ASTM D3764-23

Standard Practice for Validation of the Performance of Process Stream Analyzer Systems

SIGNIFICANCE AND USE
5.1 This practice can be used to quantify the performance of a process stream analyzer system or its subsystem in terms of precision and bias relative to those of a primary test method for the property of interest.  
5.2 This practice provides developers or manufacturers of process stream analyzer systems with useful procedures for evaluating the capability of newly designed systems for industrial applications that require reliable prediction of measurements of a specific property by a primary test method of a flowing component or product.  
5.3 This practice provides purchasers of process stream analyzer systems with some reliable options for specifying acceptance test requirements for process stream analyzer systems at the time of commissioning to ensure the system is capable of making the desired property measurement with the appropriate precision or bias specifications, or both.  
5.4 PPTMR from Analyzer Systems validated in accordance with this practice can be used to predict, with a specified confidence, what the PTMR would be, to within a specified tolerance, if the actual primary test method was conducted on the materials that are within the validated property range and type.  
5.5 This practice provides the user of a process stream analyzer system with useful information from on-going quality control charts to monitor the variation in δ over time, and trigger update of correlation relationship between the analyzer system and primary test method in a timely manner.  
5.6 Validation information obtained in the application of this practice is applicable only to the material type and property range of the materials used to perform the validation. Selection of the property levels and the compositional characteristics of the samples must be suitable for the application of the analyzer system. This practice allows the user to write a comprehensive validation statement for the analyzer system including specific limits for the validated range of application. Th...
SCOPE
1.1 This practice describes procedures and methodologies based on the statistical principles of Practice D6708 to validate whether the degree of agreement between the results produced by a total analyzer system (or its subsystem), versus the results produced by an independent test method that purports to measure the same property, meets user-specified requirements. This is a performance-based validation, to be conducted using a set of materials that are not used a priori in the development of any correlation between the two measurement systems under investigation. A result from the independent test method is herein referred to as a Primary Test Method Result (PTMR).  
1.1.1 The degree of agreement described in 1.1 can be either for PPTMRs and PTMRs measured on the same materials, or for PPTMRs measured on basestocks and PTMRs measured on these same basestocks after constant level additivation.  
1.1.2 In some cases, a two-step procedure is employed. In the first step, the analyzer and PTM are applied to the measurement of the same blendstock material. If the analyzer employed in Step 1 is a multivariate spectrophotometric analyzer, then Practice D6122 is used to access the agreement between the PPTMRs and the PTMRs for this first step. Otherwise, this practice is used to compare the PPTMRs to the PTMRs measured for this blendstock to determine the degree of agreement. In a second step, the PPTMRs produced in Step 1 are used as inputs to a second model that predicts the results obtained when the PTM is applied to the analysis of the finished blended product. Since this second step does not use analyzer readings, the validation of the second step is done independently. Step 2 is only performed on valid Step 1 results. Note that the second model might accommodate variable levels or multiple material additions to the blendstock.  
1.2 This practice assumes any correlation necessary to mitigate systemic biases between the ...

General Information

• Status: Published
• Publication Date: 30-Jun-2023
• ICS: 17.120.10 - Flow in closed conduits
• Technical Committee: D02 - Petroleum Products, Liquid Fuels, and Lubricants
• Drafting Committee: D02.25 - Performance Assessment and Validation of Process Stream Analyzer Systems

Relations

• Refers: ASTM D3606-24 - Standard Test Method for Determination of Benzene and Toluene in Spark Ignition Fuels by Gas Chromatography
• Effective Date: 01-Mar-2024
• Refers: ASTM D6708-24 - Standard Practice for Statistical Assessment and Improvement of Expected Agreement Between Two Test Methods that Purport to Measure the Same Property of a Material
• Effective Date: 01-Mar-2024
• Refers: ASTM D1265-23a - Standard Practice for Sampling Liquefied Petroleum (LP) Gases, Manual Method
• Effective Date: 01-Dec-2023
• Refers: ASTM D6299-23a - Standard Practice for Applying Statistical Quality Assurance and Control Charting Techniques to Evaluate Analytical Measurement System Performance
• Effective Date: 01-Dec-2023
• Refers: ASTM D5842-23 - Standard Practice for Sampling and Handling of Fuels for Volatility Measurement
• Effective Date: 01-Oct-2023
• Refers: ASTM D6122-23 - Standard Practice for Validation of the Performance of Multivariate Online, At-Line, Field and Laboratory Infrared Spectrophotometer, and Raman Spectrometer Based Analyzer Systems
• Effective Date: 01-Jul-2023
• Refers: ASTM F307-13(2020) - Standard Practice for Sampling Pressurized Gas for Gas Analysis
• Effective Date: 01-Apr-2020
• Refers: ASTM D5842-19 - Standard Practice for Sampling and Handling of Fuels for Volatility Measurement
• Effective Date: 01-Nov-2019
• Refers: ASTM D6122-19b - Standard Practice for Validation of the Performance of Multivariate Online, At-Line, and Laboratory Infrared Spectrophotometer Based Analyzer Systems
• Effective Date: 01-Jun-2019
• Refers: ASTM D6708-19 - Standard Practice for Statistical Assessment and Improvement of Expected Agreement Between Two Test Methods that Purport to Measure the Same Property of a Material
• Effective Date: 01-May-2019
• Refers: ASTM D6122-19a - Standard Practice for Validation of the Performance of Multivariate Online, At-Line, and Laboratory Infrared Spectrophotometer Based Analyzer Systems
• Effective Date: 01-May-2019
• Refers: ASTM D6122-19 - Standard Practice for Validation of the Performance of Multivariate Online, At-Line, and Laboratory Infrared Spectrophotometer Based Analyzer Systems
• Effective Date: 01-Jan-2019
• Refers: ASTM D6122-18 - Standard Practice for Validation of the Performance of Multivariate Online, At-Line, and Laboratory Infrared Spectrophotometer Based Analyzer Systems
• Effective Date: 01-Jul-2018
• Refers: ASTM D6708-18 - Standard Practice for Statistical Assessment and Improvement of Expected Agreement Between Two Test Methods that Purport to Measure the Same Property of a Material
• Effective Date: 01-Apr-2018
• Refers: ASTM D6299-17b - Standard Practice for Applying Statistical Quality Assurance and Control Charting Techniques to Evaluate Analytical Measurement System Performance
• Effective Date: 15-Dec-2017

Overview

ASTM D3764-23: Standard Practice for Validation of the Performance of Process Stream Analyzer Systems provides a systematic approach for evaluating and validating the performance of process stream analyzer systems. This standard is critical in industries such as petroleum, petrochemicals, and fuels, where accurate real-time or at-line analysis of process streams is essential for operational efficiency, quality control, and regulatory compliance. The practice guides both developers and end-users in confirming that analyzer systems deliver measurements with accuracy, precision, and bias within required tolerances compared to recognized primary test methods.

Key Topics

• Performance Validation: The standard outlines statistical methodologies for evaluating the agreement between analyzer system outputs and primary test method results, ensuring user-specified criteria are met.
• Precision and Bias Assessment: Provides guidance on quantifying both the reproducibility (precision) and bias (systematic error) of analyzer measurements relative to standard laboratory methods.
• Probationary and General Validation: Describes initial (probationary) validation using a limited sample set, progressing to comprehensive (general) validation over extended operating ranges once sufficient data are available.
• Quality Control Monitoring: Emphasizes the use of ongoing statistical control charts to track prediction deviations over time and trigger recalibration or model updates as needed.
• Material and Range Specificity: Validation applies only to the types and property ranges of materials used during the validation exercise, ensuring the reliability of results within the defined bounds.

Applications

ASTM D3764-23 is used extensively in:

• Initial commissioning of analyzer systems: Ensures newly installed systems are fit for purpose by validating measurement performance relative to established laboratory methods.
• Ongoing quality assurance: Supports continuous monitoring and validation to maintain measurement accuracy and detect drift or systematic errors in process stream analyzers.
• Acceptance testing: Assists purchasers in specifying performance benchmarks and acceptance criteria during system procurement and implementation.
• Industrial process control: Used in real-time or near-real-time monitoring of petroleum and petrochemical process streams for properties such as composition, purity, or additive content.
• Regulatory compliance: Facilitates adherence to industry, contractual, and regulatory requirements by demonstrating that analyzer outputs are statistically comparable to primary reference methods.
• Process system upgrades and maintenance: Applied after significant system changes or repairs to reaffirm performance consistency and suitability for continued operations.

Related Standards

To ensure comprehensive and effective validation of process stream analyzer systems, the following related ASTM standards are referenced and often used in conjunction with ASTM D3764-23:

• ASTM D6708: Statistical Assessment and Improvement of Expected Agreement Between Two Test Methods
• ASTM D6122: Validation of Multivariate Analyzer Systems (e.g., FTIR, NIR, Raman)
• ASTM D6299: Evaluating Analytical Measurement System Performance Using Quality Control Charts
• ASTM D7235: Establishing Linear Correlation Relationships Between Analyzer and Primary Test Method Results
• ASTM D7453: Sampling for Process Stream Analyzer System Validation
• ASTM E177: Use of Terms “Precision” and “Bias” in ASTM Test Methods

These standards ensure a robust framework for sampling, data analysis, system calibration, and ongoing quality assurance, maximizing the reliability and effectiveness of analyzer measurements in process environments.

---

Keywords: ASTM D3764-23, process stream analyzer validation, analyzer performance, precision and bias, statistical validation, analyzer system commissioning, quality control, petroleum analysis standards, process analyzer acceptance testing, validation procedures.

