Evaluation of consumer digital radon measurement devices: a comparative analysis

Kansas State University (KSU) Engineering Extension conducted an abridged evaluation of eight consumer grade digital radon monitors. Using the KSU secondary radon chamber, these devices were exposed to three different radon concentrations for 7 d in average household temperature and relative humidity conditions. The three different radon concentration ranges used were: 12.8 pCi L−1 to 15.5 pCi L−1 (473.6 Bq m−3–573.5 Bq m−3), 27.7 pCi L−1 to 29.4 pCi L−1 (1024.9–10 857.8 Bq m−3), and ambient room level average radon concentration of 0.6 pCi L−1 (22.2 Bq m−3). The American National Standards Institute/American Academy of Radon Scientists and Technologists Performance Specifications for Instrumentation Systems Designed to Measure Radon Gas in Air (ANSI/AARST MS-PC) (ANSI/AARST MS-PC 2022 Performance Specifications for Instrumentation Systems Designed to Measure Radon Gas in Air (AARST Radon Standards)) minimum performance metrics were used to evaluate the accuracy and precision of each model type for each radon concentration tested. The eight different device models performed within the 0 ± 25% requirement for the individual percent error (IPE) for radon concentrations between 27.7 pCi L−1 and 29.4 pCi L−1 (1024.9–10 857.8 Bq m−3). For radon concentrations between 12.8 pCi L−1 and 15.5 pCi L−1 (444–592 Bq m−3) seven of the eight monitors fell within the IPE requirement and for ambient room radon concentrations six of the eight monitors fell within the IPE requirement for the ANSI/AARST MS-PC minimum performance requirement (ANSI/AARST MS-PC 2022 Performance Specifications for Instrumentation Systems Designed to Measure Radon Gas in Air (AARST Radon Standards)) ranges. All eight device models fell within the ± 15% ANSI/AARST MS-PC minimum performance requirement (ANSI/AARST MS-PC 2022 Performance Specifications for Instrumentation Systems Designed to Measure Radon Gas in Air (AARST Radon Standards)) coefficient of variation (CV) range for radon concentrations between 12.8 pCi L−1 and 15.5 pCi L−1 (444–592 Bq m−3) and for radon concentrations between 27.7 pCi L−1 and 29.4 pCi L−1 (1024.9–10 857.8 Bq m−3). In the future, evaluating the performance of these models over time to observe their long term accuracy and precision is anticipated.


