ACTUAL TDH NOT MATCHING PUMP CURVE
ACTUAL TDH NOT MATCHING PUMP CURVE
(OP)
I have a centrifugal pump that should be working fine but isn't. According to the pump curve, this pump should be transporting the required flow, which it is. The TDH on the pump curve is about 300' but only about 1/2 of that is needed. I would expect the control valve downstream of the pump to take the pressure drop. Instead, the pump is just not putting up the discharge pressure. When the pump discharge is blocked in, the pump does shows the right discharge pressure (and the right TDH). The NPSHa is just above the NPSHr and the pump motor is large enough. The pump efficiency is poor at about 35%.
The pump has been repeatedly abused since it was installed.
What would be causing the discharge pressure to drop 50%, when suction pressure is constant and when the flow through the pump does not warrant it, according to the pump curve?
The pump has been repeatedly abused since it was installed.
What would be causing the discharge pressure to drop 50%, when suction pressure is constant and when the flow through the pump does not warrant it, according to the pump curve?





RE: ACTUAL TDH NOT MATCHING PUMP CURVE
If by "blocked in" you mean the pump is operating against a closed valve, which I call "shut-Off" head, and you mean that the shut-off head matches what the pump curve shows.
Then,
Assuming that rpm is correct, you have thereby verified that the impeller diameter is correct. Good information.
Assuming impeller is not blocked/clogged in any way, I then quickly jump to look at either wear ring clerance on a closed impeller, vane clearance against the case on open impeller. It is surprising to some people how just a little extra clearance on the wear ring/vane clearance can dramatically affect head output.
If flow rate has been verified, and then you check that flow rate against the pump Head Capacity curve, you should be reading roughly what the pump curve states for head at that flow rate minus or plus intake head, as the case may be.
Now, that is all assuming that you are pumping water if the pump characteristic is based on water. If you are pumping a lighter liquid than water, and the pump characteristic is based on water, your pressure gauage will indeed read less than the rated pressure.
PUMPDESIGNER
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
Other general thoughts
1 - Is your pump operating at rated speed?
2 - Do you trust your flow measurement instruments?
3 - Has fluid density been considered in flow measurement?
=====================================
Eng-tips forums: The best place on the web for engineering discussions.
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
Pumpdesigner, if it were a wear ring clearance problem, wouldn't that affect the shut off head as well, preventing the pump from showing true shut off pressure while 'blocked in'?
Techsupport1, one more thing to check. You give npshr and npsha information, but what is the total positive suction head. Is your fluid saturated, or do you get a significant contribution to TDH from your suction pressure.
With your pump 'blocked in' you have no line losses on the suction side, and with the pump flowing, the line losses before the pump rob you of TDH. Things get worse if you have flashing or cavitation on the suction side.
rmw
rmw
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
If as you say NPSHa is above NPSHr then there shouldn't be any suction problems (assuming here that you have an operational pressure guage at the suction of the pump, and are not just going on initial calculations made when specifying the pump)Cavitation however, should be making itself known if the suction conditions are badly wrong.
In any case if NPSHa is less than NPSHr, power absorbed by the motor would drop off. Again this can be confirmed by observation. If there is less pressure drop across the control valve then usual or expected, this would indicate that the valve is opening up to allow more flow.
The clue looks to be the comment that "the pump has been repeatedly abused since installation"
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
If NPSHA is just above NPSHR then you will be cavitating.
If the pump has been abused then it is worth looking inside before doing too many more measurements. At 35% efficiency, volt/amp readings wont tell you a lot.
I suspect its stuffed.
Cheers
Steve
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
rmw - Wear ring affect on shut-off head is variable, you are correct that it may have an effect. On flat curve pumps however there may be very little difference, and on high specific speed pumps the effects at shut off are small also unless the wear ring leakage pushes the flow rate beyond a critical point where the curve starts to drop, but still at shut-off the affect may be so small that shut-off head appears correct even with excessive clearance.
That is one problem with these forums, makes you sharp because the people posting often do not have much information about the pumps, what kind they are, stages, specific speeds, diffuser, scroll, double volute, double suction, etc. Makes it really hard and we end up speculating quite a lot.
PUMPDESIGNER
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
Thanks for the info.
I agree with the part about groping around in the dark, but often many posters do not know that much about a pump, in order to give us what we need to analyze situations.
rmw
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
The first thing that I noticed was your comment:
"The pump has been repeatedly abused since it was installed."
What manner of abuse(s) have been visited upon this beloved piece of machinery? Abrasive slurries, large damaging solids, broken/bent vanes, repairs that would be better characterized as vandalism, ...?
How well is the axial alignment being maintained? If the axial position is not well controlled, this could account for damage and excess leakage at the wear rings.
Another thought involves the accuracy of the pump curves that you have. Were the curves based on actual test performance of this pump, generic curves for the pump model, or "advertizing" curves for a series of pump models? The combination of an abused (or worn) pump and uncertainty of the true applicability of the presumed pump curves can readily account for a significant portion of the apparent performance discrepancy.
How does the actual shaft speed compare to the shaft speed that applies to the pump curves that you are using? Apply the affinity laws to see the effects of the speed differences on expected performance values.
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
I don't think that you have described your system in detail. What you are observing makes me think that you may have air (or some type of compressible gas) which is entrained in the liquid entering the pump. When flow is reduced to shut-off, possibly the air is escaping prior to the pump suction. However, if you have entrained air in significant volume (>6%) - this could result in major decrease in performance. Entrained air will cause the performance curve to drop below the standard curve at an increasing rate with increasing flow rate.
Entrained air does not result in cavitation, but would probably cause some noise.
You should be able to make a judgement on this possibility by checking the HP drawn at higher flow rates. Actual HP will be well below expected pump curve HP.
Richard
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
Drapes
RE: ACTUAL TDH NOT MATCHING PUMP CURVE
RE: ACTUAL TDH NOT MATCHING PUMP CURVE