Skip to content
Environmental Solutions

What is in the air, in becquerels per cubic metre

Three instruments for aerosols, gaseous iodine and gamma spectrometry — one wheeled, two as a walk-in cabin. Designed for monitoring networks, power plant surroundings, decommissioning and emergency response.

The task

A count is not yet a measurement

An instrument that counts pulses says little. A measurement of airborne activity only becomes usable once the sampled volume is known, the efficiency has been calculated for exactly this geometry, and every figure comes with an uncertainty, a decision threshold and a detection limit. Four principles apply in the same way in all three instruments.

Principle 01 Activity concentration instead of counts Results are given in Bq/m³, referred to 0 °C and 1013 hPa. The volume is measured — in the stations by a gas meter — and not calculated from nominal flow times time. The raw count rate stays visible — as a diagnostic value, not as a result.
Principle 02 Every figure carries its uncertainty To ISO 11929, the combined uncertainty at 95 % is stated for every nuclide, together with the decision threshold and the detection limit. If a nuclide lies below them, that is what the report says — not a zero.
Principle 03 Radon is compensated, not averaged away On every filter, the short-lived daughters of natural radon-222 exceed the artificial nuclides being looked for by orders of magnitude. Ignore that and you report alarms that do not exist.
Principle 04 The report can be recalculated Every result comes with the volume, the filter identifier, the calibration status at the time of measurement and the version of the evaluation model used. A test record is produced in under a second and is byte-for-byte reproducible from the same input data.
Physics

Twelve and a half kiloelectronvolts between the answer and the interfering line

Iodine-131 emits at 364.489 keV. The strongest line of the radon daughter lead-214 lies at 351.932 keV — 12.56 keV away, and the strongest line of all in the natural background. How far a detector separates the two decides the detection limit.

0,0 0,2 0,4 0,6 0,8 1,0 300 320 340 360 380 400 ENERGIE KEV ZÄHLRATE NORMIERT Pb-214 · 351,932 I-131 · 364,489 NaI(Tl) · FWHM 34 keV CeBr₃ · FWHM 16 keV HPGe · FWHM 1,05 keV

Fig. 1 — Two lines, three resolutions. Each curve referred to its own maximum. Schematic; full widths at half maximum as stated by the manufacturers.

Sodium iodide
34 keV FWHM
Cerium bromide
16 keV FWHM
Germanium
1.05 keV FWHM
Separation of the lines
12.56 keV

Compensation runs through the second line of the same nuclide: lead-214 at 295.2 keV stands to the line at 351.9 keV in a fixed ratio of 1.933. That is a property of the decay scheme and drifts neither with the weather nor with the radon concentration nor with the age of the air. Methods that instead work with empirical side windows or with bismuth-214 as the reference do precisely that.

Fields of use

Where the instruments stand

Monitoring networksFixed points under Euratom Article 35. Reporting to EURDEP and IRIX without an intermediate layer.
Power plant surroundingsFence line and stack. Discharge monitoring with a sample line to ISO 2889.
DecommissioningAirborne activity during dismantling, where a fixed installation would be in the way.
RadiopharmacyProduction and filling of iodine-131, where the noble gas background is the real difficulty.
Nuclear medicineTherapy wards with high activity. Tens of Bq/m³ — and rarely an instrument that measures it.
Emergency responseDrive up, extend the mast, start drawing air. A first indication in minutes, a dependable figure within the hour.
Environmental measurementLong campaigns at low concentration, where the detection limit over 24 h is what counts.
Remote sitesDesert heat, alpine winter. The cabin is the housing — the electronics never see the weather.
Common basis

What is the same in all three instruments

The instruments share the evaluation, the operating software, the data paths and the way the figures come about. Anyone who knows one of them knows the others.

01

Efficiency calculated

The efficiency curve is determined with Geant4, the Monte Carlo toolkit of CERN, on cross sections from NIST XCOM and decay schemes from ENSDF. Every run is stored as a version and remains attributable to a result.

