[mrennie]

Hello:

I have read many posts asking about how to run air lines in a shop, what kind of pipe to use, how to keep air dry, etc., so I thought I would pass along some info on what I did for my shop.

I have a 7.5 HP, 23 CFM compressor, and since I planned on sandblasting and painting I wanted to make sure the air was as dry as possible. Since I am cheap, I didn't want to spring for a refriderated dryer or anything that had an ongoing cost to maintain (the electricity to run the compressor is painful enough!!)

After some online searching I came across several diagrams and suggestions from various compressor and filter manufacturers that had the same theme:

-install an aftercooler after the pump

-build a system that will require the air to travel a minimum of 25' before it reaches a water filter

-slope the main distribution line away from the storage tank

-install water traps as far form the tank as possible (or alternately, install water traps as close to the tool as possible)

-tap off the main line by going upwards and then down, to minimize carryover of water into the regulator/filter

-provide a drain at the low point of the main line

-provide a drip leg with drain at each regulator.

To incorporate all these features takes some planning, and the material is expensive, but the remeber that the quality of the air will be directly related to how much water you can extract before it reaches the tool. After using my system for a year, I have seen no evidence of water while using my tools, and the water traps on my filter/regulators have never had water in them.

The majority of the water is condensed at the tank, and what doesn't fall out there is trapped in the water trap I built for the output of the tank. I will attach pics of each with some explanations.

[mrennie]

Here is a diagram showing the layout for an industrial air system. Note that they show a large aftercooler and water trap between the compressor and the air receiver (tank). In a typical home or shop compressor, the pump is connected directly to the tank, and there is not much aftercooling happening other than the heat radiating off the copper pipe between the pump and tank. Some bigger cast iron compressors have an aftercooler, but a dedicated aftercooler of some type will more effectively lower the air temperature coming out of the pump, which will start the water condensing process.

The earlier the air is cooled, the sooner the water will condense, and the easier it will be to trap it before it reaches the tool.

[mrennie]

Here is a pic of the copper aftercooler I made to replace the factory pipe that connected the pump to the air tank on my compressor. This is a 25' piece of 3/4" ID (7/8" OD) type ACR copper tubing, which was chosen because it was slightly larger than the original pipe and because it could be formed into a loop.

This was by far the most difficult part of the system install, as the pipe is very hard to bend without kinking. I used a sleeve that slid over top to be able to bend it, and I wrapped it around a small steel drum to get the diameter I wanted. Note it is held to the wall with several plastic U-clamps and is free to moev with the compressor.

The temperature of the pipe at the pump discharge gets quite hot when the compressor runs for any length of time, but the other end where it enters the tank is much, much cooler, barely getting warm. This difference in temperature causes most of the water vapor in the compressed air to condense, and since the receiver is so large the air velocity is much slower and the water drops out easily. This does require very frequent draining of the tank to ensure it does not promote rust.

[mrennie]

My solution for regular draining was to add an auto drain valve to the tank drain, which is plumbed into the unloader valve. This setup will open the drain for about 1/2 second everytime the compressor comes on, and again for about 6-10 seconds when the pump shuts off. You would not believe the amount of water that drains at the end of every pump cycle. This is water that otherwise would have travelled thorugh the air lines and MAY have been trapped at the filter.

As mentioned, the earlier you can condense the water the earlier you can trap it.

(I will add a pic of the auto drain later)

[mrennie]

With the majority of the water condensing in the air tank, the filters at the end of the air line should have an easier job of getting the rest.

However, I decided to build a secondary water trap that is connected to the output of the air tank, and this trap will collect almost all of the water that has not dropped out in the air tank.

It is made with 2" Schedule 40 black pipe, and is connected to the air tank with a piece of premade 3/4" ID hydraulic hose (it was easy to find and decently priced). The hose connects to a Tee fitting that has a 12" drip leg below it, and a 5' piece of black pipe above it. The air must rise from the inlet of the Tee to the outlet 5' higher, where it ties into the main distribution line near the ceiling of the garage. Gravity is our friend here, and the large diameter pipe slows the air down allowing the heavier droplets of water that are remaining to drop out and collect inthe 12" drip leg at the bottom.

To ensure regular draining, this water trap also has the same type of auto drain that I installed on the air tank, and is drained at the start and stop of every pump cycle.

