Intro: 0:09

Solar mini-grids have turned from small pilots to an electrification wave. We were there when mini-grid regulation was established and when financial transactions were closed. We saw a 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 INENSUS.

Nico: 0:35

Hello everyone. This is Nico. Welcome to our episode on mini-grid financial modeling Today. My guests are Kellie Morungi, our new head of Africa business, and Diego Perez, our head of consulting services. Kellie has just joined INENSUS and brings a lot of experience and financing renewable energy projects in Africa into our team. Diego is active in the mini-grid space for many years already. Please introduce yourselves quickly, Kellie. Do you want to start? Thanks, Nico, and hello, everyone.

Kellie: 1:08

My name, as Nico has said, is Kellie Murungi. I'm excited to be part of Enensis, but also to join the podcast. Before joining the team, I have been working with East African Power. This is a Rwanda-based Pan-Africa clean energy developer. I was in charge of the project finance portfolio, developing about half a gigawatt of projects across the continent. My role at East African Power was managing the acquisition strategy, structuring the financing and raising finances for the projects. In terms of country footprint, I have experienced Uganda, rwanda, malawi, zimbabwe, cameroon and Madagascar. Now, before that, I was in corporate finance for five years and then real estate finance. But I must mention that, besides my professional experience in finance and clean energy, I have interest in agriculture, specifically coffee farming and also regular food farming. What excites me about joining Enensis is I see the intersection between clean energy access, food security and, of course, the role of agribusiness in driving economic development, and I look forward to being part of the team and helping ourselves and our people access affordable, clean electricity but also, more importantly, bring a level of economic empowerment. Thanks, nico.

Diego: 2:34

Hi everyone. My name is Diego Perez. I've been working in this field of mini grids and rural electrification for about the past 12 years. I've been specializing in mini grids economics and mini grids financing, working in different projects, mostly in sub-Saharan Africa. These projects have imploded, both public sector support supporting regulators, developing mini grid tariff tools, setting up mini grid tenders, designing mini grid grants for donors as well as private sector support, For example, delivering trainings, increasing private sector readiness to acquire finance, et cetera. So we're happy to be here discussing mini grids financial modeling.

Nico: 3:14

Thank you, Diego. And Diego, you have already contributed to one of our earlier episodes on price elasticity, and price elasticity will again play a little role today, so I'm really looking forward to discussing with the two of you financial modeling in the mini grid sector today, Kellie, we're talking about the specificities of financial models for the mini grid space, but let us get started with a more general overview of what a financial model is and does Kellie, what are full financial models or financial models in general good for?

Kellie: 3:52

Yeah, Nico. So for any business really, not just in the energy space, a financial model is a numerical representation of all aspects of the business. So we take the operational, the technical, we take sales, we take HR and we put it into numbers and then we ask what are these numbers then telling us about the business At a global level? A financial model tells us whether the business is viable or not, if it is a new business and if it's an existing business, then we're able to see the direction that the business is taking. If you make certain decisions Now from a financing perspective, a financier is looking at you know, if I put money in this business, is it going to come out, whether it's debt, whether it's equity, so it's ability to repay, ability to give returns to both the owner of the businesses but also the financiers of the business.

Kellie: 4:44

And then, just to add, you know, the other thing a model does is it tells us what the key levers or the drivers of the business are and how sensitive the profitability or the viability of the business is to those levers. Because, of course, going into a business, we know the costs, we know the revenues, but we don't know what are the most sensitive aspects of those towards influencing the returns of the business. So, in a nutshell, generally, that's what a financial model is supposed to do, and then, of course, you have specific applications, depending on industries, but also users.

Nico: 5:16

Yeah, All right, and what is the financial model, after all, used for? You already mentioned it's a very important tool when it comes to financing. It's a very important tool for managing companies, but you can also do other things with it.

Kellie: 5:33

Yes, absolutely. You know. You could look at it to determine where is the best place to locate your project, because that of course affects the implementation costs, it affects the operating costs, which ultimately affects the viability. It could help you in terms of planning your implementation, your operations, determining how much you're going to charge your customers, for example, whatever industry you're in, and then, of course, determining the structure of how you will finance the business from the entrepreneurial perspective. But then also, Nico, even from a government strategy sort of policy formulation perspective. Business models guide policy makers in determining that and key policies in terms of enabling certain industries or discouraging other industries. So, for example, if they're looking to discourage, say, diesel consumption or certain waste management processes, looking at how the business models in those sectors run will help them. Then see what levers can they pull. Are they looking to get rid of plastic bags, for example? They look at industries that use plastic bags and see what will be the effect of this policy on the industry, on employment and other macroeconomic factors.

Nico: 6:44

Right. So financial models all over. Whoever has not worked with financial models, this is a decisive instrument you should actually learn to handle. Now we know what financial models are being used for in general. What components does a financial model from a private sector perspective entail?

Kellie: 7:06

So for private sector really, it's what we call the three statements. So it's a profit and loss, that's the income statement, the statement of cash flows and then, of course, the balance sheet, or, as the accountants like to call it, the statement of financial position, and these are the key general outputs. They are generally prepared in accordance with accepted accounting principles. But then out of this then we pull out key indicators or KPIs that we'll talk about later in the podcast. But in a nutshell, the profit and loss or the income statement is that it outlines the revenues or the income that the business is getting, and then, of course, the expenses, and here we have the direct expenses, and then you have your gross profit, and this is the amount of revenue that's available to run the operations. Then you have your operation expenses and once you've subtracted the operational expenses, then you have what we call your earnings before interest, tax and depreciation, and this basically tells us how operationally efficient is this business and we'll then be able to service the interest and to meet the long-term cost of the assets. That's depreciation.

Kellie: 8:15

So I'd say that the income statement really is the heart of a financial model for our business. And then, when we get to the cash flow, then this is the air that the business breathes. We may make profit, but it doesn't always translate to cash, and that's why the cash statement then becomes important. We're able to see how much money we are receiving, or how much money the business is getting, and where that money is being spent. And then, finally, the balance sheet. I like to see the statement of what you have as a business and where it is coming from, so it's your assets and how you have financed those assets. Have you borrowed from the bank? Is it your own capital? And so on and so forth.

