1、Designation: D6051 15Standard Guide forComposite Sampling and Field Subsampling forEnvironmental Waste Management Activities1This standard is issued under the fixed designation D6051; the number immediately following the designation indicates the year oforiginal adoption or, in the case of revision,
2、 the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon () indicates an editorial change since the last revision or reapproval.1. Scope1.1 Compositing and subsampling are key links in the chainof sampling and analytical events that must be perf
3、ormed incompliance with project objectives and instructions to ensurethat the resulting data are representative. This guide discussesthe advantages and appropriate use of composite sampling,field procedures and techniques to mix the composite sampleand procedures to collect an unbiased and precise s
4、ubsample(s)from a larger sample. It discusses the advantages and limita-tions of using composite samples in designing sampling plansfor characterization of wastes (mainly solid) and potentiallycontaminated media. This guide assumes that an appropriatesampling device is selected to collect an unbiase
5、d sample.1.2 The guide does not address: where samples should becollected (depends on the objectives) (see Guide D6044),selection of sampling equipment, bias introduced by selectionof inappropriate sampling equipment, sample collection proce-dures or collection of a representative specimen from a sa
6、mple,or statistical interpretation of resultant data and devices de-signed to dynamically sample process waste streams. It alsodoes not provide sufficient information to statistically design anoptimized sampling plan, or determine the number of samplesto collect or calculate the optimum number of sa
7、mples tocomposite to achieve specified data quality objectives (seePractice D5792). Standard procedures for planning wastesampling activities are addressed in Guide D4687.1.3 The sample mixing and subsampling procedures de-scribed in this guide are considered inappropriate for samplesto be analyzed
8、for volatile organic compounds. Volatile organ-ics are typically lost through volatilization during samplecollection, handling, shipping and laboratory sample prepara-tion unless specialized procedures are used. The enhancedmixing described in this guide is expected to cause significantlosses of vol
9、atile constituents. Specialized procedures shouldbe used for compositing samples for determination of volatilessuch as combining directly into methanol (see Practice D4547).1.4 This standard does not purport to address all of thesafety concerns, if any, associated with its use. It is theresponsibili
10、ty of the user of this standard to establish appro-priate safety and health practices and determine the applica-bility of regulatory limitations prior to use.2. Referenced Documents2.1 ASTM Standards:2C702 Practice for Reducing Samples ofAggregate to TestingSizeD1129 Terminology Relating to WaterD44
11、39 Terminology for GeosyntheticsD4547 Guide for Sampling Waste and Soils for VolatileOrganic CompoundsD4687 Guide for General Planning of Waste SamplingD5088 Practice for Decontamination of Field EquipmentUsed at Waste SitesD5792 Practice for Generation of Environmental Data Re-lated to Waste Manage
12、ment Activities: Development ofData Quality ObjectivesD6044 Guide for Representative Sampling for Managementof Waste and Contaminated MediaE856 Definitions of Terms and Abbreviations Relating toPhysical and Chemical Characteristics of Refuse DerivedFuel (Withdrawn 2011)33. Terminology3.1 Definitions
13、:3.1.1 composite sample, na combination of two or moresamples. D11293.1.2 sample, na portion of material taken from a largerquantity for the purpose of estimating properties or composi-tion of the larger quantity. E8561This guide is under the jurisdiction of ASTM Committee D34 on WasteManagementand
14、is the direct responsibility of Subcommittee D34.01.01 on Plan-ning for Sampling.Current edition approved May 1, 2015. Published May 2015. Originallyapproved in 1996. Last previous edition approved in 2006 as D6051 96 (2006),which was withdrawn in January 2015 and reinstated in May 2015. DOI:10.1520
15、/D6051-15.2For referenced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Customer Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.3The last approved version of this historical stand
16、ard is referenced onwww.astm.org.Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States13.1.3 specimen, na specific portion of a material orlaboratory sample upon which a test is performed or which istaken for that purpose. D44393.1.4 subsam
17、ple, na portion of a sample taken for thepurpose of estimating properties or composition of the wholesample.3.1.4.1 Discussiona subsample, by definition, is also asample.4. Summary of Guide4.1 This guide describes how the collection of compositesamples, as opposed to individual samples, may be used
