RETURN of the Return Air Filter
Here's the breakdown at 10min intervals measured in CPM: 1150,1070,960,760,700,580,500,435,375,350,300,270,250,230,220,200,180,175,170,160,150,145.
Do these decay rates look like radon daughters Pb214 and Bi214?
Thanks again, James
This data fits very well to
This data fits very well to the collection of both Pb214 and Bi214 with a background of 120 cpm.
The ratio of the number of initial particles with detector efficiency is 1:1.26 Pb214 to Bi214. Without knowing the geometry or type of gas, it is hard to figure out the actual concentrations in your sample. I can ball park it by using 1/E scaling: Pb214 has lines at 242keV(7.43%), 295keV(19.3%), and 352keV(37.6%),..., and Bi214 has lines at 609keV (46.1%), 1120 (15.1%), and 665keV, and 768keV at <5%. So, weighted average gives Pb214 a 322 keV mean energy and I will estimate the Bi214 with only the 609 keV. So, I expect ~89% more counts per incident photon (609/322) from the Pb214 due to inefficiency of the detector at higher energies scaled to 1/E.
So, ballpark forensics here: ratio of Pb214 to Bi214 is 1:2.38. Below is an image from Gat, et. al., (1966) , JOURNAL OF GEOPHYSICAL RESEARCH VOL. 71, No. 6 MARCH 15, 1966, "Disequilibrium between the Short-Lived Radon Daughter Products in the Lower Atmosphere Resulting from Their Washout by Rain". This clearly shows Bi214 having greater airborne activity by ~60-80%. So, I am off here, but you get the idea.
Mark, My best global minima
Mark,
My best global minima fit was:
Initial Pb214: 373 cpm
Initial Bi214: 695 cpm
Initial Po218: 0
Background: 120 cpm
This ratio of initial cpm is 86% higher for Bi relative to Pb. This agrees well to the above image from Gat 1966.
Nice Fit!
Perfect match with radon daughter decay!
Thank you for taking such good data, and for the whimsical title! I was thinking of doing this myself with one of our air filters out of curiosity, but you beat me to it! For reference, James is referring to his original post on this thread from last week: "House return air filter high total CPM".
As a little bit of background, Radon gas (primarily the longer-lived isotope Radon-222, with a 3.8 day half-life) is naturally-occurring. It is the result of the radioactive decay of naturally-occurring Uranium-238 in rocks and soil. Radon is a noble gas, meaning it is for the most part chemically inert. However, once it decays, its daughters are charged and chemically active, and they attach themselves to fine particles in the air. These particles can be collected by an air filter. By the way, collecting such particles with an air filter is precisely what we in BRAWM are doing to collect the particles that Iodine-131 and Cesium-137 attach to in order to be transported through the air from Japan.
Here is the decay chain of Radon-222:
Radon-222 (3.8 day half-life)
alpha-decays to:
Polonium-218 (3.1 minute half-life)
alpha-decays to:
Lead-214 (27 minute half-life)
beta-decays (very often with a gamma) to:
Bismuth-214 (20 minute half-life)
beta-decays (very often with a gamma) to:
Polonium-214 (0.16 millisecond half-life)
alpha-decays to:
Lead-210 (22 year half-life)
beta-decays (very seldom with a gamma) to:
Bismuth-210 (5 day half-life)
beta-decays (very seldom with a gamma) to:
Polonium-210 (138 day half-life)
alpha-decays to:
Lead-206 (stable)
Notice that the gamma-ray activity is primarily from Pb-214 and Bi-214 -- everything else is very weak in gamma-rays. So I made a numerical model that includes the decay of Pb-214 and Bi-214 -- though Po-218 would "feed" this chain, the sample length is too long to accurately calculate the contribution from Po-218. I allowed the relative quantities of these isotopes at the initial time to vary, as well as the total radioactive decay rate at the initial time. I also assumed that every single radioactive decay of the isotopes is detected, but the exact amount will depend on a proportionality constant (based on the branching ratios of the gamma-rays and the efficiencies of the detector at each gamma-ray energy). Lastly, I also added a constant background level to the model. Here is James' data (green), as well as my model (red) and the breakdown of its components:
The fit values are:
The ratio of Pb-214 to Bi-214 should be on the order of 1, which is what is expected when they are in equilibrium. As I said before, the exact activities detected for these isotopes will vary depending on the detector efficiency, so the measured value won't be exactly 1. Notice that the initial activity is a whopping 7.4 times the background level! But by observing the decay over time, we confirm that this is nothing but the decay products of radon.
James, does that background level of 145 CPM agree with your background measurements? And thanks again for taking the time to confirm this!
Mark [BRAWM Team Member]
[Statistics details, if anyone cares: I assigned error bars to the data by calculating Poisson errors based on the average measured rates, which almost certainly underestimates errors here. The formula I used for the error is sqrt(10*CPM)/10. Using these values, the model fit has a χ2 value of 95.5 with 19 degrees of freedom, or a reduced χ2 of 5. In general, the reduced χ2 of a good fit should be close to 1 if error bars are well-estimated.]
One more point -- definitely not fission products
One more point from me on this -- For completeness, I thought I would test the hypothesis that this activity could be caused by a single radioactive isotope (rather than the decay chain of radon), I tried a fit using just a single exponential with a background. The fit was worse (reduced χ2=7, as opposed to 5 for the radon model fit), and requires an isotope with a half-life of 40 minutes.
I searched through all isotopes for those with half-lives between 35 and 45 minutes. There is only one naturally-occurring isotope in this range: Pb-211 (36.1 minutes), from the decay chain of naturally-occurring U-235. It is not a strong gamma emitter, and we never see it in our spectra here.
There are only two isotopes that are produced in large quantities in reactors that also fall in this range: Sb-130 (39.5 min) and Te-134 (41.8 min). However, these are both strong gamma emitters and we would easily have detected them if they are here in any quantity. Lastly, because their half-lives are so short, these wouldn't be able to travel across the ocean without decaying away to practically zero anyway.
Therefore, because this model has a worse fit and there is no good candidate isotope, I conclude that the radon progeny hypothesis is much, much more likely.
Mark [BRAWM Team Member]
Yes - 145 was where it
Yes - 145 was where it bottomed out after 4 hours. Once again, thanks Mark and Anonymous for the outstanding help.
Be cool,
James
Thanks so much again for the
Thanks so much again for the data -- this was a great demonstration!
Mark [BRAWM Team Member]
Actually - Thanks Mark and
Actually - Thanks Mark and dchivers for the outstandig help.
James