[mrennie]

Here is a diagram of the recommended layout of distribution piping from Sharpe.

Proper design suggests that the distribution line slope away from the highest point, so that any water droplets still being carried along in the air naturally want to flow down hill to the drip legs at the ends of the horizontal runs.

As well, due to the distance we have travelled from the pump to the distribution line, the air will have cooled considerably by now and any water that is still a vapor will likely condense so it can be filtered at the water trap at the regulator.

[mrennie]

Here is a shot of the top of the secondary water trap as well as the main distribution line that runs along the 2 walls of my garage.

Note that I used 3/4" type M copper pipe. 1/2" might have been adequate since the distribution lines are about 20' long from the outlet of the water trap, but since I was sizing the system for sandblasting I used 3/4" right to the regulator.

There have been many opinions expressed about what type of copper pipe is required (Type L vs M, etc), but the "Copper Tube Handbook" published by the Copper Development Association in the U.S. reveals that the weakest link in a copper piping distribution system when calculating the safe working pressure is actually the type of solder used.

http://copper.org/publications/pub_l...e_handbook.pdf

-3/4" type M drawn (hard) copper tubing (the thinnest and cheapest) is rated for 701 PSI at 200F. This is obviously adequate for our shop air system.

-1/2" type M drawn (hard) copper tubing is rated for 850 PSI at 200F, and is also adequate.

-the safe working pressure for annealed (soft or bendable tubing) is roughly half the above numbers (about 330 PSI for 3/4" type M), which is still more than adequate for a compressed air system up to 175PSI.

NOTE THE SAFE WORKING PRESSURES DEPENDING ON WHAT TYPE OF SOLDER IS USED:

-soldering the joints with 50/50 tin/lead solder reults is a safe working pressure of only 100 PSI at 200F. THIS IS NOT ADEQUATE AND WOULD BE DANGEROUS

-soldering the joints with 95/5 tin/antimoney solder results in a safe working pressure of 505 PSI at 200F. This is the solder I used for my system and is sometimes referred to as alloy "SB5".

Please don't take my numbers as gospel, read the handbook and look it up for yourself.

I often see some debate of black pipe vs. copper vs. plastic for an air distibution system, and here is my rationale for choosing copper:

1) I didn't want to use plastic, as I read too many suggestions it was not safe unless the correct type of tubing was used.

2) Black pipe requires a threading kit or machine, which I do not have.

3) Threaded joints are more difficult to assemble without leaks compared to soldering, especially with the very poor quality black pipe fittings now sold at most hardware stores.

4) Copper is lighter and easier to install yourself.

5) Copper dissipates heat faster than black pipe which promotes water condensation, which is exactly what we want!

6) Copper doesn't rust if water sits in it.

[mrennie]

The distribution line runs along the top of the wall, sloping downwards as it moves away from the compressor as suggested by proper design.

There are 2 important items to note here:

-the service drop for the end point/regulator comes off the TOP of the distribution line, not off the bottom. (Note it is still sized at 3/4", the same as the main distribution line) While this is somewhat of a PITA to plumb and not exactly cheap due to the number of fittings used, it forces the majority of the water droplets still being pulled along with the air to have to rise up against gravity if they are to get to the regulator/filter.

-note that the end of the distribution line goes to a drip leg that allows the water droplets that didn't get pulled up at the service drop to be collected and drained away.

The drip leg at the end of the distribution line can be scaled down to smaller sized tubing (1/2" in my case) since there is no flow here and you are just creating a water collection point that can be drained at the bottom.

[mrennie]

Finally, here is a picture of the combination filter (water trap)/regulator at the end of a service drop. Note that there even though there is a water trap in the filter/regulator, I installed a drip leg below the regulator in case water came down the 3/4" service drop.

Also note the drip leg to the right of the service drop, which is there to collect any water that ran along the main distribution line and was not pulled down into the service drop. Both the service drop drip leg and the drip leg for the distribution line have 1/4 turn ball valves to allow for quick and easy draining of any water in the system. They are also smaller than the service drop (1/2" instead of 3/4") since again there is no flow here and they are just there to collect water.

As mentioned at the start, this system has worked extremely well for me and I have yet to find any noticeable amount of water in the trap at the filter/regulator, even after 2 hours of sandblasting. This would seem to indicate that the water is being condensed much further upstream and is being adequately trapped.

