Solar Water Pump Sizing: How to Get the Numbers Right Before You Buy Anything

Most solar pumps that disappoint were not badly made. They were badly sized.

The failure comes in two flavours. The undersized system delivers water for four hours a day in the dry season, exactly when the farm needs it for ten, and the diesel pump comes back out of the shed. The oversized system pulls a weak borehole down below the pump intake by mid-morning, runs dry, trips, cools, restarts, and repeats until either the motor or the borehole gives up. Both failures were fully predictable from numbers that could have been collected in a week, before any money changed hands.

This article is the sizing methodology in full: how to work out daily water demand, how to build up total dynamic head from its parts, how to convert litres and metres into pump power and array size using peak sun hours, why the borehole test matters more than any brochure, and why a water tank beats a battery bank in almost every case. It closes with a complete worked example, 30 cubic metres a day at 60 metres of head, carried through every step.

If you want the broader picture first (what solar pumping is, where it beats diesel, what a system costs), start with our overview of solar water pumping systems and come back. This piece is the deep dive on getting the size right.

Start With Water, Not Watts

Every correct sizing exercise starts at the tap, the trough or the field, never at the panel. The first number you need is daily water demand in cubic metres per day, and it has to be honest about the worst period of the year, not the average.

For people, demand is population multiplied by a per-capita supply standard. Rural supply standards commonly sit around 20 to 30 litres per person per day; the World Bank's solar pumping guidance uses 30 litres per capita per day as its planning figure, so a village of 2,000 people needs at least 60 m³/day. Add an allowance for population growth over the design life, because the system will still be pumping in ten years.

For livestock, demand is herd size multiplied by per-head consumption, which varies with animal and climate. Cattle in hot climates can take 40 litres or more per head per day; goats and sheep far less. Livestock demand also swings hard with season, so size for the hot months.

For irrigation, the calculation is genuinely more involved, and this is where most farm systems go wrong. Crop water requirement depends on the reference evapotranspiration at your site, the crop, its growth stage and the irrigation method, and the standard method for computing it is set out in FAO Irrigation and Drainage Paper 56. As a rough field orientation, peak-season crop water needs in hot arid climates commonly land in the range of 50 to 80 m³ per hectare per day for surface-irrigated field crops, and considerably less under drip. But treat those as orientation only. An agronomist's demand estimate costs a fraction of an oversized array, and the World Bank's guidance is blunt that irrigation demand is best determined by one.

Whatever the use, the output of this step is a single design figure: the required volume on the worst-case day, in m³/day. Everything downstream hangs off it.

Total Dynamic Head: Where Sizing Goes Wrong First

The second number is total dynamic head (TDH), the total resistance the pump must overcome, expressed in metres. It is not the depth of the borehole, and treating it as such is the single most common sizing error we see. TDH is the sum of four parts.

Static water level is the depth from ground level down to the water surface in the borehole when nothing is pumping. It is measured, not assumed, and it moves with the seasons; a level taken at the end of the rains can be optimistic by many metres compared with late dry season.

Drawdown is how much further the water level drops once you start pumping. Pull water out faster than the aquifer feeds the borehole and the level falls until it finds a new equilibrium. Drawdown depends on the pumping rate, which is why it must come from a proper pumping test at something like your intended design flow, not from the driller's memory. A borehole might draw down 5 metres at 5 m³/hour and 10 metres at 10 m³/hour; the relationship is rarely linear and always site-specific.

Delivery head is the height above ground to which the water must be lifted at the far end, typically the inlet of an elevated storage tank, usually 5 to 10 metres up so the network can run on gravity.

Friction losses account for water rubbing against the inside of the pipework, fittings and non-return valves. They rise steeply with flow rate and fall with pipe diameter, and are read off head-loss charts for the actual pipe run. For a competently sized pipe, a working allowance of roughly 10 per cent on top of the physical lift is the standard first-pass figure; a long horizontal run to a distant tank, or pipe chosen one size too small to save money, can push it far higher.

Add the four together at the design flow rate and you have TDH. Get it wrong by 20 per cent and the pump operates off its curve, the flow prediction fails, and the array is the wrong size, all at once.

Total dynamic head is a sum of four components, not the drilled depth of the borehole. A 20 per cent error here puts the pump off its curve and the array at the wrong size. Source: MIMAH engineering sizing method.

