The acceptance-test decision
For any pump purchase above ~$10,000, the buyer needs to decide what level of factory acceptance testing to require. The Hydraulic Institute Standard ANSI/HI 14.6 โ *Rotodynamic Pumps for Hydraulic Performance Acceptance Tests* โ defines the standard test grades:
| Grade | What it covers | Typical use | |---|---|---| | 1B | Tightest tolerance | Boiler-feed, nuclear, critical service | | 1E | Tighter than 2B | Industrial-grade chemical service | | 1U | Standard tightest | Most municipal water + standard process | | 2B | Mid-range | Standard ANSI B73.1 chemical service | | 2U | Standard | General-purpose municipal + commercial | | 3B | Loose | Construction / dewatering pumps |
The grades differ in:
- Allowable deviation of head from the published curve (1B: ยฑ2.5%, 2U: ยฑ5%, 3B: ยฑ8%)
- Allowable deviation of efficiency from published (1B: -3% absolute, 2U: -7%)
- Allowable deviation of NPSHr (1B: +5%, 2U: +10%)
For most municipal water + commercial service: Grade 2U is the standard. For most industrial chemical service: Grade 2B. For critical / high-cost service: Grade 1U or tighter.
What the test actually involves
A factory acceptance test is performed at the manufacturer's test facility before shipping. The test:
1. Pump is installed on the manufacturer's test loop 2. Calibrated instruments measure: flow (Venturi, magmeter), head (pressure transducers), shaft power (torque transducer or kW meter), speed (tachometer) 3. Pump is run at 5+ operating points across the curve 4. Data is recorded for each point 5. Test report compares measured curve to published curve at the test grade tolerance
The test takes 2-8 hours depending on pump size + grade requirements. Cost: $500-$5,000 added to the pump price.
When you should require a witnessed test
A "witnessed" test means the buyer (or a hired representative) is present at the factory to observe the test. Critical for:
- High-stakes installations (boiler-feed, fire-pump, critical-process)
- New pump models without established field history
- Specifications with unusual requirements (custom impeller trim, special materials)
- Buyer's first pump from a new manufacturer
For routine purchases, an unwitnessed test (factory-conducted, results sent to buyer) is sufficient.
NPSHr testing โ separate consideration
HI 14.6 covers head + flow + power. NPSHr testing is governed by ANSI/HI 9.6.1 and 14.6 jointly.
NPSHr testing measures:
- Head drop at 3% flow reduction as suction pressure is throttled
- The NPSHa value at which 3% drop occurs = NPSHr at that flow
- Repeat at 3-5 flows across the curve
NPSHr testing is significantly more expensive than head/flow testing because the test loop has to control suction-side pressure precisely. For NPSH-critical applications (high-temp service, low-suction-pressure systems), require a separate NPSHr test.
For routine service, accept the manufacturer's published NPSHr curve and verify in field commissioning.
Test report contents
A complete HI 14.6 test report includes:
1. Pump identification (model, serial, impeller trim) 2. Test conditions (water temperature, atmospheric pressure, fluid) 3. Flow vs. head data at 5+ operating points 4. Flow vs. shaft power data 5. Flow vs. efficiency calculated from above 6. NPSHr data (if tested) 7. Test grade (1B, 2U, etc.) and pass/fail vs. published curve 8. Calibration certificates for all instruments 9. Date + signature of test engineer 10. Signature of witness (if witnessed)
Keep the test report in the pump's installation file. It's the baseline for future field-test comparisons.
Field acceptance test (vs. factory)
Factory test โ field test. Field acceptance verifies:
- The pump as installed meets the operating point
- Installation effects (suction-piping arrangement, alignment, foundation) don't degrade performance
- Operating instrumentation is calibrated correctly
- Operating procedures function as designed
Field test sequence:
1. Bring the pump up to operating speed 2. Open the discharge valve to design flow (calibrated meter) 3. Measure suction + discharge pressure at the pump flanges 4. Measure shaft power at the motor (amps ร volts ร power factor ร motor efficiency) 5. Calculate ฮท = (Q ร TDH ร SG) / (3960 ร BHP) 6. Compare to factory test report
Acceptable field deviation: typically + 5% from factory test (additional uncertainty in field measurement). Unacceptable deviation: > 10% suggests installation problems.
Common test issues
Cold-water test, hot-water service. Pump tested in 70ยฐF water at the factory; installed in 200ยฐF service. Performance may differ due to viscosity / vapor-pressure effects. Spec the test temperature to match service if margin is tight.
Test loop differences. Manufacturer test loops have ideal suction conditions. Field installations rarely match that. A pump that performs at factory specs may underperform in the field due to suction-side issues โ which the field test catches.
Specific gravity assumed = 1.0. Test loops use water (SG = 1.0). For non-water service, head + flow are the same but BHP differs. Verify SG correction in the field.
VFD testing. Factory tests are typically across-the-line at rated speed. For VFD service, consider requiring the test at the planned minimum speed too โ sometimes performance drops more than affinity laws predict.
Cost-benefit
Test grade costs (added to pump base price):
| Grade | Typical premium | |---|---| | 3B | $0 (standard for low-cost) | | 2U | $200-$1,000 | | 2B | $500-$2,500 | | 1U | $1,500-$5,000 | | 1B | $5,000-$15,000 |
For a $10,000 pump in standard service: 2U test ($500) is cheap insurance. For a $50,000 critical-service pump: 1U test ($3,000) is essential. For a $500,000 specialty pump: 1B test + witnessed ($15,000-$25,000) is the right call.
Type test vs. unit test
Type test โ performed once per pump model (or per major design change). Establishes the published curve. Unit test โ performed on each individual unit before shipping. Verifies the individual unit matches the type-test curve within grade tolerance.
For commodity pumps (small ANSI B73.1, common Gorman-Rupp T-series), only the type test is performed; unit testing is waived because design + manufacturing are tightly controlled.
For custom pumps (special impellers, severe-service materials, large industrial), both tests are mandatory.
A common error: skipping the test for cost
A buyer specifies a 2U test, then accepts the pump without one to save $800 on a $25,000 purchase. The pump is installed, fails the field test, gets pulled, returned to the manufacturer.
Result: $5,000 in shipping + lost-time delays + reinstallation labor. The original $800 was cheap insurance.
How the calculator handles it
Headloss Calculator doesn't enforce test-grade requirements (that's a procurement decision). The pump-selection panel surfaces:
- Manufacturer's published curve at design speed
- BEP location + tolerance band
- Operating-point efficiency
These help the buyer specify the test parameters. For the actual factory test report, work with the OEM and reference HI 14.6 directly.
References
- ANSI/HI 14.6 โ *Rotodynamic Pumps for Hydraulic Performance Acceptance Tests.*
- ANSI/HI 9.6.1 โ *Rotodynamic Pumps Guideline for NPSH Margin.*
- API Standard 610 (12th ed.) โ *Centrifugal Pumps for Petroleum, Petrochemical, and Natural Gas Industries* (acceptance criteria for petroleum service).
- Karassik, I. J., et al. *Pump Handbook,* 4th ed. โ chapter on field testing.
- DOE PSAT (Pump System Assessment Tool) โ free reference for field-test methodology.