Nico: 8:52

Interesting, good, very nice description, and the financial model does not just look into one year. It looks across a wide range of years, depending on the duration of the project or the duration of the company that you're working on, right?

Kellie: 9:10

Absolutely. It could be as long as 30 years, where we are looking at a 30-year project finance model, or even as short as 10 years, where maybe a business is looking to raise equity and they know that the equity investor wants to exit in seven years, and so they would look at a 10-year financial model. And also we have models that are made on a rolling basis, what we call planning models, which keep rolling, maybe over a five-year period or a 48-month period, basically just depending on the dynamicism of the business.

Nico: 9:42

All right. So far we have talked about financial models for businesses in general. Now what is specific about mini-grid financial models? Because mini-grids are a very specific type of business with its own challenges and its own business model.

Kellie: 10:01

Yes.

Kellie: 10:02

So, nico, what's unique about mini-grid financial models compared to typical project financial models is that, like any rural electrification initiative, going into a mini-grid, we don't really know the demand until the mini-grid is up and running and we see how the demand plays out.

Kellie: 10:23

And so this means that, unlike an on-grid project whose cash flows are backed by a power purchase agreement and a fixed tariff, with a mini-grid we have that demand uncertainty, and it means that as we are going in designing the project and implementing the project, we have to know that the first one, two, three years then help us define what the mini-grid cash flows and profitability will look like. So, with this in mind, as we are modeling for mini-grids, we still want to direct long-term financing, because this is a long-term business, the demand takes time to evolve, and so we want to put in the kind of metrics that are aligned to the mini-grid infrastructure business. So we want to say that the first year we look at the PNL and then we use that to project the second and the third year, and we keep evolving as we go along. So I think that's the only element that's unique, but in all other aspects, in terms of the long-term nature of the business. It's very much similar to any other utility business and should be treated as such.

Nico: 11:27

Yeah, and that is also the difference between greenfield mini-grids and brownfield mini-grids. In some cases you have existing diesel mini-grids that you want to put on a 24-7 electricity supply, where you then add, for example, solar, photovoltaic panels and battery storage and then deliver 24-7. That is brownfield. In that case you already can assess roughly how electricity demand will be when you switch from whatever 12-hour supply to 24-hour supply. But there are also green field projects where you don't know at all how the demand will develop. Diego, do you want to come in here and maybe elaborate a little bit on how you, when you come into a green field project that does not have a mini-grid at all, assess what the demand may potentially look like and how accurate that projection can be?

Diego: 12:20

There are different methods to have a first understanding or an idea of what the demand might be.

Diego: 12:26

But, as you both are saying, this is really the most challenging part of financial modeling for a new mini-grids project.

Diego: 12:35

It's challenging because, as you said, you are making a large investment, you are committing to a 20-year time frame and you want to know how much of the service you will be offering will actually be demanded by the users.

Diego: 12:49

In terms of methods, private developers and also sometimes donors and governments typically conduct ability to pay, willingness to pay surveys asking respondents what they might be able to spend, how much they spend currently on electricity substitutes, and from there make some estimations. Our experience is that these surveys are helpful but still subject to a lot of uncertainty. Their option, which is actually supplementing the first one it's not replacing the first one can also be to look at recently electrified villages, at similar villages in the vicinity, where perhaps the grid has already re-arrived or where there are other mini-grids, and then we can already have some real demand data from which to draw some conclusions. But in any case, it's a really hard task. It's difficult to know how it will develop over time, how much business activity will there be in that community and that's, I would say, the key risk to be considered.

Nico: 13:52

Yeah, when I come into a new community as a mini-grid developer, and I'm talking from my own experience now in our mini-grids that we have set up over the years in Senegal, Tanzania and Uganda, usually you can get an idea of how phenomenically vibrant that village is. But after all, you cannot really assess if, after electrification, the economic cycle is significantly accelerated, and that usually depends on individual entrepreneurs that are either existing in the village even before electrification, in the best case or that are not existing in the village and are moving to the village after electrification. There may be cases when people from towns move back to their village because they suddenly see oh now, after electrification, there are great opportunities for certain product processing or for certain services that they would like to offer inside the community, and that individual entrepreneur can channel significant cash flows into the community, which changes the economic cycles inside the community significantly. If just one person builds up a larger company with, say, 20, 50 employees in one community and pays good salaries, then this has a significant impact on the community. If this one person does not come and does not have a good idea, or if that business of this one person fails, then this also has an impact on the economic cycles of the community and the economies of the community are strongly linked to the electricity demand. Wherever the income level is going up, people automatically purchase more electricity, not only for productive purposes, but also for consumptive purposes. And yeah, this is something that you cannot assess before you electrify. You only know a few years after you have electrified. So then there is another strong component that determines demand in the rural community before electrification, you don't know if people have enough money, have enough finance available to purchase advanced appliances for consumptive use of electricity.

Nico: 16:32

I'm talking about televisions, I'm talking about fridges, I'm talking about freezers, I'm talking about electric stoves in the best case, and these consume significantly more electricity than LED lights and mobile phone chargers. It's a multiple of that electricity that you need to light a home. And, yeah, where does the money for these appliances come from? It's usually, well, a mother having a son or a daughter in an urban area, and this daughter or son is having a good job and comes home for I don't know birthday or Christmas or whatever kind of celebrations and brings along a new television and everybody celebrates. Now we can watch TV. But this is usually a very strong source for appliances that, after all, increase the electricity consumption. Kellie, you are from a rural village in Kenya, right? Have you experienced these kinds of effects that I am just talking about?

