Evaluation of Pressure Vessel LTA Per API 579

[vragle]

I recently performed a rerate of a 55 year old vessel. Based on inspection data there was some general thinning in the feed area, and everything else was in good order. The inspection report mentioned a goove adjacent to a nozzle that was described as “0.080 deep and about 1.5 inches long” There is a picture that show it to be what looks like a mechanical gouge. It is about 2.5 inches from a nozzle and completely covered by the nozzle reinforcing pad.

I evaluated the gouge using Section 5 of API 579 (Locally thin areas) and ignored the proximity limitations imposed by the API 579 evaluation. The evaluation indicated that the gouge would not limit the MAWP. However since there is a proximity limitation imposed by API 579, the evaluation methodology cannot be used. Section 9 of API 579 has similar proximity limitation. To complete the evaluation, I used the thickness at the bottom of the gouge as the available thickness for pressure and it resulted in a MAWP reduction.

QUESTIONS: 1. Is there some other methodology that can be used for evaluation of such a flaw (short of a FEA). Or can anyone direct me to some reference that might help me address another methodology (other than API 579) to evaluate this gouge.
2. Does the fact that the external nozzle reinforcing pad covers the internal flaw contribute any “safety factor” that might over-ride the potential combination of stress (the flaw and the nozzle welds) that is apparently the concern of the API 579 methodology.

 

[pennpiper]

VRAGLE,
Your links do not work anymore.

Sometimes its possible to do all the right things and still get bad results

 

[TGS4]

Because of the discontinuity stresses added by the nozzle and repad, the Level 2 assessment is not appropriate. Your only solution is a Level 3 assessment. I have a pretty good feeling that it will pass, but there is more to such an evaluation than just stresses. A good analyst will be able to guide you through this.

 

[Duwe6]

". . but there is more to such an evaluation than just stresses"

Exactly, such as the *actual* pipe stress applied to that nozzle [nozz loading]. However, what is the Min. Thickness in that area? If the bottom of the gouge is above Tmin [with a Corr Allowance}, buff out the gouge to a taper greater than 3:1 and document that spot. If there isn't enough metal left, preheat & weld. A small 'pick-up' weld repaie is cheaper than a proper evaluation of the flaw; stop talking/analyzing and fix it.

 

[vragle]

I agree totally with the suggestion by Duwe6 for the fix. Exactly what I told the client. I am working with photos and thickness data from the recent TAR and the rerate was done after the tower was restarted. There is only about .005 corrosion allowance without the gouge. The gouge resulted in a thickness less than that required for the desired (original) MAWP. Unfortunately the vessel was designed with no specified corrosion allowance. There is little or no corrosion because the tower has been in operation since 1960 and the excess material thickness of the original plate was sufficient to give 55 year of operation. My problem now is that the thickness at the bottom of the gouge is what I set as the Min thickness. This resulted in a reduction of MAWP that gives a really tight diff press between the operating pressure and relief pressure.

 

[loilfan]

Sorry if this post is late to the party. I hate it when there is no CA in a vessel. If you have any corrosion outside of the manufacturer's tolerance, you basically have to redesign the vessel if you don't have the original calculations. Fixing it definitely seems like the best option for such a localized repair. If it's in vapor/gaseous service, consider using a pilot relief valve if you need to operate so close to MAWP.