MIMAH Logo

Solar Irrigation System Design for Farms: From Crop Water to PV Array

Author

Hisham Abdalla

Date Published

Illustration of a farm with drip irrigated fields, a canal, a solar array and a raised water reservoir, with small figures walking between the rows

Disclaimer: Research and analysis by the engineering team. Sources referenced below.

A solar irrigation system is designed backwards from the crop. Crop water demand in the hottest month sets the daily volume, the irrigation method sets the pressure, volume and pressure set the pump power, and pump power with the local sun sets the array. On a farm of 1 to 50 hectares the first surprise is the volume: a fully grown vegetable crop in a hot, dry climate uses roughly 60 to 90 cubic metres per hectare per day, before any irrigation losses. The second is that the water source, not the array, often limits how much land you can plant.

This article follows the design in order and works one 5 hectare farm from evapotranspiration to array size. It builds on our general design method for solar water pumping, and the cost side is in our solar versus diesel analysis.

Start With the Crop, Not the Array

FAO Irrigation and Drainage Paper 56 is the standard method for crop water demand. It works in two steps. The reference evapotranspiration, ETo, is the water a standard grass surface would lose in the local weather, in millimetres a day. The crop evapotranspiration is ETc = Kc x ETo, where Kc is a crop coefficient for the crop and growth stage.

For ETo, FAO-56 gives typical ranges by climate in its introductory chapter: 6 to 8 mm a day for the arid and semi-arid tropics when the mean temperature is above 30 degrees Celsius. FAO says those ranges are for orientation only, so a real design should calculate ETo from local weather records.

For Kc, Table 12 lists mid-season values of 1.15 for tomato and 1.05 for dry onion. Those values are for sub-humid conditions with a minimum humidity near 45 percent and winds near 2 m/s, and FAO-56 raises them in drier, windier places. At 20 percent minimum humidity and the same wind, its Equation 62 takes a 0.6 m tomato crop from 1.15 to about 1.21, so the demand below is, if anything, low.

The conversion to volume is simple. One millimetre of water over one hectare is 10 cubic metres. A crop using 8 mm a day therefore needs 80 m3 per hectare per day. ETo of 6 to 8 mm with these two Kc values gives 6.3 to 9.2 mm, which is the 60 to 90 m3 per hectare above.

Choose the Irrigation Method: It Sets the Pressure

The method decides how much pressure the pump must add, and pressure is array. The GIZ and FAO Toolbox on Solar Powered Irrigation Systems gives typical figures. Flood irrigation applies water at 40 to 50 percent efficiency at about 0.5 m of head, and the Toolbox says furrow and other surface methods do not pass 60 percent. Open canals reach 50 to 60 percent at 0.5 to 1 m. Sprinklers reach 70 to 80 percent but need 10 to 20 m. Low-pressure drip is about 80 percent efficient at 1 to 10 m, and high-pressure drip 85 to 95 percent at 10 to 100 m.

The Toolbox concludes that low-pressure drip suits solar pumping best, that flood irrigation by solar pump is barely cost-efficient, and that sprinklers are seldom used with solar pumps because their pressure demands a large and costly array. We agree. For a farm of 1 to 50 hectares we would start with low-pressure drip fed from a raised reservoir, and move to sprinklers only for a crop that cannot be drip irrigated and where the pressure cost has been counted. A centre pivot is a sprinkler system and sits at the expensive end.

Irrigation methods compared by typical application efficiency and head: flood 40 to 50 percent at about 0.5 m, open canals 50 to 60 percent at 0.5 to 1 m, sprinkler 70 to 80 percent at 10 to 20 m, low pressure drip about 80 percent at 1 to 10 m, and high pressure drip 85 to 95 percent at 10 to 100 m

Typical application efficiency and head by irrigation method. Low pressure drip suits solar pumping best. Source: GIZ and FAO, Toolbox on Solar Powered Irrigation Systems, 2018.

From Crop Water to Daily Volume in the Peak Month

Net demand is the crop's ETc summed over its area. Gross demand is what you must pump: net demand divided by the application efficiency of the method. In the dry season, with no effective rain, net equals the crop's need.

Keep the two apart, because drip changes one and barely touches the other. A crop near full cover transpires what the weather asks of it, and FAO-56 says Kc mid is less affected by wetting because little bare soil is left to evaporate. So at peak a full tomato crop needs about the same net water under drip as under a furrow.

Drip earns its saving elsewhere. Early in the season it wets only part of the ground: in an FAO-56 example, a trickle system wetting 40 percent of the surface takes the initial Kc from 1.05 to 0.42. It also loses far less on the way to the root. At the 77.7 m3 per hectare net of the farm below, drip at 80 percent means pumping 97 m3 per hectare, where flood at 45 percent would mean 173. Ask of any published peak figure whether it is net or gross.

