Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Cavitation in tubes of heat exchanger

  1. Nov 13, 2006 #1
    In heat exchanger (U-type) i've problem with erosion/corrosion at beginning of tubes in heat exchanger,near tube-sheet where steam enters.i suspect cavitation is problem. on tubes there's a channel of erosion in the middle of tube where material is gone and tube is very thin in that channel,as tube continues channel is smaller and smaller,and finally there is normal cross-section of tube.medium in tube is 4bar steam used for heating methanol in shell side
    anyone had experience with this? how to try solve this problem?
  2. jcsd
  3. Nov 13, 2006 #2
    Does your exchanger have an impingment plate installed at the inlet nozzle of the shellside? This keeps the steam from impacting the tubes at the shell inlet, causing caviation and errosion. If not currently installed, one could be added above the shell. This does not apply to your problem. I reread that steam is on the tubeside in your application.

    How old is the exchanger?

    Was the proper tube material selected for the fluid being handled?

    Is the errossion inside the tubes or outside?
    Last edited: Nov 13, 2006
  4. Nov 13, 2006 #3


    User Avatar
    Science Advisor

    Steam cavitation. Huh. Ya learn something new everyday. I immediately thought of impingement issues as well.
  5. Nov 13, 2006 #4


    User Avatar
    Gold Member

    Are you desuperheating your steam supply to the heat exchanger to get the maximum amount of work from it? If so, are you injecting the feedwater into the steam very close to the heat exchanger so that it does not have sufficient time to vaporize (erosion by impingement)? If you are indeed desuperheating with feedwater injection, and this is a new problem (it worked fine for a period of time without undue tube thinning) perhaps the feedwater injection nozzle has eroded to the point where it delivers a stream of feedwater instead of a fine spray, again leading to erosion from impingement. Good luck.
    Last edited: Nov 13, 2006
  6. Nov 14, 2006 #5


    User Avatar
    Science Advisor

    Somone please enlighten me; I learned about cavitation as something that happens to liquids when pressure drops too much. The liquid briefly becomes a gas. How can steam cavitate if it's already a gas?
  7. Nov 14, 2006 #6


    User Avatar
    Science Advisor

    I had never heard of it either. I'm curious myself. I did do a search and there is a website here: http://www.jlfiberservices.com/optima_online/refining/vol6/no_2/index.asp [Broken] where there is a mention of a failure mode due to steam cavitation, but that's all there is on it. I can't find anything explaining the phenomena.
    Last edited by a moderator: May 2, 2017
  8. Nov 14, 2006 #7


    User Avatar
    Science Advisor
    Homework Helper
    Gold Member

    Heat exchanger is oversized (wrongsized)? You're doing all your heat transfer in "the channel," collapsing steam bubbles and getting the erosion?
  9. Nov 14, 2006 #8


    User Avatar
    Gold Member

    malkio hasn't been back to clarify, so we're left with little but speculation. Typically, these u-tube heat exchangers have a baffle running down the center and the product in the shell flows counter to the steam in the tubes. This lets the hottest steam contact tubes surrounded by the hottest product and the coolest, wettest steam contacts tubes surrounded by the coolest product, avoiding thermal shock and allowing heat transfer to occur gradually all along the length of the tubes.

    I, too, am puzzled about the usage of the term steam cavitation, since that typically refers to the abrupt collapse of steam bubbles in a flow of liquid, like in a condensate line or perhaps a fuel injection system. Generally, when one thinks about mechanical erosion in a steam system, the culprit is "wet steam". That's why we superheat steam before passing it on to the steam distribution system and desuperheat it to near saturation just before introducing it to its loads.
  10. Nov 15, 2006 #9
    here i am,
    i study my exchanger and cut some tubes,so here's little drawing of my case and picture of part of damaged tube.
    As you see, damage is on the inner part of tubes where steam goes,damaged tubes are tubes on top of bunddle and upper part of tubes at aproximately in the middle of tube.
    i first thought about cavitation but as you said cavitation happens in liquids so now i'm not sure i can explain mechanism which cause that erosion
    as you see from temperatures my steam is not superheated ,it's probably some mixture condensate/steam
    my theory is :mixture of steam/condens. came in tubes,condens. flows in lower part of tube and in upper part of tube is steam,in the middle of tube length (pointed at picture) where methanol (in shell) has lower temp. bubbles of steam begin to colapse (turn in condensate) and in that process devastate tube,maybe tubes on top of bunndle and upper parts of that tubes are most exposed to methanol temp. so that's where most of that transition steam bubbles-condensate hapenns.

    i hope i make thing clear ,if i forget answer or mention something feel free to ask
    i think one of solutions would be different material of tubes or superheating steam
    any suggestions?

    Attached Files:

  11. Nov 15, 2006 #10
    middle of tubes means middle of length of tubes (middle of exchanger)(drawing)
  12. Nov 15, 2006 #11


    User Avatar
    Gold Member

    If that eroded channel is on the top inside surface, you may very well be looking at a cavitation problem, as you suggested. That implies that the heat exchanger is doing almost all of its work (transferring the latent heat of vaporization) in that short little section of tubes near the product outlet pipe, and the remaining length of the tubes is carrying an almost solid plug flow of condensate. The tubes at the top of the bundle are exposed to the hottest flow of product, and the steam in them might not fully condense, carrying a mix of steam and condensate to the next chamber, where the steam bubbles collapse when they come into contact with the cooler tubes (surrounded by cooler product), resulting in erosion. That could be why you see the damage concentrated in the upper tubes just beyond the first baffle. I had assumed (wrongly) that you have a counter-flow heat exchanger that would minimize thermal shock, but from your drawing, the product flow is serpentine, and there is an area that is prone to thermal shock, right where the damage is concentrated. I am afraid that I don't see an obvious solution that doesn't involve re-engineering. Superheating the steam might just push the problem further downstream without solving it.

    The most obvious solution would be to replace the heat exchanger with a vertically-mounted single-pass counter-flow unit. Steam enters the top, condensate exits the bottom, and product flows in at the bottom of the shell and exits from the top. This would minimize thermal shock and if cavitation did occur, it would not be confined to a small section of the tube interior, but would be evenly distributed.
    Last edited: Nov 15, 2006
  13. Nov 15, 2006 #12
    Just curious why this exchanger is setup with steam in the tubes and the Methanol in the shell?

    Is the steam trap operating correctly, adequately sized and the proper type for the process to allow the condensate to be removed quickly enough? What kind of trap is it?
  14. Nov 16, 2006 #13
    turbo-1 tnx
    very very helpful

    it's actually heater of methanol,why this setup i really don't know,maybe because of higher pressure of steam than methanol,hmm you make me thing,what if i switch this(steam -shell,methanol -tubes)??? i don't know

    i just working on checking steam traps along pipes ,but i think it's not main problem cause steam bubbles that condense along tubes devastate tubes not the condensate wihich is not being removed by steam traps,just thinking
  15. Nov 16, 2006 #14
    Pressure could be the reason it is reversed from the norm, which would be to have the more corrosive material in the tubes.

    If the trap does not remove condensate fast enough, this plug will grow larger backing up into the exchanger and reduce the capacity of the exchanger. If this plug moves suddenly, the pressure in the tube drops and the condensate at the steam end of the plug flashes to steam, causing hammering. The steam trap should be working at or near steam temperature, if condensate is allowed to form in the exchanger to that degree, the trap is probably operating considerably below steam temperature. Condensate allowed to cool below steam temperature forms carbonic acid, which also eats and pits piping.

    Your entire exchanger should be filled with steam, with condensate only forming near the trap end. This is all a function of the trap.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook