Can installing Solar Home Systems be financially sustainable in rural Thailand?

By Sam Shrank, 2009 MAP Fellow

Last September I came to work with the Border Green Energy Team (BGET) on a three-month fellowship. My role was to help develop a program that BGET could run independent of grant funding. I wasn’t sure it could be done, but I was excited to apply my experience in evaluating business models toward making my own. Now, entering my last month of what turned into a six-month stay, the problems inherent in making rural electrification financially sustainable are even clearer to me, but I am confident that the scheme I have designed will become a successful venture for BGET.

The first question to ask is why we would want to do this in the first place. The advantage for BGET is obvious—we would be more self-sufficient as an organization and have income that isn’t dependent on receiving grants or contracts from other organizations.

There is also research, however, showing that distributed generation projects that include financial contributions from the end users, which is what financial sustainability would require, are more successful. End users treat the equipment with more care and are more invested, pun intended, in the project. Technology dumps, once popular, are now widely acknowledged as less than ideal.

Almost immediately I zeroed in on solar home systems (SHS) as a promising technology for my project. SHS are household-sized solar power systems, in this case providing enough power for a couple lights and one small appliance. I was particularly drawn to SHS, ironically, because almost every rural household already had one. In 2004 The Thai government decided that they wanted to use SHS to bring electrification to every household not connected to the national grid, and so went to great expense to install almost 200,000. In the intervening years, however, the warranties on these systems have expired (and rural households had a very difficult filing claims anyway) and now the vast majority of these SHS are either partially or completely broken. But in almost every case the solar panel, by far the most expensive piece of equipment, is still functioning, a huge wasted asset.

With this major advantage in hand, I began thinking about how a business like this would operate. The most important consideration is the villagers’ ability and willingness to pay for solar power. Even though rural households in the developing world are often willing to spend more on electricity than people realize, their cash flow is very low. Therefore any work BGET does would have to be paid off over time in small installments. I also realized quickly that it would be critical to limit the number of trips BGET staff would need to make to each village. Even restricting the project’s area to the border districts of Tak province (where BGET currently operates) would include some villages an entire day’s drive away considering the quality of many mountain roads and others not even accessible by car.

These two considerations led me to conclude that such a scheme would need to be run using local franchisees who would do as much of the work—customer recruitment, installation and maintenance, and fee collection—as possible. It also meant that BGET could not simply repair the broken components as we originally planned. The components used by the government were old and cheap, and so even if we fixed or replaced, say, the charge controller, the battery would likely fail before long. With our goal being to travel as little as possible, we could not afford to rely on suspect equipment. Therefore the project evolved from repairing the existing systems to installing our own with components we could trust. By similar logic we also decided that we could not use flooded batteries, which require monthly or ideally even weekly maintenance. Instead we will use maintenance-free batteries that will live up to their advertised lifespan without constant attention.

Both of these decisions increase our equipment cost but we believe that they are necessary for our scheme to work. Using more expensive batteries, we believe, even decreases the amount we will need to charge because we will not need to rely on franchisees for as much maintenance work and we will have to replace fewer batteries. Increasing the lifetime of our equipment is particularly important to our bottom line because we will be renting the systems to customers. In other words, the household pays a monthly fee for the energy service we provide, lighting and power for appliances, but never owns the equipment. If a household decides they no longer want the system or stops paying we take the system away and use it in another location. Therefore we bear the entire cost when a component needs to be replaced but we also get to use each component for its entire lifetime, as even a used piece of equipment can be put into an otherwise new system and simply be replaced when it no longer functions. The main advantage of this arrangement is that we believe it will help attract customers. Households will not be tied down to a long-term ‘mortgage’ and also will more easily be able to compare the cost of the SHS to their current energy expenditures.

Because BGET will bear complete responsibility for the systems, I realized that we needed to make sure that the customers could tamper with the equipment. In many of the households I visited that currently have government SHS systems equipment had been moved, wires had been reconfigured, or components tampered with. There is rarely bad intent, but in this situation protecting our assets is important. For this reason BGET has created a design (not that we are anywhere near the first people to have this idea) where all the electrical equipment is kept within a locked container.

Thus we settled on the scheme as it stands now. We will install a SHS consisting of all BGET equipment save the panel and will retain ownership of the equipment, requiring the households to pay only a monthly fee for the energy services the system provides. We will use local franchisees as sales representatives, installers, repairmen, and fee collectors. But there are many aspects of the scheme that are still unsettled and will go a long way towards determining whether it will be successful.

