YARDPUMP · Buying Guide

How Do You Audit a Pump Test Spreadsheet? Check Raw Values, Formulas and Revisions

Audit measured and calculated values, units, formula references, sample identity and revisions in pump test spreadsheets before accepting the results.

A purchasing team receives an Excel pump report containing curves, power, speed and efficiency. Is a complete-looking workbook sufficient for acceptance? In this typical review situation, first identify measured values, manual settings and calculated results. Then check units, sample identity, formula references and revision control. A polished chart is a presentation of evidence, not a substitute for a traceable measurement and calculation chain.

Give each value an identity

Original readings should connect to an instrument, sample and operating condition. Voltage, frequency, water temperature, flow, pressure and speed are not automatically simultaneous merely because they occupy the same row. Identify any manually entered correction, assumption or default separately. Measured and assumed quantities should not share an indistinguishable label that invites a buyer to treat them as equivalent evidence.

Calculated values need a defined formula, inputs and applicable conditions. A shaft-power column does not prove that torque was measured. An efficiency column also needs to identify the efficiency boundary. This audit does not demand identical equipment for every test; it demands an honest description of what the report establishes. Where a quantity is estimated, retain the method and its limitations instead of presenting it as a direct reading.

Correct algebra can still reference the wrong cell

After a worksheet is copied, a formula may continue using a different sample's data. Fixed references, cross-sheet links, hidden rows or external workbooks can disconnect a displayed result from the current test. Reviewing the mathematical expression alone is insufficient. Trace the actual input cells and confirm that they belong to the stated pump, operating point and test sequence.

Independently recalculate several representative points, particularly unit conversions. In a hypothetical example, 6000 litres per hour is approximately 0.001667 cubic metres per second. At 8 metres of head, water near ordinary ambient temperature gives roughly 131 watts of hydraulic power. With 300 watts input, electrical-to-water efficiency is approximately 44 percent. These teaching figures are not a product report and do not reveal an unknown shaft power.

Data layerEvidence to retainUseful review question
Original readingSample, time, instrument and conditionsDoes this belong to another sample?
Setting or correctionOrigin, reason and approvalHas an assumption become a measurement?
Calculated resultFormula, references and unitsDid copying change the intended reference?
Released reportRevision, date and authorizationIs this the approved final version?

Keep more than a chart screenshot

A screenshot helps a discussion but normally conceals formulas, hidden data, calculation settings and revision history. Preserve the relationship among instrument exports, calculation workbook and issued report. A fixed-format release document can be convenient for archiving. It should not eliminate the original evidence needed to review a disputed result or reproduce the calculation later.

When revising a file, identify whether the change concerns measurements, units, formulas or presentation. Correcting an error is a normal quality activity; retaining its reason, approval and earlier revision is what makes the correction usable. Avoid silently replacing a file under an ambiguous name. Perform investigative recalculation on a controlled copy so that the review itself does not alter the original test evidence.

Turn concerns into specific requests

List missing information and distinguish issues that affect acceptance from those that only affect presentation. Missing voltage at the stated curve point may prevent a meaningful comparison; a different chart colour may not change the engineering conclusion. Request a particular missing field or a repeat under defined conditions rather than reject a report merely for looking unprofessional. Precise requests make the supplier response assessable.

If sample identity, formulas or inputs cannot be reconciled, withhold the affected conclusion until the gap is resolved. Do not automatically treat every other reading as false. Final acceptance still follows the agreed test and decision procedure. A useful report allows another competent reviewer to follow the same data chain and reach a consistent conclusion with the same limitations.

Questions about report audits

Does converting a workbook to PDF make its data reliable?

No. A fixed format can discourage later editing, but it does not validate the original readings, formulas or references.

Are maximum flow and maximum head enough to review?

No. Check the agreed operating points, simultaneous conditions and decision rules. The two maxima are not one running point.

Does one formula error automatically reject the shipment?

Determine the affected scope, correct or repeat the relevant work and apply the agreed acceptance procedure.

Related Reading