How Many Times Can You Repair A Pressure Vessel Safely

Exactly the point. GA.

If the OP goes back to the their records on time between failures, then they might be able to predict how long it will take between repairs and how many repairs can be attempted. This will certainly be a random process curve with probably a PDF exponential curve. Depending on the resolution of the data taken of these repairs (who did the repair, the nature of the repair, etc.) one maybe able to identify all sorts of trends.

But we do not have this data to work from.

If I am understanding you correctly you are welding over top of multiple prior welds. This is going to eventually lead to issues of HAZ hardness, grain growth, and depending on the chemistry of the weld members, precipitation of undesirable phases in the grain boundaries. (Probably not an issue here with this weld. Milo could speak more authoritatively however.)There may also be corrosion issues depending on whether the gas has had odorant added or has significant amounts of CO2, H2S and/or water content as well. Does the Vessel have to meet NACE MR0175 for corrosion? If so, HAZ hardness is going to be the canary in the coal mine here. If there is CO2 and water in the gas, you could end up with "ringworm" corrosion in the HAZ of the welds which will he attacked preferentially to the base metal if the area is not fully normalized (but if you only locally normalize it instead of doing the entire vessel, you will merely move the HAZ to somewhere else)

There will be very little in the way of liquids, this compressor sits on the down stream side of a Low Temperature Seperation plant and supplies pressure to a dry gas sales line. I agree that the metallurgy in the area of repair will change over the course of time with multiple repairs. The repairs are being made by a certified Code shop with the appropriate welding procedures, stamps, and AI oversite. The vessel is not very large and could easily be stress relieved, at least as easily as doing a localized stress relief. This vessel is not required to meet NACE standards at this time.

Thank you for your response.

If this is occurring on a "regular" basis one should come to the conclusion that there is a definite design problem and one "needs to go back to the drawing board," as you yourself stated..........not doing this may eventually cause a costly repair and or serious injury or death.............time should NEVER EVER replace sound engineering practice.

Thank god it was only a pressure vessel!!!!

Only a pressure vessel, what could possibly happen.

They have doing this for years, they have three identical vessels that they swap out, so they can fix one, have one inservice, and one on standby in case they don't get the one fixed fast enough. They seem to have a very good idea about how long they have before it starts cracking. They are "Saving Money" this way, till I asked how much does it cost each time you repair it, and how much would it cost to build a new one that doesn't have this problem, how much have you spent so far repairing these vessels. Then it was suddenly cost effective to buy a new one that won't crack. For people who are all about the bottom line, they can be so short sighted. They have a stringent safety program that apparently does not extend to mechanical integrity, but I think I have their attention now.

Thanks for the response

If this is the case, that they have two others that they swap out, surely time is not an issue for a re-design since they have two others in various stages of repair one would assume that could stand in for the failed vessel while a fourth, new design is being designed/built?

This type of question had been raised at many form and no concrete reply had been made by any body. Experts says till the quality of the material of construction is within accepatable range of parameters with respect to standards, material can be used and repaired.

Suction dampening bottle nozzles are subjected to vibrations and cyclic loading which causes work hardening and ultimately failure in fatigue. The repair procedure shall be such that material's physical properties are achieved after repair as per the standard material by heat treatment etc.

In normal practice standard do not intend for heat treatment for SA516-70 plates with 1/2 inch thickness fabrication, but this type of damaged material either shall be replaced of shall be reclaimed by heat treatment to get the required properties.

I had not considered the work hardening aspect of the loading, good point. I have heard this question asked myself many times, but like you I have never heard a concrete answer or really even good rule of thumb. Typically I'm against more than 2 repairs on components in pressure service, even in new construction I prefer they cut something out and replace it rather than repair it a third time. I have worked for some companies where the rule was fix it once, if it has to be fixed again cut it out.

Thank you for the response

At our plant we are following almost similar philosphy. Other then small repair of crack, we are cutting out the damaged plate and replacing by new with new reinforcement pad. This is giving satisfactory repair and good time between failures.

"Suction dampening bottle nozzles are subjected to vibrations and cyclic loading which causes work hardening and ultimately failure in fatigue." If this is the root cause of failure, then it can be fixed by just introducing a compensator between compressor and dampening bottles to absorb vibrations. Compensator may also have less life but it is easier to replace and cheaper.