we can operate off of a low number as far as years go. We could do something for even two or three years, because our designs are modular. We could just pick up the container and move it. Ok. However, we would much rather sign, like, a five-year agreement or longer, because we expect to be there for a long time, and really what we're trying to solve for is the problem that you deal with and other many good developers deal with too, which is who fills that delta between buildout and full-demand usage? Right, and what we want to do is take up that delta and, as you have a decreasing need for us, we decrease the number of miners that are located there and we move them somewhere else.

Speaker 3:

OK, yeah, so I think for us, I think it would be interesting to think about where are sites that would be really advantageous to have a data miner or data center for a long-term haul versus relative to the demand in the community, maybe again filling in that delta between the demand and the community and what the system might be able to provide during a certain window. For example, for us, even our largest sites, it will probably take a while maybe 12 months, 18 months to achieve 100% utilization. So I can see, especially in that window where we would be really beneficial to have a partner, a data center or a Bitcoin miner operating there with us. Two to five years. That could still be interesting in terms of making the project bankable to take to investors. Filling in those gaps are also really meaningful and just kind of absorbing some of the demand risk that Nico was talking about.

Speaker 3:

More rural or smaller projects I think that's where something more long-term is probably even more viable, because in those communities it's often just a telecom tower that is kind of the most stable long-term offtaker in that location. So I don't know, those are just some thoughts, but just curious how you would think about it. You talked about modular, portable. What would that look like for you to kind of start with us at one site and then move to another site as that one's coming online?

Speaker 5:

Let me first address something we've kind of talked about but not addressed directly, which is what we do as GridList is. We've written special software to for our operating system that allows us to do automated real-time demand response. What that means is, as the demand spikes in the community, it could be the local flour miller, the pochial miller, turned on his switch and now you have a spike in usage for him to start grinding and milling the flour. Or you have the normal evening spike in demand from 6 to 10 PM by the community. We automatically turn off these machines one by one, and it can be very granular so that the demand response is very tightly aligned. So your dump load is almost nil.

Speaker 5:

And what you're talking about if you're building a new site and we come in with you is exactly what we did here in Kenya. There was a new site that was built in December. We had our miners on site when it was commissioned, so they were expecting 26% usage from year one as they kind of built out and built up the community demand site and they were at 100% instead. That completely changed their financials and their numbers, and so it's a lot easier for them to raise money now for their next sites because we work with them. We go everywhere they go and we like to cluster.

Speaker 5:

So if we're in Kenya and they're going to be building out more sites in Kenya, or they're in Malawi, they're in Zambia, we just go. We go with them and we like to cluster geographically so we can more easily move machines. Sometimes it's as many as 160 machines like a container load, and sometimes it's as few as six machines 20 kilowatts worth of machines and we can easily then cycle those between sites as needed. Of course we've been doing hydro, which means our cycles are around rainy and dry seasons, but so I'm guessing they actually have a little bit different cycle than what we deal with right now. Does that answer your question?

Speaker 3:

Yeah, it does. Yeah, nico, just also to jump on what you said earlier about intermittency challenges with solar, I mean, that is why we think it is difficult to be 100% solar for mini grids and provide energy services with strong reliability and uptime. So for us, that's why we have to have diesel as a backstopping technology and, if we can, it displays diesel with other renewable energy sources like hydro, like I've mentioned. That's something that we're trying to do as much as possible. Having the diesel backstop at a 75% renewable energy fraction enables us to provide energy services with 99% uptime.

Speaker 2:

Yeah, the interesting aspect about the Bitcoin mining is that it does not require 100% availability of electricity. You can use it with, I don't know, 90%, 80%, 70%. If you go below 60% or 50%, then it probably becomes difficult regarding extended amortization times, but even that may work if you use some lower cap acts, lower efficiency miners, as Eric already explained.

Speaker 5:

Yeah, that's right. There's something interesting there in that we've had to actually write our code. We talk about automated demand response right, but we actually do automated supply response too. We plug in to the energy sector side as well to understand the supply of energy, and that way we can manage the dump load. We can manage anyway the need for our miners to become the dump load, and that changes the dynamics for both the power producer as well as the grid itself.

