Common mode failure

[MikeFerrara]

I do a lot of diving in cold water and while we see many free flows we don't have them. I have never had to call a dive because of a free flow. In theory, any reg can freeze but some don't seem to and others do.

Multiple failures...
Possible of course but I worry more about situations where there are multiple "problems" because one problem is causing the next than I do about multiple independant problems or equipment failures.

 

[ba_hiker]

i was diving with my sister and brother in law last month and they told me a story...

they are experienced divers from germany (both DMs) and dive often. once on a wreck dive in the winter in the baltic sea (cold), they had free flows that required sutting the valve on three of four 1st stages (they were both diving doubles).

soon after starting the dive both had free flow on their primary 2nd stages, switched to their backups and shut the valve on the primaries. they abandon the dive and started to ascend more or less normally. then my brother-in-law's backup started to free flow, and he had to share air with my sister using her backup backup.

they believe this was because the first stages froze up.
they often dive in very cold waters, through holes in ice, off boats in the baltic sea durring the winter, in german quarries and lakes durring the winter, stuff like that. they hadn't had a problem like this with their gear before, but apparently had called dives before, due to free flows.

this is not really to disparage that brand (or any other brand) but got me thinking (not always a good thing!). Is common mode failure an issue? for redundency reasons should we use different brands of regulators on doubles?
for instance use apeks for the primary and something else for the backup.

this would reduce the danger that a design or systematic error would cause total failure if the system was stressed.

btw my sister and brother-in-law sold all those regulators and bought new ones, with a 'membrane' that are said to be compleatly un-freeszable.

 

[John C. Ratliff]

It may be too late for them, but maybe they should have checked whether the air they had in their tanks was dehumidified or not. If there's significant water vapor, it could cause freeze-up from crystals that are generated internally. If so, then no regulator (even the almost bomb-proof for freeze-up double hose regs) would help the situation.

If there is a common mode failure, then you need to look at a common causative factor. It could be regulator design, but it could also be the air being used.

SeaRat

 

[Bob3]

said to be compleatly un-freeszable.

No such thing. When the water gets below freezing the breath can condense inside the 2nd stage & form a solid block of ice in there.
Both fresh & saltwater can get to 28°f (-2°C) if its kept moving, as in a fast current or heavy surge.
Folks diving in cold water just need to be sure they are using the proper gear in order to minimize chances of 1st stage freezes. Freeze-resistant 1st stages come in a few flavors; sealed, dry bleed, & thermodynamic heat sinks.
Proper cold weather procedure also needs to be followed. Regs need to be kept out of the mouth & the fingers off the purge and inflator buttons 'till underwater.
Double hose regs are still the most freezeproof though.

 

[donacheson]

It may be too late for them, but maybe they should have checked whether the air they had in their tanks was dehumidified or not. If there's significant water vapor, it could cause freeze-up from crystals that are generated internally. If so, then no regulator (even the almost bomb-proof for freeze-up double hose regs) would help the situation.

SeaRat

I doubt it would happen, Sea Rat. While the relative humidity (RH) of the air in the tank when full could be as much as 100%, when expanded it drops dramatically. As the pressure drops in the tank, the RH drops proportionately IF THERE IS NO LIQUID WATER IN THE TANK. For example, by the time the tank has dropped to 1000 psig in an Al80 initially filled to 3000 psig, the RH is 33% (1015/3015). This gas is further expanded to the intermediate pressure, say 150 psig, so the RH drops to (165/1015)x33% = 5%. That's why your mouth gets so dry.

Going one step further, air does cool when expanded adiabatically (without any heat flow to warm it) and it can cool enough to condense water vapor. That's why clouds form and precipitation occur. However, at the slow flow rates in regs through an orifice surrounded by metal and over metal parts, the expansion in the first stage is hardly adiabatic. However, there is still some cooling, but not enough IMHO to condense the water vapor in the tank air.

The alternative explanations offered - water on the first stage diaphragm or piston freezing or the diver's breath condensing and freezing in the second stage - are the likely culprits.

 

[pufferfish]

Mike I have often wondered as well why despite many high performance regs on the boats there still seems to be an awful lot of cold water free flows. This is despite experienced divers having diaphragm, dry bleed, and TIS designs tuned for cold water.

Recently I had a discussion on this topic with the scientist who did the reg testing at McMurdo Station in the Antarctica back in the early 90's. As you may recall the Sherwood Maximus had the lowest failure rate (close to 1%) as did the Poseidon Thor. Other brands in the same line did not fare so well. When I asked him what he saw as the major determinants of failures he suggested reg design and handling techniques. After that all the other stuff like cold water reg tuning, flow rates (depth and air consumption rates), ambient water temps, and of course tank moisture. The water temp down there was stable at -1.8 C.

