Solar mini-grid technology - trends and markets

Speaker 1: 0:09

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

Speaker 2: 0:35

Hello everyone, this is Nico. Welcome to our podcast on solar mini-grid technology trends and markets. I'm here today with my Inensos colleagues, Dipto Maggiomda and Mike Arrasarathemo, both mini-grid technology experts in our company. Dipto, do you want to start and introduce yourself quickly?

Speaker 3: 0:56

Yeah, sure, nico. Hi everyone. I am Dipto Maggiomda, working as a mini-grid expert in Innsis since 2018. I have been working in the mini-grid sector for some 10 years now, have worked a lot in Bangladesh for it call, also in the mini-grid sector, and then I did my master's in renewable energy from Germany, and since then I have been working with Innsis, mostly focusing on the Sub-Saharan African market. Looking forward to this.

Speaker 2: 1:27

Thanks, dipto Mike. Please tell us about yourself quickly.

Speaker 4: 1:31

Hello, my name is Mike Arrasar. I'm based in Nairobi, Kenya. I'm a renewable energy expert. I've been working in the solar industry for the past seven years, mainly focused in Kenya and industry. I have a master's in sustainable energy and environmental systems from Cyprus All right Dipto.

Speaker 2: 1:52

Let's talk about system components in solar mini-grids. What type of system components do we need to consider when we think about solar mini-grids?

Speaker 3: 2:02

When we are discussing about an AC mini-grid setup, the key components are, of course, the solar panels, batteries and inverters. On the generation side, a diesel generator can be there for backup purpose and for the distribution side, poles, conductors and energimeters are the key assets. A diesel generator is typically needed as a backup when we have days with limited solar generation or battery backup. But as we see that as the batteries price are decreasing over time and the diesel fuel prices are seeing volatility in the market, we are, I think, poised to transitioning towards diesel-free mini-grids in the near future.

Speaker 2: 2:45

Yeah, and there are also many other aspects. Think about climate change. Everybody wants the CO2 emissions anymore, especially when you use donations, grants from donors. They usually require you to go without the diesel generator, and other aspects are maintenance of diesel generators in rural areas. It's so, so much work and so much to follow up on. Every few hundred hours you need to exchange the oil. You need to send somebody to site. You need to bring the oil. You need to ship the diesel fuel. The diesel fuel can be stolen. You need to protect that.

Speaker 3: 3:27

And also the transportings.

Speaker 2: 3:29

Transport is very difficult indeed. Yeah, there are so many reasons why to go without a diesel generator, and you already mentioned that this may be possible in the future, especially if we find a way to mitigate the demand risk, as already explained in one of our earlier episodes, for example, with edge computers.

Speaker 3: 3:51

Right.

Speaker 2: 3:52

Mike DIPTO already introduced the generation assets of a solar mini-grid. Briefly, Do you want to talk about how the world market prices for these generation assets developed over time?

Speaker 4: 4:04

So the cost of the case solar generation assets, which are the PV panels, batteries and inverters, are significantly changed the recent years, impacting the system configuration. When we're looking at the PV panels, prices have steadily declined since the early 2000s, driven mainly by the technological advancement and economies of scale and fierce competition among manufacturers. If you look at from 2010, a typical 60 cell panel cost around two watts, while today it's closer to 20 cents per watt, representing a staggering 90% decrease over the decade. When we're looking at the battery aspect, like LFP batteries, a typical 5 kilowatt LFP battery pack costs at now around 500 to 700 dollars, compared to 1,000 to 1,500 USD just a few years ago, representing a 30% to 50% decrease For the investors. The witness decrease of 30% to 50% whereby, when you look at the 5kW inverter typically now costs 500 to 1,000 USD compared to 1,000 to 2,000 dollars a few years ago.

Speaker 2: 5:22

Okay, and there's also new technology on its way to the market. What is that the new?

Speaker 4: 5:28

technology is currently in that, as shown, the most promising the sodium-based batteries, where they use the sodium ions as charge carriers, compared to the lithium ion batteries that are commonly used. Some potential benefits of the sodium ion batteries include that sodium is more abundant and cheaper compared to the currently used lithium ion batteries, with a projected cost of around 40 to 80 USD per kilowatt hour, which is projected to go even below 15 USD per kilowatt in future, and, with this in mind, sodium batteries are projected to make battery components cheap to around 20% of the capex.