Introduction
Radon is a radioactive noble gas that has long been studied by scientists all over the world.Since the 1940s, the association between radon and lung cancer has been examined and today radon exposure is considered to be the second-leading cause of lung cancer [1].Cancers of the human respiratory system represent about 20% of cancer mortality and are the fifth leading cause of death around the world [2].Additionally, the number of lung cancer deaths have increased by 57% between 1990 and 2013 [3].Radon exposure is estimated to cause up to 14% of all lung cancers [4] and up to more than 6% of all cancer mortality [3], depending on country.Studies have been conducted to better understand the variation in radon exposure-attributable mortality around the world [3].Initially, radon exposure was only known to be an occupational hazard.It was not until 1984, when an event at the Limerick nuclear power plant, started a series of events that resulted in the creation of the 1988 US EPA Indoor Radon Abatement Act [5].This Act aimed at setting limits and standards to reduce public exposure to this naturally occurring gas.
The Indoor Radon Abatement Act was the start of regulations and standards being put into place in regards to how homes are tested and mitigated for radon in the United States and authorized the US EPA to develop model construction standards and techniques [5].The US EPA published the 'Home Buyer's and Seller's Guide to Radon' [6] and 'A Citizen's Guide to Radon' [7] as resources for homeowners on how to test their homes for radon.The US EPA currently recommends home mitigation if the radon test results indicate radon concentrations of 4 pCi L −1 (148 Bq m −3 ) or greater.As more homeowners began testing and mitigating their homes, manufacturers found a market for consumer grade digital radon monitors.
Currently, if a homeowner uses an at home approved test kit, they must send the exposed test kit to a laboratory and wait for results.This process usually takes 2 weeks for a short term test kit.A long term test kit must be exposed for a minimum of 91 d and then takes approximately a week for the lab to receive and analyze the test kit.The US EPA Radon Hotline provides daily technical assistance to the US public on radon related questions.Homeowners frequently call wanting radon test results in real time to support a mitigation decision or to indicate whether a previously installed radon mitigation system should be monitored.Interest in consumer digital radon monitors is increasing, and the devices are readily available online and at local hardware stores, however, these devices have not undergone any of the necessary testing to become a National Radon Proficiency Program (NRPP) approved testing device.Thus, it is important to investigate if these devices are reporting accurate and precise data so that homeowners are making appropriate mitigation decisions.
Consumer grade digital radon monitors are devices that have not been subjected to device performance analysis, which includes approval by a qualified local jurisdiction [8] or through a national certification or listing program and are not approved for making mitigation decisions.The NRPP uses the ANSI/AARST 'MS-PC Performance Specification for Instrumentation Systems Designs to Measure Radon Gas in Air' [9] to approve devices that are used for making mitigation decisions in homes.Approved devices include activated charcoal, liquid scintillators, electret ion chambers, or a variety of continuous radon monitors (CRMs) to be used by a certified measurement professional [10].There are two types of consumer digital radon monitors: electronic integrating devices (EID) and CRMs.An EID is an electronic radon device that displays one result for the time period specified [9].A CRM is an electronic radon device that records retrievable radon concentrations over time intervals of 1 h or less [9].Some of these devices are battery operated and others require power from an electrical outlet.
The ANSI/AARST MS-PC minimum performance requirement [9] defines six different areas for performance and testing criteria: accuracy and precision, minimum detectable concentration or integrated concentration, proportionality, temperature, humidity, and compliance.For each criterion being tested, a set of at least five representative devices of the same model are tested and evaluated.A device model must pass all portions of the performance and testing criteria listed in this standard to be listed as an NRPP approved device [9].The accuracy of the device is a measure of how closely the measured radon concentration is to the true radon concentration of the environment [11].Precision is how close the measurements of the same device model are to each other for the same exposure.
Consumer digital radon monitors have been on the market for over five years [12], but very little device performance analysis has been performed on them.Gunning et al varied the measurement period to determine if shorter measurement periods than the previously recommended 3 months could be used [13].Additionally, Fuente et al examined the response time and accuracy of a singular device [14].Both of these previous works led to the study performed by Warkentin et al in Canada in 2020 on six different consumer digital radon monitors: Safety Siren Pro Series 3, Airthings Corentium Home, Airthings 2900 Wave, Airthings Wave Plus, Radon Eye Plus, and Radon Eye RD200 [15].The authors characterized measurement error for each device model at radon concentrations near the Canadian guidance level for temperatures consistent with their winter and summer seasons.Warkentin et al also analyzed the measurement error of these six device models at twice the Canadian guidance level for temperatures similar to their winter months and five times the Canadian guidance level for temperatures similar to winter months.Using the measurement error, they then ranked the six devices on a scale from A to E with a performance grade of A having a measurement error of less than or equal to ten percent and a performance grade of E having a measurement error greater than forty percent.
The purpose of the present study is to assess the accuracy and precision of eight models of consumer grade radon monitors at three different radon exposure levels.The results were calculated and plotted and can be used by homeowners to choose a model best suited for their home.Homeowners choose to test their home for a variety of reasons and ability to access the appropriate technology needed for using some of these devices varies.Therefore, this study will not use an A to E ranking system and will instead present the results and device specifications.

