Solar pumps in Bangladesh are usually bought by horsepower and then disappoint in March. How to size on total head and daily water demand, choose surface against submersible, and secure an array standing alone in a field.
A cooperative buys a 5 HP solar pump because the diesel engine it replaces was 5 HP. The array goes up in December, the discharge at the field channel is fine, and by the third week of March it has fallen to about half of what the old diesel set delivered. Nothing has broken. The pump was matched to a horsepower badge rather than to the depth the water has fallen to and the head the delivery line demands, and by late Boro both have moved against it.
Bangladesh's agriculture sector depends on approximately 1.5 million irrigation pumps, the majority powered by diesel engines or grid electricity. During the dry Boro rice season, December to April, these pumps run for 8 to 12 hours per day, consuming enormous quantities of diesel and drawing heavily on the rural grid. Solar removes the fuel cost outright and generates at exactly the hours irrigation demand peaks. The sizing method, though, is not the rooftop sizing method, and neither are the failure modes.
A solar pumping system consists of PV panels, a solar pump controller (MPPT based), and a DC or AC submersible or surface pump. During daylight hours, the panels generate DC electricity that powers the pump directly. No battery is required for daytime-only irrigation. For farms that need to pump at night or store water, an elevated storage tank (gravity-fed distribution) or battery storage can be added.
The work a pump does is flow multiplied by total dynamic head, and total dynamic head is not the depth of the borehole. It is the lift from the pumping water level, which sits below the rest level once drawdown starts, up to the discharge point, plus every metre lost to friction in the rising main and delivery line. A design quoting a rest level measured in September will be short of head by the following March, and the shortfall appears as reduced flow rather than as an alarm.
The bands below are what most suppliers quote against. The head assumption underneath them decides whether a given array actually delivers the output shown.
| Pump size | Solar panels needed | Daily water output | Suitable for | System cost |
|---|---|---|---|---|
| 1 HP (0.75 kW) | 1.5 kWp | 40,000–60,000 litres | 1–2 acres | BDT 1,20,000–1,80,000 |
| 3 HP (2.2 kW) | 3.5 kWp | 1,00,000–1,50,000 litres | 3–5 acres | BDT 2,80,000–4,00,000 |
| 5 HP (3.7 kW) | 5.5 kWp | 1,50,000–2,50,000 litres | 5–10 acres | BDT 4,50,000–6,50,000 |
| 10 HP (7.5 kW) | 10 kWp | 3,00,000–5,00,000 litres | 10–20 acres | BDT 8,00,000–12,00,000 |
A surface pump does not lift water. It lowers the pressure above it and lets the atmosphere push, which puts a ceiling of roughly ten metres of water column on the theory and considerably less on a real site once pipe friction, a hot afternoon and a slightly leaky foot valve have taken their share. Where the pumping level stays shallow, a surface centrifugal set on a shallow tubewell is cheaper and easier to repair. Where the dry season table falls away, that pump loses prime in March and the answer has to be a submersible sitting below the water, pushing rather than pulling.
The instinct on a solar project is to add a battery. On an irrigation project it is usually wrong. An overhead tank stores the same energy as head: no cycle life, no temperature derating, no replacement bill in year five, and it gives a drip or sprinkler line the steady gravity pressure a direct-drive pump cannot hold through a cloudy afternoon. A battery earns its place where night pumping is genuinely required, where the array also carries a homestead load after dark, or where the water is for drinking supply. Otherwise it adds the shortest-lived and most stealable object on site to an unattended installation.
A factory rooftop plant works against a flat weekday load all year. An irrigation array is fully loaded for one season and idle for much of the rest, and its design point sits at the worst combination of high head and high demand, which is the tail of Boro rather than the middle of it.
| Season | Roughly | Irrigation demand | Design consequence |
|---|---|---|---|
| Boro | December to April | Highest, near continuous | The design point. Size on the March pumping level, not the December one |
| Aus | March to July | Moderate, supplementary | Overlaps the tail of Boro, so peak head and residual demand coincide |
| Aman | July to November | Low, dry spells only | Under-used array and monsoon cloud. Plan a second load or accept idle capacity |
| Rabi vegetables and pulses | November to February | Light and frequent | Small volumes, short intervals: suits a tank and drip layout |
That idle stretch is worth designing for. The same array can carry a homestead load, a small cold room at the collection point or a drinking water supply. Where the site has a utility connection and the distribution company permits it, a net-metered arrangement can absorb the surplus, but a standalone field pump usually has no connection and no way to export. Settle this before the controller is specified, not after the array is up.
A 5 HP diesel pump running 8 hours per day during the 120-day Boro season consumes approximately 1,200 litres of diesel, costing approximately BDT 2.5–3 lakh at current diesel prices. A solar pump of equivalent capacity costs BDT 4,50,000–6,50,000 installed, giving a payback of 1.5–2.5 years on diesel savings alone. After payback, the pump runs at zero fuel cost for 20+ years.
What moves that arithmetic is running hours, the distance fuel has to be carried, and whether the pump serves one holding or sells water to neighbouring plots. Around most deep tubewells the operator is in practice a water seller with a command area, which changes the question from how much land you own to how much land the pump must serve.
The Infrastructure Development Company Limited (IDCOL) has a solar irrigation pump programme that provides financing and partial subsidy for solar pump installations in Bangladesh. Under this programme, farmers can access solar pumps with a reduced upfront payment and instalment financing. The Bangladesh Agricultural Development Corporation (BADC) and Department of Agricultural Extension (DAE) also support solar pump adoption through extension services and demonstration projects. The surface and submersible sets Vvon supplies as Solar Irrigation Pump are BADC and DAE scheme compatible, with automatic water level control and remote monitoring. Before approaching any of them, assemble the command area boundaries, the crop plan, the existing fuel spend with receipts, the land tenure position and the water charge arrangement with neighbouring plots.
A factory array sits behind a gate with a guard on it. An irrigation array stands in an open field, and every component on it has a resale value. This is the part of the design missing from most quotations, and it is why so many rural solar assets stop working for reasons unconnected to the technology.
The institutional side matters as much as the hardware. An asset with no named owner gets stripped. Committee ownership with a paid operator whose income depends on the water charge survives far better than a system handed to a village in general.