When a retailer weighs out his product in front of the customer, can the customer trust the scales? It is usually up to local government to certify the accuracy of such scales, though in recent years, certification of measuring devices within much of the world has been privatised with various countries’ national laboratories overseeing the certification process.

With the advent of digital weighing devices, the increased sensitivity of devices has become an issue as has the threat of multiplicity of differing local regulations. To promote international harmony, the International Organization of Legal Metrology (OIML) has introduced various recommendations which member states can adopt into their legal structures. The EU was in the forefront of adopting these recommendations and when the United Kingdom left the EU, these rules remained in place. The three recommendations that are of interest to non-automatic weighing devices (i.e. weighing devices that you see at food and vegetable markets, at the deli counter, in the doctor’s surgery, at the Post Office etc) are:

• R 52 – Hexagonal Weights – metrological and technical requirements
• R 76 – Non-automatic weighing instruments
• R 111 – Weights of classes E1, E2, F1, F3, M1, M1-2, M2, M2-3 and M3

Recommendation R 52 – Hexagonal Weights

Recommendation R 52 was the first of these recommendations to be published. It defined the physical sizes of cast-iron weights that are sometimes seen in markets. The recommended sizes are 100 g, 200 g, 500 g, 1 kg, 2 kg, 5 kg, 10 kg, 20 kg and 50 kg. Such weights are made of cast iron and have a hollow in their base. This hollow is partially filled with lead to bring it up to the required weight. Once this is done, the lead plug is stamped by an auditing organisation to certify its weight. The allowable tolerance is ±0.05% apart from the 100 g weight where a tolerance of ±0.1% is allowed.

(Source: https://commons.wikimedia.org/wiki/File:1kg_with_creditcard.JPG)

Typical hexagonal 2 kg weight with a credit card to show its relative size.

(Source: https://commons.wikimedia.org/wiki/File:2kg_auditmark.JPG)

The same 2 kg weight showing the assayer’s mark on the lead plug.

In view of recommendation R 111, there have been moves to deprecate this standard but a counter argument that these weights are easy to manufacture and are often still used in third world markets has prevailed.

Recommendation R 76 – Non-automatic weighing instruments

Recommendation R 76 covers almost all weighing devices that display the weight in question and covers anything between jeweller’s balances and industrial weighbridges.

The OIML originally categorised weighing devices in one of four classes defined by the allowable tolerance of the device, but in recent years have expanded Class III, when it is used for weights above 4 tonnes, has been split into two classes – Class III and Class III L.

All classes must meet the following criteria:

• For a weight to be valid on the device, it must be equal to or exceed the minimum number of divisions with each division usually being equal to one increment.
• The maximum range of the device must be equal to or exceed the value given in the column “Maximum number of divisions.”
• The device must be certified by an accredited certification agency, both when it is new and at regular intervals as specified in the manufacturer’s documentation.
• Other criteria also apply, which are discussed later.

Since the recommendation covers such a wide range of devices the recommendation uses the symbol “e” to represent the minimum increment on the scale. To put this into perspective, consider two different weighing devices that I own. One is a jeweller’s scale which has a capacity of 200 g in 0.01 g increments. The value of “e” for this device is 0.01 g and the maximum value is equivalent to 20 000 times the basic increment. My bathroom scales on the other hand have a different value for “e.” In the range 50 kg to 100 kg, the display is rounded to the nearest 0.2 kg, so on this device, “e” has a value of 0.2 kg thus at 100 kg, the display is 500 times the basic increment.

