EN
fr 
es 
ru 
ar 
The constraint that decides the question is not purity, and it is not flow. It is pressure.
Purity: a smaller issue than it sounds
An oxygen concentrator separates oxygen from room air using a molecular sieve. The output meets the USP 93% oxygen monograph, typically delivering somewhere between 90% and 95%. A cylinder holds oxygen at much higher purity.
Clinically this gap rarely matters. A patient breathing 93% oxygen instead of 99% is still receiving a fraction of inspired oxygen far above room air's 21%. The remaining few per cent is mostly argon, which is inert. For routine inhalant anesthesia and oxygen therapy, concentrator output is appropriate.
Where the difference does register: calibrating equipment that assumes a known FiO₂, and any application needing a documented high-purity supply. Those are the exceptions, not the rule.
Pressure: the constraint that actually decides
Different parts of a practice need oxygen at different pressures, and this is where concentrators run into limits.
| Anesthesia machine, flowmeter only | 1–2 L/min per machine at modest pressure — well within concentrator capability |
|---|---|
| Oxygen flush valve | Depends on at least 20 psi at the inlet. A concentrator feeding a machine directly should meet this minimum, or the flush will not work as intended |
| Centralised manifold | Minimum 20 psi output |
| Ventilator | Averages around 4 L/min at 50 psi. Many concentrators cannot sustain this pressure — confirm with the manufacturer before assuming a pneumatic ventilator will run from a concentrator |
| ICU or oxygen cage | Upward of 15 L/min to flush the cage, around 5 L/min to maintain |
That ventilator line is the one that catches practices out. A gas-driven (pneumatic) ventilator uses compressed gas to drive its bellows, and needs both the flow and the 50 psi to do it. A concentrator sized comfortably for two anesthesia machines may be unable to run one pneumatic ventilator at all.
The design consequence: if your oxygen comes from a concentrator, prefer an electrically driven ventilator — piston or turbine. These need mains power and a battery but no drive gas, which removes the pressure problem entirely and also removes the drive-gas consumption from your oxygen budget.
Sizing a concentrator
Add up simultaneous demand and leave real headroom. A practice whose peak demand is 60 L/min should not install a 60 L/min unit — a concentrator running continuously at its ceiling has no margin for a second case, ages faster, and leaves nothing for a cage flush.
Two factors are easy to forget:
- Altitude. Above roughly 457 m (1,500 ft), thinner air reduces both flow rate and output pressure. A unit specified at sea level may underperform at elevation, which matters for practices in highland regions.
- Room conditions. The unit needs airflow for intake and to carry away compressor heat, and the room must not accumulate the nitrogen it rejects. Larger systems may need three-phase power and professional installation — not available everywhere.
The backup is not optional
A concentrator makes oxygen from air using electricity. Lose power and you lose oxygen, immediately, with a patient connected.
Every concentrator-based practice needs a cylinder or liquid supply available as backup, with staff who know where it is and how to switch over. This is not a theoretical precaution — it is the single most important caveat on the whole cost comparison, and it means the choice is never really "cylinders or concentrator" but "cylinders alone" versus "concentrator plus cylinders".
Comparing the options
| Cylinders | |
|---|---|
| Strengths | High purity; high pressure, so drives anything including pneumatic ventilators and flush valves; no power dependency; portable; low capital cost |
| Weaknesses | Recurring purchase or rental cost; delivery logistics and the risk of running out; storage and handling regulation; periodic pressure-vessel inspection; staff time spent monitoring and changing |
| Best for | Low-volume practices, mobile and field work, and as backup in every setting |
| Concentrator | |
| Strengths | No recurring gas purchase; immune to supplier price rises, missed deliveries and contracts; no cylinder handling; sieve life typically 10 years or more, with systems lasting 15 years-plus |
| Weaknesses | Power dependency; output pressure may not drive a pneumatic ventilator or a flush valve; performance falls with altitude; needs a ventilated space; moisture or oil in the air stream contaminates the sieve, which then needs replacement; a backup supply is still mandatory |
| Best for | Practices with steady daily anesthesia volume, reliable mains power, and electrically driven ventilators |
| Piped supply from a manifold | |
| Strengths | Oxygen at every point of use; no cylinders in clinical areas; a single managed source |
| Weaknesses | Installation cost and building work; requires at least 20 psi at the manifold; a single point of failure unless redundancy is designed in |
| Best for | Multi-theatre hospitals, referral centres and new builds where the pipework can be designed in |
A note on portable concentrators
Battery-powered portable concentrators serve a different purpose: transport, field work, recovery-area supplementation and mobile clinics. Their output flow and pressure are lower than a stationary unit's, so they are generally not intended to feed an anesthesia machine's flowmeter for a full procedure. Check the stated continuous-flow output — some portable units deliver only pulse-dose flow, which is unsuitable for a breathing circuit.
What to confirm before buying
- What is the continuous output flow in L/min, and at what outlet pressure in psi?
- Does that pressure meet the 20 psi minimum my flush valve needs?
- If I plan a pneumatic ventilator, can this unit sustain roughly 4 L/min at 50 psi? If not, am I buying an electrically driven ventilator instead?
- What is my total simultaneous demand, and how much headroom does this unit leave?
- What is the altitude at my site, and what derating applies?
- What is the sieve service life and replacement cost, and what maintains it — filters, drying, oil-free compression?
- What backup supply will I hold, how many hours does it cover, and who knows how to switch to it?
- What power does it need, and will it run on my supply including any generator?
Frequently asked questions
Is 93% oxygen from a concentrator good enough for anesthesia?
For routine inhalant anesthesia and oxygen therapy, yes. Concentrator output meets the USP 93% oxygen monograph and typically runs 90–95%. The balance is largely inert argon. Against room air's 21%, the difference between 93% and 99% is clinically minor in most cases.
Can an oxygen concentrator run an anesthesia ventilator?
Only if the ventilator is electrically driven — piston or turbine. A pneumatic ventilator uses compressed gas to drive its bellows and typically needs about 4 L/min at 50 psi, which many concentrators cannot sustain. Confirm the concentrator's outlet pressure against the ventilator's requirement before buying either.
Do I still need oxygen cylinders if I have a concentrator?
Yes, always. A concentrator depends on mains power; a power failure stops oxygen production with a patient connected. A backup cylinder or liquid supply must be available at all times, and staff must know how to switch to it.
Will a concentrator work in a practice at high altitude?
It will work, but with reduced output. Above roughly 457 m (1,500 ft) thinner air lowers both flow rate and output pressure. Ask the manufacturer for performance figures at your site's elevation rather than assuming the sea-level specification.
How long does a concentrator last?
The molecular sieve should last at least ten years under normal operation, and complete systems commonly run fifteen years or more. The main threat to sieve life is contamination: moisture or oil reaching the sieve shortens it considerably and requires replacement, so intake filtration and air quality are the maintenance items that matter most.
Which is cheaper over five years?
It depends almost entirely on case volume and local cylinder pricing. A concentrator's economics improve the more anesthesia you do, because the capital cost is fixed while cylinder cost scales with use. A low-volume practice may never recover the capital. Build the comparison on your own cylinder invoices from the last twelve months rather than on a generic payback figure — and include the backup cylinders in the concentrator column, since you will still be buying them.