Speaker 2: 6:09

Which is great, because in the past, batteries have always been the most expensive component and the component with the shortest lifetime, so therefore the largest part of the cost was actually going into batteries, and this is changing now. Batteries are becoming a commodity. With the LFP technology that you just introduced, we have already made a big step to the range of something like 100 USD per kilowatt hour on a cell level and maybe 150 to 200 USD on a battery systems level, and now, if the prices are further going down with the sodium battery, we can potentially reach price levels of 100 USD per kilowatt hour for the battery system and, as you already said, like well below 50, 40, maybe even towards 20 USD per kilowatt hour on a cell level.

Speaker 2: 7:01

Sodium batteries have been developed over the last year, and in China you can now see the first cars deploying sodium ion batteries as their energy storage, and the slight challenge for sodium ion batteries in the car industry is that they have a lower energy density compared to the LFP batteries.

Speaker 2: 7:26

So, from a car batteries perspective, you have the NMC technology, which has the highest energy density, which is used in the high performance cars and the very expensive premium cars with very long range and, for the shorter ranges, lfp batteries are being used and soon we will see in Europe, in the United States, but also probably in Africa, sodium ion batteries as the next trend, which have even a lower range in a car, but as stationary batteries in mini grids they are ideally suited and I'm really looking forward to seeing how sodium batteries will finally change the economics of mini grids. Mike, to what extent do you see Chinese inverters compete with European ones? We've also seen a lot of development on that end, and especially price reductions compared to the traditional inverters that you receive from Europe, like the SMAs, like the Studers, like the Victrons and so on. Do you want to tell us more about that?

Speaker 4: 8:37

Yes, when we're looking at the Chinese inverters, currently they have approached like a tier system, where they have brands like Huawei, sunroof and Solis, which have heavily invested in R&D and offer inverters that, comparatively, are at par with European brands in terms of quality and performance, and also price. So when? You're looking at prices for the Chinese inverters. We are looking at price difference of 15 to 25% cost advantage compared to the European brands, which is mainly to the access to the lower labor and material costs, which is a significant factor for budget cautious projects.

Speaker 2: 9:17

So when you're looking at the Chinese inverters, and then you said there is a different, a second tier and maybe a third one. Do you have some insight on what's happening on that end?

Speaker 4: 9:27

So the second tier brands typically offer the cheaper inverters but often compromise on quality and lifespan. Also, warranty is limited, like, for instance, here in Kenya we have brands like Mass, which normally typically the lifespan you find that is between two to five years. However, the warranty also you're given like one year warranty, compared to Huawei and Sunroof brands, which can give five years and also offer extended warranty options up to 10 years, like SMA, for example, or which range yes up to 10 years, and these inverters that you're talking about are battery inverters that are single phase and can be combined into three phase clusters.

Speaker 4: 10:11

Yes, the typically single phase inverters clustered into the phase.

Speaker 2: 10:17

I see. Do you have information about what the maximum system sizes are? These, like maximum system sizes, have in the past been kind of a challenge for some of the European brands For Studer, partly also for Victron, for example, compared to SMA. Have the Chinese vendors overcome that challenge?

Speaker 4: 10:39

On maximum system size, like for tier one brands like Huawei, sunroof like recently I saw Sunroof have grid inverters of up to 350 kilohertz. Yes, so on size, I would say a bit more flexible compared to the European brands because, like, for instance, most recently when I look at Schneider Electric brand, like here in the Kenyan market, we're looking at a maximum size of 50 kilowatts.

Speaker 2: 11:08

I see like clusters of 50 kilowatts or one individual inverter.

Speaker 4: 11:13

Yeah.

Speaker 2: 11:14

Clusters. Yeah, okay, good, and another challenge that I have faced in the past with some inverter brands was that these inverters could just connect to one large battery bank. Even if you form clusters, they all need to be connected to one large battery bank, and this of course makes it difficult to expand the system in stages. It's easier if each cluster has got its own battery bank and then you can add another cluster to add additional capacity. In the other option that I explained before, you would need to add new battery strings to an existing old, already cycled battery bank, which is very tricky not only with the old lead asset technology but also with the newer lithium ion technology. What is your take on this, dipto and Mike?

Speaker 3: 12:11

Yeah, some years back, the most common design for mini grids where this cluster based like, as you mentioned, niko, that they're using single phase inverters and making this like three of them and making this like a three phase system and each cluster would be tied to one or two battery strings and that's difficult to scale. But nowadays we are seeing that the mini grid developers are shifting from this idea and are rather using larger three phase inverters, as Mike mentioned, in the range of some 30 50 kilowatt.

