Walk into a car shop on a January morning and the first thing you hear is not a torque gun. It is the yard air header charging up, and somewhere down the track a carman waiting on it so he can start a brake test. When that header cannot hold pressure, nothing else in the building matters. Cars do not go out. That is the whole job.
Compressed air for railroads is a different animal from shop air anywhere else, and the reason is simple: the air is not just running your tools, it is the product you are testing. Every railcar and locomotive that leaves your facility leaves because the air system inside it passed a test that your air system administered. Get your own air wrong and you do not just lose productivity, you lose confidence in the test result.
Why Railroad Air Is Not Just Shop Air
In a machine shop, air runs equipment. In a rail shop, air does three separate jobs, and they fight each other for the same compressor.
The first job is charging equipment. You are pumping up brake pipes, main reservoirs, and auxiliary reservoirs on cars and locomotives. Brake pipe on a North American freight consist runs fully charged at 90 psi. Locomotive main reservoirs sit considerably higher, generally in the 130 to 140 psi range depending on the railroad and the unit. That means your yard air has to deliver useful pressure well above brake pipe target, with enough left over to overcome line loss across a long header, because charging a cut of cars from one drop at the end of a track is not a gentle load.
The second job is testing. Under the Air Flow Method, the brake system has to be charged within 15 psi of the highest locomotive feed valve pressure the railroad uses, and air flow has to stabilize at or below 60 CFM before the test passes. That number is worth committing to memory, because it defines your instrumentation. An air flow indicator used for this reads from 10 to 80 CFM in increments of 10 CFM or finer, and it has to stay accurate within plus or minus 3 SCFM at 60 CFM flow. Calibration intervals cannot exceed 92 days. If your supply air is swinging 15 psi between cut-in and cut-out, you are asking a technician to read a leak rate off a moving baseline.
The third job is everything else. Impact guns pulling truck bolts, needle scalers and chipping hammers on underframe corrosion, blow-off, paint, sandblasting on tank car exteriors, pneumatic jacks. This is ordinary shop demand, but it lands on top of the charging load and it lands intermittently. That combination is exactly what kills an undersized system.
Sizing the System
Start by separating your continuous load from your peak load, because they get solved with different hardware.
Continuous load is what the shop actually averages over a shift: tools in use, small leaks you have not found yet, and the steady charging air going into equipment that is already near pressure. Peak load is the initial charge on a cold cut of cars, several test racks going at once, or a blasting nozzle opening up.
| Demand | Typical draw | Notes |
|---|---|---|
| Brake test rack, per position | Up to 60 CFM during charge | Falls off sharply once the system stabilizes |
| 1 inch impact wrench | 10 to 20 CFM | Intermittent, high instantaneous draw |
| Needle scaler | 8 to 15 CFM | Often runs continuously for long stretches |
| Blast nozzle, 3/8 inch | 150 to 200 CFM at 100 psi | Dominates the system whenever it runs |
| Yard air header leakage | Varies widely | The single biggest hidden load in most rail facilities |
Those tool figures line up with our general air tool CFM chart, but the blasting number is the one that catches people out. If you do any exterior work on tank cars or hoppers, size for the blast pot or put it on its own machine. Do not try to sneak it onto the shop header.
For the base load, a rotary screw compressor is almost always the right call in a rail facility. Duty cycle is the reason. Rail shops run long, steady demand punctuated by spikes, and a rotary screw is built to run continuously without the recovery time a reciprocating machine needs. If your shop load swings a lot between shifts, a variable speed drive unit will pay for itself faster here than in most industries, because the swing is real and it is daily.
Size the trim, not just the base. A common and expensive mistake is buying one large compressor sized for peak. It spends most of its life unloaded, burning 25 to 35 percent of full power to make nothing. Two machines, a base unit and a smaller trim unit, track the load far better. Our piece on compressor control modes walks through why unloaded running costs what it does.
Storage Is Not Optional Here
If there is one thing rail shops underbuy, it is receiver capacity. The charging load is the definition of an event load: a big draw, then quiet. That is what storage exists for.
Put wet storage right after the compressor to knock down pulsation and drop out bulk condensate, then dry storage downstream of the dryer to serve the events. Size dry storage against the actual event, not against a rule of thumb. If a test rack pulls 60 CFM for two minutes and you want to ride through it on a 10 psi pressure drop, that is real gallons, and it is usually more than whatever tank came bolted to the compressor. We covered the math in sizing compressed air storage.
The practical payoff is that storage lets you run a smaller compressor and hold steadier pressure at the test rack. Steadier pressure at the rack means a cleaner leak reading, which means fewer retests.
Air Quality and the Freeze Problem
Rail shops in the northern tier have a problem most industries do not: a big part of your air goes into equipment that then sits outside at minus 20. Water that condenses in a brake pipe and freezes in an angle cock does not announce itself until a crew cannot get a release.
