Visiting a floating village energised by a mesh-grid one year after deployment

TLDR Key takeways for Okra team / project developers / product developers in the energy access space.

• I recently visited a Cambodian village, Kbal Toal - a community energised with mesh-grids under an Oxfam program (CAPFISH/Foster). It's a floating village where households actually physically move their household by tying a boat to it and finding a perfect location for fishing when the seasons change. Only 4 of the households are connected to the "mesh-grid" (ie. sharing power) even though 55 households were originally interconnected because at the this time of year most households move their locations temporarily to a place where they are not nearby to other households with an Okra system.

• Kbal Toal is a great model for energy access: the community already had an industrious nature, a high penetration of diesel generators before the Okra mesh-grid technology arrived and an LMA (local maintenance agent) that has been successful in championing further productive use since the mesh-grid technology was installed. This has lead to an average daily consumption of more than 600Wh/day for the 55 households that are connected, and there are more than an additional 100 households waiting to be connected.

• A lot of people in the community didn’t even purchase appliances like refrigerators because they didn’t know what they were capable of powering - this indicates even with a highly engaged LMA there are some end user onboarding & education issues that need to be addressed.

• Some applications for productive use shouldn't necessarily be targeted by our infrastructure if we're looking for best bang for buck electrification. For example the circular saws I saw that were only used for 1 hour/week. It’s not effective to aim the entire infrastructure development at these use cases - low frequency and high power uses should be evaluated separately.

• This was a good opportunity to look at a community-owned, self-sufficient model, where revenues collected from the mesh-grid have been enough to cover minor fixes, which the LMA has also proved to be capable of conducting. However, there are concerns over the ability for these revenues to cover major equipment replacement like batteries when the time inevitably comes and there should be contingency budget before commissioning grant funded projects, or higher tariffs set otherwise there will be challenges that will be harder to deal with later.

• Some lessons that can be taken from this community to our projects in Nigeria: What’s in the developers control is being able to select communities with high pre-existing genset use and also to select and train LMAs that can act as champions for productive use - we can do this by supporting the LMA with demo appliances at their home and spending most of the community engagement resources on the LMA so they can combine it with local knowledge to promote what’s possible in the community.

• Bonus takeaways from ACE (African Clean Energy): Regular training of rural team members such as LMAs can lead to more effectiveness and also retention. Basic KYC steps such as checking for a secondary contact and confirming an active mobile money account can go a long way to boost repayment rates. Quality of service can be standardised through a field agent / LMA app process that doesn’t leave room for error when it comes to support, upselling and after-sales.

Let's dig into the site visit - Background

I’ll begin by stating we’re not currently actively supplying mesh-grid technology to the Cambodia market. This is because currently tariffs are capped and energy must be supplied by government sponsored REEs (Rural Electrification Enterprise). This means the private sector has no place in serving rural households and households are required to get their own solar home systems or diesel gensets if they are not connected to the grid and want access to energy.

That being the community we visited, Kbal Toal, is a site that we energised about 18 months ago under the CAPFISH/Foster project funded by Oxfam. The reason why this site is a particularly interesting case study to explore is that the ownership of the assets are not by a 3rd party private DESCO (distributed energy service provider) but actually by the community itself. The community is responsible for charging for energy, collecting revenue, and also conducting operations and maintenance on the network.

I went on this trip with a party including some old colleagues:

Otteh: our old service delivery team lead, who has been on a year long journey of soul searching and finding out what it means to be alive. What does it mean to be alive? This question has been asked for thousands of years, so although I’m sincerely wishing him the best, I’m not holding my breath, and am just grateful to be able to spend small moments catching up and taking bits and pieces of wisdom he learns along the way.

Sovannara: our first local Cambodian team member who has been engaging with this community since before they were connected. We call him the naughty monk, because he used to be a monk and he's always the jolliest guy in sight, which is no wonder why people feel comfortable sharing their woes and their desires with him, it's made him the guy who understands rural community engagement better than anyone I know and it was so good to see him again after more than a year!

Special cast of team members from African Clean Energy (ACE): a company supplying improved cookstoves to address cooking with wood at home in Cambodia.

My girlfriend: I was excited to show her the first off-grid community she’s seen that is energised with Okra technology - and I knew her brain would provide a completely different perspective than my own would.

Pictured also on the boat: The LMA trained up to maintain local operations and a monk whose name we didn’t catch who was waiting for a ride back to his community.

Going to site, we were not sure what things were going to look like on the ground - and that’s because of the 52 households that were connected, 2 households were online and reporting data, but 50 households were offline…

Interviews with community members

Upon arriving to the community, there was much optimism and much to be inspired about. Even though 50 households were offline, it was only a data issue. The systems were fully operational, households were still using energy access, and using it fruitfully for that matter too!

What kind of productive use is out there?

Productive use was booming. Our boat drove straight past a laundromat - in an off-grid village! If you’ve been to as many off-grid communities as I have, this is pretty damn cool, and it sure as hell isn’t commonplace.

