Desiccator Experiment Results…and what happened to my compressor

[AlanMcD]

Those tanks from that compressor look pretty good - lots of life left in them prior to being cut up.  People always think any rust in these tanks is horrible, but it is inevitable, and as I indicated in an earlier post.  The usage / duty cycle pretty much guarantees that there will always be some amount of liquid water in these after the first use.

Very interesting to see, so thanks for that.  I understand your being concerned with the age of the compressor, and not knowing how it was used before you had it, so I understand why you decided to abandon it and get a new one - and then you might as well cut it open and take a look to see what you can learn.

[Mole2017]

Those tanks from that compressor look pretty good - lots of life left in them prior to being cut up.  People always think any rust in these tanks is horrible, but it is inevitable, and as I indicated in an earlier post.  The usage / duty cycle pretty much guarantees that there will always be some amount of liquid water in these after the first use.

Very interesting to see, so thanks for that.  I understand your being concerned with the age of the compressor, and not knowing how it was used before you had it, so I understand why you decided to abandon it and get a new one - and then you might as well cut it open and take a look to see what you can learn.

Well, having the motor burn up was one foot in the grave for this setup. After I got into them I did think I probably could have kept them around as a portable air source, but then again, I have plenty of other tanks that could do the same job and hold more air than this pair could.

[anti-squirrel]

Fun read, this thread.  For a guy that dealt with HP Air his entire naval career on Submarines and fixing submarines, I got a few observations:

1: I prefer 2-stage compressors that use oil.  Oil-free makes me sad.  My air compressor is nearly 35 years old and the receiver is older than me
2: blow-downs of the receiver were daily PMs when I worked construction.  Still do them whenever I use the compressor
3: compressors were an item I fixed literally close to a thousand times while a service tech for Sears after my Navy experience.  I prefer oil versus oil-free compressors for a reason
4: dedicated HP compressors deserve dedicated drying; my 2 HP handpumps do not have any dessicant and do not need them since they are not pressurizing fast- internal inspection of air tubes proves this, plus 8 years of Navy experience.  Slow-fill means dry air.  HP air moving fast has to be super-duper-dry on subs for a reason.  Google "Thresher SSN-593" for more info
5: my new HP Compressor WILL get an "Improved by Peter" dessicant system; the "factory" moisture trap is effective but why gamble?

Dry air is wonderful.  It makes submariners get back to port and ensures pellets keep flying.  They're the only reasons I need

[Mole2017]

As I get ready to put the Makita compressor into service, I tried my hand at removing the drain plug to get access to the interior of the tank. Fortunately, this unit is so new the threadlock they used was still sticky and it came out once I was able to get something on it to turn it.

I haven't settled on a packaging scheme for the new desiccant, but here is a composite picture of the interior of this tank:



The interior is lit with a "cob" LED I hacked out of an battery powered LED "lighswitch" at Walmart. Smaller wires would be nice--they are contributing to the blur at the bottom. By "composite," I mean this shot is actually nine overlapping shots stitched together by software, which did a pretty good job, I think. I held my phone camera up to the hole and snapped a shot, adjusted the angle for another and kept going.

I am looking through the condensate drain. That is the tank outlet at the lower right of the picture, which is the same size fitting as the condensate drain (nom. 1/2"). In use that is above the mid-line of the tank. The brightly lit area is just about the "top" of the tank when this is set on its feet properly.

You can't help but notice that there are spots of rust already on the wall and near that outlet. I bought this unit, ran it once to hear what it sounded like, and carried downstairs. The next day, I experimented with setting the regulator and left it with no hose connected to it. By the next day the tank was empty.

Now in practice, I will "shut" the regulator too when I leave the compressor after filling the gun, but I was testing this one--I guess this regulator leaks too. The Campbell Hausfeld regulator leaked so bad sometimes I think I lost as much air or more than I managed to pump into the R10 each session. I even took it apart once to see if I could do something about it.

Anyway, a few days later, I opened the tank and with a flashlight examined the inside and noted small wet spots. They were still just wet then and so small they were nearly flat--like when you wipe something with a towel that doesn't quite get it dry. But here we are a few days later and they have all turned to rust. As you can see, there was not a lot of them from that one tank full.

But that gets me wondering: will the drops form--and rust--faster than the desiccant can keep up with them? I guess we'll have to check on the tank in after a few more fill and empty cycles in the course of filling the R10. But before I can do that, I have to come up with an easier way to "contain" the silica gel for easier recovery. A 7/16" hole isn't much to work with 

[desertplinker]

would anyone be able to comment on the climate you live in, as in humidity in your geographic area? I live in a part of Arizona where humidity is like 10% on average. I always thought that would mean my compressor would be dry? Is this true?

