0:00

Greetings everyone, this is

0:04

Volts for June

0:06

5, 2024. Making

0:08

mini-grids work in

0:10

Sub-Saharan Africa.

0:23

I'm your host, David Roberts. Hundreds

0:27

of millions of people in Africa

0:29

lack access to electricity, which means

0:31

they lack access to the fundaments

0:33

of modern life. What's

0:36

more, that number, which has been

0:38

regularly lamented for as long as

0:40

I can remember, is

0:42

rising, not falling. Many

0:45

of the places lacking

0:48

electricity have no expectation

0:50

of a large, high-powered,

0:52

centrally-administered grid—the kind Americans

0:54

take for granted—arriving anytime

0:56

soon. So the

0:58

best answer for providing power

1:00

right now is mini-grids—small, self-contained

1:03

grids, usually composed of

1:06

solar panels and batteries, that

1:08

can provide rudimentary electrical service at

1:10

the village level. But

1:13

people have been talking about bringing

1:15

mini-grids to Africa for decades, and

1:17

progress on the ground has been,

1:19

to put it charitably, slow.

1:22

Why is that? What's holding them back?

1:26

To dig into these questions, I

1:28

contacted Tombow Banda, a Malawi native

1:30

with advanced degrees in engineering who

1:32

now works at Crossboundary,

1:34

an advisory firm that collaborates

1:36

with businesses and governments to

1:39

unlock capital for sustainable development.

1:42

Specifically, Banda heads up

1:44

Crossboundary's mini-grid innovation

1:46

lab, experimenting with new business

1:48

models that can bring mini-grids

1:51

to scale. We're

1:53

going to discuss the state of mini-grids

1:55

in Africa, the barriers that have

1:57

held them back, and the kinds of

2:00

innovations that can unlock their

2:02

potential. All

2:13

right then, with no further ado, Tom

2:15

Bovanda, welcome to Volts. Thank you so much

2:17

for coming. Thanks for having me, David.

2:20

It's exciting. Yes, I'm so excited

2:22

to get into this

2:24

subject and I want to start before

2:26

we get into the specifics

2:29

of particular mini-grids with

2:32

just some context about

2:35

why they're so needed. What's the sort

2:37

of kind of addressable

2:39

market here of people in

2:42

sub-Saharan Africa who lack electricity access?

2:45

Sure thing. So in sub-Saharan Africa,

2:47

over half a billion people still

2:50

don't have electricity access in

2:53

the year 2024. Half a billion. Over half a billion. So

2:57

upwards of 500 million and

3:00

that number is growing. I feel

3:02

like I've been hearing about this for years.

3:04

Is that number going down or is it

3:06

getting worse? Well, it's

3:08

getting worse and the reason for

3:11

that is, well, population

3:13

is growing, you know, urbanization

3:16

is growing and

3:19

the infrastructure is not keeping up

3:21

with population growth and the needs of

3:24

communities, people, industries and

3:26

countries. So yeah, it's

3:28

a bit depressing, right? Because you think

3:31

like surely by now we'd

3:33

be a bit further along, but

3:35

you know the World Bank reports that at

3:37

current rates by 2030 we will

3:39

not achieve universal electrification

3:42

and you know good

3:44

I think 200 million people will remain

3:46

without electricity at current rates

3:49

of progress. Here's kind of

3:51

a big picture question before we get

3:53

into it, which is What

3:56

is the long-term vision for

3:58

these mini grids? Wondering what

4:00

is the sort of relationship between these me

4:02

goods and the kind of. Main

4:05

Grid A Big Grid A centralized grid

4:07

is the idea that these many grids.

4:10

Are. Put in place as

4:12

a sort of stopgap measure,

4:14

waiting until. The. Main Grid

4:16

arrives and is the me. Is that

4:19

idea that the main grid expands and

4:21

eventually absorbs them? Or is the idea

4:23

that they themselves expand and connect to

4:26

one another and become a main grid?

4:28

Or or are they up for quasi

4:30

permanent solution? Do not. I mean like,

4:33

well, what do we expect for these

4:35

mini grids and twenty or thirty or

4:37

forty or whatever years. Were

4:39

absolutely self how most countries

4:42

and on effort the now

4:44

think about the electrification is

4:46

to list of planning and

4:48

that is is it is

4:50

done through just basal modeling

4:53

so. If a country is thinking through

4:55

hard thing. To electrify it's communities

4:57

that it's population is a

4:59

looks at it's to focusing.

5:01

It'll look at where urban

5:03

centers are and figure out

5:05

what is the best and

5:07

least cost way of electrifying.

5:10

You know everybody Internet in

5:12

everyone's says existing great which

5:14

has in a bigger generation

5:16

centers and distribution networks that

5:18

are connecting it on nearby

5:20

communities and nearby industries are.

5:22

There's also those isolated areas.

5:24

Where you know many good the more

5:26

suitable and as also other areas where

5:29

you know the smallest zola home systems

5:31

a more suitable. So if

5:33

you think about isn't this terms of

5:35

least cost then you can establish okay

5:37

this is how old this goes. But

5:39

to your question, what happens That and

5:41

twenty to thirty years I think what's

5:43

exciting about supply and are for tat

5:46

the moment is you know we. Are

5:48

really building. Disputes. A

5:50

Grid Now which is more

5:52

a network of interconnected me

5:54

know, main grid. quote unquote.

5:56

many grid solar home sister and a

5:59

lead to com charging

6:01

networks for electric buses

6:04

and it's going to be this

6:06

interwoven mesh of nodes of generation,

6:08

consumption, all connected by distribution

6:11

network. That's where I believe it's going.

