Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Condensate Pump Possible Issues 2

Status
Not open for further replies.

ChaseThisLight

Chemical
Nov 20, 2014
5
Hello, I work in a steam generation facility, and am currently having issues with condensate pumps. Forgive my ignorance, I am a recent graduate and don't know much about pumps, or the facility in general. The plant has 3 vertical condensate pumps. Two are in operation at any time, with one as a backup redundancy. The pumps transport condensate throughout the plant, including being pumped to another set of pumps for seal water. Recently, the seal water has been flashing and causing leaking of steam through the pump, meaning there is not enough discharge pressure from the pumps, and it has been necessary to operate with all 3 pumps in service. I pulled historical data, and made a calculation for pump efficiency for each pump, and each set of pumps that are in operation at each time. It looks clear to me that there are 2 pumps that are operating well below their efficiency when the flow reaches ~3500 GPM for each pump (average operation is around 3000 GPM, and max is around 4000 GPM). I tried finding a drawing of the pump, but they were installed in the 1960's and the drawings are unclear at best. When I did a calculation of the NPSH, the values I came up with are less than half of the original pump curve suggests (again a curve from the 60's or 70's). I don't really even know what I am looking for, but I have a lot of questions. Is there a resource available to learn about pumps, and how to troubleshoot issues? I don't believe there is a way to deadhead individual pumps to determine if a pump isn't operating, the piping and instrumentation isn't set up for that. It was something that was just kinda handed to me, and I am supposed to figure out what is going on. The data I've looked at includes, the total discharge pressure, flow from each pump, temperature entering the pump, and Amps of each pump. From these I've calculated total head, NPSH, and efficiency.
I haven't looked at the condensate vacuum pressure, or looked at how the pumps perform versus various load conditions, these are what I will be working on tomorrow. I don't know if I am being coherent, but I've been staring at this issue all week and haven't gotten anywhere, and I'm getting frustrated. I guess I am looking for direction or advice in general, any suggestions or guidance would be greatly appreciated. Thanks, Chase.
 
Replies continue below

Recommended for you

I am not sure where to start, but you can do a web search for many of your questions.

Please refer to these links:



Prior discussion in Eng-Tips:


A couple of points:

Not sure how you can have steam leaking from the pumps with vacuum suction pressure
You need to analyze the demand from the downstream system
Not sure how you calculate NPSHa without knowing the vacuum conditions

For our help:

What is the pumping temperature?
How did you determine the pumps were not providing sufficient pressure?
What manufacturer and model pump do you have?
 
Seal leakage will not cause lower flow from the pumps. If you have steam in your pumps, the problem is cavitation and that is what causes the lower flow and pressure.

Can you get information from the pump nameplate?
 
First you need to study the concepts involved. Previous threads have reference to good books. Then you need data. If you can run the pumps individually, that is best. If you can't, then you can run them in combination (A and B, A and C, B and C). Measuring shutoff head can be useful if it is safe to do so. If you have flow, pressure and motor amps, you should be able to get a good idea which pump is strongest and which is weakest.

If you provide more information about the system and the pumps we can provide better assistance.

Johnny Pellin
 
first off you need to work out what should be happening or what was happening before the problems appeared. Any historical instrument records would be good, but if the pumps are that old you might be working in a plant without much instrumentation.

Go talk to the operators and seeif they noticed anything different or if this has happened before, when the pumps were last maintained etc. Expect some conflicting information, but it will help.

Get data fromt he ump name plate. If you have a serial number it's surprising how long pump vendors keep original documents or saved them in microfilm or something similar.

Get a PFD or flow schematic or work one out yourself so that you can understand flows and pressure in the condensate system. Try and follow the pipework through from pumps to the seal pumps / system

Try and see what changed or has changed over time. If three pumps are needed to avoid flashing (up the pressure basically), then something is using more flow or your pumps are getting very old and "tired".

You might need to get some additional instruments installed to try and see what is going on, but that shouldn't be that hard.

Just systematically work through the system and get some longer term results, then try and se where it differs from what happened before or what was the ideal system.

