I worked at National Nuclear Corp. in the 1980s. I was designing electronic systems for nuclear contamination instruments, including walk-through portals and monitors for bagged waste and vehicles. The radiation detectors were large slabs of special fluorescent plastic, which give off flashes of visible light when struck by energetic photons from radioactive materials. The slabs were optically coupled to a photo-multiplier tube (PMT) to detect the flashes and amplify the resulting electronic pulses via high voltage avalanche multiplication. Pulses were tallied and the system calibrated to measure radioactive source levels.

Many of our vehicle monitors were sold to scrap steel recycling plants, because radioactive scrap getting melted into their mix could be a financial disaster costing over $800,000 and a week's shutdown for cleanup. A major pollution headache was scrapped medical equipment with its original radiation sources crushed before shipment. Another was imported scrap, including one customer who found 2,200lbs of table legs from Mexico contaminated with uranium.

One of our monitors, installed on a mill’s incoming train track spur about 100ft off of a main line in South Africa, was giving random false alarms even when the spur track was empty. We suspected some kind of arcing or high voltage leakage, especially since the assembly was mounted outdoors in an area prone to dampness. Heating the enclosure was no help, so the factory shipped me a new electronics assembly and PMT, which didn’t alleviate the problem.

Bananas blamed

After much head scratching, I suspected that maybe our detector was too sensitive and might be picking up radiation from the main track 100ft away. We had always tested our instruments for required minimum sensitivity to detect small radiation sources in a car full of typical scrap, but had no specification limit for highest sensitivity. After some legal wrangling, we finally had to resort to threat of subpoena to secure the shipping manifests of all main line shippers at the target times on "false alarm days." Lo and behold, we discovered the cause in all five identifiable cases was freight boxcars full of bananas. The detector couldn’t be easily relocated, so the customer had to end up building a special concrete block wall to better shield the monitor from the main line track.

The real culprit was radiation from the isotope potassium-40, or 40K, which comprises about .012% of naturally occurring potassium, including that in our bodies. The remainder is stable 39K or 41K. 40K decays by various methods, emitting combinations of electrons, positrons, gamma rays and X-rays at about 1.3 to 1.5MeV energy levels. 40K decays very slowly, with a half-life of over a billion years, but the portion in our bodies is still active enough to produce about 4,400 decay events per second in a 70kg man (without bananas). Those who might freak out hearing that their bodies are radioactive can relax. They would have to eat more than 50,000,000 in a day to show any signs of radiation poisoning.

I was later to learn that there were sometimes similar problems with our walk-through portal monitors used by workers in nuclear power plants. They often had a banana with their lunches and could set off alarms when they exited at end of shift. At some plants, workers milked the system for overtime pay for the delays caused by frisking with other instruments to determine if the radiation indeed was from potassium 40K rather than from real contamination.

There have also been problems with truckloads of bananas trying to cross at international borders. A very real hazard, of course, is that nuclear materials might be smuggled under a banana shipment. More sophisticated instruments can differentiate potential hazards by displaying and analysing an energy spectrum of ionising radiation. Some other foods have comparable radiation levels, including lima beans (from 40K) and Brazil nuts (from radium).

Stanley Pitman operates his own engineering design services lab.