The first major question is what our relationship with the local governments will be. After the government systems were installed the national government gave ownership of the systems over to the sub-district governments (a sub-district might have between 200-1000 households in it) and so we will need to have some sort of arrangement with them to work on these systems at all. We are hoping, however, to get more than just permission to operate from the local governments, because they could provide significant logistical assistance even if they cannot contribute monetarily. With their local ties they could help us find reliable franchisees, who could even be government employees. Their office would be a convenient location for equipment storage, and government officials could be involved in fee collection. We hope we can secure their help not only because will be running a program that will provide a service to their constituents, but also because we can offer to run the program as a partnership.

We also have yet to determine the best relationship we can establish the franchisees. There are two related questions: what the best way is to compensate the franchisees and how much responsibility we can safely give them. To begin with the latter issue, we are expecting our franchisees to handle customer recruitment, installation, and basic maintenance. But it would perhaps be beneficial to have franchisees be in charge of fee collection and more complicated repairs as well. While it would help us logistically, there are possible concerns about how effectively franchisees could be trained and whether we’d be willing to trust them with large sums of money. These concerns could partially be allayed depending on how the franchisees are compensated. It is important to encourage franchisees to recruit and keep as many customers as possible by aligning their incentives with our own. Therefore we will give them a portion of each monthly fee paid by their customers. The question is whether they will also receive an annual salary. A salary could make the job more attractive and perhaps remove temptation to steal money if franchisees are doing fee collection.

In the coming months we will hopefully be operating a pilot project in one sub-district and we will be able to test out different components and franchisee arrangements. We will also start to get a sense of the biggest unknown, how much households are willing to pay for electricity. Until we get a better idea of what fee we might be able to charge, however, all we can do is try to make the cost as low as possible.

Having made a financial spreadsheet that includes all of our costs and revenues over the first ten theoretical years of the project I am able to see clearly, even without exact values for some variables, where our main expenditures are and how changes to our scheme would change costs. Unsurprisingly equipment costs will be an important element throughout the project’s life, as we will need to continually be buying new equipment. By the later years, however, salary costs for our franchisees will likely far surpass equipment as our main expenditure. The conclusion that can be drawn from this is that, if franchisees are going to be paid any kind of annual salary, we will need to make sure that they each still accumulate a large number of customers. In other words we will still need to make sure their incentive to find customers is strong. To compound the incentive of receiving a portion of customer fees we could also offer prizes to top sellers or offer to raise the annual salary of franchisees who reach certain benchmarks.

Other significant costs include needing to buy a truck at the outset of the project and possibly pay back a loan over time if we need to take a loan out at first. We will probably need a loan because we will incur large equipment costs upfront and only recoup those costs in the form of monthly fees years later. Because we are an organization that pursues funding from project to project, it is important for the project not to stray too far into the red at any point, even if it will be profitable in the long run. Therefore it is important for the scheme to install some systems immediately to begin generating revenue but also grow slowly enough that by the time the money from the initial loan has been spent, enough revenue is coming in from existing systems to pay for further work.

So the ultimate question remains of whether, even after we’ve done all we can to make the systems affordable, we will be able to find customers. The prospect is a little bleaker because these households have previously received solar power for free and therefore and may not want to pay now. We are also, with good justification I believe, using higher quality components than we could and scrapping some existing equipment that could at least temporarily be operable. At this point, without receiving any grants or government assistance, my guess is that our product will only be within the reach of the wealthier village households. This obviously doesn’t mean the scheme isn’t worth running, because it still would provide power to those who do not have it currently, and maybe within a few years, as incomes grow, we might find wider interest. It does mean, however, that we will pursue funding options that will allow us to lower our price.

And even if that means that the project cannot be strictly defined as financially sustainable it will still provide many of the advantages we originally sought. The end users will still have a stake in the project’s success, and BGET will have a stable source of revenue. And, of course, households currently relying on candles or battery-powered headlamps would have access to clean, reliable, solar power, which is what we’re here trying to achieve.

Solar System Installation at Lay Tong Khu Clinic

By Sam Shrank, MAP Fellow 2009

Monday

Monday was a day of driving. We drove from Mae Sot to the last village accessible by car on the way to our final destination, Lay Tong Ku. We were headed there, the three technicians, another volunteer, and myself, to install a solar system in a medical clinic that services both the village of 1,200 and many people who come from Burma specifically to avail themselves of the clinic’s services. The system will power seven lights and a vaccine refrigerator. The refrigerator especially will allow the clinic to expand its services, as it currently has no way to keep vaccines or medications cold.

The drive took about six hours, including a long lunch break at a roadside restaurant near Umpiam refugee camp. I spent the drive dozing, trying to stop my seat cushion from sliding off the seat every time we went down a hill, and reading Che Guevara’s Motorcycle Diaries. I liked the book mostly as a travel story, though it was interesting to see the beginnings of his ideology forming (at exactly my age) and learn more about his life from the introduction.