[aPpYe]

For what it's worth, I have restored desiccant beads many times by putting them in my oven for half an hour or so at its lowest setting of 175 ...

Also,  adsorbing moisture out of the air at one pressure does not mean that there isn't still moisture that will be wrung out into your gun when you hit the higher pressures.

[Chris USA]

I have a California Air compressor feeding a Shoebox 10. I have a Campbell Hausfield water separator between the compressor and Shoebox that I put desiccant beads in. Maybe 3/4-1 cup. I replaced them awhile back when they went from orange to green. I have 1 1/2hrs of run time on the beads now and they are still orange. Not sure how long they went last time.

On humidity,.. in Ohio,... 60-80% is not uncommon in the Summer. Indoors with AC,... 40%-ish is about the best. I fill inside with AC on (or heat if winter). I drain my low pressure compressor often. I have yet to ever get liquid moisture from the moisture filter.

As I recall,... I had a choice of 5 micron or 40 micron filter on the moisture separator. I chose the 5.

I have been inside the air tube on the Maximus and M-rod and both were clean as can be.

That's all I got. 

[AlanMcD]

David,

While I love the thoroughness of of your approach to this, you are really not going to get a head of the condensation - especially if you ever use this compressor for more than feeding your hand pump.  I would say just use them as designed and take good care of them.  Mine is a 15 gallon Craftsman oil-less compressor that gets a good amount of use, and it is 25 years old, rock solid, and still going strong.  I have had to replace a few drain plugs over the years due to use, but that has been it. Take care of them as intended (aka drain them proportionate to their use) and they last a good long time (assuming it is a good unit to start with).  I bet mine has had some amount of liquid water in it for all of those 25 years - it is unavoidable.

Steve,

Relative humidity is just that - relative.  A given mass of air that is pressurized air can hold less water vapor than the same mass of air at atmospheric pressure.  Even at the humidity level you are at, there is enough water vapor in the air that, once pressurized, some water will likely condense out into liquid.  You will just have less of it - so you will not need to drain your tanks near as often.  But the inside will still get wet over time.

[Doug Wall]

The only way that you will ever prevent water from condensing on the inside of the tank is by keeping the water from getting into the tank in the first place. In a compressor tank, the water will always condense on the tank surface. If you put some kind of desiccant cartridge in the tank, the water isn't attracted to it like a magnet. The water vapor diffuses through the air, and if a molecule happens to bump into the desiccant, it will absorb.

A large desiccant cartridge on the INPUT of the compressor will keep water (or most of it) from getting into the compressor.  By large, I'm talking about something about 3" diameter. and 2 ft long. By having the air flow through a desiccant bed for a longish period of time, the water molecules have a better chance of absorbing in the desiccant.

[AlanMcD]

Doug is right.  And I will add that, in my opinion, it simply is not worth drying air before feeding it into a compressor - they are made with the expectation that liquid water will condense out in them, and if properly cared for the "compressor" will wear out before the "tank" will rust through (and even if it does, in over 99% of the cases it will be in a manner that leads to a leak not an explosion).

I would dry the air coming out of the shop compressor that you feed into your hand pump - the shop compressor will have done 90% of the work for you, by condensing out a lot of the water vapor (provided you let the tank cool before using the air). A smaller "normal" desiccant filter will do the rest (unlike the massive one you need to feed a shop compressor)

This is what most of us Shoebox users do.  Here is a post I put up years ago with the data on absolute humidity of different stages of compression (written form a Shoebox user perspective, but still applicable to using a shop compressor to feed a hand pump).  There is an attached PDF with good data in it.  https://www.gatewaytoairguns.org/GTA/index.php?topic=139845.msg1407834#msg1407834

[Mole2017]

Thanks for the input, Doug and Alan! It has been lurking my mind that I need to do some calculations about how much water my low use situation is dealing with. For example, how much air do I need to compress before I have a significant amount to drain? Is the first "tankful" of air going to just wet the inside surface, but not run down the sides and pool?

So, looking at Alan's attachment in his post he linked, let's see if I'm starting this correctly. First step, figure out how much air goes into the tank. In round numbers, suppose I fill to a gauge pressure of about 135 PSI or around 9 atmospheres. A 1st order approximation to relate the volume of air at tank pressure versus ambient pressure is to use

   V₁*P₁=V₂*P₂, or  V₁/V₂=P₂/P₁

That is, the ratio of the "free" air volume to the compressed volume is the same as the that of the two absolute pressures. In this case, P₁ is just 1 atm and P₂ is 10 atm (9 atm gauge plus that 1 atm above vacuum for 0 gauge pressure). Thus, my 2 gallons of 135 PSI air started out as 20 gallons of uncompressed air, or just about 0.075 cubic meters.

