What sort or gasket materials are you using that have a defined service life?  My first recommendation would be to select a gasket that does not have this limitation.  Second, I would question the basis for the service life.  Many common gasket materials (especially o-rings) have a defined shelf life. But, the shelf life may be based on assumed exposure to UV, ozone, oxygen or other agents.  Once in service, most gaskets would be shielded from these exposures.  A Buna o-ring may call out a shelf life of 5 years. But, installed as the head gasket in a centrifugal compressor, we would not see a problem letting it run for 20 years.  For pipe flanges, we would normally install spiral-wound gaskets with stainless steel (or Monel) wraps and high temperature filler even in services that do not require stainless steel or high temperature filler.  I am not aware of any service life limitation on these gaskets. But, please correct me if I am mistaken.  

I cannot imagine having a program that required us to open up flanges that had never leaked just to replace a gasket based on a service life recommendation unless the basis was very specific to our service conditions.    I would go even further.  Such a program to replace gaskets on a time-basis would probably result in more leaks because of the possible introduction of human error, manufacturing defects, contaminants, etc.  

Additional details may help others provide you with better information.  Raised face flanges?  Ring joints?  Spiral wound gaskets?  Flat gaskets (such as Garlock rings)?  O-rings?  Viton, Buna, Teflon, Kalrez?

I prefer to select a gasket that does not have a limited life and leave it alone.  
 

Johnny Pellin

Thanks for the response Johnny!

Since the integrity management programs are for entire systems there are many different types of gaskets, but the ones I'm most curious about are spiral wound semi-metallic gaskets for raised face flanges, metallic rings on Ring Type Joints, and various gaskets on isolation kits.  I'm not concerned about gaskets on Valves or other equipment such as compressors because they are inspected and serviced enough to assume they are fine, and because they are typically located in controlled areas where ongoing monitoring can be done should they fail.  Rather I'm more concerned about remote locations where SCADA isn't available and inspections are done far too sparse to see if they have failed.

I think that we agree that gaskets don't have a service life published, and its probably because everyone thinks they will last longer than the life of a piping system, but piping systems are getting older than most people ever expected.  I should note that when I talked with a well known gasket manufacturer, I asked them if they would feel warm and fuzzy about a gasket that had been in cyclic service for 50 years and he said that he would not.  We have plenty of piping systems that are over 50 years so I think it's time we start thinking about it.

So I'm going to play devil's advocate on your first recommendation about picking a gasket that has a service life greater than required.  Everything designed for a certain use has a probability that it will fail at that use. Therefore it's not IF it will fail, but WHEN.  The longer you wait, the more likely it will fail so since a long time has elapsed since many of these gaskets were installed my intuition tells me we are going to see an increase in failures. If we are truly aiming for a zero leak goal we should at least look into what the risk of failure is.

On your last point about installation related errors I agree.  However, I would imagine that those happen right at pressurization so could also be closely monitored for leaks.

If I can offer my lessons learned from pipeline integrity,  once one fails all the rest will be doing the same shortly there after if they share a common cause so it's important to mitigate against that so you don't have a sprinkler.

Chad

Now that I know a little more about your gasket applications, I have additional comments. If you are concerned about the failure of anything, you need to understand the failure modes and causes.  How would you expect a gasket to fail?  If the mode is corrosion, for example, it could be time dependant and may be predictable given an estimate of corrosion rate. If it is erosion, the same point may be made. If it is fatigue, then cycles and predicted fatigue life could drive one to a time-based replacement. If there is some sort of predictable chemical attack, then a time based replacement could be valuable.  But, if the failures are primarily random, then time based replacement is ineffective.  And, if there are any infant-mortality cause, then time based replacement will definitely increase risk rather than reduce it.  

Corrosion - If you are looking at SS/graphoil clamped between two carbon steel flanges, then there would be very few situations where gasket corrosion would be the limit to run length.  Unless you have a particular corrosion mechanism that would attack SS, I see no value in time-based replacements.

Erosion - The gasket should never be in a position to erode faster than the pipe.  I see no value in a time based replacement.

Fatigue - The wraps of a spiral wound gasket are compressed. But, the presence of the outer ring and properly tensioned studs prevents the wraps from seeing any significant stress cycles. If you want to worry about fatigue, you are probably better off to look for fatigue in the piping.  I see no value in a time based replacement.

Chemical Attack - I don't know of any modes for SS/graphoil that would fall under this category.  I see no value in a time based replacement.

Regarding infant mortality, I may disagree on one point. Many assembly, materials, manufacturing or contaminants issues may show up immediately.  But, others might not show up for a few days, weeks or months.  I would trust a gasket that has run for 50 years with no leaks more than one installed yesterday.  