Method
The Kansas State University (KSU) Radon Chamber (figure 1) is an American Association of Radon Scientists and Technologists (AARST) NRPP approved secondary test chamber (certificate number 1006-SC).It is a four-glove chamber with approximate dimensions of 58 in × 20 in × 24 in and is located at KSU in Manhattan, Kansas.It is used for spike services, device performance testing, and calibrations of Ecosense Radon Eye Pro devices.Qualified measurement professionals use spikes to maintain their quality assurance plans, device performance tests to obtain or renew their radon measurement certification, and calibrations to keep their devices accurate and reliable.
The KSU radon chamber uses a NIST traceable Pylon RN-1025 flowthrough radioactive gas source.This source contains dry radium ( 226 Ra) that provides calibrated quantities of radon ( 222 Rn) for use in the chamber [16].The RN-1025 source has an activity of 107.1 kBq (15 January 1996).The radon gas is distributed throughout the chamber using an Axial 1238 small fan that runs continuously with an output of 110 CFM.The relative humidity, temperature, and barometric pressure are recorded hourly using an Extech SD700 barometric pressure/humidity/temperature datalogger.
The average radon concentration for the KSU Radon Chamber is monitored using two Pylon AB-5 portable radiation monitors with LCA-2 Lucas cell adaptors.One of the Pylon AB-5 units records the number of counts over a 5 h period while the other Pylon AB-5 unit records the hourly radon concentration.This equipment is calibrated yearly at Bowser-Morner radon chamber in Dayton, Ohio.In addition, at least one grab sample is taken during each exposure.The grab sample is a sample measurement taken from the chamber using a pump and either a Pylon 300 A active cell or a Rocky Mountain Active Cell during the exposure period.The cell is connected to two ports on the back of the radon chamber and the cell is filled with gas from the chamber for 10 min.The measurement of the grab sample is made 4 h after the sample is taken to allow the sample to reach secular equilibrium [17].A yearly inter-comparison is performed with these cells through Bowser-Morner radon reference chamber to ensure accurate and reliable data.The time the grab sample is taken is recorded and the radon concentration from the hourly pylon data for the hours before, during, and after the grab are taken is collected.The percent difference between the grab sample and the hourly pylon data is calculated to ensure accuracy of results.Using the monitored values from the two Pylon AB-5 monitors and the grab sample the average radon concentration is calculated.
Using guidance from previous work [15] and a survey of devices available for purchase online in the United States, the following eight device models were chosen for the study: Detector A (name withheld by request from manufacturer), Ecosense EcoQube, Ecosense EcoBlu, Ecosense RadonEye, SunRadon Lüft, Airthings View Radon, Airthings Wave Radon, and Airthings Corentium Home.Five units of each device model type were purchased.All devices were purchased online through the manufacturer, where available, or through Amazon.These devices were purchased without notifying the manufacturer of the reason for purchase to ensure that the units received would be similar to those available to the average homeowner.Table 1 summarizes device model specifications.Additionally, two Extech Instrument Model SD700 barometric pressure/humidity/temperature dataloggers were purchased to monitor the exposure environments.
Once the units were received, they were set up according to manufacturer-provided instructions.All of the units were given the necessary time to go through automated internal calibration procedures before the devices were used to gather data.This automated internal calibration is programmed by the manufacturer and automatically begins during setup of the devices.The units were accessed using the manufacturer-provided software, where applicable.Every device model included a set-up guide within the box that instructed the homeowner on how to download and use the software for the device.The Lüft software was not used due to wireless network protocol issues; instead, the oneradon cloud [26] was used to collect and download the data.
Device models were tested at three different radon concentrations as shown in table 2. The first exposure was performed in the KSU radon chamber with a radon concentration of 12.8 pCi L −1 (473.6 Bq m −3 ) to 15.5 pCi L −1 (573.5 Bq m −3 ) for 7 d.This concentration range was chosen from the accuracy & precision criteria from the ANSI/AARST MS-PC minimum performance requirement [9].The standard states: 'The test for accuracy and precision shall be conducted using the following procedures: a set of at least five representative devices is placed in a standard test atmosphere for radon (STAR) for the minimum period of time recommended for use by the provider at a radon concentration in the range of 6-15 pCi L −1 , a temperature in the range of 65-75 • F, and a relative humidity in the range of 10%-55%.Each device shall demonstrate an individual percent error (IPE) within 0 ± 25%.The precision of the devices shall be assessed using the coefficient of variation (CV) of the set of five devices which shall be less than or equal to 15%.' [9].
Due to the nature of counting statistics, it is not anticipated for the measured and true values to be the same and the standard allows for a range of acceptable IPE and CV.There is a limited number of outlets and shelf space within the chamber, therefore, this exposure was performed in three different runs as shown in table 2. For all three runs, the exposure duration was less than the minimum duration recommended by Airthings for the View Radon and Wave Radon device models due to time constraints, chamber availability, and funding.The Airthings manufacturer recommends a 30 d exposure duration for their devices.Instead, a 7 d exposure period was chosen for all three exposures based on previous work [15].The temperature, relative humidity, and barometric pressure were kept as constant as possible and these environment conditions were collected using the Extech SD700 Instrument.
The second exposure was performed at the ambient, room level average radon concentration of 0.6 pCi L −1 (22.2 Bq m −3 ).All 40 devices were placed in a room and the test period was started.The devices were allowed to equilibrate in the room for a minimum of 12 h after the first exposure was conducted before the second exposure was started.The exposure duration was 7 d and the temperature, relative humidity, and barometric pressures were kept as constant as possible and these environment conditions were collected using the Extech SD700 Instrument.To monitor the ambient air radon concentration, two Radon Eye Pro (REP) device units were used.The REP is a NRPP approved device.The REP data were collected and the concentrations indicated by the two device units were averaged to calculate the IPE and CV for each device unit and model.
For the last exposure, the radon concentration was chosen based on previous work [15] and the proportionality component of the ANSI/AARST MS-PC minimum performance requirement [9].The proportionality criteria of the ANSI/AARST MS-PC minimum performance requirement states that 'radon concentration shall be in the range of 30-60 pCi L −1 , or at least three times the low concentration range, whichever is greater' [9].Additionally, the precision and accuracy for Detector A was listed for a radon concentration of 25 pCi L −1 .For these reasons, the radon concentration tested for the third exposure was 25-30 pCi L −1 (925-1110 Bq m −3 ).The exposure was conducted in a similar manner to exposure 1 with three separate chamber runs performed.The exposure duration was 7 d and the temperature, relative humidity, and barometric pressures were kept as constant as possible and these environment conditions were collected using the Extech SD700 Instrument.The IPE and CV were calculated for each device and model.To calculate the IPE and CV for each device unit and device model, guidance from ANSI/AARST MS-PC minimum performance requirement [9] was used.The IPE is defined as the 'degree from which a single measured value (X) deviates from the conventionally true value (T)' [9] and it measures the accuracy of the devices, The CV is defined as the sample standard deviation (s) of a set of measurements expressed as a percentage of the arithmetic mean of the measurements [9] and it measures the precision of the devices, CV = 100 ( s mean ) .

