Why Direct-Drive Solar Pump Flow Changes Throughout the Day: Irradiance, Pump Curves, and Affinity Laws

Key Formula / Key Facts Box

Governing relationship for a direct-drive centrifugal pump on a near-fixed system curve: Q(t) ≈ Q_ref × [P_PV(t) / P_ref]^1/3, with P_PV(t) ≈ η_sys × A × G(t) × (1 − L).

This article uses the most common encyclopedic scope: a battery-free, directly coupled photovoltaic (PV) pumping system in which motor-pump speed follows available solar power. Battery-buffered, grid-tied, and diesel-hybrid systems are mentioned only for comparison.

Plain-English restatement: if available solar power falls by half, a centrifugal pump on a fixed system curve does not produce half flow; it produces roughly 0.5^1/3 = 0.79 of its reference flow, until it reaches the minimum power needed to keep turning and lifting water.

Overview — What It Is and Why It Matters

A direct-drive solar pump converts sunlight into water flow without a battery between the array and the motor. The array makes DC power, an MPPT controller or inverter tracks the array’s maximum power point, and the motor-pump set turns at whatever speed can be supported by the available power. When irradiance rises, the pump speeds up; when a cloud passes, it slows down; when irradiance falls below the minimum starting power, it stops. The water tank, not a battery, is normally the energy buffer.

The flow therefore changes throughout the day because two relationships are coupled. First, PV power follows solar irradiance, temperature, soiling, and inverter or MPPT behavior. Second, a centrifugal pump’s flow, head, and shaft power follow the affinity laws as speed changes. FAO describes stand-alone direct-irrigation systems as starting in the morning when solar output exceeds the minimum power required by the pump, running while sufficient irradiation is available, and stopping before sunset when output falls below that threshold; flow changes with solar irradiation during the day. RPS Solar Pumps describes the same field behavior as a morning ramp-up, a midday plateau, and an afternoon taper, with clouds, smoke, and dirty panels causing additional variation.

The engineering consequences are practical. If a designer assumes rated flow for 24 hours, daily volume can be overestimated by a factor of four or more. If an irrigation system is sized only for midday flow, low-light hours may deliver non-uniform pressure. If the pump is oversized relative to the array, it may never start or may operate only briefly near noon. If the array is oversized without checking pump maximum speed, extra power may not produce extra flow. Understanding why direct-drive solar pump flow changes throughout the day is therefore central to pump selection, array sizing, storage planning, and operating schedules.

The Method — Derivation and Variants

Step 1 — PV power from irradiance. At a given moment, the array’s DC power can be approximated as P_PV(t) = P_STC × [G(t) / G_STC] × η_derate, where P_STC is the array rating at standard test conditions, G_STC is 1000 W/m², and η_derate accounts for temperature, soiling, wiring, mismatch, and controller losses. Equivalently, P_PV(t) = η_sys × A × G(t) × (1 − L).

Step 2 — Shaft power. The motor and controller deliver P_shaft = η_drive × P_PV. Some of the electrical power becomes heat in the motor, controller, and cable; only the remainder reaches the pump shaft.

Step 3 — Hydraulic power. The water power is P_hyd = ρ g Q H in SI units. In US customary units, P_hyd = Q × H × SG / 3960, where P_hyd is in horsepower, Q is in gpm, H is in ft, and SG is specific gravity; for water SG = 1. The pump shaft power is P_shaft = P_hyd / η_p, where η_p is pump efficiency at the operating point.

Step 4 — Affinity laws and flow. For a geometrically similar centrifugal pump, Q ∝ N, H ∝ N², and P_shaft ∝ N³. If the system curve is approximately fixed and efficiency is roughly constant, eliminating speed gives Q ∝ P_shaft^1/3, and therefore Q(t) / Q_ref ≈ [P_PV(t) / P_ref]^1/3. This cube-root relationship is why flow changes less dramatically than power at high irradiance, but it also explains why flow collapses quickly near the minimum operating power.

The cube-root variant is a useful first estimate for centrifugal pumps, but it is not universal. A system with large static head may show a steeper drop in flow at low irradiance because the pump must first overcome lift before useful flow begins. A positive-displacement pump behaves differently: flow is more nearly proportional to speed, and torque and pressure limits matter more. MPPT controllers improve energy harvest by holding the array near its maximum power point, but they do not eliminate irradiance-driven flow changes. Suehrcke, Appelbaum, and Reshef showed that pump similarity laws can be applied to variable-speed centrifugal pump performance in PV systems, with solar radiation and pipeline characteristics jointly determining flow.

Worked Example

Example 1 — SI units, centrifugal direct-drive pump. A 3.0 kWp array feeds an MPPT controller and a centrifugal pump. Assume η_sys = 0.75, a rated point of Q_ref = 12 m³/h at H = 30 m, and P_ref = 2.25 kW electrical. At the rated point, the hydraulic power is P_hyd = 1000 × 9.81 × (12/3600) × 30 = 981 W. With pump efficiency 0.55 and drive efficiency 0.85, the electrical input is about 981 / (0.55 × 0.85) = 2098 W, consistent with the assumed reference.

