1、Designation: D 7235 05An American National StandardStandard Guide forEstablishing a Linear Correlation Relationship BetweenAnalyzer and Primary Test Method Results Using RelevantASTM Standard Practices1This standard is issued under the fixed designation D 7235; the number immediately following the d
2、esignation indicates the year oforiginal adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.INTRODUCTIONOperation of a process str
3、eam analyzer system typically involves four sequential activities: (1)Analyzer CalibrationWhen an analyzer is initially installed, or after major maintenance has beenperformed, diagnostic testing will typically be performed to demonstrate that the analyzer meetsmanufacturers specifications and histo
4、rical performance standards. These diagnostic tests may requirethat the analyzer be adjusted so as to provide predetermined output levels for certain referencematerials. (2) Correlation to Primary Test MethodFor process stream analyzer systems that are notprimary test methods (PTM), once the diagnos
5、tic testing is completed, process stream samples willtypically be analyzed using both the analyzer system and the corresponding primary test method. Amathematical function will be derived that relates the analyzer output to the primary test method(PTM). The application of this mathematical function
6、to an analyzer output produces a predicted PTMresult. (3) Initial ValidationOnce the relationship between the analyzer output and primary testmethod results has been established, an initial validation is performed using an independent data setto demonstrate that the predicted PTM results agree with
7、those from the primary test method withinthe tolerances established from the Correlation activities and with no observable systemic bias. (4)Continual ValidationDuring normal operation of the process analyzer system, quality assurancetesting is conducted to demonstrate that the agreement between ana
8、lyzer and primary test methodresults during the Initial Validation is maintained. This document provides guidance for item (2)above.1. Scope1.1 This guide covers a general methodology to develop andassess the linear relationship between results produced by atotal analyzer system versus the results p
9、roduced by thecorresponding primary test method (PTM) that the analyzersystem is intended to emulate, using the principles and ap-proaches outlined in relevant ASTM standard practices andguides.1.2 This guide describes how the statistical methodology ofPractice D 6708 can be employed to assess agree
10、ment betweenthe PTM and analyzer results, and, if necessary, develop linearcorrelation to further improve the agreement over the completeoperating range of the analyzer. For instances where there isinsufficient variation in property level to apply the PracticeD 6708 multi-level methodology, users ar
11、e referred to PracticeD 3764 to perform a level specific bias evaluation. Thecorrelation relationship information obtained in the applicationof this guide is applicable only to the material type andproperty range of the materials representative of those used toperform the assessment. Users are cauti
12、oned against extrapo-lation of the relationship beyond the material type and propertyrange being studied.1.3 This guide applies if the process stream analyzer systemand the primary test method are based on the same measure-ment principle(s), or, if the process stream analyzer systemuses a direct and
13、 well-understood measurement principle that issimilar to the measurement principle of the primary testmethod. If the process stream analyzer system uses a differentmeasurement technology from the primary test method, pro-vided that the calibration protocol for the direct output of theanalyzer does n
14、ot require use of the PTM, this practice alsoapplies.1This guide is under the jurisdiction of ASTM Committee D02 on PetroleumProducts and Lubricants and is the direct responsibility of Subcommittee D02.25 onPerformance Assessment and Validation of Process Stream Analyzer Systems forPetroleum and Pet
15、roleum Products.Current edition approved Dec. 1, 2005. Published February 2006.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.1.4 This guide does not apply if the process stream analyzersystem utilizes an indirect or mathematically
16、modeled mea-surement principle such as chemometric or multivariate analy-sis techniques where results from PTM are required for thechemometric or multivariate model development. Users shouldrefer to Practices E 1655 and D 6122 for detailed correlationand model validation procedures for these types o