Results
The chamber conditions for the first exposure are listed in table 2. Two of the three radon concentrations achieved were within the 6-15 pCi L −1 ANSI/AARST MS-PC minimum performance [9] accuracy and precision range requirement, with the third chamber radon concentration being just outside the upper limit of the range.The IPE and CV were calculated for all eight device models and are listed in table 3. The IPE and CV for each consumer digital radon device was plotted and the results are shown in figures 2 and 3, respectively.Seven of the eight device models satisfied the ±25% criteria for the accuracy component of the ANSI/AARST MS-PC minimum performance requirement [9], which states that all five devices for each device model must fall within the approved IPE range.The Airthings view radon was the only device model that did not meet the ANSI/AARST MS-PC minimum performance requirement [9] IPE criteria, with four of the five devices falling outside of this range.All eight device models passed the precision criteria (CV less than 15%) as shown in figure 3.  The exposure conditions for exposure 2 are listed in table 2 and the results can be found in table 3.Each of the eight device models was compared against two radon eye pros that ran simultaneously, and the IPE and CV were calculated and plotted and are shown in figures 3 and 4, respectively.In general, the CV of the detector should increase as the radon concentration decreases due to the statistical nature of radiation counting and the central limit theorem.For exposure 2, the precision worsened for six of the eight device models.Additionally, seven of the eight device models demonstrated an average IPE within ±25%.It is anticipated that the absolute value of IPE will decrease as the radon concentration approaches the device model calibration range.Seven of the eight device models have a reported specifications for accuracy and precision near the US EPA action level of 4 pCi L −1 (148 Bq m −3 ) only Detector A has reported specifications for accuracy and precision at 25 pCi L −1 (925 Bq m −3 ).The ambient air radon concentration average was 0.6 pCi L −1 (22.2 Bq m −3 ), well below the lower limit of the assumed calibration range, and therefore it is anticipated that the accuracy will suffer.
Conditions for exposure 3 are listed in table 2 and the results are listed in table 3. The radon concentration goal for the third exposure was 25-30 pCi L −1 (925-1110 Bq m −3 ) which was achieved for all three rounds.The results of the IPE and the CV are plotted in figures 3 and 5, respectively.Due to the statistical nature of radiation detection and the central limit theorem, it is anticipated that the precision will improve as the radon concentration increases.This behavior was observed for four of the eight device models, with two other device models demonstrating precision similar to that observed in Exposures 1 and 2. The IPE fell within the 0 ± 25% for all of the device models.