6:13

I don't believe there's going to be

6:15

a, you know, giant main centralized grid

6:17

later. It's going to be this interwoven

6:19

network. Which will be kind of

6:22

a new thing in the world. That

6:25

will be something no country has done. Yeah,

6:27

it will. And I think that's what's

6:29

quite cool about the opportunity we have,

6:32

right? Because we can build for that

6:34

future now. Because we don't

6:36

have a legacy old big grid, we

6:38

can build this, you know, smarter,

6:40

nimble, more suited

6:43

grid that is interwoven

6:46

and distributed. Yeah, really interesting.

6:48

So to help us sort

6:50

of envision one of these things, one of

6:52

these mini grids, talk about

6:54

kind of the typical size of one

6:56

of these things. Kind of how big

6:59

is it? What are the parts and

7:01

how many people are hooked up to

7:03

one typically? Yeah, so I always

7:05

describe a typical mini grid as a

7:08

small generation unit. You know, typically

7:11

50 kilowatts can be 50 to 100 up to 200

7:14

kilowatts. A solar

7:16

panel array connected to a

7:18

distribution network that is

7:20

connecting about, you know, 200 households,

7:22

20 businesses

7:25

or so. And usually

7:27

it has a backup system, a

7:30

small diesel generator, or battery pack

7:33

that helps kind of balance out the cost. And

7:35

that is really the typical mini grid that we

7:38

see in the Sub-Saharan African context. But

7:40

of course, you know, you have mini

7:42

grids that are hydropowered, and

7:44

bigger, smaller. But typically,

7:47

you know, this is what we're seeing all about. And

7:50

do they typically have batteries? And is the

7:52

rate of them having batteries, is that going

7:54

up like our battery is getting more popular

7:57

to use in them? Yes, I mean... Founded

8:00

they do have batteries am so that the

8:02

solar panel. To Raise plus the batteries

8:04

plus the backup. System that is.

8:06

The quantity of battery is what

8:08

fluctuates. Answer that is a balance

8:10

of of cost, right? Because I'm

8:12

batteries are costly part of this

8:15

and so you can balance out

8:17

with having diesel generator. but of

8:19

course that spoils the renewable elements

8:21

of the have them his picture

8:23

so you want. To minimize that

8:25

by, you know, It is still hopeful when

8:27

you want to balance the cost of of the

8:29

system. And so

8:31

who typically. Owns.

8:34

These things like what's the

8:37

development model or these third

8:39

party developers and is so

8:41

are they are. African

8:43

developers? Are these developers from outside

8:45

the country? Like ooh, what's? the

8:48

typical sort of model of ownership.

8:51

Of the super interesting

8:53

question so I've presents,

8:55

it's typically third party

8:57

developers. And. The picture

9:00

of who these developers are

9:02

has been evolving. Overtime

9:04

so I think you know

9:06

this is he go back

9:08

five six years at the

9:11

beginning of in of mini

9:13

grids taking all says no

9:15

commercial venture rather than just

9:17

a community and you kind

9:20

of installations you'd find quite

9:22

a few of the many

9:24

good founders might be you

9:26

know start up entrepreneurs who

9:29

has come from different countries.

9:31

And and looking to. Kick off this

9:33

energy access business and you know

9:36

which they see as impactful and

9:38

potentially. Also profitable. Now.

9:41

You have a lot of

9:43

third party developers who are

9:45

some local. Communities local businessmen

9:47

who in also see the

9:49

opportunity in providing electricity access

9:52

to their communities and building

9:54

a business out of it.

9:57

As not say the the third model that

9:59

is a evolving. No is more

10:01

and co of cold model

10:04

where you has an ethical

10:06

company that is raising. Figure.

10:09

Amounts of money and buying or

10:11

developing many good assets that are

10:14

operated by you know, developers at

10:16

the local level through an operator

10:18

agreement where they pay them a

10:21

fee and revenues come back to

10:23

the acid full and I think

10:25

that's where the industry is is

10:28

heading more generally. And full disclosure,

10:30

that's the model that as being

10:33

by. Access, which is a

10:35

unit and the company that I work

10:37

for costs pandering, but we think generally

10:39

that squared the market is going to

10:41

tend. Is. That just because

10:43

that's where the money's meters or

10:46

way things are moving in that

10:48

direction. it's a bigger. Piece.

10:50

Capital. Yes So

10:52

at present the number

10:54

needed for investment is.

10:56

Around nine to one billion

10:58

right to achieve energy access.

11:01

The Ninety One billion is

11:03

an estimate of the total

11:05

investment needed in many grids.

11:08

To get electricity to everybody in Africa

11:10

doesn't have it. That's what that number

11:12

means. Correct. The World Bank

11:14

estimates that we need about a hundred and

11:16

sixty thousand many goods and and that's the

11:18

number that gets us that. Oh,

11:20

hundred and Sixty thousand of them.

11:23

What's to? just for context. In

11:25

west of roughly how many are

11:27

there Now, how close are we

11:29

to them? We are. So far

11:31

has about three thousand o s

11:33

three thousand for the last bit

11:35

talent and we'll have a problem

11:37

that we conceive many. Bits from space,

11:40

but that's a different problem, so it's difficult

11:42

to pronounce as far. As that's exactly how

11:44

many they are. But from Thunder Up

11:46

for that, we have a conscious level.