Standards Content (Sample)

This international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Principles for the
Development of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.
Designation: D3764 − 23
Standard Practice for
Validation of the Performance of Process Stream Analyzer
Systems
This standard is issued under the fixed designation D3764; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.
INTRODUCTION
Operation of a process stream analyzer system typically involves four sequential activities.
(1) Analyzer Calibration—When an analyzer is initially installed, or after major maintenance has
been performed, diagnostic testing is performed to demonstrate that the analyzer meets the
manufacturer’s specifications and historical performance standards. These diagnostic tests may require
that the analyzer be adjusted so as to provide predetermined output levels for certain reference
materials. (2a) Correlation for the Same Material—Once the diagnostic testing is completed,
process stream samples are analyzed using the analyzer system. For application where the process
analyzer system results are required to agree with results produced from an independent (primary) test
method (PTM), a mathematical function is derived that relates the analyzer results to the primary test
method results (PTMR). The application of this mathematical function to an analyzer result produces
a predicted primary test method result (PPTMR), for the same material. (2b) Correlation for
Material including Effect from Additional Treatment to the Material—The PPTMR in (2a) can be
used as an input to a mathematical model to predict the effect of an additive and/or a blendstock added
to a basestock material as measured by a PTM. (3) Probationary Validation—After the correlation(s)
relationship between the analyzer results and primary test method results has been established, a
probationary validation is performed using an independent but limited set of materials that were not
part of the correlation activity. This probationary validation is intended to demonstrate that the
PPTMRs agree with the PTMRs to within user-specified requirements for the analyzer system
application. (4) General and Continual Validation—After an adequate amount of PPTMRs and
PTMRs have been accrued on materials that were not part of the correlation activity, a comprehensive
statistical assessment is performed to demonstrate that the PPTMRs agree with the PTMRs to within
the tolerances established from the correlation activities. Subsequent to a successful general
validation, quality assurance control chart monitoring of the differences between PPTMR and PTMR
is conducted during normal operation of the process analyzer system to demonstrate that the
agreement between the PPTMRs and PTMRs established in the General Validation is maintained. This
practice deals with the third and fourth of these activities.
“Correlation for material including effect from additional treatment to the material” as outlined in
this standard is intended primarily to be applied to biofuels where the biofuel material is added at a
terminal or other facility and not included in the process stream material sampled by the analyzer at
the basestock manufacturing facility. The correlation shall be specific for a constant percentage
addition of the biofuels material to the basestock for each model. This practice may not apply for
physical properties where the source material for the biofuel material or the denaturant/diluent
material used with the biofuel material can significantly affect the finished biofuel’s physical property.
The user of the standard should investigate the effect of changes to biofuels material blend ratios,
biofuels material source material, and blendstock material composition when using this practice.
Limits to any of these may need to be applied when the correlation is used.
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States
D3764 − 23
1. Scope* ciple that is similar to the measurement principle of the primary
test method. This practice also applies if the process stream
1.1 This practice describes procedures and methodologies
analyzer system uses a different measurement technology from
based on the statistical principles of Practice D6708 to validate
the primary test method, provided that the calibration protocol
whether the degree of agreement between the results produced
for the direct output of the analyzer does not require use of the
by a total analyzer system (or its subsystem), versus the results
PTMRs (see Case 1 in Note 1).
produced by an independent test method that purports to
measure the same property, meets user-specified requirements.
1.6 This practice does not apply if the process stream
This is a performance-based validation, to be conducted using
analyzer system utilizes an indirect or mathematically modeled
a set of materials that are not used a priori in the development
measurement principle such as chemometric or multivariate
of any correlation between the two measurement systems under
analysis techniques where PTMRs are required for the chemo-
investigation. A result from the independent test method is
metric or multivariate model development. Users should refer
herein referred to as a Primary Test Method Result (PTMR).
to Practice D6122 for detailed validation procedures for these
1.1.1 The degree of agreement described in 1.1 can be either
types of analyzer systems (see Case 2 in Note 1).
for PPTMRs and PTMRs measured on the same materials, or
NOTE 1—For example, for the measurement of benzene in spark
for PPTMRs measured on basestocks and PTMRs measured on
ignition fuels, comparison of a Mid-Infrared process analyzer system
these same basestocks after constant level additivation.
based on Test Method D6277 to a Test Method D3606 gas chromatogra-
1.1.2 In some cases, a two-step procedure is employed. In
phy primary test method would be considered Case 1, and this practice
the first step, the analyzer and PTM are applied to the
would apply. For each sample, the Mid-Infrared spectrum is converted
measurement of the same blendstock material. If the analyzer
into a single analyzer result using methodology (Test Method D6277) that
employed in Step 1 is a multivariate spectrophotometric is independent of the primary test method (Test Method D3606). However,
when the same analyzer uses a multivariate model to correlate the
analyzer, then Practice D6122 is used to access the agreement
measured Mid-Infrared spectrum to Test Method D3606 reference values
between the PPTMRs and the PTMRs for this first step.
using the methodology of Practice D8321, it is considered Case 2 and
Otherwise, this practice is used to compare the PPTMRs to the
Practice D6122 applies. In this case 2 example, the direct output of the
PTMRs measured for this blendstock to determine the degree
analyzer is the spectrum, and the conversion of this multivariate output to
of agreement. In a second step, the PPTMRs produced in Step
an analyzer result require use of Practice D6122, hence it is not
1 are used as inputs to a second model that predicts the results independent of the primary test method.
obtained when the PTM is applied to the analysis of the
1.7 Performance Validation is conducted by calculating the
finished blended product. Since this second step does not use
precision and bias of the differences between results from the
analyzer readings, the validation of the second step is done
analyzer system (or subsystem) after the application of any
independently. Step 2 is only performed on valid Step 1 results.
necessary correlation, (such results are herein referred to as
Note that the second model might accommodate variable levels
Predicted Primary Test Method Results (PPTMRs)), versus the
or multiple material additions to the blendstock.
PTMRs for the same sample set. Results used in the calculation
1.2 This practice assumes any correlation necessary to
are for samples that are not used in the development of the
mitigate systemic biases between the analyzer system and PTM
correlation. The calculated precision and bias are statistically
have been applied to the analyzer results. See Guide D7235 for
compared to user-specified requirements for the analyzer
procedures for establishing such correlations.
system application.
1.3 This practice assumes any modeling techniques em-
1.7.1 For analyzers used in product release or product
ployed have the necessary tuning to mitigate systemic biases
quality certification applications, the precision and bias re-
between the analyzer PPTMR and PTMR have been applied to
quirement for the degree of agreement are typically based on
the model results. Model form and tuning is not covered by this
the site or published precision of the Primary Test Method.
practice, only the validation of the model output.
NOTE 2—In most applications of this type, the PTM is the specification-
1.4 This practice requires that both the primary method
cited test method.
against which the analyzer is compared to, and the analyzer
system under investigation, are in statistical control. Practices 1.7.2 This practice does not describe procedures for estab-
described in Practice D6299 should be used to ensure this
lishing precision and bias requirements for analyzer system
condition is met. applications. Such requirements must be based on the critical-
ity of the results to the intended business application and on
1.5 This practice applies if the process stream analyzer
contractual and regulatory requirements. The user must estab-
system and the primary test method are based on the same
lish precision and bias requirements prior to initiating the
measurement principle(s), or, if the process stream analyzer
validation procedures described herein.
system uses a direct and well-understood measurement prin-
1.8 Two procedures for validation are described: the line
This practice is under the jurisdiction of ASTM Committee D02 on Petroleum
sample procedure and the validation reference material (VRM)
Products, Liquid Fuels, and Lubricants and is the direct responsibility of Subcom-
injection procedure.
mittee D02.25 on Performance Assessment and Validation of Process Stream
Analyzer Systems.
1.9 Only the analyzer system or subsystem downstream of
Current edition approved July 1, 2023. Published July 2023. Originally approved
the VRM injection point or the line sample extraction point is
in 1980. Last previous edition approved in 2022 as D3764 – 22. DOI: 10.1520/
D3764-23. being validated by this practice.
D3764 − 23
1.10 The line sample procedure is limited to applications D3606 Test Method for Determination of Benzene and
where material can be safely withdrawn from the sampling Toluene in Spark Ignition Fuels by Gas Chromatography
point of the analyzer unit without significantly altering the D3700 Practice for Obtaining LPG Samples Using a Float-
property of interest. ing Piston Cylinder
1.10.1 The line sample procedure is the primary option for D4057 Practice for Manual Sampling of Petroleum and