It is very common for everyone to blame the pump when it is in fact a "process" issue, but pumps that are 40 + yrs old might be an issue.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way
 
Not to berate a young engineer, but how did you go from seeing steam flashing from the seal water to concluding "there is not enough discharge pressure from the pumps"?

The main thing we are saying is try and give some more information (with PFD or schematic) about the operating conditions and we can be more helpful.

Data for us:

Pressure at liquid surface
Height of liquid head to pump suction
Temperature of liquid pumped
Pump discharge pressure
Design differential head for pump
Pump curve (if available)
Model and manufacturer

I am suspecting some sort of NPSHa problem that wasn't there before, but don't have enough data yet
 
Thanks all for the suggestions. It has been a lot of help.
Sorry for delays, I got pulled into another project. Seems like that happens a lot. I should have been more descriptive with my post. When I began looking at this, I spoke with maintenance and operations departments to determine what the issue was and what they thought the problem could be. I began looking at pump performance, based upon discharge pressure depending on which pumps were operating. I made a calculation of pump efficiency, and graphed that over time and saw 2 of the pumps have significant drop in efficiency when the flow is over 3500 GPM. Then I looked at the discharge pressure, and calculated total head when each pump was in operation. From these, I saw distinct bands of operating pressures. At flows around 3600-3800 GPM, there is ~25 psig drop off. I have been looking at seeing if this is different with different combinations of pumps operating, and looking at when the issue of the drop-off occurred and looking at plant history. The odd thing is, this issue seems to come and go. Last week it was a huge issue, the plant was about to have to drop in load, but the issue solved itself and now people are not as worried about the issue.
GHartmann: I'm sorry if I get the terminology or description incorrect, I've been going on what the operators and maintenance staff have been telling me. There is steam leaking from boiler feed pumps, the condensate pumps supply seal water to the BFP, from what I've been told, there has been steam leaking from the BFP at the seal location. I have not seen this though. The condensate pumps supply water to several locations (more than 10), and I have not yet walked down the demands from each of these systems. The temperature of the water being pumped is typically around 80-85 F. The pump is an Ingersoll-Rand Type 25 APKD-9 Stage Pump. I have looked at the drawing for the BFP, but there is no detail of the pump where the issue is occurring and was told to focus on the condensate pump. And please berate me, otherwise how will I learn, honestly, it is very helpful. The temperature of the liquid at the pump is typically 80-85, with an average of 83 F. The vacuum pressure is around 27 inHg, and the turbine backpressure is around 1.7 inHg nominally. I measured the distance from the hotwell to the discharge of the pump to be 55 inches. The pump data sheet provided with the pump when purchased gives 121 F condensate being pumped, NPSH required 22 feet, NPSH available -1 foot at mounting; and discharge pressure 472 psig.
JJPellin- There is no way that is currently set up to run the pumps isolated, unless load is dropped considerably, and this is a very costly option. When I looked at the pumps, Pump A and C were both below the manufacture's provided pump curve at flows greater than 3500 GPM, but Pump B was mostly consistent with what the expected efficiency should be. For a brief synopsis of the system, there are three nominal 50% capacity condensate pumps (one standby); two deaerating, two-pass divided water box surface condensers; steam packing exhauster; four low pressure shell and tube-type heaters; and deaerating feedwater heater and storage tank. Condenser vacuum is held by vacuum pumps. The discharge of the condensate pumps supplies the boiler feed pump seals, condensate pumps seals, booster pumps seals, and flows through control valves for several desuperheater and drain systems.

I will take your suggestions and look at other systems and see if there has been a demand for more condensate flow over time for any of these systems. That is a very helpful suggestion that didn't even occur to me.
 
Here is a snapshot of the worst performing pump. It measures flow vs efficiency which I calculated. The black line is the pump curve provided by the manufacturer. Link
 
I am not completely understanding all your descriptions (my foggy old mind).

