Witnessed vs. observed tests
Q - What is the difference between witnessed and observed performance tests?
A - In a witnessed test the production schedule has a hold point to allow the
customer to witness the performance test. In practice there may be a preliminary
test conducted ahead of the witnessed test to avoid any surprises.
In an observed test there is no hold point in the production schedule. The test is
conducted as scheduled with, or without, the presence of the customer. The
vendor is only required to notify the customer of the test schedule.
Q - What is the difference between suppression tests by throttling valve and by
vacuum tank methods? Any advantage of one method over the other?
A - Suppression test is a test to determine pump NPSHR usually based on 3%
head drop. A baseline pump performance curve is first obtained by testing the
pump with ample NPSHA. At the flow rates under consideration the NPSHA is
then gradually reduced by reducing the suction pressure until the head drops by
3% from baseline data; that NPSHA is said to be the pump NPSHR.
In throttling valve method a valve at the suction side of the pump is used to reduce
(throttle) the suction pressure; the throttling action may cause suction flow
disturbance that may affect the accuracy of the test result.
In vacuum tank method the suction pressure comes from an elevated tank; the
pressure is reduced by creating partial vacuum in the tank (with the use of vacuum
pump.) This method produces a more accurate result because of a more uniform
suction flow velocity.
Q - What is a string test?
A - A string test refers to the testing of the complete pump package, including the
contract driver, baseplate, seal flush system including any oil pots or lube system
for dual seals, and bearing forced-feed lube system, if applicable. A string test
may also include the contract piping, instrumentation, gearbox, or variable-drive, if
specified by the customer.
The vibration acceptance criteria for string test should be negotiated and agreed
upon between the vendor and customer because of difficulty compensating for
factors such as testing on an ungrouted baseplate at the test facility that would be
grouted in the field, or the use of contract vibration monitors that do not allow for
filtered readings at discreet frequencies.
Q - I'm confused! The specifications for a pump I'm working on require that the
electric motor driver shall not be overloaded at any point in the performance curve.
I have seen many curves and there is no fixed point where the curves end; they
vary from 115% to 140% of best efficiency point. At what point does a performance
curve really ends?
A - It is common practice in the industry to show the end of a pump curve at
120% to 125% of its BEP.
Q - On a performance curve, why is the NPSHR curve cut off at about 40% of the
BEP, and not drawn all the way down to zero capacity?
A - In most cases, it is not recommended to operate the pump continuously below
the capacity point where the NPSHR curve stops.