when the validation is for (2b) materials including effect from Petroleum Products
additional treatment to the material. D4177 Practice for Automatic Sampling of Petroleum and
Petroleum Products
1.11 Validation information obtained in the application of
D5842 Practice for Sampling and Handling of Fuels for
this practice is applicable only to the type and property range
Volatility Measurement
of the materials used to perform the validation.
D6122 Practice for Validation of the Performance of Multi-
1.12 Two types of validation are described: General
variate Online, At-Line, Field and Laboratory Infrared
Validation, and Level Specific Validation. These are typically
Spectrophotometer, and Raman Spectrometer Based Ana-
conducted at installation or after major maintenance once the
lyzer Systems
system mechanical fitness-for-use has been established.
D6277 Test Method for Determination of Benzene in Spark-
1.12.1 General Validation is based on the statistical prin-
Ignition Engine Fuels Using Mid Infrared Spectroscopy
ciples and methodology of Practice D6708. In most cases,
D6299 Practice for Applying Statistical Quality Assurance
General Validation is preferred, but may not always be possible
and Control Charting Techniques to Evaluate Analytical
if the variation in validation materials is insufficient. General
Measurement System Performance
Validation will validate analyzer operation over a wider oper-
D6708 Practice for Statistical Assessment and Improvement
ating range than Level Specific Validation.
of Expected Agreement Between Two Test Methods that
1.12.2 When the variation in available validation materials
Purport to Measure the Same Property of a Material
is insufficient to satisfy the requirements of Practice D6708, a
D7235 Guide for Establishing a Linear Correlation Relation-
Level Specific Validation is done to validate analyzer operation
ship Between Analyzer and Primary Test Method Results
over a limited range.
Using Relevant ASTM Standard Practices
1.12.3 The validation outcome are considered valid only
D7278 Guide for Prediction of Analyzer Sample System Lag
within the range covered by the validation material Data from
Times
several different Validations (general or level-specific) can
D7453 Practice for Sampling of Petroleum Products for
potentially be combined for use in a General Validation.
Analysis by Process Stream Analyzers and for Process
Stream Analyzer System Validation
1.13 Procedures for the continual validation of system
performance are described. These procedures are typically D7808 Practice for Determining the Site Precision of a
Process Stream Analyzer on Process Stream Material
applied at a frequency commensurate with the criticality of the
application. D8009 Practice for Manual Piston Cylinder Sampling for
Volatile Crude Oils, Condensates, and Liquid Petroleum
1.14 This practice does not address procedures for diagnos-
Products
ing causes of validation failure.
D8321 Practice for Development and Validation of Multi-
1.15 This standard does not purport to address all of the
variate Analyses for Use in Predicting Properties of
safety concerns, if any, associated with its use. It is the
Petroleum Products, Liquid Fuels, and Lubricants based
responsibility of the user of this standard to establish appro-
on Spectroscopic Measurements
priate safety, health, and environmental practices and deter-
D8340 Practice for Performance-Based Qualification of
mine the applicability of regulatory limitations prior to use.
Spectroscopic Analyzer Systems
1.16 This international standard was developed in accor-
E177 Practice for Use of the Terms Precision and Bias in
dance with internationally recognized principles on standard-
ASTM Test Methods
ization established in the Decision on Principles for the
F307 Practice for Sampling Pressurized Gas for Gas Analy-
Development of International Standards, Guides and Recom-
sis
mendations issued by the World Trade Organization Technical
Barriers to Trade (TBT) Committee.
3. Terminology
2. Referenced Documents 3.1 Definitions:
3.1.1 accepted reference value (ARV), n—a value that serves
2.1 ASTM Standards:
as an agreed-upon reference for comparison, and which is
D1265 Practice for Sampling Liquefied Petroleum (LP)
derived as: (1) a theoretical or established value, based on
Gases, Manual Method
scientific principles, (2) an assigned or certified value, based on
experimental work of some national or international
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
organization, or (3) a consensus or certified value, based on
contact ASTM Customer Service at [e-mail]. For Annual Book of ASTM
collaborative experimental work under the auspices of a
Standards volume information, refer to the standard’s Document Summary page on
the ASTM website. scientific or engineering group. E177
D3764 − 23
3.1.2 analyzer, n—see analyzer system. appropriate bias-correction has been applied in accordance
with this practice; it is defined as the 95 % confidence limit for
3.1.3 analyzer system—for equipment in the analysis of
the difference between two such single and independent
liquid petroleum products and fuels, all piping, hardware,
results. D6708
computer, software, instrument, linear correlation or multivari-
3.1.5.1 Discussion—Within the context of this practice, R
XY
ate model required to analyze a process or product sample; the
is interpreted to be the 95 % confidence limit for the prediction
analyzer may also be referred to as the analyzer system, or the
deviation between any single Primary Test Method Result
total analyzer system.
(PTMR) and the Predicted Primary Test Method Result
3.1.3.1 Discussion—Online analyzers that utilize extractive
(PPTMR) produced by the analyzer system that is deemed
sampling include sample loop, sample conditioning system and
acceptable on the assumption that both the analyzer system and
excess sample return system (see Fig. 1 in D3764 for example).
primary test method are in statistical control, and that the
Online analyzers that utilize insertion probes include fiber
correlation relationship applied to the analyzer results to
optics and sample probes.
produce the PPTMR is fit-for-purpose.
3.1.3.2 Discussion—At-line, field and laboratory analyzers
include the instrument and all associated sample introduction 3.1.6 lag time, n—the time required for material to travel
apparatuses. from Point A to Point B in the total analyzer system (Points A
and B are user-defined).
3.1.4 analyzer unit response time, n—(see Fig. 2) the time
3.1.6.1 Discussion—Lag time is a function of an analyzer
interval between the introduction of a step change in property
system design parameters such as length and diameter of lines,
characteristic at the inlet of the analyzer unit and when the
number of fittings, flow restrictions, and the flow rate of the
analyzer output indicates a value corresponding to 99.5 % of
material (process or product stream) through the analyzer
the subsequent change in analyzer results.
system (see Figs. 2 and 1). See Guide D7278 for procedures for
3.1.4.1 Discussion—For continuous and intermittent analyz-
predicting analyzer system lag times.
ers with sufficiently short cycle times, the total analyzer
response time is the analyzer dead time plus 5.3 times the 3.1.7 line sample, n—process material that can be safely
analyzer unit time constant. For intermittent analyzers with withdrawn from a sample port or associated facilities without
long cycle times, the analyzer unit response time is effectively significantly altering the property of interest so that the
equal to the analyzer unit cycle time. For intermittent analyzers material can be used to perform analyzer system validation; the
with intermediate cycle times, the analyzer unit response time material is withdrawn in accordance with Practices D1265,
should be defined as the multiple of the analyzer unit cycle D3700, D4057, D4177, D5842, D7453, or D8009, whichever
time needed to exceed 99.5 % response. is applicable, during a period when the material flowing
through the analyzer is of uniform quality and the analyzer
3.1.5 between-method reproducibility (R ), n—a quantita-
XY
results are practically constant.
tive expression of the random error associated with the
difference between two results obtained by different operators 3.1.8 liquid petroleum product and fuels, n—in relation to
using different apparatus and applying the two methods X and process analyzers, any single-phase liquid material that is
Y, respectively, each obtaining a single result on an identical produced at a facility in the petroleum and petrochemical
test sample, when the methods have been assessed and an industries and will be in whole or in part of a petroleum
FIG. 1 Total Analyzer System
D3764 − 23
FIG. 2 Analyzer Time Units
product; it is inclusive of biofuels, renewable fuels, same sample of material, over an extended period of time
blendstocks, alternative blendstocks, and additives. D8340 spanning at least a 15 day interval. D6299
3.1.9 precision, n—the closeness of agreement between 3.1.17.1 Discussion—Site precision conditions should in-
independent test results obtained under stipulated conditions. clude all sources of variation that are typically encountered
E177
during normal, long term operation of the measurement sys-
tem. Thus, all operators who are involved in the routine use of
3.1.10 Predicted Primary Test Method Result(s) (PPTMR),
n—result(s) from the analyzer system, after application of any the measurement system should contribute results to the site
necessary correlation, that is interpreted as predictions of what
precision determination. In situations of high usage of a test
the primary test method results would have been, if it was
method where multiple QC results are obtained within a 24 h
conducted on the same material.
period, then only results separated by at least 4 h to 8 h,
depending on the absence of auto-correlation in the data, the
3.1.11 prediction deviation(s), (δ), n—calculated differ-
ence(s) (including algebraic sign) between predicted primary nature of the test method/instrument, site requirements, or
test method result and primary test result, defined as (PPTMR regulations, should be used in site precision calculations to
– PTMR).
reflect the longer term variation in the system. D6299
3.1.11.1 Discussion—This is also referred to as prediction
3.1.18 site precision, n—the value below which the absolute
residuals in Practice D6708.
difference between two individual test results obtained under
3.1.11.2 Discussion—Local validation in Practice D6122
site precision conditions is expected to exceed about 5 % of the
uses the absolute value of the prediction deviations, |δ|.
time (one case in 20 in the long run) in the normal and correct
3.1.12 primary test method (PTM), n—the analytical proce-
operation of the method.
dure used to generate the reference values against which the
3.1.18.1 Discussion—It is defind as 2.77 times σ , the
R’
analyzer is both calibrated and validated.