11:48

Ah, we think that's about three thousand

11:50

and is a commercial. Many goods. And

11:52

so you know, a hundred and sixty

11:54

seven thousand. To go right

11:56

in her article years

11:58

other. Oh yeah, it's

12:01

really a lot. One

12:03

more question about ownership, because one of the

12:05

things I wonder, obviously when people think

12:07

about developing sub-Saharan Africa, one of the

12:10

things you want to avoid is exploitation,

12:12

and I'm just wondering

12:14

about, are there community ownership

12:16

models at all being talked

12:18

about, is there a way

12:21

of keeping some ownership

12:23

in the community, building some assets

12:26

in the community, building some wealth in the

12:28

community through ownership of these things?

12:30

Is that on the table at all? Yeah,

12:33

so David, I'm in two minds about

12:35

this. And

12:37

let me explain myself. Sure. So

12:40

in my friend's mind, if

12:42

you look globally, historically,

12:45

electricity is a common public

12:47

good that is

12:50

provided usually with a lot of subsidy,

12:53

usually developed at a loss with

12:55

the realization that this is just

12:57

a core enabler to development and

13:00

to progress, and so we must

13:02

have it. So in

13:05

the United States, in Europe,

13:07

in Asia, this

13:09

has been something that government has

13:12

led, and only at a mature

13:14

stage, the private sector has come

13:16

in to play this commercial

13:18

role. So historically, that has

13:20

worked. So now when we come to

13:22

the African continent, I have a similar

13:25

view in the sense that if you go, so

13:28

I'm from Malawi, if

13:30

you go to any unelectrified

13:32

community in Malawi

13:35

and you ask them, hey, do you

13:37

want to build and own and run electricity

13:41

utility, a small one, or

13:44

do you just want to pay your electricity bill every month? I

13:47

would guess they just want to pay their

13:49

electricity bill and get on with it, you know? But

13:52

there might be some communities,

13:54

and there are communities, To

13:57

which bringing that. Electricity

14:00

and having an opportunity to own

14:02

that asset and would provide a

14:05

source of employment. Would. Provide

14:07

you know a sense of

14:09

ownership. Tons for innovation to

14:12

do that. so I think.

14:15

Honesty by and large. David it needs

14:17

to be. Sent to the

14:19

lead and you know, a commercially

14:21

driven. Past. There are some context

14:23

which. Community ownership makes.

14:26

Sense and and I know there all

14:28

models of. Community ownership

14:30

also. You know, in

14:32

Alaska? And. Scandinavia where you know

14:34

the context makes it such that it

14:37

it makes sense for that. So I

14:39

do agree that the is an element

14:41

of you know. We don't, just, you

14:43

know, some. Big T money to come

14:45

in from abroad and. Just you know,

14:48

Put. Officer to sending an. Exploit locals.

14:51

But you know it has to be. Just

14:54

a viable. Government's involved

14:56

process of. Electrifying communities because

14:58

I also believe that that is also

15:00

important to com be just purely commercial.

15:03

The has to be an element

15:05

of of government involvement in town.

15:07

Or countries electrified. Younger.

15:11

I feel like I've heard about

15:13

the promise of many grids sub

15:15

Saharan Africa for decades. You know,

15:17

you know it's like of is

15:19

a very compelling story of a

15:21

compelling idea, but I feel like

15:23

it's been floating around forever and

15:26

as you see on the three

15:28

thousand of these doesn't seem like

15:30

it's taken off at scale. So

15:32

what is the problem? What are

15:34

the barriers that has kept this

15:36

model from sort of catching on

15:38

and growing. At the kind of

15:41

speed you need. Yeah. So

15:43

the most fundamental is many goods

15:45

are not profitable. By. It's

15:47

either technology, they're a good fit the make

15:49

a lot of sense function as he. They

15:52

worked as the alleys cloth as we discussed

15:54

that. They. Don't make any

15:56

money. Usually assistance so the return

15:58

phone you know. Communities that

16:01

are connected with. Is usually

16:03

not sufficient to. Give.

16:05

A lucrative pay back to

16:08

an investor. Therefore,

16:10

You know, commercially private.

16:12

Money is not interested in investing

16:15

in that on and so you

16:17

know where the balance arm and.

16:19

And. How do you make him

16:22

more profitable? That's partly. You know,

16:24

one of the things I work on

16:26

a done many good innovation lab across

16:28

boundary is making them in a good

16:30

business model more profitable so that they

16:32

can be more private sector too many

16:34

grid and why private sector investment into

16:37

money with. Isn't the

16:39

reason. Me: This

16:41

is a dumb thing to say, but

16:43

if it's isn't a main reason that

16:45

they're not profitable, simply that there isn't

16:48

that much wealth in these communities to

16:50

get out of them. You know what

16:52

I mean. Like as a better one

16:54

hundred degrees for half an. Hour.

16:59

That is one hundred percent correct. A

17:01

half as one hundred percent correct and

17:03

you're You're coming right to the and

17:05

and I'm so happy with asthma. has

17:08

a sense at all and you're like-reason

17:10

that they don't make any money could

17:12

lead people to know communities. Are

17:14

not one thing and that's a fact and

17:17

why they're not wealthy and that's as a

17:19

whole other system that we have to look

17:21

at All says you know what to pack

17:23

to eating poverty. In this country as

17:25

a starter, have a much deeper

17:27

question and it seems like there's

17:29

a bit of a chicken and

17:31

egg here too. We need electricity

17:33

access to do wealth producing activities,

17:35

you know, so you can't. They.