The peak month matters more than the average. The Toolbox notes that peak irrigation demand is often more than twice the average. A system sized to the average leaves crops short in the hottest weeks, when they need water most. The cheapest way to shrink the array is a crop calendar that staggers planting, so the mid-season stages of different crops do not fall in the same month.

Fit the Pumping Hours to the Sun

A solar pump works the sun's shift. Peak sun hours is the day's solar energy on the array in kWh per square metre, treated as hours at full intensity. Divide the daily volume by the peak sun hours and you have the flow at full sun.

For a design month, use data for your site, and take the month with the worst ratio of water demand to sun. NASA's POWER database gives a site near Kano in northern Nigeria 6.56 kWh/m2 a day in April and 5.50 in December on a horizontal surface, with 6.00 as the annual mean. For a hot-season design month we use 6.0 and not 6.56. The World Bank derate we apply below allows only 5 percent each for dirt and heat, which is light for a hot, dusty month, so the difference is margin for both.

A tank decouples the pump from the field. For irrigation we would size it to hours of demand, not the two to three days that a village supply needs, because soil moisture is already a store and an irrigation round can slip a day. Storage is real money, so every cubic metre should earn its place. Our tank sizing guide covers the sizing in detail.

A Worked Example: 5 Hectares From Water to Array

Suppose a farm of 5 hectares beside a canal in the northern savanna: 3 hectares of tomato and 2 hectares of dry onion, both at mid-season in the hottest month, fed by low-pressure drip from a reservoir. The assumptions are stated so you can recompute them.

One of them is deliberately harsh. Mid-season Kc applies to a full stand, and FAO-56 gives tomato about 40 days at that stage in a 135-day crop. The whole farm only sits there at once if both crops went in together, so treat this as the worst case and not a typical day.

Crop water. ETo is 7.0 mm a day, the middle of the FAO-56 range of 6 to 8. Tomato: 7.0 x 1.15 = 8.05 mm a day. Onion: 7.0 x 1.05 = 7.35 mm a day.

Net volume. Tomato: 3 ha x 8.05 x 10 = 241.5 m3. Onion: 2 ha x 7.35 x 10 = 147.0 m3. Together 388.5 m3 a day, or 77.7 m3 per hectare.

Gross volume. Low-pressure drip is about 80 percent efficient, so 388.5 / 0.80 = 486 m3 a day, which is 97 m3 per hectare.

Design flow. With 6.0 peak sun hours, 486 / 6.0 = 81 m3/h, or 22.5 litres a second, at full sun.

Head. The pump sits on the bank with the water 3 m below it. Low-pressure drip works from 1 to 10 m, so we allow 8 m at the header, 3 m for the filter and main line and 1 m margin. The reservoir water level then sits 12 m above the field. Add 3 m of pipe friction: 3 + 12 + 3 = 18 m.

Energy and array. Hydraulic energy is 486 x 18 / 367 = 23.8 kWh a day. We assume a wire-to-water efficiency of 45 percent, so the array must deliver 23.8 / 0.45 = 53.0 kWh a day. We use an array derate of 0.774, the figure in the World Bank's pumping guide, which also puts a pump's best efficiency at 60 percent. The array is 53.0 / (6.0 x 0.774) = 11.4 kWp. We would specify 12 kWp.

Cross-check. The Toolbox gives array watts peak as 8 x head x daily volume / daily irradiation: 8 x 18 x 486 / 6.0 = 11,664 W. It agrees.

Pump. At 81 m3/h and 18 m, hydraulic power is 81 / 3,600 x 9,810 x 18 = 3.97 kW, so electrical input is 3.97 / 0.45 = 8.8 kW. The 12 kWp array gives 12 x 0.774 = 9.3 kW at full sun, which covers it.

Reservoir. As a working figure we take one third of the daily volume, 486 / 3 = 162 m3.

A 5 hectare worked example: crop water of 8.05 mm for tomato and 7.35 mm for onion gives 388.5 cubic metres net, 486 gross at 80 percent efficiency, 18 metres of head, 53.0 kWh a day of electrical energy and an array of 11.4 kWp, specified as 12 kWp

A hypothetical 5 hectare farm, 3 ha of tomato and 2 ha of onion, from crop water to array size, with the whole area at mid-season at once as a worst case. Assumptions: ETo 7.0 mm a day, 6.0 peak sun hours, wire-to-water efficiency 45 percent, array derate 0.774. Source: FAO-56 Table 2 and Table 12; GIZ and FAO Toolbox; MIMAH engineering practice.

The array is about 2.3 kWp per hectare, driven by the peak month. A farm that staggers its crops or irrigates less thirsty ones needs less.

