Breaking seals, bearings on centrifugal pump

[mrgu]

I have problem with two pumps i serial in a steam turbine system. The pumps are pumping
hot water. Sometimes when the turbine trips, the seal and
occasionally the bearings of the HPP fail and break. Observe that it
is the high pressure pump. When the turbine trips the pumps are running on minflow approx. 32 kg/s

BEP: Flow 137 kg/s, LPP 7-17 bar, HPP 17-37 bar. Inlet temp ca. 160 C.
When the turbin trips then pumps are cooled by a cooled (30 C) water inlet before the LPP.

The pumps looks exactly the same: overhung, end suction impeller, with a circular case. (No volute)
The impeller is recessed. After the impeller the water is passing outlet vanes and turns 90 degree
out in the "big" case. Specific suction speed 12875, specific speed 1252 (gpm,ft)

HPP-high pressure pump
LPP-low pressure pump

Ideas:
Cavitation (in the HPP ?)
Internal circulation (=> cavitation at 17 bar ?)
Radial thrust (in a circular case ?)
Axial thrust (how then, through an impulse force when the water is turning 90 degree.)

 

[vanstoja]

From your numbers I get a best efficiency flowrate of 2168 GPM at 653.7 ft. head which gives 3476 RPM running speed for a specific speed of 1252(RPM-GPM-Ft.). At that speed the required NPSH at bep flow is about 29 Ft.(4.62 bar) for a 12,875 suction specific speed. That amounts to 66% of the 7 bar suction pressure to the booster pump. At minimum flow of 32Kg/s (0.236QD), the NPSHR could be substantially higher than the booster pump suction pressure and flow breakdown could be occurring in the booster pump that also screws up the flow into the main pump and causes abnormal radial or axial hydraulic thrust loading that wrecks main pump seals and bearings. Are there any signs of damage in the booster pump, particularly impeller cavitation damage? I presumed your bep delta P and flowrate values have been measured in your facility rather than coming from manufacturer's test results and that the reported specific speed and suction specific speed are based on tests in your facility. If this is not a correct assumption, then what are the pump manufacturer's comparable design rating numbers and his NPSHR value for 23% rated flow?

 

[Scipio]

My first guess would be suction recirculation induced vibration due to your suction specific speed being too high and your minimum flow being too low, for a few reasons;

1. Seal and bearing failures are frequently due to vibrations or loads applied by axial/radial thrust (and I agree these aren't likely in your configuration) or by misalignment.

2. Typically I don't permit pumps in water service to have a suction specific speed in excess of 10,000 rpm, and impose a minimum flow of about 30% BEP. At 23% even with a 10,000 rpm pump, I would expect to see some suction side recirculation. At a 12,875 rpm pump, where I'd expect at least a 40-50% of BEP minimum flow, I'd be surprised if you weren't getting some recirculation.

3. Suction side recirculation does cause cavitation sometimes, however it can also set up some powerful eddy currents that induce a lot of vibration into the pump. Overhung construction pumps are particularly vulnerable to this kind of excitation cause shaft deflections strong enough to damage seals and eventually bearings.

Suggestions would either be to get new impellers with lower suction specific speeds and/or raise your minimum flow, if at all possible. Have you checked with the original equipment manufacturer to make see if agree with the current minimum flow?

 

[smckennz]

Hi Mrgu
if you are suddenly feeding 30 deg C water into a 160 deg C pump, then some thermal shock problems are to be expected unless the pump design allows for this.
If the pump problem occurs after trip and when the cool water is fed into the pump, then you might look further at thermal shock. Many mechanical seals are intolerant of thermal shock, and face cracking can occur. Not so sure about the bearings; possibly the casing is cooling before the shaft and reducing the bearing clearance but it depends on the bearing arrangement. I note the HPP does more work than the LPP as its pressure rise is twice as much.

Cheers

Steve

 

[mrgu]

Thanks for your answers.

Some clearance.
The pump heads are equal, i.e. LPP 7-22 HPP 22-37
During the turbine trip the system pressure is decreasing fast. The pumps are protected against cavitation by the cool water inlet (in the beginning mixed with hot water).
The explaination so far from my collegues is cavitation. But I think it is something else. The damages occur on the HPP, the axial bearing has been damaged and more often the sealing. I think it has to do with som axial thrusts. But how is it created ? Does internal recirculation creates axial forces ? Notice that the impeller is recessed axial from the pump outlet.

 

[vanstoja]

My 12/7/04 post needs the following corrected values for the HPP unit: Head-666.9 ft.,speed-3541 RPM, NPSHR-30ft(0.89bar which is 12.8% of the 7 bar suction pressure to the LPP pump). These are based on specific speed and suction specific speed values given in your original post which indicated that LPP and HPP heads were 10 and 20 bar, respectively. Now you are saying that the two pump heads are equal at 15 bar (LPP 7-22, HPP 22-37)which screws up my prior calculations and makes it even more unlikely that the HPP is cavitating with a suction pressure of 22 bar. Where are you getting these so-called BEP numbers from? I can't believe that a pump manufacturer tested two pumps in series at a water temperature of 160C(320F) with a minimum suction pressure of 7 bars(101.5psi) to the so-called LPP (now a gross misnomer if the heads are the same for the two pumps which appear to me to be feed booster and main feed pumps delivering to a boiler or heat exchanger). Your 7 bar suction pressure to the boost pump is barely enough to suppress vapor pressure at 320F (90 psia) plus some unknown NPSHR at BEP flow. If you've got a controlled coolant chemistry for corrosion control, then you need additional suction pressure to keep the dissolved gases like hydrogen, nitrogen, oxygen in solution. This can add as much as 22 to 130 psia to required minimum suction pressure for 25 and 150 cc/Kg of total gas composed of 75% hydrogen and 25% nitrogen. With insufficient suction pressure for gas suppression and a possibly large increase in NPSHR at 23% BEP flow, chances are you are choking the booster pump to the point where it is not developing much if any head and you are getting not much more than 7 bar inlet pressure to the main pump which is also choking. Cavitation damage may not be evident in either pump because released gases are cushioning the cavitation bubble collapse pressures. Radial or axial hydraulic thrust overloads are probably damaging the main pump. Another element in this process may be intense recirculation and reversed flow starting in the main pump and proceeding upstream through the booster pump into the inlet piping. This coswirling prerotation backflow would also unload the booster pump and help lower its developed head to near zero. The 12,875 suction specific speed you quoted, whereever that number comes from, is entirely too high for two (identical design ??)pumps that do not have inducers particularly when they have to operate at flows as low as 23% rated. It seems like you have been just plain lucky to have avoided more disastrous damage than what you have reported.

 

[Artisi]

Seems the question / problem asked was about seal and bearing failures.
Are you asking / saying that the bearings and seals only fail following an unsheduled shut down.
If this is the case I'm with smckennz and would suspect thermal shock as the likely cause of the seal failure- the bearing failure appears to be something else - have you had the bearing failure analysed by a bearing company. Is the bearing failure co-incidental and only apparent when the pump is dismantled for the bearing failure.



Naresuan University
Phitsanulok
Thailand

 

[Artisi]

My last post should read ---- dismantled for the seal failure.

Thinking a bit more about it, a bearing failure analysis is important as this will help point you in the right direction in what to look for from the pump operating point of view.

Naresuan University
Phitsanulok
Thailand