Gregory Shahnovsky is CEO of Modcon Systems, specializing in process analytics and AI-driven optimization for complex industrial operations.
Some technologies become famous. Others simply become essential.
Oxygen analysis belongs to the second group. It rarely appears in headlines, yet it helps enable aviation, submarines, hospitals, food packaging, emissions control, hydrogen production, natural gas processing and chemical manufacturing. In all these systems, a small change in oxygen concentration can affect safety, quality or efficiency.
The story behind this technology goes back further than many people realize. During World War II, oxygen analyzers based on Linus Pauling’s design and manufactured by Beckman were widely used in military aircraft and submarines. In those environments, oxygen was life support. Too little oxygen endangered people. Too much increased fire risk. The measurement had to be trusted.
That basic requirement has not changed. Today, oxygen sensors support medical care, packaged food safety, combustion efficiency, stack gas monitoring, gas purity, hydrogen production and hazardous process control. They have moved from specialist defense and laboratory use to being part of the everyday nervous system of modern industry.
My perspective comes from experience with industrial process analyzers and from the practical challenge of developing an inherently safer oxygen measurement approach for critical applications. My aim was to design an in-situ oxygen analyzer for direct installation in high-pressure gas pipelines without extractive sampling, because extraction can introduce delay, potential leakage points and safety compromises. For many years, this type of direct measurement was not practically available using conventional technologies.
Oxygen is chemically familiar, but technically peculiar. Its molecular behavior is shaped by unpaired electrons, which also explain its magnetic properties. It supports life and combustion, yet in the wrong concentration or location, it can become dangerous.
This is why oxygen measurement appears in so many applications. In natural gas pipelines, oxygen can indicate air ingress or leakage. In hydrogen systems, it can move quickly from quality assurance to a safety concern. In chemical reactors, it may determine whether the process is safely inerted or contaminated.
Across different industries, it makes the same point: Oxygen tells us whether reality has drifted away from assumption.
The development of oxygen analysis did not follow one straight path. I’ve observed its development across the industry. Paramagnetic analyzers use oxygen’s magnetic properties for selective measurement. Zirconia analyzers became important in combustion and stack gas applications because they can support oxygen measurement in hot and difficult environments. Tunable diode laser spectroscopy brought high selectivity and fast response for online gas analysis. Optical luminescence-quenching technology added another route, using oxygen’s effect on emitted light from a specially engineered sensing material.
Each technology has its place. The real question is not which analyzer sounds most advanced. It is which measurement best represents the process.
That is where many industrial systems fail. The weak point is often not the sensing instrument itself, but the measurement architecture. An analyzer may be accurate in the laboratory but slow or misleading in the field. A long sample line can introduce delay. A small leak can add oxygen to the sample and corrupt a trace measurement. Condensation can change the gas before it reaches the sensor. Filters, regulators and heated lines can turn a real-time process question into yesterday’s answer.
The sensor may be good. The system around it may not be.
This is why direct and in-situ measurement are gaining attention in critical applications. From what I’ve observed, measuring closer to the process can reduce sample delay, lower the risk of sample composition changes and simplify the analyzer system. That does not mean direct measurement is always the right answer, though. Temperature, pressure, contaminants, hazardous area classification and maintenance access still matter.
But the question has changed. Plants no longer ask only, “Which analyzer is accurate?” They ask, “Which measurement best represents the process?”
Hydrogen makes this question sharper. Hydrogen systems are not forgiving when oxygen is present in the wrong place. In electrolysis, compression, purification, storage and blending, oxygen measurement can support safety logic, gas purity verification, ventilation strategy, leak detection and process optimization. Many new hydrogen systems are smaller and more distributed than traditional refinery or petrochemical assets. They may not have large analyzer shelters or specialist maintenance teams.
This creates demand for oxygen measurement that is accurate, fast, robust and practical. In high-pressure gas applications, direct measurement close to the process can reduce delay and lower the risk of sample distortion. The point is to make the plant safer and easier to operate.
Leaders in this space should start with process understanding and HAZOP studies, rather than instrument selection. The first question should not be, “Which analyzer do we normally buy?” It should be, “Where is the oxygen risk, how fast can it change and how should the measurement and control strategy be integrated into the plant safety strategy?”
Several best practices follow from that. Measure as close to the process as practical and avoid sample extraction wherever possible. If in-situ installation is not possible, minimize dead volume, long sample lines and unnecessary conditioning. Design verification and calibration access from the beginning. Consider pressure, moisture, contaminants and hazardous area requirements as part of the measurement problem. Most importantly, involve process, safety and analyzer specialists together. Oxygen measurement sits between all three disciplines, and mistakes usually appear at the interfaces between them.
The remaining challenge is trust. Plants collect large volumes of process data, but operators only act confidently on data they believe. For oxygen measurement to be reliable, the industry needs robust sensors, better diagnostics, simpler field verification and installations that reduce the chance of sample distortion. A perfect analyzer connected to a poor sample system is still a poor measurement.
Reliable oxygen measurement gives operators a chance to act while the problem is still manageable. That is why it should not be treated as a minor instrument choice. It is part of process design.
In a world of tighter safety expectations, cleaner fuels, hydrogen infrastructure and rising efficiency pressure, that quiet oxygen number is becoming more important. It may not be the loudest signal in the control room, but it is often the one that tells you whether the plant is still telling the truth.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
