Cause of water hammer in a trap header? 4

[Tremolo]

I have a trap header that receives condensate from several different steam traps (220 to 240 F condensate temperature). Fluid from the header goes through a vertical rise of 8 feet and eventually discharges to a deaerator tank. The problem is that when condensate (200 F) is pumped from a flash tank into the trap header, strong water hammer events are observed.

Any ideas as to why the waterhammer events are occurring and what can be done to eliminate or reduce the water hammer?

Thanks for your help

TREMOLO.

 

[katmar]

Tremolo,

Thanks for the update. It certainly seems that you and hndler have very similar problems.

The post by Steven van Els in the thread started by hndler made me realize that there is an inconsistency in your information.

You gave the deaerator pressure as 15 psig, and the temperature of the condensate from the traps as 220 to 240F. But the saturation pressure of water at 240F is 10.3 psig. How does this low pressure condenste enter the trap header? Is the condensate sub-cooled at a high pressure?

If the condensate is coming from normal traps then it will have to be at a higher pressure than 15 psig, and at a temperature of more than 250F. In this case the condensate WILL flash and the steam will collect in the highest point, which is the horizontal section "I", and perhaps "C". When the cold condensate from the flash tank hits this pocket of steam there will certainly be hammering.

But I cannot see how this would lead to periodic (cyclical) hammering? What is the time between hammer events? Are they all of the same magnitude, or do they decrease until the pump stops?

regards
Katmar

 

[Tremolo]

Katmar,

Based on what you said, the following scenario may be the source of steam.

Condensate from the steam traps flows into the trap header because the steam line is at a higher pressure than the trap header. Presumably, the condensate is at the saturation temperature of the higher pressure steam line. Therefore, some flash steam should be produced as the saturated condensate flows into the lower pressure trap header, or as it moves down the trap header piping and enters the riser to the 617’-8” elevation (section "H").

It could be that during normal operation the steam trap condensate flow is small in comparison to the normal flow through the trap header. Therefore, the hot condensate will mix with the trap header flow and remain in a liquid state. However, during an outage, the trap header flow is low or stagnant, so as condensate flows into the trap header, the trap header fluid gradually heats up. Eventually, a hot fluid pocket could develop around the steam trap connection. Then the condensate would not be cooled as it enters the trap header and a significant amount of flash steam could be produced. Over time, the flash steam could then accumulate and form a larger steam pocket.

Periodic operation of a flash tank pump adds 200 F water to the trap header which would initiate collapse of the steam pocket and lead to a waterhammer event.

Tremolo.

 

[kirsner]

I think the key point is--does the hammer take place upon initiation of pump operation. If so, you've got column-closure waterhammer and there is no need to theorize a "rapid condensation event" due to subcooled condensate surrounding saturated steam. The hammer is strictly due to the velocity of the water in the pipe as motivated by the pump as it rushes in to a steam void. The steam void, as i understand it, will condense as it is compressed by the on-coming column of water so as to only impede it's velocity slightly.

wayne kirsner (kirsner@kirsner.org)

 

[katmar]

Tremolo,

I agree 100% with your last post. The solution to the problem has to be to eliminate the high point where the flash steam can collect. You will need to eliminate section "J", and shorten section "H" so that "I" and "K" are on the same level. Even better than having them on the same level would be to make "H" even shorter so that the line "I-K" slopes upwards towards the deaerator tank.

If you do this the cold condensate from the flash tank pump will be able to push the hot condensate ahead of it and never come in contact with a pocket of steam. However, when it hits the deaerator the cold condensate may cause problems there?

Katmar

 

[Tremolo]

kirsner,

Yes, the waterhammer occurs following pump initiation. The system operator tried a new procedure that involved throttling of a flow control valve to reduce the pumped flow rate. The pumped flow was reduced to 170 gpm into the 8” trap header. Even with this procedure in place, strong waterhamemr events were observed. So, it seemed that this was not purely a column closure event.


katmar,

Rerouting piping, naturally, is easier said then done. For instance, section “H” is a vertical riser that goes up to the ceiling and then passes through a wall into an adjacent room. The aux boiler deaerator sits on the floor of the adjacent room. There are also some asbestos concerns. It may not be easy, but we should take a closer look at this option.

Another solution we are considering is the addition of a surge suppressor to arrest the water hammer pressure wave. We have had success with surge suppressors in other applications. The surge tank dampens the peak water hammer pressure, slows the rate of pressure rise, and mitigates any pressure ringing in the system. Any thoughts on this approach would be welcome.

Thank you for your continued interest and help with this this problem.

P.S.

In response to an earlier inquiry for more information, here is a table that includes pipe diameters and lengths. The information is provided for pipe sections labeled A through K from my earlier sketch. The information is nominal diameter and length:

A – 4”, 24’
B – 4”, 30’
C – 4”, 48’
D – 4”, 10’
E – 8”, 7’
F – 8”, 4.8’
G – 8”, 5.75’
H – 8”, 13’
I – 8”, 34’
J – 8”, 2.75’
K – 8”, 1.5’


Tremolo.

 

[lilliput1]

Run your condensate to the flash tank (slope lines down to it). Then pump the condensate direct to the feedwater heater. This pumped line can go up & down providing you put a check valve at the pump discharge. You may have to throttle the pump flow. You can also pipe other condensate direct to the deaerator so long as it slopes down without rising up.