You want a rock-solid and immediate oxygen alarm for concentration less than 19.5% in the event of a cryogenic spill leading to rapid displacement of breathing air.
You don't need a false oxygen alarm situation. It is important that the oxygen concentration measurement be correct irrespective of the foreign gas. Some electrochemical sensors have been found to be inaccurate when helium gas dilutes the oxygen, where a helium spill is the hazard to be detected. The false positive reading was about 3%, the electrochemical sensor reading 19.5% for an actual concentration of 16.6%. The Oxigraf sensor is accurate to ±0.2% with admixtures of noble gases, hydrocarbons, fluorocarbons, CO2, and N2O among other gases tested.
The following paragraph is the abstract of a DOE paper on the subject. An Oxigraf Model O2 analyzer for was used as the reference standard for the test to evaluate various ODM units that were in use at the time by DOE facilities. This was just prior to the introduction of the Oxigraf Model O2iM, so a laboratory version was used for the testing.
Investigation of personal and fixed head Oxygen Deficiency Hazard Monitor performance for Helium Gas.
On May 14, 2001, the Thomas Jefferson National Accelerator Facility (JLAB) conducted a planned liquid helium release into its accelerator tunnel to study the effectiveness of the JLAB facility to vent the helium and therefore limit the oxygen deficiency hazard (ODH). During the test, it was discovered that a wide range of various oxygen deficiency monitors, of different manufacturers, were providing substantial conflicting measurements of the true oxygen level where health effects are of concern. Yet, when tested separately with nitrogen gas as the diluting gas into air, the same models performed very well. This problem, which is associated with helium displacement of air, was found for both personal oxygen monitors and fixed installation monitors from many different manufacturers. By informing other facilities of its findings, JLAB became aware this problem also exists among other national laboratories and facilities. Many manufacturers do not have data on the effects of helium displacing air for their devices. Some manufacturers have now duplicated the test results conducted at JLAB. Since both fixed installation and personal oxygen monitors have become standard safety device in many research facilities and industries in the United States and abroad, it is important that these facilities are aware of the problem and how it is being addressed at JLAB. This paper discusses the methods, procedures and materials used by JLAB to qualify its ODH sensors for helium. Data and graphs of JLAB's findings are provided.