Speaker 2: 12:43

And that is an individual inverters right and that range that are still modular and can be stacked exactly to 100, 150, 200, 250 kVA and so on.

Speaker 3: 12:54

Exactly, and with the LFP batteries. Basically each of these inverter comes with a stack and then we can basically do it like a vertical stacking of them. So that's very modular in design and can be done quite easily, compared to what previously was done for latest in batteries with battery strings and this complicated idea.

Speaker 2: 13:15

And these 50 kVA modules? At what price point do they come Roughly, indicatively, I've got information that they are not far from ten thousand dollars per piece, like a little bit above maybe, but if you compare to single phase inverters and clusters of those, this is definitely significantly cheaper. All right, and these 50 kVA modules? What battery voltage level do they usually support? When you look at these small single phase inverters 5 kVA, 6 kVA, 8 kVA they usually come with 48 volt low voltage. My understanding is that this is not being supported by these larger batteries. The currents would be simply too high, right?

Speaker 3: 13:58

Yeah, as Mike also mentioned, that nowadays LFP batteries are the preferred battery technology over LSD batteries for mini grids. When LSD batteries are mostly used, the battery string level voltage was like 48 volt on the DC side. Normally they would 24 numbers of 2 volt. Latest batteries were tied in series to make one large or long battery string and then you would need these such kind of strings in parallel to then attach to this cluster of battery inverters that we just discussed. Of course this 48 volt systems have a higher current carrying capacity requirement because it's low voltage. But you could also argue that these are comparatively safer compared to its high voltage applications. So these were the norm at some point that was used.

Speaker 3: 14:46

But nowadays we see that some 6 kilowatt hour, 8 kilowatt hour, is used on a voltage level. With LFE batteries that's more than 100. So that's already not less than 100 volt DC but more than that. And as you stack up this battery modules your DC level voltage also increases and this goes in the range of 400 to 600 volt maximum. And this has certain benefits to do it in this way.

Speaker 3: 15:15

Definitely it's able to charge and discharge faster than a 48 volt system and in mini grids where you need power for a like high power for a very short amount of time. This high voltage system is definitely better compared to a 48 volt systems. We are discussing about really industrialization and productive use of energy in mini grids and so on, and, yeah, compared to 48 volt systems, higher voltage systems are definitely better in that respect. Also, there is this issue of space utilization. So this kind of low voltage systems or 48 volt systems with latest batteries that we have discussed, of course needs some space, because there are strings of batteries that are laid on the floor, like horizontally, and now with the LFE batteries, these are more vertically integrated or stacked up on top of one another and, yeah, much more space efficient, so to speak.

Speaker 2: 16:17

What may be interesting is the auxiliary components for higher voltage, switch gears and so on, and lower currents are usually cheaper than those for lower voltage and higher currents, isn't it?

Speaker 3: 16:35

Exactly so, then this also has an impact on the cost side as well.

Speaker 2: 16:39

Yeah, but you already said that high voltage is also kind of dangerous. What precautions do you need to take care of when deploying a high voltage system?

Speaker 3: 16:49

Definitely, if you are discussing about the 400 to 600 volt range, it needs some additional safety measures nowadays. So continuous systems that come are already integrated with its own necessary safety protections. Of course, the battery management system should be integrated as we are discussing about the LFE batteries. There is also this discussion around using anti-static mat to provide like a discharge, but for the self discharge current right now, the high voltage system providers. They also ask that you do not keep the DC system on a floating level, rather you should ground it. So, for example, 48 volt system, you could keep it floating, but with high voltage LFE batteries manufacturers recommend them. They are properly grounded to make the operation safer.

Speaker 2: 17:42

They are well understood, which after all means that these very, very low cost cells, as we said, like sometimes you can get cells in the $30 per kilo watt hour range when it comes to LFP and especially when it comes to sodium ion battery technology, those cells are then available. If I understand you correctly, dipto, you wouldn't recommend minigrid companies to go and build their own 400 to 600 volt system from that, because that requires some additional expertise and safety precautions and so on.

Speaker 3: 18:17

Yeah, definitely not. The minigrid developers should do it by themselves. For LFE battery systems this was different because it was more like the approach of doing it yourself, that you install and designed by yourself. But with LFE batteries there are certain risks that we all are aware of, For example, the thermal Thermal runaway yeah.

