DC Mesh Grids

WEBVTT
00:00:09.391 --> 00:00:13.935
Solar mini-grids have turned from small pilots to an electrification wave.
00:00:13.935 --> 00:00:21.265
We were there when mini-grid regulation was established, when financial transactions were closed.
00:00:21.265 --> 00:00:25.512
We saw new technology thrive and companies fail.
00:00:25.512 --> 00:00:28.282
This is where we tell the stories.
00:00:28.282 --> 00:00:34.433
This is where we discuss the future the mini-grid business Powered by Enensis.
00:00:36.140 --> 00:00:37.142
Hello, this is Nico.
00:00:37.142 --> 00:00:39.886
Welcome to our episode on Mesh Grids.
00:00:39.886 --> 00:00:46.015
My guests are Afnan Hannan from Okra Solar and Felix Boldt from Solarworx.
00:00:46.015 --> 00:00:51.231
Afi is a co-founder and the CEO of Okra.
00:00:51.231 --> 00:01:00.027
Before founding Okra, he was working in building automation and IoT and then transitioned to energy access through mesh grids.
00:01:00.027 --> 00:01:08.694
Afi aims to unlock opportunities for last mile communities around the world with access to affordable energy.
00:01:08.694 --> 00:01:18.375
Felix is the founder of Solarworx, with 10 years of experience in solar off-grid in sub-Saharan Africa.
00:01:18.375 --> 00:01:24.147
Initially, he started his journey with Mobisol and is now with Solarworx.
00:01:24.147 --> 00:01:29.975
He and his team are manufacturing standalone off-grid solar systems and mesh grids.
00:01:29.975 --> 00:01:35.712
Solarworks also engages in EPC business in Africa and Germany.
00:01:35.712 --> 00:01:42.612
Well, now, as we're talking about mesh grids, Felix, what is a mesh grid?
00:01:43.480 --> 00:01:44.322
Yeah, hi everyone.
00:01:44.322 --> 00:01:46.450
Thanks very much for having us in the first place.
00:01:46.450 --> 00:01:49.147
So just straight to the first question.
00:01:49.740 --> 00:01:53.731
So a mesh grid is a decentralized DC microgrid.
00:01:53.731 --> 00:02:05.030
It'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 --> 00:02:22.669
So 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 --> 00:02:33.347
So 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 --> 00:02:43.689
There'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 --> 00:02:53.445
And 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 --> 00:02:57.782
Yeah it's safe to touch as long as the voltage stays below a certain level.
00:02:57.782 --> 00:03:03.100
Right, and that's what you're basically doing, exactly, yeah, but that also increases the current right.
00:03:03.100 --> 00:03:22.073
To 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 --> 00:03:30.078
we 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 --> 00:03:40.036
So 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 --> 00:04:13.092
most 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 --> 00:04:20.451
Therefore it may also be reasonable to accept the slightly higher losses due to the low voltage.
00:04:20.451 --> 00:04:23.675
But save a lot of costs, save a lot of capex after all.
00:04:23.675 --> 00:04:28.605
Right, thanks to the lower installation costs and the lower material costs.
00:04:28.906 --> 00:04:32.380
Yeah, thanks, Felix and Nico, for explaining a bit about mesh grids.
00:04:32.380 --> 00:04:48.125
And 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 --> 00:04:50.995
Let's try to imagine how these two things would be different.
00:04:50.995 --> 00:05:00.350
So 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 --> 00:05:03.706
Let's say you know one or two kilometers, as you mentioned.
00:05:03.706 --> 00:05:21.910
In 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 --> 00:05:26.567
So 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 --> 00:05:31.124
So if you had a hundred houses in the mini grid, you're sending a hundred kilowatts over the line.
00:05:31.124 --> 00:05:40.259
In 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 --> 00:05:46.124
And then finally we know that power loss is proportionate to I squared R.
00:05:46.124 --> 00:05:51.107
So in the mini grid 100 kilowatts squared is 10,000.
00:05:51.107 --> 00:05:53.208
So 10,000 times power loss.
00:05:53.208 --> 00:05:56.689
In the mesh grid one kilowatt squared is one.
00:05:56.689 --> 00:05:59.612
So basically one times power loss.
00:05:59.612 --> 00:06:04.735
So in that mini grid you would have 10,000 times more power loss over the line.