Speaker 2:

Eric, this is great. You have got the automatic demand and supply response for hydropower, where it rather is all about power response, and we have got the energy management for solar battery systems. Well, I think there is some room for us to work together. And then there is one more aspect which I touched upon briefly before but which we have not talked about in detail. Kyle, if you and you're probably being approached by your investors and potentially by the government, but maybe also by your users, with a request to reduce diesel fuel consumption and further improve renewable energy consumption, or increase renewable energy consumption and from my understanding, this request will come more often in the future and then there is another whole new range of funding opportunities.

Speaker 2:

If you go to 100% renewable energy All these CO2 mitigation and adaptation funds and so on they will suddenly become interested in funding 100% renewable mini grids. Now, if you do that, then you have a lot of surplus electricity, a lot of stranded electricity, and then Bitcoin miners or data centers which are flexible, which can work as a dump load or similar, can be a solution, because the revenue that these data centers generate for the mini grid companies are usually, in most of the cases Eric, correct me if I'm wrong, above the LCOE of your generation cost, like your generation LCOE. So therefore, you always have, even if the margin is very, very small, you never run into losses. You always have a small positive margin on that trade. And well, I think this is a good thing to have and this is something that provides you a solid business case. It provides your investors with probably a good feeling about their investment, and we hope that this, after all, will make Financing of mini grids also easier. Eric, do you want to contribute here?

Speaker 5:

So I just released a report a small report that I did last month, on mini grid energy and Bitcoin mining in Africa. It's basically this idea that there's a new way to finance, a new reason to finance mini grids. That's because you know it goes into the concessionary financing issues you were talking about earlier, kyle. The time it takes to do the financing, by the way, is for somebody new to the energy sector is, let me just say politely, a tar cake right, and there's feels like there's a new way, and the new way stems from this idea that you could have a geographically agnostic can so can be anywhere industrial offtaker by our first and last resort that can basically help you have usage of 100% from day one. That changes Considerably the financial profile of any mini grid that wants to build out.

Speaker 5:

I think that there is a whole slew of new energy Investors who will come in to that space Rather than ones who were used to operating here in Africa alone. Some of them are. Without getting into too much detail, I'll say that I think some of it. The current investors have built their models around an old system and they won't change easily and there is.

Speaker 5:

You know there's a. There's a way to bring in some new investors, to move that needle a little further and move new money and new capital into Pushing electrification to the edges in Africa where we need it the most. So I this is why I love being on this call. That's why you know, niko, when you said hey, the the podcast would be with somebody from Neurue, I immediately said yes, it took me like five seconds because Neurue is working in one of the hardest places in Africa to work building out mini grids and has been doing it successfully for Around seven, eight years. That's a solid track record. It's a company that's done it and stayed alive and Continue to grow throughout some of the hardest black swan events the world has thrown at it. So if somebody can do that with the old model, what do you think they can do with a new one?

Speaker 2:

Yeah, true, yeah, amazing thoughts. I think there is a lot of growth potential in this idea, I agree. Can I just say one? Can I jump in with one thought?

Speaker 3:

here, niko Neurue. One of the ways that we've made some of our sites bankable is because of some of these special innovative financing revenue solutions related to Renewable energy credits in particular. So we've sold quite a bit of peace renewable energy credits, which are a traditional I rec with social and environmental Co-benefits, and energy peace partners is the organization that's kind of responsible for verifying and validating and Processing those renewable energy credits. And so today we've sold renewable energy credits to help fund capex for our Two rural sites. We sold some credits to Google, for example, and then we've also sold Renewable energy credits to Microsoft to support street lights, public street lighting in communities where there were no street lights, and also to Expand to connect hundreds of new customers and harder to reach Pockets of the communities where we were already operating.

Speaker 3:

And then we're moving forward. We're planning to sell more renewable energy credits to Provide ongoing revenue for the project, to kind of make these projects bankable. Beyond renewable energy credits, there are many more things that are evolving and it's very dynamic marketplace with carbon credits and all the different Financial instruments that are coming to fruition. I think it's really exciting and kind of playing an important role in the sector.

