1、Designation: D7825 12Standard Guide forGenerating a Process Stream Property Value through theApplication of a Process Stream Analyzer1This standard is issued under the fixed designation D7825; the number immediately following the designation indicates the year oforiginal adoption or, in the case of
2、revision, 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 This guide covers and provides a workflow overview ofthe necessary steps related to generating a
3、 Process StreamProperty Value obtained from the application of a processstream analyzer.1.2 Generating a Process Stream Property Value from theapplication of a process stream analyzer requires the use ofseveral ASTM standards. These standards describe proceduresto collect a representative sample, es
4、tablish and validate therelationship to the primary test method, and calculate aproperty value with an expected uncertainty. Each standardbuilds or prepares data, or both, to be used in another standard.The workflow process culminates to produce a process streamanalyzer result that represents a user
5、 defined batch of product.The sequence in which the standards are to be utilized isdefined in this guide.1.3 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
6、 health practices and determine the applica-bility of regulatory limitations prior to use.2. Referenced Documents2.1 ASTM Standards:2D3764 Practice for Validation of the Performance of ProcessStream Analyzer SystemsD4177 Practice for Automatic Sampling of Petroleum andPetroleum ProductsD6122 Practic
7、e for Validation of the Performance of Multi-variate Online, At-Line, and Laboratory Infrared Spectro-photometer Based Analyzer SystemsD6299 Practice for Applying Statistical Quality Assuranceand Control Charting Techniques to Evaluate AnalyticalMeasurement System PerformanceD6624 Practice for Deter
8、mining a Flow-Proportioned Aver-age Property Value (FPAPV) for a Collected Batch ofProcess Stream Material Using Stream Analyzer DataD6708 Practice for Statistical Assessment and Improvementof Expected Agreement Between Two Test Methods thatPurport to Measure the Same Property of a MaterialD7235 Gui
9、de for Establishing a Linear Correlation Relation-ship Between Analyzer and Primary Test Method ResultsUsing Relevant ASTM Standard PracticesD7278 Guide for Prediction ofAnalyzer Sample System LagTimesD7453 Practice for Sampling of Petroleum Products forAnalysis by Process Stream Analyzers and for P
10、rocessStream Analyzer System ValidationD7808 Practice For Determining the Site Precision of aProcess Stream Analyzer on Process Stream MaterialE1655 Practices for Infrared Multivariate QuantitativeAnalysisE2617 Practice for Validation of Empirically Derived Mul-tivariate Calibrations3. Terminology3.
11、1 DefinitionsPlease refer to the individually citedASTM standards for definitions.3.2 Acronyms:3.2.1 FPAPV(s)Flow Proportional Average Property Val-ue(s)3.2.2 MLRMultilinear Regression3.2.3 PCRPrincipal Components Regression3.2.4 PLSPartial Least Squares3.2.5 PSPV(s)Process Stream Property Value(s)3
12、.2.6 PTMPrimary Test Method3.2.7 PTMR(s)Primary Test Method Result(s)3.2.8 PPTMR(s)Predicted Primary Test Method Result(s)3.2.9 QCQuality Control3.2.10 UAR(s)Uncorrected Analyzer Result(s)1This test method is under the jurisdiction of ASTM Committee D02 onPetroleum Products and Lubricants and is the
13、 direct responsibility of SubcommitteeD02.25 on Performance Assessment and Validation of Process Stream AnalyzerSystems.Current edition approved Nov. 1, 2012. Published February 2013. DOI: 10.1520/D7825-12.2For referenced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Customer
14、Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States14. Significance and Use4.1 The standards
15、employed in the Process Stream AnalyzerPSPV Generation Flow Diagram each have a specific deliver-able that when combined into a single system produces a PSPVenabling the representation of product by process streamanalyzer.4.2 The description of each standard in the process providesthe user with an o
16、verview of the application of the standard inthe process for developing a PSPV.5. Flow Diagram and Work Process5.1 A flow chart showing the process for generating a PSPVis shown in Fig. 1.5.2 The various standards shown in the flow chart areapplied in sequence, building on the results of the previou
