Content
Two suppliers quote 99.999 percent nitrogen at 50 Nm3/h. One needs a hydrogen feed, a heated purifier and a 3.0 MPa rated vessel. The other reaches the same figure with a carbon bed and a standard 0.8 MPa skid. Both quotes are honest and both will pass a factory test, because neither states the measurement basis, the dew point, or the turndown the figure was written against.
That gap between a headline number and a real operating guarantee is where most high purity nitrogen projects lose time. A specification becomes readable only when three things sit next to each other: the purity tier, the conditions the tier is guaranteed under, and the way purity will be measured once the plant is running.
High purity on a nitrogen spec sheet means residual oxygen and moisture measured at the delivery point, not the headline percentage on the datasheet. The percentage is shorthand; oxygen in ppm and dew point in degrees C are the numbers the process actually experiences.
Three quoting habits cause most of the confusion. Some suppliers quote oxygen on a dry basis and moisture separately, others quote a combined figure. Some quote purity at the generator outlet with the buffer tank empty, others quote it at a stable continuous draw. And many quote the best number the system reached during commissioning rather than the number it holds at 40 C ambient in August.
| Purity tier | Residual oxygen | Pressure dew point | Common duty |
|---|---|---|---|
| 99.9 percent | around 1,000 ppm | -40 C | Laser cutting assist gas, tank blanking, pneumatic transfer |
| 99.99 percent | around 100 ppm | -55 C | Reflow soldering, food flushing, pharmaceutical transfer |
| 99.999 percent | under 10 ppm | -65 C | Electronics assembly, bright annealing, glove boxes |
| 99.9995 percent and above | under 5 ppm | -70 C | Semiconductor tool feed, analytical instruments |
Read that table as a cost ladder rather than a quality ladder. Each step up changes air demand, sieve volume and purification hardware, and the process either needs the step or it does not.
A high purity system builds its number in five stages, and a weakness in any one of them shows up as a slow drift in outlet oxygen rather than a clean breakdown.
High Purity PSA Nitrogen GeneratorOn-site PSA nitrogen generator using carbon molecular sieve and alternating towers to reach up to 99.999% purity; relevant when sizing sieve capacity for real duty cycles and long-term purity.View Product →
Sizing the sieve for the real duty cycle instead of a nameplate flow is the most reliable way to protect a purity guarantee across a ten-year life.
The final two nines are added, not adsorbed. Carbon molecular sieve alone rarely holds below about 100 ppm oxygen across a full service life, so high purity systems install a purification stage downstream of the buffer tank.
There are two mainstream routes, and the choice is usually made by what the site already has rather than by the purity target. Both appear in the high purity nitrogen generator range and both can be supplied as a skid or retrofitted to an existing PSA plant.
LMFC Series Carbon Purification Nitrogen GeneratorCarbon-purification route for high-purity nitrogen, using alternating reactors and carbon-supported oxygen removal; compare consumable rates and reactor temperature when selecting a supply route.View Product →
Whichever route is chosen, the consumptions belong in the enquiry: catalyst or bed life in hours, hydrogen consumption per cubic metre of nitrogen, and the reactor temperature. Those are the costs that decide whether the last two nines are worth buying.
Inlet air quality, pressure stability, flow turndown and ambient temperature cause most purity complaints that are not equipment faults, and each one is visible in trend data before it becomes a rejection.
Turndown deserves its own note. A plant designed for 100 Nm3/h at 99.999 percent may only hold that purity between 60 and 100 percent of design flow; below that, cycle times stretch and the sieve never fully regenerates. If the process runs at 30 percent load overnight, say so in the enquiry.
Every additional nine costs compressed air, and compressed air is the largest operating cost in a PSA plant, usually outweighing sieve replacement over a ten-year life.
The practical consequence is that purity should be specified per point of use, not per plant. A site running laser cutting, packaging and a small electronics line rarely needs the same tier everywhere, and splitting the supply into two headers is often cheaper than purifying everything to the highest number. Nitrogen purity in electronics manufacturing is a good example: what matters is the level at the specific tool, not at the plant gate.
Purity should be verified where the gas enters the process, not at the generator outlet, because the delivery system is part of the purity chain.
CI-PC96 Trace Oxygen AnalyzerFuel-cell analyzer for trace oxygen measurement with multiple ppm and percent ranges, automatic calibration and RS485/RS232 output, suited to verifying purity at the process entry.View Product →
A permanent trace oxygen analyzer on the header turns purity from a claim into a trend line, and it shows when a purifier bed is exhausting, which is the maintenance signal that matters most.
No. The higher tier costs more air, more sieve and usually a purification stage. It is worth paying for where the process is oxygen-sensitive at ppm level, such as tool feed gas or bright annealing of reactive metals. For laser cutting assist gas or general blanking, 99.9 to 99.99 percent is normally sufficient and far cheaper to run.
Yes, if it uses carbon bed polishing instead of catalytic deoxygenation. Carbon polishing typically holds outlet oxygen around 5 to 10 ppm with no hydrogen supply, which covers most electronics and heat treatment duties. Catalytic systems reach lower oxygen levels but need a stable hydrogen source and dosing control.
Specify a pressure dew point at the point of use and state the pressure alongside it. Most high purity duties call for -60 C or lower at line pressure. Quoting an atmospheric dew point without the pressure reference makes the figure impossible to compare between suppliers.
With clean, dry inlet air, eight to ten years is realistic; with oil carryover or a failing dryer, two to three years is common. Sieve life is decided upstream by air treatment far more than by the sieve itself, which is why the pre-filter differential pressure log matters more than most maintenance records.