EPA RadNet spikes are due to Radon decay products

Submitted by bandstra on Sat, 2011-08-27 13:15

Berkeley Radiological Air and Water Monitoring Forum

Posted 8/27/2011

The spikes in the EPA's RadNet real-time air monitoring system have come up so often on this forum and have led to so much lasting confusion. I should have done this long ago, but here is a long explanation of why the EPA RadNet spikes are not due to radiation from Japan, but are due to natural radiation in the form of radon decay products.

First, I want to point out that, as the EPA states, sometimes there are anomalies in the data that are not due to any radioactivities. These anomalies can be spikes from electrical interference, or brief gaps when filters are switched out or the system is malfunctioning. I will not focus on any of these issues.

My comments below explain why the variation in the gamma and beta measurements by the RadNet detectors are not from radioactive isotopes from Fukushima, but rather are due to the decay products of naturally-occurring Radon gas.

Radon and its decay products

Radon gas is a naturally-occurring radioactive gas that is present in small amounts in the lower atmosphere. It has two isotopes: Rn-222 from the decay of Uranium-238 and Rn-220 from the decay of Thorium-232. Both U-238 and Th-232 occur naturally in the soil, rocks, and minerals in trace amounts. The isotope Rn-222 has a four-day half-life, and is much more abundant than Rn-220 (1 minute half-life) since it has much more time to seep out of the ground before it decays.

Radon-222 does not decay into a stable isotope, but rather it begins a "decay chain" of radioactive isotopes that eventually decay to Lead-206. Taken together, the decay products emit alpha, beta, and gamma radiation. This is because the chain contains radioactive isotopes that exhibit all three kinds of radiation:

Rn-222 alpha-decays (with hardly any gammas) to
Polonium-218, which alpha-decays (with no gammas) to
Lead-214, which beta-decays (with lots of gammas) to
Bismuth-214, which beta-decays (with lots of gammas) to
Polonium-214, which alpha-decays (with hardly any gammas) to
Lead-210

Lead-210 has a 22-year half-life, so the short-term variability in the radioactivity is determined by these first 5 isotopes.

Radon gas and its decay products are breathed in by humans every day, and this has been happening for millennia. The dose from radon and its decay products give humans an estimated 50% of our background dose. High levels of radon (e.g., in basements with poor ventilation) have been linked to lung cancer, and radon is believed to be the second biggest cause of lung cancer after smoking and ahead of secondhand smoke. An estimated 21,000 people die from radon-caused lung cancer each year. For the source of these data and more info, see the EPA Radon health risks page.

Radon Prediction #1: Beta and Gamma counts should be correlated

Something one would expect from the Rn-222 decay chain is that alpha, beta, and gamma radiation should be correlated: an increase in one should always be accompanied by an increase in the others. This indeed occurs in the EPA RadNet data. Here are about three weeks of data in June 2011 for Denver, CO. I have highlighted the correlated spikes between the gamma counts and beta counts:

So you can see there's a very good correlation there.

Radon Prediction #2: Local weather conditions can cause large variations

The concentration of radon gas and its decay products depends greatly on temperature and weather patterns. Radon gas seeps up out of the ground from the decay of naturally-occurring uranium in the soil and rocks (U-238 in particular). It is a heavy gas, and so it prefers to stay near the ground. When the air is cooler, or when there is an inversion layer "holding down" air near the ground, there can be a higher concentration of radon gas and its decay products — especially Lead-214 and Bismuth-214, which are strong gamma and beta emitters and are quite radioactive. Other weather conditions can cause large variations as well, especially rain.

Weather and temperature cycles often repeat over the course of 24 hours. If you look at any of the EPA RadNet air data, there are always up and down patterns that have a frequency of 24 hours. This can be seen in the data above, where large spikes occur nearly every day.

Here's a paragraph about the weather effects on radon from the UNSCEAR 2000 Report Vol I. Annex B, paragraph 120:

Concentrations of radon in the outdoor environment are affected not only by the magnitude of the exhalation rates in the general area but also by atmospheric mixing phenomena. Solar heating during the daytime tends to induce some turbulence, so that radon is more readily transported upwards and away from the ground. At night and in the early morning hours, atmospheric (temperature) inversion conditions are often found, which tend to trap the radon closer to the ground. This means outdoor radon concentrations can vary diurnally by a factor of as much as ten. There are also seasonal variations related to the effects of precipitation or to changes in prevailing winds. These effects must be taken into account when interpreting the available measurements, many of which are daytime samples.

So there can be huge variations in radon concentration (and also the concentration of its daughters) over a single day, just due to temperature effects. This can explain the large spikes that many people are noticing, many of which happen over the course of a day or so.

Radon Prediction #3: All Gamma energy ranges should be correlated

Gross beta count rates alone are not very helpful for distinguishing between background and non-background. This is basically why Geiger counters are not effective at finding fallout here in the US.

However, the EPA RadNet system doesn't just have gross beta — they also have gamma spectroscopy. The different channels are called Gamma Energy Ranges 2–10, and each corresponds to a portion of the gamma-ray spectrum. The gamma-ray information is much more useful than gross beta. The EPA states this same conclusion at the top of its RadNet website pages:

Two types of results from the RadNet near-real-time air monitor are presented below: gamma gross count rate and beta gross count rate. Gamma monitoring results are presented first, because they are a more useful indicator of the radionuclides associated with a nuclear power incident.

For example, Lead-214 and Bismuth-214 emit several gamma-ray lines throughout the entire spectrum. Here are some of the strongest lines from those two isotopes: 242, 295, 352, 609, 768, 934, 1120, 1238, 1377, 1764, and 2204 keV. We see these lines in our germanium detectors all the time. An increased amount of the radon decay products should correlate with increases in every EPA gamma-ray energy range, which can be seen in the following plot, where I have highlighted the spikes:

Contrast these isotopes with fission product isotopes, which happen to have fewer strong gamma-ray lines:

Iodine-131: 364 (80%), 284 (6%), and 637 (7%)
Cesium-134: 604 (98%), 796 (85%), 569 (15%)
Cesium-137: 662 (85%)

If a large amount of these were present, the gamma-ray data would show large rises in only some of the energy ranges. The signature for Iodine-131 would be a large increase in Gamma Range 3 without much of an increase in any other energies. The signature for Cesium-134 or Cesium-137 would be a large increase in Gamma Range 5 without a large increase in any other energies. I have circled those two ranges in the plots above. Both energy ranges clearly track closely with all the other energy ranges, which is exactly what we would expect from Pb-214 and Bi-214, but not what we would expect for radioisotopes from Fukushima.

Selected BRAWM Forum discussions about EPA RadNet

Mark [BRAWM Team Member]