Ask an Expert

Ask an Expert

Question and Answer Page

Survey Instruments (conventional)

Submitted 8/16/04
Q: When calculating instrument efficiency, what is used in the following equation: Total background counts or background in CPM for the CB term??

MDC = 3 + 4.65 * Square Root of CB * Count Time/Instrument Efficiency * Surface Efficiency Factor * Count Time * Proba Area In cm2/100cm2

CB = Background counts
TB = Background counting time (min)
ei = the instrument efficiency (count per particle)
es = the contaminated surface efficiency (particle per disintegration)
WA = the area of the detector window (cm2)

A: I believe your question refers to "calculating MDC" rather than instrument efficiency. The general equation for minimum detectable concentration (MDC) is presented on p. 3-5 of NUREG-1507 (PDF File). The term C_B refers to the background count in time T. That is, C_B is the total number of counts obtained during the counting interval T. For example, let's assume that the number of background counts obtained using a gas proportional detector for one (1) minute count is 300 counts. Then C_B is 300 counts and T is 1 (one minute). Now let's assume you wanted to reduce the MDC by counting background for two minutes. In this case C_B would be 600 counts (there is 600 counts collected over the two-minute count time) and T is 2 (two minutes).

There appears to be an error in the equation you provide in your question. The square root term in the numerator should either be C_B (as discussed above) or R_B * T, where R_B is the background count rate (in cpm). In either case the square root term must be unitless (just "counts").

Hope this clarifies the issue.

Submitted 4/22/04
Q: I am designing a VSAP (FSS) for surface soil. Since Pu-238 is the major COC, I'm having difficulty developing an acceptable scanning protocol since our BAT is a FIDLER/count rate meter set up. Appendix H leads one to believe that the sensitivity is in the 10 to 20 pCi/g range. Empirical studies done on site have shown that it is closer to 250pCi/g and scan MDCs are over 400pCi/g. My DCGLemc is 165 pCi/g based on RBGV. Can you provide information or cite the correlation study used for Appendix H or provide some guidance on how to reasonably scan for Pu-238? Using area factors, I'm approaching 100 samples per 2000m2. This seems a bit excessive.

A: Pu-238 emits low energy x-ray radiation (13.6 keV), which makes it very difficult to detect, even with a FIDLER. One thought is to identify a surrogate radiation to measure, such as the Am-241 59 keV. However, this is only an option if you have other radionuclides that can serve as surrogates.

As you indicate in your question, Appendix H in MARSSIM states that the sensitivity of a FIDLER for Pu-238 is about 20 pCi/g. This does seem low, even for the specific conditions stated (i.e., contamination is limited to the top 1 mm of soil). Given that the background for the FIDLER is 400 cpm (as indicated in MARSSIM Appendix H, p. H-32), and assuming a 2-second observation interval and d' of 2.32 (25% false positives and 95% detection); the minimum detectable count rate of the surveyor is 360 cpm (from MARSSIM Ch. 6). This means that an additional 360 cpm (above background) would be needed in order for Pu-238 to be detected with 95% confidence. The more difficult question is how much Pu-238, for a specific geometry, is needed to produce a net count rate of 360 cpm. To answer this question one needs both the detector efficiency of the FIDLER for 13.6 keV x-rays, as well as modeling information (e.g., from MicroShield) that provides the x-ray fluence rate at the detector from a specified geometry (areal extent and depth) of Pu-238 in soil.

To use the data provided in Appendix H, the FIDLER sensitivity is stated to range from 500 to 700 cpm/microCi/m2. Assuming that the FIDLER responds at the high end of this range, the scan MDC can be calculated for a range of soil geometries. For example, if the areal extent of contamination 1 m x 1 m, then the scan MDC can be determined as a function of depth of Pu-238. The results indicated that an optimal scan MDC of 740 pCi/g was achieved for soil contaminated to thicknesses of 0.5 to 2 cm. Contaminated soil depths greater than a few centimeters resulted in an increased scan MDC (880 pCi/g at 5 cm, and 4700 pCi/g at 15 cm) because the increased activity concentration of Pu-238 with depth did little to add to the detectability of these low energy x-rays. Conversely, a contaminated soil thickness of only 1 mm resulted in a scan MDC of 840 pCi/g. These results agree with your empirical data that scan MDCs for Pu-238 in soil are greater than 400 pCi/g. It may well be that empirical studies provide the best estimate of the scan MDC.