I hope others may find this information useful, and let me know if you have any questions.

Thanks

Michael

[SD Formula]

Your system (or portions of it) look like something I might want to try. I live in Florida & the humidity here is quite high especially in the summer months.

[69lm69gp]

Great write up.

BTW with the exception of the after cooler, I created a similar set-up and have had no issues with water and I have yet to get any water out of the line drains. I do get some water from the tank but only in the summer. I also run a extra coalescing filter when using my blast cabinet and plasma cutter.

[69Customs]

Interesting! Where did you purchase your auto-drains from and what was cost?
Thanks

[stucky2947]

You sure did your homework on that setup. It looks GREAT and I'm sure functions flawlessly!! I printed your pics and the write ups and I am going to use your suggestions when I fab up mine. Thanks for the very info.

[Cammer-6]

maybe post this as a sticky
good stuff.

[mrennie]

Hi Guys:

Thanks for the positive review! It took quite awhile for me to research compressed air systems until I decided how to build my new layout, and it really works well, so I am very pleased. My previous layout was not well designed (my bad).....back in 1999 I bought my first compressor, a 3 HP, 10 CFM single stage and plumbed my garage with horizontal runs of 1/2" copper, and had a single regulator/filter right at the tank, thinking I could same money by only having one regulator at the tank. This resulted in NO water being trapped since it left the tank still very hot (and the water vapor passed right through the filter), and there was no provision to drain water from the distribution line. The result was lots of water at my tools and into my sandblaster, which obviously doesn't help remove rust when you blast steel then coat it with water!!!

[mrennie]

Automatic drain valves.....there are 3 basic types:

1) Electric drain valve that operates on a timer, opening at timed intervals whether you run the compressor or not.

2) Spring operated valve, that is closed whenever pressure is above 5-10 PSI, and will open and drain the water when the pressure drops below 5 PSI.

3) Pneumatic drain valve (several different styles), that open on startup or shut down (or both) of compressor.

I decided not to use an electric valve, since this style would be trying to drain at timed intervals and I only use my compressor on the weekends. I suppose you could wire it in parallel with the motor and use a short time interval, but I chose to use a different style.

The spring operated valves that open below 5 psi are OK for some applications, but to remove water from the tank you would need to bleed off all the air so the valve could open. Considering that compressed air is one of the most expensive resources in your shop, this is not a good choice.

I went with a pneumatic drain valve, and as mentioned there are several different styles. The one I used is available from Harbor Freight in the US (I ordered mine off eBay from a private seller since Harbor Freight wouldn't ship to me). The kits are pretty inexpensive, but other than the valve itself, the rest of the contents are garbage and should be discarded.

http://www.harborfreight.com/cpi/cta...emnumber=46960

The supplied tubing is only good for 100 PSI, so it is basically useless....I upgrade to 300 PSI nylon air brake hose which is available at a NAPA or CARQUEST store. The manual drain valve and the 1/4" NPT TEE fitting are also junk, so spend a few bucks and get new ones when you get the air brake hose.

This valve in plumbed into the unloader line and is triggered by a pressure spike, so it will operate for a second when the pump comes on, and again when it shuts off. Because the aftercooler I built causes the water to quickly condense in the tank, it is very important to drain the tank a bit each cycle, and I am sure if you saw the water that came out you would be shocked. There is also a regular manual drain valve, in case I want to drain the tank when it is not running.

The only issues I have had are a few defective O rings inside the valve, which are easy enough to change, and I get some minor air leaks in the winter when my garage is below freezing since water in the valve will expand when it freezes and push the valve open a bit. Not much you can do for this other than put a hair dryer on it until it seals up....the joys of restoring a GTO in Northern Canadian winters!

I attached 3 pics, the first is the valve on my tank, the second is the valve on my secondary air dryer, the last is the Harbor Freight auto drain kit.

[mrennie]

BTW, another tip for helping to keep water out of your tank to prevent it from rusting, it to replace the bottom drain tube (normally 1/4" black pipe with NPT threads), with a longer pipe made with 1/2" copper pipe and soldered fittings/adapters to make the 1/4" NPT ends.

