1st Stage Freeze Up Prevention

[pufferfish]

Someone else mentioned that a while back and said -50 F below ambient achieved in the first stage air flow. Do you know if that was on the bench or actually on a real dive?
Yeah, that was me. I believe it was an article in U/W Magazine that stated that research in Panama City showed that at 3000 PSI, you have a 50 degree drop, (below ambient) in gas temperature. That is a given. This reinforces what someone above said about diving with a full cylinder in coldwater causing more temperature drop.
Norm

If that is the temperature drop in the air stream on the dry bench you can imagine what firemen with SCBA experience when the outdoor temp is -30 C. They require extremely dry air to prevent freeze failures as there is no 'warm' water around the reg for heat exchange.

As Chaos pointed out leaving your reg sitting out in the cold air before the dive only lowers the first stage temp before you get in the water such that the chance of any moisture in the tank air freezing is that much greater. Add in a high pressure tank, high air flows with depth, breathing, or BCD inflation and the risk of freeflow increases greatly.

It would be interesting if someone could repeat that experiment looking at temperature drops actually within the first stage during a real dive. It has been done with first stage wall temperatures under the ice but not actually in the chamber where the ice formation likely begins.

 

[WarmWaterDiver]

OK, I'm new to this board, and new to tech diving, and haven't done "cold" water diving, but I'm not new to gas laws or thermodynamics. So, I used some software to calculate temperature drops solely on pressure drop (not flow rate) in an adiabatic expansion of bone-dry air. I used 15 PSIG for 33 feet depth, 125 PSIG for second stage inlet pressure, and a gamut of first stage inlet pressures. Results were there is only 3 degrees Farenheit drop in temperature in the second stage from intermediate supply pressure of 125 PSIG for adiabatic expansion. Of course, the deeper one would go, the lower the difference in these two pressures (discounting the newer "overbalanced" diaphragm first stage designs), and so the lower the temperature drop from adiabatic expansion would occur.

However, there is certainly a significant drop in temperature in the first stage if this system is treated as an adiabatic system. At 1800 PSIG tank pressure (older steel tanks), to the 125 PSIG intermediate pressure, there is a 36.5 degree F temperature drop. For 2400 PSIG in the tank, there's a 46.5 degree F drop, for 3000 PSIG in the tank there's a 58.0 degree F drop, for 3500 PSIG high pressure steel tanks starting off there's a 60.3 degree F drop, and for 4350 PSIG tank pressure, there's a 67.3 degree F drop.

So, if the regulator first stage is viewed as an adiabatic system, one risks reaching 32 degrees F in a first stage with 90 degree F 3000 PSIG tanks on the first breath!

The conclusion is these systems are far from adiabatic, and in my opinion, are attempted to become as isothermal a system as possible. Has anyone had freeze-up in their first stage with 3000 PSIG tank supply pressure and tank / water temperature around 90 degrees F? How about 70 degrees F? I haven't seen a report of such posted.

From what I can see, the Scubapro TIS and AF system, and especially seeing its latest incarnation with extended heat transfer surface area (fins) and enlarged ports for easier ambient water flow, look to me to attempt to render the first stage operation as isothermal (rather than adiabatic) as possible. This appears empirically to have a mixed degree of success by the postings on this board versus the diaphragm systems with dry seals in low temperature waters. But even a dry sealed diaphragm system will depend heavily on heat transfer with the surroundings (i.e. not adiabatic operation) to not freeze up internally, and both systems depend on a low tank air moisture content as well. If a diver is moving, the approach to isothermal operation will be greater than if the diver is stationary too, as this will keep more flow of water across the first stage for heat exchange, lowering the temperature drop of the mass (or volume) of water the heat is being exchanged with at any one point in time. The same can be said for second stages like the Aqualung Glacia et. al. - they have extended heat transfer surface area in an attempt to achieve a better approach to isothermal (rather than adiabatic) operation.

I haven't even started to look at things like hydrate formation if CO2, CO, etc. is present (even in ppm quantities), although these are real-world effects well documented.

The discussion on first stage flow rates as contributing to this effect also leave me a bit puzzled, as if it's one diver using one second stage per first stage, the flow rate will be limited to whichever device has the lowest rating - and these all appear to be the second stage. This is 5 to 10 times less than the rated first stage flow rate. How then can the rated max capacity of the first stage be a root cause to the freeze-up? I just don't understand - I understand such freeze-ups do occur, just not this explanation of "why", unless the diver is also simultaneously laying hard on their BC or dry suit inflator etc.

