Solar Panel Tilt and Azimuth for Water Pumping: Aim the Array at the Season, Not the Sun

The short answer: set the array facing the equator at roughly your latitude in tilt — then break that rule on purpose, because a pump is not a rooftop grid. A pumping array answers to one question: does the water arrive when the crop, the herd or the household needs it? If your season is summer irrigation, go 10-15 degrees flatter than latitude to bank the high sun. If your season is the winter/dry-season window — where the sun sits low exactly when water is scarcest — go 10-15 degrees steeper. And if your problem is the daily curve rather than the annual one, split the array east-west and buy yourself an earlier start. Every rule below is built from the same raw material: peak sun hours by month, a 1.3x controller margin, and the specific model on the end of the pipe.

The baseline rule and why a pump gets to break it

Every generic solar guide repeats the same sentence: tilt the panels at your latitude, face the equator, and the array will collect close to the annual maximum. That advice is engineered for a different customer. A rooftop grid-tie system is paid for kilowatt-hours averaged across twelve months, so it optimizes the mean.

A solar pump is not paid for the annual mean. It is paid in water delivered during the season that matters, and the seasons rarely cooperate:

• Irrigation demand peaks in the hot, sunny months — which is also when the sun is high. A flatter array (latitude minus 10-15 degrees) tilts the annual curve toward exactly those months.
• Livestock and household demand is flat all year, but the dry season is when the well and the trough both run hardest — and in the tropics and subtropics the dry season coincides with a low sun. A steeper array (latitude plus 10-15 degrees) meets it.
• The worst month is the design month. Solar resource varies far less between months in the tropics than in Europe or North America, but “far less” is not zero — and the pump must start on the darkest morning of the darkest month, not on the annual average.

That is the whole logic. The table below turns it into numbers you can set with an inclinometer.

The tilt table: pick your water calendar first

Three reading notes:

1. Azimuth beats fine-tuned tilt. An array 10 degrees off true north/south loses around 1-2 percent of annual yield; an array tilted 10 degrees wrong for its season loses far more of the season that matters. Get the compass right first, then argue about the angle.
2. Magnetic versus true north is a real correction, not trivia. Depending on where you are, a compass can be off by more than 10 degrees from true north. Correct for local declination or set azimuth by the solar-noon shadow.
3. Row shading is the silent array killer at high tilt. At a 50-degree tilt on 30-degree latitude (a hard dry-season setting), the array throws a long winter shadow; leave row spacing at roughly 2 to 3 times the panel height so the front row never shades the back at the winter solstice.

The worst-month example: a real 750 W pump under a Sahel sky

Take a concrete machine: the 3DSC6-60-48-750 — 750 W, 6 m3/h maximum flow, 60 m maximum head, 1.25-inch outlet. The catalogue rule printed next to the A/D and HV controllers says the array must be at least 1.3 times the pump power. That gives a minimum array of 750 × 1.3 = 975 W, which in the real market of ~550 W modules means two panels, 1,100 W, and in a market stocked with 400-450 W modules means three.

Now the question the tilt table actually answers: does 1,100 W of glass start this pump on the worst morning of the year? NASA POWER long-term data for Niamey, Niger (latitude ~13.5°N) shows the shape of a Sahel solar year:

The spread between the best and worst month in Niamey is only about 14 percent — this is what “tropics are flat” means in practice. But the design question is never the average: it is whether the worst month clears the 1.3x floor with room for dust, temperature derating and cable loss. At 5.6 peak sun hours it does — barely, on paper, with a clean array at the correct tilt. Which is exactly why tilt matters: tilting steeper for the December sun, when the Harmattan dust is also easiest to wash off a steeper glass, is the cheapest kilowatt-hour you will ever buy.

The same logic scales up the ladder. A 4/6DSC36-38-300-2200-A/D — 2,200 W, 36 m3/h, 38 m head — demands a minimum array of 2,200 × 1.3 = 2,860 W before tilt is even discussed; at that wattage the array is a construction with rows, and row spacing rules from the table above stop being optional.

Facing east and west: buying shape, paying energy

An equator-facing array has one hump: a sharp peak at solar noon and weak shoulders. A pump, however, does not always want a hump.

• A hump is right when the tank is large relative to daily demand and the pump can take its time — fill from 9 to 3, done.
• A hump is wrong when the pumping day is squeezed: short winter days, afternoon-only water draws, or a controller whose start threshold needs voltage the flat-lying morning sun cannot deliver until mid-morning.

The east-west split — half the modules tilted toward the sunrise, half toward the sunset — trades about 10 percent of total daily energy for a curve with a broad plateau and an earlier start. On a pump feeding a schedule, an earlier start is not cosmetic: it is the difference between the trough being full before the herd arrives and a queue of thirsty cattle standing in front of a half-empty tank.

One hard prohibition: do not mix orientations on one string. Two panels facing east and two facing south in series share one current — and the current collapses to the weakest orientation’s. Series-strings must share a compass direction; only a multi-input or appropriately configured controller can treat two orientations as separate inputs.

Before combining orientations, check how solar pump array shading and bypass diodes can change the usable pumping window.

Why fixed beats trackers for water

A single-axis tracker promises 25-35 percent more annual yield, and on a grid farm it can be worth the mechanics. On a pumping array, it almost never is:

The tracker’s paper gain quietly assumes it tracks. When a tracker sticks in February, nobody notices a percentage — they notice the tank. Extra fixed panels at the correct tilt deliver the same margin with nothing that can seize, and they simultaneously feed the controller’s 1.3x floor. For a water asset that must run unattended for a decade, fixed wins on arithmetic that does not require a single bad day to matter.

The commissioning checklist

1. ☐ Compass corrected for declination — array azimuth within a few degrees of true equator-facing (or deliberately rotated ~10° west where afternoons matter).
2. ☐ Tilt set per the water calendar — flatter than latitude for summer irrigation, steeper for dry-season/low-sun pumping, baseline at latitude for year-round even demand.
3. ☐ Row spacing ≥ 2-3 × panel height for tilted rows so winter shadows never land on the next row.
4. ☐ Array wattage ≥ 1.3 × pump rated power verified at the controller’s input — the 3DSC6-60-48-750 at 975 W minimum, the 4/6DSC36-38-300-2200-A/D at 2,860 W minimum, before dust and derating.
5. ☐ One orientation per string — east-west splits land on separate controller inputs, never in series.
6. ☐ No trackers — the budget went into fixed watts instead.
7. ☐ Worst-month check done — the darkest month’s peak sun hours clear the 1.3x floor with dust and temperature margin.

Once the array is aimed, the rest of the system has its own geometry: the string-length and high-voltage window for larger pumps is covered in the array and string design guide, and the controller’s role in turning that morning voltage into a clean pump start lives in the controller guide.

Not sure which side of the tilt table your water calendar falls on? Send me the site latitude, the month your water demand peaks, and the pump model — I will return the exact tilt, azimuth and array size. Or put your flow and head into the sizing tool yourself. For a straight answer in one message, WhatsApp me.