Ammonia: why the 8-hour average misses the real danger

This page explains what the rules and common practice call for. Every site is different, and what a survey or a filing needs at your facility is something we work out with you after we see it.

OSHA's only federal limit for ammonia is a permissible exposure limit (PEL) of 50 parts per million (ppm), as an 8-hour time-weighted average (TWA). OSHA has no short-term exposure limit (STEL) and no ceiling limit for it.

That limit is an awkward fit, because the real danger from ammonia is sudden. What hurts people in an ammonia plant is a release: a seal fails, a hot-gas defrost goes wrong, a relief valve opens, or a hose lets go during a transfer. These exposures last minutes, at concentrations a full-shift average will barely show.

An 8-hour TWA of 12 ppm and a 5-minute spike to 400 ppm can occur on the same working day. The TWA is the number your OSHA record turns on, and the spike is the number that puts someone in an ambulance.

Exposure limits

NIOSH's limits are recommended exposure limits (RELs), which are recommendations and not legal requirements. A ceiling limit is a level that must not be exceeded at any time.

Look at how those numbers relate to each other. NIOSH sets its STEL at 35 ppm, which is below OSHA's 8-hour number. The level that is immediately dangerous to life or health is 300 ppm, only six times the PEL. The margin between “within the limit” and “immediately dangerous” is narrower than most people assume.

Where we find it

• Industrial ammonia refrigeration. Food processing plants, cold storage warehouses, breweries and ice rinks, and inside them the engine rooms, compressor decks, evaporator penthouses and the roof where the relief piping ends. OSHA maintains an online guide (an eTool) for this one industry, which is a fair sign of where its enforcement attention sits.
• Fertilizer manufacturing, blending, storage and application. Nurse tanks and transfers of anhydrous ammonia.
• Confined animal feeding operations. Poultry houses and swine barns, where the ammonia is present continuously, not in bursts, and litter management drives the concentration.
• Wastewater treatment. Digesters, sludge handling and ammonia-based processes.
• Diesel exhaust treatment systems, including selective catalytic reduction (SCR). Handling urea and aqueous ammonia.
• Chemical plants that use ammonia as a raw material or a refrigerant.

How we sample for it

The tube is treated with acid for a reason. Ammonia is a base and will not stay on plain charcoal, so the acid traps it as ammonium and the laboratory measures the ammonium. A standard tube will give a low result, and nothing in the report will show that it happened.

Ammonia sampling has its own field problem: cleaning products and urine. The analysis measures ammonium. Food plants use quaternary and ammonia-based sanitizers, and poultry and swine houses have ammonia in the air. Both are real ammonia, but so is the contamination a tube picks up if someone handles it after a sanitizer wipe-down or stores it in a barn office.

Clean handling therefore matters more here than the laboratory cost suggests.

Short-term sampling has to be targeted. The short-term tube takes 7.5 L at 0.5 L/min, which is 15 minutes, so someone has to be there with a pump running during a defrost, a transfer or a rack entry. Nobody schedules a leak, but plants do schedule defrost cycles and ammonia deliveries, and that is where the sampling plan starts.

We generally pair the tubes with a direct-reading meter, a handheld instrument that shows the ammonia concentration on the spot, and use it to screen areas. The meter does not produce your official record. It tells us where in the building to place the tubes, and sampling a refrigerated warehouse without first walking it with a meter is guesswork.

→ How many air samples do you really need?

The overlap that usually decides the budget

With ammonia, the air survey and the process safety program stop being separate purchases. Process safety work is outside the four fields we work in, but you need to know how the two connect.

OSHA's Process Safety Management standard applies when a process holds 10,000 pounds or more of anhydrous ammonia. That is not a large system for refrigeration, and a mid-sized cold storage warehouse or food plant is often over the threshold without anyone having done the math lately. The quantity counts all the ammonia in the whole process, not the amount in one tank.

If you are over that quantity, the standard brings a list of requirements:

• Process hazard analysis
• Mechanical integrity (keeping the equipment sound)
• Operating procedures
• Management of change
• Compliance audits

That is a far larger obligation than meeting a 50 ppm PEL, and an air survey does not satisfy any of it. It does draw on the same walkthrough, the same piping and instrumentation drawings and the same equipment history, and the findings feed each other. When a survey records steady low-level ammonia at a compressor deck, that is usually a mechanical integrity problem under another name.

If you don't know whether you're over 10,000 pounds, find out before you buy a survey. It changes what you need far more than the sampling does.

When you may not need this

If you only use small quantities of dilute aqueous ammonia, as in a laboratory, a cleaning task or a small water treatment dose, a formal air survey may be more than you need. General ventilation plus a written review may be the right-sized answer.

A survey earns its cost in two kinds of places. The first is where ammonia is in the air continuously, such as poultry and swine housing and enclosed wastewater processes. The second is a refrigeration engine room, where the question is whether the ventilation and ammonia detectors do what the drawings say. In an engine room, the survey usually shows which of several suspected leak points is actually adding ammonia to the air.