So, I've done my bit of science here, and if the answers were easy on how to absolutely prevent regulator freeze-up, these threads wouldn't exist on the discussion of same. But one of the key items I see is a properly dried compressed gas supply no matter what brand / design of regulator is used. The second would be technique gained by training and experience in this type of diving, which would include redundancy in equipment and lots of practice drills.

Honestly, after reading through all this, saving my diving for more equatorial latitudes rather than local now that I live in the Great lakes area looks more appleaing to me all the time - or at least keeping my eyes peeled for a transfer back to the Gulf Coast . . .

 

[DA Aquamaster]

The whole heat transfer idea works very well in relatively warm water. With a 3000 psi tank, the seat and internal parts of the reg are no doubt getting quite cold but the heat transfer from the warm water (50 degrees plus) is adequate to keep things thawed in the ambient chamber where ice can be a problem. But the whole concept encounters significant problems as the "warm" water approaches freezing. The small temperature differential just does not seem to provide enough heat transfer to keep the reg "warm" under high adibatic cooling loads. In that case low pressure tanks really help and slightly extend the cold water range that can be tolerated by the reg.

Over time with the Mk 20/25, SP has increased the size of the holes in the ambient chamber to improve water flow, has removed the plastic and rubber trim pieces to improve heat transfer and most recently has added fins to the swivel cap to increase surface area and improve heat transfer. (The fins work well until ice forms on them, then they are pretty well insulated and become essentially useless.)

I really prefer SP's older approach of using environmental silicone and a rubber boot to keep the water out of the ambient chamber and have set up my early production Mk 20's this way.

Good cold water diving protocol is important (ie don't breathe and inflate at the same time, no long blasts on the infaltor, no breathing or inflating before the reg is competely under water, etc) but a minor mishap like the reg falling out of your mouth is a potential area where flow rate is important.

Lose a second stage on a low performance reg, and the resulting freeflow and cooling are comparatively small. On a high performance reg, the brief freeflow can be enough to freeze things up. Consequently, some cold water divers set the second stage to breathe a bit harder and where possible adjust the flow vanes to inhibit an inadvertant freeflow.

 

[pufferfish]

Warmwater, those numbers on the temperature drops with varying tank pressures really hit home with how important a risk factor tank pressure can be for freeze failures. It has been a long time since I have done physical chemistry so it is nice someone can do the math. It is quite apparent that minimizing one's tank pressure is a very important freeze failure risk factor moreso than say reducing intermediate pressure although all risk factors should be minimized. Prevention is the name of the game.

You mention flow rates and the second stage being the limiting factor here. On TDS recently there has been a discussion on the time to empty a tank in event of a low pressure hose failure or freeflow. What was very apparent other than the fact one has very little time to deal with a freeflow before the tank is empty is that with the new high performance regulators like a MK25/S600 or Posieden Jetstream with the second stage wide open, the flow rates are very high and the cooling effect would be much greater than with a low performance reg. With those new setups it only takes a slight increase in air consumption at depth and an overzealous blast on the inflator to lower the first stage core temp considerably and intiate the freeze cycle.

I think you are right in that with the addition of various technologies to increase heat exchange in the various regulators one is not dealing with a pure adiabatic situation. The dry bench test may show a temperature drop of -50F in the first stage, but with water flowing through the reg and the various heat exchange technologies adopted the final temperature in water at the freezing point may only be around 15 F with good cold water technique.

If one assumes a temperature say of 15 F in the first stage and allows this to be the dew/frostpoint at those various tank pressures you used the importance of dry air becomes apparent especially with the higher tank pressures.

Assume pressure dewpoint of 14 F (-10 C) and calculate dewpoint at 1 atm

Tank Pressure.......Dewpoint (C).....(F)
2400 psi................... -57............-71
3000.........................-59...........-74
4400.........................-62...........-80

I didn't do the math there but have a nice piece of software to calculate those for me

The message though is by using a lower tank pressure not only does one lower the amount of cooling in the first stage but also increases the amount of moisture the system will tolerate before a freeze failure occurs.

It also is apparent that for diving in water at the freezing point breathing gas with a dewpoint of at least -75 F or -60 C would be prudent.

 

[WarmWaterDiver]

What I was unclear on was it seemed some folks were attributing initiation of a freeflow event to the high flow capacities of some first stage regulators. I can certainly understand how, once a freeflow is initiated, the high flow capacity would sustain the event and cause rapid compressed gas supply loss.