Discussion
Previous work [15] characterized the performance of six different device models available at the time of that study.When choosing the device models for this study, similar brands (Ecosense and Airthings) and models (RadonEye, Wave Radon, Corentium Home) were chosen where available for comparison.Some of the devices used in the previous study [15] are no longer available on the market or have upgraded to a newer model series.The device Warkentin et al found that performed the best of all the models tested was the Airthings Corentium Home.The Corentium Home fell within the 0 ± 25% IPE and 0 ± 15% CV ranges for all three exposures in this study, as well.
Warkentin et al found that the Safety Siren Pro Series 3 and the radon eye performed slightly worse than the Airthings Corentium Home [15].In comparison, this study found that the Ecosense Radon Eye performed similarly to the Airthings Corentium Home, with the IPE falling within the range of 0 ± 25% and the CV falling within the range of 0 ± 15% for all three exposures.Warkentin et al gave the lowest ratings to the Airthings Wave and Airthings Wave Plus.The Airthings Wave Plus was not available for purchase at the time of this study, so the Airthings Wave Radon and the Airthings View Radon were tested.This study found the Airthings Wave Radon IPE fell within the range of 0 ± 25% and the CV fell within the range of 0 ± 15% for two of the three exposures.Fuente et al [14] also looked at the accuracy of the Wave, but only examined a singular device while this study looked at five devices of each model, therefore a comparison of results was not performed.The Airthings View Radon performed similarly, with the IPE falling within the range of 0 ± 25% for two of the three exposures and the CV falling within the range of 0 ± 15% for all three exposures.
In this study, we used testing parameters similar to those in previous work [15].We examined eight newer device models, characterizing accuracy and precision at radon concentrations similar to those used in previous work [15].Lastly, we kept the temperature, relative humidity, and barometric pressure consistent with average living conditions.
This study aimed to examine the accuracy and precision of the eight different consumer device radon monitors.It was anticipated that the precision of the device models would increase as the radon concentration increases.For seven of the eight device models, the precision either remained the same or improved as the radon concentration increased.The Airthings View Radon was the only device model for which the CV increased substantially (approximately 3%) when the radon concentration increased, as shown in table 3. The accuracy of the device models was anticipated to be best near the respective calibration concentrations.Although detailed information on calibration procedures is not available, all of the devices  fell within the IPE range of 0 ± 25% for at least two of the exposures and seven of the eight device models fell within the IPE range of 0 ± 25% for exposure 1.

