Diesel exhaust fluid - the 32.5% aqueous urea solution sold in Europe as AdBlue and in North America as DEF - is the reagent that keeps every Euro VI truck and passenger car with selective catalytic reduction (SCR) inside its NOx envelope. The quality specification that OEMs, fleet operators, fuel distributors, and inspectors point to is ISO 22241. Under Euro 7, the specification does more work than ever: on-board monitoring must catch reagent-driven emission drift for the life of the vehicle.

The convenient assumption is that inline sensors on the vehicle or on a dispenser can confirm the fluid in the tank actually meets ISO 22241. They can do part of that job. They cannot do most of it. The gap is worth naming precisely, because misunderstanding it produces both false confidence (an “AdBlue quality sensor” reading green does not mean the fluid is compliant) and unnecessary alarm (an off-spec metals hit at a laboratory does not mean the vehicle sensor is broken).

What ISO 22241 actually specifies

The standard is a five-part series titled “Diesel engines - NOx reduction agent AUS 32”. Part 1 sets the quality requirements; Part 2 the test methods; Part 3 handling, transportation and storage; Parts 4 and 5 the refilling interfaces for commercial vehicles and passenger cars respectively.

Part 1 fixes urea content at 31.8-33.2% by mass, density at 20 degC in the 1.087-1.093 g/cm3 window, and refractive index in a matching narrow band. It then lists impurity ceilings that matter for the SCR catalyst and the injector: alkalinity as ammonia (0.2% mass max), biuret (0.3% mass max), aldehydes (5 mg/kg), insolubles (20 mg/kg), phosphates (0.5 mg/kg), and a metals table that runs down to 0.5 mg/kg for calcium, iron, copper, zinc, chromium, nickel, aluminium, magnesium, sodium and potassium. The metals limits are what make the specification hard to hit and easy to fail: a small quantity of tap water or a poorly rinsed transfer hose is enough to push sodium or calcium out of tolerance.

Part 2 is the important document for anyone measuring the fluid. It defines the reference methods: refractometry at 20.0 +- 0.1 degC and 589.3 +- 5 nm for urea content (Annex C), titration for alkalinity (Annex D), HPLC for biuret (Annex E), spectrophotometry for aldehydes and phosphates, ICP-OES or ICP-MS for the metals. In the 2019 revision the refractive index limit values and the urea-by-refractive-index equation were updated to reflect a biuret correction term. The reference methods are laboratory methods.

What inline sensors actually measure

Two families of inline devices exist. The first is the OEM in-tank quality sensor fitted by every current Euro VI heavy-duty and passenger-car OEM. These typically use one of three techniques: near-infrared absorbance around the urea combination bands, ultrasonic time-of-flight, or a combination of density plus temperature plus dielectric measurement. They read out urea concentration, temperature, level, and a coarse “contaminated / not contaminated” flag over CAN or LIN back to the ECU. Their engineering purpose is to keep the SCR dose correct across concentration drift and to detect gross misfilling (water, diesel, gasoline) that would damage the catalyst or the injector.

The second family is dispenser-side or plant-side: benchtop digital refractometers, portable NIR analysers, and process refractometers plumbed into IBC filling lines. Their purpose is to verify that what leaves the tank matches what the ticket says. On the refractometer path, calibration transfer between instruments is a live concern for chains running the same procedure across many depots. On the NIR path, model transfer and probe fouling dominate.

What both families measure well is the majority parameter - urea content - and the coarse detection of the wrong liquid being present. They see refractive index. They see, indirectly, density. They do not see biuret at 0.3% mass. They do not see aldehydes at 5 mg/kg. They see none of the metals in ISO 22241 Table 1.

The gap, stated plainly

An inline sensor reading “AdBlue OK” is telling you three things: the tank contains a liquid whose optical or ultrasonic signature is consistent with 32.5% aqueous urea; the tank does not contain a gross adulterant like diesel or plain water; the tank temperature is in a range where the SCR system can dose. That is genuinely useful information and it is why every OEM installs the sensor.

It is not telling you that the fluid meets ISO 22241-1. The specification is a laboratory certificate, not a sensor readout. The parameters that most often fail in field surveys - metals contamination from re-used containers, biuret from thermal degradation in unshaded IBCs sitting on forecourts, aldehydes from cross-contamination in a shared hose - are all invisible to the in-tank sensor and to any dispenser-side refractometer or NIR unit.

This is not a limitation of the technology; it is the correct division of labour written into the standard. Refractive index is Part 2 Annex C, laboratory equipment at controlled temperature. Metals are ICP-OES or ICP-MS in a laboratory. No sub-100 EUR sensor was ever going to substitute for either.

What that means in practice

For fleet operators, the OEM in-tank sensor and the OBM stack under Euro 7 are the compliance instruments the vehicle needs to keep operating and the inspection stack the operator will be judged against at type-approval and periodic technical inspection. They are not a quality gate on the fluid delivered by the supplier. A fleet that wants to defend itself against catalyst-warranty disputes still needs a certificate of analysis from an ISO 22241-2 laboratory per batch received, and an audit trail linking that certificate to the IBC or bulk delivery.

For fuel distributors and forecourt operators, the dispenser-side refractometer is a good gross check - it will catch a mis-branded delivery or a diluted tank - and a poor batch-quality check. The AdBlue certification schemes that exist (VDA licensing in Germany, ISO 22241 conformity assessments in most other jurisdictions) run on laboratory data, not on sensor readouts. If a fleet returns a batch citing “sensor readings”, the arbiter will still be a laboratory test to Part 2.

For regulators moving into the Euro 7 period, the standard remains the reagent gatekeeper. Under Regulation (EU) 2024/1257, from 29 May 2028 the new heavy-duty type approvals must carry on-board monitoring that watches NOx, particulate matter and ammonia over the vehicle’s life. The regulation raises the cost of reagent quality drift, not the sensitivity of the reagent sensor. The sensor is still watching optical or ultrasonic proxies; the ammonia downstream is where the OBM will show the operator whether the SCR is working.

Read the sensor as an SCR-availability signal and the certificate of analysis as the compliance signal, and the two systems stop being in tension. Read the sensor as the compliance signal, and eventually a batch will pass a green light on a dispenser and fail every metals test in a laboratory, and the resulting warranty and enforcement argument will be needlessly hard.