Do you really have a total flow of ~ 2 x 3500 GPM (7000 GPM) for pump seals and desuperheating needs?
That seems a very high demand for these requirements. Requiring a pressure of 472 psig must be for desuperheating requirements (I would have used discharge from the boiler feedwater pumps)

I am not understanding a pump that requires 22 ft NPSH, yet states an NPSHa of - 1 foot at the pump inlet (How long is the pump barrel?) I can only assume that someone entered the incorrect data. From the McNally Institute:

•A negative NPSHA is physically impossible because it implies that the friction losses exceed the available head and that cannot happen. The rule when pumping a boiling fluid is: The NPSHA equals the Static Suction Head minus the Suction friction head because the suction surface pressure and the vapor pressure equalize one another

What is the liquid height (elevation) above the pump suction inlet?

SUMMARY:

Pressure at liquid surface - 27 in Hg vacuum
Height of liquid head to pump suction - ??
Temperature of liquid pumped - 80=85 deg F (121 design)
Pump Flow - 3000 - 4000 GPM
Pump discharge pressure - 472 psig
Design differential head for pump - (~ 1120 by my calculation)
Pump curve (if available) - Flow vs head (you have flow vs efficiency)
Model and manufacturer - Ingersoll-Rand Type 25 APKD-9 Stage Pump

Apart from the NPSH issue (which to me looks like a bad selection, but has worked for 50 YEARS!) it appears once you get beyond 3000 GPM you are drastically loosing efficiency as if you were getting to the end of the pump curve:


Thus with this older pump, with worn internals, you can have a pump curve that has moved to the left and down from the original curve

There are also two other things that hurt you as you go up in flow.
The NPSHr increases (can even be drastic)​
The NPSHa decreases due to increased friction losses​

Thus at the higher flows the pumps can experience cavitation due to low NPSHa resulting in poor performance.

Not sure any of this helps; it is mostly an effort to teach

One final note, I don't see the connection between the leaking seals at the boiler feedwater pumps and how that means your condensate pumps are not working.

Your statement that seal water has been flashing just doesn't make sense as 85 deg water won't flash (I am sure you are trying to describe something else, but what?)

I apologize if I have come across as brash.
 
You are in luck. Flowserve is still manufacturing the APKD as part of their "legacy" line of pumps. It is of the IDP (Ingersoll-Dresser Pumps) heritage (after Ingersoll-Rand & Dresser merger). You can therefore probably still get documents.

Unfortunately, the impellers can be "custom" designed to get a variety of head curves. See page 5 of the attached brochure.

Regarding the NPSH, there is something wrong with your data. These are low NPSHr pumps (which to me means 3 - 5 feet), so I am not sure about your 22 feet. Regardless, the only Net Positive Suction Head you have is your liquid level, so check your operating level in your hotwell (maybe you can raise the level).

Now, back to your pumping system. Your pumps almost certainly must return your condensate back to a deaerator or similar for ultimate return to the boiler feedwater system and on to your boilers.

Unfortunately, we haven't solved your problem, which I am still not sure has been clearly identified.
 
You're amazingly helpful. I truly appreciate all of the help and guidance you have given. I walked down the system with an operator and after that and looking at the P&IDs as well as all the data I have, it looks like the main cause to the issue of steam being created from a blockage in a Y strainer. I still believe there are issues with the pump, and will continue to look into it, but for now, the main issue has been resolved. I've learned a great deal from this.
 
good, glad you got the primary issue fixed and thanks for letting us all know - too many just stop responding and let the thread die - A few "likes" wouldn't go astray though... You always learn more from mistakes / problems than when it's all going smoothly, but that's what makes us engineers. It is sometimes the pump, but don't get too blinded by what the operators think or state - they are great at operating, but often don't really understand the design issues.

If you had efficiency curves, you must have differential head curves to compare to the original? don't know what sort of condition monitoring you're doing on these units, but drop off in efficiency is a clear sign of wear - work out the potential energy saving per month on these things so the plant chiefs can see the benefit of maybe fixing the worst one first...

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way
 
First rule in pump trouble shooting is, it's probably not the pump at fault.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor