OT: Radiation - a little perspective

Unfortunately, it rather misses the point. The main risk (if there is one) to the general public isn't from direct exposure to radiation emanating from the plant, it's from radioactive pollutants.

Of particular concern in this regard is the plutonium from the MOX fuel at reactor #3. As a nearly-pure alpha emitter, the radiation is practically undetectable at a distance (the penetration depth of an alpha particle in air is a few centimetres), but can cause significant damage if ingested or inhaled, or even from contact with the skin.

Dose was informative. Thanks.

And in particulate smoke hot dust which can be ingested into the lungs. Alpha emitters though short range and easily blocked do enormous damage internally (about 20x more per particle than beta or gamma).

I am not entirely surprised that small doses of very penetrating gamma radiation are perhaps less damaging than heavy relativistic charged particles for a given dose. The tricky thing is that in reality with an open source you get some of each. And certain radionucleides which are very common in spent fuel and volatile like iodine or water soluble like caesium have a bad habit of concentrating in the thyroid gland or bones respectively.

Neither of these outcomes is good for you and if you really doubt that too much radiation is a bad thing you would do well to study the unfortunate Radium sportsman Eben Byers.

Most commonly causing lung damage. It is thought that traces of polonium in tobacco smoke plays a part in causing cancer in smokers.

Regards, Martin Brown

Black sand beaches eg Kerala, Brazil. Background radiation up to 400x normal

--
Dirk

http://www.neopax.com/technomage/ - My new book - Magick and Technology

Gabon Africa

Cheers

I don't think it misses the point at all and a bit of factual analysis does help to dispell the the more alarmist reports, which is all we see in the media at present.

As for pollutants, i'm sure they know what they are doing and the Japanese in particular are more obsessive in such areas. They will be using a wide variety of sampling methods, both near and far from the plant, with detectors including air and water sampling types...

Regards,

Chris

I'm saying that, in the context of the risk to the general public, talk of radiation levels (in terms of waving a Geiger counter around) misses the point.

Oh, I'm sure that such readings are being taken. Whether such readings are made public is a different matter. The media focus has been almost entirely on direct measurement of radiation levels, which really isn't the issue.

The only information I've seen on contamination so far is excess levels in water, milk and vegetables near the plant, and levels beyond what is (continuously) safe for infants in the water in Tokyo.

That last one is actually quite worrying. Sure, the actual levels aren't particularly high, but Tokyo is quite some distance from Fukushima. What are the contamination levels in between? It can't just be an area around the plant plus Tokyo, surely? How long will it take before contaminants disappear from the water and food supplies? Are we seeing the peak of the curve or just the rising edge?

I certainly don't know the answers. The media seems more concerned with publishing readings from Geiger counters.

How else do you measure radiation levels? And do any of these reports say anything about danger levels relative to, say, a chest X-ray, or a CT scan, or the output of a coal-fired power plant?

Thanks, Rich

If you are referring to the Japanese gov't disclosing measurements, they are doing a better job than I've seen elsewhere. (I'm sure it's not full disclosure by any stretch, but it is something.) See this link as a starting point, then drill in. You will see specific measurements of specific isotopes in some of the PDF files:

Yes, but most of the measurements I've seen reported were collected and reported by reporters who don't understand the terms nor the right questions to ask.

For example, take this report:

Near the bottom, you can read:

"The latest results accumulated and posted online by Japan's education, science and technology ministry showed slight but notable upticks in airborne radiation readings around Japan in recent days. But even the highest readings, .11 millisieverts some 30 kilometers northwest of the plant, were still considered significantly below what's considered dangerous to humans."

110 uSv? Over what time period was that taken?

If you look at the first link I gave, from MEXT, you will see that the PDFs report values that are "per hour." The reporter didn't understand that uSv/hr is probably what was being discussed. uSv is useful for total dose, but again this is usually more meaningful when the time period is known and the MEXT data reports it that way, as it should. Chances are, the ".11 millisieverts" was meant as 'per hour.' And with that is close to a rate of 100 rem/yr, which is perhaps

200 times the usual background+medical many people are more commonly exposed to in a year. Of course, if only for a short time, it's not a great concern. But if you live there and want to continue doing so, you might care more. Still, it doesn't tell you _what_ the isotopes are and the MEXT data sets 1 Gray = 1 Sievert, so you don't know the bioequivalent or how long to worry without knowing the isotopes yielding those readings.

