For the last few months I have been running an experiment to see if there is any increase in ground level radiation when known broken, radioactive satellites pass overhead. Over 30 satellites have nuclear power sources, and the long term impact of the nuclear waste has not been dealt with much. The expected orbital life (some 3,000 years in one case) is far less than the half-life of the fuel. One such satellite crash-landed in Canada in 1978, luckily in an un-populated area. Of those in orbit, some are known to have broken up due to impact with debris. I chose to study Snap-10A as it is known to be in about 50 pieces now, and the orbital passes are not hard to find (using http://www.n2yo.com). It is about 1300 km above the earth.
To make the GM-tube detector (SB20) more sensitive to satellite radiation, I wrapped the detector on 5 sides with 35mm / 1.4" of lead sheeting (11.3 kg / 25 lb). This thickness is enough to reduce the gamma interference from below and the sides only by about 30%, so it's not ideal but it will make any radiation from above a bit easier to see. It certainly blocks beta radiation. The GM tube is at the bottom of this detector, so the 'window' is a rectangular cone with a narrow width of about 25 degrees. The intent is to use it parallel to the satellite path to get basic correlation from orbit to orbit, then turn it cross-ways to see a more definite rise and fall on subsequent orbits.
This detector sends raw readings every second, which was captured to a log file and time-stamped with Tera Term software. Data analysis was done using MS Excel to chart the readings from 1 hour before the orbital pass to one hour after. Each pass overhead for this satellite is about 20 minutes. The expectation was to see a rise in the radiation readings during those 20 minutes compared to the hour before and after. Here are the charts:
When I saw this first one, I thought I was onto something, like a trailing bit of debris after the tracked main satellite body passed overhead. But as you can see, this did not repeat:
Overall, the radiation during the 'visible' times does not rise compared to when this satellite is not overhead. So I think this satellite is not dangerous to people in its present orbit. Maybe its shielding is still intact, despite being in 50 pieces; maybe it is emitting radiation but is too far away (only the Space Station or other satellites can get closer).
I would like to chart other satellites, but I don't know the passing times of any other broken nuclear satellite.