Leaders in professional radon instrumentation and analysis solutions    


Short Term Radon Monitoring: Uncertainty Analysis


The following article provides an explanation of the suitability of the RAD7 for short term radon tests.


While we have studied the (highly comprehensive) RAD7 User's Manual, I am nonetheless minded to ask you the specific question of whether you believe the (highly impressive) RAD7 instrument is suitable to sniff or short term monitor the presence of Radon in private homes here in Ireland on a purely commercial basis

To be more specific, if it were possible for you to clarify to me the time frame involved to take a RAD7 into an area with, say, three homes side by side; starting at home A, perform a Radon sniff, move to home B, perform a second sniff and likewise to home C for a third sniff - given the reset time required for the RAD7 each time plus your recommendation to conduct multiple cycles for improved accuracy in each Radon test?

I appreciate that, if testing time and money were of no issue, then my question wouldn't be of any relevance and the RAD7 would fit the bill here perfectly given its comprehensive abilities, however, given the overall economic environment here in Europe, we are pessimistic of the home owners ability or desire to pay significantly for the proposed service and therefore we are looking for an estimate on how long we should anticipate having to spend in each home in order to offer value for money combined with an accurate reading each time. Needing to dwell for excess periods of time in each house to ensure proper, semi accurate sniff results will make this an uneconomical venture for us to consider given the high staff costs we must endure - is it unrealistic for us to consider 20 minutes total per home or is 30 minutes (or longer) more appropriate?

Again, we appreciate that the more cycles in each test the more accurate the result - here, I am merely trying to establish the bare minimum time required to give an adequate result, not a comprehensive one...and hope you'll not consider me a heathen for the blatant commercial angle I am proposing with my questioning; we have a strong corporate infrastructure in place, sufficient funding and the basis of a solid business plan to create a profitable enterprise around one or more RAD7s if the test times they offer run in our favor. I believe it to be a fortuitous opportunity and your expertise in answering this small but vital piece of information is all we're missing.


I fully understand your perspective and will endeavor to address it.

1. The EPA requires a minimum 48 hours for a house test.  This is predicated partly by the poor sensitivity of charcoal test devices, that were mostly used for such tests, and partly to average out the diurnal variation that can occur, especially if the house goes from being colder than outside during the day to warmer than outside during the night.  However, if you can ignore that, and take a spot reading covering a short time period, then:

2. For a device to satisfy the EPA, the accuracy has to be no worse than +/-25%.  Let us, therefore, require that our RAD7 reading have a one-sigma statistical uncertainty no worse than +/-25% at the threshold radon concentration, above which action is recommended and below which there is no government recommendation to take action (200 Bq/m3).

3. The EPA standard requires that the house be in "closed house condition" from 12 hours before the test starts until after its completion.  This condition ensures that all houses are tested under the same conditions and also makes a radon problem more likely to manifest itself if it exists.  The location of the test should be in the lowest livable space - normally a basement.

4. The British HPA, in direct contrast, require long-term, alpha-track measurements in normal living space under normal living conditions, which is fine as a study of family exposure but pretty useless, in my opinion, as a reliable test of a house.  One family may always have the windows open, and the house therefore tests OK, then a new family moves in who closes all the windows and lives in a high radon house with a certificate saying it is OK!

5. For highest consistency I suggest that you contact the three houses you are going to test and ask them to keep all the doors and windows closed, and to not use air conditioners, from the previous evening through the night and through the day that you conduct the tests.  I also suggest that you follow the EPA guidelines and test the level in the lowest livable space.

6. For a one-sigma uncertainty of 25% you need 16 counts (Poisson statistics: sigma = SQR(N))

7. A typical RAD7 has a Sniff sensitivity of about 0.68 cpm for a radon concentration of 100 Bq/m3.  200 Bq/m3 will generate 1.36 cpm.  16 counts will be obtained, therefore, in 16/1.36 = 12 minutes.

