1、Designation: D6543 00 (Reapproved 2012)Standard Guide tothe Evaluation of Measurements Made by On-Line CoalAnalyzers1This standard is issued under the fixed designation D6543; the number immediately following the designation indicates the year oforiginal adoption or, in the case of revision, the yea
2、r 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 This guide provides techniques to be used for theevaluation of the measurement performance of on-line coalanalyze
3、rs.1.2 This standard does not purport to address all of thesafety concerns, if any, associated with its use. It is theresponsibility 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. Reference
4、d Documents2.1 ASTM Standards:2D121 Terminology of Coal and CokeD2013 Practice for Preparing Coal Samples for AnalysisD2234/D2234M Practice for Collection of a Gross Sampleof CoalD4702 Practice for Quality Management of MechanicalCoal Sampling Systems (Withdrawn 2008)3D6518 Practice for Bias Testing
5、 a Mechanical Coal Sam-pling System (Withdrawn 2008)3E177 Practice for Use of the Terms Precision and Bias inASTM Test MethodsE178 Practice for Dealing With Outlying ObservationsE456 Terminology Relating to Quality and Statistics3. Terminology3.1 Definitions:3.1.1 analyzer system, na coal quality me
6、asurement sys-tem which includes an on-line coal analyzer and which mayinclude one or more stages of a coal-sampling system.3.1.2 calibration, nmathematical modeling of analyzerand comparative coal sampling and analysis data. Factors fromthe model are used in the on-line analyzer control software.3.
7、1.3 Grubbs estimator, nan estimate of the measurementprecision of an on-line analyzer (1-3).43.1.4 on-line analyzer, nan analytical tool consisting of aninstrument and systems, which together provide measurements,or estimates, or both, of coal quality parameters.3.1.5 outlier, nan extreme value that
8、 statistical tests indi-cate to be far enough from other results in a population underconsideration to cause suspicion that the value is not a memberof the population.3.1.6 reference material, nmaterial of stable compositionthat may be used to generate static analyzer measurements.3.1.7 reference sy
9、stem, na measurement system used tomeasure the characteristics of a lot of coal that are alsomeasured by an on-line-analyzer, and against which the on-lineanalyzer measurements are compared.3.1.8 standardization, ncalibration of an instrument to areference material using static stability measurement
10、s.3.1.9 static stability, nan estimate of the measurementprecision of an instrument obtained on material that is notmoving. The estimate normally is expressed as the standarddeviation and average of the measurements for a given periodof time.3.1.10 synchronization error, nan error that occurs fromco
11、mparing measurements made by an on-line analyzer and areference system that are not measuring exactly the same lotbecause of temporal and/or spatial offsets.4. Summary of Guide4.1 This guide outlines the evaluation of the measurementperformance of an on-line analyzer using comparative mea-surements.
12、 The comparative test uses a paired comparison ofanalysis from a reference method using ASTM sampling,sample preparation, and analysis methods for a lot of coal withthe analysis from the on-line analyzer for the same lot of coal.1This guide is under the jurisdiction of ASTM Committee D05 on Coal and
13、Coke and is the direct responsibility of Subcommittee D05.23 on Sampling.Current edition approved Sept. 1, 2012. Published November 2012. Originallyapproved in 2000. Last previous edition approved in 2006 as D654300(2006).DOI: 10.1520/D6543-00R12.2For referenced ASTM standards, visit the ASTM websit
14、e, 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 standard is referenced onwww.astm.org.4The boldface numbers in paren
15、theses refer to the list of references at the end ofthis standard.Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States1The data resulting from the comparative test may be evaluatedusing graphical and statistical techniques outlined below.4
16、.2 Various techniques are recommended by on-line ana-lyzer manufacturers for standardization or static testing. Thesetechniques are useful for establishing a benchmark beforeconducting a comparative test. These techniques also may beused as diagnostic tests in accordance with methods recom-mended by
17、 on-line analyzer manufacturers and graphical andstatistical techniques included in this guide.5. Significance and Use5.1 On-line analyzers are used to provide quality data onlots of coal. The resulting quality data are used as a productiontool or for some contractual application. This guide provide
18、smeans of evaluating the system and data produced.5.2 The user should become familiar with the documentsterminology and layout. The section on test design and datacollection will provide the means by which all the analysis datawill be gathered. The test design should be carefully consid-ered so as t