Speaker 2:

Good, let's talk about some challenges Related to Bitcoin mining in Africa. One aspect is that these miners they have some value and therefore they may also be a target. Who secures these mining assets from theft and vandalism?

Speaker 5:

So, you know, we partner pretty closely with the power generation partner, and so that means that we are on site, so they have their security there, which obviously brings De-risking to the, the security of the assets. But we also provide our own security, both digital and in-person, so analog and digital, both and we do that so that we can have eyes on at all times. Yeah, we just, you know, we want to make sure things run well. Quite frankly, we haven't had a lot of issues yet. Our partners are incentivized for us being there as well, and they don't want to see anything happen either good, yeah, good luck so far.

Speaker 2:

I hope this stays like this. Have you had any issues, eric, with tariff regulation in countries you're active in? Some regulation, some legislations, do not really allow mini-grid companies to sell to third-party at At other prices than the regulated price. Have you met where have you run into, any issues in this record?

Speaker 5:

The regulations. Different jurisdiction I mean we haven't run into that yet, but we haven't been in some of the jurisdictions where that applies either. We've had a good relationship and able to structure the deals that work for both of us in the sites right right now.

Speaker 2:

What taxes apply to Bitcoin mining? Withholding tax, vat, services sold, I don't know. Financial transactions, I don't know. There are a lot of taxes that could apply. What actually applies and how do Authorities go about this?

Speaker 5:

so they're set up as data centers, so the same taxes that would apply to a data center applied to Bitcoin mining data center as well. It's a data center that's doing Bitcoin mining as a service, and so that is actually something that's transacted in another, in another jurisdiction. It really just is like a running a data center in that same country, which might be, you know, hosting Googles or my own blogs website on it. So the interesting thing is there's already rules and regulations around that that we fit right inside of nice.

Speaker 2:

There are several mining technologies on the market. What mining technology do you apply?

Speaker 5:

Yeah, there's a couple different ways of doing what we do. The cheapest and, for us, the most efficient and what's worked really well has just been ambient air, which means we just put big fans on things and push the air through faster, and that cools off the machines. You know there's also water cooling and and other types of cooling of these machines, but they actually bring extra expense on the CapEx side and the OpEx side if you're not careful. That's really what you don't want in a Bitcoin mining company. It actually is all about processes and operational efficiency, and so we have to be very tight on those type of things, and so ambient air works great. We have that running in all of our sites, have not had any issues with overheating. We use micro BT.

Speaker 5:

What's miners? There's only a few manufacturers of Bitcoin mining computers globally and we're really happy with what they offer. So what's miners? Very good for dealing with dust and bugs.

Speaker 5:

We've had bad issues where the rainy season comes and all the flying ants come up and then they get sucked into the fans and then it's just like dust and glue bug glue inside of the machines and they've kept running, which, of course, much lower efficiency, but they kept running. It took us three days with toothbrushes to clean those stupid things out, but we got it done. But you know that proved to us that what's miner is probably the best of the manufacturers thus far and Can deal with the different environmental Issues that you have in Africa that you might not have elsewhere, and gives us confidence that these things will run even under Extremely harsh conditions, including up to over a hundred degree heat. So that is important because we are right now operating in the Highlands but in sometimes, sometimes soon in the future, will be operating in the lowlands and when we do that there's gonna be a lot more heat there. We have to be able to make sure that these machines will run regardless of that.

Speaker 2:

Do you filter the air when you Basically suck it into the container or into the room where your miners stand?

Speaker 5:

Well, we at least bug netting over it now. I mean, we learned that lesson the hard way. We put dust filters on them if they're in a place that has more dust, and Basic netting, which also takes out some of the larger environmental issues like it could be bugs or it could be twigs, things like that.

Speaker 2:

Hmm, yeah, the miners rerun in our mini grids. They are also using filtered air, but what we Found is that, well, if you have a very good filter, you also need an extra fan which Improves the airflow, because if you don't do that, the miners don't run it full efficiency.