The classes are defined as follows:

ClassMinimum number of divisions at minimum weightMinimum number of divisions at maximum weightMaximum tolerance at maximum weightApplicationsI50 000Not specifiedNot specifiedHigh precision laboratory workII100 (Note 1) 3 000100 000±0.001%Laboratory work and transactions of high value goods (such as jewellery)III100 (Note 2) 50010 000±0.01%Most commercial and medical transactionsIII L2 000 (Note 3)10 000±0.01%Weighbridges etc (Weights above 4 tonnes)IIII1001 200±0.08%Portable devices used in medical clinics etc

Note 1: e <= 50 mg

Note 2: e <= 2 g

Note 3: e >= 2 kg

Outline definitions of the various classes of weighing devices

The recommendation runs to almost 100 pages as it covers a variety of weighing devices, both analogue and digital including yardarm balances, point of sale weighting devices, medical scales etc. The points that the recommendation covers include:

• Default temperature ranges of operation which varies by class. If the device is designed to operate in extreme heat or extreme cold, the ranges should be catalogued and certification tests planned accordingly.
• If a device has multiple ranges, then each range should be tested separately. For example, my bathroom scale has 0.1 kg increments for weights below 50 kg, 0.2 kg increments for weights between 50 and 100 kg and 0.5 kg increments for weights above 100 kg – this being typical of multiple-range devices.
• The default range of permitted power fluctuations is defined. Operational use must not be affected if the electrical power fluctuates within the specified limits. If the power fails or is outside limit, the device should either turn itself off or display an error message.
• Specification of printouts and displays.
• If the device has a dial and pointer the recommendation defines the spacing of graduation marks.
• The acceptable accuracy of weights used during certification and must be “e/3”.

The above list is a fraction of the criteria that are listed in the recommendation. The interested reader is referred to the recommendation itself.

Recommendation R 111 – Weights of varying classes

Recommendation R 111 covers three principal categories of weights – those used as references for the calibration of other weights, those use for the calibration of weighing devices and those used for everyday purposes in conjunction with a weighing device. The recommendation is limited to weights that are greater than or equal to 1 mg and less than or equal to 5 tonnes.

The standard defines nine classes of weights known as E1, E2, F1, F2, M1, M2, M3, M1-2 and M2-3 (in descending order of accuracy). Three classes of weights are defined for calibration purposes of weighing devices of Classes I, II and III respectively and the remainder are intended either for everyday use for scales of Classes I, II, III, III L and IIII or for calibration purposes of Class IIII devices.

Weights of Class E1 are designed to be calibrated against the national prototype kilogram and to serve as working copies for calibration of other weights. A certificate of traceability against the national prototype accompanies such Class E1 weights. For Class E1 weights of 100 g and above, the permitted tolerance of ±0.5 parts per million. At the other end of the scale, the tolerance weights of class M3 having a nominal weight of 100 g or more is 0.05%. Below 100 g, the tolerance level for all classes of weight is relaxed due to the difficulty of manufacturing of such weights.

Most of the recommendation is devoted to detailed rules concerning the physical construction of the weights, the environment in which they are to be used and the way in which tests are to be carried out. Of note is that the calibration of the weights assume that they will be used in air and that the buoyancy effects need to be considered. The recommendation states that the weight should be accurate when used in air with a density of 1.2 kg/m³ against a target object having a density of 8000 kg/m³. (The density of water is approximately 1000 kg/m³).

Other parts of the recommendation deal with the way in which the calibration is to be recorded, the requirement that the apparatus reaches thermal stability before performing the test and ensuring that magnetic fields are either absent, are negligible or that there is compensation for them.

The interested reader is referred to the recommendation for further details.

(Source: https://commons.wikimedia.org/wiki/File:Weights01.jpg)

Set of weights, range 10 mg – 100 g used for laboratory work.

References:

• R 52 Hexagonal Weights – https://www.oiml.org/en/files/pdf_r/r052-e04.pdf
• R 76 – Non-automatic weighing machines – https://www.oiml.org/en/files/pdf_r/r076-1-e06.pdf
• R 76 explained (plus US Customary extensions): https://adamequipment.com/content/post/what-are-scale-balance-accuracy-classes-divisions/
• R 111 Weights of Various Classes – https://www.oiml.org/en/publications/recommendations/en/files/pdf_r/r111-p-e04.pdf

This week saw the publication of research into the effects of reducing the standard serving size of draught beer in the UK.