Kellie: 17:44

Yes, absolutely, Nico. I mean, it's interesting the example you give of a son or a daughter bringing home electrical appliances when we just built a new home and we had not been grid connected for almost 10 years, and it's what you talk about. First it was do we put a solar panel on the roof? Because there wasn't a mini grid, and of course then my mother's house is the only one that had electricity and all the funds then would be charged there. And then with time we had a bulb, of course, a couple of bulbs in the home, and then a TV appeared and then a fridge appeared. So every time we would go home it was a matter of thinking how can we improve? And we'd call and say can the connection take these appliance? And if she says yes, then of course we do that. And so that is very real in African countries where often many children the mark of success is to leave the village and go to the city, and then once a year in December, you go home and whatever you bring has the potential to really change both the electricity consumption but also the economics, because you see, you keep advancing, you move from household and you start doing little micro-industrial sort of developments. For example, we cut the grass that the cows eat. You start with a manual chaff cutter that somebody turns physically and at some point the child from the city brings a mortar that can be connected to electricity so you can see the home slowly mechanizing. And as you write, you say that depends entirely on how much money is coming into the village from the city or even from outside the country.

Kellie: 19:23

Nico, if you allow me, I'll add another perspective, also, as government is a factor in determining how demand grows. We are seeing a lot more devolved governments in Africa, where economic development is being done a lot more at village level and really the leader of the village could change what the village looks like. Some leaders are very focused. They are setting up youth centers, they are channeling funding to if it's, agriculture, into processing, and you start seeing processing units coming up. In the northeastern of Kenya, for example, you see slaughterhouses, cold chains for animals and also in the milk value chain, the same thing, and so that could change an area. If it's an area that had some dairy production and now you have some cold chain and some processing brought by the government, of course, depending on how well it's done, it could then change and have this multiply effect in terms of driving demand for the many great.

Nico: 20:20

Sure, but these slaughterhouses and cool chains could not only be funded by government, but also by private sector, couldn't it?

Kellie: 20:29

Absolutely, there's opportunity there.

Nico: 20:33

Yeah, but indeed you never know if the government will install such a facility over the coming 15 or 20 years. So therefore, how would you possibly project the cash flow over a 20 year timeline? It's impossible, you cannot do that. So the financier would say, okay, well, I give you money. I need to amortize that investment over, or that financing over some 15 years or so. But how can I possibly project the demand of electricity over 15 years? It's not doable, is it?

Kellie: 21:10

It's it.

Diego: 21:11

I remember you, nico, making a comparison at some meeting, and the example was imagine someone that wants to set an egg selling business and he needs to order all the eggs that he's going to sell in 20 years time, the first day of operation, or even before starting. It's a bit like that.

Nico: 21:30

Exactly. Yeah, well, I keep saying that mini grids are like supermarkets, like supermarkets that sell food and the food goes bad after a few days or week, depending on the type of food. And if you don't sell the food then you don't make revenue, like you have the cost of purchasing the food and procuring it, but if you don't sell it you're left with a losses. The same is the case in the mini grid you produce your electricity, you have your cost, you have your amortization on your assets that you installed, you have your customer service cost and so on, and if you don't sell your electricity then you end up with losses. And therefore we're saying let's rather look into a shorter term horizon, like one year, two years, three years, and after that time you can more easily project the electricity demand over another one, two or three years, but probably not over 20 years. And therefore, all the KPIs Kellie, you mentioned some already, like the net present value, npv or the IRR internal rate of return they all depend on solid and accurate and reliable projections of the cash flow over 20, 25, 30 years, which is not possible in mini grids. So therefore, these KPIs in the mini grid space, they should not be used anymore because they are kind of false friends.

Nico: 22:57

You believe that this number may be correct, but it's probably not. There is a 90, 95% probability that this number is completely wrong. But what you can project is what may happen in the coming one year, or what will happen in the coming two years or three years, even, depending on how much track record you have with your village, and that is then reflected. Coming back to our financial models, this is then reflected in the profit and loss cases, because there you can look into individual years. Kellie, do you want to explain a little bit more why you can do this in profit and loss and why you cannot do that in cash flow projections?

Kellie: 23:41

Yeah, absolutely so. As you rightly say, we cannot do a 20-year cash flow waterfall. I mean, well, we can do it, but it wouldn't be meaningful for financiers. And so we want to give them information that speaks to the current viability or what the mini grid looks like for the first year, then maybe the first three years. We keep that on a rolling basis and hoping that by that time there's enough data to push the projection up five years, seven years.

Kellie: 24:11

And the reason why the income statement, or the P&L, becomes very important is what I hinted at the earnings before interest, tax and depreciation. That tells us is this mini grid operationally efficient? Are there levers in the operations of the mini grid we need to pull to make it a bit more efficient? And then, what do we need to do in terms of growing the revenue? I don't know if we'll have time to talk about the tariff dance, looking at the revenue requirement versus the kilowatt hours that can be sold, but the P&L basically allows us to see the effect of that and also establish a very important metric, which we call the break-even point, the point at which we are able to cover the costs of the mini grid. But also, I mean just to add, and it's not to say that the cash flow is entirely not useful. It's just that it tends to then become more of an operational tool for the mini grid. We are running a 12 or 24-month cash flow projection and we're using that to see what's the cash conversion cycle, what's the debt service cover ratio.

Kellie: 25:10

Again, here, from a financing perspective, a financier should then come and say the 10, 15-year debt service cover ratio should be 1.3. Well, we'll produce that, but it doesn't mean much. They should look at what does the short term look like? And then, when they look at the balance sheet, I think what's important there is the asset turnover, the leverage rates and then sometimes also the working capital ratios, because you're able to maybe negotiate your payables.

Kellie: 25:35

But also, as much as we tend to assume that the mini grid business is cash-based in terms of receivables, we are talking about enhancing mini grids, including productive use business, other businesses. We are talking about other revenue streams that the mini grid could bring in, and it's important to make sure that the working capital for those also makes sense. So you're not accumulating a lot of productive use assets, for example, that have a pay-go model that's running over years, which of course increases the indigents and losses. So, looking at the three financial statements, I think each has its own use, but in terms of viability I think the P&L. The income statement then becomes very key.