17:38

Have wealth to build the electricity before

17:40

they have the electricity I guess. Is

17:43

the way to put it Says that lane

17:45

exactly the just seems like of. a

17:47

real puzzler have a problem so what

17:50

do you what do you do about

17:52

that i mean what do you you

17:54

can't conjure up wealth in these communities

17:57

before the electricity gets there so how

17:59

do you you solve that

18:01

chicken and egg problem? How do you boost

18:03

profits? So I think that's

18:05

a bottom-up and top-down way of doing

18:08

it. So my work is kind of

18:10

focused on the bottom-up way of doing it.

18:13

So what you do is say

18:15

we're going into a community, we're building a mini

18:17

grid, and we're going to figure

18:19

out ways of increasing the

18:22

revenues that we get from

18:24

each consumer in that community.

18:27

And how we do this is

18:29

providing them with electrically powered, income-generating

18:31

machinery that will allow them, you

18:34

know, if they are farmers to

18:37

do agro-processing on our

18:39

electrically powered machinery, that'll make them consume more,

18:41

so that's good for us. And the processing

18:43

will make them make more money, and that's

18:45

good for them because they can afford to

18:47

pay for electricity and pay for other things.

18:50

So from the mini grid perspective, we look

18:52

at income-generating machinery,

18:56

electrically powered machinery that can be

18:58

deployed in those communities using

19:01

financing. So, you know, can we

19:03

get them to get these machinery

19:05

on credits that they can also afford it,

19:08

and really working with the community to

19:10

provide that. So that's, you

19:12

know, the bottom-up way of looking at it to

19:14

say, well, come with the electricity, figure out what

19:16

businesses are there, what businesses can be electrified

19:18

at the same level

19:21

of cost as what they're doing

19:23

presently or cheaper, and what can

19:25

make them more money. So that's one way of

19:27

looking at it. So the

19:29

idea is a business person would take out a

19:31

loan, buy a machine, use the machine

19:35

to start a business. The

19:37

business makes money so it can pay back

19:40

the loan for the appliance and pay

19:42

the electricity, pay the mini grid.

19:45

So then the entrepreneur is making some money,

19:48

the appliance gets paid back, and

19:50

the mini grid is making more money.

19:52

That's the idea there. That's

19:55

the basic idea. But I think one clarification

19:57

I would make is in the bottom-up model.

20:00

the majority of consumers

20:03

that end up purchasing

20:05

income generating machinery are

20:08

already existing entrepreneurs that are maybe

20:10

using an alternative or maybe not

20:12

using an efficient alternative. So

20:15

for example, it could be a mill

20:17

operator who's using a diesel powered grain

20:19

mill. If you can

20:21

offer him a more efficient, electrically

20:24

powered mill that's going to help

20:26

him process faster in

20:28

a day at a cheaper cost

20:30

than diesel, then that's the

20:32

typical business that you would be

20:35

working with. There are of course

20:37

new entrepreneurs that see

20:39

the opportunity to start new things. There's

20:42

a lot of carpenters, even people

20:44

who just own bars

20:47

in communities who will

20:49

get TVs, fridges, and

20:51

attract more business to

20:53

their establishment. So

20:55

it's simple things like that. So

20:57

converting existing entrepreneurs and there's also

20:59

the creation of new entrepreneurs and

21:02

getting them onto income

21:05

generating machinery that they can afford

21:07

and access through credit. And

21:10

ensuring that they are reducing costs that

21:12

they had before and making more profit

21:15

is generally the sustainable way

21:17

of ensuring that you have these revenues that

21:20

can be maintained by the community. Interesting.

21:23

So you're just sort of like trying to create

21:25

more demand for electricity basically as a

21:27

way of making the mini grid more profitable. That's

21:30

exactly it. Okay. And you said there was another

21:32

way? Yeah. So the other way is a

21:34

more top down way, right? So in

21:37

the method that I described, the

21:39

onus is really on mini grid

21:41

developers to act as

21:44

the activators in those communities

21:46

for switching to these electrically

21:49

powered machinery and getting

21:51

into more agro-processing, et cetera. But

21:53

in the top down way, if electrification

21:56

and energy access is thought

21:59

about... More realistically in the

22:01

context of economic development and I think

22:03

you know this is what you were

22:06

looting to earlier and thinking through you

22:08

know many good the not possible. Because

22:10

the community that they're serving don't have

22:13

you, don't have sufficient money. So

22:15

if we think about electrification,

22:17

surface and the in the

22:19

context of economic development than

22:21

you think Hey so this

22:23

community. if we electrify this

22:25

community they have. You know

22:27

they have timber and. The

22:29

communities they need:

22:32

Woodworking shop. And

22:34

you know what does that take?

22:36

They have a school on behalf

22:38

hospitals. These order to. Be electrified and

22:40

in of these are the vaccine fridges

22:42

we need in the. The hospitals to

22:45

make sure everybody's healthy to. Then

22:47

go to school and by the desks. For

22:49

the school and exports doors and window

22:51

frames to make additional money for the.

22:54

Community or you know, is

22:56

it? Say. A more agrarian

22:58

community where they are. Trolling

23:00

a cast cop that needs to be processed.

23:03

You know, maybe they are growing mangos? I'm

23:05

in that. Village or they're

23:07

growing. but what? What's a bit

23:09

of cash crop actually to them?

23:12

You know, Can we

23:14

then provide the community or

23:16

can be said is up

23:18

to that such as as

23:20

processing the access to markets

23:22

and it's electricity am this

23:24

refrigeration as all the you

23:26

know other things that are

23:28

needed to have that community

23:30

thrive apart from to sell

23:32

it to supervision and looking

23:35

at a moralistic li and

23:37

coordinating across many good developers

23:39

government institutions. Community organizations,

23:41

agricultural cooperatives, And

23:44

making that a more realistic way

23:46

of looking at the new economic

23:49

development. So I hope that's the

23:51

direction that we're going. Rather than

23:53

you know, kind of relying on the

23:55

bottom. Up or maybe another summer where

23:57

they. They meet in the middle halfway.