The Energy Balance: Litres and Metres Into Watts

With demand and TDH fixed, sizing becomes an energy balance: the electrical energy the array delivers in a day, degraded by every loss between sunlight and shaft, must equal the hydraulic energy needed to lift the day's water.

Hydraulic energy is the physics end and is mercifully simple. Lifting Q cubic metres per day through a head of H metres requires Q × H × 1,000 × 9.81 / 3,600,000 kilowatt-hours per day. A convenient shorthand: hydraulic kWh/day is Q × H divided by 367.

Wire-to-water efficiency converts that into electrical energy at the controller input. It bundles the motor, the pump end and the drop cable, and for a well-matched submersible set operating near its best-efficiency point it tops out around 60 per cent; away from the ideal operating point, 40 to 50 per cent is more representative. Divide the hydraulic energy by this figure to get the electrical energy the array must deliver each day.

Array derating covers everything between the module nameplate and the controller. Modules are rated at standard test conditions that a dusty 45-degree afternoon in Sudan does not resemble. The World Bank's sizing guidance stacks the typical factors: manufacturing tolerance, soiling around 95 per cent, temperature around 95 per cent, cable and power-conditioning losses, arriving at an overall subsystem efficiency of roughly 77 per cent. In other words, a 1 kWp array behaves like 770 W of deliverable capacity before cloud is even considered.

Array size then falls out directly: required electrical energy per day, divided by (0.77 × peak sun hours), gives the array rating in kWp. The pump and motor rating follows from the design flow rate and TDH via the manufacturer's curves, which we come to below.

The logic is the same energy-balance discipline we apply when sizing a commercial solar system for a business; the only difference is that here the load is water instead of kilowatt-hours on a bill.

Peak Sun Hours and Designing for the Worst Month

Peak sun hours are simply daily solar insolation expressed in kWh/m²/day: a site receiving 5.5 kWh/m²/day has 5.5 peak sun hours, as though the sun delivered full standard intensity for 5.5 hours and nothing the rest of the day. This is the number that turns an energy requirement into an array size, and also the number that sets the design flow rate: daily demand divided by peak sun hours gives the flow the pump must sustain while the sun is up. A 30 m³/day system at 5.5 peak sun hours must pump about 5.5 m³/hour, roughly double the flow rate a mains-powered pump running 12 hours would need for the same daily volume. Solar pumps work a short, intense shift.

The critical design decision is which month's insolation to use. Sizing on the annual average produces a system that fails to meet demand for several months of the year. The conservative discipline is worst-month design: take the month with the lowest insolation at your chosen panel tilt, check the water demand in that same month, and size for the worse combination. In much of Sudan, Egypt and northern Nigeria the resource is generous, typically in the range of 5 to 6.5 peak sun hours even in weaker months, which is exactly why the region is such natural solar pumping territory. But the Harmattan season in Nigeria, with its dust haze, and cloudy stretches on the coast can knock the resource down meaningfully, and demand for irrigation often peaks in a different month than the solar minimum. The design month is wherever the ratio of demand to sun is worst.

Tilt matters here too. Fixing the array at a tilt that favours the winter sun raises worst-month output at a small cost to the annual total, which is usually the right trade for a water system, where the penalty for a bad month is an empty tank rather than a slightly smaller annual figure.

Sizing a pumping system for a farm, a factory or a water scheme? Our renewable energy team designs and installs solar water pumping systems from 2 to 200 HP, and the sizing study comes before any equipment is specified.

The Borehole Sets the Ceiling

Everything so far assumed the water source can supply the design flow. That assumption kills more solar pumping systems than any electrical fault.

A borehole has a tested safe yield, the rate it can sustain without the water level collapsing. The design flow rate must sit below it with margin, full stop. The temptation runs the other way: solar pumps must move the daily volume in five or six sun hours, so the design flow is already double what a longer-running pump would need, and a buyer who then "adds a bit of safety" by going one pump size up can end up demanding two or three times what the aquifer will give.

What follows is mechanical and hydrogeological damage at the same time. The water level is pulled down to the pump intake and the pump runs dry or gulps air; submersible motors are cooled by the water flowing past them, so interrupted flow cooks the windings. Deep drawdown also degrades the borehole itself: the World Bank's guidance notes that overpumping drives chemical changes in the aquifer, oxidising iron compounds and forming ochre that clogs the pump and screen, raising service costs and shortening the life of both pump and borehole. And because a solar pump's marginal running cost is effectively zero, there is no fuel bill to discourage overuse; IRENA's policy work on solar irrigation flags exactly this risk of groundwater over-abstraction at programme scale.