18、to:more precisely estimate the mean concentration of a wasteanalyte in contaminated media, reduce costs, efficiently deter-mine the absence or possible presence of a hot spot (a highlycontaminated local area), and, when coupled with retestingschemes, efficiently locate hot spots. Specific procedures
19、 formixing a sample(s) and collecting subsamples for transport toa laboratory are provided.5. Significance and Use5.1 This guide provides guidance to persons managing orresponsible for designing sampling and analytical plans fordetermining whether sample compositing may assist in moreefficiently mee
20、ting study objectives. Samples must be compos-ited properly, or useful information on contamination distribu-tion and sample variance may be lost.5.2 The procedures described for mixing samples and ob-taining a representative subsample are broadly applicable towaste sampling where it is desired to t
21、ransport a reducedamount of material to the laboratory. The mixing and subsam-pling sections provide guidance to persons preparing samplingand analytical plans and field personnel.5.3 While this guide generally focuses on solid materials,the attributes and limitations of composite sampling applyequa
22、lly to static liquid samples.6. Attributes of Composite Sampling for WasteCharacterization6.1 In general, the individual samples to be compositedshould be of the same mass; however, proportional samplingmay be appropriate in some cases depending upon the objec-tive. For example, if the objective is
23、to determine the averagedrum concentration of a contaminant, compositing equalsvolumes of waste from each drum would be appropriate. If theobjective is to determine average contaminant concentration ofthe waste contained in a group of drums, the volume of eachsample to be composited should be propor
24、tional to the amountof waste in each drum. Another example of proportionalsampling is estimating the contaminant concentration of soiloverlying an impermeable zone. Soil cores should be collectedfrom the surface to the impermeable layer, regardless of corelength.6.2 The principal advantages of sampl
25、e compositing in-clude: reduction in the variance of an estimated averageconcentration (1),4increasing the efficiency of locating/identifying hot spots (2), and reduction of sampling andanalytical costs (3). These main advantages are discussed in thefollowing paragraphs. However, a principle assumpt
26、ion neededto justify compositing is that analytical costs are high relativeto sampling costs. In general, appropriate use of samplecompositing can:6.2.1 Reduce inter-sample variance, that is, improve theprecision of the mean estimation while reducing the probabilityof making an incorrect decision,6.
27、2.2 Reduce costs for estimating a total or mean value,especially where analytical costs greatly exceed sampling costs(also may be effective when analytical capacity is a limitation),6.2.3 Efficiently determine the absence or possible presenceof hot spots or hot containers and, when combined withrete
28、sting schemes, identify hot spots, as long as the probabilityof hitting a hot spot is low,6.2.4 Be especially useful for situations, where the nature ofcontaminant distribution tends to be contiguous and non-random and the majority of analyses are “non-detects” for thecontaminant(s) of interest, and
29、6.2.5 Provide a degree of anonymity where population,rather than individual statistics are needed.6.3 Improvement in Sampling PrecisionSamples are al-ways taken to make inferences to a larger volume of material,and a set of composite samples from a heterogeneous popula-tion provides a more precise e
30、stimate of the mean than acomparable number of discrete samples. This occurs becausecompositing is a “physical process of averaging.” Averages ofsamples have greater precision than the individual samples.Likewise, a set of composite samples is always more precisethan an equal number of individual sa
31、mples. Decisions basedon a set of composite samples will, for practical purposes,always provide greater statistical confidence than for a com-parable set of individual samples.6.3.1 If an estimated precision of a mean is desired, thenmore than one composite sample is needed; a standard devia-tion ca
32、nnot be calculated from one composite sample.However, the precision of a single composite sample may beestimated when there are data to show the relationship betweenthe precision of the individual samples that comprise thecomposite sample and that of the composite sample. Theprecision (standard devi
33、ation) of the composite sample isapproximately the precision of the individual samples dividedby the square root of the number of individual samples in thecomposite.6.4 Example 1An example of how a single compositesample can be used for decision-making purposes is givenhere. Assume a regulatory limi
34、t of 1 mg/kg and a standarddeviation of 0.5 mg/kg for the individual samples. If theconcentration of a site is estimated to be around 0.6 mg/kg,how many individual samples should be composited to haverelatively high confidence that the true concentration does notexceed the regulatory limit when only
