It is a reasonable worry, and a very common one. The cylinder the hospital used said 99%. The concentrator you were sent home with says 93%. It feels like a downgrade — like you are being given a weaker, cheaper version of the real thing. You are not. For home oxygen therapy, those two numbers are clinically equivalent, and the reasons are worth understanding properly so you can stop second-guessing the machine.
Why a concentrator makes 93% and a cylinder makes 99%
They are made in completely different ways. A medical oxygen cylinder is filled at an industrial plant that liquefies air by cooling it to around −183°C and then distils oxygen out of it — a process that can reach 99%+ purity. A home concentrator does something far cleverer for a bedside box: it draws in ordinary room air and uses a zeolite sieve to adsorb the nitrogen under pressure, then vents it, leaving concentrated oxygen behind. This is pressure swing adsorption (PSA), and we cover the mechanism in why 93% is the ceiling and how PSA oxygen concentration works.
The catch is that the sieve removes nitrogen efficiently but cannot remove argon, which makes up just under 1% of air and behaves enough like oxygen in the PSA process that it concentrates right alongside it. That is what caps a home concentrator at roughly 95.5% in theory and 93% ±3% in honest published practice. It is not a quality shortfall or a cost-cutting compromise — it is a hard physical limit of the chemistry.
What the “missing” 7% actually is
People hear “93% oxygen” and picture 7% of something harmful. It is not. The remainder is overwhelmingly argon, with a trace of residual nitrogen. Argon is a noble gas — completely inert, already present in the air you are breathing right now (about 0.93% of it), and biologically harmless at these concentrations. You are not inhaling a contaminant or a pollutant; you are inhaling oxygen plus a little of an inert gas that does nothing in your body. Beware retailers advertising “96%” or “99%” home concentrators — at a home scale that claim is almost always a measurement trick or a misrepresentation, a point we make in why 93% is the ceiling.
Does the difference matter clinically?
For home therapy through a nasal cannula, no — and the reason is in how oxygen is actually prescribed. Your prescription is written in litres per minute, not in percent purity, because the dose your body receives is governed by how much oxygen flows to you, not by the final few percent of concentration. A patient on 2 L/min of 93% oxygen and a patient on 2 L/min of 99% oxygen receive almost the same effective oxygen dose, because in both cases the oxygen is further diluted by the room air you also draw in around the cannula prongs with every breath. That dilution swamps the small purity difference entirely.
This is exactly why long-term oxygen therapy guidelines worldwide — the standards that go back to the landmark NOTT and MRC trials — are written around concentrators delivering 93% ±3%, and why that figure is the accepted clinical standard rather than a tolerated compromise.
When higher purity genuinely matters
There are settings where the last few percent counts — but home cannula therapy is not one of them. Higher-purity oxygen matters for:
- High-flow nasal cannula or ventilator circuits, where oxygen is the carrier gas and the fraction delivered is controlled precisely.
- Certain surgical and anaesthetic uses.
- Specific clinical situations your physician will name explicitly, usually in hospital.
If you are on a standard home prescription of a few litres per minute by cannula, you are not in any of these categories.
When you might still want a cylinder — for reasons that are not purity
A cylinder still has a real role in a home-oxygen setup, just not because of its purity:
- Power-cut backup. A concentrator stops in a blackout; a small backup cylinder at the bedside buys you 2–3 hours of runtime. In many Indian cities, overnight load-shedding makes this essential — see concentrator night-use considerations.
- Short acute needs before a concentrator is arranged.
- Prescribed flows above what your concentrator delivers at rated purity.
- True portability away from mains power, where a pulse-dose portable concentrator or a small cylinder is the practical choice.
The full trade-off is laid out in oxygen cylinder vs concentrator. The point is that the decision is about runtime, backup, and mobility — never about the 93-versus-99 number.
The number you should actually watch
The purity figure that matters is not 93 versus 99 — it is whether your concentrator still holds 93% at your prescribed flow rate. Two things erode it:
- Higher flow. Purity naturally falls as flow rises; a 5 L/min unit reading 93% at 2 L/min may read 86–89% at its full 5 L/min, which is normal physics, covered in why oxygen purity drops at high flow.
- An ageing sieve bed. Over thousands of hours the zeolite degrades and the whole flow-vs-purity curve drifts downward — the pattern explained in sieve bed lifespan and degradation.
This is where a live purity readout earns its keep. The Home Medix HM-KX (up to 10 L/min) and the quieter HM-KV (up to 5 L/min) both display real-time oxygen purity and flow, so you can confirm the unit is sitting inside 93% ±3% at your setting rather than trusting the spec sheet — and catch a degrading bed early, before purity quietly falls below 90%.
Takeaway
A concentrator’s 93% and a cylinder’s 99% are clinically equivalent for home oxygen therapy, because your dose is set by flow rate, not by the final few percent of purity — and the part that is not oxygen is inert argon, not a contaminant. The 93% figure is the honest physical ceiling for a home machine and the worldwide standard for long-term oxygen. Keep a cylinder for backup and mobility if your situation needs it, but for the therapy itself, the thing worth monitoring is simply that your unit holds its rated purity at your flow rate over time.
This is general information, not medical advice. Follow the flow rate your physician prescribed, and raise any concern about your oxygenation — breathlessness, a falling pulse-oximeter reading — with them directly rather than adjusting therapy yourself.