WEBVTT
00:00:09.391 -->
00:00:13.935Solar mini-grids have turned from small pilots to an electrification wave.
00:00:13.935 -->
00:00:21.265We were there when mini-grid regulation was established, when financial transactions were closed.
00:00:21.265 -->
00:00:25.512We saw new technology thrive and companies fail.
00:00:25.512 -->
00:00:28.282This is where we tell the stories.
00:00:28.282 -->
00:00:34.433This is where we discuss the future the mini-grid business Powered by Enensis.
00:00:36.140 -->
00:00:37.142Hello, this is Nico.
00:00:37.142 -->
00:00:39.886Welcome to our episode on Mesh Grids.
00:00:39.886 -->
00:00:46.015My guests are Afnan Hannan from Okra Solar and Felix Boldt from Solarworx.
00:00:46.015 -->
00:00:51.231Afi is a co-founder and the CEO of Okra.
00:00:51.231 -->
00:01:00.027Before founding Okra, he was working in building automation and IoT and then transitioned to energy access through mesh grids.
00:01:00.027 -->
00:01:08.694Afi aims to unlock opportunities for last mile communities around the world with access to affordable energy.
00:01:08.694 -->
00:01:18.375Felix is the founder of Solarworx, with 10 years of experience in solar off-grid in sub-Saharan Africa.
00:01:18.375 -->
00:01:24.147Initially, he started his journey with Mobisol and is now with Solarworx.
00:01:24.147 -->
00:01:29.975He and his team are manufacturing standalone off-grid solar systems and mesh grids.
00:01:29.975 -->
00:01:35.712Solarworks also engages in EPC business in Africa and Germany.
00:01:35.712 -->
00:01:42.612Well, now, as we're talking about mesh grids, Felix, what is a mesh grid?
00:01:43.480 -->
00:01:44.322Yeah, hi everyone.
00:01:44.322 -->
00:01:46.450Thanks very much for having us in the first place.
00:01:46.450 -->
00:01:49.147So just straight to the first question.
00:01:49.740 -->
00:01:53.731So a mesh grid is a decentralized DC microgrid.
00:01:53.731 -->
00:02:05.030It's still an interconnection of systems, similar to an AC mini grid, but everything runs on DC, as the name indicates, and it's meshed because it's decentralized.
00:02:05.030 -->
00:02:22.669So you have multiple solar systems across a community and not everyone necessarily has to have a solar system, but in our case, usually 10 or so of the connections within the village have a solar system and that aggregated power of those solar systems is shared across the community.
00:02:22.669 -->
00:02:33.347So let's say, 90% of the users just have a mesh grid box and 10% actually consuming energy from the solar systems that are feeding in the power into the grid.
00:02:33.347 -->
00:02:43.689There's a couple of reasons why we're doing DC versus AC, but the main driver here is the cost, because DC technology it's safe to touch and therefore there's a lot of less safety hazards involved in that technology.
00:02:43.689 -->
00:02:53.445And therefore there's a lot of less safety hazards involved in that technology and therefore we can drop the cost of it and make it an easier entry point for rural off-grid communities across Africa.
00:02:53.466 -->
00:02:57.782Yeah it's safe to touch as long as the voltage stays below a certain level.
00:02:57.782 -->
00:03:03.100Right, and that's what you're basically doing, exactly, yeah, but that also increases the current right.
00:03:03.100 -->
00:03:22.073To transfer a decent amount of power at a low voltage, you need a relatively high current, and that causes losses doesn't it solar home system company that we are as well?
00:03:22.112 -->
00:03:30.078we are also sourcing fridges and freezers and tvs, and the consumption of a ac freezer in our case would be only 50 watts round about there.
00:03:30.078 -->
00:03:40.036So the peak load capacity of those dc appliances is just significantly lower and therefore we can tolerate higher degrees of losses yeah, and there's probably another advantage.
00:03:40.336 -->
00:04:13.092most of our listeners are of course familiar with ac mini gr, where you have a central generation station somewhere and then from there the electricity is distributed to all the customers over a relatively long distance, up to 1.3, 1.5 kilometers over low voltage lines Now in DC, and the mesh grids you basically generate in a decentralized manner within the community, so the distance from the production of electricity to the consumer is not that long.
00:04:13.092 -->
00:04:20.451Therefore it may also be reasonable to accept the slightly higher losses due to the low voltage.
00:04:20.451 -->
00:04:23.675But save a lot of costs, save a lot of capex after all.
00:04:23.675 -->
00:04:28.605Right, thanks to the lower installation costs and the lower material costs.
00:04:28.906 -->
00:04:32.380Yeah, thanks, Felix and Nico, for explaining a bit about mesh grids.
00:04:32.380 -->
00:04:48.125And just to add a little bit to kind of imagine how the cost savings happen and why it happens from a bit of a kind of engineering perspective, imagine you've got, let's say, a 100 kilowatt mini grid and a 100 kilowatt mesh grid.
00:04:48.125 -->
00:04:50.995Let's try to imagine how these two things would be different.
00:04:50.995 -->
00:05:00.350So in a 100 kilowatt mini grid you've got 100 kilowatts at the powerhouse, all the panels, all the batteries at one point, and then you're sending that power.
00:05:00.350 -->
00:05:03.706Let's say you know one or two kilometers, as you mentioned.
00:05:03.706 -->
00:05:21.910In the mesh grid you've still got 100 kilowatts, but those 100 kilowatts would be distributed amongst all of the rooftops in the community and given, let's say, an average distance of like 20, 25 meters, the power is not only traveling a much shorter distance in the mesh grid, but because all the generation is distributed, it's only a fraction of the power that's going.
00:05:21.910 -->
00:05:26.567So if you had 100 houses in the mesh grid, but because all the generation is distributed, it's only a fraction of the power that's going.
00:05:26.567 -->
00:05:31.124So if you had a hundred houses in the mini grid, you're sending a hundred kilowatts over the line.
00:05:31.124 -->
00:05:40.259In the mesh grid, let's say a hundred kilowatts with one kilowatt on each roof, you're only sending one kilowatt over the line in that simple kind of example.
00:05:40.761 -->
00:05:46.124And then finally we know that power loss is proportionate to I squared R.
00:05:46.124 -->
00:05:51.107So in the mini grid 100 kilowatts squared is 10,000.
00:05:51.107 -->
00:05:53.208So 10,000 times power loss.
00:05:53.208 -->
00:05:56.689In the mesh grid one kilowatt squared is one.
00:05:56.689 -->
00:05:59.612So basically one times power loss.
00:05:59.612 -->
00:06:04.735So in that mini grid you would have 10,000 times more power loss over the line.
00:06:04.735 -->
00:06:12.247And that's why in the mini-grid you need higher voltages, thicker cables, bigger poles, and that's where all the distribution costs add up.
00:06:12.247 -->
00:06:17.812And in the mesh grids you can use that lower voltage because you're having less power over the line.
00:06:17.812 -->
00:06:21.750So hopefully that explains a little bit more as well on the distribution cost savings.
00:06:22.360 -->
00:06:33.286I think there's maybe, if I can jump in also there's benefits in terms of mesh grids operational costs, because energy theft is a major concern for utilities in Africa and also for mini grid companies.
00:06:33.346 -->
00:06:41.428I can imagine, and if I'm not mistaken, it can yield up to 30% of the entire energy generated in utility grids and AC mini grids being stolen.
00:06:41.428 -->
00:06:46.288You know, after some time of installation of course I wouldn't say that's the first day, but it's going to develop that way.
00:06:46.288 -->
00:06:56.166So with a mesh grid for being a DC grid, the power line is pretty useless for the end customer, so they cannot just bridge the meter.
00:06:56.166 -->
00:07:04.750If they do that, they're going to fry all their DC low voltage appliances because they're taking that 60, 70 volt DC, jump that over and just destroy everything they have.
00:07:04.750 -->
00:07:06.413They try that once but never try it again.
00:07:06.413 -->
00:07:12.283So there's not really a high degree risk of theft and I think that's a major part here and again.
00:07:12.283 -->
00:07:24.449The security aspect is also very important because if you can inclusively involve the community in the installation initially because it's safe to touch then that drops the cost of installation also dramatically.
00:07:24.839 -->
00:07:25.300Yeah, it's safe to touch.
00:07:25.300 -->
00:07:26.523Then that drops the cost of installation also dramatically.
00:07:26.523 -->
00:07:59.190Yeah, we are going into direction where we're discussing DC mesh grids versus AC mini grids, Whereas I see that there is a lot of complementarity even between those two, where AC mini grids, for example, could be established where there is demand for I don't know tens of kilowatts of power for small industrial production or large workshops or these kind of things, where most likely AC three phase power supply is well suited.
