OXYMAT 6E — Oxygen Analyser (Paramagnetic, also for Corrosive Gases)

OXYMAT 6E combines paramagnetic oxygen measurement with the possibility of choosing a version for a specific process matrix. The addition "also for corrosive gases" should be read as a requirement for the appropriate configuration. The public range includes 7MB2021-0CC00-1AA1 and 7MB2021-1EA00-1AA1. Universal chemical resistance of its gas path cannot be inferred from the listing of the model in this category.

Principle and two different compensations

In a variable magnetic field, paramagnetic oxygen causes a pressure difference relative to the reference. The microflow is sensed by an element in the reference flow. Another sensor compensates for vibration; see the ULTRAMAT 6 / OXYMAT 6 manual.

Siemens catalogue distinguishes between a flow and a non-flow compensation branch. Vibration compensation is not a composition correction: the magnetic properties of admixtures can shift zero.

Material compatibility is assessed along the entire route

Siemens in process analysis guides make use in corrosive mixtures conditional on appropriate materials. For the specification, create a list of parts in contact with the gas: sampling, lines, filters, fittings, seals and the measuring section. Assess each one against the composition and operating states. A suitable measuring chamber alone does not verify the seal of an upstream valve.

The sample must be dust-free and must not condense in the chamber. For corrosive media, the manual requires consultation and approval by the Siemens specialist department.

In the assembly project, separate the requirements for the environment around the device from the requirements for the gas inside. During maintenance, it must be clear which part can be disconnected, where the sample is discharged and how the measuring path is verified after reassembly. Also enter the seals in the parts list so that any service replacement is not selected just by size.

Author's example: time-mismatched correction

Imagine an assembly in which the oxygen reading is to be corrected according to another separate analyser. In the model acceptance record, a change in the matrix reaches the oxygen channel after 30 seconds and its further processing adds 10 seconds. The correction channel passes on the same change after 15 seconds of transport and 5 seconds of processing. The total model delays are 40 and 20 seconds; their difference is 20 seconds.

If the software combined the data just delivered without taking time into account, for part of the transition it would compare two different moments of the process. All times are the author's assumptions, not the responses of OXYMAT 6E or a prescription for its settings. The demonstration does not calculate the actual correction coefficient. It justifies the requirement that acceptance of any external correction should include the transient and the time alignment of both signals.

For such verification, design a common time record of the inputs, the calculated result and the validity flag. For each signal, it must be clear whether the time belongs to sampling, reception by the control system or archiving. In the evaluation, first compare the uncorrected data and only then the result of the algorithm. This separates the question of time alignment from the question of the correctness of the conversion itself.

How to formulate the acceptance test

Select a representative steady state and at least one transition that is important for the technology. Determine in advance which quantities will be recorded, what time stamp they have and who will assess the result. Also store the configuration used during the test. Without it, a later service intervention is difficult to compare with the original measurement.

A common mistake is to mistake a stable result on a calibration mixture for proof of correct correction of the entire process matrix. Another is to assess material suitability only by the name stainless steel and omit other wetted components. Both questions belong in the technical specification before ordering.

FAQ

Is a sensor outside the sample proof of resistance of the entire assembly?

No. The location of the sensing element does not eliminate gas contact with the other parts. Suitability must be confirmed for the specific combination of medium and version.

What to compare when the measuring task changes?

For a standard oxygen application, look at OXYMAT 61. For trace concentrations, continue to OXYMAT 64, which uses a different physical principle.