The practical rules are short. Commission a proper pumping test before sizing anything. Set the design flow below the tested yield. Fit the controller's dry-run protection and set it, because it ships disabled or misconfigured more often than anyone admits. And if the borehole cannot yield the daily demand in the available sun hours, the answer is a second borehole or reduced demand, never a bigger pump. We covered the wider pattern of installations undone by shortcuts like this in why solar systems fail early.

Tanks Are the Battery

A solar pump without storage delivers water only while the sun shines. The question is whether to buffer with batteries or with water, and for pumping the answer is almost always water.

An elevated tank is a battery with no electronics, no charge controller, no degradation curve and a service life measured in decades. Energy is stored as lifted water and recovered by gravity, for free, at night and on cloudy mornings. Batteries add cost, heat-sensitive chemistry and a replacement cycle to a system whose entire appeal is low operating cost; they earn their place only in niche cases such as constant-pressure household supply or where an elevated tank is genuinely impossible.

Size the tank at two to three days of daily demand. That is the World Bank's planning rule, and its field surveys found that undersized tanks are a common defect in real installations: the tank overflows at midday, wasting pumped water, then runs out in the evening. For our 30 m³/day example that means a 60 to 90 m³ tank, and the delivery head to its inlet is already sitting inside the TDH calculation from earlier. The tank also buys operational slack: a cloudy day, a controller trip or a maintenance stop no longer means a village or a herd without water that evening.

Why pumping systems store water rather than electricity, and the two-to-three-day tank sizing rule. Undersized tanks overflow at midday and run dry by evening. Source: Solar Pumping: The Basics, World Bank, 2018; MIMAH engineering practice.

Matching Pump, Motor and Controller

With flow, head and array size known, the equipment has to be selected as a matched set, not as three separate purchases.

Pump type follows the duty point. Centrifugal pumps suit high flows at low to medium heads, broadly 10 to 120 metres; positive displacement types, and helical rotor pumps in particular, suit lower flows at medium to high heads, roughly 30 to 250 metres. Helical rotor pumps have a further advantage under solar power: they keep pumping efficiently at low speed, whereas a centrifugal pump below its threshold speed delivers nothing at all, which matters on hazy and intermittently cloudy days.

Motor choice follows system size. DC motors couple naturally to PV arrays with minimal power conditioning and are economical at the small end, but are generally unavailable above about 5 kW. Larger systems use AC motors fed by a solar inverter or variable-speed pump controller. Long cable runs at low DC voltage lose unacceptable power, so array placement and motor choice interact.

The controller is not an accessory. It performs maximum power point tracking so the array delivers what it can as irradiance swings, soft-starts the motor, and carries the protections: dry-run cutout from a water sensor, over- and under-voltage, and locked-rotor. The pump and motor pair should be verified against the manufacturer's performance curves, characterised to IEC 62253, at your actual TDH and design flow, not read off a headline "maximum head" figure that the pump only achieves at zero flow.

A set matched this way runs for years on almost nothing: panel cleaning, an annual inspection of the drop cable and clamps, and periodic checks of drawdown against the original test.

A Worked Example: 30 m³/day at 60 m Head

Now the whole method, end to end, for a representative farm borehole system in the Sudan or northern Nigeria solar belt.

The brief. A horticultural farm needs 30 m³/day in the peak irrigation month. The pumping test shows a static water level of 40 m, a drawdown of 9 m at 5.5 m³/hour, and a sustainable yield comfortably above 6 m³/hour. Water goes to a tank whose inlet is 5 m above ground. Worst-month insolation at the chosen tilt is 5.5 kWh/m²/day, so 5.5 peak sun hours.

Design flow. 30 m³/day across 5.5 sun hours is 5.5 m³/hour. That sits below the tested yield with margin, so the borehole passes.

Total dynamic head. Static 40 m, plus drawdown 9 m at design flow, plus delivery head 5 m, gives 54 m of physical lift. Add 10 per cent for pipe friction: 59.4 m. Call it 60 m TDH.

Hydraulic energy. 30 × 60 / 367 = 4.9 kWh/day. That is the physics floor; no system can do the job on less.

Electrical energy. Assume a realistic wire-to-water efficiency of 50 per cent for a well-matched submersible set: 4.9 / 0.50 = 9.8 kWh/day at the controller input.

