Because without it the system loses purity whenever demand peaks, and purity loss is invisible.

A generator has a fixed production rate at a given purity. Real processes do not consume at a fixed rate - a packaging line draws gas in pulses, a laser cutter draws heavily while cutting and nothing between parts, a blanketing system draws on tank movements.

When the instantaneous draw exceeds production, gas is pulled through the generator faster than it can be separated properly, and what reaches the process is off-specification. The generator does not stop or alarm; it simply delivers worse gas.

A receiver decouples the two. It fills during the quiet intervals and discharges during the peaks, so the generator experiences a steady average demand.

This is why sizing must come from the demand profile including peak magnitude and duration. Sizing from average consumption produces a receiver that is too small and a system that intermittently fails specification.

From the peak demand, how long the peak lasts, and how far the pressure may fall - not from average consumption.

The calculation asks how much gas the process will draw in excess of the generator's output during the worst peak, and then how large a vessel is needed to supply that volume while its pressure falls only as far as the minimum the process can accept.

That means three inputs: the peak flow rate, the duration of the peak, and the acceptable pressure band between the generator's cut-in and the process's minimum working pressure. A wider allowable pressure band lets a smaller vessel do the same job.

In practice the demand profile is often not known precisely, which argues for generosity - a receiver is a relatively cheap component and an undersized one undermines an expensive generator.

Where demand is genuinely unpredictable, logging actual consumption for a period before sizing is worth the delay.

Smooths the generator's own cyclic air draw so it neither disturbs the plant air system nor degrades its own performance.

A PSA generator does not consume air steadily. As it switches between vessels it draws in surges - one vessel repressurising while the other vents - and that pulsation propagates back into the compressed air system.

On a marginal air system, that surge can pull the pressure down momentarily, which affects other users and can also mean the generator itself is fed at less than its design pressure at exactly the moment it needs it.

An inlet receiver placed close to the generator supplies each surge locally and refills between them, so the plant system sees a smooth average draw.

It also provides a settling volume where any remaining liquid can drop out before the air reaches the treatment train. Membrane systems, which draw air steadily, need this far less than PSA machines do.

A purity analyser with alarm, an automatic vent, and pressure protection - with the vent being the item most often left out.

Continuous purity measurement is the only way to know the system is doing its job, because an underperforming generator produces gas at the right pressure and the wrong composition. The analyser should alarm on low purity and its output should be trended, since a slow decline is the earliest warning of sieve or valve problems.

An automatic vent valve dumps product to atmosphere while purity is below specification - principally at startup, but also during any upset. Without it, off-specification gas goes straight to the process, and on a packaging or cutting application that means scrap before anyone notices.

Where the consequence of off-specification gas is severe, the analyser is interlocked to stop the process rather than merely alarm.

Standard pressure relief and regulation complete the arrangement, sized to the receiver and the delivery pressure.

Coalescing filtration and activated carbon ahead of the generator, changed on a schedule rather than on appearance.

The separation media - carbon molecular sieve or membrane fibre - is destroyed by oil and water rather than worn out by use, so the filtration protecting it determines the generator's life. A typical train is a coalescing filter to remove aerosol and particulate, a dryer to the required dewpoint, and an activated carbon filter to remove oil vapour that a coalescer cannot catch.

The carbon filter is the critical and least visible one. It has no visual indicator of exhaustion - it simply stops adsorbing, and oil vapour passes to the sieve with no outward sign.

Elements are therefore changed on elapsed time or throughput as specified, not when a differential pressure gauge suggests. A blocked element shows on the gauge; an exhausted carbon element does not.

Record changes. A generator with unexplained purity loss and no filter history is very hard to diagnose.

Downstream of the generator and upstream of the regulator - storing at the higher pressure gets far more usable gas into the same vessel.

The mass of gas a receiver holds scales with its pressure. A vessel charged to the generator's full delivery pressure and feeding a regulator therefore holds substantially more usable product than the same vessel sitting downstream at the lower process pressure.

That stored quantity is precisely what covers the demand peaks, so the arrangement can decide whether a system holds specification - with no change of hardware at all.

The gain is larger the greater the difference between generator delivery pressure and process pressure.

On an existing installation that struggles with peaks, check this before assuming the receiver is too small. A receiver installed downstream of the regulator is a common design error and moving it is far cheaper than replacing it.

Yes - the receiver remains a live pressurised nitrogen source after the generator is switched off, and treating the off switch as isolation is a genuine hazard.

The reserve that makes the system work also means the distribution pipework stays charged when the generator stops. Anyone opening the system for maintenance meets stored gas at pressure.

Nitrogen released into an enclosed space displaces oxygen with no smell, no irritation and no sensation of breathlessness, so a vent into a plant room is dangerous in a way a compressed air vent is not.

The requirements are therefore a proper isolation and venting procedure that names the receiver explicitly, venting routed to a safe outdoor location, and oxygen depletion monitoring where personnel may work near the equipment.

Write the receiver into the permit-to-work rather than relying on the generator being off.