00:06:04.735 --> 00:06:12.247
And 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 --> 00:06:17.812
And in the mesh grids you can use that lower voltage because you're having less power over the line.
00:06:17.812 --> 00:06:21.750
So hopefully that explains a little bit more as well on the distribution cost savings.
00:06:22.360 --> 00:06:33.286
I 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 --> 00:06:41.428
I 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 --> 00:06:46.288
You 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 --> 00:06:56.166
So 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 --> 00:07:04.750
If 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 --> 00:07:06.413
They try that once but never try it again.
00:07:06.413 --> 00:07:12.283
So 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 --> 00:07:24.449
The 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 --> 00:07:25.300
Yeah, it's safe to touch.
00:07:25.300 --> 00:07:26.523
Then that drops the cost of installation also dramatically.
00:07:26.523 --> 00:07:59.190
Yeah, 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 --> 00:08:17.156
And 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 --> 00:08:21.576
So why not operating two systems under one management?
00:08:22.117 --> 00:08:22.779
Yeah for sure.
00:08:22.779 --> 00:08:25.166
So I could jump in there.
00:08:25.166 --> 00:08:28.863
I might do a bit of a segue and then come back On the segue.
00:08:28.863 --> 00:08:31.550
It's basically about DC mesh grids.
00:08:31.550 --> 00:08:43.389
So 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 --> 00:08:57.490
But 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 --> 00:09:05.267
I 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 --> 00:09:08.630
So I think that's maybe something interesting we can touch on in a bit.
00:09:08.691 --> 00:09:35.379
But 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 --> 00:09:38.549
We're talking, you know, thousands of households or more than that.
00:09:38.549 --> 00:09:40.184
Households are close to each other.
00:09:40.184 --> 00:09:50.394
There'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 --> 00:09:55.731
Whenever an individual user needs some energy, they can get it, and they can get it reliably.
00:09:55.731 --> 00:09:57.023
That's the benefit of the mini grid.
00:09:57.023 --> 00:10:00.292
And then on the other side you've got solar home systems.
00:10:00.292 --> 00:10:09.365
If 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 --> 00:10:11.407
We're trying to take the benefits of both.
00:10:11.807 --> 00:10:25.934
We'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 --> 00:10:37.283
So 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 --> 00:10:46.865
And 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 --> 00:10:51.471
And then you're looking at SHS making sense for everything that's isolated.
00:10:51.471 --> 00:11:13.668
And 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 --> 00:11:16.527
I think it is very exciting to be thinking about the future there too.
00:11:16.960 --> 00:12:00.491
Yeah, 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 --> 00:12:18.152
And 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 --> 00:12:20.886
Is that the way that you're also working in africa?
00:12:21.269 --> 00:12:21.428
well?
00:12:21.448 --> 00:12:33.735
For 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 --> 00:12:39.371
Right, not exactly an ac mini grid, but definitely larger than what any private household in africa would ever consume.
00:12:39.371 --> 00:12:46.433
So 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 --> 00:12:50.171
That's basically saying mesh grids are the middle child between SHS and AC mini grids.
00:12:50.171 --> 00:13:06.863
I 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 --> 00:13:14.409
So 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 --> 00:13:17.947
Municipality has to be convinced, exclusivity agreements have to be signed.
00:13:18.048 --> 00:13:55.610
It'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 --> 00:13:56.631
Am I correct?
00:13:56.631 --> 00:14:00.196
Maybe I can make a start if you can add to that, our systems are very modular.
00:14:00.196 --> 00:14:06.092
So 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 --> 00:14:08.889
So there's a very cabler, seamless integration possible.
00:14:08.980 --> 00:14:19.696
But 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 --> 00:14:24.091
So you can even have customers that already have been paying off their standalone system for a long time.
00:14:24.091 --> 00:14:28.171
They've bought it in any random hardware shop in town and they installed themselves.
00:14:28.171 --> 00:14:36.145
And 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 --> 00:14:44.504
Of 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 --> 00:14:46.366
Right, it must be within proportion.
00:14:46.366 --> 00:14:55.753
But 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 --> 00:15:06.490
Yeah, 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 --> 00:15:14.039
No, 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 --> 00:15:21.183
We 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 --> 00:15:22.586
You can really use anything.
00:15:22.586 --> 00:15:23.807
That's the great flexibility.