Diego: 26:14

Just to add to the discussion. Prospective lenders or prospective investors do have this longer-term perspective. They need to recover their money, hopefully after eight, 10 years. So what I just wanted to add to the discussion is that, in order for those financial projections to be more accurate, for them to have a higher sense of actually being true, we need the right regulatory approach to mini grid tariff setting and mini grid revenues. And that's how you can mitigate some of those risks and increase the likeliness that those cash flows indeed materialize, because there is the right regulatory framework, perhaps also the right cross-subsidy framework in case this is needed, so that in the end, those lenders, those financiers, have some comfort, because you cannot tell them give me money for 10 years, but we will just look at the one-year projection. They need to see, okay, what happens next. And that's what the model itself cannot deliver. You need to have this government side of things, this revenue requirement, cost of service model, in place with the right mitigation mechanisms.

Nico: 27:21

Let's get back to that later, diego. I would first like to quickly go one step back to what Kellie just said, that the cash flow model in a full financial model has its role to play. I fully agree. Usually, what I see where the best value of the cash flow models is, if you really look at high granularity, monthly, maybe even weekly patterns for working capital planning, for example, and to make sure that the company never runs out of liquidity. That again is then something that you wouldn't do for 20 years or so, but, as you already said, Kellie, you would do that for maybe 12 months, maybe six months, maybe 24 months. That is definitely also something that helps mini-grid managers, the CEOs, the CFOs of the companies to plan properly and also helps financiers and investors to get a better idea of how the mini-grid is actually working.

Nico: 28:31

My experience is that financiers and investors often think of well, we are from the electricity business. What we know and what we have long-term experience in is independent power producer financing where they have large PV systems, tens or even hundreds of megawatts, or wind farms Kellie, you mentioned this subject briefly already. The great difference between those two, like the wind and the solar IPPs independent power producers on the one hand, and the mini-grids on the other, is exactly what we have been discussing so far. It's the demand risk. The demand risk is not existing, almost not existing, on the IPP side.

Nico: 29:20

As a wind farm, as a solar, photovoltaic, industrial scale grid connected system, you just feed All the kilowatt hours you produce into the main grid and the main grid absorbs all the kilowatt hours you produce. This is not the case in the mini grid. In the mini grids, you produce a megawatt hour over a day and sometimes you just sell out of this megawatt hour. You just sell 100 kilowatt hours. You lose 90% of what you produced.

Nico: 29:51

So that is the challenge that needs to be addressed here, and I think that there are many investors and financiers out there which still think that mini grids are just smaller IPPs, which is not the case. But coming back to this analogy of the supermarket, there are also investors and financiers financing and investing into supermarkets. So why shouldn't we find investors and financiers who invest in, finance mini grids? And yeah, I think we need a change in mindsets. And probably it's not the energy specialist that should actually take up the mini grid finance, but they should at least join hands with their colleagues from other departments which are more familiar with retail risk demand risk. Kellie, what do you think about this?

Kellie: 30:43

There is, as you rightly say, some certainty and I am putting air quotes as I'm saying this in the IPP, traditional project finance, because we usually say this is no recourse financing. The power purchase agreement is 20 years. We're giving 15 years, so we're even leaving some five years in case we have to restructure the loan, and we're saying we have a tariff that is fixed for the period. And so you find many financiers around the clean energy space are coming in with that in mind In the just energy transition. Really, they're not differentiating between on grid projects, off grid projects, mini grid and CNI even, and yet the structure is quite different. And so, in terms of who then should be looking?

Kellie: 31:26

Yes, of course, energy people, but also, as you rightly say, people who are more accustomed to analyzing retail demand, and maybe I mean even behavioral economists, because we're saying that we have no way to predict behavior. You know, once people receive power and, yes, every African village is different, but then I wonder, you know if there was some behavioral study of villages across Africa? You know, and they receive power, what happens typically, how much money is coming in and such. It would not, I guess, improve the financial modeling, but it would improve, I want to say our understanding of what we can expect going into a certain village. So to your demand analysts, I've added the economists. I hope they are welcome and I think we could use them in this case.

Nico: 32:12

All right, yeah, I fully agree. Now we've talked about the demand risk a lot, and the demand risk, for sure, is by far the largest risk by magnitudes, probably followed by other risk. But there are also other risk, like the political risk, the foreign exchange risk, the environmental risk, the technical risk. How can those be reflected in financial models, Kellie?

Kellie: 32:37

So traditionally you know a political risk for example foreign exchange risk those two risks specifically would be covered with some form of political risk insurance. But again, this is not viable for a mini grid. The project size is fast as small. And then for the other, you know environmental risk and then, on technical risk, you know construction quality of materials used. I think those two environmental and technical those are fully on the developer's plate.

Kellie: 33:05

On FX risk, what we are seeing, nico, that might be interesting, is that we're seeing some guarantee funds coming into guarantee local currency lending and my hope is that then we will start seeing more local currency money flowing into mini grids. But as it is now, you know, tools like swaps, which usually are good for big projects, may not work for mini grids. Political risk insurance, as it is now, it may not work. But I think MIGA, ati and others, they should really be thinking about this. You know, as we're talking about development funding and development support for mini grid operators, especially because, I mean, we all agree, as as an industry, that we cannot really electrify, you know, without decentralized electrification, which is where mini grids play.

Nico: 33:55

Yeah, coming back to the financial model, talking about technical risk, I guess this would mainly influence the CAPEX and the OPEX. Going into the financial model, higher CAPEX, higher OPEX would probably reduce the technical risk. On the political risk and the foreign exchange risk side, probably even the environmental risk side, you can run sensitivity analysis. You could say okay, what happens if we suddenly need to reduce the tariff by 50% or so due to political interference? You could probably say what if we lose one mini grid for a few months because there is political unrest in a certain geographic area or so? And this is how you would probably reflect these risks in financial models, wouldn't you? And then, of course, as you said, the insurances come at a cost, they come at a premium, and that would then also add to the operational expenditures the ongoing expenditures that need to be considered.