standard deviation of results obtained under site precision
3.1.13 primary test method result(s) (PTMR), n—test re-
conditions. D6299
sult(s) produced from an ASTM or other established standard
test method that are accepted as the reference measure of a 3.1.19 total analyzer system, n—see analyzer system.
property.
3.1.20 total analyzer system response time, n—(see Fig. 2)
3.1.14 repeatability conditions, n—conditions where inde-
the time interval between when a step change in property
pendent test results are obtained with the same method on
characteristic at the sample loop inlet and when the analyzer
identical test items in the same laboratory by the same operator
output indicates a value c corresponding to the 99.5 % of the
using the same equipment within short intervals of time. E177
subsequent change in analyzer results; the total analyzer
3.1.15 reproducibility conditions, n—conditions where test
system response time is the sum of the sample loop lag time,
results are obtained with the same method on identical test
the sample conditioning loop lag time, and the total analyzer
items in different laboratories with different operators using
response time.
different equipment. E177
3.1.21 validation, n—for equipment in the analysis of liquid
3.1.16 sample conditioning unit lag time, n—the time re-
petroleum products and fuels, the statistically quantified judg-
quired for material to travel from the start of the sample
ment that the analyzer system or subsystem, in conjunction
conditioning unit to the analyzer unit inlet.
with any correlation applied, can produce acceptable precision
3.1.17 site precision conditions, n—conditions under which
and bias performance on the prediction deviations (δ for
test results are obtained by one or more operators in a single
materials that were not used to develop the correlation).
site location practicing the same test method on a single
3.2 Definitions of Terms Specific to This Standard:
measurement system which may comprise multiple
instruments, using test specimens taken at random from the 3.2.1 Analyzer System Items:
D3764 − 23
3.2.1.1 analyzer output, n—a signal (pneumatic, electrical, reference value or site assigned value can be calculated or
or digital), proportional to the property being measured that is measured.
suitable for readout or control instrumentation external to the (1) Discussion—A composition-specific VRM may be a
analyzer system. commercial standard reference material (SRM) having a cer-
tified accepted reference value.
3.2.1.2 analyzer system result, n—the measured property
reading, in the accepted property measurement units, that is 3.2.3.2 continual validation, n—the quality assurance pro-
cess by which the bias and precision performance determined
displayed by the analyzer unit readout instrumentation or
transmitted to end user of the analyzer system. during initial validation are shown to be sustained.
3.2.3.3 direct measurement, n—a quantitative measurement
3.2.1.3 analyzer unit, n—the instrumental equipment neces-
result obtained using a principle or principles that express the
sary to automatically measure the physical or chemical prop-
characteristic property of interest in its defining units.
erty of a process or product stream sample using either an
intermittent or a continuous technique.
3.2.3.4 indirect measurement, n—a correlated quantitative
measurement result obtained using a measurement principle
3.2.1.4 analyzer unit repeatability, n—2.77 times the stan-
that produces values that do not express the desired character-
dard deviation of results obtained from repetitive analysis of
istic property but which can be modified empirically, using
the same material directly injected into the analyzer unit under
mathematical modeling techniques, to estimate the necessary
repeatability conditions.
defining units of the property of interest.
3.2.1.5 continuous analyzer unit, n—an analyzer that mea-
(1) Discussion—Methods that utilize chemometric or multi-
sures the property value of a process or product stream on a
variate analysis are indirect measurements for generating
continuous basis and dynamically displays the instantaneously
correlative characteristic property measurement results.
updated analyzer output.
3.2.3.5 process-derived VRM, n—a validation reference ma-
3.2.1.6 intermittent analyzer unit, n—a cyclic type analyzer
terial derived from an isolated batch of process or product
that performs a measurement sequence on samples from a
stream material with chemical or physical characteristics, or
process or product stream and displays a new analyzer output
both, that is suitable for determination of an accepted reference
at the conclusion of each cycle.
value or site assigned value for the property of interest.
3.2.2 Time Unit Items—General Terms:
3.2.3.6 site assigned value (SAV), n—a property value of a
3.2.2.1 analyzer unit cycle time, n—for intermittent
reference material that is based on multiple results from either
analyzers, the time interval between successive updates of the
the analyzer unit or a primary test method, obtained under site
analyzer output.
precision conditions.
3.2.2.2 analyzer unit dead time, n—the time interval be-
3.2.3.7 validation reference material (VRM), n—for valida-
tween the introduction of a step change in property character-
tion and quality assurance testing, a material having an
istic at the inlet of the analyzer unit and the initial indication of
accepted reference value or site assigned value for the property
analyzer response to this change.
of interest.
(1) Discussion—For intermittent analyzers, if the analyzer
dead time is less than one analyzer unit cycle time, the analyzer
4. Summary of Practice
unit dead time cannot be directly measured.
4.1 PPTMRs from the total analyzer system or its subsystem
3.2.2.3 analyzer unit response time, n—(see Fig. 2) the time
are compared to the corresponding PTMRs on at least 15
interval between the introduction of a step change in property
materials. PPTMR and PTMR are statistically assessed relative
characteristic at the inlet of the analyzer unit and when the
to each other using the methodology of Practice D6708,
analyzer output indicates a value corresponding to 99.5 % of
recognizing that this is only a preliminary Practice D6708
the subsequent change in analyzer results.
assessment. Precision and bias statistics on the prediction
3.2.2.4 analyzer unit time constant, n—(see Fig. 2) the time deviations (δ) are generated and the bias is assessed against
interval between the initial response of the analyzer unit to a pre-specified performance criteria. The system or subsystem
step change in property characteristic and when the analyzer performance is considered to be probationary validated for
output indicates a value corresponding to 63 % of the subse- materials and property ranges representative of those used in
quent change in analyzer results. the validation if the prediction deviations are in statistical
(1) Discussion—For intermittent analyzers, if the analyzer control, and bias performance statistic meets pre-specified
unit time constant is less than one analyzer unit cycle time, the criterion.
analyzer time constant cannot be directly measured.
4.2 After probationary validation is achieved, continued
3.2.2.5 sample loop lag time, n—the time required for
statistical quality control chart monitoring and analyses on δ
material to travel from the process takeoff point of the sample are carried out with new production samples to ensure on-
loop to start of the sample conditioning unit.
going prediction performance of the PPTMR meets the levels
established from the probationary validation.
3.2.3 General Terms:
3.2.3.1 composition-specific VRM, n—a validation reference 4.3 Once the total number of samples with completed
material consisting of a single, pure compound, or a known, datasets (PPTMR , PTMR , δ) from probationary and continual
reproducible mixture of compounds for which an accepted validation reaches 30, a general validation is conducted using
D3764 − 23
the statistical methodology of Practice D6708. The objective of pure compounds or simple mixtures of pure compounds may
the general validation is to demonstrate performance with at not be representative of that achieved on actual process or
least 30 samples over a wider operating envelope, or, to product samples.
confirm outcome from probationary validation with more
accrued data.
6. System Components
4.4 If the variation among the 30 samples is inadequate to 6.1 Fig. 1 illustrates a total analyzer system incorporating a
conduct the Practice D6708 assessment, a level specific vali-
selection and arrangement of components that are typical but
dation may be performed to validate the agreement between not specific for any particular analyzer system. A total analyzer
PPTMR and PTMR over a narrow operating range. As addi-
system design addresses the chemical and physical properties
tional (PPTMR / PTMR / δ) datasets are collected covering a of the process or product stream to be measured, provides a
wider operating range, the general validation may again be
representative sample, and handles it without adversely affect-
attempted. ing the value of the specific property of interest. Included are
a sample loop, piping, hardware, a sampling port, sample
4.5 After general validation has been achieved, continue to
conditioning devices, an analyzer unit instrumentation, any
monitor δ using statistical quality control charts at a frequency
data analysis computer hardware and software, and a readout
commensurate with the criticality of the application.
display.
5. Significance and Use 6.2 Sample Loop—Piping connected to the main process
stream to deliver a portion of the stream to a location close to
5.1 This practice can be used to quantify the performance of
the analyzer system with minimum lag time and return the
a process stream analyzer system or its subsystem in terms of
unused material to the main process stream.
precision and bias relative to those of a primary test method for
the property of interest. 6.3 Sampling System—Sample probes, valves, lines,
containers, pressure regulator, and gages that constitute the
5.2 This practice provides developers or manufacturers of
equipment employed to obtain a proper sample from the
process stream analyzer systems with useful procedures for
sample loop and introduce either it or a validation standard
evaluating the capability of newly designed systems for indus-
sample to the analyzer.
trial applications that require reliable prediction of measure-
ments of a specific property by a primary test method of a 6.4 Sample Conditioning Unit—A collection of devices to
flowing component or product. properly treat a portion of the sample from the sample loop so
that it meets the requirements for testing by the process
5.3 This practice provides purchasers of process stream
analyzer. These components can incorporate temperature or
analyzer systems with some reliable options for specifying