35、 one composite sample is4The boldface numbers in parentheses refer to a list of references at the end ofthis guide.D6051 152used? Assuming the composite is well mixed, then the preci-sion of a composite is a function of the number of samples asfollows:Number of IndividualSamples in CompositePrecisio
36、n (standard deviation n)of One Composite Sample2 0.353 0.294 0.255 0.226 0.20Thus, if six samples are included in a composite, thecomposite concentration of 0.6 mg/kg is two standard devia-tions below the regulatory limit. Therefore, if the compositeconcentration is actually observed to be in the ne
37、ighborhood of0.6 mg/kg, we can be reasonably confident (approximately95 %) that the concentration of the site is below the regulatorylimit, using only one composite sample.6.5 Example 2Another example is when the standarddeviation of the individual samples in the previous example isrelatively small,
38、 say 0.1 mg/kg. Then the standard deviation ofa composite of 6 individual samples is 0.04 mg/kg (0.1 mg/kgdivided by the square root of 6 = 0.04 mg/kg), a very smallnumber relative to the regulatory limit of 1 mg/kg. In this case,simple comparison of the composite concentration to theregulatory limi
39、t is often quite adequate for decision-makingpurposes.6.5.1 The effectiveness of compositing depends on therelative magnitude of sampling and analytical error. Whensampling uncertainty is high relative to analytical error (as isusually assumed to be the case) compositing is very effective inimprovin
40、g precision. If analytical errors are high relative tofield errors, sample compositing is much less effective.6.5.2 Because compositing is a physical averaging process,composite samples tend to be more normally distributed thanthe individual samples. The normalizing effect is frequently anadvantage
41、since calculation of means, standard deviations andconfidence intervals generally assume the data are normallydistributed. Although environmental residue data are com-monly non-normally distributed, compositing often leads toapproximate normality and avoids the need to transform thedata.6.5.3 The sp
42、atial design of the compositing scheme can beimportant. Depending upon the locations from which theindividual samples are collected and composited, compositescan be used to determine spatial variability or improve theprecision of the parameter being estimated. Fig. 1 and Fig. 2represent a site divid
43、ed into four cells. Composite all sampleswith the same number together. The sampling approach in Fig.1 is similar to sample random sampling, except they are nowcomposite samples. Each composite sample in this case is arepresentative sample of the entire site, eliminates cell-to-cellvariability, and
44、leads to increased precision in estimating themean concentration of the site. If there is a need to estimate thecell-to-cell variability, then the approach in Fig. 2 is suitable. Inaddition, if the precision of estimating the mean concentrationof the cell is needed, multiple composite samples should
45、 becollected from that cell.6.6 Effect on Cost ReductionBecause the compositesamples yield a more precise mean estimate than the samenumber of individual samples, there is the potential forsubstantial cost saving. Given the higher precision associatedwith composite samples, the number of composite s
46、amplesrequired to achieve a specified precision is smaller than thatrequired for individual samples. This cost saving opportunity isespecially pronounced when the cost of sample analysis is highrelative to the cost of sampling, compositing, and analyzing.6.7 Hot Container/Hot Spot Identification and
47、 RetestingSchemesSamples can be combined to determine whether anindividual sample exceeds a specified limit as long as theaction limit is relatively high compared with the actualdetection limit and the average sample concentration. Depend-ing on the difficulty and probability of having to resample,
48、itmay be desirable to retain a split of the discrete samples forpossible analysis depending on the analytical results from thecomposite sample.6.8 Example 3One hundred drums are to be examined todetermine whether the concentration of PCBs exceeds 50mg/kg. Assume the detection limit is 5 mg/kg and mo
49、st drumshave non-detectable levels. Compositing samples from tendrums for analysis would permit determining that none of thedrums in the composite exceed 50 mg/kg as long as theconcentration of the composite is 5 mg/kg, one or more drums may exceed 50mg/kg and additional analyses of the individual drums arerequired to identify any hot drum(s). The maximum number ofsamples that can theoretically be composited and still detect ahot sample is the limit of concern divided by the actualdetection limit (for example, 50 mg/kg 5 mg/kg = 10).6.9 Example 4Assume background lev
