Boiler Feed Pump Issue

[hunter321]

Hi, i am having an issue with one of our boiler feed pumps. Randomly the pump is tripping due to low outlet flow (this is a protection in out control system). The pump is a Torishima multistage (10 stage) horizontal axis and is fed from a deaerator. I've chcked all the temperatures/pressures and the water is not at saturation. We've installed some instrumentation and the first thing noted before the pump tripping is a drop in balance line pressure, this is followed a few seconds later by a drop in suction pressure and then outlet pressure. All vibrations are fine. The suction strainer has been checked and is clean. If anyone has experienced this problem before or has some suggestion i would be very interested to hear. Thanks you, Mark.

 

[pennpiper]

Just thinking:
a) are there two 100% pumps or three 50% pumps?
b) does the number and location of the pumps cause more fittings in the suction line of the subject pump?
c) are the suction lines to all of these pumps undersized?

prognosis: Lead or Lag

 

[hunter321]

Hi pennpiper.

a) Two 100%. Standby pump starts and keeps process running when duty pump trisp.
b) Not sure what you are asking here
c) probably should have said. THe pumps were installed five years ago and have ran without issue until a few weeks back. So pipes are sized correctly.

 

[JJPellin]

If these were my pumps, I would gather data and evaluate the pump performance. I would verify that they were still performing well relative to the original test curve.

I would ask our instrument techs to check the proper operation of the level control on the deaerator. An intermittent loss of level could explain the problem.

You said that the first indication was a drop in balance line pressure. I assume you mean balance line differential pressure. Does the pump use a balance disk or a drum? Does the pump have a thrust bearing or does it depend on the balance device? If it has a thrust bearing, is the bearing spring loaded like some of the big Flowserve barrel pumps? A shift in axial position could change the clearance in a balance disk arrangement and change the balance line differential pressure.

Is your vibration data from a continuous monitoring system? Are you actually looking at vibe data during the event? Or, are you just looking at vibe data taken by a portable system during normal operation?

It sounds to me like you could be losing level in the deaerator because of a faulty level controller. At some point, you draw vapor into the pump suction. The vapor passing through the pump causes a loss of performance and affects the operation of the balance disk or drum. The lag in indication could be related to the sample rate of the DCS data points.

We could provide better input if you could provide a flow schematic and describe your control system in detail. A more complete description of the pump would be helpful to me since I am not familiar with this manufacturer.


Johnny Pellin

 

[hunter321]

Hi JJPellin,

I'll try answer all your questions.
- We've checked the pump performance and it looks fine.
- The deaerator level is also fine. This is verified by three level transmitters. DCS is 1 second interval, level is flat before, duraing and after the event.
- The balancing devide is a disk.
- We also have a thrust bearing, i don't think is is spring loaded, balancing device takes most of the axial forces.
- The balance line pressure is not differential. The balance line comes off the final stages of the pump and feeds back to the deserator. There is a pressure control valve on the line and a standard pressure gauge.
- Our vibrations are continuously monitored, as i said, they all look fine prior to the event.
- Some pump details below


Pump manufactures TORISHIMA
Product Number AP345059
Number of supply 2
Type &Size Single suction mul ti-stage diffuser horizontal
No of Stages 10
Rotation CW
Design Pressure 218.7 barG
Pumping Liquid BOILER FEED WATER
Total Head HP 1670 m
Total Head IP 605 m
Capacity (Discharge) 320 m3/h
Capacity (Extraction) 110 m3/h
Speed 2980
Pump Efficiency. 79%
NSPH 8.8m
Pump ing Temp. 140.4 DegC

 

[JJPellin]

Thank you for the additional information.

I don’t understand what you are describing as a balance line. I am familiar with a balance line as the line that takes the leakage from the balance disk or drum and routes it back to the pump suction line or supply vessel. What you described is a line coming off the discharge of the pump and spilling back to the DA. I would describe this as a minimum flow spill-back line. A balance line would not normally have a pressure control valve in it. A spill-back line would have some sort of flow or pressure control. So, just to clarify, is this a balance line (based on my definition) or a minimum flow spill-back line?

Rather than wait for an answer, I will address both options.

If this is a balance line, then I can’t imagine why it has a pressure control valve in the line. I don’t understand what parameter would be used to tell this valve to open or close. But, in any case, I would expect that there is some problem with the control of that valve. If some signal (false indication or true) tells that valve to open, then the “balance line pressure” would drop. With lower pressure on the downstream side of the balance disk, the pump might thrust in reverse (depending on the thrust bearing arrangement). The faces of the balance disk would open up and the leakage through the balance line would increase. With more flow through the balance line, the flow available for the outgoing process line would be reduced. Then low flow is registered and the pump trips off.

If this is a minimum flow line, the scenario is very much the same. The pressure control valve on the spill-back line seems like a strange choice. I would normally set the minimum flow control based on flow, not pressure. But, if this is trying to control the outgoing stream to a constant pressure, I guess this could work. If this pressure control were to malfunction, the pressure control valve on the spill-back line could open, diverting flow back to the DA. The outgoing flow to the process line would drop and the pump would trip on low flow. This assumes that the pump would trip on low outgoing flow and not total pump flow (spill-back flow plus outgoing flow).

I am not familiar with any sort of an extraction pump with a single suction and two discharges. I assume that is what you are describing. So, there could be something particular to this type of pump that I know nothing about that could explain your problem. If this pump has all of the impellers in one orientation, then the pressure on the balance line is very, very important for balancing the high thrust. If the impellers are in an opposed configuration, then the net thrust would be much lower and the balance line pressure would not be as critical.