24:00

But I think that's the two ways to think about it. Well,

24:03

that's a matter for policy and regulation,

24:06

is it not? Yes. Are

24:08

the governments of

24:11

that region, are they clued

24:13

in to mini-grids? Are they moving in that

24:15

direction? Do they get it? I mean, are

24:17

they moving in that direction that you want

24:19

them to be moving? I think

24:21

many are now. So, you

24:24

know, five, six years ago, the

24:26

story was different. You know, mini-grids

24:28

were a new concept to many,

24:31

including some people in government. I

24:34

think now there is a greater awareness

24:36

across governments, across regions. But

24:39

of course, you know, the level of development, the

24:41

level of acceptance is different from country

24:43

to country, from region to region. But

24:46

I'll say the general trend is, you know,

24:48

governments are more clued up to this

24:51

and governments are more

24:53

cognizant of, you know, this element of

24:55

needing to think about electricity

24:57

together with income generating

25:00

machinery, productive uses of energy and tying

25:02

it into markets. So, there is

25:04

greater awareness of that now, for

25:07

sure. So, I think we're moving in the right direction. One

25:10

thing that I've heard a lot

25:12

about the mini-grid market, one

25:14

of the problems with it is that these

25:18

deals for development of mini-grids tend

25:20

to be quite bespoke.

25:23

Like everyone is different from

25:25

village to village. You know,

25:27

you're building slightly different ways, you're servicing

25:29

slightly different. You know, you

25:31

have a different amount of sort of local

25:34

labor skill that you can tap into, etc.,

25:37

etc. That seems to me, that

25:39

seems like something that's difficult for developers,

25:42

just a lot of puts a lot of weight

25:44

on developers. Is there

25:46

movement to try to standardize this

25:48

process or is that even a

25:51

thing that could happen? Yeah.

25:54

So, David, I'd say that there is

25:56

movement. There's a lot of innovative

25:59

solutions out there. there that are kind of

26:01

mini grid in a box sort

26:04

of set up where you have the

26:06

containers with all the electric equipment inside

26:09

and panels can be set up and

26:11

all you have to do is now set up the distribution network.

26:14

And I think set ups like that

26:16

and standardizing or set ups like that

26:19

makes it quick and easy to deploy at scale.

26:21

And I think what's happening

26:23

now is as well, David, I know

26:25

we mentioned that we only have 3000

26:28

mini grids or thereabouts, this

26:31

huge investment, well, bigger investment going into

26:33

mini grids now in the hundreds

26:35

of millions. And we're seeing

26:38

those players who started out as

26:40

developers with five to

26:42

10 mini grids now go into the

26:45

hundreds of mini grids, you know,

26:48

scale or level of

26:50

magnitude. So, you know, with that,

26:52

they learn how to deploy quicker, they

26:54

know what works, they know the teething

26:56

problems to be found in

26:58

specific kinds of communities. And,

27:01

you know, we work with them, the African Mini

27:03

Grid Developers Association. And

27:05

one of the things that they actually try

27:08

to do, and are working

27:10

on now is passing

27:12

on, you know, the wisdom from the elders

27:14

to the generation of mini grid

27:16

developers, because they don't want

27:18

them to make the same mistakes that they

27:21

did, they tried out a whole bunch

27:23

of things to figure out how to do

27:25

it right and how to do it most cost

27:27

efficiently and how to do it quickly. So

27:30

they are also investing their time into training

27:33

up the new developers that

27:35

are coming up in across

27:37

different countries on best practices

27:39

to roll out mini grids easily,

27:42

more cheaply, and more efficiently.

27:45

Right. Well, one of the other problems I've also heard

27:47

about is it has to do with ongoing sort of

27:49

operation and maintenance. Like, you know, you

27:52

did, I think in the early, early

27:54

days of this sort

27:56

of the mini grid thing, you heard a lot

27:58

of stories about, you know, like a couple of

28:00

panels and a battery would get installed and they would just

28:03

break and there wouldn't be anybody around to fix them

28:05

and they just kind of, you know, molder. So

28:08

what is the typical model? Who,

28:10

does the developer operate this

28:13

thing or is it a local agency

28:15

of some kind that operates it and

28:17

maintains it? How is that typically done?

28:19

Yeah, so typically the developer

28:21

will operate the mini grid.

28:24

They will do the operations

28:26

and maintenance. They'll collect the revenues. They'll

28:29

do the customer service. They will do

28:31

the community engagement. So that

28:33

model is a lot more common.

28:35

I think, you know, what can

28:37

be challenging is, you

28:40

know, if you have big countries, you

28:42

know, for example, in Sierra

28:44

Leone where, you know, sites

28:46

are far apart, hours

28:49

apart, you know, might be tricky and

28:51

take long to get there is, you

28:54

know, you might not have the maintenance

28:57

necessarily on the ground, but,

29:00

you know, usually these systems are

29:02

monitored remotely and they

29:04

have, you know, a technician, someone looking at,

29:06

you know, the condition of the

29:08

equipment on the ground. And if they do see

29:10

a problem, you know, it might be some

29:13

hours or a day or two before maybe

29:15

the right technician can get there to fix

29:17

it. But usually it's the

29:19

developers on these commercial mini goods who

29:22

are operating and maintaining to

29:24

make sure that they're maintaining at least a decent

29:26

level of service for the communities. And

29:29

just a quick question about technology.