pressure adjustment, change of state (liquid, vapor), or removal
acceptance test requirements for process stream analyzer
of contaminants.
systems at the time of commissioning to ensure the system is
capable of making the desired property measurement with the 6.5 Inlet Port—Appropriate piping with selector valve(s) for
appropriate precision or bias specifications, or both. placement either at the inlet to the analyzer unit or, when
dictated by the measurement specifications, at the inlet to the
5.4 PPTMR from Analyzer Systems validated in accordance
sample conditioning unit. The purpose of this inlet port is to
with this practice can be used to predict, with a specified
allow injection of validation standards or other calibration
confidence, what the PTMR would be, to within a specified
material into the analyzer system with quick switching between
tolerance, if the actual primary test method was conducted on
these typically containerized materials and the flowing process
the materials that are within the validated property range and
stream.
type.
6.5.1 For many analyzer systems the inlet port requires a
5.5 This practice provides the user of a process stream
manifold arrangement for validation or quality assurance
analyzer system with useful information from on-going quality
studies. Such a manifold, with suitable valving, provides a
control charts to monitor the variation in δ over time, and
means to use a containerized supply of standby material when
trigger update of correlation relationship between the analyzer
a flowing process stream is not available for the purpose. It also
system and primary test method in a timely manner.
permits quick switching between different validation standards
when that is desirable.
5.6 Validation information obtained in the application of
this practice is applicable only to the material type and property
6.6 Sample Port—An appropriate probe or fitting in the
range of the materials used to perform the validation. Selection
piping to permit collection of representative samples for
of the property levels and the compositional characteristics of
laboratory analyses using a primary test method.
the samples must be suitable for the application of the analyzer
6.7 Analyzer Unit—Instrumentation designed to automati-
system. This practice allows the user to write a comprehensive
cally measure the chemical or physical property of a process or
validation statement for the analyzer system including specific
product stream sample and provide either an intermittent or a
limits for the validated range of application. This practice does
continuous output signal representing the measurement result.
not recommend extrapolation of validation results beyond the
material type and property range used to obtain these results. In 6.8 Readout Instrumentation—If it is not an integral com-
addition, users are cautioned that for measurement systems that ponent of the analyzer system, a device to display or record or
show matrix dependencies, bias information determined from both, the property measurement analyzer result.
D3764 − 23
7. Preparation of Analyzer System 9.1.3 This practice recommends articulation of precision
performance of δ as a between-method reproducibility (R ).
XY
7.1 Implementation of this practice requires that the process
stream analyzer system operates under conditions specified: 9.2 The line sample procedure directly fulfills the validation
objective since the validation results for both the process
7.1.1 Meets all applicable electrical and safety codes.
system and the primary test method are obtained on process
7.1.2 Meets the supplier’s recommendation.
samples. However, if line samples covering the composition
7.1.3 Complies with operating conditions specified by the
and property range of interest cannot be acquired within a
manufacturer.
reasonable length of time once the validation process begins,
7.1.4 Includes a predicted PTM algorithm, if necessary.
consider using either process-derived or composition-specific
7.2 After installation or major maintenance, conduct such
validation reference materials (VRMs) to extend the composi-
diagnostic tests as recommended by the manufacturer to
tion and property range of the validation sample set. A suitable
demonstrate that the analyzer meets the manufacturer’s speci-
process-derived VRM may simply be a batch of material
fications or historical performance levels, or both. If necessary,
obtained at a time prior to the start of the validation procedure
adjust the analyzer system components so as to obtain recom-
but that was not used in calibrating either the analyzer or the
mended analyzer output levels for specified reference materi-
primary test method. In general, the composition of a VRM
als.
used for validation should be similar to a composition that is
7.3 Inspect the entire analyzer system to ensure it is in-
anticipated for the process stream at some future time.
stalled properly, is in operating condition, and is properly
9.2.1 In cases where it is necessary to include the sample
adjusted after completion of the initial commissioning proce-
loop or the sample conditioning unit (Fig. 1), or both, in the
dures.
validation procedure, VRMs should not be used to the exclu-
sion of lines sample unless it is practical to use the VRMs to
7.4 Application of Practice D6708 mathematics requires
validate both sample system and analyzer (this is generally not
that the site precision of the analyzer system be known. If the
practical). The sample system can be excluded from the
analyzer site precision is not known, then it should be
validation procedure if it is known that the sample system does
determined prior to initiation of the validation procedure.
not materially alter the composition or condition of the sample
Methodology described in Practice D6708 can be employed.
presented to the analyzer and if the sample system response
time can be estimated with reasonable certainty. Guidance on
8. Pre-Validation Analyzer Calibration Check
how to meet these conditions is beyond the intended scope of
8.1 When an analyzer is initially installed, and after major
this practice. If these conditions cannot be met and if VRMs are
maintenance has been preformed, diagnostic tests should be
needed to extend the property and composition range of the
conducted to demonstrate that the analyzer meets manufactur-
validation set, it is recommended that the user conduct two
er’s specifications and historical performance standards. These
probationary validations, one using line samples and the other
diagnostic tests may require that the analyzer be adjusted so as
using VRMs, to demonstrate that VRM procedure adequately
to provide predetermined output levels for certain reference
reflects corresponding performance for actual process materi-
materials. Such adjustment may be done in hardware, software
als. Once demonstrated, the statistical quality control charting
or both.
for continual validation can be done using VRM procedures,
with a periodic line sample procedure mixed in over time to
8.2 Description of specific calibration procedures for the
demonstrate that both procedures continue to provide similar
numerous analyzer types is beyond the scope of this practice.
and acceptable performance.
9. Validation Procedure
NOTE 3—If the process analyzer system is not based on identically the
same measurement principle as the primary test method, then the analyzer
9.1 The objective of the validation procedures is to quantify
system may react differently to variations in the sample matrix than does
the precision and bias performance of prediction deviations (δ)
the primary test method. In such case, analyzer results for process samples
between PPTMR produced by a process stream analyzer
might be biased relative to primary test method results even when the
VRM procedure results shown no such bias unless the VRM is process-
system (or its subsystem) versus PTMR for materials spanning
derived. The bias can be minimized by using a process stream (test)
the intended operating range for the analyzer system. The user
sample for which an ARV or SAV was determined as the VRM. The test
must specify acceptable precision and bias performance criteria
sample used in this fashion should be representative of the current process
before initiating the validation. These criteria will be depen-
stream.
dent on the intended use of the analyzer.
NOTE 4—If, due to differences in sample pretreatment, the sample
analyzed by the process stream analyzer and the sample analyzed by the
9.1.1 For analyzer systems used in product certification,
primary test method are not identically the same, then the use of the VRM
precision performance acceptance criteria for δ will typically
procedure may not accurately reflect agreement between the process
be based directly on the published reproducibility (R) of the
analyzer and the primary test method. The VRM may not be affected in the
primary test method. Bias criteria will typically be based on
same manner as process samples by the different sample pretreatments.
regulatory or contractual requirements. It is a general perfor-
Again, this effect can be minimized by using current process stream (test)
samples as VRMs.
mance expectation that no bias correction can further improve
the precision of δ statistically.
9.3 Probationary, General and Level Specific Validation
9.1.2 For analyzer systems used in other types of service, using the Line Sample Procedure:
precision and bias criteria must be developed based on the 9.3.1 This procedure is applicable for analyzer systems that
intended use of the analyzer results. are equipped with sample ports anywhere within the system
D3764 − 23
that can facilitate the safe collection of material intended for is unknown, the repeatability of the primary test method can be
analysis by the analyzer unit without significantly altering the used as the reference for data comparison.
property of interest. The subsystem from the sample port up to
9.3.3.2 After steady state has been verified, begin collecting
and including the analyzer subsystem (see Fig. 1) is considered
the process line sample from the sample port. Refer to
to be validated for current process stream samples if the δ
Practices D1265, D4057, D4177, D5842, D7453, or F307 for
results are in statistical control, and the precision and bias
procedures for sample collection. Record the time, t , corre-
s
statistics meet user-specified requirements.
sponding to the start of sample collection. Record the analyzer
9.3.2 Line Sample Procedure Requirements:
system result A (t ) observed at t . Collect the volume of
0 s s
9.3.2.1 Select point of line sample withdrawal.
sample required for PTM analysis. Record the time, t , when
e
9.3.2.2 Determine the total lag time of the system or
sample collection ends.
subsystem from the sample withdrawal point (see Figs. 2 and
9.3.3.3 If the sample collection interval t – t is less than the
e s
1 for guidance) up to and including the analyzer.
lag time o
...