Independent of the scan MDC for Pu-238, you might consider compositing as a way to reduce the number of soil samples needed to satisfy hot spot concerns. Remember that the additional measurements are used to show that there are no “hot spots” that exceed the DCGLemc. The additional measurements are not needed for the statistical tests. As an example, lets say that 18 soil samples are needed to meet the requirements of the Sign test, but that a high scan MDC for Pu-238 meant 90 samples were needed. Obviously these 72 extra analyses for Pu-238 would be costly (a bit of an understatement). Fortunately, it is only necessary to show that the levels of contamination in these 72 additional samples are not above the DCGLemc (165 pCi/g in your case).

Here's how it works. Take the 90 soil samples - and let's assume you plan to composite three samples at a time. Each of the 90 samples is split in thirds (ensure that you have collected sufficient material for analysis requirements); the remaining soil from each sample is set aside as a backup and might not need to be analyzed. Groups of three adjacent soil samples (keeping track of soil sample location for each grouping) are composited with one other (note: it is important that appropriate compositing procedures be followed). We now have 30 composites. These 30 are analyzed and the statistical test is performed on the data. Let's assume that we pass the statistical test. The next step is to ensure that none of the original 90 samples exceeded the DCGLemc of 165 pCi/g. This can be shown if the 30 composites had less than 55 pCi/g each, i.e., the DCGLemc divided by the number of samples in each composite. If one of the 30 composites exceeded this value, say one sample has 80 pCi/g, the three backups from the original samples used to produce the composite would then be analyzed to determine whether one of them actually exceeded the DCGLemc, and if so, which one(s). The net effect of this compositing example is to reduce the number of samples analyzed from 90 to 30.

Submitted 2/14/04
Q: My colleagues and I are in debate about smear counting in the field and ISO 7503. Specifically, how does the ISO 7503 approach to developing total efficiencies (instrument efficiency, X surface efficiency) apply to a Ludlum model 2929 w/43-10-1 probe? [Are] the accepted values of .5 and .25 still applied as surface efficiencies?

A: The ISO-7503 ["Evaluation of surface contamination - Part 1: Beta-emitters (maximum beta energy greater than 0.15 MeV) and alpha emitters"] does indeed address the issue of smear counting. The standard provides an equation where the net count rate from the smear count is divided by three factors: instrument efficiency, surface efficiency and removal factor. The values of surface efficiency are the same as for direct measurements of surface activity in the field (i.e., 0.25 for alpha emitters, and 0.25 and 0.5 for beta emitters, depending on the beta energy). The removal factor should either be experimentally determined, or a default value of 0.1 used.

Submitted 1/29/04
Q: When reporting stationary measurements made with various hand-held instruments, how do you reconcile significant figures? We are ending up with an assortment of numbers with 1 or 2 significant figures (9, 22, 160, 10, 20,30). Depending on the scale that the reading was made on you might very well be able to read to the ones place, while on a higher scale maybe to the tens place. I've looked through many survey reports and it looks to me like it is not uncommon to see tables of data with varying number of significant figures.

A: A widely cited reference for reporting environmental (and D&D) data is the U.S. Environmental Protection Agency, Upgrading Environmental Radiation Data, Health Physics Society Committee Report HPSR-1 (1980), EPA 520/1-80-012, August 1980. Among other things, the report states "the uncertainty should be reported to no more than two significant figures, and the value itself should be stated to the last place affected by the qualification given by the uncertainty term." This guidance is most relevant to laboratory measurements-- given that laboratory data usually have uncertainties reported along with the value.

Considering your specific question, it appears that a radiation detector is coupled to a ratemeter as the measuring instrument rather than a scaler (to integrate the counts). Therefore it is necessary to assess the uncertainty in the measurement system. A common rule of thumb is to report the value to the first estimated figure. For example, assume we are measuring elevated surface activity with a gas proportional detector coupled to a ratemeter. Our ratemeter scale goes from zero to 5 and we are on the 1000 scale. Assume the needle is fluctuating between the 2 and the 3 --i.e. between 2000 and 3000 cpm. The best we could do is estimate that second significant figure (the hundreds place). After observing the needle we estimate the count rate averages around 2400 cpm - this is the value to report.

We could change the measurement system now and improve our precision by taking a measurement with a scaler set to count for one minute. For this example let's assume the result is 2365 cpm (four significant figures). In this case all four significant figures would be recorded (on field data sheets) and used in the subsequent calculation of surface activity. If surface activity level is the ultimate quantity being reported, then this value would be rounded to the appropriate number of significant figures (usually two). The rule to apply is that when multiplying or dividing, the result is rounded to the least number of significant figures of any one term. The terms in the surface activity equation include the gross count rate, background count rate, detector probe area, and detector efficiency. Let's assume that the total efficiency factor is 0.09 and the background count rate is 350 cpm. Therefore, the surface activity is given by [(2365 - 350)/(0.09)(1.26)], or 17,769 dpm/100 cm2. The significant figures range from 2 to 4, but our efficiency as is usually the case, is our limiting term. Therefore we would only report two significant figures: 18,000 dpm/100 cm2.