00:07:59.190 -->
00:08:17.156And then there may be areas where there are more residential customers, where there are markets, where there are shops, where there are maybe hairdressers and these kind of small businesses, where DC systems like mesh grids may be very well suited.
00:08:17.156 -->
00:08:21.576So why not operating two systems under one management?
00:08:22.117 -->
00:08:22.779Yeah for sure.
00:08:22.779 -->
00:08:25.166So I could jump in there.
00:08:25.166 -->
00:08:28.863I might do a bit of a segue and then come back On the segue.
00:08:28.863 -->
00:08:31.550It's basically about DC mesh grids.
00:08:31.550 -->
00:08:43.389So I think the name can in many cases be misleading because the distribution is in DC for the reasons mentioned safety, power theft, reduced cost.
00:08:43.389 -->
00:08:57.490But in our mesh grids that Okra provides, every single output is configured with AC power output so that households can use the appliances that they're used to purchasing at the market, and so not all mesh grids are the same.
00:08:57.490 -->
00:09:05.267I think the SolarWorks mesh grids are slightly different to the Okra mesh grids and we also have slightly different philosophies on how to build it.
00:09:05.267 -->
00:09:08.630So I think that's maybe something interesting we can touch on in a bit.
00:09:08.691 -->
00:09:35.379But I guess, for the Acra side of things, I heard recently somebody say that mesh grids are the middle child of mini grids and solar home systems, and so I think that's actually a good way to look at it, that basically, you know you need to make sure that all of your kids are doing a good job and serving the family appropriately, and I think where mini grids really thrive is dense areas.
00:09:35.379 -->
00:09:38.549We're talking, you know, thousands of households or more than that.
00:09:38.549 -->
00:09:40.184Households are close to each other.
00:09:40.184 -->
00:09:50.394There's big productive loads and it obviously makes sense because you're not using much distribution and you're getting the benefit of high power and that shared kind of asset.
00:09:50.394 -->
00:09:55.731Whenever an individual user needs some energy, they can get it, and they can get it reliably.
00:09:55.731 -->
00:09:57.023That's the benefit of the mini grid.
00:09:57.023 -->
00:10:00.292And then on the other side you've got solar home systems.
00:10:00.292 -->
00:10:09.365If there's an isolated house in the hardest to reach part of the world, a solar home system can still connect that household, and so mesh grids really fit in the middle.
00:10:09.424 -->
00:10:11.407We're trying to take the benefits of both.
00:10:11.807 -->
00:10:25.934We're trying to be reliable, we're trying to meet a varying load and we're also trying to get clusters of households or communities that are more sparsely populated, which actually make up a massive part of the addressable market.
00:10:26.296 -->
00:10:37.283So I think, if we're to use a general rule, I'd say that if there's less than 15 or 10 meters between households and there's like a thousand households or more, a mini-grid makes a lot of sense.
00:10:37.283 -->
00:10:46.865And if we're looking at going from 10 to 15 meters, all the way up to 75 meters of distance between households, mesh grids make a lot of sense.
00:10:46.865 -->
00:10:51.471And then you're looking at SHS making sense for everything that's isolated.
00:10:51.471 -->
00:11:13.668And I guess a final note, maybe Felix has some input here as well but SHS is actually becoming more and more competitive because as panels come down in price, as batteries come down in price and distribution costs isn't coming down in price, you can start making bigger SHS Still not going to get that benefit of, like, the variation of load and the shared asset base, but, yeah, shs.
00:11:13.668 -->
00:11:16.527I think it is very exciting to be thinking about the future there too.
00:11:16.960 -->
00:12:00.491Yeah, and as we are talking about a collaboration between companies working on the different technologies, I would like to refer to a podcast episode that we recorded a few months ago with the title the Mini-Grid and Solar Home System Multi-Technology Approach A Way Towards More Finance for the Sector, with Lynn Wiesonga and Gillian Huart Lynn from Cross Boundary Access and Gillian from NG Energy Access and in that episode we already discussed how the different technologies could potentially be used in a harmonized approach by one company.
00:12:00.491 -->
00:12:18.152And then there are so many synergies by having all the different customers covered and supplied with electricity, all the customers that belong to one village, having them all covered and basically doing that all through one team that knows all three technologies.
00:12:18.152 -->
00:12:20.886Is that the way that you're also working in africa?
00:12:21.269 -->
00:12:21.428well?
00:12:21.448 -->
00:12:33.735For us definitely, I mean, for we are being a solar home system company and a mesh grid company and at the same time an epc company, where you could say we're also serving the upper system segment in the higher kilowatt range, like 40, 50 kilowatts or so.
00:12:33.735 -->
00:12:39.371Right, not exactly an ac mini grid, but definitely larger than what any private household in africa would ever consume.
00:12:39.371 -->
00:12:46.433So I totally agree with this multi-tier approach, you know, and also the of Afi, I think, really hits the nail on the head.
00:12:46.433 -->
00:12:50.171That's basically saying mesh grids are the middle child between SHS and AC mini grids.
00:12:50.171 -->
00:13:06.863I totally echo that statement For us, though I mean when I'm saying we are doing it, I mean actually our developers are doing it but I think it's a convincing job to do, because either the developer is a mini grid company or it's a solar home system company and they operate quite different.
00:13:06.883 -->
00:13:14.409So it's a totally different child to sell an shs with us having an internet connected system and running it in a community, from a regulatory standpoint as well as from, like you know, a community standpoint.
00:13:14.409 -->
00:13:17.947Municipality has to be convinced, exclusivity agreements have to be signed.
00:13:18.048 -->
00:13:55.610It's completely different than just talking to a just one end customer right Now talking about the three technologies solar home system, mesh grids and AC mini gr example, the approach that is being used in Bangladesh, where you have similar types and characteristics of solar home systems that then could be upgraded to a mesh grid, but your companies are not doing that.
00:13:56.091 -->
00:13:56.631Am I correct?
00:13:56.631 -->
00:14:00.196Maybe I can make a start if you can add to that, our systems are very modular.
00:14:00.196 -->
00:14:06.092So there's a sliding mechanism in our mesh grid where you can slide in battery extension boxes and solar home systems of our type.
00:14:06.092 -->
00:14:08.889So there's a very cabler, seamless integration possible.
00:14:08.980 -->
00:14:19.696But in general we have designed our mesh grid systems to be taking any 12 volt or 24 volt standalone system that might be a solar home system and incorporate that in the grid.
00:14:19.696 -->
00:14:24.091So you can even have customers that already have been paying off their standalone system for a long time.
00:14:24.091 -->
00:14:28.171They've bought it in any random hardware shop in town and they installed themselves.
00:14:28.171 -->
00:14:36.145And if the developer wants to, they can basically integrate that existing fleet of shs that they find in the village and integrate that into the mesh grid.
00:14:36.145 -->
00:14:44.504Of course it goes without saying that you wouldn't want to take a pico system that has a 10 watt hour battery capacity and pump that into a grid where it's like hundreds of watts of power.
00:14:44.504 -->
00:14:46.366Right, it must be within proportion.
00:14:46.366 -->
00:14:55.753But I would say if the SHS has an installed capacity of from above 200 watt hours or so, you may consider integrating that unit into the mesh grid.
00:14:55.860 -->
00:15:06.490Yeah, but that means that the small standalone system does not necessarily need to be from SolarWorks to be integrated into your systems.
00:15:07.153 -->
00:15:14.039No, not at all, and in fact, actually, if we're building our mesh grids from scratch, we would not even use our own solar home systems.
00:15:14.039 -->
00:15:21.183We would just source off-the-shelf lead carbon batteries, canadian solar panels, you know, off-the-shelf charge controllers, all of those kind of things.
00:15:21.183 -->
00:15:22.586You can really use anything.
00:15:22.586 -->
00:15:23.807That's the great flexibility.
00:15:23.807 -->
00:15:28.053So I don't think if you're a developer, you want to be dependent on just one manufacturer only.
00:15:28.053 -->
00:15:29.196You want to have that flexibility.
00:15:29.196 -->
00:15:36.869As an EPC company, we also like choosing between Goodwill and Huawei inverters or whatever is on the market, whatever is currently most available and most cost effective to us.
00:15:36.889 -->
00:15:38.091Yeah, nice approach.
00:15:38.091 -->
00:15:40.086Afi, how are you handling this?
00:15:40.368 -->
00:16:20.865Yeah, no-transcript asset into the system to the energy as a service model, where they're going to be purchasing energy from the developer who owns the assets.