02

Spectrum unfolded

A solver with a non-negativity constraint evaluates a full spectrum against the entire library in under two seconds. True coincidence summing comes from the decay scheme — without it, Co-60 reads about 18 % too high.

03

Two calibrations

In the works against traceable sources in filter geometry, with a certificate. At the installation site, a spherical calibration over eight or fourteen source positions. The on-site calibration corrects for the location; it does not replace traceability.

04

One way out

ANSI/IEEE N42.42 in all three instruments, plus REST and a documented manufacturer interface with an example client. The two cabin stations additionally report to EURDEP 2.1 and IRIX and speak Modbus TCP, OPC UA on request. We implement customer-specific protocols.

c(DL) = 2 · ( 2.71 + 4.65 · √B ) / ( ε · p · η · V · t )

The factor of two is not a safety margin but bookkeeping: sampling and counting run at the same time, so the filter only fills up during the measurement. At a constant concentration, the mean activity on the filter is half the final value. Leave the factor out and you get detection limits that look better on paper and are not reached in the field.

Background Predicted instead of assumed A compact model — two hidden layers with 64 and 32 neurons — couples four principal components of the spectrum with pressure, humidity, temperature and a radon proxy. The mean squared error of the background estimate falls by 35 to 50 % compared with a fixed baseline. The values reported to ISO 11929 remain classically calculated.
Safeguards Five checks before the model counts Completeness of the input quantities, distribution drift via a Kolmogorov-Smirnov test, spread over fifty runs, physical bounds and confirmation by the supervisor — each stage with a defined fallback to the classical evaluation, every decision in the audit log.
Operation One interface for all instruments 1920 × 1080, control areas no smaller than 20 mm, operable with gloves. Three roles, eleven individually assignable rights, an audit log entry for every action that touches a result — and read access for the assessing body.
Data Everything stays in the instrument Raw data on a two-minute grid, a report every ten minutes — both adjustable; the cabin stations additionally hold a ring buffer covering at least three years. If the connection drops, nothing is lost — the instrument backfills as soon as it is up again. The encrypted USB stick is a handover copy, never the original.
Two further platforms

Measuring where no filter can go

Two instruments measure the gamma radiation where it arises instead of drawing air through a filter: one from the air, one under water. For neither of them is there yet a product description with substantiated figures — the pages say so explicitly and therefore state no detection limits.

In the air and in a vehicle

ENVIRO AIR

Gamma spectrometry as a drone payload, vehicle-mounted or fixed in place — for an overview and for mapping across an area.

  • Mapping of contamination, coupled to position data
  • Rapid situation assessment during deployment
  • Detector freely selectable, multichannel analyser on board
  • No sampling — no result in Bq/m³

To the page →

Under water

ENVIRO NAUTA

In-situ gamma spectrometry, in fresh water and sea water, without sampling on land.

  • Operating depth to 100 m
  • On-board data storage up to 5 years
  • Power consumption under 2 W
  • Detector freely selectable, multichannel analyser on board

To the page →

The figures for both instruments come from the previous website and have not been checked against a data sheet.

Where this stands today

The ENVIRO GO has not been built yet. The cabins of the two stations exist in this design — they can be seen in the photographs — but the measuring chains inside them have not yet passed a type test. Every performance figure on these pages is a design value or an experience value from comparable systems; detector figures come from the data sheets of the component manufacturers.

On the standards we write “design to follow” — IEC 60761-1, -2 and -4, IEC 61171, ISO 2889, ISO 18417 and ISO 11929. “Tested to” appears only where the type test has been passed. We claim no conformity we have not demonstrated. CE marking to EN 61326 and EN 61010 through an external test house.

Enquiry

Tell us what is to be measured

Four details are enough for a sound first conversation: which nuclide, which detection limit in what time, where the instrument is to stand and which reporting system the data have to reach.