Statistical methods

  1. 4.1
    Statistical methods for NM Control & Accounting (introductory course)

    The trainees are given general knowledge of statistics applications for MC&A — from terminology and basic definitions to specific statistical approaches to MC&A problems.

    Theoretical lessons

    • Role of statistics in MC&A. Statistical approaches in regulatory documents of national MC&A systems and that of IAEA; in reporting documents.
    • Statistical terminology and basic notions.
    • Statistical principles and methods of measurement quality control, of inventory difference variation, detection of NM losses/no losses/excess inventory under conditions of data uncertainty due to measurement errors.
    • Algorithms and computer codes for statistical sampling calculation and random sampling planning.
    • Plans of NM parameter measurements to reveal total and partial defects.

    Practical exercises

    • Calculation and analysis of inventory difference for a hypothetical facility which fabricates nuclear fuel.
    • Exercises with conventional nuclear material in application of statistical methods for initial and periodical PITs. The weighing equipment and RMs of mass are widely used.
    • The methodological material includes excerpts from the IAEA guidelines on MC&A statistical principles.
    • The course is accompanied with demonstration of computer software for statistical sampling, videos about different MC&A issues (prepared by US specialists) and PIT performance at Russian nuclear facilities.

    Potential audience

    Technologists, metrologists, engineers, scientists engaged in MC&A field.
    Specialists of state bodies and operators involved in inspection activity.

    Preliminary training

    Hide
  2. 4.2
    Statistical methods for measurement quality control

    The course covers the detailed information about the application of statistical methods for MC&A measurement quality control.

    Theoretical lessons

    • Application of statistics in MC&A systems.
    • Statistical terminology and basic notions.
    • Statistical approaches in regulatory documents of national MC&A systems and those of the IAEA; Russian regulatory documents.
    • Models of measurement. Theory of measurement errors.
    • Systems of control and measurement quality assurance, statistical analysis of measurements of weight, mass, volume, statistical models for NDA and DA in MC&A.
    • Sharing the experience in statistical methods in MC&A measurement quality control among representatives of Russian facilities.

    Practical exercises

    • Exercises with NM item models in application of statistical methods for measurement quality control in case of NM shipment/receipt, PIT performance at a hypothetical facility with the use of weighing equipment, reference materials, computers and special software.
    • Practical exercises at the Laboratory for scale and volume calibration.
    • Methodological materials available for the students comprise excerpts from the IAEA guidelines on MC&A statistical principles for measurement quality control.

    Potential audience

    Technologists, metrologists, engineers, scientists engaged in MC&A field.
    Specialists of state bodies and operators involved in inspection activity.
    The course was significantly updated in 2005, it is meant for the specialists trained in this subject several years ago.

     Preliminary training

    Hide
  3. 4.3
    Statistical methods for inventory difference (ID) calculation and analysis

    The trainees will acquire skills in the use of statistical methods and computer codes for ID calculation and analysis.

    Theoretical lessons

    • Application of statistics in MC&A area.
    • Terminology and basic concepts.
    • Statistical approaches to ID determination and analysis in regulatory documents of national MC&A systems and those of the IAEA.
    • Russian regulatory documents.
    • Measurement uncertainty propagation.
    • Detailed consideration of measurement errors, ID dispersion analysis for the cases with averaged and non-averaged data flows.
    • Defect detectability, MUF-D method being used by IAEA inspectors, software to determine and analyze ID and its components.
    • Statistical analysis of NM balance at the Siberian Chemical Plant (Seversk) and the Machine Building Plant (Electrostal).

    Practical exercises

    • Modeling of measurements and calculation of ID variance.
    • Use of statistic methods during PIT performance at a hypothetical facility with NM item models, with the use of weighing equipment, reference materials, computers and specific software.
    • Planning of NM parameters measurements to determine total and partial defects.
    • Methodological materials available for the students comprise excerpts from the IAEA guideline on MC&A statistical principles for ID analysis.
    • The course is accompanied with demonstration of videos about different MC&A issues prepared by US specialists.

    Potential audience

    Technologists, metrologists, engineers, scientists engaged in MC&A activity field.
    Specialists of state bodies and operators involved in inspection activity.

    Preliminary training

    Hide
  4. 4.4
    Statistical methods for accountancy performance assessment (advanced course for experts and instructors)

    The course considers the detailed information about the application of statistical methods for MC&A system quality control — from the theory of sampling for final populations up to accountancy effectiveness assessment in quantitative terms.

    Theoretical lessons

    • Sampling theory for final population.
    • Plans of sampling for verification measurements, specific features of processing the results of measurements performed with the HLNCC.
    • MUF-D procedure which can be used by inspectors (from a facility, Rosatom, Rostechnadzor).
    • Error models, measurement co-variation for the HLNCC, mathematical equation for actual balance assessment and accountancy effectiveness assessment.
    • How «International Performance Values. IPV — 2000» can be used for accountancy quality assessment.
    • Lectures are accompanied with demonstration of the software for statistical sampling.

    Practical exercises

    • Preparation of statistical sampling for verification.
    • Planning of NM parameter measurements with rough and precise instruments to determine total and partial defects.
    • Processing of the given results measured by the HLNCC.

    Potential audience

    Technologists, metrologists, engineers, scientists engaged in MC&A field or those involved in development of MC&A elements, instructions, procedures or in MC&A inspection activity. Instructors.

    Preliminary training

    Hide