Array size. 9.8 kWh/day divided by (0.77 × 5.5 sun hours) = 2.3 kWp. Specify 2.5 kWp to hold margin for soiling between cleans and module ageing.

Pump and motor. Hydraulic power at the duty point is 1,000 × 9.81 × (5.5/3,600) × 60, about 0.9 kW. At 50 per cent wire-to-water efficiency the electrical input is roughly 1.8 kW, so a nominal 2.2 kW (3 HP) solar submersible, selected from curves to deliver 5.5 m³/hour at 60 m, is the right class of machine. Note the head: at 60 m we are in the band where a helical rotor pump deserves consideration alongside a multistage centrifugal, especially given hazy-season performance.

Storage. Two to three days of demand: a 60 to 90 m³ tank.

The instructive part is what happens if one input is sloppy. Assume the borehole is "about 50 m deep so call TDH 50 m" and the array comes out at 1.9 kWp, undersized by nearly 20 per cent; the farm loses a fifth of its water in the worst month. Guess the wire-to-water efficiency at 60 per cent when the real matched point is 45 per cent and the same thing happens again. Small input errors do not average out; they multiply.

Weighing a solar pump against the diesel it replaces? Run your own numbers through our solar payback calculator to see what the energy cost difference does over a ten-year horizon.

What one sloppy input does to a sizing. Assuming head from drilled depth rather than measuring total dynamic head undersizes the array by nearly 20 per cent. Source: MIMAH worked example.

Frequently Asked Questions

Can I just oversize the pump to be safe? No, and with a borehole source it is actively harmful. A pump larger than the tested yield drags the water level down to the intake, causing dry-running, motor overheating and long-term damage to the borehole screen and aquifer chemistry. Safety margin belongs in the array (a few extra per cent of kWp) and in the tank (an extra day of storage), never in pumping capacity beyond the source's tested yield.

What are peak sun hours and where do I find them for my site? Peak sun hours equal your site's daily solar insolation in kWh/m²/day. Long-term monthly averages are available from public satellite-derived databases and national meteorological records, and any competent designer will pull site-specific monthly figures at your panel tilt rather than a single annual number. In most of Sudan, Egypt and northern Nigeria the worst-month figure still sits around 5 to 6 sun hours, which is a generous starting point by world standards.

Do I need batteries with a solar water pump? Almost never. An elevated storage tank holding two to three days of demand stores the same energy as lifted water, costs less over life, and has no replacement cycle. Batteries are worth considering only for constant-pressure applications or where an elevated tank is impossible. If a vendor's first proposal includes a battery bank for a straightforward borehole-to-tank scheme, ask why.

My borehole cannot yield my daily demand within the sun hours. What are my options? In order of preference: reduce demand (drip instead of flood irrigation changes the arithmetic dramatically), add storage and accept a lower delivered volume, or develop a second water source and split the duty. What is not an option is a bigger pump on the same borehole; the yield is a property of the aquifer, and no pump changes it.

How accurate are manufacturers' free sizing tools? The reputable ones are good at their core job, matching their own pump curves to your stated flow, head and insolation, and serious suppliers should present a simulated monthly water output with their offer. But the tools are only as good as the inputs, and the inputs are exactly the things this article covers: honest worst-month demand, a measured TDH with real drawdown, and a tested yield. Garbage in remains garbage out, whatever the software.

Roughly what pump size do I need per hectare of irrigation? There is no honest single number, because crop, climate, season and irrigation method move the answer by a factor of five or more. As orientation, peak-season surface-irrigated field crops in hot arid climates can demand 50 to 80 m³ per hectare per day, while drip-irrigated horticulture may need far less. Compute it properly from crop water requirements per FAO's standard method, or have an agronomist do it, before sizing anything electrical.

Get the Sizing Right Once

A solar pumping system is a twenty-year asset whose entire economics were decided in the week the sizing was done. Undersize it and the shortfall recurs every dry season for two decades. Oversize it against a weak borehole and the repair bills recur instead. The sizing study, a pumping test, an honest demand figure, a TDH built from measurements and a worst-month energy balance, is the cheapest engineering the project will ever buy.

We do this work across the region: MIMAH designs and installs solar water pumping systems from 2 to 200 HP, with operations in Nigeria, Egypt, Sudan and the UK, and the method above is the one our engineers apply on real boreholes, not a textbook exercise.

Planning a solar pumping project and want the sizing done properly? Talk to our engineering team. Bring your pumping test data if you have it; if you do not, that is the first thing we will organise.