00:15:23.807 --> 00:15:28.053
So I don't think if you're a developer, you want to be dependent on just one manufacturer only.
00:15:28.053 --> 00:15:29.196
You want to have that flexibility.
00:15:29.196 --> 00:15:36.869
As 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 --> 00:15:38.091
Yeah, nice approach.
00:15:38.091 --> 00:15:40.086
Afi, how are you handling this?
00:15:40.368 --> 00:16:20.865
Yeah, 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 --> 00:16:35.328
So 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 --> 00:16:46.264
However, 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 --> 00:16:52.749
But 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 --> 00:16:59.813
And 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 --> 00:17:02.482
From a technical perspective, it has been working greatly.
00:17:02.482 --> 00:17:20.827
However, 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 --> 00:17:34.601
So 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 --> 00:17:35.784
In fact, the country was there.
00:17:35.784 --> 00:17:41.728
I 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 --> 00:17:49.127
So 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 --> 00:17:54.432
Okay, let's talk about business models in more detail later.
00:17:54.432 --> 00:17:56.367
I have one more question.
00:17:56.367 --> 00:18:13.296
Like 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 --> 00:18:25.709
Could 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 --> 00:18:28.259
Yeah, taking a step back.
00:18:28.259 --> 00:18:45.219
To 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 --> 00:18:52.202
So really the end consumer is getting the same thing that they would get from accessing energy from a mini grid.
00:18:52.202 --> 00:18:53.905
Where are the limitations?
00:18:53.905 --> 00:19:07.282
So 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 --> 00:19:09.577
Both single phase, right, yeah, single phase AC.
00:19:09.577 --> 00:19:19.565
And 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 --> 00:19:27.739
But 99.9% of all the energy use cases in last mile communities is covered by 3.6.
00:19:27.739 --> 00:19:34.259
But 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 --> 00:19:46.957
This 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 --> 00:19:52.895
So 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 --> 00:20:00.128
And that productive hub can be connected with an AC to DC converter into the mesh grid.
00:20:00.128 --> 00:20:02.404
So the excess power from the mini-grid is actually supplied into the mesh grid.
00:20:02.404 --> 00:20:08.419
So 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 --> 00:20:12.218
That's actually something that we're working on designing with a couple of our developers.
00:20:12.218 --> 00:20:12.640
Right now.
00:20:12.640 --> 00:20:16.296
We've got the prototypes out and getting excited for a field deployment.
00:20:16.316 --> 00:20:17.559
Sounds good, sounds good.
00:20:17.559 --> 00:20:23.842
So now, what is the difference between the Okra solar technology and the SolarWorks technology?
00:20:24.035 --> 00:20:30.409
I 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 --> 00:20:31.718
We believe in that.
00:20:31.718 --> 00:20:40.948
You 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 --> 00:20:48.535
So 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 --> 00:20:54.903
Inverters 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 --> 00:20:58.897
For the devices that have a significantly higher demand, we are quite open end.
00:20:58.897 --> 00:21:01.763
So, like we have no limitation in terms of how many kilowatts can be used.
00:21:01.763 --> 00:21:28.818
Again, 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 --> 00:21:31.426
So that's one difference, I think, between us and okra.
00:21:31.776 --> 00:21:40.906
Another 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 --> 00:21:49.974
In 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 --> 00:21:57.980
So 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 --> 00:22:07.604
But 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 --> 00:22:25.063
I 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 --> 00:22:31.682
On the low side, as we call it, which is usually 12 volt power output, I see listening to you, felix.
00:22:32.064 --> 00:22:56.230
Would 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 --> 00:22:58.075
Is that a correct summary?
00:22:58.395 --> 00:23:05.635
well, afi and felix, well I think that both systems are still very flexible in components that can be used.
00:23:05.635 --> 00:23:08.624
I would disagree to this part of the explanation.
00:23:08.624 --> 00:23:11.538
Both of us we are aiming for economies of scale.
00:23:11.538 --> 00:23:17.039
I 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 --> 00:23:19.836
We are operating mesh grids at the moment in Cameroon and Zambia.
00:23:19.836 --> 00:23:26.749
We 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 --> 00:23:28.657
But I think Afi can speak more to that.
00:23:29.200 --> 00:23:54.696
I 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 --> 00:24:00.218
I think it really stems from that and that actually feeds into a couple of design decisions.
