Cavitation during pump startup 1

[processeng01]

Hi all,

Can anyone point me in the right direction as to how I should approach below problem I am facing:

I am designing a pumping system where condensate transfer pump discharges water directly to boiler feedwater pump in a power plant with no tank in between. (My condenser is deaerating type. I do not want to put additional atmospheric tank nor a pressurized deaerator in between). My concern is that when the boiler pump is started up, its suction side may cavitate because there is no real reservoir per se. (At this time condensate pump would be running on minimum recirculation line.) I have seen in another project that a pulsation dampener is used in suction side of boiler pumps. I surmised that this equipment is used as a means of preventing cavitation due to above mentioned phenomenon. Am I correct? If so, how do we size a pulsation dampener for such a case?

Thanks in advance

 

[BigInch]

Instead of a tank or a pulsation dampner, you may be able to use a large diameter suction line between both pumps.

The best way to design that larger diameter line is to simulate the start-up of the system including transient flows and watch the drawdown of pressure, if any, in that suction line. If it goes below NPSHr at the inlet to the second pump, you will have cavitation. In that case, you should consider a higher minimum flow for the transfer pump, or include a pulsation dampner.

The larger suction line, pulsation dampner must effectively hold enough fluid to sustain the flow during start-up of the feed pump, while the transfer pump is accelerating to a higher RPM. If you know how long that will take, a conservative design might simply provide a volume reservoir of a size to hold that feed pump flowrate x transfer pump acceleration time. I wouldn't think acceleration of the transfer pump will take more than a second and a half or so, which is why I think your best bet might be to try a large suction line diameter in a simulation and see how it does.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[chicopee]

Sound like you may have a NPSH problem with warm water entering the pump intake.

 

[JRLAKE]

What is the water temp and how long is the run?

Reading your description it sounds like the suction of the boiler feed would be under pressure at start up, since the condensate pump is running on the min flow recirc. I don't think I understand your concern. I don't think it will hurt the boiler feed to cavitate for a few moments upon startup.

 

[processeng01]

Dear All,

Water temperature is between 120-140 deg F. The piping between two pumps is around 200 ft. The condensate pump will keep the line pressurized at approximately 30 psig.

The more I think about it, the more I believe that a water hammer is more of a concern than the cavitation.

What I don't know is how to consider/model initial transient when the second pump accelerates and sucks water. (Will it cause any pressure fluctuation, or worse a water hammer?) I can have the pump supplier to provide me with some data but I don't know where to start. Appreciate any insight.

Thanks

 

[Artisi]

firstly you need to establish what are tying to solve, either the likelyhood of cavitation or the possibility of water-hammer - 2 distinct and unrelated problems.

From the limited info it is unlikely that any meaningful answer can be given - it will depend on the relationship of the start-up proceedure of the boiler feed pump and the change from a by-pass situation to full supply of booster pump.

 

[BigInch]

Waterhammer during pump starts can be a problem too. Best thing to do is to keep the velocity down as low as possible by use of large diameter pipes.

The real beauty of a transient simulation is you get to see all the problems and correct them before its built, instead of the usual after the fact makeover.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[processeng01]

I have done some homework by surfing on internet. There is a device called suction stabilizer used mostly in positive displacement pumps (not centrifugal) Also acceleration head calculation is mentioned on the net. But all of these do not seem to be 100% relevant to my original question. With a pressurized suction line, most probably a transient simulation may not be needed.

 

[BigInch]

You do have centifugals, right?

You can't really say for sure that transients will not cause problems, unless you somehow look or analyze the acceleration times of each pump, the transfer of fluid into the suction and the take of fluid from the suction. If you have a high pressure suction line, your chances for maintaining a pressure above NPSHr go up, but your potential for exceeding line pressure maximums also go up with a pump trip, etc. Best bet is first trying a medium pressure large diameter suction line. If you do a simulation, its a lot cheaper fixing that input file then it is to fix the real piping system. Your choice.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[Artisi]

What size and type of pumps are we discussing? You are asking involved questions without giving any real data.

 

[processeng01]

Artisi,

The first pump's (condensate transfer) capacity is 130000 lb/hr with a head of 70 feet. Boiler feed pump which is downstream of the first pump has a same flow capacity with much higher head, namely 1500 psig. Operating temperature is bet. 120-130 deg F. Two pumps are separated by a distance of 200 feet. Pipe size is mostly 4" except for boiler pump suction pipe spool section which is 6".

 

[Artisi]

This is not a area in which I have any first hand experience but would think this is nothing very difficult- it must be an application that is used around the world in in numerous power plants.

My approach would be:- prior to the boiler feed pump (BFP) coming on online the condensate transfer pump (CTP) which you say will be on by-pass needs the by-pass valve to close down and at the same time the discharge valve opened allowing it run up its curve pumping the condensate to the BFP - at some point, once the CFT is running up its curve the BFP can be called into service - as it runs up to speed the CFT will increase flow (head will reduce) to meet the demand of the BFP . I guess the trick is that the CTP needs to be delivering demand flow at the same time the BFP come to full speed producing full flow and pressure.


Can the BFP be forced to run up its curve for a short while the CTP gets upto flow and pressure - this will reduce the likelyhood of cavitation in the BFP.

Provided somewhere near to full flow into the BFP is available as it comes to full speed I don't see any cavitation problems - and if there were it would only be for a few seconds.


There are a couple of regular posters here who will have experince in this area - cross your fingers that they see and respond to this for you.

 

[BigInch]

Cavitation will not be the problem. Even if it happens, it will/should stop as soon as the transfer pump winds up.

Waterhammer, on the other hand can be a severe problem. Starting electric driven pumps is a matter of a few seconds, but waterhammer happens in a split second. In this case it will be important to check start-up, failed start, and a trip of each pump. Do you have a relief valve in the suction line? Do you need one? Do you have check valves? Do you need them? IMO, checking waterhammer is not optional. Ask Wisconsin Electric.

**********************
"Pumping systems account for nearly 20% of the world’s energy used by electric motors and 25% to 50% of the total electrical energy usage in certain industrial facilities." - DOE statistic (Note: Make that 99.99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[Artisi]

As you are in the design stage think about increasing the line size between CTP and BFP from 4" to either 6" or even 8" - this will have the effect of increasing the reserve available which was your initial concern.