Speaker 5:

Yeah, right, the first fan that you need to add is the exhaust fan. That just sucks out more of that hot air faster, yeah, and you have to make sure you have a really good wall between your hot and cold side so that you don't have that hot air kind of mixing back in and making your job harder true, true.

Speaker 2:

And then we had an issue with lightning strikes. Of course there are ways of protecting sides against lightning, but what we found is especially the power supplies of the miners are a little bit sensitive against Overvoltage. Have you seen the same?

Speaker 5:

Yeah, of course we have to deal with the same kind of thing. We have to keep an eye on our power delivery units to make sure that you know fuses aren't frying. We have recently moved to a new power delivery unit supplier that allows us to change each energy sources fuse one by one, so that way it has a lot more granularity and we can fix it easier.

Speaker 3:

Lightning is actually a huge issue in Congo. I think Eastern Congo have get some of the most highest rates of lightning strikes in the world. Yeah, having a good lightning protection system is absolutely critical. We've definitely lost some street lights because of lightning strikes and now all of our street lights have lightning protection. It's amazing how we learn, isn't it Kyle?

Speaker 5:

Yes, it is.

Speaker 4:

We have a friend of ours who I think you know as well.

Speaker 5:

They do some Bitcoin mining for the Varunga National Park and the Congo as well, and they talk all the time about those lightning strikes. They have had major issues trying to deal with that and, yeah, it's real. So most people think of Africa as a country. They don't realize how different each park can be. So in one part, you'll be dealing with high heat, desert-like environment. In the next one, you're dealing with electricity strikes in the mountains and everything in between. And here's the beauty of it building that energy can be anywhere Solar and hydro and geothermal and wind can be in any one of these locations and you have to be able to just roll with it and deal with it and be creative in fixing the problems.

Speaker 2:

Let's talk about Bitcoin and its reputation. For a while, bitcoin does not have the best reputation, unfortunately. What People say that's being used for illegal businesses Fraud. It consumes a lot of electricity, which, for many, is also not a good thing. African governments are warning their citizens from using Bitcoin, for example, in Uganda. We've seen that recently. Some countries do not allow the ownership of Bitcoin at all. In most countries, bitcoin mining, however, is not forbidden, and some countries go into the exact opposite direction. Do you want to tell a bit more about how various governments think about Bitcoin and the mining opportunities in the various countries? Eric, yeah, I mean I can speak a little bit about it.

Speaker 5:

I'll say that, while Bitcoin mining and there's a larger crypto ecosystem Of other types of cryptocurrencies that have even made things a lot worse, bitcoin mining stands apart in that it is a decentralized cryptocurrency. It's not like the other ones In that it is something that is decentralized. It allows for nobody to be a single owner of it, which, as an asset, makes it a commodity, unlike the other cryptocurrencies, which are securities and therefore they fall under different regulatory bounds. What that means is, while there's different regulations in every country, bitcoin is a really hard thing to stop people from using. Now the funny thing about people bringing up how it can be used for crime or fraud. The truth is, it's a lot easier to use cash for the same thing, because Bitcoin mining is done on an open ledger which anybody in the world can see every transaction. So it's really not the best currency to use for that type of activity, and it makes it a little bit harder to cover your tracks, whereas you know cash. By the way, if cash was created today, if we hadn't had cash for the past 80 years or whatever you know our fiat cash then it would probably be made illegal too. Right If it was just created today, because it's something that is hard for the government to control, and I think we're moving to a space where, you know, the government is trying to not just control everything around currencies. They're trying to deeply understand how everybody is using it and then control the usage of it by individuals. With centralized bank, digital currencies and CBDCs are coming We've seen them in Nigeria and China and other places to more or less affect. I think anything in that space gets a little bit scary.

Speaker 5:

What I think is interesting about it, though, is this is that Bitcoin mining is something that, yes, it does take a lot of energy, but it is exactly what we need in Africa. So in other parts of the world, you might say well, bitcoin mining is taking energy that should be used for something else. I won't get into the morality of that question, because you could make the case that Bitcoin mining takes as much energy as it does as clothes dryers do every year in the US. Is one more morally right than the other? You decide.