00:24:00.218 --> 00:24:10.036
And 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 --> 00:24:24.797
We 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 --> 00:24:26.922
The boxes are in every household, just to clarify.
00:24:26.922 --> 00:24:30.067
But, like you know, you have one solar system every four or so.
00:24:30.194 --> 00:24:33.726
Yeah so imagine like where the battery and the PV is installed.
00:24:33.726 --> 00:24:45.691
In 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 --> 00:25:04.805
So 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 --> 00:25:06.131
So those are some of the kind of trade-offs between the two techs.
00:25:06.151 --> 00:25:13.641
Well, 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 --> 00:25:25.324
Obviously, 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 --> 00:25:40.174
So 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 --> 00:25:52.326
Right, 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 --> 00:26:02.063
I 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 --> 00:26:12.116
Yeah, 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 --> 00:26:12.878
Exactly so.
00:26:12.878 --> 00:26:15.384
The argument of safety indeed is gone.
00:26:15.384 --> 00:26:20.746
You'd have to introduce a lot more safety features, mcbs, residual current arresters and all that kind of stuff.
00:26:20.746 --> 00:26:29.501
I agree, but at the same time you still have that big benefit of decentralization.
00:26:29.501 --> 00:26:32.333
So 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 --> 00:26:46.319
So 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 --> 00:26:51.660
You can retrofit or use them and then make something like a mesh grid, but on an AC basis.
00:26:51.660 --> 00:26:57.855
But 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 --> 00:27:46.519
But 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 --> 00:27:56.142
This 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 --> 00:28:08.342
Do 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 --> 00:28:09.124
Good question.
00:28:09.124 --> 00:28:12.199
Indeed, we haven't done the calculation at that point.
00:28:12.199 --> 00:28:30.964
As 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 --> 00:28:41.309
And 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 --> 00:28:46.557
Ac network and support the voltage at a certain point.
00:28:46.557 --> 00:28:58.761
But 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 --> 00:29:06.681
Beyond these 1.3 kilometers you would anyway not connect any heavy loads to the AC network anymore.
00:29:07.241 --> 00:29:09.867
And therefore, a DC network could do the job as well.
00:29:10.736 --> 00:29:23.845
And 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 --> 00:29:26.101
That's why it should be configured with AC at the output.
00:29:26.101 --> 00:29:31.768
But these are philosophical things that I guess we will be continuing to discuss and iron out.
00:29:31.768 --> 00:29:50.037
Another 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 --> 00:30:04.670
Yeah, 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 --> 00:30:07.625
Depending, of course, on the national standards.
00:30:07.625 --> 00:30:09.121
Do you need steel poles?
00:30:09.121 --> 00:30:11.042
Can you install wooden poles?
00:30:11.042 --> 00:30:15.066
What is the voltage drop that is acceptable?
00:30:15.066 --> 00:30:20.086
And well then, you design your system and the capex results accordingly.
00:30:20.487 --> 00:30:20.788
Gotcha.
00:30:20.788 --> 00:30:23.964
So that's like $140 per connection roughly.
00:30:24.336 --> 00:30:32.747
Well, 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 --> 00:30:33.957
So 140,.
00:30:33.957 --> 00:30:38.488
So if I get that right, if that's like 10 meters of distance between households.
00:30:39.297 --> 00:30:40.521
Which is very dense already.
00:30:40.815 --> 00:30:43.321
Which would be very dense between households, which is very dense already, which would be very dense.
00:30:43.321 --> 00:31:00.144
And 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 --> 00:31:02.522
Is that aligned with what you're looking at, felix?
00:31:02.855 --> 00:31:05.083
Well, if it's just the cost of distribution.
00:31:05.143 --> 00:31:13.125
Yes, 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 --> 00:31:19.704
But 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 --> 00:31:23.857
They come at around 80 to 100, depending on which country you're operating in, per pole.
00:31:23.857 --> 00:31:33.442
So 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 --> 00:31:42.847
But 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 --> 00:32:01.029
Even 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 --> 00:32:26.526
Yes, 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 --> 00:32:29.699
They're like six meter long galvanized steel poles or something like that.
00:32:29.699 --> 00:32:34.721
It'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 --> 00:32:35.743
It's really a pipe.