Just for interest, some of the readings reported were as high as 300 mSv:

"officials who briefed congressional staffers Friday said they had earlier in the week detected radiation levels above reactor No. 4 in the Fukushima complex spiking to 300 millisieverts"

Now, it's possible they meant microsievers. Who can be sure? But let's say the figure is right. And since the reporter didn't get the time and since the MEXT measurements are given in "per hour," we can assume the reporter failed to include that time period ("spiking" implies time) and we should assume that 300 mSv/hr was meant. That's about 83 uW/kg. Every kg of body mass experiencing 83uW of some kind of radiation (probably gamma and maybe some beta, though without knowing the instruments involved it's hard to be sure on that, either.) It's also 1/4 million rem/yr, if it is to be believed. The report also doesn't tell you where this "spike" was measured -- how far "above" the core, itself. It might be nice to know that, too.

Take a look at the MEXT site. Some of the "actual levels" seemed pretty high to me, though I'm no expert on this stuff. I'm certainly not accepting quoted qualifications by gov't officials of "miniscule" as being definitive. The US is making those kinds of statements when discussing what is being blown across the ocean -- and they are probably right about it -- but I'd rather know what they are using to make these measurements, how often, what thresholds they are using to initiate further testing, and exactly what isotopes are being checked or what ranges of energies they can measure and distinguish and what the digested readings actually are so that I can hire an independent specialist and ask what they think. When I'm able to get enough to do that, I'll trust the broader statements as tentatively honest and true.

150 miles, they often repeat. Osaka is another 250 more miles, often stated in news reports.

Read the MEXT site. For example, in this:

they report that Tochigi measured 540 MBq/km^2 for Iodine

131. I haven't checked again. But a few hours ago, China detained two Japanese travellers said to have radiation levels that "seriously exceeded limits." That doesn't bode well:

Just been reading, lately. So what follows isn't informed gospel, just what I think I managed to glean:

Iodine 131 decays very rapidly. So it's a concern for some, but less so for those of us half a world away. But Cesium

137 and 135 are other matters. As is Strontium 90. Actually, I can't even find out what all the nasty isotopes might be in these systems -- even Wiki can only report about 50% of the mix and doesn't know what the other 50% contains, in practice. I suspect some of the details are confidential.

But with neutrons jamming about, even stable isotopes are only temporarily stable.

There are far too many variables right now to know an answer to your last question. I wonder if anyone can put a bow on it. They don't know how much salt there is, caking up, or where it is caking. They don't know the corrosion status. They didn't even know the temperatures, until recently, and only then from just a few measurement points -- which were very concerning as the values were higher than hoped. I think this is a huge, nasty brew and only time will tell the results. Only the most vague predictions are possible right now. Main thing is to keep punching and hope a lot. Japan isn't so big it can afford to lose a lot of territory. Chernobyl's restriction zones cover three countries and a lot of land.

They don't even know how to publish what they are being told.

What I'd like to see are specific details about the instruments being used, locations used, when used, for how long, how they prevent contamination of the instruments from interfering from the measurements, what particles and energy ranges are their instruments sensitive to, how are they calibrated and how often, what thresholds do they use for deciding if more work is needed (for example, if they are using microfiber filters followed by carbon granule traps, what levels will trigger more analysis?) As full disclosure as possible, enough one hopes that independent, informed opinions can be formed from the reports with some chance at being similar to others'.

For some examples of instruments, look up alpha spectroscopy with an MCA (multi-channel analyzer) or as just low resolution, gas-flow proportional counter, liquid scintillation spectrometry, alpha scintillation and Lucas cells, to name a few. For hand-held devices, there are ionization chambers (not so good really), Geiger-Muller (I have a couple of these, as they are easy to design and build), and a host of scintillation types like ZnS, NaI, CsI, plastic, BF3, and 3-He, to name a few. (The last two are mostly good for neutrons and poor to useless for anything else.) Thin mylar may be used to enhance detection of alpha or beta. 20 years ago, mercuric iodide was also being talked about though I can't recall seeing it around, of late.

Geiger counter readings are pretty crappy for gamma and neutron, decent with beta, and vaguely okay with alpha (with thin windows.) Geiger-Muller doesn't usually tell you much about energies, just acts as detection only and can mostly just count events within detectability boundaries (which only gets up to about 2MeV and probably no lower than 30keV, memory serving.) Moderators, like 3-He, help in detection of fast neutrons. I haven't ever seen a geiger-muller with a

3-He moderator sheath, though. Wonder if anyone makes such a tube.

Bottom line is that some Japanese folks, and not folks who supposedly were nuclear power workers working recently at Fukushima, were detained in China for serious levels of radiation emitting from them. I don't consider that a good sign. Vegetables and fish, with excessive levels, have been stopped and refused in China and Taiwan. Also not a good sign. Black smoke, 4 reactor buildings demolished in hydrogen explosions, open pools of spent fuel running hot, cracked reactors (several, it appears), salt caking up on the rods by the ton, ... and all of this just after a terrible earthquake, loss of many thousands of lives and infrastructure from a tsunami, .... it's the kind of thing where you really hope for a lot of good luck.

