I suspect that as time goes well past skipping a recommend service interval, your risk of failure goes up. Then again, it seems like shortly after servicing, your failure risk goes up a good deal. Wonder where the sweet spot is?
Richard.
Typical failure rate distribution.
The typical statistical failure rate follows a curve that we call (in engineering) a bath tub distribution. Think about the shape of a bath tub. On the far left and the far right the raised walls represent high number of failures. In the center the flat bottom represent very low failure rate.
The large numbers of initial failures is called "infant mortality" and it occurs to new or recently serviced equipment/ machinery. There are a number of reasons for infant mortality on equipment, just a few examples: human error during assembly, new unproven parts, defective new parts, new parts that don't play well together (due to tolerance build up, etc), parts that need to break-in together, and a number of combinations of any of the above.
Once a piece of equipment is working properly and is past the infant mortality period (also called the break in period in some equipment) the failure rate normally drops dramatically (this is typical for most well design equipment/ machinery). This is the reason why, in engineering, we have the saying: "if it ain't broke, don't fix it".
Note: the actual break-in period (or period for potential "infant mortality") varies a lot depending on the type of equipment.
Then at some point the failure rate starts to rise again due to end-of-life types of failures. In the case of pneumatic equipment (like scuba regulators), this types of failures are caused by material degradation (like rubber O-rings and the soft seats). Also normal wear is another typical end-of-life type of failure.
Any of the rubber components, like the exhaust valve, diaphragm, hoses, have a life expectancy that can range from a few years to well over 40 years if stored carefully and properly maintained. But they will decay with exposure as a function of time, and can cause "end-of-life types of failure.
Typical end-of-life type of failure will take many years to occur on pneumatic equipment like regulators, unless there is accidental contamination (from salt water flooding, rust from a tank, etc) or other abuse. Just a few drops of salt water (from not drying the tank valve connection) can accumulate and cause enough corrosion to totally disrupt the normal life expectancy of the regulator.
Note: most end-of-life type of failure can be avoided by monitoring and inspecting the regulator (see: http://www.scubaboard.com/forums/regulators/346813-regulator-inspection-checklist-rev-7-a.html)
The risk of failure normally doesn't go up gradually as a function of time. It normally flattens with a very low rate until materials start to degrade (normally takes a long time). But accidents and poor normal maintenance (cleaning the salt water off) can ruin the failure rate distribution.
The failure rate described above, (what should be typical) can be totally ruined due to a number of accidents caused by human error, etc (like forgetting to use a dust cap and flooding the regulator) or from poor rinsing and fresh water soaking the regulator.
Notice that my discussion did not address the type of failure (minor free flow, versus catastrophic air shot-off, etc) and that in some parts I was trying to describe failure rates as it applies to most generic pneumatic equipment. Scuba regulators are just one type of pneumatic regulator.
Last edited: