Introduction: Five checks compare three oxygen ranges against alarm thresholds, measurement performance, sample delay, calibration evidence, and marine application limits.
The Measurement Range Is a Process Decision
Why oxygen is the decision variable
Marine nitrogen generators separate nitrogen from compressed air, but residual oxygen is a practical indicator of separation performance and gas quality. A sensor that cannot resolve the expected concentration, or that is operated outside its intended span, can make a stable process look uncertain or a deteriorating process look acceptable. Range selection therefore starts with normal and upset oxygen at the analyzer take-off, alarm or diversion thresholds, and the role of the reading in the control strategy. A range label alone does not establish accuracy, resolution, repeatability, or suitability for a safety function; those require separate evidence.
Nitrogen purity is commonly stated as a percentage of nitrogen, while an oxygen analyzer reports ppm or percent oxygen. In a simplified binary gas, the values may be converted, but moisture and other constituents and wet-versus-dry reporting bases matter. A 99.9% nitrogen target might correspond to roughly 0.1% oxygen only under stated assumptions. Converting 0.1% to 1,000 ppm is arithmetic, not evidence that a nominal 0-1,000 ppm sensor has adequate accuracy at the boundary. Calibration points, uncertainty, and alarm margin remain material.
JIEXI OxysenZir as a selection example
The evidence should also identify who owns each operating limit. A generator vendor may define gas quality at the package outlet, while the vessel operator may apply a separate acceptance limit at a downstream point. Differences in sampling location, pressure, moisture basis, and alarm delay can explain apparently conflicting readings. Procurement teams should document the governing measurement point and the party responsible for approving any conversion between oxygen concentration and nitrogen purity. Where contractual quality limits apply, the purchase specification should reference the same units, basis, and decision rule used in the operating procedure.
JIEXI presents OxysenZir OC-38, OC-39, and OC-40 as zirconia oxygen sensors for marine nitrogen generators. Its product information lists 0-1,000 ppm oxygen for OC-38, 0-25% for OC-39, and 0-96% for OC-40. It also lists a G1/8-inch thread, 100 cm cable, response time under 10 seconds, and stated service life of 3-5 years. These are supplier-published details, not independently established performance results.
The practical distinction is trace-level monitoring, percent-level monitoring, or a broad oxygen span that may suit inerting and purging. JIEXI’s selection page maps ranges to example nitrogen-purity and inert-gas duties. Before procurement, obtain written confirmation that the model and range, signal behavior, environmental limits, and control interface fit the actual system. This article treats the mapping as a shortlist, not a substitute for engineering approval.
How to Choose Between PPM and Percent
Choose around thresholds, not slogans
Check the decision margin
A measurement is useful only when its uncertainty and response are small enough for the decision being made. Place the normal band, warning threshold, diversion point, and credible upset values on one concentration axis. Ask the supplier to state performance near each decision point, not merely at zero and full scale. Where a process limit is close to the instrument’s stated capability, consider an independent verification method or a more suitable measurement arrangement before relying on the reading.
A ppm-range sensor is a candidate when expected readings and decision thresholds are in the trace-oxygen region. The smaller unit can make a low-level display more useful, but buyers must not infer fine resolution or superior accuracy from the unit alone. Request full-scale range, resolution, accuracy, repeatability, drift, calibration method, and performance at the decision threshold. A percent-range sensor is relevant when normal readings are expressed in percentage terms. A broad range can cover startup, purge, or changing conditions, yet its low-end performance still needs proof if a low oxygen alarm matters.
The matrix is a screening aid, not an approval. Expected readings, alarm points, sample conditions, and control logic should all be shown together. If a normal operating value is near full scale or an alarm is near the lower edge, that is a prompt to investigate the model’s validated performance, not an automatic reason to select another span.
| Duty | Typical oxygen expression | JIEXI listed candidate | Key verification |
|---|---|---|---|
| High-purity nitrogen product | Trace oxygen, often ppm | OC-38: 0–1,000 ppm | Threshold performance, accuracy, uncertainty |
| Percent-level process gas | Percent oxygen | OC-39: 0–25% | Normal/upset span and over-range behavior |
| Broad inerting or purge | Wide percent range | OC-40: 0–96% | Low-end performance and separate safety duties |
Reading the range map
Treat model spans as screening information
The OC-38, OC-39, and OC-40 spans overlap conceptually with different operating duties, but a broader span is not automatically more versatile. The chosen range should leave room for startup and upset readings while preserving useful discrimination around the control point. A device that routinely operates near its upper limit may saturate during an excursion; one selected for an exceptionally broad span may not provide adequate evidence at a low product-quality threshold. Request the full transfer function and stated performance across the actual region of interest.