Speaker 3: 18:36

Yeah, so this kind of issues to avoid that, and even thermal propagation, so it can easily get ignited from one cell to the another. There are some IC standards, for example the 62619, which clearly outlines that what test lithium battery should go through before it's used. So there are strict requirements on their operating range and how the battery management system should work. So therefore, I would not recommend that system integrators, particularly minigrid operators, do it by themselves, rather than take more integrated and commercially available systems that are in the market with the correct certifications and all the safety precautions.

Speaker 2: 19:20

And that is what our subsidiary, volta Labs, also based out of Germany, has special expertise in. Volta Labs tests lithium ion battery systems for the European car industry, and it's a very interesting business basically shoot iron blocks into the battery and into the cell and then film with high speed cameras and infrared cameras how the battery burns or explodes and then help the car manufacturers to improve the safety. So, yeah, well, therefore, we know quite a bit about lithium ion batteries here in the Nensos, and lithium ion is definitely a field that will develop further, and it's good to say that LFP batteries are already much safer than NMC batteries, but the sodium ion battery that is coming in the near future will even be safer like those. Things like propagation are very unlikely to happen in sodium ion and therefore they are really well suited for the minigrid sector with a relatively low risk of explosion and fire. Very interesting. So after all, if I understand you correctly, mike and Diptole, minigrid operators have the choice of either going for extremely low cost batteries getting the cells, getting a battery management which you can buy off the shelf for some $60 or so for a battery system from China or from Europe or the United States, and then basically set up everything themselves, but then only on the 48 volt level, and there they can either go for the more expensive single phase inverters and build clusters. That has got the advantage that even in the small ranges these clusters are quite modular and you can add clusters and you can reduce clusters with their respective battery strings, or alternatively they go for the self made kind of battery system plus a low cost Chinese inverter, which usually depends heavily on fans for cooling and is very sensitive to dust.

Speaker 2: 21:39

And, mike, you already said that your usual lifetime and in Kenya is what did you say? Two to five years or so, and that is probably mainly due to the dust that settles on the power electronic parts and due to the inverters that cool these parts. And yeah, and then the third option would be to go for a low cost large scale inverter like the one from a BB, the 50 kva modules that, for example, husq Power is using, and then go for a professional battery system that is manufactured, but then the battery system does not come at $100 per kilowatt hour anymore, but probably a multiple of that, because all the safety related aspects and all the certifications and all the testing and everything has to also be paid for. Yeah, so a difficult decision to take. What do you think, mike and Dipto? In which setups, in which countries, which developers should use what technology?

Speaker 4: 22:49

Okay, for the, for the technologies, especially for the inverters. I would say the bigger 50 kilowatt inverters makes more financial sense compared to smaller single face. However, there is a downside when you're using the single face inverters, however smaller, at the end of the day the balance of system and also the components in the overall system design is more compared to 50 kva inverter, thank you. So one downside I can say for the bigger Inverters, like the 50k VA in modules, in case for the inverter does happen to fail, the impact will be bigger in the 50k VA compared to one of the single-phase inverters.

Speaker 2: 23:37

That's true indeed, diptoe. What do you think? Which technology is best suited for whom?

Speaker 3: 23:42

for the large mini crit Developers who are working in multiple countries are really looking for scale.

Speaker 3: 23:50

They can really take the benefit of this standardized or large capacity Inverters, so then all integrated with the battery management systems and high voltage Lithium and batteries so they can really help on Reducing the price or keeping the price because, as we have discussed, it's still a multitude of a hundred dollar per kilowatt hour.

Speaker 3: 24:10

But if we are talking about scale for large Developers, they can definitely go on this route. So I will say very small developers who are just starting out and have not got the capacity or the skill that's needed to do this kind of clustering technology by themselves, they can also go for this Standardized, like from a VB or from Hawaii, that this kind of plug-and-play kind of systems that comes with inverters, with batteries and everything, this kind of systems. But those who are in the medium scale, that are not on the same level as, for example, power chain and others, they can go for building it by themselves. That would probably be most cost efficient for them and also provides them the flexibility to work with them. But I will say the other two extremes, like very small or Upcoming developers who are yet to get some experience, and the large ones. They should go for more standardized works.

Speaker 2: 25:05

Hmm, yeah, after all, probably each company has to find its own way forward. There are pros and cons for each option and, yeah, well, I wouldn't say large or small, or large pipeline or not, and it's difficult to give clear recommendations for sure. But probably there are mini grid sizes. Definitely, if you just have enough demand for a 20 KVA mini grid, you wouldn't go for a 50 KVA inverter. Most probably, if you need to be a Modular in the small ranges, you probably rather go for the single phase inverters and clusters rather than the large modules.