Johnny Pellin

 

[hunter321]

Sorry, maybe i misled you, you are correct the balance line is taken from the balance disk and routed to the DA. The pressure control valve on this line is a manual valve. Its is set during commissioning and should not need to be changed. It is usually set about 1-2 bar higher than the DA pressure. It basically sets the pressure at the balance disk. In our case it is fully open, so in effect is doing nothing. Our balance pressure is a bit on the high side, about 3 bar higher than the DA, the OEM does not think this is a problem. I can't understand how the balance line pressure can drop suddenly, unless the balance device sticking or passing excessively.
Thanks for your help.

 

[JJPellin]

The picture is getting clearer. I have never seen a pressure control valve in a balance line, but we don’t have any boiler feed water pumps for pressures this high. This would be most comparable to our big coker jet water pump (270 m3/hr, 248 barg, 2570 kW, 10 stages)

A drop in balance line pressure (given a fixed restriction in the balance line) suggests a reduction in balance line flow. That suggests that the faces of the balance disk closed which would normally happen for thrust movement in the active direction.

A drop in suction pressure would not be easily related to a change in thrust position. This is especially true if the thrust occurs first and is followed by the drop in suction pressure. You mentioned a suction strainer. I don’t like to have suction strainers at my boiler feed pumps. Unless the DA is chronically sending scrap iron to the pump, it should not be needed. Unless there was a clearly understood problem that required a suction strainer, I would remove it.

The drop in discharge pressure comes after the drop in suction pressure. This makes even less sense to me. I can see a situation where a drop in discharge pressure causes an increase in thrust load which closes up the balance disk which drops the balance line flow which reduces the balance line pressure. But, I can’t come up with an explanation that works the other way around.

Do you monitor thrust position? Does the thrust position shift during this series of events?

Otherwise, I am stumped. Sorry.


Johnny Pellin

 

[hunter321]

Thanks a lot Johnny for your help. There is one other possibility. We had no pressure transmitters installed on the suction line or balance line. We installed temporary transmitters and connected them to seperate recorders. My instrument guy tells me the times are synced on the recorders, but if they were out by a few seconds we could be getting false times and the suction pressure may be dropping before the balance.
So this raises the question as to what could cause a sudden drop in suction pressure when it operates fine most of the time. We are seeing these events on average once every two weeks.

 

[JJPellin]

I assume you have some sort of a pressure controller for the vent off the top of the DA. If that controller was sticky, I could see a scenario that might make sense. Perhaps, the DA pressure increased because of a sticky pressure control and then suddenly dropped in pressure. The loss of pressure would cause the water to want to flash all the way down to the pump. The pump would cavitate and might be passing vapor bubbles. The pump loses performance, dropping the discharge pressure. The flow passing up the balance line also cavitates, reducing the pressure drop across the pressure control valve.

Perhaps this is all a bit of a stretch. But, whatever is used to control the pressure in the top of the DA should be checked. If it is automatically controlled, the controller could be malfunctioning. If it is manually adjusted, you could have an operator or board man with a heavy hand.

We have a pressure controller on the 70# steam injected into our newest DA.

At this point, I should point out that we are getting far away from the pump and far away from my area of expertise. There are a lot of people on this site that know more about DA systems than I do.


Johnny Pellin

 

[electricpete]

Your daerator level indicator may have reference leg flashing. This causes level to indicate higher than actual, so actual level gets controlled low...but looks fine on dcs.

=====================================
(2B)+(2B)' ?

 

[rmw]

How does your boiler drum level controller operate? Is it modulating, open/closed, what? I suspect low boiler water flow that is flashing the fluid in the pump. Could the spill back line be plugged in some way or partially blocked not allowing adequate minimum pump flow through your pump?

Things change over the years and a valve set at commissioning might not have the right setting 5 years later.

rmw

PS: I always shudder when I see the words "suction strainer" - but that's just me.

 

[ccfowler]

hunter321,

Am I interpreting your information correctly that your pump has a nominal discharge flow of 320 m3/hr with an intermediate extraction (bleeder port) with a nominal 110 m3/hr flow with a resulting nominal suction flow of 430 m3/hr?

If this is the case, variations in the extraction, or bleeder, flow could be acting in combination with the other potential items discussed above. Do the discharge and bleeder flows tend to vary consistently, or do they fluctuate relatively independently from each other? Since the two discharge flows are additive at the suction, nasty things can be happening at the suction when the two flows combine for particularly large momentary flows. Since the work done to produce the bleeder flow (per unit of volume) is only a fraction of the work done on the discharge flow, relatively large increases in bleeder flow can appear as seemingly trivial increases in the work done by the motor. Are both flows monitored?

How long has it been since the inlet stage has been inspected? I have been amazed at how well badly damaged impellers (from cavitation) can seem to function until a critical amount of damage has been sustained. Then, performance gets worse quickly. I recall an old boiler feed pump that was to be retired in a relatively short time (impending plant closure) that swas starting to perform badly. All stages had identical impellers, but the suction impeller was nearly destroyed from cavitation. All of the other impellers were in excellent condition. The expedient solution was to swap the suction and discharge impellers and hope for the best. If the deteriorated impeller did disintegrate, at least the debris would not be fed to the other stages. When reassembled with a good impeller at the suction, apparently normal operation resumed. Since cavitation was not an issue for the discharge stage, that dreadfully damaged impeller performed beyond my wildest hopes.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

 

[mirghaffari]

please check the leakage from drain line. somtimes due to leakage from drain valve,water flow from high pressure side to low pressure side and cused a low flow in discharge side. if you have a same line to dran both high and low perssure side please ceck my suggestion