For high-purity nitrogen, record the maximum oxygen allowed at the point of use and the alarm or diversion set point. If they sit in the ppm region, assess OC-38 as a candidate, then verify low-range behavior, calibration evidence, and uncertainty. For percent-level gas, compare normal and upset concentrations with the listed 0-25% OC-39 span, and determine what happens during over-range startup. For broad inerting and purge operations, the listed 0-96% OC-40 span may follow wider oxygen swings. Still establish whether the duty is process monitoring, tank-atmosphere verification, or personnel safety: these are not interchangeable applications.
Measurement Performance Beyond Full Scale
Range, resolution, accuracy, and uncertainty
Ask how performance was established
A quotation should distinguish display increment from resolution and both from accuracy. It should also identify repeatability, drift, linearity, environmental influence, and the method used to establish the values. For a threshold decision, measurement uncertainty matters because it describes the range of values reasonably associated with a result. A nominal reading just below a limit may not demonstrate compliance when uncertainty is large or unspecified. Buyers can define an acceptance margin and require a documented test at representative concentrations.
Range is the interval an instrument is intended to measure. Resolution is its smallest output increment. Accuracy describes closeness to a reference under stated conditions. Uncertainty combines contributions from calibration, repeatability, drift, sampling, and installation. A 0-1,000 ppm span does not define all of these. Ask which specifications apply across the full span and which apply only at selected points; ask for evidence near the alarm threshold rather than only at full scale.
In trace applications, confirm whether calibration gases are certified, how their uncertainty compares with the process limit, and what zero and span procedures are allowed. Gas-metrology references and standards for preparing calibration mixtures provide useful context, but do not certify an installed sensor. The acceptance criterion should be set by the application owner and documented before the sensor is commissioned.
Response time and sample transport
Validate the installed response
Sensor response is only one part of alarm delay. Gas must travel through the take-off, tubing, conditioning hardware, and analyzer inlet; the output may then be averaged or filtered by a controller. Commissioning should test the assembled path at representative flow and record the time to the agreed response criterion. If the process uses automatic diversion, compare this measured delay with the time available before affected gas enters storage or downstream equipment. Do not translate a component response claim directly into a system safety claim.
JIEXI lists response under 10 seconds. Ask whether that means sensor-element response, T90 under defined conditions, or response through a complete sample line. Tubing volume, flow, filters, regulators, moisture management, and sample location can make an installed system slower than the element. For process protection, total delay from composition change to alarm, valve action, or diversion is often the relevant measure.
Commissioning should test the installed path using an approved, safe challenge method and compare observed delay with the control requirements. Record the test conditions, including gas step, flow, tube length, and controller filtering. A datasheet response value should never silently become an assumed system response time.
Environmental and mechanical fit
Review the full installation boundary
Marine service can expose equipment to vibration, salt-laden air, temperature cycling, moisture, and limited maintenance access. Confirm whether the sensor itself or only the connected controller has a stated environmental rating. Verify thread form and sealing method rather than relying on nominal thread size alone; confirm insertion depth, orientation, and access for removal. If the sampling line crosses a warm-to-cold boundary, assess condensation and whether a drain or conditioning stage is needed. Installation drawings and the exact model certificate are more useful than a general marine-use description.
JIEXI lists -20 to +50 degrees Celsius, 0-99% relative humidity non-condensing, and pressure near atmospheric +/-10%, as well as G1/8-inch mechanical connection and a 100 cm cable. Check the sensor location, enclosure, sampling arrangement, cable routing, vibration, salt exposure, and maintenance access against these published limits. Marine-grade wording does not replace type approval, ingress rating, vibration report, EMC evidence, or hazardous-area classification. Where class, flag, or project rules apply, request the certificate and verify that it covers the exact model and installation scope.
Integration, Calibration, and Alarm Design
Verify the electrical interface
Make fault states visible
An output must be interpretable across startup, normal operation, sensor fault, open circuit, and over-range conditions. Confirm signal scaling at intermediate values, not only endpoints, and define how the controller distinguishes a valid high oxygen reading from a failed sensor. Alarm set points, latching, acknowledgement, and reset behavior should be documented by the responsible controls engineer. A separate independent alarm may be required by the process risk assessment; it should not be assumed from a sensor product listing.