Speaker 2: 25:46

At the same time, if you have very rural settings, you may not want to use higher voltages and run the risk of the higher voltages. And especially, like, if you handle higher voltage, you probably also need more trained Personnel. And if there's more trained and more costly personnel that needs to travel there's also incur some cost for high voltage systems. But after all, yeah, there are so many Aspects to consider on the operation side, on the capex side, on the OPEC side and so on, that each company probably needs to take its own decision. All right, let's look downstream in a mini grid system, which is the distribution network, dipta. How are distribution networks designed? There are low voltage, their medium voltage? When do you deploy what type of distribution network?

Speaker 3: 26:39

Thanks, nico. Usually in a low voltage scenario the electricity is then distributed in a radial structure. That means with a low voltage distribution network it's a certain distance that we can go without Significant voltage drop. So if the households and all the customers are quite close together, then we can definitely go for a low voltage scenario. But if you are talking about a five kilometer radius, where the customers are Dispressed and not so close by themselves, then in that case low voltage network may not work. Or In the case where there are two villages nearby, but let's say distance by some two kilometer, then also you have this limitation from low voltage just to have one Generation plant and then distribute the electricity to those two villages, mm-hmm in Africa and Asia and developing countries in general, you don't build the distribution networks with a lot of redundancy in.

Speaker 2: 27:45

In Europe, for example, we have a lot of mashed or ring type of networks when if one string fails, then the other string can take over the entire supply to the end customers. This also helps in maintenance. You can, for example, switch off a certain part of the distribution network without disconnecting the end customers from the network. This, however, also means that both lines, both angles of the ring need to have the full capacity to Transfer the full amount of electricity to the customers, which means way too much capex for a developing country. You want to keep costs down and therefore, as you said diptoe, in developing countries, africa and so on, you rather builds array type of Networks. Can you explain diptoe quickly how we at the Nensos design these types of Distribution networks, what we recommend, like what line distance from the power station the productive user should have, like the maximum line distance, what maximum line distance commercial users should have with their respective power consumption, and to what extent we allow how is all residential customers to be connected to the low voltage network?

Speaker 3: 29:12

Yeah, sure, normally when we are designing low voltage networks from the generation side, particularly in a radial manner, the idea is that we want to keep all the connections within one kilometer radius.

Speaker 3: 29:26

It can go to 1.5 kilometer radius as well, but then of course we will have to increase the conductor cross-section just to see if we are within the voltage drop level.

Speaker 3: 29:37

But we definitely recommend that the productive use and all the commercial uses are as close, maybe less than 500 meters from the generation plant. Sometimes it's not possible, and then we have to find a way that how we can use like larger cross-section cables to reduce the voltage drops. Just to provide an example, like sometimes in different localities, for example in Namibia, the distribution side, we can go up to 6% voltage drop from the nominal level of 230 volt, and until the service connection that means at the household level or the customer level, it can be another 2%. So in total, 8% voltage drop is allowed from this nominal value of 230 volt, and then all the design criteria then are around that that we adhere to this particular requirements, either in a local jurisdiction or in as per the IEC standards. So then that's the idea that we take and what kind of loads there are. If it's an productive users with three phase line, the voltage level is much more sensitive compared to a household level one, and they should be as close as to the generation plant as possible.

Speaker 2: 30:55

Yeah, we usually say, within a range of 500 meters or even less, we would connect productive users like mills or welding workshops or so, and then up to something like 1000 meters we allow commercial users to be connected, and above that it can only be residential customers. Right, Just as a rule of thumb. The problem is that in many cases villages extend to beyond this 1000 to 1500 meter radius. And what do you do then? You can only go for medium voltage. How does that look like? You need to build another network right on top of the existing low voltage one, isn't it?

Speaker 3: 31:35

Exactly. Of course, for the medium voltage network we need additional assets, like you need to first step up the voltage level from 400 volt to a higher level, like, let's say, 11 kV, and then distribute the systems there. So you need additional line as well. What we do is to use the higher pole and then use the same pole, just to take also the low voltage in a different height, of course, with the safe distance. That's one way we do it, but definitely needs longer poles 9 meter, 11 meter poles and then we distribute to the next level.

Speaker 2: 32:15

What type of material is being used here? It's like poles for overhead lines its lines. Have you ever seen any underground networks in mini grids?

Speaker 3: 32:26

I have not come across underground lines myself for mini grids, but it's mostly overhead lines, as I said, with some 13 meter poles.

