Heat Exchanger Tube Failure

[Bilal7590]

Hi All,
We recently carried out Eddy Current Testing (ET) of tubes of a heat exchanger. It had 512 tubes in total. But ET revealed that 11 tubes had wall loss beyond acceptable limit. Hence these tubes were plugged.
My question is that how can I find whether the heat exchanger will perform its intended function satisfactorily after losing 11 of its tubes. What type of calculation is required for this purpose?

Regards,
Bilal

 

[SNORGY]

Is this exchanger in condensing service or is it single phase on both sides? If the former, then specifically *which tubes* were plugged may impact performance somewhat (eg if they were all in the last pass of a multi-pass unit). Otherwise, a 2% reduction in area might have no impact at all since exchangers are often designed with some excess area available. Overall fouling might be a larger concern.

You could use Aspentech to check-rate the unit, if you have access to it; or Promax or Unisim. You could also do some textbook hand calculations, starting with:

U*A*LMTD = m'*Cp*dT (if your LMTD is defined; if not, apply l'Hopital's rule or use NTU-eff method)

and run some scenarios.

Depending on the tubeside service and available pressure drop, if you are shy of desired performance you might be able to use tube inserts to improve your heat transfer somewhat.

 

[EdStainless]

I would suggest that you learn some about ET also.
Was the reference standard supplied with the bundle, or is it from a different lot of material?
How well were the tubes cleaned before testing?
Have they been tested before? Was this done the same? do the results make sense?
Were the artificial indications in the reference similar in size and shape to the ones "found" or was this just and ASME standard?
How many tubes were removed for analysis to verify the accuracy of the ET calls?
Errors of a factor of 3 are easy if the standard does not match the actual indications.
What this means is that you may only have a couple of bad tubes, or you may have 50.

You also need to look at where the tubes are in the bundle and where along the lengths they are the worst.
But in general 2% tube loss will have no impact on heat transfer. Usually the tube side is the limiting side, and this just raises flow velocity in the other tubes which helps offset the loss of surface area.



= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[metengr]

This is why ECT is done for in-service inspection. Contact the OEM of the equipment for determination of tube pluggage effects on performance.

 

[EdStainless]

This is why you should do ET before the unit ever goes into service. That way you have a baseline to compare to.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[chicopee]

From past experiences with boiler inspections and if I remember correctly as this information is over 30 years, utility companies were not concerned about performance unless there was 5% to 10% of the tubes plugged.

 

[MJCronin]

http://www.eng-tips.com/viewthread.cfm?qid=203299

http://www.eptq.com/qandaquestion.aspx?q=32237996-1e1c-4eb8-85ac-8ab308b0f3f9

MJCronin
Sr. Process Engineer