Online inspection of Heat Exchanger Tube

[Anton Sudiantoro]

Dear All,

Are there any method to inspect the tube of heat exchanger correlation of thickness loss?
Right now, inspection done by ECT during offline.
Considering plant availability, studying about if any method to determine the thickness loss of tube during heat exchanger in service

Thank you

 

[DBreyer]

There are tubes with embedded sensors available (

https://www.materials.sandvik/en/products/services/digital-monitoring/intelligent-tube-system/),

they allow some measurement of the wall loss.
Alternatively you could analyse the process stream before and after the exchanger for iron content (if you use steel tubes). This allows a qualitative estimate if wall loss is happening.

Daniel Breyer
Inspection Engineer

http://lnkd.in/dFcQA-w

 

[Strong]

You might be able to detect change in tube vibration properties (damping, amplitude and natural frequency) as tube wall reduction occurs whether at mid span or at tube sheets and supports.

Walt

 

[Anton Sudiantoro]

Dear Daniel and Walt,

Thank you very much for your reply.
Looking at the link which you are given, it is quite interesting.
The problem is, heat exchanger now in service which not possible to install the sensor and others.

I am interesting about detection of damping, amplitude and natural frequency which Walt's mentioned.
Could you give us a detail about that?
What method for checking and so on

 

[EdStainless]

Anton,
So you are positive that your tubes are experiencing uniform general corrosion?
No pitting and no erosion damage?
Then why wouldn't simply inspecting at turnarounds give you enough data to predict life?
There are no satisfactory on-line HX tube assessment methods.

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

 

[Anton Sudiantoro]

Dear Ed,

Sorry for not complete explain about the problem.
This is the case of Ti tube.
Damaged not coming from erosion corrosion inside the tube, but in outside tube due to vibration
Baffle is carbon steel. Inspect for the other exchanger with same characteristic, baffle is heavy corroded
So, gap between tube and baffle established and vibration occurred --> made damaged outside of tube

We have plugged the damaged tube as the immediate countermeasure.
But, there is also same heat exchanger with the same characteristic at our second train
We want to understand the condition of that exchanger's tube during plant operation since it is not possible to stop right now

So far, we hear about the detection by using acoustic
But, not yet come into the detail

 

[EdStainless]

Is the failure related to wear or fatigue cracking?
Since you plugged tubes I presume that you extracted a couple for analysis.
Ti gr2 is soft, has a very low modulus, and is very prone to fatigue.
You really need to see if you can reduce shell side flow enough to reduce the vibration of the tubes.
You can use simple acoustic sensors to listen for the vibration.
And I hope that you have re-designed these and have the replacements on order.


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

 

[Strong]

Anton,

I can't advise you how to do it in a forum like this. There are several techniques that could be considered that depend a lot on type and size of heat exchanger, access to shell surfaces, and background sound and vibrations produced by connected pumps and other nearby machines. Sensors can be of several types: accelerometer on shell (contact 10-20-kHz), audio microphone close to shell surface (20-18-kHz), ultrasound microphone close to surface (40-kHz), or ultrasound contact sensor on shell (40-kHz or higher). Noncontact sensors can be scanned along defined paths, while contact sensors can be fixed or moved to discrete locations on an array arrangement. Measurements can be with portable hand-held, temporary, or permanent analyzers. A lot of the analysis would be to compare 2 or more heat exchangers, different points on same exchanger, varying flow/temperature conditions, and over time (trending). The "need to know" heat exchanger condition would have to exceed the work and cost required for the measurements and expertise to analyze the data! I say this from direct experience.


Walt