Most of what we saw is that productivity is being used self-reliance : soldering irons, drills to fix broken electronics or household construction and of course rice cookers to ensure ample amounts of rice are available at the local restaurants. However some of the households were also making boats, furniture and small bits like cutting boards with the aid of their power tools.

Here are some paraphrased remarks from some of the people we interviewed -

Safety & Security

Village Chief - Before the Okra system, we used to have 3 light bulbs, I have people working at my place to make Prahok (dry fermented fish) into the evening. And I also have about 400,000 catfish fingerlings growing under our house. Not having much lighting was a big security concern and 3 times thieves have come and taken catfish from under my house. Now we have 9 bulbs and we leave them on all night, there’s no more theft and we can feel safe investing in more catfish farming which is what we’re doing now.

Livelihood & E-Learning

Leng Bou - Our whole livelihood is dependent on fishing. When we go fishing we don’t come back till the evening. Before I would wake up at 3am to fetch wood, and then start a fire to cook rice before we go out to go fishing at sunrise. Now I just wake up and turn on the rice cooker, and when the rice is ready we take it with us on the boat and enjoy the rice during the day. Also my children use the energy from Okra Solar systems to charge our phones. The 6 year old one uses it to watch cartoons and the 9 year old one is using it to watch videos learning English off youtube. We think the phones are good for them.

Microenterprise

Onnam - The main change to my life has been the freezer, I had my own solar (home system) before for lighting and phone charging and it allowed me to run my shop where I have a small pharmacy and sell other things like charging cables and engine oil. Because it's hot, the fish spoils quickly. After getting the Okra system, I run the freezer using my old solar system until the battery runs out and then I move to powering the freezer using the Okra system which powers my freezer through the entire night. I have to pay for energy from the Okra system so I prefer to use my own system first. I keep my vegetables and fish that I catch in the refrigerator and I make ice everyday that I use to take out on the boat so I can keep the fish cool and fish for longer

The mesh-grid technical operations

In addition to checking in with the community members, it was important to understand how the technology was functioning. We’ll go through a few questions I was asking along the way.

What’s happening with the households that were not reporting data?

As mentioned before, 50 of the households weren’t reporting any data, so until visiting them we had no idea if they were actually still receiving power or if something had gone wrong. Fortunately, our hardware had not malfunctioned - households were still getting access to energy, the battery charge control also seemed to be working completely fine, and households that were using up to the daily limit of 1.8kWh were still getting cut off at 1.8kWh of daily usage.

As expected, all the logic on the devices was still operational.

Although we still haven’t resolved the the data issue, our hypothesis is that we have an issue with our global sim provider. Outside of not being able to monitor, another issue is that because our algorithms do not bill households if they are offline for the majority of the day, and because we have no keypad-based solution to manage billing, offline houses have been granted free power. This is great for the end users in the short term, but bad for the community in the long term because the community is not able to collect revenues, which are to be used for managing operations and maintenance.

We need to resolve this offline billing issue in the future: one solution that we’ve already deployed at a pilot site is our Wi-Sun based communication solution. This is now getting rolled out in Nigeria and once it’s deployed at scale we won’t be so dependent on each house being online in order to get a data connection to them.

Were there use end-user energy cases that were not feasible with the Okra systems installed?

🟠 Blenders would trip the system when connected with other appliances (the blenders I saw were 900W appliances)

🟠 The Circular saw would trip the system (1.05kW appliance)

⚪ Some end consumers didn’t think the Okra system would be powerful enough to run their freezer, even though it is. They would still be paying more than $1 per day for ice instead.

How is the mesh-grid helping out reliability and productive use at the floating village?

Only 4 households are inter-connected in mesh-grid configuration. Although a lot of the households were initially located near to each other and interconnected into a mesh configuration, because it’s a floating village a lot of the households have moved far apart from each other during the rainy season. The interconnected systems have then been disconnected from mesh-grid to standalone configuration. This indicates the households are not actually benefiting from improved reliability that you get from interconnected systems and would have had the same experience from large standalone solar home systems.

Are there any other takeaways for improving the mesh-grid setup?

🟠 A number of inverters we had supplied have been failing in the network, the failure mode appears to be an interesting one - short circuit by fallen gecko. Here’s an image of the failure mode after opening up one of the inverters. Inverters are something that we have a priority on optimising in any case so another bit of feedback to take into the mix.

🟠 Not all of the households are using close to the max of the daily energy availability. From looking at the Harvest platform, I was able to confirm this (by checking data older than the last 2 months data after which we have no visibility) the average load is at about 600Wh per day, with the average daily energy availability being about 1.5kWh per day, which means less than 50% of the total energy availability is utilised and there is the potential for nearby households to benefit by interconnection into mesh.

Could this really be a practical solution especially for this floating village?