Speaker 5:

However, in Africa, we don't have that problem. What we have is a problem of more than half the world's population that doesn't have electricity is here, so if we want to push electricity out to the edges, where those people who need it most are, then we need somebody, a type of industry, that can be the anchor tenant and the buyer of last resort that allows these energy companies to expand and spread further. So us being large users of energy and being creative and reliable, and be modular in the way we can move around is actually exactly what we need, and it allows for energy to push to the edges. It allows for it to be financed differently. It allows it to be profitable in a way that it hasn't ever been before.

Speaker 2:

Bitcoin miners fall under the category of computers which carry high import duties in Texas. This makes miners quite expensive when they arrive on site. In an African country, if one found a business line that the government considered purely beneficial for the country, tax holidays and similar may apply. At Inensis, we're currently thinking about exchanging Bitcoin miners for other cloud computers which deliver cloud services for AI applications or similar. You already mentioned this, eric. This may attract interest of governments and therefore may attract tax and duty alleviations and similar, which may finally attract international investors more easily. Kyle, what is your experience with import taxes and DRC?

Speaker 3:

Yeah, thanks, nico. So you know, congo is one of the most challenging places to operate a business in the world and, for sure, issues related to taxes and import duties are a significant aspect of that. Your costs in terms of importing and clearing logistics even because there aren't a lot of great roads and infrastructure in Congo, it's just enormous costs for the business. So importing, clearing, you can plan to spend, you know, 40% plus on that. Taxes, also very significant VAT, 16%. It's definitely a really challenging aspect of the business and not only paying the taxes that you owe but also making sure you only pay the taxes that you owe can also be a challenge. So making sure that you spend a lot of time to really understand what you owe to whom at what times, and then having the team ready to kind of navigate any other additional inquiries to additional taxes is definitely a big part of managing a business in a place like Congo.

Speaker 2:

Eric, have you got any experience in decentralized cloud computing services outside of the Bitcoin space? I think you talked about AI applications already. Have you tried this? Who would buy these kind of services?

Speaker 5:

Yeah. So first of all, I'd say that what Kyle was just saying about import duties and everything is true we also have to pay attention to that, but it's actually more for our components, for the Bitcoin mining containers, than for the computers. Depending on the jurisdiction, those can come in even duty-free. Speaking of AI kind of machine learning, those kind of data centers, rather than a Bitcoin mining data center, it should be noted that you know, the caveat there is that they can't have as much downtime. Bitcoin mining makes money every 10 minutes approximately in blocks, and so it's always making money.

Speaker 5:

But if you turn them off on the AI kind of data center side, then you do risk having some problems with somebody who's running a process and is doing some computational work at that time when it goes off, so you do have to just bear that in mind. However, I think that it is a really interesting space to look into. I think, specifically because it uses less energy, it can be really useful on the solar side. It's at least worth a try. I think somebody needs to do some testing on that and just run a pilot on something that's, you know, 20 to 100 kilowatts and see what happens.

Speaker 5:

Now the problem is, besides that reliability of the energy supply to the computers for AI, you also don't have a well-formed market for computational work in the world yet. It's growing and there's something there. We modeled this out two years ago and it's come a lot further since that time to see that there is more places to sell that computation than there was before. But you do have to have a different type of team that does that and you do have to pay attention to the reliability and your uptime compared to Bitcoin mining.

Speaker 2:

Thank you, Kyle. You're expanding Nuro's footprint in DRC heavily. Do you want to tell us a bit more about your plans and if cloud computing or Bitcoin mining will play a role in that?

Speaker 3:

Yeah, thank you, nico. I really appreciate the question. So Nuro is in a very exciting moment of our business. We've been working on Series B financing for the past two years, about to scale our business. From the current four mini-grids at a total capacity of 1.69 megawatts, we're planning to add an additional 13.7 megawatts across three new sites, so a 3.7 megawatt site in Goma, a two megawatt site in Kendo and an eight megawatt site in Buña, all in the eastern part of the Democratic Republic of Congo. So we're really excited about that.