19、o assure meeting the users requirements.5.3 The procedures defined in this guide can be used to testthe accuracy and precision of an on-line analyzer, for accep-tance after its installation, to check precision and accuracyduring routine use (quality control), when changes are made tothe system, when
20、 the nature of the coal being tested changes,and to determine mathematical factors to be used for calibra-tion of the on-line analyzer.6. Selection and Conduct of Performance Evaluations6.1 Introduction:6.1.1 Several techniques can be used to evaluate the perfor-mance of an on-line analyzer. These t
21、echniques provide datathat can be evaluated by using the graphical and statisticalmethods described in Section 7 of this guide.6.1.2 At time of installation, all of the graphical andnumerical methods outlined in this may prove useful. On aroutine basis, conducting any of the instrument stability che
22、cksand comparative evaluations that do not disrupt normal opera-tions may prove useful. Control charts may be applied to all theperformance measures that are gathered on a routine basis, forexample, mean analysis value of reference material, RMSD,and so forth.6.1.3 In the event that there is a chang
23、e to the operationalparameters or the equipment associated with the analyzer or thereference system, comparative checks should be performed. Inaddition to comparative checks, standardization or staticchecks, or both, as recommended by the on-line analyzermanufacturer may be helpful.6.1.4 Changes in
24、the following may impact performanceevaluations. Coal characteristics, such as particle size, sourceof coal, mining techniques, degree of preparation, and so forth,which if changed from previous test periods and are not in theanalyzer calibration database, may affect analyzer precisionand accuracy:6
25、.1.4.1 Analyzer components;6.1.4.2 Coal-handling system;6.1.4.3 Laboratory services;6.1.4.4 Sampling technique;6.1.4.5 Coal flow rates; and6.1.4.6 Power disturbances.6.2 Static Stability Measurements for BaselineAssessmentA reference material may be used to provide abaseline assessment of static mea
26、surement precision. Thereference material may be used to compare current mean andstandard deviation values with previous mean and standarddeviation values, previously collected in the same manner. Theresulting comparative data may be used to assess whether anychange in the mean or standard deviation
27、 of the static mea-surements may be attributed to a change in the operatingcharacteristics of the analyzer, in the absence of the influenceof sampling and analysis.6.2.1 Reference materials may include actual coal in asealed container, cast high-carbon solids, or other materialswhich may be provided
28、 by the analyzer manufacturer. Theprimary characteristic of these materials is that they not besubject to changes in composition.6.2.2 The results of this evaluation can indicate whetheranalyzer performance has significantly degraded or whether ashift may have occurred. If so, it may be possible to
29、adjust theanalyzer to restore initial performance. If the user wishes onlyto measure current static repeatability, any available coal maybe used in the analysis zone of the analyzer. Note, however,that the actual standard deviation in static repeatability testsmight be influenced by the composition
30、of the coal beingexamined or analyzer factors, such as the strength of theradioactive sources used by the analyzer or condition ofanalyzer electronic components.6.2.3 It is essential that the length of the analysis period bedefined, for example, 1, 2, or 5 min, and be constant in thestatic stability
31、 test. The static standard deviation resulting fromthe static stability test decreases as the length of the analysisperiod increases. Static stability testing may be conducted onthe same day(s) as comparative tests. One type of static testinvolves a comparison of the analyzer to static coal over man
32、yhours.6.3 Comparison of Analyzer System to Reference SystemMeasurements:6.3.1 Once an analyzer installation has been completed andcalibration adjustments have been made, the analyzer ownermay require acceptance testing. Also, the analyzer owner maydecide to relocate the analyzer. In these cases, co
33、mparisontests, that is, to compare the analyzer system results toconventional sampling and analysis techniques, will providethe user calibration verification data and/or data that could beused for recalibration of the analyzer.6.3.2 Depending upon the comparative technique chosen,there may be need f
34、or more than one comparison. Thesecomparisons may be conducted in a batch over several hours ordays or continuously throughout the operational life of theanalyzer system.6.3.3 If two independent conventional coal-sampling andlaboratory analysis measurements can be made from each of aseries of batche