We stayed at the (very nice) house of our ‘guide’. I watched the technicians play caneball—think volleyball with no hands, except the people here are still able to spike! They wanted me to play but it would have just been embarrassing. I spent the evening playing the Thai version of gin rummy with two of the technicians. Once we cleared up some rules misunderstandings we had a good time. Going on these trips is when I interact with the technicians the most, and that is definitely a highlight. They are—and this is characteristic really of most people in Mae Sot and elsewhere here—very happy and friendly. Even listening to their conversation without understanding anything you notice how much they laugh and how animated they are. And even with their broken English they are always joking around with me about anything and everything.

Tuesday

Tuesday was a day of walking. It took probably 4 hours to get to LTK from the end of the auto road. We went up and down two mountains, but it was the heat that really made the hike hard. There were nice views at times, but most of the time our view was obstructed even at the top of the mountain by the dense vegetation. When there were gaps, though I could see forested mountains in what must have been Burma, and occasionally the outlines of villages. We met a group of maybe 30 Karen men going in the opposite direction. They stopped to talk with us, and I repeatedly heard them say one of the two only Karen words I know: “Tahb-luh” or “thank you.” It turns out they were our porters, men from LTK who were carrying our equipment, especially the behemoth vaccine refrigerator. Though, as we soon found out, they decided the refrigerator was such a behemoth that they could not carry it on the trail, so it would have to be brought to LTK by tractor over a much longer route.

Anyway, the particularly interesting thing about these Karen is that they are not Christian or Buddhist as most are, but have a belief system that is some combination of spiritualism and animism. This leads to many differences. They do not drink alcohol, or eat domesticated animals. Second, the men mostly have very long hair that they tie in a bun at the front of their head similar to a samurai’s topknot. In most ways they are like all other Karen though—incredibly friendly, agile and fearless when it comes to manual labor, fanatical about spicy food and betel nut, and generous with their food and time.

Wednesday

Wednesday was a day of work. The refrigerator still had not arrived but we installed the rest of the system so that if the refrigerator arrives tomorrow we will still be on track to leave Friday first thing in the morning. The five 130W solar panels are also powering six fluorescent lights and one LED light for the clinic, kitchen, outhouse, and doctor’s house. Except for a short time carrying wood in the morning, I spent the first part of my workday helping assemble wire ties (very unskilled labor) and following M (the third technician) around passing him tools as he climbed up to the rafters to hang lights.

Soon though I was doing my own tasks. When you think of installing solar-powered lighting you probably think of setting up the panels, or maybe hanging up lights. But actually the vast majority of the work required is laying and connecting wires and setting up switches. Wire ties have to be nailed into the wooden columns and rafters every three inches or so, differing in size depending on how many wires need to be held in place. There are wires that go from each switch to each light, but also a main line that comes from the batteries/panels, and in some locations that main line must be ties in to lines going to the outer buildings (nurse’s house, outhouse, kitchen). Therefore in some places four wires needed to be run in parallel and tied down. The particularly tedious work is nailing in the ties. I spent most of the time nailing in the 1-wire ties, which have a nail the size of the top of a grain of rice and themselves are less than half the length of my pinky finger. To nail them in you fold them over the nail to hold the nail in place until it is firmly in the wood, but still holding the tie in place with my fat fingers means that there is almost no room for error in where I hit the nail. And because the nails are so small even one hit in the wrong direction will bend the nail severely.

Now I believe I possess many talents but fine motor skills are not chief among them. This task would have been hard enough for me had I been doing it standing on the ground. But most of the time I was in the rafters, sitting either on a wooden crossbeam an inch thick or on the bamboo doorframe (a little thicker but much creakier) of the nurse’s bedroom. Though it was fun to climb all over the buildings and I never felt like I was putting myself in danger, it meant that I could never give my complete focus to my hammering, and increased my mistakes. I later had to climb back up to tie the ties around the wires, which often meant I had to stand precariously on a thin piece of wood without touching the wire so that I could pull it taut.

The other main job I had was making switches, which again is more complicated than it sounds. In the simplest switch, the wire to the load (lamp) is connected to the main line from the power source. The wires must first be stripped to expose the positive and negative wires, which in turn must again be stripped at their ends to reveal the copper that actually conducts the electricity. The negative wires are then twisted together to allow current to flow freely, while the positive ends are screwed into the switch. In this way, when the switch is closed (in other words, the light is turned on) a complete circuit exists, and the light will receive power while when the switch is open there is no circuit.

After we finished work for the day we took a trip to the local waterfall, which must have been 75 feet high or so, full of deep pools, almost vertical drops over craggy rock, and funnels with violently rushing water. I was just looking forward to swimming in the pool at the bottom and taking some pictures until the technicians started literally running up the falls. They ran up the steep parts, climbed up the vertical stretches, and waded through the fastest flows. They could even run and jump down the falls without a second thought. I was persuaded to try to climb the falls myself and so began my very slow ascent. I never felt at all out of control, but that is only because I tested every foothold and handhold obsessively. It was a pretty big thrill when I got to the top, and we all spent maybe 30 minutes fooling around in the water and relaxing. After all, we had definitely earned it.