This document is not an ASTM standard and is intended only to provide the user of an ASTM standard an indication of what changes have been made to the previous version. Because
it may not be technically possible to adequately depict all changes accurately, ASTM recommends that users consult prior editions as appropriate. In all cases only the current version
of the standard as published by ASTM is to be considered the official document.
Designation: D3764 − 22 D3764 − 23
Standard Practice for
Validation of the Performance of Process Stream Analyzer
Systems
This standard is issued under the fixed designation D3764; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.
INTRODUCTION
Operation of a process stream analyzer system typically involves four sequential activities.
(1) Analyzer Calibration—When an analyzer is initially installed, or after major maintenance has
been performed, diagnostic testing is performed to demonstrate that the analyzer meets the
manufacturer’s specifications and historical performance standards. These diagnostic tests may require
that the analyzer be adjusted so as to provide predetermined output levels for certain reference
materials. (2a) Correlation for the Same Material—Once the diagnostic testing is completed,
process stream samples are analyzed using the analyzer system. For application where the process
analyzer system results are required to agree with results produced from an independent (primary) test
method (PTM), a mathematical function is derived that relates the analyzer results to the primary test
method results (PTMR). The application of this mathematical function to an analyzer result produces
a predicted primary test method result (PPTMR), for the same material. (2b) Correlation for
Material including Effect from Additional Treatment to the Material—The PPTMR in (2a) can be
used as an input to a mathematical model to predict the effect of an additive and/or a blendstock added
to a basestock material as measured by a PTM. (3) Probationary Validation—After the correlation(s)
relationship between the analyzer results and primary test method results has been established, a
probationary validation is performed using an independent but limited set of materials that were not
part of the correlation activity. This probationary validation is intended to demonstrate that the
PPTMRs agree with the PTMRs to within user-specified requirements for the analyzer system
application. (4) General and Continual Validation—After an adequate amount of PPTMRs and
PTMRs have been accrued on materials that were not part of the correlation activity, a comprehensive
statistical assessment is performed to demonstrate that the PPTMRs agree with the PTMRs to within
the tolerances established from the correlation activities. Subsequent to a successful general
validation, quality assurance control chart monitoring of the differences between PPTMR and PTMR
is conducted during normal operation of the process analyzer system to demonstrate that the
agreement between the PPTMRs and PTMRs established in the General Validation is maintained. This
practice deals with the third and fourth of these activities.
“Correlation for material including effect from additional treatment to the material” as outlined in
this standard is intended primarily to be applied to biofuels where the biofuel material is added at a
terminal or other facility and not included in the process stream material sampled by the analyzer at
the basestock manufacturing facility. The correlation shall be specific for a constant percentage
addition of the biofuels material to the basestock for each model. This practice may not apply for
physical properties where the source material for the biofuel material or the denaturant/diluent
material used with the biofuel material can significantly affect the finished biofuel’s physical property.
The user of the standard should investigate the effect of changes to biofuels material blend ratios,
This practice is under the jurisdiction of ASTM Committee D02 on Petroleum Products, Liquid Fuels, and Lubricants and is the direct responsibility of Subcommittee
D02.25 on Performance Assessment and Validation of Process Stream Analyzer Systems.
Current edition approved April 1, 2022July 1, 2023. Published June 2022July 2023. Originally approved in 1980. Last previous edition approved in 20192022 as
D3764 – 19.D3764 – 22. DOI: 10.1520/D3764-22.10.1520/D3764-23.
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States
D3764 − 23
biofuels material source material, and blendstock material composition when using this practice.
Limits to any of these may need to be applied when the correlation is used.
1. Scope*
1.1 This practice describes procedures and methodologies based on the statistical principles of Practice D6708 to validate whether
the degree of agreement between the results produced by a total analyzer system (or its subsystem), versus the results produced
by an independent test method that purports to measure the same property, meets user-specified requirements. This is a
performance-based validation, to be conducted using a set of materials that are not used a priori in the development of any
correlation between the two measurement systems under investigation. A result from the independent test method is herein referred
to as a Primary Test Method Result (PTMR).
1.1.1 The degree of agreement described in 1.1 can be either for PPTMRs and PTMRs measured on the same materials, or for
PPTMRs measured on basestocks and PTMRs measured on these same basestocks after constant level additivation.
1.1.2 In some cases, a two-step procedure is employed. In the first step, the analyzer and PTM are applied to the measurement
of the same blendstock material. If the analyzer employed in Step 1 is a multivariate spectrophotometric analyzer, then Practice
D6122 is used to access the agreement between the PPTMRs and the PTMRs for this first step. Otherwise, this practice is used
to compare the PPTMRs to the PTMRs measured for this blendstock to determine the degree of agreement. In a second step, the
PPTMRs produced in Step 1 are used as inputs to a second model that predicts the results obtained when the PTM is applied to
the analysis of the finished blended product. Since this second step does not use analyzer readings, the validation of the second
step is done independently. Step 2 is only performed on valid Step 1 results. Note that the second model might accommodate
variable levels or multiple material additions to the blendstock.
1.2 This practice assumes any correlation necessary to mitigate systemic biases between the analyzer system and PTM have been
applied to the analyzer results. See Guide D7235 for procedures for establishing such correlations.
1.3 This practice assumes any modeling techniques employed have the necessary tuning to mitigate systemic biases between the
analyzer PPTMR and PTMR have been applied to the model results. Model form and tuning is not covered by this practice, only
the validation of the model output.
1.4 This practice requires that both the primary method against which the analyzer is compared to, and the analyzer system under
investigation, are in statistical control. Practices described in Practice D6299 should be used to ensure this condition is met.
1.5 This practice applies if the process stream analyzer system and the primary test method are based on the same measurement
principle(s), or, if the process stream analyzer system uses a direct and well-understood measurement principle that is similar to
the measurement principle of the primary test method. This practice also applies if the process stream analyzer system uses a
different measurement technology from the primary test method, provided that the calibration protocol for the direct output of the
analyzer does not require use of the PTMRs (see Case 1 in Note 1).
1.6 This practice does not apply if the process stream analyzer system utilizes an indirect or mathematically modeled measurement
principle such as chemometric or multivariate analysis techniques where PTMRs are required for the chemometric or multivariate
model development. Users should refer to Practice D6122 for detailed validation procedures for these types of analyzer systems
(see Case 2 in Note 1).
NOTE 1—For example, for the measurement of benzene in spark ignition fuels, comparison of a Mid-Infrared process analyzer system based on Test
Method D6277 to a Test Method D3606 gas chromatography primary test method would be considered Case 1, and this practice would apply. For each
sample, the Mid-Infrared spectrum is converted into a single analyzer result using methodology (Test Method D6277) that is independent of the primary
test method (Test Method D3606). However, when the same analyzer uses a multivariate model to correlate the measured Mid-Infrared spectrum to Test
Method D3606 reference values using the methodology of Practice E1655D8321, it is considered Case 2 and Practice D6122 applies. In this case 2
example, the direct output of the analyzer is the spectrum, and the conversion of this multivariate output to an analyzer result require use of Practice
D6122, hence it is not independent of the primary test method.
1.7 Performance Validation is conducted by calculating the precision and bias of the differences between results from the analyzer
system (or subsystem) after the application of any necessary correlation, (such results are herein referred to as Predicted Primary
Test Method Results (PPTMRs)), versus the PTMRs for the same sample set. Results used in the calculation are for samples that
D3764 − 23
are not used in the development of the correlation. The calculated precision and bias are statistically compared to user-specified
requirements for the analyzer system application.
1.7.1 For analyzers used in product release or product quality certification applications, the precision and bias requirement for the
degree of agreement are typically based on the site or published precision of the Primary Test Method.
NOTE 2—In most applications of this type, the PTM is the specification-cited test method.
1.7.2 This practice does not describe procedures for establishing precision and bias requirements for analyzer system applications.
Such requirements must be based on the criticality of the results to the intended business application and on contractual and
regulatory requirements. The user must establish precision and bias requirements prior to initiating the validation procedures
described herein.
1.8 Two procedures for validation are described: the line sample procedure and the validation reference material (VRM) injection
procedure.
1.9 Only the analyzer system or subsystem downstream of the VRM injection point or the line sample extraction point is being
validated by this practice.
1.10 The line sample procedure is limited to applications where material can be safely withdrawn from the sampling point of the
analyzer unit without significantly altering the property of interest.
1.10.1 The line sample procedure is the primary option for when the validation is for (2b) materials including effect from
additional treatment to the material.
1.11 Validation information obtained in the application of this practice is applicable only to the type and property range of the
materials used to perform the validation.
1.12 Two types of validation are described: General Validation, and Level Specific Validation. These are typically conducted at
installation or after major maintenance once the system mechanical fitness-for-use has been established.
1.12.1 General Validation is based on the statistical principles and methodology of Practice D6708. In most cases, General
Validation is preferred, but may not always be possible if the variation in validation materials is insufficient. General Validation
will validate analyzer operation over a wider operating range than Level Specific Validation.
1.12.2 When the variation in available validation materials is insufficient to satisfy the requirements of Practice D6708, a Level
Specific Validation is done to validate analyzer operation over a limited range.
1.12.3 The validation outcome are considered valid only within the range covered by the validation material Data from several
different Validations (general or level-specific) can potentially be combined for use in a General Validation.
1.13 Procedures for the continual validation of system performance are described. These procedures are typically applied at a
frequency commensurate with the criticality of the application.
1.14 This practice does not address procedures for diagnosing causes of validation failure.
1.15 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility
of the user of this standard to establish appropriate safety, health, and environmental practices and determine the applicability of
regulatory limitations prior to use.
1.16 This international standard was developed in accordance with internationally recognized principles on standardization
established in the Decision on Principles for the Development of International Standards, Guides and Recommendations issued
by the World Trade Organization Technical Barriers to Trade (TBT) Committee.
D3764 − 23
2. Referenced Documents
2.1 ASTM Standards:
D1265 Practice for Sampling Liquefied Petroleum (LP) Gases, Manual Method
D3606 Test Method for Determination of Benzene and Toluene in Spark Ignition Fuels by Gas Chromatography
D3700 Practice for Obtaining LPG Samples Using a Floating Piston Cylinder
D4057 Practice for Manual Sampling of Petroleum and Petroleum Products
D4177 Practice for Automatic Sampling of Petroleum and Petroleum Products
D5842 Practice for Sampling and Handling of Fuels for Volatility Measurement
D6122 Practice for Validation of the Performance of Multivariate Online, At-Line, Field and Laboratory Infrared
Spectrophotometer, and Raman Spectrometer Based Analyzer Systems
D6277 Test Method for Determination of Benzene in Spark-Ignition Engine Fuels Using Mid Infrared Spectroscopy
D6299 Practice for Applying Statistical Quality Assurance and Control Charting Techniques to Evaluate Analytical Measure-
ment System Performance
D6708 Practice for Statistical Assessment and Improvement of Expected Agreement Between Two Test Methods that Purport
to Measure the Same Property of a Material
D7235 Guide for Establishing a Linear Correlation Relationship Between Analyzer and Primary Test Method Results Using
Relevant ASTM Standard Practices
D7278 Guide for Prediction of Analyzer Sample System Lag Times
D7453 Practice for Sampling of Petroleum Products for Analysis by Process Stream Analyzers and for Process Stream Analyzer