Speaker 2: 32:34

In 2009, 2010,. We built some underground lines in our first mini grids in Senegal because the ground was very sandy and poles didn't last for long and they bend all the time. So it was better to build underground lines and digging was quite easy in sand. So that was one example. I'm aware of all the other mini grids I've seen all have overhead lines according to standards. What types of poles are being used in Africa? Dipto.

Speaker 3: 33:05

Right now we see that wooden poles are quite used at this point of time. So also steel poles are used as well as the concrete poles, but not so much. It's mostly a choice between cost. Just to give an example, like a 9 meter wooden pole would cost you some hundred dollars, whereas steel poles of the same height will cost you twice as much, almost.

Speaker 2: 33:32

Depending on the international steel prices right.

Speaker 3: 33:35

Of course it's also varies and if good quality like treated wooden poles are available, it's definitely the most cost efficient option for the market.

Speaker 2: 33:45

Which is not the case everywhere. Right In East Africa there are a lot of pole manufacturers in country. Basically in Tanzania we have some, for example, in Sierra Leone. There were none. Wooden poles would have needed to be imported, and in that case it was cheaper and lower cost to actually bring steel poles or concrete poles into the country.

Speaker 3: 34:09

Exactly. Some companies also tried with bamboo poles at some point to reduce the cost, but I think, considering the sustainability of these poles and more stringent requirements from governments, and ultimately people are now switching to wooden poles and, as you said, in some cases steel poles and concrete poles.

Speaker 2: 34:27

Have you experienced any price trends in steel poles, wooden poles, concrete poles over time?

Speaker 3: 34:34

Over time. I think their prices are to some extent constant. We have not seen like a significant price change in the market over time of this kind of poles. The conductor prices of course slightly depend on whatever you are using it's an aluminum or copper but the prices of the poles remain more or less concrete. On the other hand, concrete poles themselves like and they would if they're locally built they would depend on the energy prices. But other than that I have seen, for example in Namibia I can give an example like 9 meter wooden pole costs something around like 85, 90 dollars, whereas over the last three, four years we have seen similar in other countries like Kenya, sierra Leone as well. So it's in that range.

Speaker 2: 35:25

All right, and we've discussed medium voltage and there are some mini grids with medium voltage, but according to our experience and Jumeima, our subsidiary in Tanzania, is operating medium voltage networks for some eight years or so now and we've seen that this really incurs a lot of op-ex. The transformers break, need to be repaired, oil needs to be exchanged, lightning strikes into the medium voltage network, always cause a lot of harm and destroy networks, cut off certain parts of the consumers for a longer period of time until everything's repaired. So therefore, there is now a new technology that's coming up, which is a technology of basically distributing the generation asset across the network and letting these inverter solar battery nodes communicate with each other. How does that exactly work?

Speaker 3: 36:26

Yeah, right now, as you said, nikot. So this kind of mesh network is coming up and the idea is that these nodes will provide support to the rather large, like a central plant, and these will help in many ways. We have discussed the voltage drop issue and it also provides ways to keep on providing the electricity to the customers, if needed, isolating them and providing them, but also reducing the conductor cross sections that needed and also the need to upgrade to a medium voltage line potentially. So that's what we have discussed, that with medium voltage line we always have just different issues, but reducing the cross section and also adhering to the voltage drop regulations. So then this kind of mesh network can provide like a way to do that.

Speaker 2: 37:20

Yeah, and that brings us back to our earlier discussion. If you go for smaller individual nodes, you may then use the lower cost 48 volts self made lithium ion battery systems and connect them to one cluster of battery inverter and then AC couple the PV generation plant of whatever 510 kilowatt peak or so, or even 20 as a node, and this significantly reduces, as you said, the cross section of the aluminum cable, the conductors, and that significantly again reduces the overall capex for the distribution network. This mesh network, or some call it swarm approach, is really a very nice idea that is currently being developed. Bottom up is more or less coming from the solar home systems range and the idea was to interconnect solar home systems initially. Then AC mesh networks were developed and now I understand that the nodes that are being offered are getting larger and larger. The interesting part would be to investigate more on which inverters can be used for this mesh network, but I guess that is what we'll have a separate episode on and discuss the new mesh network approach in more detail.

Speaker 3: 38:44

Yeah, I also actually did my master thesis on this, on doing this mesh networks, and we have tested this and simulated different options. But I also agree that there are now more companies that are looking into this and initially this way like as you said, in solar home systems was done. So these were kept with them, like less than 100 volt DC level, so it could be distributed quite easily. Doesn't need that many quality poles with stringent requirements with AC we will need that. But this is a very interesting topic and we should look forward to that how this will change the distribution landscape.