Kellie: 34:56

But I think you've brought in the portfolio aspect, where the model is not being done for one project, but it's the entire portfolio. I think I would also want to add another way of managing political risk is, of course, determining the locations you know that an operator or a developer is going into and ensuring that there is some diversification. It could be by country, if they are big enough, or it could even be geographically within one country, to make sure that they understand the country very well and that they are well diversified for all their plans.

Nico: 35:27

Coming back to the demand risk, uh Diego, what solutions or approaches has the mini grid sector developed so far to overcome the demand risk?

Diego: 35:40

Cap on why this is so important. As we said, you have a fixed capacity typically, especially if it's renewal. You have a number of panels installed, you have a given storage capacity and therefore the profitability window, as we could call it, is rather narrow, right? So if you sell too little, you're not making enough revenue. If demand is too high, you cannot meet that demand. So therefore, the window where you can operate successfully is relatively narrow, and what the sector is exploring now is how this window can be expanded, either because we fill the gaps when demand from consumers is too low and we use this excess energy to make some additional revenue, or because we are able to span beyond what would have been our 100% capacity and serve this extra demand.

Diego: 36:34

So, for the first part, for those cases where demand is lower than expected and we have excess capacity, there are some solutions, such as using this excess electricity for specific applications, one of these being edge computing and bitcoin mining, which is something INENSUS has been testing in mini grids. Another option is, acknowledging that demand grows over time, to try to implement mini grids in a stage way. So start with lower capacity and then gradually build capacity as demand develops, and then also based on how demand develops and then, on the other side of things, if demand happens to grow faster or if there are some periods in which big consumption is higher than what you can meet with your renewable sources, there's still a very legitimate use, I would say, for diesel generators. Diesel generators can still play this role of matching those consumption peaks and ensuring, on the one hand, that people have better power availability and then that no revenues are lost because of the inability to meet those consumption peaks 

Nico: 37:59

whereas investors and financiers are strongly pushing for 100% renewable mini grids, reducing diesel generator operation and also deployment in general. And diesel generators are also very difficult to manage. From my experience like JUMEME is operating diesel generators in Enersa in Senegal has operated diesel generators we're seeing many mini grid companies really struggling with diesel generator operations. Diesel generators means you frequently, periodically, need to deliver diesel fuel to sites, and diesel fuel logistics in rural areas of Africa is not that easy and that also increases the cost of the diesel fuel. When arriving on site, just transporting drums on a truck, it's not that easy right. Then you have fuel theft. You need to store fuel on sites, and fuel can used for many different things and it disappears somehow at some point. Diesel is just used for other things in the community when other sources of diesel are not available. And there are also maintenance tasks for diesel generators that you need to take care of, like lubrication oil, exchange filter, exchange for the air, for the fuel, for the oil. All these kind of things need some special training.

Nico: 39:19

And the last aspect is, of course, the environmental aspect, where everything needs to be handled in a way that diesel fuel and also lubrication oil doesn't get into the ground and even in danger, for example, the freshwater resources of the community. And, of course, the other part of the environmental consideration here are the CO2 emissions, and we want to move away from CO2 emissions, not into CO2 emissions when we electrify rural areas in Africa. So, therefore, I don't think diesel generators will be deployed very often in the future. I believe that the mini-grid sector is better off with these two other opportunities or options that you just mentioned, Diego staged implementation and edge computing. But stage implementation sounds really like this is a good solution, but why is it not happening? It's not being used very widely, is it?

Diego: 40:24

no, it's not.

Diego: 40:26

I would say the main reason probably is grant design subsidies allocated by donors, which usually are available just as one of things, even if there's more than one payment milestone, but they typically have a short cycle, let's say one year to year, and then there's no certainty that they will be available later on.

Diego: 40:48

And then developers then to install everything they think they might need already in the first place, because then they can secure 50 percent, 60 percent, 70 percent of that CA PEX in form of grants, and I would say there is an opportunity for more flexible mini grid subsidy design, grant design that allows for this stage implementation. And then, of course, it's a bit linked to your description of the issues with diesel generators there are also some logistical issues. The cost of installing assets in remote locations is also high. So if you have to go there three times, you have to have containers come into the site three times instead of just once. So those are additional costs as well, but I do believe the benefits could outweigh the costs in this case and edge computing, after all, doesn't deliver a revenue as high as electricity sold to end customers, does it?

Diego: 41:45

no, it doesn't, and it has its own sets of risk. It's relying on what is being paid for this service, which is also something that can fluctuate very widely over time, but at least it ensures that the operator can get some us stands out of those kilowatt hours which would otherwise be lost exactly so.

Nico: 42:05

After all, we must say, there is no silver bullet to mitigate or avoid the demand risk completely, but I think a combination of instruments let it be staged implementation with an upgraded, improved grant scheme, edge computing that uses the surplus electricity that is still there even after the staged implementation approach being implemented, yeah, and then, of course, a clever fostering of productive use and maybe even rural industrialization. That we have already discussed or will discuss in the future in other episodes. Diego, in another episode we've also discussed price elasticity. How does price elasticity come in here, when we're considering demand risk, when we're considering tariffs, when we're considering demand of electricity in general?

Diego: 42:58

well, I would say it's a bit like your first question around financial modeling. Price elasticity is a general feature, it's like a generic economics concept that applies to any service or good and it's basically the relationship between the price and then the demand. When things become more expensive, people tend to consume less of that thing. And in the mini grid space it's particularly critical because, as we said, the capacity is rather limited and any changes in tariffs will have an impact on how people consume electricity, how much they consume, and therefore financial modeling needs to take into account this. The consumption, the demand in kilowatt hours, cannot be an input in the model because it will change when another input, which is the tariff, the price, changes. So we need to be very careful when modeling consumption, when modeling demand, and keeping always in mind this link.

