The below figure shows a basic centrifugal pump test circuit. The Centrifugal Pump Performance Test procedure is carried out as follows, but it may slightly change based on pump vendor test bed condition and internal procedures.
The following test are normally carried out in centrifugal pumps
- Pump performance test
- Mechanical Run test
- NPSHr test
In that NPSHr test is usually carried out whenever the NPSH margin is less than 1 meter.
Centrifugal Pump Performance Test procedure
Step 1. The Flow Vs Head(q/H) test
Normally the number of points required for measurement is based on the Code used for pump manufacturing ie., API 610, ASME, NFPA 20, Hydraulic Institute (HI 1.6).
In the case of API pump, six point measurement are required.
These points will range from a sub-minimum flow (Shut off) to a minimum of 120% of BEP flow sub-minimum
- Shut off point,
- Minimum stable flow
- An intermediate flow (flow between minimum and BEP/ rated)
- A point between 95% and 99% of rated
- Between rated and 105% of rated flow, BEP (if rated is not within 5% of BEP),
- At the end of allowable operation region (120% of BEP)
As per HI 1.6, the performance test points are based on pump types A & B.
Pump Type A:
These pumps are manufactured for special cases. For these pumps, the “Seven” measuring points are required.
Pump Type B:
These pumps are manufactured by mass production. For these pumps, the “Five” measuring points are required.
Note: All ASME pumps testing shall be as per HI 1.6.
During the performance test, the following values shall be recorded at each test point: suction pressure, discharge pressure, flow, RPM and power.
Refer below “sample” performance test sheet
Based on the flow rate (test points), the corresponding head is measured and plotted in the Flow Vs Head curve.using the measurement accuracy levels given for the class of pump, drawing the q/H measured band;
Other Points check during Performance test:
One more major point to ensure the test accuracy is pump speed. As per API, all tests shall be run at the rated speed + 3% tolerance. In case the pumps at 50 Hz (ie., Operating condition) but the test facility is 60 Hz. The pump manufacturer as to have a necessary arrangement for reducing the frequency from 60Hz to 50Hz and achieve the operating speed. But in practical some of the pump manufacturer does not have the facility for step down the frequency, in those cases, the test speed at 60 Hz shall be at the same pole speed as 50 Hz provided the pump is rated for that operating speed. If it is not, then the pole speed must be reduced to the next lowest value.
Normally job seal is used during the performance test. Seals shall be check for leakage during performance testing. If visible leakage is observed the seal must be removed, inspected and repaired as required. The repaired and installed seal must be hydro tested with water at 100 PSI or observed during a subsequent running test to have no visible leakage.
Step 2. The efficiency test
The efficiency guarantee is checked using the same set of test measurements as the Flow Vs Head test. Pump efficiency is shown plotted against the flow as in below figure. In most cases, the efficiency guarantee will be specified at the rated flow rate (q).
Step 3. Vibration measurements
At each performance test data point (except shutoff point) pump bearing housing vibrations will be measured and recorded.
For between bearings pumps, vibration readings shall be measured on each bearing housing in the vertical and horizontal directions and on the thrust bearing housing in the axial direction.
For overhung pumps, vibration readings shall be taken in the vertical, horizontal and axial directions
in the location of the thrust bearing. These readings will be taken at the pump drive and non-drive end bearing housings in the horizontal and vertical positions.
Mechanical run Test:
The pumps will be run at the rated flow until bearing temperatures have stabilized. Stabilization will be achieved when the bearing temperatures have changed at or less than 2 deg F in 20 min. After stabilization has been achieved, a four-hour mechanical run will be performed at the specified rated flow condition.
The bearing temperature test:
Usually, this test is carried out prior to all performance test. This is the right time to discuss the test.
For pumps with flood oil or ring oil lubrication, sump oil temperatures are monitored during the performance test. The temperature rise values are applicable to the duration of the standard performance test, typically one hour or less of operation.
