Failure during Hydro Testing of 36" Ball Valve

[Inchtain]

Valve Data:
Borsig Ball Valve 36" ANSI 600 RJ
Type: Trunnion mounted/Top Entry
Process Fluid: Gas
Operating Pressure: 490 psi
Max. Allowable Working Pressure: 830 psi
Max. Allowable Working Temperature: 110 Degrees Celsius
Ball & Seat Rings Material: F316
Year Built: 1977
Test Pressure: 2175 psi

This valve was initially hydro tested in 2020 after general overhauling following long period of operation to keep it as spare (Body test Pressure kept at 1500 psi due to fabricated blind flanges leaking, while ball was tested at 1585 psi from both sides) Test Passed.

The same Valve was tested again in July, 2024 for confirmation at 1250 psi and 940 psi for the body and ball, respectively. Test Passed.

It was retested again due to the third party witness request in February 2025 at 1250 psi Shell, 940 psi Seat (High Pressure Closure) and 80 psi Seat (Low-pressure Closure). Test Passed.

After all these tests and before installation of this valve to replace an existing valve at an export line, but within the concession of another company, this company request to test the valve again at their premises in the presence of our representatives and the witness of their engineers.

The body test was conducted with no leakages at 1900 psi.
The ball first side test was also successful at 1600 psi.

Shockingly, and during the final stages of the gradual increase of the test pressure of the ball second side (at 1200 psi), a loud voice was heard and the pressure started to fall very slowly.

After releasing the pressure and subjecting the valve to visual inspection, the ball was horribly cracked (photo attached).

Personally, I can say that this is the first time I encounter such failure and I would like your support in analyzing such incident and investigate its probable causes.

Thanks and best regards,

 

[goutam_freelance]

Since the valve was tested repeatedly at comparatively high pressures(max allowable 830 psi, first test pressure 1580 psi), it is possible that the ball had yielded at such high pressures. With repeated loadings, the ball might have cracked due to low cycle fatigue (LCF).

 

[LittleInch]

We don't have much to go here.

Two photos with no real idea where they are.

Need more photos and a diagram showing where in the ball it is.

Has the ball been removed and examined?

It looks? to me like an internal excess pressure inside the ball, but difficult to say.
What types of seats have you got?

But if the MAWP was only 830psi, why were you testing the seats and ball to such a high pressure? Was the valve actually designed for a higher pressure in terms of the ball?

 

[Inchtain]

Thanks for your feedback

I have attached a video for better understanding of the crack location where the ball is not dismantled yet.

The valve has seat rings (F316) with seal rings (Lauramid).

According to API 598, the test pressure was calculated in accordance with the valve rating @ 38 degrees Celsius. (600 #)

Best regards,

 

[LittleInch]

Some crack.

Do you have a diagram of the valve /sectional drawing?
A data sheet?
Why were the previous two tests only done at 940 psiu - which seems a more reasonable pressure to me

 

[Inchtain]

The crack was at the upper side of the ball and we have recently discovered a tiny crack at the center (does not extend to the sides) of the lower part of the ball.

I have attached the valve sketches

The latter pressure value was requested by the partner Company where the valve was to be installed.

Do you think that there is a relation of the SCC of SS316, as there is shallow pitting were noticed on the internal surfaces of the ball?

We also think of the probability of that the ball was not perfectly set on position and during the actuator operation, the damage was occurred?

or might subsurface cracks opened up to the surface?

Thanks

 

[LittleInch]

What were the seat types? SPE? DPE?

Was there a body pressure relief?

This to me looks more like at some point the ball was subjected to an internal pressure it couldn't handle, whether that was trapped gas which couldn't escape fast enough or liquid at some point in its 50 year life.

Yes rotation of the ball in excess of max torque or if the ball wasn't centred may have had a n impact.

Time for the scrap yard.

 

[Inchtain]

What were the seat types? SPE? DPE?

Was there a body pressure relief?

This to me looks more like at some point the ball was subjected to an internal pressure it couldn't handle, whether that was trapped gas which couldn't escape fast enough or liquid at some point in its 50 year life.

Yes rotation of the ball in excess of max torque or if the ball wasn't centred may have had a n impact.

Time for the scrap yard.

The seats are Double Block and Bleed

The pressure was depressurized after the body test