The University of Cambridge study was described by its leader as being “the first real-world study to look at this”. It shows that reducing the current standard beer glass size in bars, pubs and restaurants from the current pint (568 ml) size has the potential to reduce the total amount of alcohol consumed in the UK, and should be given consideration as part of the Government’s fight to reduce the effects of alcohol consumption on health.

Health benefits of reducing alcohol consumption

Alcohol consumption is the fifth largest contributor to premature death and disease worldwide. In 2016 it was estimated to have caused approximately 3 million deaths worldwide.

Alcohol misuse can cause long-term consequences, such as higher risk of brain damage, liver disease, dementia, and a number of cancers. This has a particular effect for health services. In 2022/23, there were over 320 000 alcohol-specific hospital admissions. NHS England estimate that up to 15% of A&E attendances are alcohol related.

Alcohol-specific deaths in the UK were over 10 000 in 2022, a 70% increase from 2002.

Methods of reducing alcohol consumption

Until now, governments have focused on reducing alcohol misuse by reducing the availability of cheap, high-strength drinks. This has been done by the use of a duty, or tax, regime on the sale of alcohol. This regime was updated as recently as August 2023.

However, the Government hasn’t yet sought to use another option available to it, which is the reduction of the serving sizes of alcoholic drinks in pubs and restaurants. This is in spite of two recent academic studies that have both shown that reducing the largest available serving size leads to a reduction in the amount of alcohol consumed.

The results of a University of Cambridge study, published in January 2024, showed that removing the largest wine glass size led to an average reduction in the amount of wine sold in pubs and bars of just under 8%.

The results of a similar study, published this week, showed that removing the pint (568 ml) glass in favour of a two-thirds pint glass (379 ml) led to an average reduction in the amount of draught beer sold in licensed premises of just under 10%.

Published in PLOS Medicine, on 17 September 2024, the University of Cambridge research article concluded, “Given the potential of this intervention to reduce alcohol consumption, it merits consideration in alcohol control policies.”

The Guardian quoted Prof Dame Theresa Marteau, study leader and director of the behaviour and health research unit at the University of Cambridge, as saying, “Does this have the potential to contribute to population health? I’d say definitely, yes.”

Government reaction

The only published Government reaction to the study, so far, appears to be in a BBC article, in which Labour MP Josh Simons, an ally of prime minister Sir Keir Starmer, says he would not back any plans to remove pints as the top measure of drinks. The article quotes him as saying, “I love a pint and leader of the Labour Party Keir Starmer loves a pint.”

The elephant in the room – Metrication

The University of Cambridge study was restricted to the use of legally available draught beer serving sizes, and so was unable to test more subtle reductions in size from the one pint (568 ml) glass size. The two-thirds of a pint (379 ml) glass size used in the trial was the available option with the smallest reduction in size – but even this option represented a large reduction of 33%. The consumer-acceptability of this reduction was an obvious obstacle in conducting the study – researchers had asked more than 1700 pubs, bars and restaurants to participate in the trial, but only 13 accepted, despite compensation offers for lost sales.

A more subtle reduction in glass size would probably have led to a more subtle reduction in alcohol consumption, but it might also have led to a proposition more acceptable to consumers and therefore less politically problematic for the Government to consider. Any possible reduction in alcohol consumption should be something to be taken seriously, rather than rejected outright.

The sizeable 33% reduction necessary for this trial is reminiscent of the situation in Wales when a reduction of 33% had to be made when the default speed limit was lowered in 2023 from 30 mph, using imperial units – a more finely-tuned adjustment could have been made using metric speed limits in km/h.

The current range of prescribed sizes for draught beer is quite coarse:

⅓ pint, ½ pint, ⅔ pint, 1 pint, multiples of ½ pint.

All sizes are specified in imperial pints and fractions only. It might surprise some readers to learn that draught beer cannot currently be sold in metric quantities.

Given that the fluid ounce ceased to be a legal unit for trade long ago, any new imperial glass size intermediate to the two-thirds pint and one pint would have to be specified in another awkward-looking fraction of a pint. Two of the prescribed quantity sizes for draught beer already lead to recurring numbers when written as decimals.