00:16:20.865 -->
00:16:35.328So ultimately it just comes down to does the developer have an agreement with this household to repurpose their assets, put it into the mesh grid and then sell that household energy again, which I don't think is not achievable.
00:16:35.328 -->
00:16:46.264However, so far, because there's such small capacity deployed in most of the communities that are being served, like the Pico Solar which Felix was mentioning, there really hasn't been the need for this.
00:16:46.264 -->
00:16:52.749But I can really see it in the future, as the solar home systems start getting cheaper, more and more demand for this kind of thing popping up.
00:16:53.200 -->
00:16:59.813And we can speak for like an anecdote from Zambia, where we installed a system like that, where we initially incorporated existing solar home systems.
00:16:59.833 -->
00:17:02.482From a technical perspective, it has been working greatly.
00:17:02.482 -->
00:17:20.827However, it becomes a social and maybe also customer education issue and at some points where you need to prove to the customer that you're not actually excessively taking energy out of their batteries, you're actually stabilizing their own batteries and therefore increasing battery lifetime, because you're not deep cycling their batteries as much as they would have been if they were still standalone.
00:17:20.827 -->
00:17:34.601So we had cases where we had integrated, I think, five or so existing solar home systems that we found in the village into the mesh grid and out of those five, two wanted to be disconnected at some point, to be honest, because they thought that we would deplete their batteries.
00:17:34.601 -->
00:17:35.784In fact, the country was there.
00:17:35.784 -->
00:17:41.728I was trying to show them the evidence of the data, but you know it comes down to customer education and mutual understanding.
00:17:41.728 -->
00:17:49.127So therefore, I would say that in the majority of the cases, the solar systems that power the mesh grid are owned by the developer that deploys the mesh grid.
00:17:49.700 -->
00:17:54.432Okay, let's talk about business models in more detail later.
00:17:54.432 -->
00:17:56.367I have one more question.
00:17:56.367 -->
00:18:13.296Like we now talked about the transition from standalone systems solar home systems into mesh grids, now is there a similar transition from mesh grids to AC mini grids or from AC mini grids to mesh grids?
00:18:13.296 -->
00:18:25.709Could I, for example, install an AC mini grid and on the edge of that AC distribution network I say, okay, now from here I extend further with a DC mesh grid.
00:18:27.257 -->
00:18:28.259Yeah, taking a step back.
00:18:28.259 -->
00:18:45.219To summarize that first bit, the benefit of the mesh grid is that in sparse areas we can reduce the cost per connection substantially by reducing the distribution cost, and we can provide DC output or AC output as well, which is how we provide the energy service.
00:18:45.219 -->
00:18:52.202So really the end consumer is getting the same thing that they would get from accessing energy from a mini grid.
00:18:52.202 -->
00:18:53.905Where are the limitations?
00:18:53.905 -->
00:19:07.282So with Okra, right now we have 1.2 kilowatt AC output limitation from the standard box and then you can scale that up to 3.6 by going parallel but beyond 3.6.
00:19:07.282 -->
00:19:09.577Both single phase, right, yeah, single phase AC.
00:19:09.577 -->
00:19:19.565And then after that you've basically got a limitation with the Ocropod where, like, it's not making sense to add so much power electronics just to get higher power.
00:19:19.684 -->
00:19:27.739But 99.9% of all the energy use cases in last mile communities is covered by 3.6.
00:19:27.739 -->
00:19:34.259But if you look at the higher consumption than 3.6, let's say you've got a 10, 20 kilowatt mill.
00:19:34.259 -->
00:19:46.957This mill is going to generate a lot of income for the community, which is great, and it's also going to generate a lot of income for the energy developer, which they care about.
00:19:46.957 -->
00:19:52.895So our products aren't really scaling up to that point and I think that's where there's a really good fit for a centralized AC mini-grid to power that productive hub.
00:19:52.895 -->
00:20:00.128And that productive hub can be connected with an AC to DC converter into the mesh grid.
00:20:00.128 -->
00:20:02.404So the excess power from the mini-grid is actually supplied into the mesh grid.
00:20:02.404 -->
00:20:08.419So the excess power from the mini grid is actually supplied into the mesh grid, which is lower cost distribution networks supplying energy to all the households.
00:20:08.419 -->
00:20:12.218That's actually something that we're working on designing with a couple of our developers.
00:20:12.218 -->
00:20:12.640Right now.
00:20:12.640 -->
00:20:16.296We've got the prototypes out and getting excited for a field deployment.
00:20:16.316 -->
00:20:17.559Sounds good, sounds good.
00:20:17.559 -->
00:20:23.842So now, what is the difference between the Okra solar technology and the SolarWorks technology?
00:20:24.035 -->
00:20:30.409I think there's a philosophical difference in terms of if every system should be equipped with an AC inverter or not.
00:20:30.409 -->
00:20:31.718We believe in that.
00:20:31.718 -->
00:20:40.948You know, the main reason why we do mesh grids over AC mini grids is really the cost, and therefore adding an inverter to every existing mesh grid connection would be adding too much to the cost.
00:20:40.948 -->
00:20:48.535So the majority of our systems within the mini grids or our mesh grids don't have an inverter, especially the inverter is not inbuilt.
00:20:48.535 -->
00:20:54.903Inverters can generally be connected, but it's an external inverter, usually working with victron, and that's totally fine for our purposes.
00:20:54.903 -->
00:20:58.897For the devices that have a significantly higher demand, we are quite open end.
00:20:58.897 -->
00:21:01.763So, like we have no limitation in terms of how many kilowatts can be used.
00:21:01.763 -->
00:21:28.818Again, like as we're working with victron, we have developed a what we call a victron dongle, where the victron is controlled, similar to the mesh grid power output, but we're just controlling the ac output on the end of the inverter and therefore we can directly plug it onto the battery side, and that system then is literally independent of the capacity of the mesh grid box, even though from the customer experience for them it looks like as if they're paying the same thing, so paying for ac output of the inverter or paying for the mesh grid is all the same.
00:21:28.818 -->
00:21:31.426So that's one difference, I think, between us and okra.
00:21:31.776 -->
00:21:40.906Another difference is that, again for trying to drop the cost per connection as low as possible, we do not integrate charge controller into every of our mesh grid box.
00:21:40.906 -->
00:21:49.974In fact we're working with external, off-the-shelf charge controllers because for us usually only every 10th connection or so gets a solar system.
00:21:49.974 -->
00:21:57.980So we only have the charge controller in the house where that PV capacity is needed and where the charge controller ultimately is needed.
00:21:57.980 -->
00:22:07.604But in 90% of the cases where the charge controller that would be inbuilt would be idle and being unused, we just don't put it into the mesh grid box in the first place.
00:22:07.604 -->
00:22:25.063I think those are the major differences in terms of like feed-in capacity on a dc side from or into the grid, we are between 250 watts and 500 watts that can be either extracted from the battery and pushed into the grid or taken from the grid and made available to the customer.
00:22:25.063 -->
00:22:31.682On the low side, as we call it, which is usually 12 volt power output, I see listening to you, felix.
00:22:32.064 -->
00:22:56.230Would you agree to the following summary that the solar works system is probably more modular and customized and that in the okra solar system you try to drive down cost through having a large number like economies of scale in equal parts?
00:22:56.230 -->
00:22:58.075Is that a correct summary?
00:22:58.395 -->
00:23:05.635well, afi and felix, well I think that both systems are still very flexible in components that can be used.
00:23:05.635 -->
00:23:08.624I would disagree to this part of the explanation.
00:23:08.624 -->
00:23:11.538Both of us we are aiming for economies of scale.
00:23:11.538 -->
00:23:17.039I think the reason why some of the differences are there are because of the markets that Okra and us are working in.
00:23:17.039 -->
00:23:19.836We are operating mesh grids at the moment in Cameroon and Zambia.
00:23:19.836 -->
00:23:26.749We don't really have a need for that many inverters, while Okra might have that need because operating to a large extent in Nigeria.
00:23:26.749 -->
00:23:28.657But I think Afi can speak more to that.
00:23:29.200 -->
00:23:54.696I think the major difference it all stems from, philosophically, what Felix mentioned, that he doesn't believe, as of now, that every house needs to have AC power output, whereas we believe that every house should be equipped with AC output because that's what they're used to using and that's going to enable them to use more appliances, which is going to boost income and then sustainability of the projects.
00:23:54.696 -->
00:24:00.218I think it really stems from that and that actually feeds into a couple of design decisions.
00:24:00.218 -->
00:24:10.036And so where Felix said they have one box per 10 households to really get the cost down low as much as possible makes a lot of sense, but that's a trade-off.
