Repetitive Firetube leaks on 2 Steam Boiler serving a Laundry

[EidEEid]

We are experiencing repetitive firetube leaks on two steam boilers serving a laundry.
(York-Shipley 100HP, 150 PSI design, burning Diesel Oil)
This has been dragging for some time, and we are sure that we have no issues in Water Treatment, no short-cycling, no batch loading, no thermal shocks.

The boiler are turned on in the morning, and are operated until early afternoon, then turned off. To make sure that we are not feeding the boiler with cold water at the morning startup, we have added an electric heater in the feedwater tank, to keep it at 60C.

All the leaks ever experienced were tube attachment leaks (tube attachment weeping), never tube ends cracks. There is no correlation in the location of the tube failures.

- Boilers were started up Nov.2014
- Feb. 2017 : Reroll boiler A (few tubes)
- Mar. 2017 : Reroll boiler B (few tubes)
- Apr. 2017 : Reroll Boilers A and B (few tubes)
- May. 2017: Reroll Boiler A (few tubes)
- Feb. 2018 : Reroll Boiler A (All 2nd pass of tubes)
- Dec. 2018: Electric Heater installation in feed unit.
- Dec. 2019: Reroll Boiler B (All 2nd pass of tubes)
- Dec. 2020: Replacement 2nd pass of tube Boiler A
- May 2021: Replacement 2nd pass of tube Boiler B
- Dec. 2021: Reroll Boiler B (10 tubes)
- Apr. 2022: Reroll boiler A (5 tubes).

We are really out of arguments as to why this could happen. Would anyone around think of any possible cause?

 

[Guest102023]

Unless you pull a sample and have it investigated by a failure analyst knowledgeable about boilers we can only speculate.

"If you don't have time to do the job right the first time, when are you going to find time to repair it?"

 

[EidEEid]

Thank you @brimstoner

Actually, we do not have access to a failure analyst where the jobsite is. (Lebanon).
In all cases thank you, and if anyone else can think of anything, that would really be helful.

 

[Cobra17]

Were the firetubes designed by code calculations or FEA? and if code calculations did you use BS5500 methodology or go with VIII-1 and apply it as close as possible to the design?

If you run FEA on a VIII-1 code designed firetube, there is a fair bit of deflection on the flange face that might be causing the leaks. And in some cases if you calculate it to VIII-1, it won't pass VIII-2, mainly due to the stresses at the nozzle to shell joint. BN5500 isn't much better than VIII-1 in terms of relating it to an FEA and VIII-2.