By doing this you can create a longer/larger drain pipe that will hold much more water than the factory drain, which will mean that once the water drains into the drain pipe, if you forget to drain it right away the pipe would have to fill up completely before standing water would be in the bottom of your tank.

On my new compressor replacing the factory drain pipe with 1/2" copper of the same length added at least 3-4 ounces of capacity, which means less water is sitting in the tank trying.

As mentioned in an earlier post, make sure to use a solder rated for the pressure you will be using (I used 95%/5% tin antimoney). DO NOT USE 50%/50% lead/tin solder!!!!

[SD Formula]

mrennie, Just recently I tried some of your compressor ideas. I have a 20 some year old 220V, Sanborn 5 HP 2 stage with an 80 gal. tank.
Lets call this part phase one. I started with the Harbor Freight drain kit & eliminated the plastic line, as you suggested. Then I took 1/4" copper line & a Tee & ran it from the line at the top of the tank. Then I connected the other end to the HF valve that is connected to the drain line on the bottom of the tank. I ran it for a while & noticed a small amount of water comming out or the HF valve. It didn't seem like enough water coming out so after a few days of running, I opened the manual drain valve & got probably a quart of water. The valve seems to be working ok. If I put a piece of tape over the hole it will blow off when the compressor starts & stops. Any suggestions?
Anyway, phase two. I got a piece of 2" ID galv pipe 50" long, threaded & capped both ends, then drilled & taped the caps to accept air fittings. I also put a fitting on the side or the pipe approx. 10" from the bottom. Then leak tested it. I then ran a 3' flex air hose from the top side hole in the compressor tank to the pipe & attached the pipe vertically to the wall with the top of the pipe well above the height of the compressor. I haven't really tested this out yet. But was thinking that I probably should have connected the pipe directly to the tank (as I think you did) & used the bottom side air hole instead of the top one. Any thoughts on one idea vs. the other.
I bought a 2nd HF valve kit with the intention of putting it on the bottom of the 2" pipe as a secondary drain. But If the compressor bleed off air isn't working with one valve, two valves probably won't make things better. What do you think?

[mrennie]

Hi:

When you say the drain valve is connected to the line at the top of the tank, do you mean the line attached to the unloader valve? For the drain valve to work properly, the "signal" line has to be connected to an air source that goes from 0 psi to full pump pressure and back to 0 psi as the pump cycle starts and stops.

I have found with the valves I have that they stay open longer when the pump shuts off than they do when it first comes on; when the pump starts, they open for maybe 1/2 a second, but when the pump shuts off, they stay open for maybe 3 or 4 seconds. They don't cycle exactly the same either; onbe o fthem stays open alot longer than the other, possibly due to a variation is internal dimensions (made in China!!)

Note that the way the valve doesn't actually just open and close; there is a slug inside with an o-ring at each end and when the pump comes on it slides to one end of the valve body, and when the pump shuts off it slides back. As the slug moves from one end to the other it unblocks the high pressure port from the compressor tank and bleeds off air and water.

Several times over the past year I have had a problem with one or the other o-ring popping off the valve and the valve starts to leak, but that doesn't seem to be an issue for you.

For the water trap you built, the air should enter at the side near the bottom, and the air exit should be at the top. The goal is to slow the air down so the water droplets being carried by the air will drop out of the air stream by gravity and settle in the bottom 8"-10" of the trap. If your air enters from the top and exits at the side near the bottom, it will be easier to carry water droplets with it and won't be very effective.

The air connection from your tank to this trap should be fairly large; using a 3/8 flex air hose is too small, and will be a bottleneck for your system. If you have a true 5 hp motor and an 80 gallon tank you probably have a 2 stage pump running 16-18 CFM and ideally you would connect the water trap to the tank with a 3/4" line for least resistance.

If you have more questions fire away!!

[Birdguy]

I also have some experience with the Harbor Freight drain valve, and it's not very positive. The first thing to go was the crappy black plastic tubing coming off the unloader valve that blew out the first time the compressor cycled. Next I ran copper from the upper tee down to the valve under the tank.

What I have observed is that it seems work just fine initially. Over time the condensed water and oil seems to want to back up into the valve and render it useless, even with me manually draining the tank periodically. The only thing I can think of is that the cycle time for the valve is too short and not enough gets expelled when the compressor cranks up and shuts off. At least that's my experience.