System Validation
D7808 Practice for Determining the Site Precision of a Process Stream Analyzer on Process Stream Material
D8009 Practice for Manual Piston Cylinder Sampling for Volatile Crude Oils, Condensates, and Liquid Petroleum Products
D8321 Practice for Development and Validation of Multivariate Analyses for Use in Predicting Properties of Petroleum
Products, Liquid Fuels, and Lubricants based on Spectroscopic Measurements
D8340 Practice for Performance-Based Qualification of Spectroscopic Analyzer Systems
E177 Practice for Use of the Terms Precision and Bias in ASTM Test Methods
E1655 Practices for Infrared Multivariate Quantitative Analysis
F307 Practice for Sampling Pressurized Gas for Gas Analysis
3. Terminology
3.1 Definitions:
3.1.1 accepted reference value (ARV), n—a value that serves as an agreed-upon reference for comparison, and which is derived
as: (1) a theoretical or established value, based on scientific principles, (2) an assigned or certified value, based on experimental
work of some national or international organization, or (3) a consensus or certified value, based on collaborative experimental work
under the auspices of a scientific or engineering group. E177
3.1.2 analyzer, n—see analyzer system.
3.1.3 analyzer system—for equipment in the analysis of liquid petroleum products and fuels, all piping, hardware, computer,
software, instrument, linear correlation or multivariate model required to analyze a process or product sample; the analyzer may
also be referred to as the analyzer system, or the total analyzer system.
3.1.3.1 Discussion—
Online analyzers that utilize extractive sampling include sample loop, sample conditioning system and excess sample return system
(see Fig. 1 in D3764 for example). Online analyzers that utilize insertion probes include fiber optics and sample probes.
3.1.3.2 Discussion—
At-line, field and laboratory analyzers include the instrument and all associated sample introduction apparatuses.
3.1.4 analyzer unit response time, n—(see Fig. 2) the time interval between the introduction of a step change in property
characteristic at the inlet of the analyzer unit and when the analyzer output indicates a value corresponding to 99.5 % of the
subsequent change in analyzer results.
3.1.4.1 Discussion—
For continuous and intermittent analyzers with sufficiently short cycle times, the total analyzer response time is the analyzer dead
For referenced ASTM standards, visit the ASTM website, www.astm.org, or contact ASTM Customer Service at [e-mail]. For Annual Book of ASTM Standards
volume information, refer to the standard’s Document Summary page on the ASTM website.
D3764 − 23
FIG. 1 Total Analyzer System
FIG. 2 Analyzer Time Units
time plus 5.3 times the analyzer unit time constant. For intermittent analyzers with long cycle times, the analyzer unit response
time is effectively equal to the analyzer unit cycle time. For intermittent analyzers with intermediate cycle times, the analyzer unit
response time should be defined as the multiple of the analyzer unit cycle time needed to exceed 99.5 % response.
3.1.5 between-method reproducibility (R ), n—a quantitative expression of the random error associated with the difference
XY
between two results obtained by different operators using different apparatus and applying the two methods X and Y, respectively,
each obtaining a single result on an identical test sample, when the methods have been assessed and an appropriate bias-correction
has been applied in accordance with this practice; it is defined as the 95 % confidence limit for the difference between two such
single and independent results. D6708
3.1.5.1 Discussion—
Within the context of this practice, R is interpreted to be the 95 % confidence limit for the prediction deviation between any
XY
single Primary Test Method Result (PTMR) and the Predicted Primary Test Method Result (PPTMR) produced by the analyzer
system that is deemed acceptable on the assumption that both the analyzer system and primary test method are in statistical control,
and that the correlation relationship applied to the analyzer results to produce the PPTMR is fit-for-purpose.
3.1.6 lag time, n—the time required for material to travel from Point A to Point B in the total analyzer system (Points A and B
are user-defined).
3.1.6.1 Discussion—
Lag time is a function of an analyzer system design parameters such as length and diameter of lines, number of fittings, flow
restrictions, and the flow rate of the material (process or product stream) through the analyzer system (see Figs. 2 and 1). See Guide
D7278 for procedures for predicting analyzer system lag times.
D3764 − 23
3.1.7 line sample, n—process material that can be safely withdrawn from a sample port or associated facilities without significantly
altering the property of interest so that the material can be used to perform analyzer system validation; the material is withdrawn
in accordance with Practices D1265, D3700, D4057, D4177, D5842, D7453, or D8009, whichever is applicable, during a period
when the material flowing through the analyzer is of uniform quality and the analyzer results are practically constant.
3.1.8 liquid petroleum product and fuels, n—in relation to process analyzers, any single-phase liquid material that is produced at
a facility in the petroleum and petrochemical industries and will be in whole or in part of a petroleum product; it is inclusive of
biofuels, renewable fuels, blendstocks, alternative blendstocks, and additives. D8340
3.1.9 precision, n—the closeness of agreement between independent test results obtained under stipulated conditions. E177
3.1.10 Predicted Primary Test Method Result(s) (PPTMR), n—result(s) from the analyzer system, after application of any
necessary correlation, that is interpreted as predictions of what the primary test method results would have been, if it was conducted
on the same material.
3.1.11 prediction deviation(s), (δ), n—calculated difference(s) (including algebraic sign) between predicted primary test method
result and primary test result, defined as (PPTMR – PTMR).
3.1.11.1 Discussion—
This is also referred to as prediction residuals in Practice D6708.
3.1.11.2 Discussion—
Local validation in Practice D6122 uses the absolute value of the prediction deviations, |δ|.
3.1.12 primary test method (PTM), n—the analytical procedure used to generate the reference values against which the analyzer
is both calibrated and validated.
3.1.13 primary test method result(s) (PTMR), n—test result(s) produced from an ASTM or other established standard test method
that are accepted as the reference measure of a property.
3.1.14 repeatability conditions, n—conditions where independent test results are obtained with the same method on identical test
items in the same laboratory by the same operator using the same equipment within short intervals of time. E177
3.1.15 reproducibility conditions, n—conditions where test results are obtained with the same method on identical test items in
different laboratories with different operators using different equipment. E177
3.1.16 sample conditioning unit lag time, n—the time required for material to travel from the start of the sample conditioning unit
to the analyzer unit inlet.
3.1.17 site precision conditions, n—conditions under which test results are obtained by one or more operators in a single site
location practicing the same test method on a single measurement system which may comprise multiple instruments, using test
specimens taken at random from the same sample of material, over an extended period of time spanning at least a 15 day interval.
D6299
3.1.17.1 Discussion—
Site precision conditions should include all sources of variation that are typically encountered during normal, long term operation
of the measurement system. Thus, all operators who are involved in the routine use of the measurement system should contribute
results to the site precision determination. In situations of high usage of a test method where multiple QC results are obtained
within a 24 h period, then only results separated by at least 4 h to 8 h, depending on the absence of auto-correlation in the data,
the nature of the test method/instrument, site requirements, or regulations, should be used in site precision calculations to reflect
the longer term variation in the system. D6299
3.1.18 site precision, n—the value below which the absolute difference between two individual test results obtained under site
precision conditions is expected to exceed about 5 % of the time (one case in 20 in the long run) in the normal and correct operation
of the method.
3.1.18.1 Discussion—
It is defind as 2.77 times σ , the standard deviation of results obtained under site precision conditions. D6299
R’
D3764 − 23
3.1.19 total analyzer system, n—see analyzer system.
3.1.20 total analyzer system response time, n—(see Fig. 2) the time interval between when a step change in property characteristic
at the sample loop inlet and when the analyzer output indicates a value c corresponding to the 99.5 % of the subsequent change
in analyzer results; the total analyzer system response time is the sum of the sample loop lag time, the sample conditioning loop
lag time, and the total analyzer response time.
3.1.21 validation, n—for equipment in the analysis of liquid petroleum products and fuels, the statistically quantified judgment that
the analyzer system or subsystem, in conjunction with any correlation applied, can produce acceptable precision and bias
performance on the prediction deviations (δ for materials that were not used to develop the correlation).
3.2 Definitions of Terms Specific to This Standard:
3.2.1 Analyzer System Items:
3.2.1.1 analyzer output, n—a signal (pneumatic, electrical, or digital), proportional to the property being measured that is
suitable for readout or control instrumentation external to the analyzer system.
3.2.1.2 analyzer system result, n—the measured property reading, in the accepted property measurement units, that is displayed
by the analyzer unit readout instrumentation or transmitted to end user of the analyzer system.
3.2.1.3 analyzer unit, n—the instrumental equipment necessary to automatically measure the physical or chemical property of
a process or product stream sample using either an intermittent or a continuous technique.
3.2.1.4 analyzer unit repeatability, n—2.77 times the standard deviation of results obtained from repetitive analysis of the same
material directly injected into the analyzer unit under repeatability conditions.
3.2.1.5 continuous analyzer unit, n—an analyzer that measures the property value of a process or product stream on a continuous
basis and dynamically displays the instantaneously updated analyzer output.
3.2.1.6 intermittent analyzer unit, n—a cyclic type analyzer that performs a measurement sequence on samples from a process
or product stream and displays a new analyzer output at the conclusion of each cycle.
3.2.2 Time Unit Items—General Terms:
3.2.2.1 analyzer unit cycle time, n—for intermittent analyzers, the time interval between successive updates of the analyzer
output.
3.2.2.2 analyzer unit dead time, n—the time interval between the introduction of a step change in property characteristic at the
inlet of the analyzer unit and the initial indication of analyzer response to this change.
(1) Discussion—For intermittent analyzers, if the analyzer dead time is less than one analyzer unit cycle time, the analyzer unit
dead time cannot be directly measured.
3.2.2.3 analyzer unit response time, n—(see Fig. 2) the time interval between the introduction of a step change in property
characteristic at the inlet of the analyzer unit and when the analyzer output indicates a value corresponding to 99.5 % of the
subsequent change in analyzer results.
3.2.2.4 analyzer unit time constant, n—(see Fig. 2) the time interval between the initial response of the analyzer unit to a step
change in property characteristic and when the analyzer output indicates a value corresponding to 63 % of the subsequent change
in analyzer results.
(1) Discussion—For intermittent analyzers, if the analyzer unit time constant is less than one analyzer unit cycle time, the
analyzer time constant cannot be directly measured.
3.2.2.5 sample loop lag time, n—the time required for material to travel from the process takeoff point of the sample loop to
start of the sample conditioning unit.
3.2.3 General Terms:
3.2.3.1 composition-specific VRM, n—a validation reference material consisting of a single, pure compound, or a known,
reproducible mixture of compounds for which an accepted reference value or site assigned value can be calculated or measured.
(1) Discussion—A composition-specific VRM may be a commercial standard reference material (SRM) having a certified
accepted reference value.
3.2.3.2 continual validation, n—the quality assurance process by which the bias and precision performance determined during
initial validation are shown to be sustained.
3.2.3.3 direct measurement, n—a quantitative measurement result obtained using a principle or principles that express the
characteristic property of interest in its defining units.
D3764 − 23
3.2.3.4 indirect measurement, n—a correlated quantitative measurement result obtained using a measurement principle that
produces values that do not express the desired characteristic property but which can be modified empirically, using mathematical
modeling techniques, to estimate the necessary defining units of the property of interest.
(1) Discussion—Methods that utilize chemometric or multivariate analysis are indirect measurements for generating correlative
characteristic property measurement results.
3.2.3.5 process-derived VRM, n—a validation reference material derived from an isolated batch of process or product stream
material with chemical or physical characteristics, or both, that is suitable for determination of an accepted reference value or site
assigned value for the property of interest.
3.2.3.6 site assigned value (SAV), n—a property value of a reference material that is based on multiple results from either the
analyzer unit or a primary test method, obtained under site precision conditions.
3.2.3.7 validation reference material (VRM), n—for validation and quality assurance testing, a material having an accepted
reference value or site assigned value for the property of interest.
4. Summary of Practice
4.1 PPTMRs from the total analyzer system or its subsystem are compared to the corresponding PTMRs on at least 15 materials.
PPTMR and PTMR are statistically assessed relative to each other using the methodology of Practice D6708, recognizing that this
