Here is the thing roasters get backwards when they start pricing nitrogen. The number that matters is not the purity of the gas you buy. It is the residual oxygen left inside the sealed bag, and those are two different problems with two different solutions.

You can run 99.999 percent nitrogen through a badly set flush and still seal in four percent oxygen. You can run 99.5 percent through a well set one and come in under one. Purity buys you the ceiling. The packaging machine decides where you actually land.

That distinction is why nitrogen generators for coffee roasters make sense earlier than most roasters expect, because coffee needs far less purity than people assume and the cheap part of the purity curve is exactly where generators are economical.

This goes deeper on the nitrogen section of compressed air for coffee roasters. Start there if you have not sized your air system.

The Number You Are Actually Buying

Oxygen is what stales coffee. It oxidises the lipids that carry most of the aroma, and the reaction does not stop because the bag is sealed.

Common practice is to get residual oxygen in the sealed bag under about three percent. Specialty roasters running premium or extended shelf life programs push considerably lower, and the Specialty Coffee Association threshold often cited for oxidative stability sits around 0.1 percent, with some programs targeting well under that.

So before you price a generator, answer one question: do you know what your residual oxygen actually is? A handheld headspace analyser is not expensive, and roasters are routinely surprised. If you are at four percent today, fixing the flush timing on the bagger will do more for your shelf life than any purity upgrade, and it costs nothing.

Why Coffee Sits in the Cheap Part of the Curve

Nitrogen purity and cost do not scale linearly. They scale brutally.

Making 95 percent nitrogen is cheap. Making 99.5 percent costs more. Making 99.999 percent costs a great deal more air per unit of nitrogen produced, because you are stripping out progressively harder-to-remove oxygen.

Coffee lives in the easy part:

Application	Typical purity	Residual oxygen in the gas
Standard whole bean and ground retail bags	About 99.5 percent	Roughly 5,000 ppm
Specialty single origin, pods, long shelf life	About 99.9 percent	Roughly 1,000 ppm
Laser cutting, for comparison	99.999 percent	About 10 ppm

Both coffee tiers are comfortably inside what an on site generator produces economically. That is the whole argument. A roaster is not asking the generator to do the expensive thing.

PSA or Membrane

Both work at coffee purities and the choice usually comes down to scale and how you run.

• Membrane is simple, quiet, has no valves cycling, starts instantly and suits lower flows and the 99.5 percent tier. Good for a roastery that packages in bursts.
• PSA uses carbon molecular sieve beds that cycle, scales better to higher flow, and reaches the higher purities more efficiently. Better if you package most days or want 99.9 percent.

Our PSA vs membrane comparison covers the technologies properly, and the nitrogen generator guide covers the general economics.

The Part That Catches Roasters Out

A nitrogen generator does not make nitrogen out of nothing. It makes it out of your compressed air, and it is a real load.

Depending on technology and purity, you are looking at several standard cubic feet of compressed air per cubic foot of nitrogen delivered, and that consumption rises steeply if you ask for more purity. A generator sized for your packaging line is therefore also a compressor decision.

Three things to check before you commit:

1. Does your compressor have genuine headroom on a packaging day? Not on an average day. The day you are running the bagger and the sorter together.
2. Is your air clean and dry enough for the generator inlet? Membranes and CMS beds both have inlet specifications, and oil carryover will shorten their life expensively. The generator manufacturer will state what it needs.
3. Do you have storage? Both a receiver for the compressed air feed and a nitrogen receiver downstream, so the flush demand is not fighting the generator in real time.

If the answer to the first is no, the generator is a two part purchase. We would rather tell you that up front than sell you a unit that starves your color sorter.

Sizing It Against Your Bagger

Work it from the packaging line rather than from a horsepower rule.

1. Get the flush gas consumption per bag from your bagger manufacturer. This is the real number and it varies a lot by machine and flush method.
2. Multiply by your bags per hour at peak, not average.
3. Add any other nitrogen use. Some roasters blanket a hopper or purge a grinder.
4. Add a margin and a nitrogen receiver so short bursts come out of storage.
5. Then work backwards to the compressed air the generator needs, and check that against your compressor.