Submitted 12/16/03
Q: When developing a survey and sampling plan for MARSSIM you are required to determine the theoretical MDC (LLD) for the instrumentation and methods to be used during the survey. For this calculation one needs to determine the theoretical background count rate for an ideal uncontaminated surface. Since all materials contain some quantity of radioactive material are there standard numbers that are used for the background count rates. For example, if alpha static measurements will be performed on a concrete surface, what is the "ideal" background count rate?

A: The background level used for calculating the minimum detectable concentration (MDC) for a survey instrument is usually based on actual (empirical) measurements. There are no standard numbers that should be used in lieu of actual measurements of background. For example, the background level for a concrete surface should be determined by a number of static measurements on a concrete surface in a non-impacted area. The average background value is then used in the calculation of the instrument's MDC.

Given that the background level can vary as a function of surface material construction, instrument MDCs can be calculated and reported for a number of surface types. According to the MARSSIM (p. 6-35), "MDC values should be calculated for each type of area, but it may be more economical to simply select a background value from the highest distribution expected and use this for all calculations."

Submitted 3/17/03
Q: I'm looking for a copy of NUREG 1506, "Measurement Methods for Radiological Surveys in Support of New Decommissioning Criteria", Draft 1995. I would prefer an electronic copy. I wonder if the NRC has superceded the report with a replacement. Any suggestions?

A: Draft NUREG-1506 "Measurement Methods for Radiological Surveys in Support of New Decommissioning Criteria" (1995) has not been superseded, nor finalized. This NUREG report, as stated in the abstract, "contains a description of proposed methodologies for measuring low-level radiation and radioactivity that could be used in conducting surveys associated with decommissioning of licensed NRC facilities." Perhaps most noteworthy about this report is its coverage of in situ gamma spectrometry that can be used for both indoor and outdoor survey applications. Draft NUREG-1506 will soon be electronically available on the DDSC Guidance on Selection and Use of Survey Instruments page.

Submitted 2/25/03
Q: I am designing a survey for an old Quonset Hut style building. As you may know, the majority of the building is made of heavily corrugated galvanized steel. As such, direct measurements (fixed points and scans) with typical probes (eg. L 43-20 GFP) presents some unusual challenges. I also have the added excitement of having primarily alpha emitting nuclides in my COC. I have empirically determined an effective probe area which represents the surface area within 1/4" of the probe. I'm procedurally on safe ground here, but I'm wondering if there is a better way to approach this short of painting a sample piece with Th230 and calculating a new Ei. NUREG 1507 isn't much help here. Any ideas? Are there any other probes better suited to this type of surface? I have about 12,000 sq. foot to survey.

A: Surface activity measurements performed on corrugated steel pose a challenge due to the variable source-to-detector distances. Having to measure alpha contamination only complicates the matter. NUREG-1507 might be able to offer a reasonable solution.

First, the total efficiency using the ISO-7503 approach is calculated by multiplying the instrument efficiency by the surface efficiency (0.25 for alpha). The instrument efficiency should be determined using a NIST-traceable source. However, the resultant instrument efficiency must be corrected for the source-to-detector distance. [The instrument efficiency is typically determined at contact with the source, yet the corrugated steel measurements present varying source-to-detector distances]. This is where NUREG-1507, Section 4 may help.

Table 4.6 provides the reduction in gas proportional detector response (normalized) for Th-230 (alpha emitter) as a function of distance:

Detector Response
Contact 1
0.5 cm 0.76
1 cm 0.58
2 cm 0.10

Let's assume that the corrugated steel is known to have a peak-to-valley distance of 2 cm. The detector response from alpha contamination in the valley would only be 10% of that at the top, in contact with the detector. The idea is to determine the average source-to-detector distance offered by the corrugated steel. For simplicity, let's assume that the midpoint of the peak-to-valley distance represents the average distance (1 cm in this case). So from Table 4.6 in NUREG-1507 the normalized response at this distance is 0.58. Therefore, the total efficiency for the alpha measurements would be given by (0.58)* (instrument efficiency)* (surface efficiency).

Incidentally, the other approaches mentioned in the submitted question may also work well, particularly the application of Th-230 NIST-traceable material to the corrugated steel and empirically determining the appropriate instrument efficiency.

Submitted 11/27/02
Q: Ref. source calibration of 100 (126 active) gas probe: If one only has 2 inch diameter sources (Tc99, Th230, C14), and no 100 or 126 cm2 planar sources, please suggest how they can do MARSSIM type probe calibration. Can one move the 2 inch sources around all locations that cover the active 126 cm2 area of the probe (probably 12 different locations or so to cover all probe face), and then just take average of all cpm readings and call it a 126 cm2 cpm calibration? No funding for planar sources, only have 2 inch diameter sources, what can one do?