17、sstandards. The end result is a PSPV which is expected to agreewith a PTMR for the same material to within the user-specifiedrequirements.FIG. 1 Process Stream Analyzer PSPV Generation Flow DiagramD7825 1226. Supporting Practices6.1 While practices D6299 and D6708 do not appear di-rectly as part of
18、the Flow Diagram/work process, the method-ologies described in these two practices are incorporated byreference.6.2 D6299 Practice for Applying Statistical Quality Assur-ance Techniques to Evaluate Analytical Measurement SystemPerformance:6.2.1 The statistical quality control procedures and toolsdes
19、cribed in D6299 are an integral part to the validation ofanalyzers by D3764 or D6122.6.3 D6708 Standard Practice for Statistical Assessment andImprovement of Expected Agreement Between Two Test Meth-ods that Purport to Measure the Same Property of a Material:6.3.1 D6708 covers statistical methodolog
20、y for assessingthe expected agreement between two standard test methods thatpurport to measure the same property of a material, and fordeciding if a simple linear bias correction can further improvethe expected agreement.6.3.2 Practices D3764, D6122 and D7235 which are part ofthe PSPV generation wor
21、k process all make use of thestatistical methodology described in D6708.7. Sampling and Sample Delivery7.1 D7453 Practice for Sampling of Petroleum Products forAnalysis by Process Stream Analyzer System Validation:7.1.1 Sampling is the initial process in the generation of aprocess stream property va
22、lue, equally important as any otherprocess in the chain it provides the material that everything willbe based on. The sampling standard provides guidance on howto collect a representative sample from the sample stream anddeliver it to the desired sample destination (line sample point,on-line analyze
23、r, or composite sampler).7.1.2 This sampling method is focused on sample streamdelivery without the contamination that can be found in processstreams. The end use of the sample is to determine physicalproperties of the sample so filtering and coalescing are requiredto protect the on-line analytical
24、system. Sampling methods likeD4177 are designed to collect a representative amount ofsediment and water so are not conducive to proper analyzersystem operation.7.2 D7278 Standard Guide for Predicting and MeasuringLag Times for On-Line Sampling:7.2.1 The time it takes for a sample to travel from thep
25、rocess stream to the analyzer inlet or line sample point,including the required flush volume, is a critical piece ofinformation when establishing the relationship between theon-line analyzer and the primary test method (D7235) and alsothe validation of the on-line system (D3764 and D6122).7.2.2 The
26、lag time needs to be taken into account whencollecting samples and recording analyzer reading to correlatewith lab sample results and process conditions.8. Analyzer System Site Precision8.1 D7808 Standard Guide for Determining the Site Preci-sion of a Process Stream Analyzer on Process Stream Materi
27、al:8.1.1 To properly apply D7235, and as an extension ofD7235 the application of D6708, the user needs to havedetermined the analyzer site precision over the expectedprocess stream operating range. The process employed todetermine the Site Precision of a Process Stream Analyzer onProcess Stream Mate
28、rial results can be utilized as part of a QCprogram as described in D6299.8.1.2 D3764 and D6122 require that the user has determinedanalyzer site precision in order to verify that the processanalyzer system is at steady state during the validation process.8.1.3 To reliably calculate the analyzer sit
29、e precision, infra-structure must be in place to repeatedly introduce aliquots ofone or more bulk samples under the same conditions as thesample stream.9. Analyzer Calibration9.1 D7235 Standard Guide for Establishing a Linear Cor-relation Relationship between Analyzer and Primary TestMethod Results
30、using Relevant ASTM Standard Practices:9.1.1 Guide D7235 covers a general methodology to de-velop and access the linear relationship between uncorrectedanalyzer results (UARs) produced by a total analyzer systemversus results produced by the corresponding primary testmethod (PTMRs) that the analyzer
31、 is intended to emulate.9.1.2 Guide D7235 describes how the statistical methodol-ogy of Practice D6708 is employed to access the agreementbetween the PTM and the analyzer results, and if necessary,develop a linear correlation to improve the agreement over thecomplete operating range of the analyzer.