If you are currently on cylinders or a dewar, you already have the usage data. Pull twelve months of delivery invoices, total the gas, and you have your annual volume and your current cost per unit without doing any estimating at all. That is the single best input to a payback calculation and most roasters have it sitting in accounts payable.

A Coffee-Specific Wrinkle Worth Knowing

Freshly roasted coffee off-gasses carbon dioxide for days, which is why the bag has a one way degassing valve.

That means the headspace in a coffee bag is not static after you seal it. CO2 comes out of the beans and displaces gas through the valve. For your purposes this is mostly helpful, because the CO2 is inert and it continues to push oxygen out after packing.

Two practical consequences. First, measuring residual oxygen immediately after sealing and measuring it a week later can give different numbers, so be consistent about when you test. Second, do not let it become an excuse. The CO2 helps, but it does not rescue a badly set flush, and the oxidation you care about happens fastest right at the start when the bean is hottest and freshest.

Running the Payback

The honest framework, which you can do on one page:

• Current annual spend. Gas, cylinder or dewar rental, delivery fees, and demurrage if you pay it. Rental and fees are often a surprising share.
• Generator capital plus any compressor capacity you need to add, plus installation.
• Running cost, which is almost entirely the electricity to make the compressed air the generator consumes.
• Maintenance, which is filters and eventually CMS or membrane replacement.
• The thing that does not show in the spreadsheet: removing a delivery dependency from your production schedule. Roasters who have run out of nitrogen on a packaging day tend to weight this heavily.

Generators tend to look good for roasters who use gas steadily. They look worse if you package in occasional bursts with long gaps, where cylinders remain reasonable. We will tell you which side you are on.

What Roasters Get Wrong

• Buying purity to fix a flush problem. Measure residual oxygen in the bag first. Higher purity gas will not fix bad flush timing.
• Specifying 99.999 percent because more sounds better. It is dramatically more expensive and coffee does not need it.
• Forgetting the generator is a compressor load. The most common reason a generator disappoints.
• Feeding the generator wet or oily air. Expensive, and the damage is gradual.
• No nitrogen receiver. Then the generator is chasing every flush burst in real time.
• Estimating annual usage when twelve months of delivery invoices are sitting in accounting.
• Never measuring the bag again after commissioning. Flush settings drift and machines get adjusted.

Frequently Asked Questions

What nitrogen purity does coffee packaging need?

Roughly 99.5 percent, about 5,000 ppm residual oxygen, covers typical whole bean and ground retail bags. Specialty single origin, pod formats and extended shelf life programs commonly specify 99.9 percent, about 1,000 ppm. Both sit comfortably in the range an on site generator produces economically, which is why coffee is a good candidate compared with applications needing 99.999 percent.

What residual oxygen should I target in the bag?

Common practice is under about three percent, with specialty programs pushing much lower and the oxidative stability threshold often cited around 0.1 percent. The important point is that bag residual oxygen is set by your flush method and timing as much as by gas purity, so measure the bag with a headspace analyser before you spend money on purity.

Should a roastery use a PSA or membrane nitrogen generator?

Membrane is simple, quiet and starts instantly, which suits lower flows and roasters who package in bursts at around 99.5 percent. PSA scales better to higher flow and reaches 99.9 percent more efficiently, which suits roasters packaging most days. Both are proven at coffee purities.

Will a nitrogen generator overload my air compressor?

It can, and this is the most common disappointment. A generator consumes several standard cubic feet of compressed air per cubic foot of nitrogen delivered, and more as purity rises. Check your compressor has headroom on a packaging day rather than an average day, and treat the generator as a two part purchase if it does not.

How do I work out whether a generator will pay back?

Pull twelve months of cylinder or dewar invoices, which gives you annual volume and current cost including rental and delivery fees without any estimating. Compare that against generator capital, any compressor capacity you need to add, and the electricity to run it. Steady users tend to do well. Occasional burst packaging often does not.

Where to Start

Before you price anything, do two things that cost almost nothing. Measure the residual oxygen in a sealed bag off your line, and pull a year of gas invoices. The first tells you whether you have a purity problem or a packaging problem, and the second gives you the only payback input that is not an estimate.

Browse nitrogen generators, or send us your bags per hour, your flush consumption per bag and your current gas spend and we will size it and run the numbers with you.