Jon

Sure, all the time:

That's a UK rag, an Arab rag, and a US/LA rag. Plenty more where those come from, and each and every single day, too. Over and over and over, again. Across the world, in every country, from every political stripe.

So .. yes, they do.

Jon

Yes, after running a reactor for some time, it is a devil's soup of isotopes, but the most worrisome are well known.

CsI, NaI and BGO (Bismuth Germanate, desirable because of great density) can be used for Gamma scintillators, and there are hand-held instruments that can take a spectrum and identify isotopes within minutes, they just need to accumulate enough counts for clear peaks to emerge.

Beryllium and Mylar-window GM tubes are pretty common, and work fine to pick up the Alphas from typical contaminant isotopes. Even the metal-window ones work fine to pick up Uranium emissions. Geiger-Muller doesn't usually tell you much

They do make 3He tubes very much like a Geiger-Mueller tube, metal shell (doesn't have to be thin) with a hair-thin wire down the center. They do require a LOT more voltage than typical G-M tubes, something like 2 -

2.5 kV. We have some here. The signals out are pretty small, so it takes careful design to get valid readings. We had a problem with ceramic capacitors giving off "micro-seismic" signals due to the electrical field stress that looked exactly like real signals, it took weeks for that to settle down.

Jon

I'd suppose so. Even better than that, since IAEA investigators can take a look at small samples and decide whether or not Iran is lying about where they get their fuel supplies. I expect there is rather precise knowledge to some. Just not the rest of us.

But I take your point, too, that the ones of reasonable public health concern are probably well enough known.

It appears that it is mostly the volatiles and not the rest that is being emitted from the Fukushima plant. I suppose that's as good a sign as can be hoped.

MCA? In any case, the solid state versions bring a question to mind. Perhaps you have an answer.

Some years ago, I was setting up an ice-bath cooled Hamamatsu Si detector (1cm^2 at the time) using an ACF2101 integrator design to do some experiments on pyrometry detectability of temperature. Set up a hot plate, fiber optics to point at it feeding back to the cooled detector (and electronic package, too, underwater) and had this run for a week while collecting ADC data at several integration rates. Thin film filtering of the light was also used to limit the band (+/- 20nm around

880nm, memory serving.)

The results were excellent. We could easily detect a 1C change at 180C on the hot plate (controlled to oscillate up and down a couple of degrees.) However...

We'd get periodic spikes. These were, by appearances, random and occurring on order of sub-minutes to minutes. And they had different 'heights.' After back-calculating, somewhere in the rough range of a few hundred electrons to perhaps a few tens of thousands of electrons. Since astronomy is a hobby of mine, I ran around looking for existing data on cosmic rays at sea level (where we were testing) and found some excellent reports on Si detectors done in France that were consistent with our observations. So I grabbed up some autunite I had at home and we ran some tests with it. No question, the stuff produced similar results but at higher event frequencies.

Turns out that if the detector is "on edge" relative to the source I had, I'd get higher but less frequent spikes. (More electrons stripped up as the emissions stripped up electrons long-ways through the depletion region.) And "on face" I'd get lower but more frequent spikes. Which brings me to my question.

I'm considering the idea of using a relatively cheap and durable high back bias PIN diode for the detector of a radiation counter. But if a solid state detector is used, it seems to me that the pulse height depends on orientation to the target and that then characterizing the energy, even if many pulses are accumulated to get the distribution, is a bit of a problem in both calibration AND use. Is calibration done using pin-holes at a carefully constructed distance, with the same being then true for testing samples?

Can you point me to any manufacturers with good descriptions of the use of any kind of Si detectors used this way?

It sounds like a fun experiment to get back doing.

Yeah. The one I built for myself 35 years ago used metal and it did really well with the autunite mineral sample (that I still have around, though my wife forces me to keep it a LONG way away from us, as she can easily see how it destroys the styrofoam base I've got it sitting on.)

Any source of these I might consider buying from that you can point towards? I'd be very interested in re-trying my hand at building another geiger-muller but using such tubes. I'd need a supplier, though.

Thanks for your comments, Jon

Yes, that's right, a scintillator and MCA, with a computer to find and match the peaks to known spectra.

Ah, yes, I've been bitten by this. This is the deadly 1/f noise, to which certain Silicon devices, especially, have problems with. It could be trapping of ions produced by cosmic ray showers, but it could also be just flaws/impurities in the chips or detector, too.