Dinner was ready soon after we got back from the waterfall and finished showering. I think now is a good time to describe what it is like to eat a Karen meal. Each person is served a plate of rice and each dish is placed in the center and eaten family style. Before eating you use your spoon (or hands when especially traditional or hungry) to break up your rice, which is often clumpy or stuck together. You then eat a bite of plain rice before you combine with anything else, I assume as a check on the rice’s quality. Then you are free to eat from the shared bowls. It is customary to take only a small amount at once, enough for maybe two or three bites. Fish dishes are very common, as are chicken and pork (except in this village). They do have decent selection of green vegetables that they eat, mostly leafy greens that are put in soups. If there is a dish that looks like it might be spicy is almost definitely is—of the 3 or 4 dishes served at a meal there are never more than 2 I can handle and 1 often is only vegetables. I have acquired a taste for many new vegetables in Thailand, including cucumbers, kale, and many leafy greens that many not even have English names. Because all of the meat here is heavy on bone and gristle it is acceptable to pick up the pieces, which are always small, with your hands, or to put them in your mouth and play with it to remove the meat. The bones and waste are put in a pile next to your bowl to be cleared at the end of the meal. I am always completely full after a meal, even if with mostly rice on occasion.

Thursday

Thursday was a day of training. For most of the day that meant that I sat and watched as the technicians trained the nurses and villagers in Karen, following along by the diagrams drawn on the blackboard. I had been informed a couple days ago that I would be doing the refrigerator training and Thursday evening the time finally came. The refrigerator arrived at dusk by way of some long tractor journey in Burma, and we quickly installed it, though we had to break down a wall of the clinic because the refrigerator didn’t fit through the door. Doing this was much easier than it would have been in the US—we just took out a few nails and the vertical wood planks came out, and then we sawed off a piece of the horizontal support and voila, we had a hole big enough for a refrigerator. Anyway, over the past couple days I had been reading the manual and not quite understanding some of the instructions. I was consequently extremely nervous about this training, considering that not only was I going to have to explain everything, but I would have to do so with a translator that spoke only relatively simple English (for example did not know the words dial, mold, or metal). As soon as I got a look at the refrigerator and the electronic controls on the top, though, everything clicked into place. The diagrams that were unclear suddenly made perfect sense and I felt a lot better going into the training. Though I had some translation problems, as far as I can tell it went off without a hitch. If we get a call in a week saying that they tried to fix a broken thermostat with a sledgehammer I’ll need to revisit my confidence, but for now I am happy with how things went.

The other interesting part of the day was our visit to a local shrine during the afternoon. As I mentioned before most of the people in this area are adherents of a spirit religion of which I do not know specifics. At this shrine they worship a pair of enormous elephant tusks, intricately carved with animals and other designs. I’d never seen tusks even close to as big as the ones in this shrine. There was also a Buddha in the shrine, underneath the tusks. I asked our guide if they worshipped the Buddha as well. He said no, that it was only there so that Buddhists who come to visit “feel more comfortable.” Bizarre, no?

Friday

Friday was a day of traveling home. We woke up at 5 am to get an early enough start to get home by the evening and to avoid hiking during the hottest part of the day. We left by about 630—part of the delay was due to our discovery that one of the two porters who were coming with us to carry our equipment was a strict adherent to the spirit religion of the area and therefore refused to carry—or even touch—the backpacks holding the gear. Maybe he thought that there were some animal products in it (leather?) and so it was related to not being allowed to eat meat? He brought a bamboo basket, however, and they loaded as much of the gear as they could into it. I volunteered to carry the backpack along with some of the lighter gear. Only after a couple minutes did I discover that one of the arm straps was broken, and held in place by being tied to the chest strap. This meant the bag was hopelessly lopsided and I couldn’t tighten the arms. Still the bag wasn’t too heavy and having a waist strap was nice, so it wasn’t much of an issue.

Probably because we left so early, the hike out wasn’t nearly as tiring as the hike in, and I think was quicker as well. When we got back to the truck (after an incredible meal at the house where we spent the first night) I volunteered to sit in the truck bed because it would have been really crowded in the back seat with three people and I like feeling the wind and seeing the view. I really enjoyed the ride, though it was incredibly long (almost 7 hours) considering I had nothing to do but look at the view. The mountain scenery was incredible, especially as the sun was setting, and I got great views of villages and refugee camps. For the first couple hours the roads were extremely bad, and sometimes we hit potholes that made me feel like I’d punctured a lung, but once we hit the highway it was smooth sailing.