00:32:35.743 --> 00:32:40.537
It's really yeah, yes, exactly okay, good, now let's go into the business models.
00:32:40.537 --> 00:32:42.778
You already started talking about this.
00:32:42.778 --> 00:32:48.820
So both of you consider yourselves as the manufacturers of the technology.
00:32:48.820 --> 00:33:08.833
So 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 --> 00:33:09.313
Is that correct?
00:33:09.313 --> 00:33:10.192
Yeah, that's largely correct.
00:33:10.434 --> 00:33:24.815
For 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 --> 00:33:28.190
So we are integrated with software as a service companies in a sector like Agaza, paygops, payg, opia and so on.
00:33:28.190 --> 00:33:33.585
Our philosophy we don't want to necessarily bind the developers fully on our software.
00:33:33.585 --> 00:33:39.804
However, of course, there is a certain dependency by our developers to us because they purchase our mesh grid technology.
00:33:39.804 --> 00:33:43.682
But principally yes, the explanation is correct we're working on a B2B basis.
00:33:43.682 --> 00:33:44.746
We manufacture hardware.
00:33:45.507 --> 00:33:52.880
Our 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 --> 00:34:21.802
But 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 --> 00:34:26.762
So they are selling energy as a service to end customers.
00:34:26.762 --> 00:34:28.730
That can be done on an upfront basis.
00:34:28.730 --> 00:34:36.661
There'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 --> 00:34:50.978
And 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 --> 00:35:05.505
That'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 --> 00:35:19.025
But 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 --> 00:35:21.342
our developers not so much.
00:35:21.342 --> 00:35:31.717
Both 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 --> 00:35:42.101
I 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 --> 00:35:44.498
So there's a subsidy for mini grids, they do mini grids.
00:35:44.498 --> 00:35:47.512
If there is a tender for salome system, they do that right.
00:35:47.512 --> 00:35:53.295
But without, without certain degree of subsidy very few of those off-grid businesses are generally profitable, I see.
00:35:53.775 --> 00:35:55.438
Is that the same on your end, afi?
00:35:56.018 --> 00:36:12.835
um, 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 --> 00:36:15.701
They have a very, very, very tough job.
00:36:15.701 --> 00:36:29.737
They 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 --> 00:37:02.101
Just 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 --> 00:37:05.585
The local developers are going to do a much better job than we can.
00:37:05.585 --> 00:37:16.065
They 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 --> 00:37:26.538
So 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 --> 00:37:30.467
But maybe some of the things that we have specifically focused on.
00:37:30.467 --> 00:37:33.925
One is to make sure that all of the billing data can come through.
00:37:33.925 --> 00:37:50.804
There'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 --> 00:37:57.648
So, yeah, I think the software part is also a pretty important aspect of making sure that operations can scale.
00:37:58.775 --> 00:38:13.891
And 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 --> 00:38:19.288
Let's talk about average revenue per user that you have seen in your systems.
00:38:19.288 --> 00:38:24.317
I guess they are very similar to what we know from AC systems, aren't they?
00:38:24.516 --> 00:38:30.358
I keep the information high level because I don't want to give away data from any of our partners.
00:38:30.358 --> 00:38:47.068
But 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 --> 00:38:55.025
So like $16, $17 ARPU, which I think would be viable for even higher CapEx solutions.
00:38:55.025 --> 00:38:57.818
So there's the entire range and what's the lesson there?
00:38:57.818 --> 00:39:09.195
The 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 --> 00:39:22.909
And 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 --> 00:39:26.971
Otherwise things are going to balloon out if the currency depreciates.
00:39:27.440 --> 00:39:32.833
Let's look at the average operations cost per year of about $15 to $20.
00:39:32.833 --> 00:39:37.393
And let's look at a realistic ARPU of $5 to $10.
00:39:37.393 --> 00:39:42.769
That doesn't change in local currency, Like it's in local currency and it stays there from today's date.
00:39:42.769 --> 00:39:53.751
And 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 --> 00:39:59.813
This can actually lead to a sustainable seven-year payback period without subsidies.
00:39:59.813 --> 00:40:07.487
So 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 --> 00:40:14.420
And 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 --> 00:40:20.108
That's massive and that's going to make all these projects even more viable and more attractive.
00:40:20.108 --> 00:40:31.463
And 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 --> 00:40:32.826
Yeah, thank you Afi.