Check the reservoir against your ground. On a natural rise, 162 m3 with its water 12 m above the field is cheap to place. On flat land it means a serious tower, so price a lower reservoir, or a lower drip pressure with shorter runs, before you commit.

Now see what pressure costs. Rerun the farm with sprinklers. At 75 percent efficiency, the middle of the 70 to 80 percent range, the gross volume is 388.5 / 0.75 = 518 m3 a day. Take 20 m of nozzle pressure, the top of the Toolbox range, with the same 3 m lift and 3 m of friction: 26 m of head. The array becomes 518 x 26 / 367 / 0.45 / (6.0 x 0.774) = 17.6 kWp, about 54 percent bigger than the 11.4 kWp for drip.

Drip Design in Practice: Zones, Pressure and Fertigation

A reservoir gives drip a steady pressure, because gravity does not follow the clouds. The Toolbox notes the price of that choice: the field must be divided into small sections so the flow is shared out under control. Size each zone so the header never serves more emitters than it can supply.

Connecting the pump directly to the drip lines saves the reservoir but loads the system with changing flow and pressure through the day. The Toolbox lists the remedies as automatic valves, smaller fields and solar tracking. We prefer the reservoir where the ground allows it, because it removes the problem instead of managing it.

Fertigation needs a decision of its own. The Toolbox does not recommend a Venturi injector on a solar pump, because it loses a lot of pressure and the pump's pressure is not constant, so the fertiliser concentration would swing. It suggests water-driven dosing pumps, or the simplest method of all for low-pressure drip: mixing soluble fertiliser into the reservoir. Flush the lines with clean water after liquid fertiliser to limit clogging.

Water Sources: Borehole, River or Canal

The example needed 486 m3 a day, and the source must provide it in the sun hours.

Borehole. A borehole is limited by its safe yield. Suppose the pump is capped at 12 m3/h and runs for 7 hours: 84 m3 a day. At 97 m3 per hectare, that supplies 84 / 97 = 0.87 hectare at peak. That is why the pumping test comes before the array. Our guide to well yield testing explains how to measure it, and our article on solar pump cost shows how the yield and head shape the price.

River or canal. A canal can supply 22.5 litres a second where a borehole cannot. The trade-offs are level changes between seasons, silt, and the right to abstract. The Toolbox lists water rights and concessions among the first things to confirm, and says surface water needs filtering much more often than borehole water.

Whichever source you choose, check the supply in the driest weeks, not at the best moment. The Toolbox makes availability of water the first consideration for any design.

Check the Water: Salinity and Filtration

Irrigation water quality is a design input. FAO Irrigation and Drainage Paper 29 sets out guideline limits. Electrical conductivity of the water below 0.7 dS/m (total dissolved solids below 450 mg/l) carries no restriction on use. From 0.7 to 3.0 dS/m (450 to 2,000 mg/l) the restriction is slight to moderate, and above 3.0 dS/m (above 2,000 mg/l) it is severe. Test the water before you choose crops. Saline water also needs extra volume to carry salt past the roots, which the example above leaves out.

Drip emitters clog. The Toolbox calls a filter essential in any drip or sprinkler system. It advises against screen filters on solar pumps because their pressure loss is high. Sand filters need high pressure to operate and backflush, which makes them less suitable, and automatic backflush systems do not suit a low-pressure design either. It recommends disc filters, fitted one size larger than the flow suggests or two in parallel, to reduce pressure loss.

Water checks before design: salinity limits of 0.7 and 3.0 dS/m from FAO Irrigation and Drainage Paper 29, and filter guidance for solar pumps recommending disc filters, advising against screen filters and rating sand filters and automatic backflush as less suitable

Irrigation water salinity guideline limits and filter choices for a solar pumped drip system. Source: FAO Irrigation and Drainage Paper 29; GIZ and FAO, Toolbox on Solar Powered Irrigation Systems, 2018.

Do Not Pump the Aquifer Dry

A solar pump costs almost nothing to run, so nothing in the bill tells you to stop. The World Bank's guide warns that near-zero marginal pumping cost can lead to excessive groundwater extraction. The Toolbox defines over-exploitation as pumping beyond long-term recharge.

The protection is practical. Meter the water. Design to the crop's requirement, not to the most the borehole will give. Cap the pump's maximum output below the safe yield. Measure the static level each year, and repeat the step test, as our yield testing guide describes.

What to Decide Next

Take the decisions in this order. Fix the crop plan and calendar, and find the peak month. Choose low-pressure drip unless you have a reason not to. Test the source and check the water quality. Compute the volume, head and array as above, and check the result with the Toolbox rule. Then ask for quotes on one sheet.

Our guide to solar for agriculture in Nigeria covers the wider farm picture, including cold storage and financing. If you are planning a farm system and want the numbers checked, our renewable energy team can work through the design with your crop plan, water source and site data. Contact us with the area, crops, water source and head.