For general shop tools, a refrigerated dryer holding a pressure dew point around 38 to 40 degrees F is fine. For air going into equipment that will sit outdoors in freezing weather, it is not. You want a desiccant dryer on that leg, targeting a pressure dew point at or below minus 40. That is the same target the rest of industry uses for outdoor and instrument air, and the logic is identical: get the dew point below the coldest temperature the air will ever see, with margin.
You do not have to dry the whole plant to minus 40. Split it. Refrigerated dryer for the shop header, desiccant for the charging and test legs. That split usually costs less than one oversized desiccant unit and it costs a lot less to run, because desiccant purge air is real money. If you want the full comparison, see refrigerated vs desiccant air dryers.
Oil carryover matters too, though less than in food or electronics work. Brake valves and control portions are precision pneumatic devices with elastomer seals, and oil aerosol accelerates seal degradation. A coalescing filter ahead of the charging leg is cheap insurance. Our air filters catalog covers the element grades, and ISO 8573-1 decoded explains what the class numbers actually promise.
The Header Is Where the Money Leaks Out
Yard air headers are long, old, and usually threaded steel that went in before anyone in the building was hired. They run down a track, they take vibration, they get hit, and they leak from every drop and every union.
Two things are worth doing. First, meter the plant so you know what you are actually using; a flow meter on the main will tell you in a week what a decade of guessing has not. Second, run an ultrasonic leak survey on the header specifically. In rail facilities it is common to find that leakage alone accounts for a third or more of total production, and the header is where most of it lives. Our leak detection guide covers how to run the survey and how to tag what you find.
When you do replace header sections, size for the run length and not for the fitting you have in your hand. Pressure drop across a few hundred feet of undersized pipe is the reason the drop at the far end of the track never seems to have enough air. Start with pipe sizing and work backward from the farthest drop.
Mistakes Rail Shops Make
A few patterns show up over and over.
Sizing off nameplate horsepower instead of measured demand. The compressor that has been there for 30 years is not evidence of what you need. It is evidence of what somebody bought in 1994.
Treating the test rack like a tool drop. A rack pulling 60 CFM through a half inch drop with a quick coupler on it will never give a stable reading. Hard pipe it, size the drop up, and put dedicated storage near it.
Skipping the dryer on the charging leg to save capital. This one comes back in February as frozen brake pipes and bad orders, and it costs multiples of the dryer.
Letting the blast operation share the shop compressor. Blasting is a continuous, enormous, unforgiving load. It will starve everything else in the building.
Ignoring the 92 day calibration interval on flow indicators. An out of calibration indicator is worse than no indicator, because people trust it.
What We Can and Cannot Help With
We sell the plant side: rotary screw compressors, dryers, filtration, receivers, piping, and air tools. That covers the yard air system, the shop header, and the treatment train feeding your test racks.
We do not sell FRA test equipment, single car test devices, or the brake valves and control portions themselves. Those come from the rail supply houses, and they should. What we can do is make sure the air feeding that equipment is clean, dry, and steady enough that the test tells you the truth.
Frequently Asked Questions
What pressure does yard air need to run at?
High enough to charge brake pipe to 90 psi at the farthest car on the track, with margin for line loss. Most facilities run the header well above that, often 110 to 125 psi, and regulate down at the rack. If your header pressure at the far drop is barely above brake pipe target, charging will be slow and tests will drag.
Do I need a desiccant dryer, or will a refrigerated dryer do?
It depends on where the air goes. Air for shop tools is fine on a refrigerated dryer at roughly a 38 to 40 degree F pressure dew point. Air that gets charged into equipment leaving the building into freezing weather needs a desiccant dryer at minus 40 or better, or you will find the water later as ice in a brake pipe.
How much CFM does a brake test rack actually use?
During initial charge it can pull up to the 60 CFM ceiling the Air Flow Method allows. Once the system stabilizes the flow drops off sharply, which is the whole point of the test. Size for the charge, not the steady state, and use storage to cover the difference.
Is a rotary screw worth it over a reciprocating compressor for a rail shop?
Usually yes. Rail shop demand is long duration with spikes on top, which is the duty cycle rotary screws are built for. A reciprocating machine sized for the same load will run hotter, need more recovery time, and cost more per year in maintenance. The exception is a small satellite shop with light intermittent demand.
Where should I put the receiver tank?
Both places. Wet storage right after the compressor, before the dryer, to handle pulsation and bulk water. Dry storage after the dryer and close to the biggest event load, which in a rail shop is usually the test rack area. One tank bolted under the compressor is not enough for a facility with charging loads.
How do I know if my header is leaking badly?
Shut every tool off on a weekend, charge the system, and watch how often the compressor cycles. If it is loading at all with nothing running, that is leakage, and you can estimate the rate from the cycle timing. Then run an ultrasonic survey to find where it is going.
If you are rebuilding a yard air system or replacing a compressor that has been limping for years, start with measurement: know your actual demand and your actual leak rate before you size anything. Then build the treatment train to match where the air is going, not to match a single number on a spec sheet. Compressed air for railroads is one of the few applications where the air itself is part of the safety case, and it is worth doing properly.
