Hammer in Reactor Dome Dimple Jacket

[RJB32482]

Hello,
We think our reactors are experiencing hammer in the dimple jacket that encompasses our reactor dome. The dome cooling liquid is treated water. Is there any considerations that might be helpful to preventing this hammering from occuring (less or more water flow into jacket)? My guess is that steam pockets might be forming in the jacket and cool water hitting the steam is causing the hammer.

Thanks.

 

[Latexman]

Put some backpressure on the CW in the dome jacket, or increase flow?

Whats T in the shell? Whats T and P in the dome jacket? Is the controlling heat transfer coefficient on the shell or jacket side? Or is it the metal?

Has it always done this or just started? How long has it been in service? Has the flow changed?

Good luck,
Latexman

 

[RJB32482]

I don't know exactly the parameters. Isn't a dimple jacket just a plate with dimples in it to control flow and induce more heat transfer? So you mean shell as cooling side and tube by the fluid inside of the reactor?

Would adding a backpressure valve on the outlet of the cooling water helpto decrease this hammer? Is this option because this would increase P in the jacket thus the water would have to reach a higher T to transform into vapor?

My though was to increase flow so the flow characteristics in the jacket so there would be no dead spots? Would this help?

THanks for the reply.

 

[rmw]

It would still be helpful for us if you gave us some parameters to aid us in our evaluation of your question.

rmw

 

[RJB32482]

OK here's some parameters from my estimates:

Jacket cooling water comes in about 70 degrees F. The temperature in the reactor is around 250 F. Pressure if I remember correctly is around 10 psig. Heat transfer coefficient is on the jacket side (I believe tube side then).

Thanks

 

[rmw]

If you raised the operating pressure of the jacket to 16 psig, then your 250 degree process would never boil the water. However, it sounds to me like just increasing the throughput of the jacket until the hammer stopped would solve the problem. It appears that your water flow is so slight that the process is boiling the water in the jacket.

Before you do increase the water flow, which would increase the cooling, however, you should check the process side in the dome to be sure that adding cooling water would not be detrimental to whatever is going on inside the 250 degree dome.

rmw

 

[RJB32482]

Thanks for the reply. If somehow the water input to the dome is not able to be increased, can adding a backpressur valve to the outlet of the water help increase pressure in the reactor head, thus decreasing boiling of the water in the reactor head? Any other ways to increase pressure in the head or remove the water hammer?

Thanks.

 

[katmar]

Hi RJB32482,

Two things to note:

If there are no baffles inside the dimple jacket then any bubbles that form will tend to rise to the highest point. Can you put the water outlet at this highest point so that any bubbles will be removed as soon as they form and they will not have a chance to join up and make large noisy bubbles?

Be careful of increasing flowrates or back pressure valves or anything else that will increase the pressure inside the jacket. Jacketed vessels fail a lot more frequently due to external (i.e. jacket) pressure than because of anything else. Check the allowable design pressure in the dimple jacket before you make any changes.

 

[RJB32482]

Do you mean the external pressure inside the jacket or external pressure on the jacket. Are dome jackets usually designed for internal pressure + external vacuum (14.7 psi)

 

[katmar]

I mean the pressure inside the jacket, which is external to the vessel shell. A dome will support much less pressure on the convex side than on the concave side.

I have seen failures where the jacket still looks perfect from the outside, while the shell inside is destroyed. The outer shell of the jacket obviously has the pressure on its concave side, but for the vessel shell it is on the convex side.