32、9.1.3 Guide D7235 applies two either of the following twocases: (1) the process stream analyzer system and the primarytest method are based on the same measurement principle(s),or, (2) the process stream analyzer system uses a direct andwell-understood measurement principle that is similar to themea
33、surement principle of the primary test method.9.1.4 If the process stream analyzer system uses a differentmeasurement technology from the PTM, this practice alsoapplies provided that the calibration protocol for the directoutput of the analyzer does not require use of the PTM.9.1.5 Procedures are de
34、scribed to ensure that the sample setused to generate the linear correlation are representative of thematerial type and property range for the intended analyzerservice, and to ensure that the set provides adequate variationin property level.9.1.6 Preferably, line samples are collected in accordancew
35、ith Practice D3764 from a sampling point after the sampleconditioning system. Taking into account the analyzer lag time(D7278), corresponding analyzer results are obtained. The linesamples are measured by the PTM. The UARs and PTMRs areaccessed using the statistical methodology of D6708 to deter-min
36、e if there is an adequate linear relationship to allow PSPVsvalues to be estimated based on UARs.9.1.7 Alternatively, the statistical methodology of D6708can be used to correlate PTMRs obtained from compositesamples to FPAPVs generated from the analyzer results usingthe methodology of D6624.9.1.8 PP
37、TMRs are generated by applying the linear corre-lation to the measured analyzer result.D7825 1239.2 E1655 Standard Practices for Infrared MultivariateQuantitative Analysis:9.2.1 Practice E1655 may be used to develop the calibrationif the process stream analyzer system utilizes an indirect ormathemat
38、ically modeled measurement principle such as ch-emometric or multivariate analysis techniques where resultsfrom PTM are required for the development of the chemomet-ric or multivariate model.NOTE 1While the practices described within E1655 deal specificallywith mid- and near-infrared analysis, much
39、of the mathematical andprocedural detail contained therein is also applicable to other analyticalmethods. The user of E1655 is cautioned that typical and best practices formultivariate quantitative analysis using data from other multivariateanalytical techniques may differ from that described in E16
40、55 for mid- andnear-infrared spectroscopies.9.2.2 E1655 describes procedures for collection and treatingdata for developing multivariate calibrations. Multivariatemathematics are applied to correlate spectra measured for a setof calibration samples to PTMRs for this same set. Theresultant multivaria
41、te calibration model is applied to theanalysis of the spectrum of an unknown sample to estimate thePPTMR for that sample. Application of E1655 is limited tomodels developed by MLR, PCR or PLS.9.2.3 E1655 describes procedures for validating the calibra-tion model. This validation tests for bias in mo
42、del predictions,and for the expected agreement between the PPTMRs andPTMRs. Validation of the model is intended to demonstrate themultivariate models capabilities, but it is not intended as ameasure of analyzer performance. The performance of multi-variate analyzer systems must be validated using pr
43、oceduresdescribed in practice D6122.9.2.4 E1655 describes statistical tests which are employedto detect when samples being analyzed exceed the range forwhich the multivariate model has been validated. The analysisof such “outlier” samples represents an extrapolation of themodel, and there is lower c
44、onfidence that PPTMRs predictedby extrapolation will agree with PTMRs.9.3 E2617 Standard Practice for Validation of EmpiricallyDerived Multivariate Calibrations:9.3.1 If the multivariate model is developed using tech-niques other than MLR, PCR or PLS, then the model itselfshould be validated using E
45、2617.As with E1655, the validationof the model is intended to demonstrate the multivariatemodels capabilities, but it is not intended as a measure ofanalyzer performance.9.3.2 For calibrations developed using Practice E2617,Practice D6122 is still used to validate analyzer performance.10. Validation
46、 of Process Analyzer Performance10.1 D3764 Practice for Validation of Process Analyzers:10.1.1 For analyzers where Guide D7235 was used toestablish the linear correlation relationship between UARs andPTMRs, Practice D3764 is applied for the validation ofperformance.10.1.2 After the analyzer system h
47、as been calibrated per themanufactures requirements, and the correlation between thesystem and a PTM has been completed (D7235) andimplemented, the analyzer system is ready for validation.Analyzer validation is proving that all the previous steps in thework process have been successfully carried out
48、 and that theanalyzer system produces PPTMRs that predict the PTMresults to within the users expectations and requirements.10.1.3 After an analyzer is installed, or major maintenanceis conducted, a probationary validation is performed to dem-onstrate that the PPTMRs agree with the PTMRs to within th
49、euser-specified requirements for the analyzer system. The pro-bationary validation is conducted using a limited set ofmaterials which were not used in developing the linearcorrelation. Once the analyzer passes the probationaryvalidation, it may be put into service to generate PPTMRs.10.1.4 General and continual validation of the analyzerperformance is conducted using the Quality Assurance meth-odology of D6299.10.2 D6122 Practice for Validation of Multivariate ProcessInfrared Spectrometers:10.2.1 For analyzers where E1655 or E2617 was used toestablish the relationship b