A: Your proposed solution sounds reasonable. In fact, while ISO-7503 states that "the dimensions of the calibration source should be sufficient to cover the window of the instrument detector", it goes on to state that in rare circumstances "sequential measurements with smaller distributed sources of at least 100 cm2 active area shall be carried out. These measurements should cover the whole window area or at least representative fractions of it and shall result in an average value of instrument efficiency." The only deviation is that you plan to use a 2-inch diameter source to accomplish the averaging, rather than a source of at least 100 cm2.

The larger issue at play here is that calibrating detectors to smaller sources results in higher instrument efficiencies as compared to the efficiency obtained when calibrating to larger (distributed) sources. This difference is illustrated in NUREG-1507, Table 4.11, where the smaller source efficiencies were 6 to 42% greater than those obtained with distributed sources. Therefore, it is recommended to calibrate the detector to larger area sources since it is a time-consuming matter to determine the distribution of the contamination being measured in the field. That is, calibrating to the larger calibration source size provides a degree of conservatism in the efficiency determination when the field distribution of surface activity is unknown. Depending on the circumstances, one may choose to characterize the contamination distribution (via scanning) in an effort to know when the higher efficiency for small hot spot areas may be warranted.

Submitted 10/28/02
Q: Our company is having a debate over the application of surface efficiencies to the calculation of MDCs and activity. We understand the science behind it, and we agree with that does add costs to decommissioning projects. SO, we are afraid that our competitors are not using surface efficiencies yet but we are. We would like to know what is going on out there. Are people using this MARSSIM-based surface efficiency to calculate MDCs and activity? Is it yet required by regulators? If not, then why not?

A: Many D&D professionals are using the ISO-7503 approach for calculating total efficiencies for portable survey instruments, and it is a technical issue evaluated by the NRC. The appropriate section of MARSSIM that specifies use of ISO-7503 is section 6.5.4 Instrument Calibration. In NUREG-1727 (Sept 2000), Sec 14 on Radiation Surveys states (under Evaluation Criteria), "The final status survey design is adequate if it meets criteria in" and in the 6th bullet, "MARSSIM Sections 6.5.3 and 6.5.4 for selection of acceptable survey instruments, calibration..." The NRC's consolidated decommissioning guidance, NUREG-1757, vol. 2 (draft) contains the same guidance.

On the other hand, not everyone is using ISO-7503. Those sites that have NUREG/CR-5849 as the basis of their D&D plan are typically not using it. In general, the larger D&D sites are, such as the power reactor D&D sites. My recommendation is to get explicit approval from the regulator if you plan to deviate from the MARSSIM guidance.

Submitted 10/16/02
Q: Would you please clarify the following sentence presented in paragraph 2, page 208 of the Decommissioning Health Physics, MARSSIM user handbook: "Now the reader should be cautioned that these results are for the low gamma energy Am-241; depth of contamination might be more significant for higher energy gammas like those from Co-60."

A: The context of your question deals with the calculation of scan MDC, and specifically the impact that the depth of contamination has on the scan MDC value. Note that contamination depth refers to the contamination thickness as measured from the surface to the stated depth. The depth of contamination was varied from 12 cm to 18 cm (15 cm is the nominal value used) in this evaluation. Page 208 of DHP reports the results for the low-energy gamma radiation from Am-241: scan MDC equals 44.8 pCi/g for a depth of contamination of 12 cm, and 44.7 pCi/g for a depth of 18 cm. These results can be interpreted to mean that when the contamination is present at a greater depth (at the same concentration) the scan MDC for Am-241 is marginally better (44.7 vs. 44.8 pCi/g).

The statement is then made that the "depth of contamination might be more significant for higher energy gammas like those from Co-60." The interpretation here is that if the same concentration of Co-60 is present at a deeper depth, then more of the Co-60 gamma radiation will be contributing to the radiation level at the surface (as detected by the meter), effectively lowering the scan MDC. This was confirmed by the following evaluation of the depth impact for Co-60:

The scan MDC for Co-60 for a nominal depth of 15 cm: 5.8 pCi/g
The scan MDC for depth of 12 cm: 6.4 pCi/g
The scan MDC for depth of 18 cm: 5.4 pCi/g

Therefore, the depth of contamination, for equal concentrations, has a greater impact for higher energy gamma emitters than for low energy gamma emitters. The scan MDC is reduced as the depth of contamination is increased, at least until one approaches an 'infinite slab thickness'.


Please visit this page again soon for additional questions submitted to our D&D experts and the answers that may assist you!

Return to Q&A Page