29:32

I just sort of am assuming that

29:34

these things are pretty standardized at

29:36

this point and that the pieces

29:38

are pretty simple. Like

29:41

solar panels are pretty simple, batteries are pretty simple, you

29:44

know, stringing wires to houses is relatively simple. Is

29:46

there any technological advancement happening

29:48

or anything exciting to say about the

29:51

technology or is it just, is it

29:53

good that it's standardized

29:55

and boring? So it's good

29:58

that it's standardized. and

30:00

boring. But you know what

30:02

we try to do is you know

30:05

make the mini-grid ever more lucrative

30:08

for investment right and ever more

30:10

affordable for consumers. So that

30:13

requires some innovation in

30:15

the technology to reduce the cost

30:17

or to get more for less. So

30:20

I think in terms of you know

30:22

the panels not really much

30:24

innovation going on there and

30:26

in the batteries but in terms

30:28

of innovations to make things cheaper we

30:31

are seeing you know second

30:33

life batteries come onto mini-grids

30:37

and that can reduce the battery

30:39

cost significantly. Oh yeah. These second

30:41

life batteries yeah you know come

30:44

from electric vehicle applications where they

30:46

were used in dynamic settings but

30:48

now they can used in stationary

30:50

settings. So you know

30:52

we have some middlemen

30:55

I guess is the right term who

30:58

take these EV batteries, disaggregate

31:01

them, look at the

31:03

individual cells, make sure they are you know

31:05

still performing well and still have sufficient life

31:07

in them, reconfigure them, snap

31:10

on a new battery management system

31:12

and redeploy these to mini-grid sites.

31:15

So that's an interesting technology

31:17

innovation in that you know the battery

31:19

chemistry itself is not new but

31:22

the application is new and how it's been set up is

31:24

new and it's you know resulted in a 40%

31:28

cost cut. Oh interesting. I think another

31:30

innovation that is interesting for us

31:32

is you know smart

31:34

inverters and mesh grid. Oh

31:37

yeah. So yeah so this you

31:39

know is basically using the same components

31:41

but using you know software

31:44

to allow individual units

31:46

to talk to each other and

31:48

share power you know across different

31:50

consumption centers. So

31:53

I'll say the core technology remains the same but

31:55

how we use it is

31:58

changing and the use of software

32:00

to share power, even

32:02

track and predict power

32:05

is making it interesting. Another

32:07

one more thing I would talk about is on

32:10

the metering side. We're

32:13

doing quite a cool experiment with

32:15

a local Kenyan company here called

32:17

PowerPay who's

32:19

doing disaggregated,

32:22

non-intrusive load monitoring for

32:24

individual households. And we also want

32:26

to test that at community level. And

32:28

simply what this is, David, is

32:31

actually it's used in commercial applications

32:34

in industrial settings, so you probably

32:36

know what it is. But

32:38

using one meter and connecting

32:40

different loads to that individual

32:42

meter and then

32:45

disaggregating where the load

32:47

or consumption is originating from. So in a

32:49

house, you'd be able to tell from that

32:51

one meter that 60% of the

32:54

consumption was from the

32:56

cook stove and 40% was from bulbs. And

33:01

that's important because we

33:03

want to be able to charge

33:05

things like cooking terrace to encourage

33:08

clean cooking as opposed

33:10

to cooking using

33:12

charcoal or other dirty sources.

33:15

And it's also important because if you think

33:17

about it on a site level, meters

33:21

can be up to 10% of

33:23

a site's capex because developers

33:26

use smart meters to ensure that they

33:28

can charge different customers

33:31

accurately. They can also track

33:33

what they're actually generating

33:35

versus consuming. So they use smart meters

33:37

and those are expensive. But if

33:40

you can use one smart meter across

33:42

five or six houses and accurately build those

33:44

five and six houses because you can disaggregate

33:47

their loads using AI,

33:49

that's a cost reduction. Interesting.

33:52

So there is some cool technology stuff

33:54

happening. I'd say that the physical hardware

33:56

probably the same, but how you

33:59

use software. and then how you reconfigure

34:02

those innovations is changing.

34:04

Really cool. One more money question.

34:07

I watched your TED Talk. Everybody should

34:09

go watch it. It's great. But

34:12

you said you're trying to make these things more profitable, and

34:14

right now they're in the 5% to 6% return

34:17

range, and you want to get them up

34:19

to the 10% to 12% range

34:22

to attract more private capital.

34:25

And I guess I get that. I get

34:27

why you need those kind of high returns

34:29

to attract private capital. But is there really,

34:32

on another level, it's bothersome that

34:35

people are going without electricity because

34:38

someone can make 6% returns and not 12%?

34:40

You know what I mean? Is

34:43

there no patient capital

34:45

out there? Is there no

34:47

pool of infrastructure funding that's

34:49

more patient and more tolerant

34:51

of lower, longer term results?

34:55

Is there none of that to tap

34:58

into? David, that's such a great

35:00

question. And I think the answer is there is,

35:03

but we're not seeing it fast enough. So

35:06

in the model I talked about earlier, where it's

35:08

more asset co-op

35:10

co-invest in a larger portfolio of

35:12

mini grids, that attracts the

35:15

kind of money you're talking about.