From Alan's table, then, hypothetical "worst case" saturated ambient air at 68 degrees F has 17.5 grams of water per cubic meter. My tank of 2 gallons of air at 135 PSI then could have as much as 1.3 grams of water.

Hmm, that was easy. That also isn't much. Even if I could get all of it to condense, we can all see from the pictures of the insides of the tanks, I'd be lucky to get a couple drops to blow out with the air. If I were using it on a big project and it cycled many times, I'd start getting enough water collecting to drain it at some point.

So in my experiment, I loaded dry silica gel into my compressor tank and used it infrequently over a year to fill my air rifle. That is, I ran it about every second or third time I needed to fill my air rifle. I actually don't shoot a lot, so I guessed at maybe having run the old compressor a dozen times over the year. That's about 16 grams total or just about 1 tablespoon of water (at the hypothetical ambient conditions).

The "used" silica was recovered, dried, and found to have been holding 23 grams of water at the time I dried it. However, even some of that 23 grams probably was adsorbed between the time I got it out of the compressor, sifted out the rust and started drying it. My notes from monitoring the "dry" silica exposed to room air afterwards indicate that it may have been able to adsorb as much as 2 to 5 grams in an hour from the air, so I'd guess it came out of the compressor with something less than 20 grams. Between what water may have still been in the tank and what got in there over the year, that seems about right. Neat.

Assuming I did the math correctly, it worked out ok for me. But if I was running that compressor every week, my silica gel would get swamped.

Picking up on that 0.075 cubic meters. My 200 cc buddy bottle at 220 atm is holding 0.044 cubic meters of air. If I've figured that correctly, that works out to two "less then full" buddy bottles for the R10 from a 2 gallon tank of air at 135 PSI. But, I'm not starting from empty each time either...filling 200 cc from 100 atm to 220 atm is just 0.025 cubic meters of air at 1 atm. So, if I haven't botched the numbers, that is about 3 refills from full tank of compressed air.

[Chris USA]

Years ago I worked at a radiator shop that did re-core ) replace unit, re-use tanks. They also did gas tank repairs that had small leaks. The gas would be drained. The interior was steam cleaned. Dried thoroughly.

Then,.... a plastic type liquid/epoxy? was poured in and the tank rolled around until the inside was coated. Any access was drained, let dry and that was it. As far as I know,.... it worked quite well. It was a red liquid and about the consistency of syrup.

If I were to line a new tank,.. or even an old one,... this would be the approach I would take.

[Mole2017]

Funny that you should mention that. I had lining the tank on my mind too this morning.

First difficulty/inconvenience of lining this tank is that is has fittings on it already that use some really tough sealant/adhesive. The stuff is still fresh, so it may be now or never if I were to disconnect anything, but then again let's think about what has to be disconnected...Here's a view of the connections:



From front to back:

• Inlet. You can see the red sealant on the threads into the tank. I don't know if they did any to that compression fitting. (That sure is a tempting target. That tube has a compression fitting on both ends. If one were to insert an aftercooler and moisture trap, this tube is where it would happen, just not under the compressor like this one does...) On the backside of this is a  1/4" poly line to the on/off/safety switch.
• Outlet. This just goes up to the regulator and gauge manifold.
• On/Off/Safety switch. I think this is just a diaphragm switch that cuts of at the preset fill pressure, or, that failing, blows the bleed line coming from the Inlet.

If I were going to pour and swirl any coating in this tank, that third connection is the one I'd worry about fowling. But, maybe with care it could be done--just don't let it all slop into that port and don't get too careless with the other two for that matter. Or just man up and see about taking that part off.

That leaves what to use for a coating that I can get reasonably easily. Ideas:

• Epoxy mix or paint
• other paints--no solvents 
• wax
• other?

Wax was what was on my mind this morning. It isn't going to cure before you finish. It wets easily once a surface is heated. Certainly the slowest application process too. But maybe someone has a suggestion? For the life expectancy of the system (oil-less compressor) maybe not worth the trouble either, but it seems too tempting not to consider it.

[AlanMcD]

David,

As one that also lived much of my life by the motto "if you think it ain't broke, you haven't looked hard enough," I commend your desire to make this compressor better.  But I have to ask:  just how many decades do you want this thing to last?

You bought a fine little compressor there that without doing anything other than draining the water out every several months (based on your stated duty cycle), probably will last you well over 20 years - and when it does finally fail, it almost certainly won't be due to the tank rusting out.