00:24:10.236 -->
00:24:24.797We typically have one box per every four to six households and that means sorry when you, when I jump in there, but when you say box, you mean the system with a solar system connected Exactly yeah, like a hub or a pressure, you have the boxes.
00:24:24.797 -->
00:24:26.922The boxes are in every household, just to clarify.
00:24:26.922 -->
00:24:30.067But, like you know, you have one solar system every four or so.
00:24:30.194 -->
00:24:33.726Yeah so imagine like where the battery and the PV is installed.
00:24:33.726 -->
00:24:45.691In the SolarWorks system it's roughly, let's say, one per every 10, just for illustrative purposes, and with the Okra mesh grid it's one for every four to six households, and so obviously there's a trade-off there.
00:24:45.691 -->
00:25:04.805So SolarWorks is trying to push down on the costs, and then what we're trying to do while always pushing down on costs, but to make sure that there's slightly shorter distances, which means you can get more power availability, which means you can run inverters without tripping the system, which means you get AC power output, means you can run inverters without tripping the system, which means you get ac power output.
00:25:04.805 -->
00:25:06.131So those are some of the kind of trade-offs between the two techs.
00:25:06.151 -->
00:25:13.641Well, there's also there's another reason, apart from cost, that I believe that it I mean like it makes sense, right to provide a lot of capacity to an end customer.
00:25:13.701 -->
00:25:25.324Obviously, and that's why I think, in certain degrees ac mini grids from the total power output can under certain circumstances be superior to a mesh grid, but in many of the villages the energy demand is simply not that high.
00:25:25.324 -->
00:25:40.174So you can achieve the same quality of service with just a dc connection and no one, like even in our mesh grids, prevents the end customer from buying an inverter and then powering those less efficient ac appliances and ultimately paying more for the electricity bill plus the upfront cost of the carpex of the inverter.
00:25:40.174 -->
00:25:52.326Right, there's pros and cons, but I think if I'm categorically equipping an mesh grid connection with an inverter, then I might just use a bunch of micro inverters in every household and just run them all in parallel mode.
00:25:52.326 -->
00:26:02.063I mean, I know that many of the micro inverters as they are in the market don't provide the same degree of regulation that your mesh grid and our mesh grid has yeah, in terms of droop control regulation, all that.
00:26:02.262 -->
00:26:12.116Yeah, but then you've got like a 230 volt distribution line, so it's no longer safe and it's adhering to safe extra low voltage, which is a bunch of the stuff that you're talking about at the start, right?
00:26:12.377 -->
00:26:12.878Exactly so.
00:26:12.878 -->
00:26:15.384The argument of safety indeed is gone.
00:26:15.384 -->
00:26:20.746You'd have to introduce a lot more safety features, mcbs, residual current arresters and all that kind of stuff.
00:26:20.746 -->
00:26:29.501I agree, but at the same time you still have that big benefit of decentralization.
00:26:29.501 -->
00:26:32.333So you can still have that argument of dropping costs because you can cluster systems as opposed to we have a centralized location.
00:26:32.333 -->
00:26:46.319So I think there would be very good reasons to look into and compare and cost in like would it make sense to design a sort of like micro inverter, as we know it, from the pv panels, like you know, those balcony systems and microinverters that you can find, always connecting those two panels within a larger system.
00:26:46.319 -->
00:26:51.660You can retrofit or use them and then make something like a mesh grid, but on an AC basis.
00:26:51.660 -->
00:26:57.855But as I said, like, I really believe in that DC because again, it drops cost, it's decentralized, with all the benefits given.
00:26:58.056 -->
00:27:46.519But the moment I'm moving AC on every household, I really would consider personally if I retrofit 1.2, 1.3 kilometers low voltage line we are setting up another node with another AC inverter and PV panel and a battery to support the voltage at that node at that point, and then extend the AC network by another 500, 600 meters to be able to then cover a total length of 1.8 kilometers just instead of 1.2.
00:27:46.519 -->
00:27:56.142This is basically a mesh AC grid, so to say, but it's in an early pilot stage and we will know more by Q2 next year.
00:27:56.694 -->
00:28:08.342Do you see that the developer would be better off designing with DC or AC, because you want to grow the load growth and you think this is what's going to unlock that?
00:28:08.442 -->
00:28:09.124Good question.
00:28:09.124 -->
00:28:12.199Indeed, we haven't done the calculation at that point.
00:28:12.199 -->
00:28:30.964As you know, we're coming from the AC world and the idea was well, usually you install medium voltage networks on top of your low voltage network, which is extremely expensive to extend your AC network, and then you have transformers and everything.
00:28:30.964 -->
00:28:41.309And the idea was maybe we can bring down costs by having a decentralized generation within that AC network and support the voltage at a certain point.
00:28:41.309 -->
00:28:46.557Ac network and support the voltage at a certain point.
00:28:46.557 -->
00:28:58.761But of course, deploying that point in the network to install an AC to DC converter and then from there go on with a 60 volt DC distribution would probably also make a lot of sense and may even be more efficient.
00:28:58.761 -->
00:29:06.681Beyond these 1.3 kilometers you would anyway not connect any heavy loads to the AC network anymore.
00:29:07.241 -->
00:29:09.867And therefore, a DC network could do the job as well.
00:29:10.736 -->
00:29:23.845And then our philosophy would be that all of those customers, just because they're more than a kilometer away from the powerhouse, doesn't mean that they now only want DC, as opposed to everybody else who still wants AC output.
00:29:23.845 -->
00:29:26.101That's why it should be configured with AC at the output.
00:29:26.101 -->
00:29:31.768But these are philosophical things that I guess we will be continuing to discuss and iron out.
00:29:31.768 -->
00:29:50.037Another question that I have for you, nico, while I got you on the spot, is do you have a rough gauge for distribution cost per connection on the AC mini grids If you're looking at like within a kilometer from the powerhouse, and then maybe the costs if you're looking at beyond a kilometer from the powerhouse?
00:29:50.277 -->
00:30:04.670Yeah, like a kilometer of AC network could be anywhere between, say, $8,000 and $20,000, with an average probably around $14,000 or so Okay.
00:30:04.670 -->
00:30:07.625Depending, of course, on the national standards.
00:30:07.625 -->
00:30:09.121Do you need steel poles?
00:30:09.121 -->
00:30:11.042Can you install wooden poles?
00:30:11.042 -->
00:30:15.066What is the voltage drop that is acceptable?
00:30:15.066 -->
00:30:20.086And well then, you design your system and the capex results accordingly.
00:30:20.487 -->
00:30:20.788Gotcha.
00:30:20.788 -->
00:30:23.964So that's like $140 per connection roughly.
00:30:24.336 -->
00:30:32.747Well, yeah, there are different densities of population, of course, and then, after all, you can connect more or less customers to this one kilometer line.
00:30:33.335 -->
00:30:33.957So 140,.
00:30:33.957 -->
00:30:38.488So if I get that right, if that's like 10 meters of distance between households.
00:30:39.297 -->
00:30:40.521Which is very dense already.
00:30:40.815 -->
00:30:43.321Which would be very dense between households, which is very dense already, which would be very dense.
00:30:43.321 -->
00:31:00.144And then, for context, on the mesh grid side, you're looking at about 30 meters average distance and you're looking at about $40 to $60 on the distribution side, and I think we can get lower than that as well if we're using, like, aluminum conductors.
00:31:00.144 -->
00:31:02.522Is that aligned with what you're looking at, felix?
00:31:02.855 -->
00:31:05.083Well, if it's just the cost of distribution.
00:31:05.143 -->
00:31:13.125Yes, and I'm surprised, nick, you said like that you consider, like you know, wooden poles cheaper than metal poles, because I think the opposite might be the case, depending on the height of the poles.
00:31:13.434 -->
00:31:19.704But if we would be forced to use CCA treated wooden poles those are like nine, 11 meter high gum tree poles.
00:31:19.704 -->
00:31:23.857They come at around 80 to 100, depending on which country you're operating in, per pole.
00:31:23.857 -->
00:31:33.442So that means that I'm still having the exact same cost in terms of distribution line, maybe saving slightly maybe on the cables because they can be a bit thinner than the ac mini grids.
00:31:33.442 -->
00:31:42.847But I wouldn't say there is so much of a cost saving, especially on the poles, unless and that is where I think the great benefits can come in is I can break slightly with the regulation and say, okay, look, it's DC.
00:31:42.847 -->
00:32:01.029Even in the worst case, if I have like 60 volt touching the pole and the pole is electric, since no one can get hurt from that, the safety regulations, you could allow me maybe to use smaller, thinner, lower poles and then, yes, I would agree with you that the cost for, like you know, 30 meters or so could drop to like maybe 60 or so.