For the second stage to initiate the freeflow event is less complex to me. The second stage, downstream of the demand valve, is a subcooled moist environment after the first exhaled breath in any waters less than 98.6 degrees F. So, in waters below 40 degrees F, the 3 to 4 degrees F cooling from adiabatic expansion in the second stage comes right to the regime for the freeze point of water. Such things as metal second stages, extended heat transfer surface area on the second stage inlet, etc. are all ways to try to reduce this impact and make the environment more of an isothermal than adiabatic experience. I think it's interesting designs like Sherwood's with moisture retention designed into their second stages still have a good cold water (if not deep water) performance reputation.

It's also interesting that CGA Grade E breathing air doesn't have a specified numerical moisture content spec, and even things like ANDI SafeAir have a spec of -50 degree F dew point (I'm assuming at 1 ATA). I've found the following sites to be of interest here. Chaos, at 2 mg/M3 moisture spec, your fill station must have some of the driest compressed breathing gas delivered on the planet, as according to this, NASA only specifies a -40 degree dew point (F or C, it doesn't matter at -40), the USN requires a - 63 degree F dew point, and 2 mg/M3 water content equates to nearly -95 degree F water dew point at 1 ATA.

http://www.airanalysis.com/Technical/oxygen_compatible_air.htm
http://www.lawrence-factor.com/images/lf0011.pdf
http://www-nehc.med.navy.mil/ih/respirator/breathingAirQuality.htm

 

[quimby]

With all the information bantered on especially this thread I have to sneak back on and say that essentially keeping the dew point down is to me the actual key. I have said before that we never saw the kind of problems talked about here but I know my air was virtually dry and most of the (if not all) was diaphragm type regulators and some double hose again dia but dry chamber.
The other point beat on is the external icing within the "end cap". If this is a problem on dia type regs, some phenolic or non metal separator replacing the housing nut, as a cold path break, would or should break any icing in the external chamber.
Piston type is a little more complex as the stem is being super cooled and the piston itself is as stated by warmwaterdiver dropping upto 65 degrees.
(me in fl and warmwaterdiver talking about icing oh well) Those temps are now conducive to icing in the ambient chamber and to my simple thinking would just not use piston regs for potential icing situations.

 

[WarmWaterDiver]

My take is what each tech or cold water diver defines as "properly dried compressed gas breathing supply" is the most critical item - as you say in your first sentence. Chaos, on page 2 of this thread, gave us his empirical results using piston regulators (with 2 mg/M3 spec on his gas supply) and I know a number of Canadian divers have posted similar experiences on this board, asking others what the moisture content of their air fills were when they had free flows (see Pufferfish on page 2 and above for an example). The thermodynamics involved are independant of regulator design - delta pressure is delta pressure, piston, diapragm, air conditioner, refrigerator / freezer, propane barbeque grill regulator, you name it. The point of pressure change is exposed to the same effects.

I used to love 20/20 on ABC when John Stossel had his "Junk Science" feature each week - so I guess some of that has crossed over into this hobby of mine. I expected technical divers to be more, well, technically broadened and cognizant than the average recreational diver (and looks like that's largely true). And I am new to "tech" diving having done our ANDI TSD 3 certification dives on Grand Cayman over Thanksgiving week in 2003 (hence my chosen "handle" here).

This endeth my techno-geek-speak. I'll sit back and have more fun here going forward.

 

[UWSojourner]

DA Aquamaster -

Ok. Until now I thought it was GOOD practice to make sure your reg was actually working.

How do you ensure a working reg before decent if you can't breath off it?

I dive an Apeks ATX200 (if that matters).

I would not use the reg at all for any purpose until you and the reg are completely underwater. So orally inflate if needed and then do not breathe of the reg at all until after you are completely under water.

New cold water divers often have a bad habit of breathing off the reg 2-3 breaths before decending and then very quickly experience a freeze up. The colder the air temp, the bigger the problem will be with doing this. If the outside air temp is 50 degrees you may get away with it with an unprotected reg. If the air temp is colder than this you probably won't. If the air temp is below freezing already, you definitely won't.

 

[WarmWaterDiver]

Ahhh . . . perhaps now the variable definition of "properly dried breathing gas supply" would come into the equation . . . how else could rescue workers in SCBA be effective in -30 degree C air environments as alluded to elsewhere in this thread?

Also, I found the link showing the US Navy tests used breathing gas supply with a moisture content of 23 ppm which is a dew point of -65.5 degrees F at 1 ATA while immersed in 28 degree F simulated polar seawater conditions, at least for their Scubapro MK10 & MK 20 testing.

I wonder if that breathing gas moisture spec is the best the USN can do, or whether it's just the "mil-spec" for USN compressed breathing gas supply for scuba. Anyone out there that knows the answer offhand?

