Size the pump for the system you actually built.

A pump curve relates flow to head. The headline flow printed on a box is commonly measured under a favorable condition and cannot tell you what reaches the grow beds after vertical lift, pipe friction, fittings, valves, filters, and fouling. Start with the required flow at the point of use, then calculate the system resistance the pump must overcome.[1]

Total dynamic head combines static lift with friction losses. Every design assumption should be visible: pipe diameter and length, elevation, fittings, expected filter condition, bypass flow, and the operating points you intend to verify. Select against the manufacturer's curve for the exact model, not a family brochure.[1]

Rules of thumb can help with an early sketch, but final circulation depends on feed load, tank geometry, solids removal, biofilter design, plant system, oxygen delivery, and species. A universal “tank volumes per hour” number hides those mechanisms. Document who established the design flow and what biological and hydraulic assumptions support it.[2]

Commission the system by measuring actual flow at critical branches. Balance valves and record their positions. Repeat after filters have accumulated a realistic load. A design that works only when every screen is freshly cleaned is a maintenance emergency waiting to happen.

A sump must hold normal operating variation plus the drain-back that occurs when pumps stop, while keeping the pump submerged at the lowest operating level. Mark the normal, low, high, and power-off levels. Test a real shutdown while someone watches every vessel and overflow path.[2]

• Measure delivered flow at the design operating point.
• Record elevations, pipe sizes, fittings, valves, filters, and branch flows.
• Test dirty-filter and partially blocked conditions that are safe to simulate.
• Mark sump operating levels and observe a controlled power-off drain-back.
• Keep an accessible bypass, isolation plan, and spare strategy for service.

A spare pump on a shelf is not redundancy if installation takes longer than the system can tolerate. A battery is not backup power until it has carried the actual load for a measured duration. Decide which function must continue first—often aeration rather than full production flow—then test transfer, runtime, alarms, fuel, charging, and the person responsible.[3]

Climate changes the consequence. Heat can reduce oxygen capacity and raise cooling demand; freezing conditions can damage exposed plumbing; unreliable grids change the value of alternate power. Those are site constraints to calculate, not proof that any single technology is universally sustainable.

A pump schedule should show expected flow at real head, the last measured flow, service events, valve changes, abnormal noise or current draw, and the last backup test. When performance falls, the farm should be able to distinguish fouling, restriction, wear, and changed system demand.