00:32:01.069 -->
00:32:26.526Yes, when I talk about steel poles, I usually then consider the standard poles that are used in the countries, and those usually come at around 200 or so per piece, cross-section of maybe 20 centimeters or 15 or so, like really heavy stuff that you bring into rural areas and install okay, the steel poles I was referring to come, like you know, off the factory for about 25 to 30 dollars.
00:32:26.566 -->
00:32:29.699They're like six meter long galvanized steel poles or something like that.
00:32:29.699 -->
00:32:34.721It's really just relatively thin kind of things because the slack and more like a pipe that you put into the ground.
00:32:35.021 -->
00:32:35.743It's really a pipe.
00:32:35.743 -->
00:32:40.537It's really yeah, yes, exactly okay, good, now let's go into the business models.
00:32:40.537 -->
00:32:42.778You already started talking about this.
00:32:42.778 -->
00:32:48.820So both of you consider yourselves as the manufacturers of the technology.
00:32:48.820 -->
00:33:08.833So you're not deploying and operating the technology yourself, if I understand correctly, but you're working with operators that then purchase the technology from you and operate the technology and integrate the technology with mobile money systems and these kind of things to then get their revenue.
00:33:08.833 -->
00:33:09.313Is that correct?
00:33:09.313 -->
00:33:10.192Yeah, that's largely correct.
00:33:10.434 -->
00:33:24.815For us in particular, especially on the software side, we provide mostly tools in terms of an Android app and a front end that mostly deals with the technology and analytics and operation of it, not so much on the collection of payments with the technology and analytics and operation of it, not so much on the collection of payments.
00:33:24.815 -->
00:33:28.190So we are integrated with software as a service companies in a sector like Agaza, paygops, payg, opia and so on.
00:33:28.190 -->
00:33:33.585Our philosophy we don't want to necessarily bind the developers fully on our software.
00:33:33.585 -->
00:33:39.804However, of course, there is a certain dependency by our developers to us because they purchase our mesh grid technology.
00:33:39.804 -->
00:33:43.682But principally yes, the explanation is correct we're working on a B2B basis.
00:33:43.682 -->
00:33:44.746We manufacture hardware.
00:33:45.507 -->
00:33:52.880Our core business really is also making one-time money with selling that hardware and then potentially there could be a service contract for maintenance as well.
00:33:52.954 -->
00:34:21.802But it's largely focused on selling hardware to developers and, of course, for us it becomes a business model only really if we're not only selling our mesh grid boxes, which of course are the core of a mesh grid, but also all the components that accompany the mesh grid, which are poles and cables and buckles and piercing connectors and whatnot to make a mesh grid happen, yeah, but so from the developer side, I would say the business model is very much that of a, I would say, regular ac, mini grid developer escrow company or desco company.
00:34:22.443 -->
00:34:26.762So they are selling energy as a service to end customers.
00:34:26.762 -->
00:34:28.730That can be done on an upfront basis.
00:34:28.730 -->
00:34:36.661There's different models of how you actually sell the kilowatt hours to the end customer, but ultimately you're selling energy on a regular basis to the end customer.
00:34:36.661 -->
00:34:50.978And one major difference to the paygo model is that if a customer doesn't pay, then they don't pay, but they're not really defaulting, while if I'm selling a solar home system and they're not selling for, let's say, one month or two months and they've exceeded their grace period, it's considered a default and the system needs to be picked up.
00:34:50.978 -->
00:35:05.505That's not really the case for a mini grid, because you'd rather dealing with the community as a whole and therefore, even if you have a certain percentage of customers not regularly paying, it's about the average payment rate of the entire community, not so much of the individual right.
00:35:06.126 -->
00:35:19.025But ultimately every mini grid company has their own little nitty-gritty models in place the developers you're talking about, that you're currently working with, are they coming from the solar home system sector mainly?
00:35:19.507 -->
00:35:21.342our developers not so much.
00:35:21.342 -->
00:35:31.717Both of them have had previous experience with mini grids and one of them also has has had some sales in standalone systems, but honestly, they do a bit of everything.
00:35:31.717 -->
00:35:42.101I have the feeling that those developers that really are reaching some degree of profitability in the off-grid sector which is hard enough to achieve really do everything as as as the demand comes right.
00:35:42.101 -->
00:35:44.498So there's a subsidy for mini grids, they do mini grids.
00:35:44.498 -->
00:35:47.512If there is a tender for salome system, they do that right.
00:35:47.512 -->
00:35:53.295But without, without certain degree of subsidy very few of those off-grid businesses are generally profitable, I see.
00:35:53.775 -->
00:35:55.438Is that the same on your end, afi?
00:35:56.018 -->
00:36:12.835um, yeah, I mean there's a few kind of tweaks, but at the base of it, we are also a B2B technology supplier, and so I think maybe it's worth digging into some of the why and also giving a bit of kudos to the developers.
00:36:12.835 -->
00:36:15.701They have a very, very, very tough job.
00:36:15.701 -->
00:36:29.737They have to access the hardest to reach communities in the world under very challenging kind of political or security conditions and then bring energy access to them and do it in a profitable way so that they can keep scaling.
00:36:29.737 -->
00:37:02.101Just to give you an anecdote, one of our partners in Haiti right now, alina Energy, doing a great job, have energized, I think, more than 15,000 people there now with mesh grids and near one of the mesh grids, just yesterday there was a news report that 70 people were murdered by gangs within 20 kilometers of a mesh grid, and so, yeah, developers are operating in hard to access areas, and the kind of mission that we have is to scale energy access globally and scale mesh grids globally.
00:37:02.641 -->
00:37:05.585The local developers are going to do a much better job than we can.
00:37:05.585 -->
00:37:16.065They know how to manage these operations, they know how to engage with the community and do it well, and so we're trying to play to our strengths and what we're trying to do is provide the best technology possible.
00:37:16.065 -->
00:37:26.538So not only on the hardware side in terms of providing low-cost energy availability and modularity, but also on the hardware side in terms of providing low cost energy availability and modularity, but also on the software side, analogous to Felix.
00:37:26.538 -->
00:37:30.467But maybe some of the things that we have specifically focused on.
00:37:30.467 -->
00:37:33.925One is to make sure that all of the billing data can come through.
00:37:33.925 -->
00:37:50.804There's insights available for the developers to see, like where are unutilized assets, where can they boost more productive loads, which customers are generating sustainable revenues, so that they can keep focusing on those areas to make sure that their portfolios are ultimately profitable, so they can keep scaling.
00:37:50.804 -->
00:37:57.648So, yeah, I think the software part is also a pretty important aspect of making sure that operations can scale.
00:37:58.775 -->
00:38:13.891And after all, with mesh grid technology, these developers and operators now have a new, very cost-effective technology at hand that may help them become more profitable earlier.
00:38:13.891 -->
00:38:19.288Let's talk about average revenue per user that you have seen in your systems.
00:38:19.288 -->
00:38:24.317I guess they are very similar to what we know from AC systems, aren't they?
00:38:24.516 -->
00:38:30.358I keep the information high level because I don't want to give away data from any of our partners.
00:38:30.358 -->
00:38:47.068But interestingly, with the mesh grids that have been deployed, we can see from 50 cents APU, which is definitely not going to be a viable project under any circumstances, not even going to cover the cost of operations, all the way to high teens.
00:38:47.068 -->
00:38:55.025So like $16, $17 ARPU, which I think would be viable for even higher CapEx solutions.
00:38:55.025 -->
00:38:57.818So there's the entire range and what's the lesson there?
00:38:57.818 -->
00:39:09.195The lesson there is that site selection and community engagement and activation through productive use to make sure that the assets that you deploy there are actually utilized is incredibly important.
00:39:09.195 -->
00:39:22.909And in terms of how we look at it, let's model for developers who are going to purchase the equipment in USD, but they should be accessing local currency debt financing.
00:39:22.909 -->
00:39:26.971Otherwise things are going to balloon out if the currency depreciates.
00:39:27.440 -->
00:39:32.833Let's look at the average operations cost per year of about $15 to $20.
00:39:32.833 -->
00:39:37.393And let's look at a realistic ARPU of $5 to $10.
00:39:37.393 -->
00:39:42.769That doesn't change in local currency, Like it's in local currency and it stays there from today's date.
00:39:42.769 -->
00:39:53.751And then we found with our technology which we're able to service this ARPU of $5 to $10 at, you know, roughly $400 per connection delivered.
00:39:53.751 -->
00:39:59.813This can actually lead to a sustainable seven-year payback period without subsidies.