Now all of this may sound as if I am disagreeing across the board regarding time-based replacements of gaskets.  However, that may not be the case.  Even if the gasket is not going to fail, I have seen localized corrosion of the raised face in the stagnant gap just inside the wraps of the gasket.  I have seen cracks or other failures of the flanges.  And, normal inspection methods probably can't detect these impending failures (smart pigs, UT, etc.)If a flange is truly remote and if a leak would result in a severe risk from safety, environmental impact or legal liability, it may be worth setting some limit.  After 50 years, it might be a good idea to open up the flange, clean and inspect the flange face and reassemble the joint with new gaskets and studs.

I work in an oil refinery where almost none of our flanges are so inaccessible as to fit these criteria.  But, I have to allow for situations in a pipeline system that are completely outside of my knowledge and experience.  I should also note that all of my comments above are directed toward spiral wound gaskets in raised face flanges. We have relatively few ring-joint flanges and I know much less about them.  I suspect that many of my points apply there, as well.  But, I should leave comments on ring-joints to others who know more than I do.  
 

Johnny Pellin

I had one other thought on this subject.  You described a situation where one gasket fails and others are not far behind. I would doubt that this is likely from a cause that can be addressed with a time based replacement. I would suspect that the piping system was subjected to a water hammer or some sort of off-design mode of operation (temperature or pressure excursion, etc.).  If that is the case, having a freshly installed gasket probably won't avoid the failure.  A water hammer can open up a freshly assembled flange just as well as it can open up an older one.  I would think through the potential causes of gasket failure and recognize that most of them do not fit a long-term wear out pattern where time-based replacement can be effective.  Conduct an FMEA (Failure Modes and Affects Analysis) to explore the possible failure modes in your application.   

Johnny Pellin

Marine exhaust riser gaskets are exposed to fire or seawater or condensed water, or all three.  Fiber gaskets, graphite 'tang', and spiral wound gaskets are all expected to last five years, at which time the risers themselves are probably corroded through anyway.  All survive better if installed with water glass.  Most are replaced long before five years because of turbo failures or maintenance.

Some manufacturers continue to use and specify a bellows-style low force ring gasket on the turbo flange.  I guess it works perfectly on their home planet.  On Earth, it leaks from zero time, stops leaking after fifty-ish hours when it gets clogged with soot, and blows out at some indeterminate time later, usually under full load in the middle of some damn ocean.

 

Mike Halloran
Pembroke Pines, FL, USA

This is an interesting question with a lot of good response.  I think I once heard that some flanged gaskets from hundreds of years ago may have even been made out of stuff like available animal hides (leather), that at least after a few hundreds of years inside of some otherwise incredibly durable pipe joints may have suffered some (I guess thus understandable) significant degradation/deterioration.  Out of curiosity, I thus did a quick internet search and found that "leather" is perhaps not surprisingly (after such aforementioned successful service) nevertheless apparently still being employed for gaskets and packing in some industries (see e.g. http://www.twinleather.com/twngaskt.htm ), and of course there have also been other bio source gaskets like cellulose fibre, cork, etc.!  It is my understanding other materials such as lead and asbestos have also been employed and the latter at least for very a long time, but have fallen out of favor for reasons not necessarily related to lack of durability.  
At some point, I think most/more modern joiners of course gravitated to more resilient materials for most common at least low temperature flanged gaskets e.g. water and sewer services, first natural rubber (variously processed from raw gum elastomer, from rubber trees) and then all flavors of synthetic and composite rubbers from other sources as they were eventually successfully developed in the early 20th century on.  With regard to the former, and as I believe as noted in Olmec ("rubber" people?), Mayans and/or Aztecs etc. historical accounts, some rubber objects also apparently with various interesting processing methods) hundreds of years old but maybe still in pretty good shape have been found in their well etc. archaeological excavations.  
While I think some modern synthetic rubber pipe gaskets (e.g. those per formulations meeting all requirements of AWWA C111 standard) are now required when produced to meet arguably pretty rigorous minimum aging, ozone exposure and mechanical properties requirements (such as compression set etc.), though that may not necessarily be true for some other flanged gaskets (that nevertheless as one might guess from reading the preceding might still be providing reasonable service in many applications, or covered by other standards such as ASTM).  E.g., while I have never analyzed same, I once heard third-hand many years ago that some flat "red rubber" sheet-type gaskets might in reality be composed not mostly of rubber, but with up to nearly 70% clay filler, and with some understandable far lesser physical and exposure some exposure properties than molded gaskets.  While this does not mean such gaskets won't work with, one would think, really little exposure area up inside flanged joints (and maybe even for long exposure in many common services such as water and sewer), it is probably best to direct questions any questions in that regard to the manufacturers of such gaskets for the specific exposure involved.
Other special services such as very high pressures and/or temperatures, and/or more critical fluids/containment as you and others describe, are admittedly a whole other ball game, and I will yield to those with no doubt much more experience than I in those regards.