00:40:32.826 --> 00:40:39.190
That indeed makes a lot of sense that the ARPUs are very similar to what you see in AC mini grids.
00:40:39.190 --> 00:40:49.715
It'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 --> 00:40:53.447
So therefore expecting the same ARPUs is very reasonable.
00:40:53.447 --> 00:41:07.304
But 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 --> 00:41:14.364
So 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 --> 00:41:19.695
So now what is an average cost per connection in a mesh grid?
00:41:20.681 --> 00:41:37.726
It'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 --> 00:41:48.275
In 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 --> 00:41:57.291
So 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 --> 00:42:12.271
You 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 --> 00:42:14.360
So there's no poles and nothing of that required.
00:42:14.360 --> 00:42:16.226
So that is the lowest entry point cost.
00:42:16.226 --> 00:42:22.432
It's about 150 dollars, for us at least, and then it goes as high as maybe 300, 350 dollars.
00:42:22.432 --> 00:42:35.327
If 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 --> 00:42:40.635
Yeah, I'm just looking at some numbers right now.
00:42:40.635 --> 00:42:47.311
I think getting at the $300 per connection and even below, that's quite impressive, felix.
00:42:47.311 --> 00:42:53.115
On our side, the way we look at it is that it's not just a short-term thing.
00:42:53.115 --> 00:43:00.731
It 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 --> 00:43:03.922
I'm just saying that we design the systems.
00:43:03.922 --> 00:43:16.465
For 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 --> 00:43:20.184
So the DES program says 83 watt hours per consumer per day.
00:43:20.503 --> 00:43:36.634
We 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 --> 00:43:52.512
So 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 --> 00:43:59.393
But 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 --> 00:44:10.487
And 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 --> 00:44:15.311
So yeah, I think we can get the cost down, but it comes at a cost to reduce the cost.
00:44:15.311 --> 00:44:18.489
So I would say let's keep it there.
00:44:19.119 --> 00:44:20.105
Well, afi, one question.
00:44:20.105 --> 00:44:22.288
There is the inverter included in that package.
00:44:22.380 --> 00:44:28.411
Yeah, that's all included inverters, batteries, distribution screws as well.
00:44:28.411 --> 00:44:33.112
Ddp past customs to the developer in Nigeria.
00:44:33.112 --> 00:44:33.251
Yeah.
00:44:33.724 --> 00:44:36.847
I think that would explain also why there's a slight cost difference here.
00:44:36.847 --> 00:44:43.534
Right, 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 --> 00:44:49.947
It 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 --> 00:44:55.813
So, in order to create the offtake and create the demand, we would want to sell appliances.
00:44:55.813 --> 00:45:00.469
So the developer should be selling appliances to the clients, because only that really they generate enough revenue.
00:45:00.469 --> 00:45:10.740
Because, 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 --> 00:45:15.612
So there's an absolute necessity for grants if you don't have that $5 per customer.
00:45:15.612 --> 00:45:20.619
And then the additional revenue you have to generate really from selling appliances to the same customer base.
00:45:21.000 --> 00:45:42.507
Yeah, 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 --> 00:45:56.072
This 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 --> 00:45:58.706
Yeah, but again, I can tell you where this comes from.
00:45:58.706 --> 00:46:05.519
So 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 --> 00:46:07.086
So that drives the cost.
00:46:07.086 --> 00:46:16.231
And 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 --> 00:46:19.630
Right, I'm talking about markets that are wall to wall.
00:46:19.630 --> 00:46:22.077
The next customer is zero meters away.
00:46:22.137 --> 00:46:24.288
Basically, yeah, which is a nice scenario to have.
00:46:24.288 --> 00:46:26.585
Well, but it's not a very uncommon scenario.
00:46:26.585 --> 00:46:29.344
Let's put it this way there's tons of those markets in.
00:46:29.364 --> 00:46:33.324
Nigeria that you can totally electrify like that and I'm sure that you know Afi.
00:46:33.324 --> 00:46:44.208
I 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 --> 00:46:46.318
But 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 --> 00:46:48.603
Average for us would be about 250 to 300.
00:46:48.603 --> 00:46:50.646
That's really also our cost all right.
00:46:51.288 --> 00:47:00.432
Where 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 --> 00:47:02.501
Yeah.
