Standards for grain testing (analysts training)

This page hosts guidance and training modules that aim to help improve the standard of grain testing across the UK. It covers equipment, methods and operator standards, as well as targets for testing.

Note: This information is not designed to supersede any appropriate quality systems already in place and is subject to third party audit and accreditation.

This page describes the elements needed for best laboratory practice for grain testing. It gives straightforward guidance on the requirements for equipment, methods and operators. It is not designed to supersede any quality systems already in place and is subject to third party audit and accreditation. However, it provides a framework for those who do not operate to these standards.

In addition to the framework, the targets for testing are described. These are measures of the proficiency of a laboratory and are given as the limits of variation that are acceptable for repeat testing by the laboratory. They are also given as limits of deviation from defined mean/median values or reference values when participating in proficiency test schemes and limits of variation for use with Internal Quality Control checks. Limits of variation for staff competency are also provided. These values have been derived from a very large pool of actual testing data. An explanation of terminology and derivation of standard value is given, followed by tables of standard values.

These standards are for all UK laboratories that test grain for trading purposes. Registers of participating laboratories (along with their testing protocols) can be found on the relevant trade association websites.

Ideally, the following tests should be included within proficiency (ring) testing schemes. Laboratories that do not undertake the full range of tests will have a reduced scope.

Equipment

1. All testing equipment must be fit for purpose, serviced and maintained in compliance with the laboratory’s quality system and manufacturer’s instructions. Requirements for equipment (e.g. balance accuracy) are given in the test method
2. Equipment shall be calibrated according to routines within the method and/or against a defined schedule
3. Performance checks must be done before use. These checks should be performed daily for contractual testing or weekly, as a minimum, for non-contractual testing
4. When all the above (1–3) have been fulfilled, the equipment is ready for use

Method

5. The method shall be documented – this can be an official method or an in-house procedure. If the latter is used, the relationship to the official method must be known. Procedures must be in place to ensure that only current methods are used
6. Is the method accepted by the grain industry? Ensure you only use methods that give results that others recognise, for example the SKCS scale for wheat hardness

Operator

7. All staff must be competent in those tests that they are expected to undertake
8. All staff (full-time and temporary) shall be trained in the methods they are required to perform. A record of competence will be held by the laboratory manager and reviewed regularly. Preferably, staff will have undertaken the necessary laboratory training modules
9. All staff must ensure that any instrument of test performance checks are completed before routine testing starts. These checks should be performed daily for contractual testing or weekly, as a minimum, for non-contractual testing

Sample, testing and reporting

10. Only when all the above elements have been fulfilled can the testing start. Ensure that the test sample is representative of the sample provided to the laboratory (either in bag or from the laboratory-taken sample). Rules about sub-sampling, grinding and mixing must be followed
11. Undertake the test according to the documented method. Observe all the critical control points
12. Results shall be reported in a format agreed with the customer. All results shall be reported with units of measurement recognised by the grain industry

The measures of the proficiency of a laboratory are given as the limits of variation that are acceptable for repeat testing by the laboratory. They are also given as limits of deviation from defined mean/median values or reference values, when participating in proficiency (ring) test (PT) schemes and limits of variation for use with Internal Quality Control (IQC) checks. Limits of variation for staff competency are also provided. These values have been derived from a large pool of actual testing data.

Standard values: terminology and derivation

1. Repeatability – permitted difference between two results on the same sample conducted by the same person on the same instrument (within a short timeframe)
2. Reproducibility – permitted difference between two results on the same sample from different testing facilities
3. Standard Deviation (SD) for Proficiency Testing Schemes (PT) – maximum permitted value to be used to calculate z-scores. Providers are encouraged to use lower values for SD, if participants are regularly achieving better performance than the standard set by these values
4. Standard Deviation (SD) for Internal Quality Control (IQC) – maximum suggested value to be used to calculate monitoring (±1SD), warning (±2SD) and action limits (±3SD) for IQC sample
5. Staff competency – dependent on methodology, a combination of repeatability (‘r’ – column 1) and reproducibility (‘R’ column 2) values for the test are used. Reproducibility can be checked by labs being members of a proficiency (ring) check scheme. There are two main steps for checking staff competency:
   1. Repeatability – repeat-testing one sample a minimum of five times, ensuring all results fall within the repeatability value
   2. Competency – testing a minimum of five samples, representing a range of values, and comparing the results obtained with those of a trained operator/established value for the sample. Difference between results should not exceed the competency figure (column 5)

z-score =          (lab value – mean value)

                            standard deviation

Regardless of the sign (+/-):

• |Z| ≤ 2: performance is satisfactory
• 2 < |Z| ≤ 3: is a warning (occurs 1 in 20) and suggests attention to equipment or procedures may be required
• |Z| > 3: requires immediate action and attention to equipment, procedure or operator

Standard values for commodities

Figures for a variety of tests across a wide range of commodities are presented in the tables. These are based on proficiency testing schemes data provided by end-users where significant amounts of data, often a year’s worth, was available. Reference has also been made to standard deviation data published in IOB, EBC, CCAT, BSI and ISO standards.

For some tests, compromises have had to be made. These are based on the knowledge and experience of members of the Grain Sampling and Analysis Laboratory Working Party. This reflects the fact that the data available from different sources covered different ranges of test results.

Balances are checked with known value weights in the critical range, upper and lower action lines are set using the figures in the table below.

Dispensers are checked by weighing the volume of dispensed water (based on 1ml = 1g).

Includes skinned barley and broken grains.

Includes pea insect damage.