Cooling Water Service Heat Exchanger Hydrotest Tube Leaks

[Mechinspqoal]

Hello,

I have a question about a heat exchanger which is tube side service cooling water and shell side propylene. It is fixed tubesheet heat exchanger the shell side's hydrostatic test pressure is 38 kg/cm^2 and the tube side's hydrostatic test pressure is 9 kg/cm^2

During maintenance activities we clean the tubes first then make them dry and after that we apply shell side the test pressure and we detect the leaked tubes, we plug and weld the leaked tubes and retest. When there is no leak anymore at the pressure test we finish the shell side test and step into tube side test. After we successfully complete the tube side hydrotest we start up the equipment. The problem is starting this phase, after start up, operation team detect tube leaks and we go back to maintenance activities. This time at 20 kg/cm^2 we see a few leaked tubes and we apply all the routine activities as I mentioned above.

So basically we plugged leaked tubes and the test passed at 38 kg/cm^2 after start up we detect more leaked tubes. Operation pressure is around 20 kg/cm^2. What is the reason that we could not detect the last leaked tubes in the first place? Do you think that it can be related with the tube side service which is cooling water and can leads to some scales inside the tubes? Have you ever experienced it? Can a scale hold a hydrostatic test pressure such as 38 kg/cm^2? What can we do to prevent this kind of maintenance returns?

Thank you very much,

 

[EdStainless]

You did cut the tubes that you plugged, didn't you?
It is possible that rapid heating and/or pressurization is causing enough stress to open leaks that you failed to detect earlier.
To start with you need a more sensitive leak test.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Mechinspqoal]

@EdStainless

Thank you for your reply. We weld the plugs to the tubesheet.

 

[shvet]

What is the reason that we could not detect the last leaked tubes in the first place?
Thermal expansion

 

[georgeverghese]

What is TEMA configuration of this fixed HX ? AEU, BEU or NEN or ? How many passes on tubeside ?
What is operating temp range on both sides ? And design temp on each side ?

 

[Mechinspqoal]

@georgeverghese

It is AJM and the number of pass is 2 on tubeside.

Op. Temps; Shell: 51 °C and Tube: 42 °C
Design Temps; Shell: 90 °C and Tube: 70 °C

 

[georgeverghese]

What is minimum CW feed temp, and what is max propylene feed temp ?

Are tubes seal welded to tubesheet or only roller expanded ?

Divided flow shell TEMA J is used for low shellside pressure drop application. If differential temp expansion is part of this problem, may be order a new AJU unit and ask for seal welded tube to tubesheet joints. Poor quality open loop cooling water with high risk of MIC may also be part of this problem ?

 

[Mechinspqoal]

@georgeverghese

Thank you for your contribution.

Minimum CW feed temp is ambient temperature which is around 10-15 °C and for propylene feed we may consider it is max 70 as inlet line's flex. temp which is also shell's design temp.
Tubes are strength welded to the tubesheet.

 

[georgeverghese]

Standard design practice is to spec the same design temp for both shellside and tubeside for obvious reasons. Open loop cooling water systems are always a challenge to run and maintain due to scaling, MIC, and can lead to serious safety issues when there is migration of flammable fluids into cooling water through internal leaks.