For pumps with oil mist, outer bearing ring temperature shall be taken. If there are no provisions for bearing temperature detectors, bearing housing skin temperature is measured.
For pumps with forced oil lubrication, temperature measurement is taken in the outlet connections on the bearing housings.
Temperatures shall be measured prior to the start of the performance/mechanical run test and at the conclusion of the test prior to shutdown. Baseline temperature shall not exceed ambient temperatures.
Maximum allowable temperature rise above ambient temperature is as follows: (The values may change slightly with different pump manufacturers)
Pumps with flood oil or oil ring – 70 degrees F (21C) oil temperature.
Pumps with oil mist – 70 degrees F (21C) bearing metal or bearing housing skin temperature.
For pressurized systems, the bearing oil temperature rise shall not exceed 50 degrees F (10C).
The NPSH test:
These are two common methods are used to doing the NPSH test.
- Vacuum Suppression
- Suction Valve Throttling
Vacuum suppression testing
The pump is supplied from a closed tank with a constant liquid level. Suction conditions are varied by the use of shop air to increase pressure above atmospheric and a vacuum pump to reduce pressure below atmospheric.
Suction valve throttling
This method is only utilized if the pump capacity exceeds that of a closed test loop and the pump is tested on a sump. The pump is supplied from an open sump with a constant liquid level. A booster pump may or may not be utilized to simulate suction pressure above atmospheric pressure depending on the characteristics of the pump. Suction pressure is reduced by throttling a valve on the suction piping of the pump in conjunction with shutting off the booster pump.
In the above two methods, vacuum method results are more accurate and scientific result. unobscured by the presence of the valve. Refer above flow diagram for vacuum method.
During NPSH testing, a flow is set and held constant while the suction pressure is reduced incrementally. The total developed head is monitored to determine the suction pressure point where a 3% drop in the total developed head is observed. The temperature of the test water is taken at this point to determine vapor pressure. NPSH is calculated in accordance via the latest revision of the Hydraulic Institute test standard.
Correction factors (applied to flow, Head, Power, and NPSH) need to be used if the test speed of the pump does do not match the rated speed. They are:
- flow q (corrected) = q (measured) × (nsp/n)
- head H (corrected) = H (measured) × (nsp/n)^2
- power P (corrected) = P (measured) × (nsp/n)^3
- NPSH (corrected) = NPSH (measured) × (nsp/n)^2
n = speed during the test
nsp = rated speed
After centrifugal pump performance test, the results are checked with the rated conditions or guaranteed point. The acceptable tolerance range is based on applicable pump manufacturing points
For API 610:
The acceptable tolerance is
- Rated point: +/-3%
- Shutoff condition: +/-10%
- Rated power < 104%
- Efficiency is not including in pump’s rated performance evaluation
- Overall vibration on the bearing housing measurements shall be less than 0.12 in/s RMS. For data points outside the POR (Preferred Operating Range) but within the AOR (Allowed Operating Range), there shall be an allowable increase in vibration of 30%.
For Hydraulic Institute standard
|Acceptance level||Type A pumps||Type B pumps|
|Under 60 m (200 ft)
o to 680 m3/h (2999 gpm)
|+ 8%, – 0 +||+5%, -3%|
|Under 60 m (200 ft)
681 m3/h (3000 gpm)
|+ 5%, -0||+ 5%, -3%|
|From 60 m (200 ft) to
150 m (500 ft),
any flow rate
|+ 5%, -0
|+ 5%, -3%|
|150 m (500 ft) and over, any flow rate||+ 3%, – 0
|+ 5%, -3%|
|Minimum efficiency at rated
rpm and rate of flow
|np or nOA||100 / ((120/np)-0.2)
For pump type A: np – Contractor pump efficiency
For pump type B: np – Published, nominal efficiency
As mentioned in the beginning of this article, Centrifugal Pump Performance Test procedure may vary slightly with respect to pump manufacture internal test procedures. Almost all pump manufacturer will meet the acceptance criteria mentioned in this article.