00:39:59.813 -->
00:40:07.487So that's no subsidies and that's really what we're focused on, because we have to design this so that it can keep scaling after the subsidies are gone.
00:40:07.487 -->
00:40:14.420And obviously right now we've got the $300 per connection mesh grid subsidy that's been announced in the DAS program in Nigeria.
00:40:14.420 -->
00:40:20.108That's massive and that's going to make all these projects even more viable and more attractive.
00:40:20.108 -->
00:40:31.463And then, when that subsidy is gone, I think we need to keep focused on making sure that we can keep hitting these ARPUs and keep the cost optimizing further so the market can scale.
00:40:31.764 -->
00:40:32.826Yeah, thank you Afi.
00:40:32.826 -->
00:40:39.190That indeed makes a lot of sense that the ARPUs are very similar to what you see in AC mini grids.
00:40:39.190 -->
00:40:49.715It's the same type of customer, after all, in the very similar community, just a different technology, basically providing very much comparable services.
00:40:49.715 -->
00:40:53.447So therefore expecting the same ARPUs is very reasonable.
00:40:53.447 -->
00:41:07.304But if the mesh grid technology comes at a lower cost, then of course it's very attractive for mini grid companies in Africa to look into this technology and maybe pick this up, as already indicated.
00:41:07.304 -->
00:41:14.364So now we said this technology is cheaper compared to AC mini grids and we've already looked at the distribution network costs.
00:41:14.364 -->
00:41:19.695So now what is an average cost per connection in a mesh grid?
00:41:20.681 -->
00:41:37.726It's a very varying kind of number depending on the density of the community, quite frankly, because obviously transmission line poles and cables and the effort of erecting that site is one of the main drivers, not so much the capex of the actual mesh grid controller and the batteries and the panels.
00:41:44.199 -->
00:41:48.275In comparison, right, we have about 35 to 40% contribution of cost on carpex for the distribution line and all the rest is all the other stuff.
00:41:48.275 -->
00:41:57.291So to bring it down to an actual number in US dollars, we experienced it could be anywhere from as low as $150 in a very special setup.
00:41:57.291 -->
00:42:12.271You know there's a lot of off-grid markets across african communities that are entirely off-grid, obviously, and so you can run the cables in the ceiling board and just like have the service line, literally like one meter of service line, and put the mesh grid box just right there.
00:42:12.271 -->
00:42:14.360So there's no poles and nothing of that required.
00:42:14.360 -->
00:42:16.226So that is the lowest entry point cost.
00:42:16.226 -->
00:42:22.432It's about 150 dollars, for us at least, and then it goes as high as maybe 300, 350 dollars.
00:42:22.432 -->
00:42:35.327If it's a very low density village and with low density I mean like one, one to two customers on a transmission line of 30 meters that would be considered a low density village afi, do you want to add to that?
00:42:37.090 -->
00:42:40.635Yeah, I'm just looking at some numbers right now.
00:42:40.635 -->
00:42:47.311I think getting at the $300 per connection and even below, that's quite impressive, felix.
00:42:47.311 -->
00:42:53.115On our side, the way we look at it is that it's not just a short-term thing.
00:42:53.115 -->
00:43:00.731It has to be a long-term thing in terms of commercial viability and in terms of providing value to the community, and I'm sure you also do this, felix.
00:43:00.731 -->
00:43:03.922I'm just saying that we design the systems.
00:43:03.922 -->
00:43:16.465For example, the DAS program has come out and they've said that the minimum consumption is 83 watt hours per consumer per day as the average load to be eligible for the milestone two of the subsidy.
00:43:16.465 -->
00:43:20.184So the DES program says 83 watt hours per consumer per day.
00:43:20.503 -->
00:43:36.634We design our smallest systems to be about three times that capacity because we know that the subsidy is not always going to be here and you need to sell enough energy to generate enough revenue to have these projects be viable in the long term.
00:43:36.634 -->
00:43:52.512So our smallest designs are coming around $350 to $400 with 30 meters of average distance between connections and I think that is a very compelling cost point because it's covered mostly by the subsidy.
00:43:52.512 -->
00:43:59.393But at the end of the day, if the subsidy is not there, there's enough consumption there for it to continue generating that revenue.
00:43:59.393 -->
00:44:10.487And of course our systems, both SolarWorks and OCRES, are also modular, so that when you identify the pockets of productive use and increased demand, you can scale things up there over time as well.
00:44:10.487 -->
00:44:15.311So yeah, I think we can get the cost down, but it comes at a cost to reduce the cost.
00:44:15.311 -->
00:44:18.489So I would say let's keep it there.
00:44:19.119 -->
00:44:20.105Well, afi, one question.
00:44:20.105 -->
00:44:22.288There is the inverter included in that package.
00:44:22.380 -->
00:44:28.411Yeah, that's all included inverters, batteries, distribution screws as well.
00:44:28.411 -->
00:44:33.112Ddp past customs to the developer in Nigeria.
00:44:33.112 -->
00:44:33.251Yeah.
00:44:33.724 -->
00:44:36.847I think that would explain also why there's a slight cost difference here.
00:44:36.847 -->
00:44:43.534Right, like why I'm saying like we're stopping $3 to $350 per connection is because we don't consider the inverter part of the standard setup.
00:44:43.534 -->
00:44:49.947It can be additionally purchased, but I would add the cost and I would consider the inverter as an appliance, and that's another point I wanted to make.
00:44:49.947 -->
00:44:55.813So, in order to create the offtake and create the demand, we would want to sell appliances.
00:44:55.813 -->
00:45:00.469So the developer should be selling appliances to the clients, because only that really they generate enough revenue.
00:45:00.469 -->
00:45:10.740Because, like going back to the question of how much money you can generate or you should generate from a client, I think if you're making less than $5 per client per month, then is no way that you're ever going to amortize the equipment.
00:45:10.740 -->
00:45:15.612So there's an absolute necessity for grants if you don't have that $5 per customer.
00:45:15.612 -->
00:45:20.619And then the additional revenue you have to generate really from selling appliances to the same customer base.
00:45:21.000 -->
00:45:42.507Yeah, and it was a little bit surprising also to me, felix, when you said, oh, we all can go below $300, because Afi just said, when Inensos calculated the required grant level for the Nigeria dares, and we came up with this $300 per connection, we actually ran financial models, I think, with data from both of your companies.
00:45:42.507 -->
00:45:56.072This is how we basically came up with a number and, after all, there must be some skin in the game, some private investment from the operators, from the developers, and this is how the $300 were set.
00:45:56.552 -->
00:45:58.706Yeah, but again, I can tell you where this comes from.
00:45:58.706 -->
00:46:05.519So the benchmarks that we have provided, indeed, are higher because the program in Nigeria requires inverters to be in place, indeed, in every household.
00:46:05.519 -->
00:46:07.086So that drives the cost.
00:46:07.086 -->
00:46:16.231And secondly, the edge case that I gave you of the $150 is really the lowest and it might be quite an alien scenario for 90% of the target communities.
00:46:16.231 -->
00:46:19.630Right, I'm talking about markets that are wall to wall.
00:46:19.630 -->
00:46:22.077The next customer is zero meters away.
00:46:22.137 -->
00:46:24.288Basically, yeah, which is a nice scenario to have.
00:46:24.288 -->
00:46:26.585Well, but it's not a very uncommon scenario.
00:46:26.585 -->
00:46:29.344Let's put it this way there's tons of those markets in.
00:46:29.364 -->
00:46:33.324Nigeria that you can totally electrify like that and I'm sure that you know Afi.
00:46:33.324 -->
00:46:44.208I mean, maybe you have not run a market like this, but I'm sure the numbers, even for your system, might drop below 300 in this particular case.
00:46:44.208 -->
00:46:46.318But of course I really want to make it very clear this is not what you can expect in an average scenario.
00:46:46.318 -->
00:46:48.603Average for us would be about 250 to 300.
00:46:48.603 -->
00:46:50.646That's really also our cost all right.
00:46:51.288 -->
00:47:00.432Where do you see bottlenecks for a rollout of mesh grids in africa, in the markets you're in, and how could these be overcome?
00:47:02.141 -->
00:47:02.501Yeah.
00:47:02.501 -->
00:47:11.704So I'm always trying to identify bottlenecks, and you have to do things thoroughly when there's a bottleneck and get it out of the way once and for all.
00:47:11.704 -->
00:47:22.010So I think some of the obvious bottlenecks have been that mesh grids have had like a disadvantage from not being kind of known in the industry.