Speaker 3:

That's gonna take us from about roughly the 2,800 customers that we have today to an additional 30,000 customers, hopefully in the next 18, 24 months or so. So we're really excited about that, and then, even beyond that, we're hoping to close Series C, maybe late this year or early next year. That will add an additional 25 megawatts and serve tens of thousands of more customers as well. I do think there is a real opportunity to work with data centers and Bitcoin miners. We don't have any of those currently in our pipeline, but that's something that we would be very eager to like. We talked about demand risk before having really solid customers anchor loads, and so finding clients like data miners and data centers is really critical to our business model, and so we would definitely be interested in finding more to work with.

Speaker 2:

Thanks, Kyle. Eric, what are your plans for Gridless over the coming five years or so?

Speaker 5:

Well, at first it could be a sign of deal with Colin Neuru. Obviously I mean that's the lowest hanging fruit for everybody.

Speaker 5:

Yeah, listen, we have the Kenya, malawi, zambia sites going and we are gonna continue to cluster and grow in each of those countries. We're gonna probably add one or two more over the course of the next 12 months as well. I'm really interested in this space of combining and bundling energy plus Bitcoin mining in the financing of new sites, and so I will be pushing more into that space in the coming I don't know six months. You know. I think that that's something that can work with current energy generation partners as well as new ones, and I think it works with the brownfield sites that are maybe expanding as well as greenfield sites that are new. So I'm gonna be looking for partners on that as well. That, I think, work well, because we might know Bitcoin mining and data centers and internet connectivity, but we don't pretend to be the energy experts at all.

Speaker 2:

All right. Thank you, kyle. Thank you Eric. We have had a very nice discussion today. I really enjoyed it. Thanks to you both and see you soon.

Speaker 5:

Yeah, nico, thanks a lot for having me on and Kyle, Great to talk to you as well, nico thank you so much for hosting us, Eric.

Speaker 3:

great to get to know you a little bit. Really excited to hear more about what you're doing.

Speaker 4:

Hi, I am Vanessa data analyst at Inensys. In this podcast we dive deeper into the topics of stranded electricity and demand risk. Demand risk comes from over-sizing the mini-grids based on how the future demand is projected. Using reliable data is a challenge in the sector. Mini-grids companies like Nuru use surveys and data logging to estimate the demand. For large sites. Nuru aims at 100% utilization within 12 to 18 months of operation, as the demand needs some time to pick up.

Speaker 4:

Mini-grids in general have a lot of unused electricity or stranded electricity. To reduce stranded electricity, gridlers offers decentralized Bitcoin miners to consume the excess power, generating extra revenue for mini-grids. They are exploring the use of these miners in solar PV battery mini-grids too. Bitcoin miners do not need uninterrupted electricity, as each process is completed in few minutes. This makes it suitable to take up the stranded electricity from the mini-grids. When the demand is high, these miners can be switched off immediately.

Speaker 4:

Bitcoin mining in rural, remote places comes with some new challenges, like theft. Gridlers takes both digital and analog ways to secure these miners. Another key challenge is the tariff, which really depends on the country and its own regulatory framework. Bitcoin miners can also get very expensive due to high taxes and duties To facilitate tax, holidays or GT alleviation is needed. These miners themselves may have performance issues in hot and humid weather. This can be resolved by using fans and air filters. Unfortunately, bitcoin has its fair share of criticism, from fraudulent activities to consuming too much energy. On the flip side, with this open ledger, it is also very transparent. The mini-grid sector is changing dynamically. Players are looking to tie up decentralized cloud computing services with stranded electricity in mini-grids. One example is the data centers built around AI-based services. Although it needs uninterrupted supply, innsos has been developing innovative AI-powered solutions to tackle this challenge. We foresee next generation mini-grids will integrate cloud computing services in new mini-grids from the beginning. This will reduce the demand risk and make the best use of stranded electricity.

Speaker 1:

This episode of the mini-grid business has been brought to you by Innsos, your one-stop shop for sustainable mini-grids. For more information on how to make mini-grids work, visit our website, innsoscom, or contact us through the links in the show notes. The mini-grid business powered by Innsos.

People on this episode

Head Shot

Nico Peterschmidt

Host