Most of the households already have some form of solar. A lot have 100W+ systems and the lowest end households have small 15W system for charging phones / powering lightbulbs - that being said another 150Wh/day energy availability (what you can expect from a Spoke) using the Energy as a Service model (pay for what you consume) in my opinion would definitely be welcome from all households.

There are already another 100 households that have signed up and want to get connected to Energy as a Service from the community, so I suspect they would be happy to get connected as a Spoke.

The community ownership model

An interesting facet of this whole project was the community ownership model. During the implementation phase which was supported by Oxfam, we trained up an LMA (Local Maintenance Agent) with the necessary tools and expertise to engage the community, get signups, maintain the systems and collect revenue from end-users of the mesh-grid.

The LMA gets paid $150 a month and after salary payments, and paying for small fixes such as fuses, tools and fuel for visiting households, currently he has $1,500 in the pot to cover other O&M. This amount in the pot is enough to cover minor fixes such as blown fuses, but it’s not enough to cover the cost of major equipment failures such as battery failures that will eventually take place.

The LMA has done such a great job of showing community members what is possible. At his house, the LMA has a setup of productive appliances, he has a knack for knowing how appliances can be used to meet community needs such as using power tools for fixing and constructing boats. He also has the trust of the community making him a perfect LMA / energy access champion in the community.

That being said I’m personally quite worried regarding the long term sustainability of the project. I can easily see capex replacement costs adding up as batteries start hitting end of life in a few years. Everyone is smiling when the sun is shining and systems are working, but with only $1,500 in the pot it’s going to be hard for the community to ensure everyone’s equipment is working perfectly if these major issues arise.

I would recommend the community actually increase the tariff to a bit more than 2x the current tariff - so to approximately 60c/kWh. And if there are leftover “profits” after the first battery replacement cycle then use those profits to purchase more systems to energise more households. Although implementation of a tariff increase is a huge challenge that would require intervention from the donor and Oxfam, and may not be possible in the regulatory climate of Cambodia.

If we could achieve the tariff increase, I would also recommend connecting additional, if not all the households in the community given there are already more than 100 signups and a high demand for productive use of energy.

How can we replicate this level of productivity in other markets like Nigeria?

One of the things that I’m always seeking to understand is how can we crack the nut of economic productivity along with the energy access that we’re delivering.

Turns out this community, Kbal Toal, has cracked the nut - compared to about 200 - 300Wh/day average load which is what we see in Nigeria, this community had incredibly high productive use of energy - 600Wh/day avg load - power tools, refrigeration, electric cooking along with a very highly functioning local economy. I have 2 major takeaways that I think can be useful going forwards, and even replicated in other markets like Nigeria.

For developers looking for communities that are highly economically sustainable and also for our customer success team, I would recommend applying both of the above lessons. Finding communities with high penetration of diesel generators, and finding high impact LMA champions to empower. Because that in turn can empower the community.

Bonus takeaways relevant to last mile operations from discussion with ACE team

Speaking with Dan from ACE (they’re pushing about 2-3 k cookstoves per month globally) I got a few takeaways from how they’re trying to manage last - mile operations, some of which will be relevant for us →

• Similar to how we were able to see really successful outcomes with a highly trained and effective LMA, ACE leverages a rural sales team that they are proud of empowering - the more the sales team is trained up, the more confident and capable they are, the better the job they do and the better the retention. As a result ACE does regular rural area sales team training to boost motivation, knowledge and retention - they also do regular performance reviews of each of the sales team against their targets on a regular cadence - it seems to be working out well for ACE.

• Because ACE is effectively an asset financing business, when signing up end users, they make sure end users pass the ACE KYC, this means if end users don’t have mobile money they can’t be connected, if they don’t provide a secondary number for someone in the community to vouch for them (ideally a next of kin) then they can’t be connected. These small steps go along way to ensuring higher repayment rates.

• When it comes to ensuring quality of service and after sales is controlled, ACE tries as much as possible to standardise it through their own field app. ACE can only accept deposits and payments through the app, to accept a customer they must first have geo tags for each user and QR code scanning for each product ID. The signup flows, ongoing revenue collections, data synching, and fuel purchases from ACE staff must all be done through the App. Minimising touch points outside the app standardises the experience and minimises theft and poor service.

And some bonus photos to finish it off -

They had a crocodile farm in the community (no electricity from Okra systems was used as far as I could see) !

And a bonus, bonus photo - this is sovannara aka the naughty monk and our very first Cambodian team member who was “walking the plank” cheerfully to the crocodile farm. Don’t fall in mate…

These trips are the moments that I live for - I left Tonle Sap with with a full heart and a healthy dose of inspiration for how we can keep optimising for our efforts elsewhere. Super grateful to Oxfam for supporting this interesting (community owned and floating village) project. Grateful to the community for being so welcoming to us and sharing so much feedback and insights with a bunch of Barangs, and of course especially grateful for the Okra team for working tirelessly to try and solve energy access in a sustainable manner full of twists, turns and ongoing innovations along the way. Onwards and upwards 🚀