8. It takes the RAD7 about 12 minutes, in SNIFF mode, for the count rate in window A to approach equilibrium.  24 minutes, therefore, is the absolute minimum to get a reading that would have a one-sigma uncertainty of +/-25% if the radon concentration was 200 Bq/m3.  You could get that by choosing a cycle time of 12 minutes, starting a measurement in the house, waiting until the second cycle was complete and using that second reading as your measurement.

9. I would, however, prefer a slightly longer measurement and suggest that you a) set the protocol to Sniff (Setup, Protocol, Sniff [ENTER]) and then change the cycle time to 10 minutes (Setup, Cycle, 00:10).  Now start a test and let it continue for 30 minutes, or 3 cycles.  Ignore the first cycle and take the average of the second and third for your measurement.  This will give you a precision slightly better than 25% at 200 Bq/m3.  If you have the time to let it run for more than three cycles the precision will improve significantly.

10. Let the measurement continue while you walk to the next house.  This will purge the RAD7 of the air from the previous house so that you start the next house with a RAD7 full of outside air.  The count rate in window A, once the radon in the chamber is removed, will decay with a half life of 3.05 minutes.  That means that it will have dropped to 10% of what it was in the first house by the time you reach the second house, say ten minutes later.  You are then going to start a new measurement and ignore the first 10-minute cycle.  So even if you stop the purge and start the new measurement straight away, the previous house's count rate in window A at the end of the first cycle will only be 1% of what it was in the previous house and will make no contribution to the subsequent cycles in the new house.  Considering that we are looking at uncertainties of 25%, you can completely ignore the residual count rate from the first house and start your next house measurement immediately you get there.  You should, however, remember to purge the house air out of the RAD7 while you are walking from one house to the other, either by using the Purge command or just continuing the measurement you started in the first house.

11. The above analysis shows that you can get better than 25% 1-sigma precision measurements, at 200 Bq/m3, with a 30-minute run time and that the recovery from a high reading in one house will take less time than it will take you to walk to the next one, so that within these constraints the time spent for each house will be 30 minutes plus walking time plus setup and take-down time plus time spent chatting with house owner about result.  You should be able to do one house per hour with that precision.

12. If you were able to use a one-hour run in each house, or six 10-minute cycles, then ignoring the first cycle, at 200 Bq/m3, you would have a total of 50 * 1.36 = 68 counts.  Sigma = SQR(68) = 8.25, or 12%.  This is twice as precise as your 30-minute reading, and much better than a typical charcoal test reading.  Depending on the separation of the houses, it may take you four hours to test all three with this precision.


Please remember that taking a spot reading does not satisfy any government protocol. Nor does it attempt to average out diurnal or longer term variations. In our practical experience, however, and from reports back from Home Inspectors in the USA using this technique, we know that invariably, when a spot reading like this is followed by a 48-hour, EPA-protocol test, the results are the same within the uncertainties of the readings. A spot test has the distinct advantage that it is almost certainly free of tampering by anyone who wants the reading to be low!

Of more significance than short-term variation is the difference between summer and winter measurements.

As a working principle, I would suggest that, after making a spot measurement such as described above:

a) If the concentration exceeds 300 Bq/m3 you recommend immediate mitigation. There is no need for further testing. The house undoubtedly has a radon problem.

b) If the concentration lies between 100 Bq/m3 and 300 Bq/m3 you recommend further measurement - I would suggest a 48-hour test done under EPA protocol, using either a charcoal-based test device (not alpha-track), or the RAD7.

c) If the measurement was made in the summer and the result is less than 200 Bq/m3, you recommend to the client that the test be repeated in the winter, when the heating is on.

Related Products



Price List (HTML)

Products Accessories Services Software Support More
RAD7 Radon Detector
Big Bottle System
Water Probe
Soil Gas Probes
Range Extender
Emission Chambers
CO2 Logger
Cool Trap
Duty Cycle Controller
Indoor Faucet Connector
Natural Rock Sample
Relay Box
Serial Bluetooth Adaptor
USB to Serial Adaptor
Water Switch
RAD7 Calibration
Thoron Calibration
RAD7 Repair
Firmware Upgrades
Software Drivers
Product Manuals
Support Contacts
Video Tutorials
Research Papers
External Radon Links
Company Information
Customer Testimonials
Contact Us