00:47:22.010 -->
00:48:01.106But I think over the last couple of years, with a bunch of developers in Nigeria picking it up, including Engie, taking a risk to understand what's going on with this technology, along with developers scaling in really challenging scenarios, like in Haiti, getting more than 15,000 people connected where there's no power even for the telcos to provide coverage there, but people are getting access to 99% energy availability, I think we've managed to unblock one of the biggest bottlenecks, which is basically understanding from the industry at a high level, buy-in from some key stakeholders such as the DAS program, and so now I'm looking at, okay, what are the next bottlenecks that are coming up?
00:48:01.106 -->
00:48:22.751And two more bottlenecks One that we're in the process of kind of navigating right now, which is when mini grids and mesh grids are rolled out because they're under the energy as a service aspect and because traditionally they're seen as infrastructure as opposed to solar home systems, which is just like set it and then forget it.
00:48:22.751 -->
00:48:31.983Well, for lack of a better term, it's just like set it and forget it and go to the next place, find the next customer, cherry pick whoever you can, whereas mini grids and mesh grids are more big, community-based solution.
00:48:32.443 -->
00:48:59.322So one of the bottlenecks that I had identified was, well, with mini grids as a complex kind of commissioning and verification aspect of it that requires a bunch of the regulators to come and sign off on everything and this is really adding so much time, as far as I understand, to the mini grid development process, whereas when you look at the mesh grid technology, you're like, hey, it doesn't have high voltage, it doesn't have land acquisition, displacement of people.
00:48:59.322 -->
00:49:01.389Like, do we need to go through this same process?
00:49:01.389 -->
00:49:12.210So that's one of the things that I'm actually discussing with stakeholders such as Rural Electrification Agency and NEMSA in Nigeria to ensure we don't have that bottleneck.
00:49:12.210 -->
00:49:17.570So when there are sites and they're suitable, we know it's low voltage, we know it's easy to deploy Just quickly get it done.
00:49:17.570 -->
00:49:20.838So that's one of the things I'm actively working on right now.
00:49:20.878 -->
00:49:47.291I'm pretty optimistic that we'll be able to have a lightweight process and then another final bottleneck that I think the industry is going to face and this is interesting, like if you look at SHS companies and like why so many people are into it, you can scale it quick Like companies doing hundreds of thousands of systems a year, whereas energy as a service or mini grid companies, they're doing thousands or tens of thousands, if at best, a year.
00:49:47.291 -->
00:49:55.887So for us to scale this industry, how can we have the developers rolling out in the hundreds of thousands or ultimately millions a year?
00:49:55.887 -->
00:50:07.646And I think it's going to come down to the developers really figuring out how to scale their distribution and their community engagement and their after sales support teams.
00:50:07.646 -->
00:50:20.865And honestly, I think this is going to be tough and the industry needs to figure out how to do that, because when we do, then we're going to see these infrastructure solutions rolled out at scale, so doing a bunch of stuff to try to help out there.
00:50:20.865 -->
00:50:23.733But I'll stop there and pass it back to you guys.
00:50:23.900 -->
00:50:30.641Yeah, but, afi, I guess you would not be able to escape tariff regulation after all with mesh grids.
00:50:30.641 -->
00:50:36.572But maybe we need a mesh grid specific approach to tariff regulation.
00:50:36.572 -->
00:50:37.014Let's see.
00:50:37.014 -->
00:50:40.402This would be a subject for a next episode, maybe.
00:50:41.164 -->
00:50:47.887New technology, new approaches yeah, actually, I mean to the tariffs, and we've been facing challenges there, with particular for uganda, right?
00:50:47.887 -->
00:50:59.641So there's this tariff saying that mini grids and under which mesh grids would also fall is limited to 25 us dollar cents per kilowatt hour, and that is definitely insufficient to refund any mini grid or mesh grid.
00:50:59.641 -->
00:51:02.068So again, then there comes in the need for subsidy.
00:51:02.068 -->
00:51:06.706I would largely also echo the entry barriers and challenges that Afi was mentioning.
00:51:06.706 -->
00:51:16.945I also wanted to thank you, afi, by the way that you know you've been driving this as a promoter of mesh grids, especially in Nigeria, and bring this now into World Bank funded program of a significant scale.
00:51:16.945 -->
00:51:18.588So that's been a job well done.
00:51:18.628 -->
00:51:22.534Thanks, thanks to the whole ecosystem, thanks to Enansis as well.
00:51:22.534 -->
00:51:26.528Yeah, indeed, yeah, yeah, it's a forward looking ecosystem, I think.
00:51:26.548 -->
00:51:26.809Yeah.
00:51:26.809 -->
00:51:39.134Well, there's one anecdote, though I can say about Zambia surprisingly, is that when we started our first mesh grid there, we like thought, okay, like look, so we played this just under the radar and sell it as a standalone system, and we actually then played it open book.
00:51:39.134 -->
00:51:48.543We went to the Energy Regulation Board and the Ministry of Energy and informed them about it, and in fact, they said, yeah, it's no problem, we've even regulated it, and the regulation says it's unregulated, basically.
00:51:48.543 -->
00:52:23.143So in Zambia, there is a good blueprint of a regulation saying that if you have a grid that's on DC between 45 and 72 volts, the only thing you have to do, and also another limitation is that it has to be below, I think, 50 kilowatts or so, which is most mesh grids are below 50 kilowatts anyways then you just have to inform the Energy Regulation Board of the existence and you can take whatever tariffs you want, though the Energy Regulation Board has the right to intervene in case they find tariffs are unreasonably high, and I think that's a very nice and lax regulation.
00:52:23.164 -->
00:52:25.592You know that if you're overdoing it, they come and challenge you, but if not, if you're behaving, you can do what you want.
00:52:25.592 -->
00:52:30.985Okay, so you were already talking about Cameroon and Uganda and Nigeria and Zambia.
00:52:30.985 -->
00:52:35.442What other markets and what other countries are most interesting to you?
00:52:35.802 -->
00:52:36.425I think there is.
00:52:36.425 -->
00:52:57.023It's unfortunate to say, but there's a growing refugee community, especially in sub-Saharan Africa, and I think, as bad as it sounds, but there is a certain customer base for us as manufacturers to serve refugee settlements because the density is very high, the consumption is quite low, so no one really put an AC mini-grid in those communities and also many of them are very scattered.
00:52:57.023 -->
00:53:11.619If you look into the northwestern areas of Uganda, there's camps called like Bidi Bidi and it's not really like a typical type of refugee settlement, how you would know it like house after house, but it's like clusters of refugee communities mingled with local communities.
00:53:11.619 -->
00:53:16.887So they're really like villages, but small villages with high density, so they are predestined for mesh grids, if you ask me.
00:53:17.309 -->
00:53:19.172Afi which countries are you looking at?
00:53:19.172 -->
00:53:20.382What?
00:53:20.402 -->
00:53:21.445markets.
00:53:21.445 -->
00:53:29.688Yeah, well, I mean, ultimately we need to look at every single market where there's off-grid people that need access to affordable energy.
00:53:29.688 -->
00:53:33.971But right now we have like a couple of criteria that we look at.
00:53:33.971 -->
00:53:43.731So we need to have supportive regulatory framework and we need to have supportive regulatory framework and we need to have developers who are our partners, who can actually scale there and access those communities.
00:53:43.831 -->
00:53:57.467So, yeah, for example, in Nigeria, where we have Engie as a partner and a very kind of forward-looking company with the multi-technology framework in mind and also the REA and the DAS program, nigeria makes a lot of sense.
00:53:57.467 -->
00:54:01.467There's also the most off-grid developers globally in the world in Nigeria right now.
00:54:01.467 -->
00:54:03.132So Nigeria makes a lot of sense.
00:54:03.132 -->
00:54:05.585But, as you know, we're also in Haiti.
00:54:05.585 -->
00:54:13.211So whenever somebody is like, hey, I need the tech and I'm going to go make it happen in an area that other technology can't go, we're going to supply them.
00:54:13.211 -->
00:54:29.516There's a bit of a battle internally because we want to be going to so many markets at once, but I also think there's a lot of good that comes out of focusing and doing it really, really well in one country and then letting that be the blueprint for global expansion.
00:54:30.018 -->
00:54:35.326Yeah, Are you setting up maintenance and service structures in each country you're going to?
00:54:35.545 -->
00:54:59.583Yeah, so that's one of the things that kind of slows us down, because our strategy is to ensure that every country that we're operational in, we set up in-country customer success, we set up in-country warehousing, we set up in-country after sales support, and all that is an investment of time and resources, but it makes sure that mesh grids have a good kind of experience and reputation in the market.
