I have limited experience with pneumatics, so it would be pretty nice if someone could read this and provide some criticism. If the opinion is positive, I will build this system, and publish the results, performance and issues that arise with it.

I was researching the possibility of most cost-effective, expandable and flexible application of pressurized air in a workshop oriented around multiple tasks, such as a Hackerspace. I want to use this for:

-pneumatic tool operation (better than electric - lighter, more powerful, less moving elements, easier to provide air than electricity)

-sanding (most universal abrasive processing method)

-vacuum generation (electronics, but also moulding)

-injection (moulding, oiling)

-spray-coating

-hi-quality spray painting

-refrigeration through vortex tube

I want this to be as simple as possible, with fewest elements, and ability to power the setup through different means (in case there is no electricity available)

Generation of huge volumes of air require a lot of power, and use of screw compressors. They are good for continuous work, and the variable rotation speed compressors are prohibitively costly. Additionally, some screw compressors have only one screw powered, the other being powered by friction from the first one, increasing wear. I can;t be bothered to research which are which. For on/off work, we should use a piston compressor. It is also the easiest to service, and most machine shops can manufacture parts (even more durable than the originals).

I decided to go a different path - contain huge volumes of highly compressed air, generated by underloaded hi-pressure compressor.

The solution fits within 16000 PLN net sum total.

Compressor: Atlas Copco LT 3-30 or 4-30 (10000-12000 PLN, 2000 PLN net price difference, higher model has doubled FAD (Free Air Delivery - the amount of air the compressor can generate))(brochure: http://www.trident.on.ca/PDF/AtlasCopco/L_Series.pdf)

Why: It's an industrial compressor, fit for continuous work. It should last forever in our application, when it works under maximum loads only on certain occasions (sanding, refrigeration, vacuum). It generates 30 bars of pressure, and the FAD is 0.17 and 0.29 square metres per minute @30bar. It can be powered with a variety of means, including belt drive.

Tank: http://www.komnino.com.pl/of1.asp 300 litre 35 bar tank, vertical, 140 Kg of weight, ø612 x 1530, 3700 PLN. It gets filled by the compressor to 30 bar in 2 minutes (LT3), or 1 minute(LT4). It is lighter than 25 bar tank, I assume a different alloy/steel type is used.

Why: The tank is relatively small (compared to the volume of air it can house), and is relatively mobile - two people (or a single stron person) can load it onto a vehicle in a horizontal position by simply tipping it. It is not the best money/capacity ratio, which is reached around 400-700 litres, but these are cumbersome. What is really impressive is the amount of air it can contain in its volume. Under 6-8 bar, the pressure for pneumatic power tools, it can sustain a continuous operation at 220L/min (180mm angle grinder, full power) for about four minutes, without the necessity to turn on the compressor

Benefits:

-Dry air. The pressure in the installation is suited for power tools and other general tasks, therefore it's lower than the pressure in the tank. Because the pressure is reduced, the water is not gathering in the pneumatic plumbing/installation, because air does not contain enough water for condensation. The tank serves the purpose of a desiccant, without moving parts.

-No air pulsation

-If inlet of air is at the bottom of the tank, the water present in the tank will act as a filter.

-Heavier particles (oil) will have time to settle down into the water during the time the tank is unused.

Additional elements: Tank exit air warming.

The working air that exits onto your tools is heated up with heat from compression, that is normally wasted, therefore it does not cool to below atomspheric temperature during decompression, increasing efficiency of the whole system. Heat recuperation is used in some desiccant devices, so it's nothing new. An insulated oil container can be used as a way to hold heat when the compressor is not used simultaneously with working air.

The aim of this setup should allow for a very good performance in short tasks, and relatively cheap and sturdy expandability if prolonged use is required (buying additional tanks).

Additionally, it is breakdown-proof. If your main compressor breaks down, and you cannot afford to repair it, but still need to use compressed air, you can still run on the capacity of the tank(s) with a borrowed or used compressor. We're basically focusing on accumulation of air rather than generating huge amounts of it on-demand.

Opinions please.

UPDATE: I was doing a quick comparison between screw compressors from different companies and piston compressors, and the 30 bar LT series from Atlas Copco seem really impressive (unless there's a catch somewhere…). The LT-5 version can deliver 4.4 litres per second of air at 30 bar (264l/h), that's comparable to some screw compressors that are twice as expensive as the LT series, and operate in the 8-10 bar range! In addition, these compressors can work at different frequencies - I'm giving the data for 50Hz, at 60Hz the FAD gets boosted to 5.5 litres per second, and that means 330 litres per hour.

I was also doing volumetric calculations - the suggested 300 litre tank @ 30 bar holds: 1125 litres of 8 bar air (8 bars per litre) 1500 litres of 6 bar air (6 bars per litre) and feed from compressor can counter up to 330 litres of air per hour. If we would like to contain such volumes in appropriate tanks, the available tanks are slightly more cost-effective for 8 bar, and less cost-effective for 6 bar, at the same time, taking up considerable space (2,5 metre height, and almost a metre in diameter, weighting 1/3 and 1/2 as much). Increasing the pressure further to 55 bar, increases the compressor related costs, while providing not much benefit in terms of storage or efficiency.

Ensuring large capacities for air, the compressor won't have to start and stop every time you knock down the air a few bars like it is in case of normal compressors. It will be doing long cycles of work, to counter the exiting air, and replenish the air in the tank.

I also expect that reheating of the working air using compressor heat will dramatically increase efficiency because of the high compression rates. It would be ideal if we could completely cool the compressor with working air in air-heavy applications such as sanding (at least for short periods of time, when the amount of expelled air exceeds the compressor's FAD).

The amount of heat created could also be used to generate energy that can, in turn, provide several different tasks (turbocharge the compressor, precool air entering the compressor, etc). I have no data regarding temperatures generated bu the compressor, but it's a two-stage compressor. While it is crucial to maintain as low temperature as possible of the air that enters the second piston, beyond that, we have place for experimentation.

I have a crazy idea regarding hi-volume sanding using this setup. Obviously, with large volumes of air running out of the sander's nozzle, we will be running out of air really quickly. The alternative here is to burn small amounts of heavily compressed (efficiency!) air/propane mixture (could also be oil or wood gas) in a chamber, and to use the generated pressure as a motive fluid to power an injector. Heat generated will allow for production of steam, that can be delivered to the motive fluid prior to its entrance into the injector, but after the fuel has burned. Such a device should also be an effective vacuum pump for gasses and liquids (a blessing if your basement ever gets flooded, ask me how I know). I've got no calculations on this one, though, as such calculations exceed my current abilities. I did however convert a Karcher pressure washer into a pretty efficient injector for expelling water out of a basement. It involved a PVC pipe, and could spew water out at rather impressive volumes, even though the injector nozzle was misaligned.