Assessment of Agent Monitoring Strategies for the Blue Grass and Pueblo Chemical Agent Destruction Pilot Plants (2012) / Chapter Skim
Currently Skimming:
5 Statistical Methods and Measurement
5 Statistical Methods and Measurement
Pages 95-112
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 95... ...
For example, using some of the new analytical technologies identified in Chapter 4, there are now quick and reliable means to detect and, if appropriate calibration standards are available, to quantify agent contamination on surfaces and hidden in crevices or other occluded places in machinery or building materials. Using these methods, it is now possible to more efficiently identify and delimit local hot spots of contamination, allowing a more efficient sorting of waste streams from the deconstruction process into segregated "contaminated" and "not contaminated" streams, each of which could get appropriate handling.
|
From page 96... ...
Current airborne monitoring methods can detect any significant agent vapor concentrations. However, new surface analysis technologies may be useful in directing efficient decontamination activities, thus reducing the possibility of airborne agent contamination.
|
From page 97... ...
REVIEW OF EXISTING AGENT MEASUREMENT APPROACHES In reviewing the Assembled Chemical Weapons Alternatives (ACWA) documents, there is little detailed description of the statistical methodologies guiding current agent monitoring and measurement methods.
|
From page 98... ...
can yield a wide range of possible detection limits based solely on what a particular analyst expects the result (i.e., the MDL)
|
From page 99... ...
Furthermore, using this statistical methodology, estimates of the detection and quantitation limits can be obtained. The detection limit describes the point at which the analytical method allows a binary decision on whether the analyte is present in the sample, and the quantitation limit describes the concentration at which the signal-to-noise ratio is 10:1 (i.e., a relative standard deviation of 10 percent)
|
From page 100... ...
This analysis reveals that the binary detection decision for VX in deionized water can be made at 2.7 ppb with 99 percent confidence, with a detection limit of 5.4 ppb. Quantification (i.e., the concentration at which the signal-to-noise ratio is 10:1)
|
From page 101... ...
SOURCE: Constructed from data provided by H Dupont Durst, Edgewood Chemical Biological Center, CMA.
|
From page 102... ...
The use DART, DESI, or related new analytical methodologies for surface area measurements at the Pueblo Chemical Agent Destruction Pilot Plant or the Blue Grass Chemical Agent Destruction Pilot Plant requires that the quality of measurements be determined and related calibration studies be performed for relevant matrices.
|
From page 103... ...
The development of an internal standard may or may not be practical for some surface analysis applications at the Pueblo Chemical Agent Destruction Pilot Plant or the Blue Grass Chemical Agent Destruction Pilot Plant. In the absence of an internal standard, the precision of the quantitative measurements may decrease.
|
From page 104... ...
STATISTICAL SAMPLING ISSUES Measurement Bias, Precision, and Detection Limits The "Analytical Measurements Issues" section in this chapter describes statistical approaches to understanding the information about a measurand (the quantity of interest) that can be inferred from one or more measurements.
|
From page 105... ...
However, where the challenge is to identify the location of quantities of adsorbed, absorbed, or trapped condensed-phase agent for decontamination, air concentration values are indirect and spatially indistinct indicators of the quantities of greatest interest. Conversely, while surface measurement techniques may accurately characterize the degree of contamination at a spatial point, such measurements are not efficient as a basis for screening a larger area or volume for contamination due to the number of such measurements that would be 2 In statistical modeling, measurement data that may take on numerical values but may also result only in an indication that a threshold value has been exceeded (e.g., "below detection limits" or "above saturation level")
|
From page 106... ...
If direct surface and/or materials wipe sampling analysis methods are adopted, appropriate statistical methods for characterizing the extent of contamination of surfaces, machinery, and/or materials should be employed. Spatial Modeling In simple analytical applications where the measurement device is known to be well calibrated, it may be sufficient to limit analysis of measurement variation to random measurement error, considering only one measurand at a time.
|
From page 107... ...
As a result, there will be uncertainty stemming from spatial variabilitythe actual variation of the measurand across the area being studiedin addition to the uncertainty associated with measurement imprecision at any point. Specification of the nature of spatial variability is a critical step in deciding how samples should be collected, and how the resulting data should be analyzed for the purpose of monitoring or characterizing agent concentration in an area.
|
From page 108... ...
However, within any specific application and for almost any realistic characterization goal, most reasonable sampling plans include: Measurements taken at a sufficient number of locations to provide reliable characterization of the spatial variation of the measurand across the region of interest and A sufficient number of replicate measurements at some locations to determine the magnitude of measurement errors being encountered or to validate the assumed precision of the instrument. While the particular details of an efficient sampling plan must depend on specific goals and the details of the statistical model to be used, most reasonable models will lead to sampling plans with these two characteristics.
|
From page 109... ...
. Determining an optimal balance requires some a priori generator knowledge of the relative sources of uncertainty associated with spatial variation of the measurand and with measurement error and also with the type of statistical model to be used for characterizing variability.
|
From page 110... ...
Once the resulting expert sampling protocols have been developed, ACWA headquarters monitoring staff or their contractors should then proceed to develop detailed standard operating procedures to guide monitoring technicians. Hot Spot Detection Another agent deposition pattern that may be especially relevant in the ACWA context could be formulated to describe one or a few hot spots in a region that otherwise has no agent or only a very low concentration of agent.
|
From page 111... ...
For this purpose, sequential sampling plans that utilize multiple sampling bases may be most effective, as described below. Sampling Plans for Hot Spot Detection The fixed and sequential sampling plans described above can sometimes be useful in detecting isolated hot spots of agent concentration in spatial domains.
|
From page 112... ...
Arranging the order of measurements so that the elevated agent concentration is more likely to be found earlier rather than later can shorten inspection, at least in cases where a single reservoir can be tentatively assumed to be the source of contamination. Finding 5-7.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.