Diego: 43:55

And even if Data is not yet very exhaustive in the mini grid space in terms of how different groups consume, we are having more and more understanding of this and we see, for example, that households Typically tend to have a fixed budget. They typically have five, ten dollars per month available for electricity, and that's how you should then reflect this in the financial model. You should then reflect this as a fixed revenue, not as a fixed consumption that generates more revenue when you increase the price. And we could also discuss about the other groups and we. There's this episode that you mentioned where we deal with this in more detail. But just to summarize, we just need to be aware of these interactions between the tariff level, the demand level and then, of course, the size of the mini grid. If we are discussing this before Installing the assets, then all this interplay also affects our sizing and our technical design.

Nico: 44:58

Yeah, which after all means that if you have gone through all the process of Evaluating your demand, going to site, asking the future electricity customers how much do you want to consume, and you got a good understanding of willingness, ability to pay, of the revenue that you would potentially generate in the future, and Then you do your technical design, you feed everything into one of these Sizing tools and then you come up with okay, I need that much photovoltaic, I need that much hydro power, I don't know. I need that much diesel generator power, that much battery Capacity. Convert it into CAPEX. Okay, all these assets cost me that much, and the Maintenance of these assets cost me that much, and the service to the customers cost me that much, and the head office Cost that much, and so on, and all of that goes into the financial model. And then, after all, you think, hmm, but my tariff is a little high, maybe I should reduce it by 10%, right?

Nico: 45:58

What happens is, basically, you then need to consider price elasticity and you have to do all the things that you have done again. So you may not need to go back to the village community, but you could apply these principles that you just introduced, diego, of the fixed budget for households, whereas productive users have rather a constant electricity demand and kilowatt hours per week or per month. And then you basically need to go into the technical design tool, size your system again. Probably, if you've decreased your tariff, then you need more a generation capacity to meet the demand, and then you put these new numbers into the financial model again and then you get new results. After all, you cannot work with Financial models without connecting directly the design tools. It's a little bit tedious, but that cycle needs to be maintained and you cannot work without Diego. What do you think?

Diego: 47:06

No, absolutely. There needs to be this constant feedback loop between both sides. I believe we will be talking later. But even also the regulatory aspect. It is pointless to prepare a financial model Without having an understanding of what tariff level might be acceptable, might be approved later on, because, as you said, you might be going in a completely wrong direction. So I think this also needs to be triangulated, so to speak, with the regulatory authority or with, and a good understanding of what tariff levels might be feasible after all, and then you can already do your projections based on that more realistic assumption to wrap up, the private sector perspective on financial modeling, everything that Kellie introduced profit and loss, cash flow models, balance sheet, Kellie.

Nico: 47:59

We said earlier that cash flow related KPIs are not really well suited for mini grids, and these are we already mentioned those the net present values, NPV, the internal rate of return let it be the equity rate of return, equity IRR or the project IRR, but also the levelized cost of electricity, LCOE All those are KPIs that are based on cash flow models that are Usually the main KPIs for the electricity sector. Everybody would ask for okay, what is your IRR, what is your LCOE? But for mini grids they don't count. So what KPIs are left like? What KPIs should we talk about in the mini grid sector then?

Kellie: 48:49

Nico. So in the mini grid sector really we are talking about profitability, so the EBITDA and earnings before interest tax and depreciation and amortization, you know we are talking about the debt service cover ratio that remains and the cash conversion cycle ratio and of course, the break even point is important, I believe, both for financing but also for operations planning and capacity planning.

Nico: 49:14

Which are other what probably need a long-term projection also, right?

Kellie: 49:19

Yes, we need to be able to know you know at what point do we break even, you know how many Connections or how much revenue, based on just the sort of demand analysis that has been done. Our equity and debt ratios unfortunately can't run away from those, because they determine how to structure the financing of a mini grid. It's accepted and that we would need some level of subsidization. So these ratios would help us see, you know, what is the extent of the subsidy that we need for the mini grids to make sense for the financiers, and then also what can they look at in terms of the returns and in how long do they get their money and what does that look like, given the limitations, of course, on the demand projection side, we did mention working capital ratios.

Kellie: 50:04

You know, payable days, receivable days, those are operational ratios, but also that matter to financiers because then they affect cash availability and, of course, asset turnover. This is just efficiency. You know, has the developer over invested in assets at the beginning and therefore, you know, not operating very efficiently. Is there a way to optimize that better, especially at group level? So I'd say those are the metrics from the financial model that financiers and private sector players would look at. And yeah, again with an acknowledgement that where we have long-term ratios, there is some nuance there. On the uncertainties around long-term projections for mini grids All right, good Diego.

Nico: 50:46

now let's go to the other side of financial modeling Opportunities, let's know, look into how governments use financial models for the mini grid sector, and we are specifically looking here at the cost of service model. Do you want to explain to our audience what a cost of service model is and how it works?

Diego: 51:11

Sure, nico. So, unlike the cash flow models that we were describing and the financial models from a private sector's perspective that we have been discussing, the focus in cost of service model is more the short term, and this is a tool typically used by regulators to assess what allowed revenues does a company a utility have? And Then try to ensure that the actual revenues generated match this revenue requirement or this cost of service. So what is typically included under cost of service is the depreciation of the assets that were privately financed, the return on those assets based on an agreed rate of return, and Then the operational costs for the period being considered, and you can also add taxes or other Specific expenses. And the idea is that under cost of service models these items are computed and we have a total revenue requirement for a given year and then from there Regulators derive tariffs to be applied based on the expected level of demand, based on the expected sales of electricity. So it's just a very simple operation. If you need to recover $1,000 and you are selling 5,000 kilowatt hours, then you need to sell each kilowatt hours at $20 cents. That's the basic core calculation and their cost of service models and, as I said, this privileges a more short-term perspective and it actually at least theoretically, at least from a conceptual perspective it actually eliminates or mitigates demand risk because we are considering the actual or the expected demand level and hopefully, as the project evolves, regulators keep using the latest demand figure, perhaps from the year before, and therefore it should actually be successful in yielding the expected revenue.