Conclusion
The purpose of this study was to characterize the accuracy and precision of eight different consumer digital radon device models available to homeowners.These device models are not evaluated by the ANSI/AARST MS-PC minimum performance requirement [9] standard and are not required to be calibrated or undergo any routine maintenance to assess their continued performance.The device models were exposed to three different radon concentrations in typical household environment conditions.The exposure duration for each radon concentration was 7 d.The IPE and CV were calculated for the different device models at different exposure conditions.Overall, the eight different device models performed within the 0 ± 25% for the IPE for radon concentrations between 27.7 pCi L −1 (1024.9Bq m −3 ) and 29.4 pCi L −1 (1087.8Bq m −3 ).For radon concentrations between 12.8 pCi L −1 (473.6 Bq m −3 ) and 15.5 pCi L −1 (573.5 Bq m −3 ), seven of the eight monitors fell within the IPE ANSI/AARST MS-PC minimum performance requirement [9] ranges.The Airthings View Radon was the only device that did not meet the ±25% IPE range with four of the five device units falling outside of these limits for radon concentrations between 12.8 pCi L −1 (473.6 Bq m −3 ) and 15.5 pCi L −1 (573.5 Bq m −3 ).For ambient room radon concentrations detector A fell outside the ±25% IPE range for four of the five devices as well and Airthings Wave Radon had one device fall outside the ±25% IPE range.All eight device models fell within the ± 15% ANSI/AARST MS-PC minimum performance requirement [9] CV range for radon concentrations between 12.8 pCi L −1 (473.6 Bq m −3 ) and 15.5 pCi L −1 (573.5 Bq m −3 ) for radon concentrations between 27.7 pCi L −1 (1024.9Bq m −3 ) and 29.4 pCi L −1 (1087.8Bq m −3 ).It was anticipated that the precision would worsen as the radon concentration decreased.For the ambient room radon concentration, only three of the eight device models did not fall within the ±15% CV range and two of the eight device models did not fall within the ±25% IPE range.This study could be improved by allowing more time between chamber runs to allow for the devices to have adequate ramp up and ramp down time.Although we allowed for 12 h between Exposures 1 and 2, there was still a significant amount of ramp down time for a few of the detectors.It took 24-36 h for the detectors readings to come to equilibrium with the air they were being exposed to.Additionally, this study could be improved by conducting each exposure for 30 d or more as per the manufacturer recommendations.Future work will characterize performance of the detectors over longer periods of time to better understand how homeowners can expect these consumer-grade device models to perform during their operational lifetimes.
There are many consumer grade radon detectors available in the United States to homeowners.This study indicates that the consumer grade devices evaluated mostly meet the accuracy and precision requirement applied to professional grade devices when new.However, consumers are likely to use these devices for long periods of time.At present, we do not have data to characterize how these devices perform over their usable lifetimes.Additionally, user error is more likely with consumer grade devices, since users are not trained on operation.Users should follow proper measurement testing protocols and device specifications to ensure they are collecting reliable data.These monitors can be useful for continuous monitoring of radon reduction systems that have already been installed in homes, but devices that have been approved by qualified local jurisdiction [8] or through a national certification or listing program should still be used every two years for system performance diagnostics.Although there were a few device models that did not perform within the required accuracy range, they provide homeowners a general understanding of their daily exposure to radon.Even with these consumer radon device monitors producing fairly accurate and reliable results, the ANSI/AARST MAH 2023 [8] standard still requires an approved device be used to make mitigation and post-mitigation decisions.

Figure 2 .
Figure 2. IPE for exposure 1.Average and standard deviation are shown for each device model.Fewer than five data points are shown for some device models due to overlapping data.

Figure 4 .
Figure 4. IPE for exposure 2. Average and standard deviation are shown for each device model.Fewer than five data points are shown for some device models due to overlapping data.

Figure 5 .
Figure 5. IPE for exposure 3. Average and standard deviation are shown for each device model.Fewer than five data points are shown for some device models due to overlapping data.

Table 1 .
Consumer digital radon monitor device model specifications.Electronic integrating device • 48 h needed before an accurate reading can be displayed with a recommendation of a longer timespan to give a more accurate year-round average radon level • Displays short-term average with a maximum of 7 d with hourly updates • Displays long-term average with a maximum of 365 d with hourly updates • Uses diffused-junction photodiode for sensor • Accuracy/precision at 25 pCi L−1

Table 1 .
(Continued.) Displays short term average that alternates between showing radon values for last day and for the last seven days • Displays long term average that represents the average radon value for the ongoing measurement • Uses alpha spectrometry for detection • Accuracy/precision at 5.4 pCi L −1 * After 7 d ∼ 10% * After 2 months ∼ 5%

Table 2 .
Environmental conditions for each exposure.

Table 3 .
Individual percent error and coefficient of variation for three radon Exposures.