00:54:59.583 -->
00:55:01.327So yeah, Good Felix.
00:55:01.467 -->
00:55:02.572anything to add from your side?
00:55:02.572 -->
00:55:03.922How are you handling this?
00:55:04.324 -->
00:55:07.844No, I think, no, I think that's, I think, covered it most Same, thing, same.
00:55:07.844 -->
00:55:08.005Thing.
00:55:08.505 -->
00:55:10.612Now a quick feedback from your sides.
00:55:10.612 -->
00:55:16.231How many systems have you installed already and how many customers have you connected, each of you?
00:55:16.481 -->
00:55:16.702Nice.
00:55:16.702 -->
00:55:34.347So, with mesh grids, we have about 40 000 people energized so far, which is pretty exciting stuff, and half of that, basically, is happening in haiti, where it's really exciting to see these hardest to reach communities get energized, and I think it's the tip of the iceberg.
00:55:34.567 -->
00:55:43.193Things are ramping up a lot more now yeah, for us, I think you know, as we are a blended company between standalone systems and mesh grids, we are significantly smaller than Okra on the mesh grid connection side.
00:55:43.193 -->
00:55:55.594So all in all we are serving about say, 600, 700 individuals as it stands right now, but we are having about 30,000 standalone systems that are equipped with our technology or fully our technology.
00:55:55.594 -->
00:56:03.168That's 120,000 people directly affected by solar work standalone system technology nice, okay, good.
00:56:03.228 -->
00:56:07.601So now what uh is up next for mesh grid technology?
00:56:07.601 -->
00:56:09.485What are you currently developing?
00:56:09.666 -->
00:56:09.867you know.
00:56:09.867 -->
00:56:11.409Firstly, no secrets.
00:56:11.409 -->
00:56:31.105Like felix and I, we speak, we share about tech specs and how we see the industry evolving, and I think we just need to give a shout out to you, nico, and also the industry, like the REAs, the world banks, like the developers, like it's a very collaborative space and we need to be collaborative in order to solve this really challenging problem.
00:56:31.105 -->
00:56:36.041Yeah, so that doting on the industry asides we're working on a couple of things.
00:56:36.041 -->
00:56:38.284So let's start with the software side.
00:56:38.284 -->
00:56:54.211With Dares, we're expecting a massive ramp up of mesh grids in Nigeria and we're finding, like, some of the bottlenecks actually being like internal operations, basically getting all of the sites simulated and having a bill of materials for all the sites.
00:56:54.211 -->
00:56:59.289It might sound a bit boring but we're building an internal product called the Network Planner 3000.
00:56:59.289 -->
00:57:10.762And basically it takes all the geo-coordinates and allows tweaking to optimize the cost to the demand and spit out a site-by-site bill of materials Very useful for the developers.
00:57:10.762 -->
00:57:26.871Another software we have so much data that's coming from the field that we have more than a hundred dashboards showing everything from energy consumption by user type, uptime in community versus non-payment rate, looking at potential demand for productive appliances.
00:57:26.871 -->
00:57:32.266But having more than 100 dashboards is not really that practical for any of our partners.
00:57:32.266 -->
00:57:45.570So we're currently investigating some AI-enabled smart solutions that can basically feed the right insights to the right stakeholders so that they can actually action that data to keep their portfolios operating effectively.
00:57:46.239 -->
00:57:50.494And I might just quickly drop one update on the hardware side.
00:57:50.494 -->
00:58:02.652But we had some challenges with communications because the grid, basically the telco towers, were not getting enough energy from the diesel so that they weren't providing connectivity to our smart devices in the field.
00:58:02.652 -->
00:58:21.563So because we have a bunch of data that's coming through, we've actually upgraded from being all 2G based communication on every device to now being Ysun based communication, and Ysun basically allows high data throughput but it's unlicensed, so we don't have to pay that fee.
00:58:21.563 -->
00:58:26.880And then it hops that data to a point where there is 2G connection and we send it out from there.
00:58:26.880 -->
00:58:35.329So it's really reducing the cost of getting the data out and sending remote firmware updates and controlling these systems An exciting under the hood update, yeah.
00:58:35.570 -->
00:58:40.050Yeah, I think from our side we also consistently develop the software in particular further.
00:58:40.050 -->
00:58:45.585We are quite intensely working on our Android app, because we also do have a front end for Windows and all those kind of things.
00:58:45.585 -->
00:58:54.188But we believe that especially when you're a developer in the field, you would like to have that flexibility and ease of accessing your data directly from an Android app.
00:58:54.188 -->
00:59:00.731So currently we're consolidating most of the stuff that we have initially programmed in our grafana front end and move that over into the android app.
00:59:00.731 -->
00:59:01.862That's one of the things.
00:59:01.862 -->
00:59:04.934We're also constantly improving the hardware further.
00:59:04.934 -->
00:59:13.451So we're currently developing stuff like line breakers so that we can isolate folds on the line so that not the entire network goes down if someone shorts the grid or something like that.
00:59:13.451 -->
00:59:16.829And there's also other little little add-ons to the grid.
00:59:16.829 -->
00:59:22.034So we have laura controlled street lighting that sits on every pole of the mesh grid.
00:59:22.034 -->
00:59:24.061So laura, just to say what about laura?
00:59:24.061 -->
00:59:35.661That's like an industry standard that allows low bandwidth, low data communication, usually between iot devices to share stuff, like you know token data, consumption data, technical data, maybe even firmware updates.
00:59:35.661 -->
00:59:37.605We actually run firmware updates via laura.
00:59:37.605 -->
00:59:39.628So those things we are constantly improving.
00:59:39.668 -->
00:59:43.242But you know those little switches that control the street lights.
00:59:43.242 -->
00:59:46.090We do already have them in the field, testing them currently in cameroon.
00:59:46.090 -->
00:59:55.088We also like want to have a device that allows direct consumption from the power line, as opposed to having a converter in between.
00:59:55.088 -->
01:00:00.175Because there's certain productive use appliances if I'm looking at, for instance, axles, it's another company between because there's certain productive use appliances.
01:00:00.175 -->
01:00:00.679If I'm looking at, for instance, axol, it's another company.
01:00:00.679 -->
01:00:05.626They're doing a great productive use equipment, like a milling machine for instance, that could be directly fed from the power line.
01:00:05.626 -->
01:00:11.052There's no need to step down the voltage for those, like you know, stronger DC brushless motors.
01:00:11.052 -->
01:00:13.505So those kinds of things are currently in the pipeline or development.
01:00:14.019 -->
01:00:19.983Interesting, interesting, yeah, and development, interesting, interesting, yeah.
01:00:19.983 -->
01:00:20.365Thanks a lot, felix.
01:00:20.365 -->
01:00:20.746Thanks a lot, afi.
01:00:20.746 -->
01:00:21.469This was highly interesting.
01:00:21.469 -->
01:00:31.356I really liked discussing this innovative new technology with you, and I believe it has a great future also among our listeners, who are mainly AC mini grid operators.
01:00:31.356 -->
01:00:33.561So thanks a lot to both of you.
01:00:33.561 -->
01:00:42.108Before we close, please tell our audience quickly how they can reach you if they're interested to learn more about your technology.
01:00:42.380 -->
01:00:49.666In our case, just check on our website, solarworksio, or just send an email to info at solarworksio and feel free to reach out to us.
01:00:49.666 -->
01:00:54.365I'm going to send you some further details and manuals and explanation about our mesh grids and standalone systems.
01:00:55.003 -->
01:00:55.980Yeah, similar.
01:00:55.980 -->
01:00:59.442So okrasolarcom, O-K-R-A.
01:00:59.442 -->
01:01:05.293Solar, and then you can just email me directly at afnan at okrasolarcom.
01:01:05.293 -->
01:01:10.090I'm always keen to talk to anybody who's interested in doing energy access better.
01:01:10.389 -->
01:01:11.552All right, thanks a lot.
01:01:11.552 -->
01:01:22.128I have the feeling that we will talk more often in the future because this is such an interesting subject that should get more space on media in our sector.
01:01:22.128 -->
01:01:22.889Bye-bye.
01:01:23.010 -->
01:01:23.972Cheers guys, Bye-bye.
01:01:24.193 -->
01:01:24.793Thanks a lot.
01:01:24.793 -->
01:01:33.891This episode of the mini-grid business has been brought to you by INENSUS, your one-stop shop for sustainable mini-grids.
01:01:33.891 -->
01:01:44.231For more information on how to make mini-grids work, visit our website, inensus.com, or contact us through the links in the show notes.
01:01:44.231 -->
01:01:56.753The mini grid business powered by INENSUS.