Diego: 53:09

However, there are some problems with that. On the one hand, link to the political acceptability of tariff levels. So, for example, if the exchange rate changed very dramatically over a period of one year, then your depreciation costs, which are in hard currency, might have doubled in local currency and same with your return. And that would, if you just apply the cost of service model, purely mathematically that would perhaps lead to a tariff twice the level of the year before, because you have to recover those additional costs and mitigate these foreign exchange rates. However, in many cases this creates some political or social issues and let alone affordability issues in the community. And the second aspect here to consider is price elasticity, because from what we discussed before we know that there is no linear relationship between tariffs and revenue and you cannot make up for a revenue shortage indefinitely by just increasing tariffs. So it's a complex exercise in the end and it has significant, but limited, not unlimited, amount of power or leverage to readjust revenues in the immediate space.

Nico: 54:22

Yeah, you just explained that there is little opportunity for regulators to actually support mini-grid companies which are falling short on profitability, like basically running losses through tariff adjustments. So what options do regulators have to save mini-grid companies from bankruptcy?

Diego: 54:44

I think in those cases where the gap persists, even when you have perhaps fine-tuned tariff levels so that an extra 10% revenue can be generated but we are talking about that kind of margin, not a 50% margin I think we need to really introduce some additional forms of support or risk mitigation in the equation in the way mini-grids are typically regulated, and I would see cross subsidies really play in a role here in plugging those gaps where they exist. So I would see this as a way to balance projects and then ensure that this revenue requirement is actually made or, if not at 100%, at a reasonable percentage by the developer, and that doesn't mean there won't be incentives for performing good work. There can be very different ways of setting up the right incentives so that it's not just free money for developers who didn't do their job. It can be really linked to their good performance.

Nico: 55:44

Yeah, and that is then performance-based regulation, which you see in Europe and most of the European countries, I believe, and that is what basically drives down tariffs in the long run. But I fully agree there must be some lever for the regulator to actually support companies which have been struck by, I don't know, demand decreases due to economic developments in a certain region or, as you already mentioned, foreign exchange fluctuations that cannot be covered by just adjusting the tariff. Cross subsidies may be one thing. Another aspect may be that we already mentioned staged implementation, or even shifting certain assets to other mini-grids, electrifying new mini-grids from assets of villages that do not need the assets anymore because demand is not growing to the level that was initially projected. These kind of things that should be also instruments that regulators should be able to use, and I see two options here. Like, cross subsidies would mean a connection to the national utilities and, after all, national utilities would need to transfer money through the regulator to the mini-grid sector, which is rather unlikely to happen, I would say. But what could definitely be done is structure grant funds in a different manner, so that they are more long-term and that you actually can implement staged implementation, so that you start with a small mini-grid that just covers the electricity demand for the first one or two years, and then in year three, you add another component of that same generation asset or even strengthen your distribution network a little bit to support the higher currents.

Nico: 57:43

And yes, in this regard, that is also a position that INENSUS is strongly promoting and really arguing for. We would like to see regulators being more strongly involved in grant level decision making and also get the opportunity to inject additional grant into an existing village where it's necessary to maintain the business, of course, at the same time making sure that the company is efficient and motivating, incentivizing the company to become even more efficient. But yeah, I think that is the future of mini-grid regulation, but this will take some years, maybe a decade or so, until this has been integrated into the existing regulatory frameworks. Kellie, some years back, not too long ago three, four, five years back people kept telling me yeah, but we are regulating IPPs through cash flow models. I want to do the same thing for mini-grids. Can you repeat to our audience once more why cash flow models cannot be used to regulate mini-grids?

Kellie: 58:57

Yeah, I'm sure happy to do so, Nico. I think the level of uncertainty is quite high, and so the 20 year, 25 year, 30 year outlook is not really informative, it's not useful in any way. And so even for policymakers you know this again their approach has to have a lot more nuance in determining how to regulate the sector, how to look at permitting tariffs, how to look at, say, metrics such as the return on equity, which factors into the, of course, the evaluation of the business. So, as they regulate and set these metrics, they cannot, of course, look at a 30 year document, because the business that they're regulating is more short term in terms of availability of data and also the level of dynamism. And so it's important for them to realize, you know, if they are coming, and typically you will find, you know, in some countries the rural electrification body maybe is stuffed differently, but mostly they will be coming from the main grid sector, where they're more used to this sort of tools.

Kellie: 1:00:03

I think there is need to educate and change the thoughts and the views insofar as to how to assess mini-grids. And just to add, I think one other thing that we, a financial model, a cash flow model, cannot tell us is the social benefit. You know we say in development finance it's beyond the monetary benefit. You know how do we value this, how do we factor this into the model you know. So it cannot only just be these are the cash flow projections for 20 years. We have to have a lot more nuance into the inputs that go into regulatory decisions.

Nico: 1:00:35

True, and when I have been working with the World Bank team some years back we also did kind of macroeconomic projections of how the economy would change if we electrified large parts of the rural population with mini-grids, and that is another way of running financial models from a government perspective which I would not want to go into too much today. Let us rather discuss Diego about whether the idea of the regulator influencing subsidies and grant funds for mini-grid projects, whether this idea is actually being considered anywhere already.

Diego: 1:01:18

I think it has been considered Well. I have seen some cases, some examples, where it was not considered in the first place and then stakeholders in the country became aware that next time they run a mini-grid tender or a mini-grid program they need to coordinate more closely. Because of exactly what you were saying. Typically, donor comes and offers some grants this might be channeled through the rural electrification agency and they run a call for proposals at the end there and then someone is selected with a financial model, as we have been discussing today, and with a tariff of X solar fence, and then the regulator only comes in at the latest stage, usually just to approve that. But usually problems happen at that stage. And the regulator? Well, they were not informed, they are not familiar, the tariff level might be higher than expected or they might have comments on the CAPEX numbers, on some ratios. So that's where countries start realizing OK, we should involve the regulator from the start, from the design phase of those programs, and really make them part of the subsidy design but also in those competitive processes, part of the process itself. Have the regulator in the evaluation committee, give them the chance to ask any questions and to solve any doubts they might have and really couple the grant funding competitive process with the tariff application process so that at the end there's much more certainty that things will happen as expected.

