First Stage Regulator Failure Rates

[grf88]

My son experience a first stage failure on a reg the only had about 20 dives since new. Here is my response to a leak in another thread. I have since come to believe that it was actually a seat failure that set it in motion.

Sounds like the primary diaphragm is leaking and building up pressure in the environmental chamber. This happened with my son's D1 and resulted in the primary diaphragm letting go followed by massive air loss. The environmental seal was shredded in the process. We were on a wreck in high current with a virtual overhead and it was not a lot of fun. As previously stated do not dive this reg until it has been rebuilt, just resetting the environmental seal will not correct the problem. There have been changes to the rebuild kit that modify the way the primary diaphragm is captured either a thicker diaphragm or a plastic ring depending on the kit.

 

[Angelo Farina]

My standard interval service has always been 2 years.
Typically I just change the dynamic O-rings, clean, lubricate and re tune.
If tuning the second stage is difficult, than I change its seat.
If the IP creeps, then I change the first stage seat.
Only this year, during the lockdown, I made a complete rebuilt of them, changing a lot of parts and spending more than 300 euros.
My oldest reg was built in 1975. My newest one in 1980.
In all those years I never changed the diaphragm, the exaust valve, the hoses...

 

[cerich]

. The yoke O-ring failure is a common issue, which deserves to be prevented with additional inspections and maintainance, or converting the reg to DIN, as I finally did.
.

Completely disagree, you said it was because the first stage inlet face was scratched that caused the issue, that is exceedingly uncommon. A oring on the face of a yoke tank valve wearing out is exceeding common.

 

[cerich]

I don't want to derail this tread or give you a hard time, but this statement really bothers me, depending on what you mean by it.

If you mean by service annually as in inspect and do a detail check of the regulator, make sure there is no detrimental corrosion and everything is working correctly, the IP is solid, etc., then I am OK with that statement. But if it is the industry recommendation of a complete tear down, a complete rebuilt, adding new unproven parts that may need to be broken in to play well with the existing equipment, then this goes against one of the most basic engineering principles:
If it ain't broke, don't fix it!

It is well documented that most failures occur just after a rebuilt (or new equipment). It is called "infant mortality". And it happens for a number of reasons (examples: introducing new unproven parts, introducing new parts that could be out of tolerance, and most important the potential for human error, etc.).

In engineering we have a failure rate versus time curve we call the "bathtub curve distribution". It is a well know pattern of failure rates. The highest number of failure occur due to "infant mortality" at the left of the curve and then far to the right you have end of life, due to wear, etc.

You restart the clock on infant mortality every time you tear down a regulator, introduce new parts, and allowing any human hands mess with the piece of machinery.

This is the reason we buy stuff with a warranty. Not because it is not going to fail, but because it will hopefully be replaced before the infant mortality period.

Extremely good quality control can mitigate infant mortality to a large degree, but it is always something that we need to deal with.


You can Google "Bathtub curve" and you will find lots of information.


View attachment 599139


The Bathtub Curve and Product Failure Behavior
Part One - The Bathtub Curve, Infant Mortality and Burn-in

Bathtub curve - Wikipedia

Bathtub Curve Reliability and How Repairs Affect It

what he said

 

[tbone1004]

@Luis H that's what @cerich was commenting against my post. I'm well familiar with the bathtub however my understanding was that the OP wanted to know about failure rates after infant mortality which in this case is almost always due to improper service, so trying to get the random failure rate value.
I definitely subscribe to the "if it ain't broke, don't fix it" on my regulators, and for that I expect "wear out" failure to show up eventually, but it does certainly take a while and I usually see random failures long before something is actually worn out.

 

[Luis H]

I figured you knew what a bathtub curve distribution means... I didn't think you got a mail-order degree from Sears and Roebuck...
But if you are going to make a statement like that...


My comments are also for the benefit of others.

 

[Open Ocean Diver]

Useful life will be extended when you apply preventative and predictive maintenance bingo.

 

[Angelo Farina]

Completely disagree, you said it was because the first stage inlet face was scratched that caused the issue, that is exceedingly uncommon. A oring on the face of a yoke tank valve wearing out is exceeding common.

The reg was damaged due to improper design of the yoke (too thin) and of the rubber cover (a sphere, which did leave the external border of the cylindrical surface unprotected).
The manufacturer addressed these problems with a wider yoke and with a new rubber protector, embracing completely the cylindrical surface. Hence the O-ring failures were caused by a faulty design of the reg, which did leave a delicate surface too much exposed, so it was scratched. I think that this specific model was in production for a short time, So I was unlucky getting one. Subsequent versions did not suffer of this, so probably the number of failures was not great. But it was enough for triggering the modification...

 

[joewr]

Hi,

Do you have the bathtub curve for, specifically, regulators—either or both stages?

I’ve read that 2nd stages, with their thin diaphragms, are the most susceptible to failure, but again, I cannot find the specific data.

Whatever the case, it appears that quite few deaths occur annually due to regulator failure.

That said, I would like to thank EVERYONE for their posts and contributions. My personal opinion: you all sound like the kind of divers I’d be happy to dive with.

Joe

 

[tbone1004]

@joewr that's the curve that I was talking about that I don't believe has ever existed and I highly doubt ever will