As far as reg handling he was saying this made a big difference in failure rates. They often would not rinse the regs until the end of the day and only under pressure as the fresh water used would freeze at a higher temp. It only took a drop of water in either stage to cause problems. They found that one should leave the first stage cap off overnight as this would allow any water in the first stage to evaporate in the warm dry building air. As mentioned above no breathing off second until submerged.

The mean internal first stage temp was about -4.0 C but at the start of the dive this could get down to about -6.0 C. This temp seemed to be dependent on depth and air consumption.

While in the Antarctic environment dew point is usually not a factor they still checked this and other potential contaminants every two weeks after the first week (tested more frequently) with Drager tubes. The ambient intake air is so dry down there that moisture is just not an issue with reg failures. Up here I am not so sure and think that given all these high end regs supposedly designed for cold water failing often on the first dive, moisture must be part of the equation. I had to chuckle on TDS when one of the divers from Norway said that after numerous cold water reg failures in the ocean environment subsequent air analysis revealed Norwegian 'trimix' should be labelled as containing mainly air, oil, and water.

They are hoping to rerun the cold water reg testing program in the next year or two. Hopefully then we will have some better data on the latest crop of cold water regs so as to test manufacturer's claims against real cold water performance.

 

[pufferfish]

Going one step further, air does cool when expanded adiabatically (without any heat flow to warm it) and it can cool enough to condense water vapor. That's why clouds form and precipitation occur. However, at the slow flow rates in regs through an orifice surrounded by metal and over metal parts, the expansion in the first stage is hardly adiabatic. However, there is still some cooling, but not enough IMHO to condense the water vapor in the tank air.

Maybe someone else can do the math but I believe adiabatic expansion and cooling is a big factor in first stage temperature. That is why the temp drops with depth and air consumption as flow rates are higher. A drop from 3000 psi to 125 psi IP is hardly insignificant. Last summer at 30 C air temp I drained a full tank in 90 seconds and had ice on the first stage,..pretty good evidence.

Dewpoints or the temperature where moisture will condense and freeze if ambient temps are the same and below freezing is pressure dependent. Pressure dewpoints in the compressor at 3000 psi are always warmer (less negative) than atmospheric dewpoint for a given temp. One wonders if this would influence the point of freezing in a first stage. Where exactly does ice usually form lets say if a diaphragm reg does freeze? If it is on the high pressure side the actual pressure dewpoint there would be much higher (less negative) than what is reported for atmospheric dewpoint and make a cold water failure more likely as water vapor content increased. Anybody have any ideas on where the icing does happen in a diaphragm reg?

 

[ba_hiker]

well all the points about free flows are good but not really the point if was trying to inquire about. rather free flows were just an example.

I was more interested in whether people thought that using regulators of two different brands might reduce the chances of common mode failures and whether people were experiencing them.

equipment failures can occur as random failures, based on use, misuses manufacturing or material defects are mostly random. other manufacturing defects and design defects affect many or all items in the lot.

most of the current regulators are well designed and likely to work in common situations. but as you approach extremes, in temps, depths, breathing load, or other things or if you dive under unusual conditions; very salty water, silt and stuff in the water, it may be your regulators encounter unanticipated conditions and fail.

using regulators from two different makes (on doubles) would reduce the chance of common mode failure.

does any one worry about this?

 

[MikeFerrara]

using regulators from two different makes (on doubles) would reduce the chance of common mode failure.

does any one worry about this?

I don't think this is a true statement.

the common mode failures you're talking about just don't seem to be a big problem especially with free flows. If you get regs that are well suited to the environment you dive (I've definately seen some brands free flow more than others) and keep them in good shape problems will be extremely rare and double problems non-existant far all practical purposes.

 

[Norm]

Great post Mike and you have a lot of great replies as well. Nearly all of my dives are coldwater and in nearly 23 years of diving I have personally experienced only one freeflow and that was predive, (operator error - mine).
I too have seen many freeflows and I can attribute almost all of them to operator error. First stages do freeze, but this seems to be extremely rare. Second stage freeze up seems to be the norm, (for those that have them), and as you alluded to - things always seem to snowball. "One thing seldom goes wrong during a dive." Naturally, the equipment you choose does have a bearing, as does the gas you breathe and especially predive care and use of your gear. I have only seen one reg that froze consistantly, on every dive, at the same depth. A guy a *@$ reg that froze every time he hit 80'. That was weird. Good gear, good technique, and good predive habits seem to alleviate most problems. I have seen ice all over second stages, first stages, and even hoses, but the regs just kept right on chugging with no problems!
Norm