00:47:02.501 --> 00:47:11.704
So 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 --> 00:47:22.010
So 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 --> 00:48:01.106
But 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 --> 00:48:22.751
And 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 --> 00:48:31.983
Well, 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 --> 00:48:59.322
So 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 --> 00:49:01.389
Like, do we need to go through this same process?
00:49:01.389 --> 00:49:12.210
So 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 --> 00:49:17.570
So 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 --> 00:49:20.838
So that's one of the things I'm actively working on right now.
00:49:20.878 --> 00:49:47.291
I'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 --> 00:49:55.887
So 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 --> 00:50:07.646
And 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 --> 00:50:20.865
And 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 --> 00:50:23.733
But I'll stop there and pass it back to you guys.
00:50:23.900 --> 00:50:30.641
Yeah, but, afi, I guess you would not be able to escape tariff regulation after all with mesh grids.
00:50:30.641 --> 00:50:36.572
But maybe we need a mesh grid specific approach to tariff regulation.
00:50:36.572 --> 00:50:37.014
Let's see.
00:50:37.014 --> 00:50:40.402
This would be a subject for a next episode, maybe.
00:50:41.164 --> 00:50:47.887
New 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 --> 00:50:59.641
So 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 --> 00:51:02.068
So again, then there comes in the need for subsidy.
00:51:02.068 --> 00:51:06.706
I would largely also echo the entry barriers and challenges that Afi was mentioning.
00:51:06.706 --> 00:51:16.945
I 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 --> 00:51:18.588
So that's been a job well done.
00:51:18.628 --> 00:51:22.534
Thanks, thanks to the whole ecosystem, thanks to Enansis as well.
00:51:22.534 --> 00:51:26.528
Yeah, indeed, yeah, yeah, it's a forward looking ecosystem, I think.
00:51:26.548 --> 00:51:26.809
Yeah.
00:51:26.809 --> 00:51:39.134
Well, 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 --> 00:51:48.543
We 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 --> 00:52:23.143
So 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 --> 00:52:25.592
You 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 --> 00:52:30.985
Okay, so you were already talking about Cameroon and Uganda and Nigeria and Zambia.
00:52:30.985 --> 00:52:35.442
What other markets and what other countries are most interesting to you?
00:52:35.802 --> 00:52:36.425
I think there is.
00:52:36.425 --> 00:52:57.023
It'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 --> 00:53:11.619
If 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 --> 00:53:16.887
So 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 --> 00:53:19.172
Afi which countries are you looking at?
00:53:19.172 --> 00:53:20.382
What?
00:53:20.402 --> 00:53:21.445
markets.
00:53:21.445 --> 00:53:29.688
Yeah, 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 --> 00:53:33.971
But right now we have like a couple of criteria that we look at.
00:53:33.971 --> 00:53:43.731
So 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 --> 00:53:57.467
So, 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 --> 00:54:01.467
There's also the most off-grid developers globally in the world in Nigeria right now.
00:54:01.467 --> 00:54:03.132
So Nigeria makes a lot of sense.
00:54:03.132 --> 00:54:05.585
But, as you know, we're also in Haiti.
00:54:05.585 --> 00:54:13.211
So 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 --> 00:54:29.516
There'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 --> 00:54:35.326
Yeah, Are you setting up maintenance and service structures in each country you're going to?
00:54:35.545 --> 00:54:59.583
Yeah, 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 --> 00:55:01.327
So yeah, Good Felix.
00:55:01.467 --> 00:55:02.572
anything to add from your side?
00:55:02.572 --> 00:55:03.922
How are you handling this?
00:55:04.324 --> 00:55:07.844
No, I think, no, I think that's, I think, covered it most Same, thing, same.
00:55:07.844 --> 00:55:08.005
Thing.
00:55:08.505 --> 00:55:10.612
Now a quick feedback from your sides.
00:55:10.612 --> 00:55:16.231
How many systems have you installed already and how many customers have you connected, each of you?
00:55:16.481 --> 00:55:16.702
Nice.
00:55:16.702 --> 00:55:34.347
So, 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 --> 00:55:43.193
Things 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 --> 00:55:55.594
So 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 --> 00:56:03.168
That's 120,000 people directly affected by solar work standalone system technology nice, okay, good.
