Long Internal Inspection Intervals - via RBI or Code 1

[robsalv]

Folks,

My plant is in the middle of an RBI review program.

The regulatory regime is a self regulating safety case based environment where we are held accountable to our own engineering judgement.

We see some great opportunities to apply long internal inspection intervals on critical process vessels (24 - 30year intervals) which RBI reviews have come up with very low risks - all things remaining the same.

We're struggling with the concept of not inspecting something for so long, given that you don't know what you don't know. We're not talking air receivers here, we're talking stainless and carbon still vessel in propylene and ethylene refrigeration circuits that do not experience corrosion or other degradation mechanisms e.g. too cold for CUI.

Is anyone else in eng tips land running with long inspection intervals?? Tell us about it! How'd you get there. What kind of controls do you have around these intervals? What's the largest interval you allow?

Thanks in advance.

Rob



p.s. Apologies in advance, but I will thoughtfully cross post this request in various forums in due course. Thanks.

- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."

 

[BigInch]

Inspection intervals are frequently set by determining a particular limit of a defect that you can find with your inspection equipment, or by your inspection proceedure, against an estimated rate of growth of that type of defect. Then, starting with an assumed undetected defect (ie. with a size or some value just under your last inspection's detection limit) project an estimated time to failure. Back up along that time path enough to include an adequate safety margin for the equipment's degree of criticality, risk level, errors in possible growth rates, etc. and you reach the time interval within which you should conduct your next inspection.

**********************
"The problem isn't finding the solution, its trying to get to the real question." BigInch

http://virtualpipeline.spaces.live.com/

 

[SJones]

http://news.hse.gov.uk/2009/09/23/r...rnal-examination-of-high-hazard-process-plan/

Steve Jones
Materials & Corrosion Engineer

http://www.oilandgaspeople.com/cv/11499664

http://www.linkedin.com/pub/8/83b/b04

 

[dcasto]

Think about air seperation plants. I worked at a He recovery plant that was built in 1963. All the vessels were in a sealed cold box. 45 years without an inspection. An other plant was a cryogenic NGL recovery, built in 1966, no inspection, ever on the cold end (below 20F).

 

[robsalv]

BigInch, yep, that is fully understood and is correct practice for a time based degradation mechanism. The question is where you can't determine a defect/degradation mechanism!

SJones, that report seems like exactly what I'm looking for. It's just downloaded. I'll be getting a coffee and making time for a good read as soon as I hit submit. Thank you very kindly sir. Star for you.


dcasto, your experience is exactly what I'm looking for. Did you step into a plant and inherit the indefinite intervals or did you deliberate over them? In my own mind (15+yrs PV asset reliability experience) I figure that once the vessel has been shaken down and had at least one inspection to see whether a manufacturing defect has broken through in the initial part of service, there's little need to inspect any further when there's no degradation driver. Have your vessels had any kind of inspection since being placed in service?

Not all the vessels in question on our plant are cryogenic service - some just don't have any internal degradation mechanisms that we can identify by a review of operation or available inspection history. However, the efficacy of past inspections isn't what you'd expect today and the "you don't know what you don't know" part of the equation bothers me.


Just out of interst, this question comes from Australia, where the inspection code is known as AS3788. It has an appendix which deals with cryogenic storage tanks and recognises that bringing them out of service for an inspection could damage the tank, so inspection is not recommended (providing certain criteria are met). The code is just not as clear on cryogenic service or non degrading service vessels.

Thanks for the replies thus far.

Cheers

Rob


- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."

 

[dcasto]

The RBI system I created is home grown of sorts. It is based on API and the Best practice for risked based inspection as a part of plant integrity management Contract research report 363/2001 HSE Books.

After assembling your pipe and equipment list you develop a series of items that affect the increase or decrease in the probability of a failure of the equipment and a list of things increase and decrease the consequences of a failure. Things like H2S in a vessel increase the probability and the consequence.

We then assigned point values to each of these items and added up the points. For example, equipment with free water has a higher potential for a leak and a resulting release. If you have vibration, H2S, cyclic service, pressure, all kinds of things of that sort, you add points. Likewise with consequence. Smaller low stress lines get fewer points and high pressure highly flammable mixtures get more points.

We then use a cut off of points to determine inspection period. The higher the points, the more frequent. ASME has a 10 year maximum interval and good engineering practice would support that. HOWEVER, lines that have only consequence points have no means for a probable failure, so the risk is almost nil. equipment with no know failure point because of metallurgy have no potential for failure.

You can argue about the cut off point, but in any risked based analysis you can, the goal is to manage the risk and drive to zero, not have absolute zero at day 1.

The fact that you document this is what regulators are looking for.