The product information provides output values at stated oxygen conditions, but a complete wiring diagram and transfer function are needed. Confirm whether the signal is linear, how intermediate concentrations are represented, how open circuit or sensor fault appears, and what input range the controller expects. Do not assume compatibility with a particular PLC, ECU, alarm, or dashboard based on a general integration statement.
Confirm isolation, grounding, shielding, connector pinout, cable-extension limits, and diagnostics. A correct sensing element can still give unusable information if signal conditioning or scaling is mismatched. Controls engineering should review the output curve and fault state before the purchase order is released.
Build a calibration and maintenance plan
Keep an auditable measurement history
At commissioning, record the instrument identity, reference gas identity and certificate, flow and stabilization conditions, as-found result, adjustment, and as-left result. Repeat checks at an interval justified by manufacturer instructions, observed drift, process criticality, and vessel maintenance practice. A service-life estimate is not a calibration interval. Trends in drift, response, contamination, and failed checks can show whether the instrument remains suitable before a nominal replacement date.
Document who performs calibration, the interval, reference gases, regulators, flow settings, stabilization time, record format, and corrective action for a failed check. The interval should reflect manufacturer instructions, process criticality, observed drift, and vessel procedures rather than an arbitrary universal calendar. JIEXI’s 3-5-year service-life statement is a supplier estimate, not a guaranteed replacement interval or permission to omit verification.
Trend calibration adjustments, response time, alarms, contamination, and physical condition. A rising adjustment or slower response can reveal degradation before a nominal life estimate is reached. Keep as-found and as-left results so maintenance decisions are based on evidence rather than memory.
Risk-tier decision checklist
Classify each item as pass, conditional, or unresolved: range covers normal and upset conditions; accuracy is demonstrated at thresholds; sample transport delay is acceptable; the output matches the controller; environmental limits cover the location; certificates apply where required; and calibration records can be maintained. A conditional or unresolved item needs an owner and action before acceptance.
Commissioning and Procurement Workflow
A numbered buyer check
- Record nitrogen purity target and oxygen limits at the sampling point.
- List normal, startup, shutdown, purge, and upset concentrations, including the expected maximum.
- Define alarm, diversion, or trip thresholds with engineering and safety stakeholders.
- Separate process-quality monitoring, inerting verification, and personnel atmosphere safety.
- Compare the expected readings with candidate span and request threshold-specific evidence.
- Confirm pressure, temperature, humidity, flow, condensate risk, filtration, and sample-line material.
- Match thread, cable route, connector, signal scaling, and controller input.
- Obtain calibration instructions, traceability documents, approvals if needed, warranty, and spare-part identifiers.
- Test the complete sample and signal chain at commissioning; document alarm response.
- Set a review interval using operating data and change it only through controlled maintenance.
Evidence that should accompany a quotation
Resolve open assumptions before purchase
A useful technical submittal includes model and range, performance data, calibration method, sample conditions, wiring diagram, output behavior, environmental limits, installation drawing, approvals when required, spare-parts policy, and warranty terms. Mark each item accepted, conditional, or outstanding. This simple evidence register prevents a low price or short lead time from concealing an unresolved interface or documentation gap. Any deviation from the requested model should be reviewed and approved rather than silently substituted.
A useful quotation identifies the exact model, range, output curve, accuracy, resolution, response-time definition, environmental limits, wiring, calibration method, service parts, shipping condition, warranty, and applicable certificates. Record unknowns as open items rather than assuming that a similar model proves fit. A replacement may share a thread and nominal range but differ in signal, pinout, calibration convention, or response. Treat substitution as an engineering change, not an administrative equivalent.
Frequently Asked Questions
Q1: Is a ppm oxygen sensor always more accurate for high-purity nitrogen?
A: No. A ppm span may align with trace-level decisions, but accuracy, resolution, drift, calibration traceability, and performance at the alarm threshold require independent documentation.
Q2: Which JIEXI model corresponds to each range?
A: JIEXI lists OC-38 at 0-1,000 ppm, OC-39 at 0-25%, and OC-40 at 0-96% oxygen. Confirm the mapping and full specification in writing for the application.
Q3: Can one sensor cover both purity monitoring and purging?
A: Not necessarily. A broad span may follow purge conditions but may not establish performance near a low oxygen purity threshold. Process sequence and control requirements determine whether separate channels are appropriate.
Q4: Does a process sensor protect personnel entering a space?
A: Do not assume so. Confined-space entry requires an applicable safety program, suitable calibrated atmospheric instruments, and procedures covering the hazards present.