35、s of coal interrogated by the analyzer, methodsdeveloped by Grubbs (1-3) can be used to provide unbiasedestimates of the measurement precision of the analyzer and ofthe conventional sampling and analysis methods.D6543 00 (2012)26.3.4 Any two series of measurements are independent iftheir measurement
36、 errors are uncorrelated. Correlation ofmeasurement errors can be avoided and independence assuredby use of a true random selection of physical increments orsamples of material or by using different schemes and equip-ment for collection, preparation, and laboratory analysis of thesamples, or both. A
37、 complete treatment of the subject ofindependence of measurements and the various means ofassuring independence is beyond the scope of this guide.6.4 Sampling Considerations:6.4.1 Selection of Appropriate Sampling and Sample Prepa-ration MethodsThe following increment collection methods(see Test Met
38、hods D2234/D2234M) are listed in general orderof preference; this preference is not rigid. Often, practicalconsiderations may supercede increment selection decisions.The method to be used for sample preparation should bedetermined before the beginning of increment collection.Sample preparation techn
39、iques should remain consistent (seeMethod D2013). Before installation of the analyzer, consider-ation should be given to the ability to obtain representativesamples for comparison to analyzer measurements and theregimen for sample handling and analysis. For the use ofmechanical sampling systems, ins
40、pection (see Guide D4702)and bias testing (see Practice D6518) are good methods forevaluation of the systems suitability for the test.6.4.1.1 Full Stream CutsWhenever possible, the fullstream cut method of increment collection should be used inthe evaluation of an on-line analyzer. This is imperativ
41、e fordrawing inference using statistical techniques.6.4.1.2 Stopped Belt CutAlthough using “stopped belt”increments allows drawing statistical inferences from the data,the stopped belt increment has limited applicability in theevaluation of on-line analyzers. This especially is true incollection of
42、a sample from the lot interrogated by an analyzeroperating in conjunction with a mechanical sampling system.The stopped belt increment may have applicability in instancesin which the analyzer is used in conjunction with mechanicalsampling, and the discharge of the analyzer may be meteredonto a belt
43、that is subject to stopping without interference withthe analyzers operation. The stopped belt increment may beuseful in evaluating analyzers that are not associated withmechanical sampling and that are being evaluated for veryshort-term performance, that is, less than 1-min results. In suchinstance
44、s, each stop of the belt is used to collect a sample froma relatively long portion of the belt. As an example, theanalyzer might collect data for 30 s and the belt stopped in sucha manner as to allow access to the entire segment of materialinterrogated by the analyzer (typically, 300 ft for a belt r
45、unning600 ft/min for 30 s). A sample may now be collected byremoving increments from the stopped belt throughout theentire stopped portion. This method reduces the number of beltstops required to assemble a database but is limited inapplicability by the short analysis time and physical require-ments
46、 of the facility wherein the equipment is installed.Nevertheless, this technique may be considered when noreliable mechanical technique is available and the alternative isa manual part stream cut.6.4.1.3 Part Stream IncrementsThis mode of sample col-lection may be required in the evaluation of analy
47、zers notassociated with mechanical sampling or in some tests in whichmore than one comparison will be made. Since this type ofsample does not allow statistical inferences to be drawn in arigorous manner, its use is discouraged. Indeed, the only use ofthis type of comparison might be for control char
48、ting purposesrelative to an historical reference.6.4.2 Selection of Reference Sample Point(s):6.4.2.1 Comparative tests can be used to evaluate theperformance of either the analyzer itself or the “analyzersystem” (consisting of the analyzer and the sampling systemthat feeds it). The comparative eval
49、uations of the analyzersystem can be used to determine the ability of the analyzersystem to measure the characteristics of the main coal stream.The selection of the reference system sample point(s) deter-mines whether the comparative tests will assess the analyzer orthe analyzer system.6.4.2.2 The most direct and practical two-instrument test,when the analyzer is fed the secondary reject of a mechanicalsampling system, uses the final save to compare directly to theanalyzer. A manual or mechanical sample collected from theanalyzer discharge may provide an independen
