Follow along as we investigate a couple considering buying land and building a home in El Salvador. Before they buy, we’re putting the property through a series of real-world tests, from what’s beneath the ground to whether the land can actually support the home they imagine. Will they ultimately buy it? The file’s still open.
The next investigation starts roughly 150 million kilometers away.
The sun.
There is already electricity at the road. The buyer could connect the house to it and be done. But this house is being designed from scratch, so before deciding how much electricity to buy, there is a better question:
Can the property make enough of its own to get through the rainy season, without building an enormous solar system?
That sounds easy. This is El Salvador. There is sun, there is land, there is a house to power. But sunshine somewhere over El Salvador isn’t the same as sunshine on this property, and it doesn’t arrive evenly all year.
Long-term climate data for the La Libertad coast, derived from NASA’s MERRA-2 dataset, puts average daily solar energy at roughly 6.9 kWh per square meter in March and 4.5 in September. That is a drop of about a third. But it isn’t a drought. The strongest months fall in the dry season, production weakens as the rains take hold, and even the weak end of the year delivers plenty of sunshine to work with.
Regional data can’t see this property, though. Trees, slope, panel angle, temperature and equipment losses all sit between the sun and the electricity. So the first test only establishes that there is enough sunlight to keep going.
The other side of the equation is how much the house will use. There is no electrical plan yet, but the Rainwater investigation already set a scale: a representative roof of 150 square meters. That isn’t a design, but it is enough to model one.
House size turned out not to be the variable that mattered. Air conditioning was. An efficient house built around natural ventilation came in at roughly 9 kWh a day. A full-time house with regular but controlled cooling reached about 18. Heavy air conditioning pushed the model toward 30. For context, SIGET reported that 93% of Salvadoran electricity users averaged under 300 kWh a month in early 2024, less than about 10 kWh a day.
The investigation used the middle case: 18 kWh a day, about 540 kWh a month, or 6,570 a year.
The topographic survey has already mapped the trees, so the buyer doesn’t have to guess what stands between the sun and a ground-mounted array. Most of the heavier tree cover sits away from the open part of the slope, which leaves substantial ground to work with. Shade appears to be a low constraint.
That doesn’t make it irrelevant. A large tree doesn’t need to stand beside a panel to shade it, and the sun’s path shifts with the seasons. So the array starts on the open ground, keeps the significant existing trees, and gets modelled for whatever shadows remain.
If one tree does conflict, cutting it isn’t automatically the answer. On steep land, a mature tree’s roots may be holding the slope and managing water. The question then isn’t whether the buyer wants it gone. It’s whether the hillside needs it. That investigation comes later.
The trees weren’t the obvious problem. The hill might be.
The slope gives the panels good exposure. It also complicates everything else: foundations that must hold through decades of wind and rain, rows spaced so one doesn’t shade the next, access for cleaning and repair, cable runs back to the house. And on this property, each of those decisions touches water. The Drainage investigation established what intense rain does once it starts moving downhill. Grading a large platform into the slope, compacting the soil and stripping the vegetation would make that worse, and concentrated runoff would scour the very foundations holding the panels up.
The alternative is to let the array follow the hill. Smaller sections step down the existing terrain on posts or engineered foundations whose exposed height changes as the ground falls away, with racks holding the panels at the angle the solar model calls for. The ground follows the hill while the panels follow the sun. Vegetation can stay between and beneath the rows.
Whether that works depends on one number the survey can supply, how quickly the ground falls across the proposed area, and on the soil, which has conveniently already been tested. The same report that decided whether the house could be supported gives an engineer what is needed to anchor the structure.
The goal isn’t just to hold the panels up. It’s to leave the hill functioning as a hill.
Sizing the array to a year of electricity treats twelve months of sunshine like one long sunny day. But the house doesn’t consume electricity annually. It consumes it every day, including in September, when rainfall in the regional data averages more than 400 millimeters.
So the property went into a photovoltaic production model. For each kilowatt of installed capacity, PVGIS estimated roughly 150 kWh in March, 118 in September and 1,576 across the year. Divide the house’s 540 kWh a month by September’s 118 and the array needs about 4.6 kWp. With 450-watt panels, that lands just above ten.
Ten became eleven, about 4.95 kWp.
That is a better answer than sizing to the annual average, which would leave the house leaning on the street every rainy season. Built for the weakest month, the array does its job when the sun is least generous. And eleven panels is a small footprint. The aim was never to cover the slope, only to find the smallest practical layout that meets the need, with room around it to grow if a well pump or another large load arrives later.
