2nd Chamber challenge

[Hoosier]

I am building the 2nd chamber (a.k.a. stack).

Here are all parts that I assemble:

32 inch AL chamber
Check valve
Priority valve (1800~2000 psi regulation)
Gauge
Vent (bleed) valve
Relief valve

What is the exact order of the parts?

Here is what I am planning… Is this a right order?

Compressor Triplex Chamber - 2nd Chamber (outlet) – Check valve – Gauge -Vent valve - Relief valve - Priority valve - Hose - Din Kit - Tank


The compressor capacity is 3.5 SCFM.


It doesn't seem to be an economical project instead of buying a pre-built one, but it will be fun.


Please share your expertise and insight.

Thanks in advance,

 

[Blue Abyss]

Why are thinking you need a relief valve on the filter assembly? Do you not have one on the compressor? If so that is plenty.
Does your compressor currently have a check valve in between the current filter and the compressor? Since you are planning to maintain pressure on the filter media through the use of a Priority valve you will not want the pressure to push back against the compressor when it is not running. You may want to install the check valve there.

I would think you would want to put the priority valve just down stream of the final filter assembly.
As far as vents go, what do you want to vent? If you want to vent the filter assembly it will need to be upstream of the priority valve. If you want to vent the filler hose it will need to be downstream of the priority valve. (The main thing to keep in mind is, make sure that you can depressurize all parts of the system when necessary)

As far as the placement of the gauge it is the same as the vent. What do you want to measure? If you want to see the pressure inside the filter stacks you will want it upstream of the priority valve. If you want to see the tank fill pressure it will need to be down stream of the priority valve.

 

[pescador775]

I hope you are patient as it will take 20 minutes just to charge up the condenser/filter set. Anyway, place the relief valve as close as practical to the compressor. Installation on or before canister #1 will be OK. The back pressure valve can be installed between the two canisters or on the output of canister #2. Explanation of the trade offs is too complex for purposes of this thread. All the other stuff, the check valve, the bleed valve, the gauge, etc, can be installed on the whip at the fill station(manifold). Basically, build a standard whip with addition of check valve at position closest to output of #2. Alternately, hard mount it to #2. However, mounting the backpressure valve at this output will eliminate the need for a check valve. Back flow in a valve of this type is severely limited. I wouldn't worry about back flow into the hose itself but if you are, then you know where to put the check valve--at the fill manifold upstream of the flow control valve and bleed valve. I don't know what a vent valve is unless you are talking about that thing Bauer uses on the PO. A drain must be installed in the base of the can #2 unless you want to drain 30 cubic feet of gas through the fill manifold. That would be prevented by the backpressure valve unless it is mounted between #1 and #2. If you reply, "what happens if I place it here instead of there", it will probably require 3 or 4 additional changes and a three paragraph reply. Such is life.

 

[rjack321]

putting the backpressure valve in between the canisters would be a mistake

 

[pescador775]

It depends. In this case, the relative size of the final filter to that of the compressor suggests that there is potentially large enough volume for the concept of dwell to actually mean something. So, if absolutely dry air is necessary then it would be better put the BP valve on the output of that cylinder.

I've done a few lab tests and desk calculations of dwell and adsorption rate of Linde 13X. The information was provided to SB some months ago. A hypothetical filter cartridge has about 15 ci of internal volume. Assume the weight of dessicant at 6 ozs. Calcs indicate that the weight of water which would be adsorbed due solely to dwell (2000 psi BP, 15 sec dwell) is on the order of 0.04 g / cf air. A backpressure valve maintains a set pressure in the filter set. If a backpressure valve is used upstream of the final filter it means that the final filter will be significantly pressurized only about 1/2 of the time (tanks are rarely dead empty when the fill is started). Thus, the advantage to the setup with BP valve at the final filter is about 50% of the number given above or about .02 g/ cf air. Thus, .02 X 80 cf = 1.6 g water vapor, not an insignificant amount. However, actual fills performed with the BP valve on the condenser (upstream of the final) indicate no visible water or rust in the tank. Air exiting a backpressure valve undergoes significant expansion and cooling. That is the reason for the better than expected performance; Linde 13X is sensitive to temperature and so the "dwell" advantage pretty much disappears. I should note that only fresh dessicant was used in my tests. Used dessicant has a significantly slower uptake. Also, temperature plays a role as mentioned. There are other complicating factors which are difficult to analyze but I believe this cursory analysis conveys some flavor to the issue.

putting the backpressure valve in between the canisters would be a mistake

 

[pescador775]

