Pump Selection Software: What It Can and Cannot Decide

Reader question: “Does anyone know of a good pump selection program that I can use for importing pumps and performing selections on them?”

Yes—but treat software as a calculation and documentation tool, not as the final engineering decision. A program can organize pump curves, calculate system resistance and compare candidates. It cannot make a purchase-ready industrial selection when the duty data, liquid properties, suction conditions, materials, seal requirements or installation constraints are missing.

AI-generated concept illustration: a generic process pump beside an abstract curve-analysis screen and duty-sheet workflow.

Image disclosure: AI-generated concept illustration. It is not a screenshot of real selection software, a field photograph, a pump test result, or a supplier selection approval.

Start by choosing the job the software must do

Tool type	Useful for	Not enough by itself for
System-calculation or energy-assessment tool	Building a system curve, estimating operating point and documenting scenarios	Selecting materials, seals, physical fit or a supplier’s final configuration.
Manufacturer catalogue selector	Narrowing that manufacturer’s catalogue to a hydraulic candidate	Comparing another maker’s data without a common data basis.
Spreadsheet or internal database	Normalizing units, preserving assumptions and comparing quoted options	Replacing verified performance curves, drawings and engineering judgement.
Supplier application review	Checking a defined duty against a documented configuration	Inventing missing process information on the buyer’s behalf.

The DOE’s Pumping System Assessment Tool (PSAT) illustrates the right scope. DOE says PSAT uses achievable pump-performance data from Hydraulic Institute standards and motor data to calculate potential energy and cost savings; it also stores system curves for analysis sharing. That is valuable system analysis—not proof that every catalogue import is suitable for every process liquid or installation. DOE Better Plants, updated page accessed 25 August 2026

The minimum data pack before any program can give a useful result

Do not begin with the product catalogue. Begin with the process.

Flow range: minimum, normal, maximum and duration at each case.
Pressure/head basis: source and destination pressures, elevations, pipe sizes/lengths, fittings, equipment pressure drops and control-valve cases.
Liquid basis: composition, concentration, density, viscosity, temperature, vapour pressure, entrained gas and solids description where applicable.
Suction basis: tank pressure/level range, suction layout, strainer condition and calculated losses.
Mechanical basis: power supply, preferred speed control, area/site constraints, nozzle/baseplate dimensions and maintenance limits.
Containment basis: wetted materials, seal arrangement, gaskets/elastomers and any customer/project requirements.
Acceptance basis: curve format, inspection/test requirement, drawings, data book and documentation expected with the quote.

If the software asks only for flow and head, it is making a preliminary hydraulic screen—not a complete process-pump selection.

What the math should show

For centrifugal pumps, the installed duty is found where the pump curve intersects the system curve. DOE describes the system curve as the sum of static and frictional head components. DOE, Identifying Energy Savings Opportunities in Industrial Pumping Systems, 2006

A competent workflow should therefore let you:

preserve the assumptions behind each system curve;
plot minimum, normal and maximum operating cases;
overlay head, efficiency, power and NPSH-required curves for each candidate;
record pump speed and impeller diameter rather than comparing anonymous points; and
export the data and assumptions for independent review.

For preliminary speed changes, affinity-law relationships are commonly used to estimate the change in flow, head and shaft power with rotational speed. A 2023 peer-reviewed pump-simulator paper presents the familiar proportional relationships and warns that the standard relationship is not fully applicable for excessively viscous liquids. Treat affinity-law output as a checked estimate, not permission to ignore liquid effects. Bartík et al., Processes, April 2023

The verification step software cannot skip

ISO 9906:2012 specifies hydraulic performance acceptance tests for rotodynamic pumps and limits its stated application to liquids that behave as clean, cold water. ISO 9906:2012, published May 2012

That is why the final selection file must identify:

the actual liquid versus the curve-test liquid;
who reviewed material and seal suitability;
which duty cases were checked for NPSH;
the exact model, speed, impeller and driver quoted; and
the required acceptance/documentation package.

A simple, auditable workflow

Step	Output
1. Collect the data pack	A dated duty sheet with unknowns clearly marked.
2. Model the system	System curves for each real operating case.
3. Import only traceable pump data	Curve source, speed, impeller diameter, revision and units retained.
4. Screen candidates	Duty range, power and NPSH checked against the same model.
5. Review construction	Materials, seal, dimensions and auxiliary needs confirmed for the actual service.
6. Freeze the selection	Quote, drawing, test/inspection requirement and acceptance criteria agreed.

Where YSM fits

YSM does not ask customers to choose a chemical-process pump from a flow/head number alone. Send the completed data pack and existing-equipment information, if this is a replacement. We can then review whether a configuration from the Summit 2196 ANSI Pumps or D Mark III ANSI chemical-process-pump product lines is a documented candidate and return the questions that still require customer or process-engineer confirmation.

The best selection program is the one that makes assumptions visible, retains the source curves and exposes missing inputs before a purchase order—not the one that produces the fastest model name.