35:17

So infrastructure, patient capital, willing

35:20

to wait 25 years for a return. So

35:22

I think the fact that we're evolving towards

35:25

that sort of model is one of the

35:27

reasons why there's more money coming to the

35:29

sector, and we expect more money to come

35:31

towards the sector that way. But

35:35

in the short term, we

35:37

have seen this need to

35:40

at least get that return at the

35:42

more basic level

35:44

to get the bigger

35:46

infrastructure funds interested. The

35:49

patient capital, confessional capital, the

35:52

impact capital, sometimes they're satisfied just

35:54

to get their money back. Sadly,

35:57

if you want some of the money back, pension

36:00

funds and things like that of the

36:02

very countries that we're serving to invest

36:06

in these mini grids, they

36:09

do want to see that kind of return sometimes. So

36:12

it's happening, I would say, and I

36:15

wish it was more. I find it

36:17

frankly outrageous. That's

36:19

kind of what I was getting at. Yeah.

36:23

It's outrageous and shocking. And

36:25

if you think of a number like 91 billion

36:28

for an entire continent, that's

36:31

not a huge number. I

36:33

know. That's pretty small. It's a huge

36:35

number. Yeah. In

36:38

the global scale, that's a

36:40

couple of billionaires knocked off one

36:44

and two, but not even one and two,

36:46

like position 19 and 20. And

36:50

that's a whole continent electrified. So

36:53

yeah, it's really not

36:56

satisfactory, but we have to

36:58

work in the context of capitalism.

37:01

So here we are. We're working

37:03

to make it more. Here

37:05

we are. Well, the last kind of thing I wanted to

37:07

ask about was the last barrier

37:10

I hear to a lot of

37:12

capital coming in is governance.

37:15

I think one of the reports

37:17

you sent, I was looking at some of the

37:19

reports you sent, and one of the reports said

37:21

that it takes the average mini grid in Africa

37:24

58 weeks to get all

37:27

the licenses and approvals it

37:29

needs to build. And

37:31

this is like, as you say, like

37:34

a couple of containers, like I'm sure

37:36

it's like in terms of physically building one of these things, you

37:38

can do it in a couple of days,

37:40

you know, or a week, but like 58 weeks

37:43

to get through the bureaucracy

37:45

seems kind of extraordinary. So

37:47

how's that, is that getting

37:49

any better? I guess

37:52

I mean, two things. One a regulatory structure that sort

37:54

of is faster and makes it

37:56

more sense and it's more predictable, but also just

37:58

sort of governance, you know, just general. like, can

38:01

I trust that I, you know, contracts

38:03

will be fulfilled and this

38:05

kind of thing. Like how big a problem is all

38:07

that stuff? Yeah. I'll reference

38:09

again the African Mini-Good Developers Association who

38:11

we work with. You know, these two

38:13

things are top of their agenda to

38:16

address. So they work

38:18

with governments and they

38:20

work with regulators to, you know,

38:22

streamline these two processes. And

38:24

I think, you know, it definitely has

38:27

been a challenge for developers

38:29

and especially in new geographies where maybe,

38:32

you know, Mini-Goods are new to

38:34

the country and, you know, regulators are

38:36

just getting accustomed to it. They have

38:38

to, you know, figure out a whole

38:40

new tariff regime and tariff models and,

38:42

you know, socialize that, you know, internally.

38:45

And you know, the people issuing licenses

38:48

have never issued that kind of license

38:50

before. So in new geographies, there's

38:52

a lot of firsts, but

38:54

I think, you know, similarly to what they

38:57

do with, you know, developers passing

38:59

on wisdom, what we've seen now is a

39:01

lot of standardization

39:03

across what, you know, regulatory

39:05

regimes should look like across countries

39:07

and countries really cooperating with

39:10

each other and, you know, not trying to reinvent the

39:12

wheel, but, you know, looking at, you

39:14

know, best practices and adopting those as

39:17

they are figuring out what their

39:20

regulatory regime looks like. So there's

39:22

a lot of talking that's happening

39:24

between regulators across these countries, which

39:27

is very helpful to developers and

39:29

very helpful Mini-Good developing, Mini-Good

39:32

developers in those countries. But,

39:35

you know, there's still an element of ramping

39:37

up that has to go so that that's

39:39

still not there. But I think in terms

39:41

of structuring, you know, as long as there

39:43

is that communication and they're learning from each

39:46

other and they're not making the same mistakes,

39:48

we can expect that this improves.

39:52

And then governance, I mean, this

39:57

is, I mean, it varies from country

39:59

to country. I think there's an element of

40:01

risk there. There's an

40:03

element of risk there, but I think any

40:05

country, a lot of them,

40:07

I think the good thing about electricity is,

40:10

most politicians, all governments know the value of

40:12

electricity, and all governments

40:14

know that if they electrify

40:16

communities and increase access, communities

40:18

are happy, and that can translate to votes.

40:22

So they tend to try not to stand

40:24

in the way of getting electricity votes, and

40:26

in fact, they can say,

40:28

under my watch, look how much electricity

40:31

access has improved. That's a pro for them.

40:34

So there's that element that you can

40:36

rely on governments for, but

40:38

on the other hand, things always

40:41

change. There's pressures from

40:43

other sources that may cause things

40:45

to change, so it's

40:48

a risk that just has to

40:50

be mitigated as far as possible. Is

40:52

there any particular policy that

40:56

you would sort of single out as

40:58

particularly helpful for a country to implement

41:00

that would make this whole thing go

41:03

faster? I

41:05

don't know what the policy environment is

41:07

like in these places

41:09

at all. I don't know if there are barriers that

41:12

could be removed by policy or anything,

41:14

or this is mostly just about the

41:16

markets. I

41:18

mean, so on the policy side, I would say

41:21

a lot of governments blanket, I

41:23

would say number one is

41:25

removing the taxes on

41:28

renewable energy projects.