Diego: 1:02:55

That's the first part, which is related to when the project is awarded. The second part is finding the right way for regulators to have access to funds and then, over time as we were discussing before being able to plug those gaps that might happen in different parts of the mini-grid sector in a given country. I think it's a bit, if I may make an analogy in the IPP sector there is this concept of take or pay. There's like a take or pay clause. It would be the take or pay version for mini-grids, in which if there's enough demand, then fine. If there's not enough demand, someone has to pay for that, and that would be this cross-subsidy we were discussing. So it's a way to make sure that the revenue requirement that was approved and stated in the first place is actually happening.

Nico: 1:03:48

Well, after all, probably a cross-subsidy would not finance the profits but would help the minigrid company to survive. I guess that would be something that governments and regulators and minigrid companies could potentially agree on.

Diego: 1:04:04

In a more detailed way. It can be said and I have seen some discussions on this idea already as a percentage of the gap that needs to be covered, so that it does not fully cover the gap, and it also incentivizes the company to do better because they know they will not cover the full gap, but they will receive some support.

Nico: 1:04:22

Do you want to talk about the Liberia example that you're currently developing with the INENSUS team?

Diego: 1:04:28

Sure, in Liberia we are helping the World Bank develop a call for proposals for minigrids, with World Bank funding and in collaboration with the Liberia Relatification Authority and the regulator, and we are exactly trying to do this in a more integrated manner, as discussed before, with the regulator already being part of the team from the start, trying to also have this very strong link between the financial models that applicants need to submit and the tariff tool that will be used later on to approve the tariff. So really trying to ensure that there are no misunderstandings or no misalignments in the process, and then hopefully we can also convince the World Bank to assign some funds for a longer time perspective and perhaps pilot these kinds of mechanisms we are discussing today.

Nico: 1:05:19

Yeah, but not going into the cross-subsidies, right?

Diego: 1:05:22

No, not yet. 

Nico: 1:05:23

All right. Okay. Now, in a financial model may it be from a government's perspective, may it be from a private sector's perspective there are a lot of inputs, and some of these inputs are capital expenditures and operational expenditures. The private sector could just take this information, these numbers, from their own history or from other mini-grid companies that they talk to, that they share data with. But how do governments actually get a hold of this data to verify that a certain CAPEX or OPEX that a mini-grid company claims is true and valid? Sure?Diego: 1:06:08

So in various ways, I would say.

Diego: 1:06:08

So in various ways, I would say.

Diego: 1:06:10

On the one hand there is a growing database, a public database of studies and reports prepared by different actors, from AMDA, from the African mini-grid developers association to ESMAP, to UNIDO, to IRENA, etc.

Diego: 1:06:27

Discussing the economics of mini-grids and collecting real data from a variety of projects. We already have a starting point there and I would say an uncertainty is reducing in that sense. But still, beyond those studies, there are a lot of country-specific factors to be considered and I would say regulators and REAS would be very right to try to create national databases of projects where they really can track the efficiency of the different projects, of the different operators, can draw comparisons. Perhaps those comparisons can be explained and there might be legitimate reasons. Economists of scale play a big role in the mini-grid space, so it is logic that larger projects have cheaper costs on a per kilowatt peak or per kilowatt hour basis than smaller projects, but at least having the information allowing them to ask the relevant questions whenever they are needed. And also in connection with a discussion before around cost of service, what is the legitimate cost of service? What is due to operators' inefficiencies?

Nico: 1:07:34

So this can really help in that discussion and in Nigeria and soon also in Ethiopia, INENSUS had developed this online application tool. And this online application tool that the regulator runs and that collects the license applications from the mini-grid companies, that has a specific feature of recording all the CAPEX and OPEX numbers for projects for which terrors have been approved, and now that creates a curve over time which can be used as a benchmark for future projects. Of course, the regulator still needs to keep the market development into consideration. I take it into consideration, but at least it provides a solid starting point for the negotiations between the regulator and the private sector investor and operator about CAPEX and OPEX values. Let's wrap our session up for today and, as a last question, I would like to ask each of you how likely is it that financial models will be used in the mini-grid sector in the future, considering the aspects discussed today?

Kellie: 1:08:52

I don't think we can run away from financial modeling. It's a key tool, as we've discussed. It's not a perfect tool in many ways, especially in the context of project finance transactions. What I think is likely to happen is, you know, we will get better at modeling, we will get better at predicting certain things as we get market experience, as regulators, get better at policy and in such, but ultimately, no, there is no running away from financial models, a tool for business viability analysis, which is key, operational planning, but also just financial structuring of transactions.

Diego: 1:09:32

I think that there is an opportunity for financial modeling to be more aware of sector-specific characteristics, as we discussed today, and also be humble when it needs to be humble in terms of what can be useful, what can be relied upon or not. As we said, those long-term projections without any further context or mitigation strategies can be misleading. So, yeah, indeed, more attention to specific things such as price elasticity of demand, such as economies of scale on the supply side, on the cost side, and then, I would say, a closer link between financial modeling and the regulatory context that applies in a given country. So you need to take into account how your tariff will be regulated, whether you will have the chance to get it reviewed or not, whether regulator will have access to additional subsidies, in case those exist, and then reflect those in your financial model and try to be as closely coordinated as possible with that key element of your business because in the end, it's a regulated business.

Nico: 1:10:43

Alright, thanks a lot. Thanks a lot, Kellie. Thanks a lot, Diego. This was very insightful. Thanks for your contributions and talk to you soon. Thank you.

Speaker 1: 1:10:52

Thank you. This episode of the mini-grid business has been brought to you by INENSUS, your one-stop shop for sustainable mini-grids. 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. The mini-grid business powered by INENSUS.