A standard without a practical test

Oil content sits in almost every anhydrous ammonia specification; metallurgical grade: 2 parts per million maximum; refrigeration grade: 3 ppm; commercial grade: 5 ppm. Those numbers are printed on product data sheets, written into supply contracts and copied onto certificates of analysis that travel with the cargo. Ask how routinely those numbers are actually measured, however, and the picture gets less clear. In my experience a significant share of plants do not routinely measure oil at all. That is not carelessness – it is a reasonable response to what the reference method asks of them.

ISO 7106

The international reference is ISO 7106, the determination of oil content in liquefied anhydrous ammonia for industrial use. A sample of liquid ammonia is evaporated off, the residue is taken up in solvent, and the oil is either weighed or measured by infrared absorbance near 3.4 microns. Written down it sounds ordinary. Run at the scale the method requires, it is not.

To recover enough residue to work with, you evaporate on the order of a litre of liquid ammonia. At a liquid density of about 0.68 g/ml, that is roughly a cubic meter of ammonia vapour to be safely handled or discharged per test. Supplier literature is candid about it: Tanner Industries’ storage and handling guide tells customers that, where quality control testing is deemed necessary, an analysis can be made using an evaporation method, then adds that analysing anhydrous ammonia by this method “is difficult and hazardous and the procedure should not be attempted by unskilled personnel.” That is striking language for a routine quality check.

The hazard is not the worst of it; the arithmetic is. Oil limits are quoted in parts per million by mass, and 1 ppm by mass is 1 milligram per kilogram, so limit and method compare directly. A litre of liquid ammonia weighs about 682 grams. At a 5 ppm limit, the oil to be recovered and weighed from that entire litre comes to 3.4 milligrams. At the 2 ppm metallurgical limit, 1.4 mg. ISO 7106 states that its gravimetric determination applies to products with an oil content of 10 mg/kg or more. In that same litre, that is 6.8 mg.

Put those figures side by side. The gravimetric route in the reference method is declared applicable only above a concentration roughly twice the commercial specification limit, and five times the metallurgical one. A plant running that route against a 5 ppm limit is working below the range that the method claims for itself. The infra-red finish in the same standard reaches lower, to about 1 mg/kg, and is the appropriate choice at these concentrations, but it still needs the same litre of ammonia and the same extraction to get there.

Solvent

That brings up the solvent. ISO 7106 specifies carbon tetrachloride. Under modern health, environmental and regulatory controls, including the US Environmental Protection Agency’s 2024 risk management rule under the Toxic Substances Control Act, it has become increasingly difficult for a laboratory to justify holding or using routinely. In practice, some laboratories substitute solvents such as hexane. That may well be sound chemistry. I have not found a published equivalence study behind it, and without one the substitution lacks a common, independently documented basis for comparison with the reference method.

The paperwork underneath is thin. ISO 7106 dates from 1985 and was reviewed and confirmed as current in 2025, so a method built around a solvent now under regulatory restriction remains the international reference for another cycle. In the United States, Federal Specification O-A-445 was cancelled without replacement more than a decade ago, and it continues to appear in supplier and industry documentation.

None of this is a failure of any one plant or laboratory. It is drift. The limits written into ammonia specifications and the framework meant to demonstrate compliance with them have moved apart, in concentration range and in what a laboratory can reasonably keep on the shelf. The predictable consequence is less routine measurement, and trend data that is rarely discussed in public. Oil introduced into an ammonia system can accumulate, coat heat-transfer surfaces and gradually degrade performance. That decline can be slow enough to be absorbed as normal equipment ageing rather than investigated as contamination.

Solutions

My suggestion is unglamorous. Start measuring, and start trending, before the standards question is settled. For trend monitoring, repeatability may matter as much as absolute agreement with a reference method. A consistent quarterly measurement can reveal accumulation long before an occasional reference test provides useful history.

There are candidates. The infrared finish already inside ISO 7106 is the obvious starting point, because it is part of the standard and reaches the concentrations being specified. Extraction, chromatographic, infrared and online techniques all deserve evaluation against this matrix, and I would not prejudge which earns the job. What the industry can usefully agree with first is what a modern method must achieve: the range it covers, the sample handling it demands, and the reproducibility it shows between laboratories.

Answer that with data, and there is something concrete to take to ISO/TC 47: a measured gap between the written method and modern laboratory practice, with evidence for what should replace it. Until then, the problem is real, but the case for changing the standard remains incomplete.