41:30

So in most countries, actually, if

41:33

you import the panels, the

41:35

batteries, that stuff

41:37

tends to be duty-free, tax-free.

41:40

So I would say maintaining that is essential

41:42

because that's a cost cut. I

41:45

would say in countries

41:48

like Kenya, now you can apply

41:50

for blanket licenses. So

41:52

if you are a mini-grid developer

41:54

and you want to develop a portfolio

41:56

of, say, 20 mini-grids, you

41:58

don't have to apply for 20. 20 licenses, you

42:01

can apply for one license for your portfolio.

42:04

So things like portfolio licenses help.

42:07

Um, this, this is a really

42:09

big one, actually, David, I

42:12

am a big believer in

42:14

subsidies for electricity. So I would say

42:17

a big policy

42:19

incentive could be how

42:22

you think about sources of subsidy

42:24

for electricity. So, you know,

42:26

in a typical utility, you have

42:28

your cross subsidies from your industrial customers

42:31

who will be cost subsidizing the lower

42:33

paying one in mini grid

42:35

that happens to some extent, right? Cause

42:37

you have the businesses who are really

42:40

holding up the business and paying most of the bill,

42:43

but you know, there's also just the element of straight

42:45

up subsidy. And, you know,

42:47

actually before this podcast, I

42:49

was, you know, looking up, uh,

42:52

subsidies in the U S and

42:54

there are so many subsidies that you

42:56

guys enjoy for electricity. Really, really are.

42:58

We just passed a bunch more. Um,

43:02

and then one that caught my eye was

43:04

the low income home assistance programs. Right.

43:08

So that basically pays low income that

43:10

helps low income households to

43:12

cover the electricity bill. If

43:14

there was a straight policy or

43:17

an understood policy across most governments

43:19

that all consumers

43:21

of electricity require some level

43:23

of subsidy. I

43:25

think the level of deployment

43:28

of mini grids would shoot up immediately

43:30

because that delta of just covering

43:32

the revenues really makes up for

43:35

all this other work that you have to do

43:37

in, you know, developing businesses

43:39

and, you know, making sure you pick

43:41

the most economically viable communities first. So

43:45

I will rewind all my answers and

43:47

say that it's a top policy recommendation. I'll make.

43:50

Excellent. Okay. Final question. Um, your

43:52

job is your team

43:54

develops sort of innovations in business

43:58

models for these many. grids and

44:00

tests them out and tracks them

44:02

and then tries to expand on

44:04

the ones that work, etc. So

44:06

I thought by way

44:08

of a last question I would just ask, is

44:10

there a particular innovation that

44:13

you have developed or tried, a particular trick

44:15

you've tried or a thing you've figured

44:18

out that worked particularly well, that

44:20

you're particularly proud of? Anyone you'd

44:22

want to single out? Ooh,

44:25

shout out.

44:28

I guess to be clear, we work

44:30

through developers and other technology providers and

44:32

kind of quarterback the process of trying

44:35

these innovations and tracking the data and

44:37

making the analysis to figure out what

44:39

works and doesn't work. So

44:42

the most high impact innovation far and

44:45

away is appliance financing, David. And I

44:47

think that's what we were talking about

44:49

earlier when you provide appliances

44:51

and income generating machinery on

44:53

credit to end users and a

44:56

bunch of developers across

44:59

different countries. It's almost standard now

45:01

that they provide that as part

45:03

of their offering when they bring a mini grid to

45:05

a community.

45:07

So hard to single out a

45:09

single individual there. But I think

45:12

a couple of cool technology providers I'd like

45:14

to give a shout out to. The

45:17

first is a company called SLS

45:19

Energy in Rwanda. I spoke

45:22

about Second Life batteries earlier. So

45:24

they take Second Life

45:26

batteries from, well, one of their

45:29

sources is a cool electric two

45:31

wheeler company called Ampersand. So they

45:33

take old batteries from

45:36

electric motorbikes that were in use

45:39

in Rwanda. And they're

45:41

converting them to Second Life batteries for

45:44

mini grids and for telco towers. And

45:47

so I think they're doing a very cool job. And

45:50

I like, you know, as a

45:52

Malawian, I like kind of homegrown solutions

45:54

and these are an indigenous random company

45:57

doing that. So

45:59

I like working with them. Another company

46:01

I'd like to shout out, I think we're

46:03

doing cool stuff, where there's a company called

46:05

PowerPay here in Kenya. They

46:08

are an Internet of Things company. They

46:10

are the company that is developing the

46:13

meter that does non-intrusive load

46:15

monitoring to disaggregate loads at

46:17

household level and site level.

46:21

They developed the AI algorithm. They

46:23

are training it within their

46:26

lab. They're quite

46:28

cool and they're doing cool stuff. So, yeah, I'd

46:30

like to give those two companies a shout out.

46:33

Awesome. Thank you so much, Tom. This

46:35

has been absolutely fascinating, super, super

46:37

educational for me. Thank you for taking

46:40

the time. Thanks so much, David. Appreciate

46:42

it. Thank

46:49

you for listening to the Volts

46:51

podcast. It is ad-free, powered entirely

46:54

by listeners like you. If

46:56

you value conversations like this, please

46:58

consider becoming a paid Volts subscriber

47:01

at Volts.wtf. Yes,

47:04

that's Volts.wtf so that I

47:06

can continue doing this work.

47:10

Thank you so much and I'll see you next time.