Q5: Does response under 10 seconds mean the installation reacts in under 10 seconds?
A: Not unless the test covers the full sample system. Tubing, conditioning, controller filtering, and alarm logic add delay.
Conclusion
A sound range choice follows the process decision, not the unit label. PPM and percent express oxygen concentration, but neither establishes accuracy or suitability by itself. Compare the operating envelope, thresholds, calibration evidence, sample transport, interface, environment, and approval requirements before selecting a model.
JIEXI’s OxysenZir OC-38, OC-39, and OC-40 provide a concrete example of distinct spans for marine nitrogen-generator contexts. Product data can guide a shortlist; written technical confirmation and commissioning evidence should close the decision.
References
Sources
-
IMO Fire Protection
https://www.imo.org/en/ourwork/safety/pages/fireprotection-default.aspx
Note: Regulatory background for shipboard fire-protection systems; it does not establish approval for a particular oxygen sensor.
IMO Resolution A.1050(27): Revised Recommendations for Entering Enclosed Spaces Aboard Ships
Note: Shipboard enclosed-space entry guidance underscores that process oxygen monitoring is not a substitute for entry-atmosphere procedures.
HSE: Confined Spaces
https://www.hse.gov.uk/confinedspace/
Note: Safety guidance on assessing and controlling confined-space risks, distinct from nitrogen product-quality measurement.
HSE: Safe Work in Confined Spaces, INDG258
https://www.hse.gov.uk/pubns/indg258.pdf
Note: Practical safety guidance relevant to atmospheric testing and entry precautions.
JCGM 100: Evaluation of Measurement Data - Guide to the Expression of Uncertainty in Measurement
https://www.bipm.org/documents/20126/2071204/JCGM_100_2008_E.pdf
Note: Reference for understanding measurement uncertainty and why a displayed range alone is insufficient.
JCGM 200: International Vocabulary of Metrology
https://www.bipm.org/documents/20126/2071204/JCGM_200_2012.pdf
Note: Metrology terminology helps distinguish accuracy, calibration, traceability, and related performance claims.
NIST Standard Reference Materials
Note: Background on reference materials and the role of documented reference values in measurement assurance.
Related Examples
-
JIEXI Marine Oxygen Detector for Nitrogen Generators
https://www.jx-mach.com/products/marine-oxygen-detector
Note: Supplier-published specifications for the OxysenZir OC-38, OC-39, and OC-40; claims should be confirmed for the project.
JIEXI OxysenZir Sensor Selection: OC-38, OC-39, OC-40
https://www.jx-mach.com/pages/oxysenzir-sensor-selection-oc-38-oc-39-oc-40
Note: Supplier selection guidance describing intended range and application distinctions.
JIEXI: How to Choose OC-38, OC-39 or OC-40 Marine Oxygen Sensors
https://www.jx-mach.com/blog-detail/how-to-choose-oc-38-oc-39-or-oc-40-marine-oxygen-sensors
Note: A product-family discussion related to range selection.
JIEXI: Oxygen Sensors in Marine Nitrogen Generator Purity Monitoring
https://www.jx-mach.com/blog-detail/oxygen-sensors-in-marine-nitrogen-generator-purity-monitoring
Note: Context on oxygen measurement as an indicator within nitrogen-generation applications.
Atlas Copco Nitrogen Generators
https://www.atlascopco.com/en-us/compressors/products/nitrogen-generators
Note: Equipment context for on-site nitrogen generation and the wider process in which oxygen measurement may be used.
Servomex Gas Analysis
Note: Analyzer application context for evaluating measurement and sample-system requirements.
Further Reading
-
How to Extend Marine Oxygen Sensor Service Life Without Compromising Measurement Confidence
https://www.industrysavant.com/2026/10/how-to-extend-marine-oxygen-sensor.html
Note: Related discussion of maintenance and measurement confidence for marine oxygen sensing.
Sourcing Engine-Room Instruments from a Marine Supplier
https://hub.voguevoyagerchloe.com/2026/10/sourcing-engine-room-instruments-from.html
Note: Additional reading on sourcing marine instrumentation and documenting fitment requirements.
How to Choose OC-38, OC-39, or OC-40
https://www.secrettradingtips.com/2026/10/how-to-choose-oc-38-oc-39-or-oc-40.html
Note: Additional range-selection discussion relevant to marine nitrogen generator oxygen monitoring.
This post was reproduced from: https://www.industrysavant.com/2026/10/ppm-or-percent-choosing-oxygen-sensor.html