17、f analyzersystems.NOTE 1For example, this guide would apply for the comparison ofbenzene measurements from a mid-infrared process analyzer system basedon Test Method D 6277 to those obtained using PTM Test MethodD 3606, a gas chromatography based test method. For each sample, themid-infrared spectru
18、m is converted into a single analyzer result usingmethodology (Test Method D 6277) that is independent of the primary testmethod (Test Method D 3606). However, when the same analyzer uses amultivariate model to correlate the measured mid-infrared spectrum toTest Method D 3606 reference values using
19、the methodology of PracticeE 1655, this guide does not apply. In this case, the direct output of theanalyzer is the spectrum, and the conversion of this multivariate output toan analyzer result require results from the primary test method.1.5 This guide assumes that the analyzer sampling system isfi
20、t for use, and both analyzer and lab systems are in statisticalcontrol during the execution of the required tasks. Proceduresfor testing for proper function of the analyzer sampling system,and ascertaining whether the systems are in statistical controlare beyond the scope of this guide.1.6 Software
21、program CompTM Version 1.0.21(ADJD6708) performs the necessary computations recom-mended by this guide.1.7 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 h
22、ealth practices and determine the applica-bility of regulatory limitations prior to use.2. Referenced Documents2.1 ASTM Standards:2D 3606 Test Method for Determination of Benzene andToluene in Finished Motor and Aviation Gasoline by GasChromatographyD 3764 Practice for Validation of Process Stream A
23、nalyzerSystemsD 4177 Practice for Automatic Sampling of Petroleum andPetroleum ProductsD 5191 Test Method for Vapor Pressure of Petroleum Prod-ucts (Mini Method)D 6122 Practice for Validation of Multivariate Process In-frared SpectrophotometersD 6277 Test Method for Determination of Benzene inSpark-
24、Ignition Engine Fuels Using Mid Infrared Spectros-copyD 6299 Practice for Applying Statistical Quality AssuranceTechniques to Evaluate Analytical Measurement SystemPerformanceD 6624 Practice for Determining a Flow-Proportioned Av-erage Property Value (FPAPV) for a Collected Batch ofProcess Stream Ma
25、terial Using Stream Analyzer DataD 6708 Practice for Statistical Assessment and Improve-ment of the Expected Agreement Between Two TestMethods that Purport to Measure the Same Property of aMaterialE 1655 Practices for Infrared Multivariate QuantitativeAnalysis2.2 American Petroleum Institute Documen
26、t:3API TP-550 Manual on Installation of Refinery Instrumentsand Control Systems, Part II, Process Stream Analyzers2.3 ASTM Adjuncts:ADJD6708 CompTM Version 1.0.2143. Terminology3.1 All of the terminology as defined in Practices D 3764,D 6122, and D 6708 are adopted for this guide.4. Significance and
27、 Use4.1 This guide is intended to be used in conjunction withPractice D 3764 (Case 1) and Practice D 6122 (Case 2).Methodology in this guide can be used to determine if a linearcorrelation can improve the performance of the total analyzersystem in terms of its ability to predict the results that the
28、 PTMwould have been if applied to the same material. This meth-odology, which is based on the same statistical data treatmentas Practice D 6708, is use to derive the parameters of the linearrelationship and to assess the degree of improvement.4.2 This guide provides developers or manufacturers ofpro
29、cess stream analyzer systems with useful procedures fordeveloping the capability of newly designed systems forindustrial applications that require reliable prediction of mea-surements of a specific property by a primary test method of aflowing component or product.4.3 This guide provides purchasers
30、of process stream ana-lyzer systems with some reliable options for specifying perfor-mance requirements for process stream analyzer systems thatare used in applications requiring reliable prediction of mea-surements of a specific property by a primary test method of aflowing component or product.4.4
31、 This guide provides the user of a process streamanalyzer system with useful information on the work processfor establishing the PTM prediction relationship and predictionperformance.4.5 Prediction (correlation) relationship obtained in theapplication of this guide is applicable only to the material
32、 typeand property range of the materials used to perform the study.Selection of the property levels and the compositional charac-teristics of the samples must be suitable for the application ofthe analyzer system. Users are cautioned against extrapolationof the prediction relationship beyond the mat
33、erial type andproperty range used to obtain the relationship.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.