 

[pufferfish]

CRDiver, This site your link takes one to is GREAT!!! All the techno-babble and hard science experiment results one could think to ask for. The tests on the Scubapro MK10 & MK20 details the dew point of the breathing gas mixture used was -65.5 degrees F, or 23 ppm. From there, applying a little thermodynamics, the gas supply used can be seen as an element of the test regimen not seen as a variable, but it certainly is a factor (understandable when one is using a standard set of test criteria for more than one piece of equipment, whether of the same make / model or different makes / models). I wonder if that's the best the USN can do in breathing gas supply dehydration, or whether that's simply what someone wrote the breathing gas mil-spec to be . . . It certainly dovetails with the results showing better performance against freeze-up at 500 PSig supply pressure versus 1500 PSIG.

Now, if the same commercial scuba regulator makes / models were then tested at variable breathing gas supply moisture contents as well as different supply pressures to the first stages (for all brands / makes / models), I would think that would be of interest - but would probably need to be performed by a non-governmental agency that would have an interest in such testing.

Here's an article that ought to interest anyone asking about whether their first stages should be rendered O2 clean as well as O2 compatible for breathing mixtures enriched in O2, or those who switch the same first stage back and forth between Grade E air and enriched O2 mixtures to ponder on...

http://scilib.ucsd.edu/sio/nsf/diving/nedu.html

Warmwater I hope you don't mind but I moved this from the Apek reg discussion over here as I didn't want to highjack that thread with this geeky moisture and dewpoint stuff

You hit the nail of the head here where the testers did not consider the gas supply as a potential variable in their experiment. In other words depending on regulator design and the testing protocol some will freeze up as the 'dry' air at -65 F was not dry enough. The pressure dewpoint of that air at 1500 psi in the first stage is a balmy 13 F (-11 C) so with an ambient water temp of 0 C or 32 F and the flows achieved at the depths they tested, the inner first stage chamber temps might easily have reached 13 F and so freezing would be initiated. I have a copy of a study done in the Antarctic where they placed temperature probes deep into the wall of first stage (not in air chamber) and recorded temps or -6 C and these were research divers with good cold water dive technique. It is obvious then that within the airstream -11 C could be easily achieved so that 'dry' air at -65 F atmospheric dewpoint would likely freeze. Even more likely if they had used tanks at 3000 psi.

If you search on that site under regulators you will see a Canadian DCIEM test where they made the same assumption that air at - 85F was dry enough for their protocol. They then proceeded to drain a tank from 3000 psi to 1500 psi with the purge button 5 seconds on/5 seconds off. The pressure dewpoint of that air at 3000 psi is only -19 C or -2 F again a temperature achievable with the testing protocol and ambient water at 0 C. As 75% of the regs froze up in both the wet and dry air they assumed it was 'material performance' rather than maybe the air itself was responsible for the freeze failures given their extreme test protocol.

As mentioned earlier on this thread if we simply take those 'hard data' Antarctica temps recorded in the wall of the first stage at -6 C and lower this 5 C (10 F) for a margin of safety at a pressure dewpoint of 3000 psi, the dewpoint of air required to avoid freeze up at 1 atm is -60 C or -75 F! This would be a dewpoint where one might feel comfortable jumping under the ice with any reg. This is still a minimum standard as I see it.

I have been unable to find a reference to the required US Navy air requirements for moisture. I suspect they just chose the SCBA spec for Grade E which is - 65 F. As far as producing air drier than this it is very easy with frequent molecular sieve changes. At many dive shops after a recent filter change dewpoints around -80 or -90 F are not uncommon. Our shop had -100 F a few hours after a filter change. Have a look here
http://www.fillexpress.com/library/pedigree.shtml#h2o
but the problem occurs three months down the road when the filters have not been changed and no moisture test is available and you decide to go diving in water under the ice at - 2 C with air that has an unknown dewpoint.

Anyhow here is a nice review on some of this material for those interested.
http://www.airanalysis.com/Technical/moisture in SCUBA & SCBA.htm

It is quite clear though that the Scubapro regs seem to be much more sensitive to moisture in the air than say the dry sealed Apeks. Freeze up prevention is the key so if one can't be sure of moisture levels in the air they are purchasing I'd say get a dry sealed first stage with reports from the field saying the reg has good cold water performance characteristics. The Apeks, Poseidons, Atomics, and certain Sherwoods seem to fit this bill. Otherwise you will have to carry around some Drager tubes to check those dewpoints before jumping in under the ice.