A fast screening method for the control of EU prepackages referred to Directive 76/211/EEC

Giuseppe Ardimento

Head of Market Regulation, Naples Chamber of Commerce, Italy

Citation: G. Ardimento 2026 OIML Bulletin LXVII(3) 20260311

Introduction

Metrological market surveillance, in accordance with the international document OIML D 9 'Principles of metrological supervision' [1], requires the locally competent metrology departments to implement appropriate enforcement actions at the premises of prepackagers and at points of sale to ensure the compliance of manufacturers and retailers with the legislation in force.

The surveillance of prepackages manufacturers involves the control of filling systems through control charts and similar systems, as well as the control of batches of manufactured prepackages through the so-called reference test of the Prepackages Directive 76/211/EEC [2] at the manufacturer premises.

The strict application of the Reference Test of the Prepackage Directive to the batches under control can prove to be a burdensome activity for metrology authorities.The need to optimize time across various surveillance activities has prompted the Naples Chamber of Commerce, the competent metrology authority for the province of Naples, to design a preliminary screening strategy for prepackage batches undergoing control. This aims to reduce inspection times while remaining fit for purpose to issue statistically valid judgments on the performance of prepackage manufacturers.

Assuming an average quality that is good or satisfactory for the requirements of the Directive, the preliminary rapid screening yields positive results for most of the checks carried out, thus speeding up routine field controls. Obviously, this type of rapid control must be applied uniformly and continuously, year after year, so as to ensure a supervisory frequency adequate for the purpose of the inspection programme.

Three Packer's Rules

As outlined below, Annex I, Section 1 of Directive 76/211/EEC [2] mandates that prepackages must satisfy three metrological conditions, colloquially referred to by WELMEC as the "Three Packer's Rules". In the European Union, these "Three Packer's Rules" constitute the foundational principles of the average quantity system for “℮-marked” prepackaged products, and the requirements are extensively interpreted and harmonized through the relevant WELMEC Guides:

WELMEC Guide 6.3 [3] provides guidance for the harmonized implementation of Directive 76/211/EEC.
WELMEC Guide 6.4 [4] functions as the primary manual for packers and importers to establish adequate quantity control systems.

In brief, a batch of prepackaged goods must comply with the following three conditions at the time of production:

Rule 1: The Average Requirement

The actual contents of the prepackages, on average, must not be less than the nominal quantity Qn stated on the label.

Rule 2: The Tolerable Negative Error (TNE)

The proportion of prepackages having a negative error greater than the Tolerable Negative Error (T) specified in the Directive's table must be sufficiently small.
WELMEC documents define this as being that no more than 2.5% of the production batch items fall below Qn – T (a "T1" error).

Rule 3: No Defective Prepackages

No individual prepackage may have a negative error greater than twice the TNE. Any package falling below this absolute limit Qn – 2 T constitutes a "T2" error and is strictly prohibited from being placed on the market.

Preliminary screening

This initial phase serves to quickly assess the compliance of the batch without destroying the product. It is carried out as follows:

Sample: Randomly select 10 prepackages from the batch at the end of the filling chain.
Measurement: Determine the gross mass of each piece and subtract the estimated tare weight.
Apply the Packer's Rules under the Directive: Verify the 10 obtained values by applying the three fundamental criteria:
1. Mean: The mean of the actual content of the 10 pieces must be equal to or greater than the nominal quantity This verification is carried out using the inequalitywhere Xm is the sample mean, Qn is the nominal quantity, and Sn is the sample standard deviation. The numerical ratio represents the Student's t for a sample of 10, which guarantees a 0.005 reliability for the statistical prediction of the sample mean X_m.
2. T1 error: The maximum number of products below the threshold TU1 = Qn – T must represent at most 2.5% of the prepackage batch under control. This condition is ensured by calculating the area under the Student's distribution to the left of T1 = –T /Sn is no greater than 0.025.
3. T2 error: No prepackage must be filled with a quantity lower than TU2 = Qn – 2 T. In practice, this is ensured by checking that the area under the Student's distribution to the left of the abscissa T2 = – 2 T / Sn is no greater than 0.001.

Outcome of the screening

The outcome of the screening determines the subsequent action.

Positive outcome: The batch passes the screening and is accepted.
Negative outcome: If any of the three packer's rules fails, a control phase using the directive's reference test is activated.

Reference test

In the event of a negative outcome, the official method provided by Directive 76/211/EEC is applied. In order to optimize time, the method defined as "by destructive control" is used. To avoid destroying or wasting the product, the determination of the actual mass via the average tare is adopted as follows.

Tare Sampling: Select 10 empty packages (tares) representative of the production.
Average tare calculation: Weigh the 10 empty packages and calculate the average tare value.
Batch control:
1. Take the official statistical sample of full packages equal to 20.
2. Weigh each intact prepackage to determine its gross mass.
3. Calculate the net weight of each piece by subtracting the previously determined average tare.

Final Evaluation: Apply the official statistical compliance tests (mean test and defective units test) on the net values thus calculated to decree the final approval or rejection of the batch.

Conclusions

Although the statistical screening method is less reliable than the reference test for the Prepackages Directive, given the average quality of prepackage manufacturers, it allows for the optimization of continuous surveillance timing across the territory for such metrological activities