is only a preliminary Practice D6708 assessment. Precision and bias statistics on the prediction deviations (δ) are generated and
the bias is assessed against pre-specified performance criteria. The system or subsystem performance is considered to be
probationary validated for materials and property ranges representative of those used in the validation if the prediction deviations
are in statistical control, and bias performance statistic meets pre-specified criterion.
4.2 After probationary validation is achieved, continued statistical quality control chart monitoring and analyses on δ are carried
out with new production samples to ensure on-going prediction performance of the PPTMR meets the levels established from the
probationary validation.
4.3 Once the total number of samples with completed datasets (PPTMR , PTMR , δ) from probationary and continual validation
reaches 30, a general validation is conducted using the statistical methodology of Practice D6708. The objective of the general
validation is to demonstrate performance with at least 30 samples over a wider operating envelope, or, to confirm outcome from
probationary validation with more accrued data.
4.4 If the variation among the 30 samples is inadequate to conduct the Practice D6708 assessment, a level specific validation may
be performed to validate the agreement between PPTMR and PTMR over a narrow operating range. As additional (PPTMR /
PTMR / δ) datasets are collected covering a wider operating range, the general validation may again be attempted.
4.5 After general validation has been achieved, continue to monitor δ using statistical quality control charts at a frequency
commensurate with the criticality of the application.
5. Significance and Use
5.1 This practice can be used to quantify the performance of a process stream analyzer system or its subsystem in terms of
precision and bias relative to those of a primary test method for the property of interest.
5.2 This practice provides developers or manufacturers of process stream analyzer systems with useful procedures for evaluating
the capability of newly designed systems for industrial applications that require reliable prediction of measurements of a specific
property by a primary test method of a flowing component or product.
5.3 This practice provides purchasers of process stream analyzer systems with some reliable options for specifying acceptance test
requirements for process stream analyzer systems at the time of commissioning to ensure the system is capable of making the
desired property measurement with the appropriate precision or bias specifications, or both.
5.4 PPTMR from Analyzer Systems validated in accordance with this practice can be used to predict, with a specified confidence,
what the PTMR would be, to within a specified tolerance, if the actual primary test method was conducted on the materials that
are within the validated property range and type.
D3764 − 23
5.5 This practice provides the user of a process stream analyzer system with useful information from on-going quality control
charts to monitor the variation in δ over time, and trigger update of correlation relationship between the analyzer system and
primary test method in a timely manner.
5.6 Validation information obtained in the application of this practice is applicable only to the material type and property range
of the materials used to perform the validation. Selection of the property levels and the compositional characteristics of the samples
must be suitable for the application of the analyzer system. This practice allows the user to write a comprehensive validation
statement for the analyzer system including specific limits for the validated range of application. This practice does not recommend
extrapolation of validation results beyond the material type and property range used to obtain these results. In addition, users are
cautioned that for measurement systems that show matrix dependencies, bias information determined from pure compounds or
simple mixtures of pure compounds may not be representative of that achieved on actual process or product samples.
6. System Components
6.1 Fig. 1 illustrates a total analyzer system incorporating a selection and arrangement of components that are typical but not
specific for any particular analyzer system. A total analyzer system design addresses the chemical and physical properties of the
process or product stream to be measured, provides a representative sample, and handles it without adversely affecting the value
of the specific property of interest. Included are a sample loop, piping, hardware, a sampling port, sample conditioning devices,
an analyzer unit instrumentation, any data analysis computer hardware and software, and a readout display.
6.2 Sample Loop—Piping connected to the main process stream to deliver a portion of the stream to a location close to the analyzer
system with minimum lag time and return the unused material to the main process stream.
6.3 Sampling System—Sample probes, valves, lines, containers, pressure regulator, and gages that constitute the equipment
employed to obtain a proper sample from the sample loop and introduce either it or a validation standard sample to the analyzer.
6.4 Sample Conditioning Unit—A collection of devices to properly treat a portion of the sample from the sample loop so that it
meets the requirements for testing by the process analyzer. These components can incorporate temperature or pressure adjustment,
change of state (liquid, vapor), or removal of contaminants.
6.5 Inlet Port—Appropriate piping with selector valve(s) for placement either at the inlet to the analyzer unit or, when dictated
by the measurement specifications, at the inlet to the sample conditioning unit. The purpose of this inlet port is to allow injection
of validation standards or other calibration material into the analyzer system with quick switching between these typically
containerized materials and the flowing process stream.
6.5.1 For many analyzer systems the inlet port requires a manifold arrangement for validation or quality assurance studies. Such
a manifold, with suitable valving, provides a means to use a containerized supply of standby material when a flowing process
stream is not available for the purpose. It also permits quick switching between different validation standards when that is
desirable.
6.6 Sample Port—An appropriate probe or fitting in the piping to permit collection of representative samples for laboratory
analyses using a primary test method.
6.7 Analyzer Unit—Instrumentation designed to automatically measure the chemical or physical property of a process or product
stream sample and provide either an intermittent or a continuous output signal representing the measurement result.
6.8 Readout Instrumentation—If it is not an integral component of the analyzer system, a device to display or record or both, the
property measurement analyzer result.
7. Preparation of Analyzer System
7.1 Implementation of this practice requires that the process stream analyzer system operates under conditions specified:
7.1.1 Meets all applicable electrical and safety codes.
D3764 − 23
7.1.2 Meets the supplier’s recommendation.
7.1.3 Complies with operating conditions specified by the manufacturer.
7.1.4 Includes a predicted PTM algorithm, if necessary.
7.2 After installation or major maintenance, conduct such diagnostic tests as recommended by the manufacturer to demonstrate
that the analyzer meets the manufacturer’s specifications or historical performance levels, or both. If necessary, adjust the analyzer
system components so as to obtain recommended analyzer output levels for specified reference materials.
7.3 Inspect the entire analyzer system to ensure it is installed properly, is in operating condition, and is properly adjusted after
completion of the initial commissioning procedures.
7.4 Application of Practice D6708 mathematics requires that the site precision of the analyzer system be known. If the analyzer
site precision is not known, then it should be determined prior to initiation of the validation procedure. Methodology described
in Practice D6708 can be employed.
8. Pre-Validation Analyzer Calibration Check
8.1 When an analyzer is initially installed, and after major maintenance has been preformed, diagnostic tests should be conducted
to demonstrate that the analyzer meets manufacturer’s specifications and historical performance standards. These diagnostic tests
may require that the analyzer be adjusted so as to provide predetermined output levels for certain reference materials. Such
adjustment may be done in hardware, software or both.
8.2 Description of specific calibration procedures for the numerous analyzer types is beyond the scope of this practice.
9. Validation Procedure
9.1 The objective of the validation procedures is to quantify the precision and bias performance of prediction deviations (δ)
between PPTMR produced by a process stream analyzer system (or its subsystem) versus PTMR for materials spanning the
intended operating range for the analyzer system. The user must specify acceptable precision and bias performance criteria before
initiating the validation. These criteria will be dependent on the intended use of the analyzer.
9.1.1 For analyzer systems used in product certification, precision performance acceptance criteria for δ will typically be based
directly on the published reproducibility (R) of the primary test method. Bias criteria will typically be based on regulatory or
contractual requirements. It is a general performance expectation that no bias correction can further improve the precision of δ
statistically.
9.1.2 For analyzer systems used in other types of service, precision and bias criteria must be developed based on the intended use
of the analyzer results.
9.1.3 This practice recommends articulation of precision performance of δ as a between-method reproducibility (R ).
XY
9.2 The line sample procedure directly fulfills the validation objective since the validation results for both the process system and
the primary test method are obtained on process samples. However, if line samples covering the composition and property range
of interest cannot be acquired within a reasonable length of time once the validation process begins, consider using either
process-derived or composition-specific validation reference materials (VRMs) to extend the composition and property range of
the validation sample set. A suitable process-derived VRM may simply be a batch of material obtained at a time prior to the start
of the validation procedure but that was not used in calibrating either the analyzer or the primary test method. In general, the
composition of a VRM used for validation should be similar to a composition that is anticipated for the process stream at some
future time.
9.2.1 In cases where it is necessary to include the sample loop or the sample conditioning unit (Fig. 1), or both, in the validation
procedure, VRMs should not be used to the exclusion of lines sample unless it is practical to use the VRMs to validate both sample
system and analyzer (this is generally not practical). The sample system can be excluded from the validation procedure if it is
known that the sample system does not materially alter the composition or condition of the sample presented to the analyzer and
D3764 − 23
if the sample system response time can be estimated with reasonable certainty. Guidance on how to meet these conditions is beyond
the intended scope of this practice. If these conditions cannot be met and if VRMs are needed to extend the property and
composition range of the validation set, it is recommended that the user conduct two probationary validations, one using line
samples and the other using VRMs, to demonstrate that VRM procedure adequately reflects corresponding performance for actual
process materials. Once demonstrated, the statistical quality control charting for continual validation can be done using VRM
procedures, with a periodic line sample procedure mixed in over time to demonstrate that both procedures continue to provide
similar and acceptable performance.
NOTE 3—If the process analyzer system is not based on identically the same measurement principle as the primary test method, then the analyzer system
may react differently to variations in the sample matrix than does the primary test method. In such case, analyzer results for process samples might be
biased relative to primary test method results even when the VRM procedure results shown no such bias unless the VRM is process-derived. The bias
can be minimized by using a process stream (test) sample for which an ARV or SAV was determined as the VRM. The test sample used in this fashion
should be representative of the current process stream.
NOTE 4—If, due to differences in sample pretreatment, the sample analyzed by the process stream analyzer and the sample analyzed by the primary test
method are not identically the same, then the use of the VRM procedure may not accurately reflect agreement between the process analyzer and the
primary test method. The VRM may not be affected in the same manner as process samples by the different sample pretreatments. Again, this effect can
be minimized by using current process stream (test) samples as VRMs.
9.3 Probationary, General and Level Specific Validation using the Line Sample Procedure:
9.3.1 This procedure is applicable for analyzer systems that are equipped with sample ports anywhere within the system that can
facilitate the safe collection of material intended for analysis by the analyzer unit without significantly altering the property of
interest. The subsystem from the sample port up to and including the analyzer subsystem (see Fig. 1) is considered to be validated
for current process stream samples if the δ results are in statistical control, and the precision and bias statistics meet user-specified
requirements.
9.3.2 Line Sample Procedure Requirements:
9.3.2.1 Select point of line sample withdrawal.
9.3.2.2 Determine the total lag time of the system or subsystem from the sample withdrawal point (see Figs. 2 and 1 for guidance)
up to and including the analyzer.
9.3.3 Procedure—Collect analyzer unit results from at least 15 implementations of the line sample procedure under site precision
conditions, with at least 8 to 12 h between each implementation, as follows:
9.3.3.1 Observe the analyzer unit output until the change between readings over at least three lag times for the subsystem
(associated with the sample port)
...