It does not need desiccant in it, and it does not need the insides sealed up.  The better use of that desiccant is on an output filter into your hand pump (or whatever you decide to feed with it).  And if you are worried about rust as a contaminant in your air feeding your pump, I would add a fine filter after your desiccant filter.  Personally use a 0.1 micron filter after my desiccant filter in the feed system into my Shoebox - and the desiccant filter is after probably 70 feet of plumbing and multiple water traps along the way, so it sees pressurized air at below 70 degrees F in my basement, so I have condensed out as much as possible before it does its job (it is large too, so I get over 150 hours of run time on my Sheobox before I have to recharge it).

All the while that 25 year old shop compressor does it's thing with condensed water (and probably some rust) sitting in the tank.  Mostly due to this thread, I did go ahead and drain it this morning - probably had a few cc's in it from use over the winter and this spring - and tapped around the tank with a wrench listening for changes in tone and heard none.  So it goes back to its duty, most likely until the compressor pump fails at some point and I'll get a new one.

Feel free to keep working to make yours better though - as I said, I get it.  I did lots of things like this when I was younger, and while I enjoyed doing them at the time, I never once got to see the anticipated functional benefit or economic payoff for doing it.  Just one last thing to consider: if you do this, and the compressor pump fails for some reason in 6 to 9 months, there is no way that Makita will honor any warranty on it . . .

Have fun!

[Mole2017]

Well, it has been a while. Since I was already into things this far, I finished the job and got some wax in there. As Alan points out above, it really isn't needed. And I could put the desiccant in an inline unit instead. However, not having an inline unit available and not wanting to shop for that and its fittings, I've continued with the plan which has shown itself to be successful.

Again, don't forget that this is a compressor that isn't going to see much use. I am certain many of you run compressors bigger than this through more fill cycles in a day than I will in a year (easily less than a dozen, I estimated). That isn't much moisture in the first place--not even enough to drain. But the water vapor in the tank would nevertheless condense to the sides of the tank and rust given an occasion. And yes the motor or something else will fail before the tank does, but, well, it seemed like something to mess with. Like they say, so know you know...

Here's a view of the now blue tinted interior of my compressor tank:



Those few spots of rust are from the moisture of a single filling to test the thing after I bought it. The big orange spot at the bottom is a blurry yellow wire that goes to the LED illuminating the inside of the tank. I can almost see both end welds on this tank from the drain port. This tank is about 7 or 8 inches in diameter and the two ports you see at each edge of the photo are about 12" or 14" apart in real life. This is a composite picture made from about a dozen shots through the port. I could have used a few more in the upper corners...

Anyway, as to how it was done, should you ever try to do the same or similar...
I found some dark blue wax on a pair of candles. Now, most candles are not solid colored all the way through. These had a just a layer of blue wax on the outside; the rest was just white wax. Fortunately, that last layer was from a single dipping on a cooler candle and didn't really fuse well. It was easily chipped off. Melting it in a tuna can confirmed it flowed well at around 170 degrees and froze up near about 140, as much as wax does such a thing. I had maybe a couple tablespoons; 1/8" or 3/16" deep in the can.

So how to heat a tank to get the wax to roll around nicely and wet everything? That I thought a while about and tried a few ideas. In the end, the solution was simple: put it in the back of the car on a sunny day. 150 degrees, piece of cake, and higher still if the sun is shining on it too. Even sitting in the sun in the driveway was enough to get it to 135 degrees. Then a quick play with the heat gun easily pushed it past 170 and I just tipped it and sloshed it, heating areas as needed to keep the wax moving and making sure my last steps were to coat the side with all the ports. By then the layer of moving wax was thin enough to not go over the inside lip of the ports.

So, let's see; over a month to pull this off? I haven't been very efficient, for sure. I've been tied up with other stuff and haven't been able to shoot much. Which is good in a way--the R10 is running on fumes now and there are finally some pests to take care of (it has been a quiet summer).  Furthermore, notwithstanding that the hottest part of the summer is nearly upon us, I've been getting an itch to go after the woodchucks again and needed to get things back together.

Eventually, I will get to the other part of this experiment: the desiccant pack that is inside the R10 buddy bottle. That is there to catch the little bit of moisture that gets through my scheme as it is now. Any day now, I'll have to take that bottle off to get the action out of the stock to get my hammer tension back in spec, and that might be a good time to have a look inside of things.

And a final note to DIY folks: don't disassemble that combination on/off pressure limit switch screwed to the side of the tank. Just disconnect the wires, strain reliefs included, and unscrew that thing if you are worried about it. (The thread sealant on this Makita is tough stuff and I had wanted to avoid messing with that, but had to do it anyway in the end in order to put it back together.) I had only wanted to get it out of the way to make sure I didn't accidentally get wax into it. It wasn't easy to get back together...You think that was foolhardy, wait to you see my next post...