34、3Available from American Petroleum Institute (API), 1220 L. St., NW, Wash-ington, DC 20005.4Available from ASTM International Headquarters. Order Adjunct No.ADJD6708.D72350524.6 The degree-of-agreement assessment promoted in thisguide is based on the statistical principles articulated inPractice D 6
35、708, which is purely statistical in nature. Noattempt is made in assessing the degree of similarity in theanalytical technique between the process analyzer and thePTM; hence, results between the PTM and analyzer unit can behighly correlated, but their measurement principles may becompletely differen
36、t, and may not be the principal cause forcorrelation. Users are therefore cautioned that a high degree ofcorrelation between results does not necessarily imply a highdegree of similarity in the measurement principles; nor does itimply a similar degree of agreement can be expected in futuremeasuremen
37、ts. In general, if sample-specific biases are de-tected, it suggests that the measurement principles may bedifferent, and may affect the degree-of-agreement in future useof the scaling/bias-correction equation. Presence or absence ofsample-specific effect can be used as a measure of therobustness of
38、 the correlation equation to sample composition ormatrix differences.4.7 Implementation of this guide requires that the processstream analyzer system complies with the following condi-tions:4.7.1 Meets the principles set forth in PART II ProcessStream Analyzers of API TP-550,4.7.2 Meets the supplier
39、s recommendation,4.7.3 Complies with operating conditions specified by themanufacturer,4.7.4 A predicted PTM algorithm has already been estab-lished if necessary, and4.7.5 Meets applicable quality assurance, data collectionand data telemetry protocols.4.8 After installation or major maintenance, con
40、duct suchdiagnostic tests as recommended by the manufacturer todemonstrate that the analyzer meets manufacturers specifica-tions, historical performance levels or both. If necessary, adjustthe analyzer system components so as to obtain recommendedanalyzer output levels for specified reference materi
41、als.4.9 Inspect the entire analyzer system to ensure it is in-stalled properly, is in operating condition, and is properlyadjusted after completion of the initial commissioning proce-dures.5. Calibration5.1 When an analyzer is initially installed, and after majormaintenance has been preformed, diagn
42、ostic tests should beconducted to demonstrate that the analyzer meets manufactur-ers specifications and historical performance standards. Thesediagnostic tests may require that the analyzer be adjusted so asto provide predetermined output levels for certain referencematerials. Such adjustment may be
43、 done in hardware, softwareor both. This should not be confused with the development ofcorrelation to a PTM, which, is described below.6. Correlation to Primary Test Method6.1 General Approach:6.1.1 Define the sample set to be used for assessment.6.1.1.1 The material type and property range for whic
44、h theanalyzer system results are to be assessed versus the primarytest method is defined.6.1.1.2 The recommended sample set design criteria for thisassessment are:(1) A minimum of six replicates at each major product/property level combination,(2) The range of major product/property levels exceed at
45、least two times the published reproducibility of the PTM, and(3) A minimum of 30 total samples.6.1.1.3 Replication at a specific level/product combinationcan be different batches of production material that are nomi-nally similar in property level (within 1.2 times the ASTMreproducibility of the PTM
46、) and composition.6.1.2 Obtain site precision information for the analyzersystem and PTM for the material type and range defined inaccordance with the procedures outlined in Practice D 6299.Ifanalyzer system precision cannot be obtained using the meth-odology in Practice D 6299, manufacturers publis
47、hed repeat-ability precision or site precision from other similar systemsmay be used as a surrogate. Users are cautioned to ensureconsistency in statistical definitions between the manufactur-ers published repeatability and the site precision statistic asdefined in Practice D 6299 for the PTM.SAMPLI
48、NG METHOD6.1.3 Line Sampling (Preferred)Samples meeting the re-quirements of the sample set design criteria above are taken inaccordance with Practice D 3764, Line Sample Procedure, atthe crosscheck sample points of the analyzer system (see Fig.1), after the sample conditioning subsystem, at a frequ
49、ency ofno more than once per day. Avoid taking this sample at thesame time of day to ensure any time-of-day related effect iscaptured in the dataset.6.1.4 Automated Composite SamplingFor installationsthat have automated composite sampling systems meeting therequirement of Practice D 4177 (or equivalent), and, theapplication is intended to provide a predicted PTM result for abatch of homogenous production material using an Flow-ProportionedAverage Property Value (FPAPV) calculated fromthe analyzer system as per Practice D 6624, the correlationequation can be established usin