At G = 600 W/m²: P_PV = 0.75 × 3.0 kW × (600/1000) = 1.35 kW. The power ratio is 1.35 / 2.25 = 0.600. The estimated flow is Q = 12 × 0.600^1/3 = 12 × 0.843 = 10.1 m³/h. At G = 900 W/m²: P_PV = 2.025 kW, ratio = 0.900, and Q = 12 × 0.900^1/3 = 11.6 m³/h. At G = 300 W/m²: P_PV = 0.675 kW, ratio = 0.300, and Q = 12 × 0.300^1/3 = 8.0 m³/h, although the pump may stop if this is below its minimum operating power. If the site has 5 effective peak sun hours at the reference point, daily volume is roughly 12 m³/h × 5 h = 60 m³/day before cloud and soiling derating.

Example 2 — US customary units, centrifugal direct-drive pump. A 4.0 kWp array feeds a pump rated at Q_ref = 50 gpm and H = 100 ft, with P_ref = 2.0 kW electrical and η_sys = 0.75. Check the hydraulic power: P_hyd = 50 × 100 / 3960 = 1.26 hp for water. With pump efficiency 0.55 and drive efficiency 0.85, electrical input is 1.26 / (0.55 × 0.85) = 2.69 hp = 2.01 kW, consistent with the reference.

At G = 700 W/m²: P_PV = 0.75 × 4.0 × 0.7 = 2.1 kW. Ratio = 2.1 / 2.0 = 1.05. Q = 50 × 1.05^1/3 = 50 × 1.016 = 50.8 gpm. At G = 500 W/m²: P_PV = 1.5 kW, ratio = 0.75, and Q = 50 × 0.75^1/3 = 45.5 gpm. At G = 300 W/m²: P_PV = 0.9 kW, ratio = 0.45, and Q = 50 × 0.45^1/3 = 38.3 gpm, subject to the pump’s minimum-power cutoff. A 5-peak-sun-hour day at 50 gpm gives about 50 gpm × 60 min/h × 5 h = 15,000 gal/day reference volume.

Calculator

For hydraulic head, flow, and pump power estimates, use the PumpCalcs Total Dynamic Head calculator: http://pumpcalcs.com/calculators/total-dynamic-head/. Pair the result with local peak-sun-hour data, array derating, and the manufacturer’s pump curve. The calculator estimates hydraulic duty; it does not directly model PV irradiance, MPPT tracking, or motor speed limits.

Reference Values & Typical Ranges

Rule of thumb: daily volume ≈ flow at total dynamic head × peak sun hours. Do not multiply rated flow by 24 hours; the pump is not a constant-speed machine when directly coupled to PV.

Application Guidance

In real systems, the daily flow curve is smooth rather than a simple on/off square wave. Irradiance rises after sunrise, the motor accelerates when power exceeds static-head and friction losses, flow climbs toward a midday plateau, and the pump tapers or stops before sunset. Clouds produce short reductions and recoveries; panel soiling and high module temperature reduce the whole curve; seasonal sun paths change both the peak and the duration. Because direct-drive systems use power as it is generated, water storage or a tank at elevation is usually the practical way to decouple pumping from demand.

For selection, match the pump curve to the actual total dynamic head, not to a shallow test-bench condition. Check the pump’s minimum starting power, maximum speed, and allowable dry-run condition. A common array-sizing rule is PV watts at least about 1.3 times pump watts, but the final check must use the manufacturer’s controller and motor limits. Oversizing the array can extend the low-light operating window, but it will not necessarily increase midday flow if the pump has reached maximum speed. For high-head, low-flow wells, positive-displacement pumps may be preferable; for lower-head, higher-flow applications, centrifugal pumps are common. In drip irrigation, variable flow and pressure can affect emission uniformity, so pressure-compensating emitters, a storage tank, or scheduled subunit operation may be needed. Clean panels periodically, protect against dry running, and use float switches to stop pumping when the tank is full.

Common Mistakes, Limits & Safety Notes

1. Using rated flow for 24 hours. Daily volume should be estimated from effective peak sun hours and the pump curve at the actual total dynamic head, not from a continuous-duty assumption.
2. Ignoring static head. The cube-root approximation is optimistic when a large static lift must be overcome; low-irradiance flow may be lower or zero.
3. Mixing US and SI constants. The 3960 denominator applies to gpm, ft, and horsepower for water; SI calculations use ρ g Q H with Q in m³/s and H in m.
4. Assuming MPPT removes variability. MPPT improves harvest but the array still produces less power under cloud, soot, smoke, or low sun angle.
5. Sizing the pump without the array. An undersized array may never reach the pump’s minimum starting power; an oversized pump may run only near noon.
6. Oversizing the array without checking pump limits. Extra PV power cannot force a pump beyond its maximum speed or hydraulic limit.
7. Neglecting soiling and temperature. Dirty panels and hot modules reduce power; RPS recommends cleaning roughly quarterly or every couple of months in many field conditions.
8. Applying centrifugal affinity laws to positive-displacement pumps. Positive-displacement flow is more directly tied to speed, and pressure relief protection is essential.
9. Omitting dry-run and tank-full protection. These controls prevent pump damage and overflow; they are operating features, not optional accessories.
10. Safety limits. PV arrays produce hazardous DC voltage whenever illuminated; use lockout/tagout, arc-rated equipment, and qualified electrical work. Water and electricity require grounding, moisture protection, and lightning/surge consideration. Pressurized lines need relief protection, and open wells need fall protection and confined-space procedures.