The obvious answer to a weak rainy season isn’t a bigger battery. A battery doesn’t make electricity. It only moves it from one hour to another. More panels might.
Monthly averages hide one more problem. A rainy month isn’t thirty equally cloudy days, and several poor ones can arrive together in the middle of an otherwise adequate month. That is a job for the battery.
At sunset the panels stop. The house doesn’t. Refrigeration, lights, cooking, cooling and water all carry on.
An 18 kWh battery would still be the wrong calculation, because the battery doesn’t have to carry everything. The pool pump, laundry and other discretionary loads can move into the solar day. If a well is eventually drilled, its pump can fill a tank while the sun is shining, storing solar energy as water instead of in a battery. Some cooling can shift into daylight too. The cheapest kilowatt-hour to store is the one the battery never has to hold.
What remains is the overnight load, and the battery has to cover it in usable capacity, after inverter losses and a reserve. Start with one normal night. Then one bad solar day. Then the expensive question: what about two, or three? Each extra day of independence costs far more storage than the last, and eventually the buyer would be paying for a large battery bank to solve a problem the property already has another answer for.
That answer is the wire at the road. Not as the primary source, but as the backup.
It raises a question the buyer hadn’t needed to ask: how reliable is it? The buyer’s own experience of outages is an observation, not data. Regulators measure it with two numbers: SAIFI, how often the average customer loses power, and SAIDI, how many hours in total. SIGET’s figures for 2020 put rural customers at 7.08 interruptions and 26.41 hours without electricity, and urban customers at 3.97 and 8.48.
Those numbers are historical and don’t describe this circuit. But they mean the buyer doesn’t have to guess. The distributor can be identified, current reliability checked and the infrastructure at the road inspected. Where is the nearest pole and transformer? What serves the neighbors? How far is the run to the house? What does the distributor say is actually available? Published rules put small-demand low-voltage service at 10 kilowatts or less, and a new connection requires an application and an inspected installation. Only then does “electricity at the road” become an electrical connection rather than a line on a checklist.
One more catch: panels don’t keep a house running in a blackout by default. A conventional grid-tied system shuts down when the grid does. If outage protection is part of the reason for installing a battery, the system has to be designed from the start to disconnect safely and keep supplying selected loads.
Electricity also flows the other way. On a bright afternoon, with the battery charged, the tanks full and the pool pump finished, the panels are still producing. Where does the surplus go?
In October 2025, El Salvador approved Decree 436, the Ley de Fomento para el Uso de la Energía Renovable, creating a framework for renewable self-consumption. By 2026, an official process was running through the Dirección General de Energía, Hidrocarburos y Minas, with homes, batteries and grid connection all included. The system being considered here isn’t an unusual workaround. There is now a formal path for it.
But the word at the center of that framework is autoconsumo. It is built for producing power for the property, not for turning the hillside into a miniature power station. And there is a major difference between being allowed to connect solar to the grid and being paid for what goes into it. The first is clear. The second, covering surplus, metering and billing, has to be confirmed for the specific installation before the system is sized.
That changes the question one last time. Not how much can the hill hold, but how much should the property make?
The pieces now behave like one system. Solar first, doing the daytime work. Battery when the sun isn’t enough. Street when the first two can’t do the job. Water tanks and batteries both store the sun. The grid hasn’t disappeared. Its job has changed.
The goal was never to disconnect from the street. It’s to need it less.
BUILDABILITY CHECK
The investigation didn’t find a reason solar couldn’t work here. There is enough solar resource. There appears to be enough usable ground. And the property doesn’t have to choose between solar and the electricity already at the road.
What remains is engineering: final array size and layout, foundations, battery capacity, the electrical connection and the treatment of surplus generation. Those decisions may change the size of the system. They don’t change the finding.
SOLAR — BUILDABILITY: PASS
The property doesn’t need to become independent of the grid. It can become far less dependent on it.
The investigation started 150 million kilometres away, with the sun. It ended on a hillside in El Salvador. But one part of the solar field has deliberately been left alone: the ground underneath it.
If it doesn’t need to be cleared, it doesn’t need to be wasted.
So what should grow there?
INVESTIGATION 06 · THE SEPTIC—COMPLETE
2 QUESTIONS REMAIN
This is Investigation 05 of CAN I BUILD HERE?, a Build El Salvador series following the real-world due diligence behind a potential land purchase in El Salvador. Interested in sponsoring this investigation or a future installment? Become a Build El Salvador sponsor