Speaker 2: 39:25

Yeah, indeed. All right, let's move one step further downstream and talk about the meter market. Mike, we've seen some movement on the meter market over recent years, from classic postpaid meters to prepaid meters and then to smart meters. Do you want to talk more about that?

Speaker 4: 39:46

Yes, sure so. Meters in the context of many grids, crucial for measuring the electricity consumption and facilitating the actual billing of power generated. So, as Nico mentioned, we have the conventional, which are the postpaid and the prepaid meters, and smart meters. When choosing the meter, it's influenced by several factors the payment model the mini grid developer chooses, the visibility on the energy consumption of the data and the cost.

Speaker 4: 40:16

Smart meters currently are favored due to the ability to measure and communicate payments and consumption data remotely.

Speaker 4: 40:24

However, it's typically more expensive compared to the conventional meters and dependent on the reliable mobile network connectivity. So cost on the meters for mini grids varies based on the type and the capabilities, as mentioned, that's for the postpaid and prepaid models are generally less expensive and simpler in design. Smart meters, on the other side, offer advanced features like remote monitoring, data communication and tend to be more costly. The choice between these depends mainly on the budget constraints of the developer and also the desired features. So currently in the market we have park meters, steam echo and curling meters which, on the cost side, curling is the cheapest among the three, around $22, and the steam echo is around $50, so that's something dollars on average and the cost of the meters expected to lower because steam, echo and spark meter currently are shifting to Chinese large manufacturing hubs. In essence, we have been working with meter since 2015. So we have the micro power manager platform which can be able to integrate with the three meter models mentioned.

Speaker 2: 41:47

And even more. We are integrated with many more meters indeed.

Speaker 4: 41:52

Yes, and when you're looking at reliability of the saved meters in the mini grid system depend on the type and environment in which they operate. Conventional meters, being simpler in design, often have fewer failure points and generally reliable in various conditions. However, they lack advanced features and require manual routine and maintenance. Smart meters, offering remote monitoring and management, can enhance operational efficiency. Still, their reliability is contingent on factors like stability of the internet or the mobile network connections and susceptibility to technical issues. We can generally say the choice between the conventional meters and smart meters will depend on a balance between the desired features and the practical aspects of the particular operating environment.

Speaker 2: 42:45

And when it comes to smart meters, some 10 years, eight years, five years back maybe even the main issue was the last my connection through Lora networks, through mesh networks, zigbee and other type of proprietary networks. That has not worked very reliably, but I understand that this challenge has now been overcome and now smart meters are connected quite reliably, even if they are a little bit dispersed and distributed within a relatively large village. So that is not an issue anymore. Other than that, of course, there's a big price range between on the lowest, and prepaid. Meters for $15 can be purchased from China single phase, whereas the smart meters are more in the range of $50. And as the average revenue per customer in a mini grid is usually somewhere at maybe four to six dollars, of course it takes a lot of time to amortize a $50 meter.

Speaker 2: 43:46

Anyway, there are big advantages, as you already said, mike. Last but not least, there are different types of powerhouses. Some companies really count on solar containers that can be shipped, readily equipped, readily installed, more or less, to put them onto a foundation, and then you connect and off you go. Others say, well, this too expensive. I am cheaper if I build my powerhouses myself, and then we can also involve the local population, and even others go for prefab structures where you basically ship complete walls and then you mount the walls on a ready made foundation on site. What is your experience here, dr, and what works where?

Speaker 3: 44:30

Yeah, nowadays we see that more and more developers are using this kind of solar containers. We have mentioned this kind of integrated system designs with LFP batteries and so on, so this is getting more common in Nigeria and Uganda also. Just to provide an idea to the audience, a 50, 60 kilowatt system can easily fit into a 20 feet container, whereas, like if we're talking about some 100, 120 kilowatt systems, this will need more spacious 40 feet containers. And these containers are standardized containers. So then, nothing special about them. Of course, this comes with a much cheaper price compared to other type of constructions. They can few thousand dollars, can. You can easily get them.

Speaker 3: 45:14

Yeah, and this is getting quite standardized in use and we see more and more developers using them, but also few developers, like PowerGen, for example, started with this containerized idea and now rather shifting towards locally building the powerhouses themselves, because with fully integrated container, we see this is very difficult to transport to the project sites.

