YARDPUMP · Technical Article

Which Pump Delivers More When Its Curve Crosses Another Pump's Curve?

Read both pumps against the same system curve. Water level and hose resistance can reverse which one delivers more.

Imagine a buyer comparing two candidates for the same drain. They overlay the head and flow curves of two small submersible pumps. Pump A offers more head toward the low flow side; Pump B has the higher curve on the right. Which one moves more water from the same pit? Compare each pump curve with the same system curve. The intersection predicts its operating flow and head for that arrangement. A different water level or discharge route can reverse the ranking. This is why selecting from a maximum flow number alone is a poor way to fill the purchase specification.

Make sure the curves describe comparable versions

Use curves for the actual frequency, electrical arrangement, wet end and liquid conditions. Confirm the axes and units, and whether each curve is a typical curve, a reading from one sample or backed by a guaranteed range. If Pump A was measured through its carton hose and Pump B at a bare discharge in a different rig, the plotted results need a common basis before comparison. For the system, record static head, flow dependent losses and any required outlet condition. Static head can change as the source level falls. Hose, fittings and flow affect losses. An open discharge and a nozzle with a required pressure do not have the same outlet condition.

Run two illustrative systems

Take a hypothetical open discharge example with Q in litres per minute and head in metres. Near the points of interest, let the simplified curves be H = 8 − 0.10Q for Pump A and H = 6 − 0.05Q for Pump B. They cross at 40 L/min and 4 m. For an example system with H = 1 + 0.001Q², Pump A intersects at about 47.5 L/min and Pump B at 50.0 L/min. If the static head increases to 3 m and the loss term stays the same, the respective flows become about 36.6 L/min and 35.2 L/min. Neither formula represents a product or fits a whole curve. Use the illustration to see why the conclusion changes, and compare real pumps within their approved operating ranges.

Bring the supplier both high and low pit levels, the real hose bore and length, fittings and the outlet requirement. Check at least the common condition and the most demanding one that still belongs to the intended duty. Near the crossover, test uncertainty and production variation can be larger than the apparent difference. If necessary, ask for both models to be tested on the same rig and distinguish minimum supported flow from typical observed flow. A maximum or average can conceal the slow final part of a drawdown. If the customer wants a known amount emptied in a set time, consider the changing water level across that job.

Turn the comparison into a decision

Write down whether the job is to clear a high pit quickly, reach a lower water line, or serve two discharge conditions. Use the same hydraulic basis for both candidates and record a minimum acceptable flow for each condition. Then check the liquid, cooling, float stop position and installation space. A good curve cannot make up for unsuitable intake or control conditions. If both pumps perform the duty and their estimated difference is small, compare the installation and spare options. If only one installation was verified, limit the conclusion to that installation.

A useful supplier response includes enough working points to show the curve on either side of the relevant crossing. If the curves touch only because of rounding or are close through part of the range, request actual test results instead of declaring a winner from the plot.

Frequently asked questions

Is the curve with more head to the left better in every installation?

No. Look at the intersection on each system and confirm the usable operating range.

Can I draw a straight line between maximum head and maximum flow?

Those endpoints cannot describe the actual curve near the duty. Obtain working points there.

Should I choose a larger model when the two predicted results barely differ?

First check the required minimum and the uncertainty. Then consider the real fittings and operating conditions.

Related reading