I have a system we have built using custom chips made by MOSIS (USC) for us. One of the problems with this design is it needs REALLY stable power supplies to reach its best noise performance - as in stable over hours to tens of uV. We had a lot of jumping around, where a steady input signal would record at a particular level for minutes, and suddenly jump a percent or so, and then hold again for some minutes. I was using a fairly low-noise op-amp in the power regulator, but it wasn't tested at low frequency. I searched the analog Devices data book for one that was individually TESTED for a low 1/f corner frequency, and the AD706 was the best conventional (not chopper-stabilized) op amp there, with tested performance to .001 Hz. Solved the problem COMPLETELY!

Yup, can't calibrate the energy deposited, IF :

1. some particles can punch all the way through the detector
2. you can't control the angle of incidence

If all particles are stopped (may require VERY thick Silicon, ie. a mm or more) then angle of incidence matters less, it is only affected by the "dead layer" or passivation on the front.

Hamamatsu (might not want to order anything from them for a while) Micron in the UK and, of course, TONS of literature in the Physics research journals. The Fermi gamma ray telescope (formerly called GLAST) has several thousand large silicon strip detectors in it.

Jon

Okay. Maybe I'm not entirely in the dark, then.

I don't believe we were observing near-DC 1/f and it wasn't just me, the data was examined by electronics design engineers and two physicists (one a well-known NIST specialist), as well. The FPGA allowed us to operate the ACF2101 over quite a sampling range, too. And using autunite produced quite similar looking events but more, of course.

Well, as has been brought up by Martin here recently, it may have been muon events. What I uncovered in my literature search was work at a French observatory (which if memory serves was located at about 3km ASL) and their extensive work with Si diode detectors, plus their estimates that took their own measurements and made predictions about sea level expectations. Their predictions matched well with our experiences and it helped convince others, along with some additional tests we did at the time.

We got that, as well. Sudden shifts to new plateaus. (Keep in mind we were measuring in 10s to several 100s of fA with theoretical detector shot noise limits very close to 800 aA, not taking into account the ACF2101 and I forget at which bandwidth, now.) The weird thing about the shifts were that there were several quite specific floors and the level would drop back pretty much exactly onto a prior one. It was not that it just went to some random level, but that it went exactly to some prior level. There were a few "sticky" places it would go to. But with timing that appeared to be 'at random.' This was NOT the spikes, though. They rode on top of the shifted levels and were typically within about one order of magnitude of each other, roughly 10,000 electrons' worth as a mean, if memory serves (20 years and more.)

The ACF2101 is an integrator. Take a look at the data sheet. The FET switches are carefully balanced so that charge injection is nicely balanced, on/off. It's got a bit of a nasty offset voltage, which is partly why we wanted to run this cooled. The ice bath was cheap and quick to do and support, though we talked about doing something even cooler. The cooling both stabilized temperature effects on the ACFC2101 and did lower noise slightly. But the main thing we gained was a very much higher dV/dI for the Hamamatsu detector we were using -- very non-linear rise with cooling, so it greatly reduced the dark currents produced even for a modest 20C reduction in temp.

Well, my brain hasn't completely gone to mush, then. It seemed as though that would be important.

But they'd need to be swept (collected) so we are talking mostly about the depletion region here, yes? (A very few outliers get collected, right?)

Laser dyes are (or were) used by astronomers to gain some down shift of wavelength into the better part of the quantum sensitivity for a particular Si diode, is my recollection for UV stuff. Also, if I remember, sometimes a very thin gold layer might be used on the front side to "drive" electrons produced by shorter wavelengths nearer the surface into the depletion zone to get collected. Forget what that was called, now. Like I mentioned elsewhere, using 3-He and other methods are probably "similar in basic concept" schemes used for radiation detection. To me, it's all new so I enjoy thinking about this, but to folks who work in this area I'm sure it is all very old-hat and boring.

I have quite a collection of Hamamatsu detectors that I keep "on file." But none of them are PIN, as I never needed (nor wanted) to use back-bias. Everything I did was designed to keep leakage to a dead minimum at room temps, which meant looking especially at opamp offset voltages, ways to jack up and calibrate the other side of an opamp to zero that out, and looking at the diode impedances (smaller is better, but then optics impinges.)

Thin film filtered, near-visible and visible pyrometry at

150C requires dealing with tiny currents. Letting it work all the way up to 1850C (sapphire light pipe) and more (focused camera) means eight orders of magnitude. But once you get enough temperature going, you are just shot noise limited and things get lots easier. (Until the flux gets to the point where the diode heats up and starts leaking more.) It was mostly the low end and the dynamic range that hurt so much. What was fun was in reaching the low end, began to "see" infrequent muon or cosmic ray events.

Jon