00:56:03.228 --> 00:56:07.601
So now what uh is up next for mesh grid technology?
00:56:07.601 --> 00:56:09.485
What are you currently developing?
00:56:09.666 --> 00:56:09.867
you know.
00:56:09.867 --> 00:56:11.409
Firstly, no secrets.
00:56:11.409 --> 00:56:31.105
Like 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 --> 00:56:36.041
Yeah, so that doting on the industry asides we're working on a couple of things.
00:56:36.041 --> 00:56:38.284
So let's start with the software side.
00:56:38.284 --> 00:56:54.211
With 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 --> 00:56:59.289
It might sound a bit boring but we're building an internal product called the Network Planner 3000.
00:56:59.289 --> 00:57:10.762
And 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 --> 00:57:26.871
Another 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 --> 00:57:32.266
But having more than 100 dashboards is not really that practical for any of our partners.
00:57:32.266 --> 00:57:45.570
So 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 --> 00:57:50.494
And I might just quickly drop one update on the hardware side.
00:57:50.494 --> 00:58:02.652
But 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 --> 00:58:21.563
So 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 --> 00:58:26.880
And then it hops that data to a point where there is 2G connection and we send it out from there.
00:58:26.880 --> 00:58:35.329
So 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 --> 00:58:40.050
Yeah, I think from our side we also consistently develop the software in particular further.
00:58:40.050 --> 00:58:45.585
We 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 --> 00:58:54.188
But 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 --> 00:59:00.731
So 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 --> 00:59:01.862
That's one of the things.
00:59:01.862 --> 00:59:04.934
We're also constantly improving the hardware further.
00:59:04.934 --> 00:59:13.451
So 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 --> 00:59:16.829
And there's also other little little add-ons to the grid.
00:59:16.829 --> 00:59:22.034
So we have laura controlled street lighting that sits on every pole of the mesh grid.
00:59:22.034 --> 00:59:24.061
So laura, just to say what about laura?
00:59:24.061 --> 00:59:35.661
That'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 --> 00:59:37.605
We actually run firmware updates via laura.
00:59:37.605 --> 00:59:39.628
So those things we are constantly improving.
00:59:39.668 --> 00:59:43.242
But you know those little switches that control the street lights.
00:59:43.242 --> 00:59:46.090
We do already have them in the field, testing them currently in cameroon.
00:59:46.090 --> 00:59:55.088
We 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 --> 01:00:00.175
Because 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 --> 01:00:00.679
If I'm looking at, for instance, axol, it's another company.
01:00:00.679 --> 01:00:05.626
They'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 --> 01:00:11.052
There's no need to step down the voltage for those, like you know, stronger DC brushless motors.
01:00:11.052 --> 01:00:13.505
So those kinds of things are currently in the pipeline or development.
01:00:14.019 --> 01:00:19.983
Interesting, interesting, yeah, and development, interesting, interesting, yeah.
01:00:19.983 --> 01:00:20.365
Thanks a lot, felix.
01:00:20.365 --> 01:00:20.746
Thanks a lot, afi.
01:00:20.746 --> 01:00:21.469
This was highly interesting.
01:00:21.469 --> 01:00:31.356
I 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 --> 01:00:33.561
So thanks a lot to both of you.
01:00:33.561 --> 01:00:42.108
Before 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 --> 01:00:49.666
In 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 --> 01:00:54.365
I'm going to send you some further details and manuals and explanation about our mesh grids and standalone systems.
01:00:55.003 --> 01:00:55.980
Yeah, similar.
01:00:55.980 --> 01:00:59.442
So okrasolarcom, O-K-R-A.
01:00:59.442 --> 01:01:05.293
Solar, and then you can just email me directly at afnan at okrasolarcom.
01:01:05.293 --> 01:01:10.090
I'm always keen to talk to anybody who's interested in doing energy access better.
01:01:10.389 --> 01:01:11.552
All right, thanks a lot.
01:01:11.552 --> 01:01:22.128
I 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 --> 01:01:22.889
Bye-bye.
01:01:23.010 --> 01:01:23.972
Cheers guys, Bye-bye.
01:01:24.193 --> 01:01:24.793
Thanks a lot.
01:01:24.793 --> 01:01:33.891
This 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 --> 01:01:44.231
For 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 --> 01:01:56.753
The mini grid business powered by INENSUS.