I forgot to mention that the hypothetical internal air volume of "15 ci" is unusual for a small setup like the Junior or Alkin portable. A typical filter cartridge has much less air space than generally assumed. Originally, I used the number as a "best case" mind game. I was attempting to design a more efficient filter and discovered there is no free lunch. Sorry, I'll try to undo any confusion. The only "dwell" that means anything to the subject of filtration efficiency applies solely to the volume of air inside the chemical cartridge. A 15 ci air volume indicates a small but identifiable advantage for purposes of water removal when using a 4 cfm compressor and is actually closer to what Hoosier is using (by coincidence). He can take full advantage by using a large volume of media, much more than 6 ozs. For example, if he uses 12 ozs, the adsortion advantage for the downstream valve would be twice that indicated above or 0.04 g/ cf. Dwell is related to air flow, and larger compressors can't extract as much dwell time as a small compressor. Thus, everything else the same, an 8 cfm compressor would have half the dwell time as my test compressor. In any case, the air volume inside a small filter cartridge like the one used in the Bauer PO is probably closer to 3 cubic inches. After all, the air space is packed with stuff leaving little room for air. Thus, "dwell" is mostly a myth. There are exceptions such as the unusual setup we have been addressing.

 

[pescador775]

Bauer indicates that the PO filter can process 3,200 cf of air before it is saturated. I wondered what this is based on. My tests of Vaporshell (zeolite) dessicant indicated that a small sample of this material will become 95% saturated after a period of 15 hours exposure to the atmosphere. Conditions were 85F, 60% RH. It so happens that 3,200 cf air volume corresponds to 15 hours run time of the Bauer Junior compressor. Coincidence? I don't know but it looks like Bauer had access to similar data and simply translated hours into cubic feet (liters, or whatever) and applied this number to different compressors regardless of fill rate.

To the compressor operator, it can mean several things. It is a cautionary note to these operators to avoid exposure of the filter to atmospheric air. Hand loading of media should be undertaken expeditiously. When the compressor is not in use, no air should be allowed to enter any of the filter ports. Leaving the drain open is deadly to the expensive filter. Many filters like the PO are constructed as a hybrid device with the condenser in the bottom. Leftover moisture in the condenser can raise the humidity in the upper part of the canister including the filter. Thus, a cartridge left in this type of device is dated. Even if the hours of operation are not met, the filter in a hybrid condensator will become degraded over time.

 

[oxyhacker]

Since it is easy to miss it while pondering all the heavy duty math and theory in this thread, it's worth highlighting the point Pesky makes below: too big a filter on a small system can be very aggravating, since it will take so long just to bring the system to pressure.

If you are going to be filling banks, it's no problem since the compressor will be doing a long runs whenever you fire it up. But if you are using it as a typical small home compressor, to fill 1-4 tanks at a time, and often just to top up a tank or two, it can drive you mad as well as wasting compressor life and electricity.

With a intermittant duty compressor like a Coltri MCH6 or Bauer Junior the situation is even worse, since you have to let the compressor rest every couple tanks.

I hope you are patient as it will take 20 minutes just to charge up the condenser/filter set.

 

[pescador775]

Surprisingly, the main obstacle to understanding dwell and filtration efficiency is the nature of zeolite itself. How does this stuff work? The answer to that could confirm my calcs or throw them into a cocked hat. Zeolite is an "adsorber" meaning it doesn't exactly soak up water, it "traps" the moisture. Thus, this stuff is generally referred to as a molecular sieve. One interesting side effect, every time a water molecule is trapped it gives up energy, a lot of energy. Never let your filter get wet, it will become hot enough to melt the cartridge.

Anyway, zeolite seems to function in different ways that I call "modes". It can act like a sponge or as a sieve. Differentiating the different modes is the focus of my difficulties in arriving at a good analysis. This dessicant has unusual ability to draw out water even if the RH is low. It appears that this ability is related to the total number of molecules encountered. The factors which influence this are humidity and air flow. Thus, whether acting in sieve mode or sponge mode, the 13X dessicant in an air compressor filter encounters the same number of molecules over time regardless of "dwell". Apparently, even without significant dwell, this dessicant is pretty efficient in either mode.

 

[captain]

About 15 year ago I worked as an operator at a Linde air sepertion plant. We used molecular seive to remove water vapor from the incoming air before the temperature dropped below the freezing point. If all the water wasn't removed the system would plug up with ice. We had two seive beds, one in service and one being regenerated which rotated on a timed basis. We were passing 17,000 cubic feet per minute through them and switched every four hours