Speaker 3: 45:40

Rather, building themselves bottom up is easier and, in some cases, cost effective as well, whereas for in NCEs, we actually did the opposite way. We actually started with building the powerhouses, but they're all like field realities that we see that it's difficult to get everything done. Sometimes the workmanship also suffers. But yeah, in my opinion there is no one size fits all solution here. It depends a lot on the field, reality, the road infrastructure, if everything is in order, or not. At some point in Sierra Leone we actually also tried this prefab shelters, as you mentioned, as powerhouses, but we also see from our experience that, after all, these were not very highly suitable for rainy areas and so on, particularly Sierra Leone. But yeah, there are three options. It's a matter of cost optimization but at the same time, how you will install it and what difficulty will face to do that, considering the sustainability of them later on.

Speaker 2: 46:41

Exactly Logistics. How do you offload the container on site? How heavy is the container when you offload it on site? Do you need a crane, as you said? On the islands in Lake Victoria where Jumeim installs their systems, bringing a complete container is not an option in most cases. The harbors are not made for that. In some cases, islands don't have harbors at all. There we had to definitely construct powerhouses with the local people, and that was also well received by the local community. That then could earn some money by providing labor to that construction site. Mike, we have not talked about DC systems like direct current systems. Are you aware of any direct current systems in larger mini grid these days?

Speaker 4: 47:27

Yes, we have the Okra mesh grid system that utilizes DC system and what we have seen is, as the system size gets smaller, you'll be able to witness more of the DC systems, and normally DC systems work best where there is a dense customer base and they utilize smaller systems like solar home systems and normally, using the solar home system, they use the peer to peer trading networks. For instance, the Okra network have access to 600 and 1200 watt DC power by plugging into the Okra pods and using this system. Okra provides power on an energy as a service basis. This system is normally designed to connect to form a network and ensure the developers who deploy these networks are incentivized to keep the network operating smoothly and sustainably. So when you're looking at, the main advantage of DC systems, I would say will be the cost per connection, which is usually around 120 USD, while you are looking at AC system can go for around 1200 USD per customer connection.

Speaker 2: 48:35

So, if I understand you correctly, mike, we may be seeing a revival of DC systems as the mini grid design is moving from a central generation to a decentralized generation setup with individual nodes which can be potentially designed at a lower cost if you select DC systems, smaller DC systems Interesting. Looking forward to see what the future brings, and that is my last question to you, dipto and Mike what do you think the future holds for mini grid technology? What do you think will the mini grid technology look like some five years from now?

Speaker 4: 49:15

I would say, for starters, when you're looking at the main component, that's the panel, battery and inverters, the efficiency of the solar panels on a yearly basis it's increasing and also the cost is continuing reducing. So I would say, on the components aspect will be more efficient equipment and lower costs. And also on the technology aspect, especially when you're talking about the Chinese and European brands. So due to the competition, we'll witness a specialization, especially because European brands might prioritize high-end residential and commercial application, while Chinese brand, which is currently more common, would focus on large-scale utility projects. And also look at the emerging markets, which are mainly focused on the cost aspect, due to the client base and I would say on meters, which the market is evolving rapidly, smart meters, particularly the solar specific ones, we'll see an increased adoption due to the advanced features and ability to support growth of renewable energy. All right, Dipto.

Speaker 2: 50:28

What is your take on this?

Speaker 3: 50:30

I will say that we get more and more standardized systems that can really work plug and play, and also what we in answer say, like running the systems on autopilot. I think we will move towards that. We will move towards in five years, hopefully, in like a fully diesel generator, free mini grids running mostly with LFP batteries and possibly, as we have discussed, sodium batteries will be its competitors. So the other way around it was before for legacy and LFP batteries and at the same time we will possibly see this kind of mesh networks so that we can connect more customers with low cost of connection. That's how I see it. But yeah, I think we are moving towards rather a cost effective solution and see where the limit is.

Speaker 2: 51:18

Cool Sounds good. Dipto sounds good. Mike. This sounds very promising, just like there are big cost reductions and efficiency gains ahead, and this will not only reduce the tariff, but it also will help rolling out mini grids across the continent, potentially reducing the grants required and convincing more governments to deploy mini grids operated by private sector across the continent, and I'm really looking forward to experiencing that in the coming years. And the Nensos will, of course, continue to play a strong role on the communication side and also on the design side. Thanks a lot, mike. Thanks a lot, dipto. This was great fun. Talk to you soon, thank you.

Speaker 3: 52:07

Yeah, thanks for having me.

Speaker 1: 52:09

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