Of course, the obvious solution would be to bring the sale of draught beer and cider into line with beer and cider already sold in bottles and cans, by switching all prescribed quantities of draught beer and cider to metric units, with sizes based on multiples of 100 ml.

The International Organisation Of Legal Metrology, OIML, already has such a standard metric range. The range provides for two round metric sizes that are both smaller than the imperial pint and also larger than the awkward two-thirds pint size.

…, 200 ml, 250 ml, 300 ml, 400 ml, 500 ml, 1 litre, …

The 500 ml size would be an obvious replacement for pint glasses in pubs, and the 400 ml size would probably be an option more welcome in a restaurant setting.

Beer glasses in these sizes can already be found in bars and restaurants across our continent. These standard sizes come in a variety of designs. The adoption of standard metric beer glass sizes will give bar and restaurant owners a greater choice of glass designs, and ought to lead to cheaper glasses in the long run.

The 500 ml (0.5 litre) size is about 12% smaller than an imperial pint, and the 400 ml (0.4 litre) size is about 30% smaller than an imperial pint. As a replacement for pint glasses, either size should lead to a reduction in national alcohol consumption, and so should be of interest to any government with a serious commitment to reducing the impact of alcohol consumption on the NHS.

Other opportunities and benefits

If new glasses switch to the lined variety, rather than the brim-measure type currently used, new 500 ml lined glasses would actually be larger than the current brim-measure pint (568 ml) glasses, and would allow room for a proper head of beer, with fewer spillages. A typical 500 ml lined glass has a to-the-brim capacity of over 600 ml.

Apart from doorstep deliveries of milk, the sale of draught beer and cider in pints represents the last holdout of imperial units for trade. Switching to metric units for this purpose is long overdue. The need to know one less conversion factor will benefit consumers.

Will metric mean the end of the “pint”?

Brits abroad already have no problem when ordering “a pint” of beer in a bar serving beer in 0.4 litre or 0.5 litre glasses. In such circumstances, the word “pint” is always understood by both parties to mean “large beer”. Similarly, selling draught beer in 500 ml glasses in the UK need not prevent Sir Keir Starmer, or anyone else, from enjoying “a pint”. The word “pint” will live on for a long time after metrication, just as the word “dram” has already done when referring to whisky served in 25 ml or 35 ml glasses.

References

Removing pint glasses could reduce beer sales by almost 10% – 18 September 2024
https://www.cam.ac.uk/research/news/removing-pint-glasses-could-reduce-beer-sales-by-almost-10

Removing largest wine glass serving reduces amount of wine sold in bars and pubs – 18 January 2024
https://www.cam.ac.uk/research/news/removing-largest-wine-glass-serving-reduces-amount-of-wine-sold-in-bars-and-pubs

Impact on beer sales of removing the pint serving size: An A-B-A reversal trial in pubs, bars, and restaurants in England
https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004442

Impact on wine sales of removing the largest serving size by the glass: An A-B-A reversal trial in 21 pubs, bars, and restaurants in England
https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1004313

The new alcohol duty system – 1 August 2024
https://commonslibrary.parliament.uk/research-briefings/cbp-9765/

Small beer: Study calls on government to shrink pints
https://www.bbc.co.uk/news/articles/c4gl737zr79o

Pint of no return? Two-thirds measure could boost English health – study
https://www.theguardian.com/science/2024/sep/17/pint-two-thirds-beer-measure-english-health-study-pubs

Vessels for commercial transactions – OIML
https://www.oiml.org/en/files/pdf_r/r138-e07.pdf

(Acknowledgement to Martin Vlietstra for the OIML reference)

The Criminal Justice Act 1988 (Offensive Weapons) (Amendment, Surrender and Compensation) Order 2024 is a Statutory Instrument that was made on 30 April 2024 to tackle knife crime. Parts 1 and 3 of it came into force on 26 June 2024, Part 4 came into force on 26 August 2024 and Part 2 will come into force on 24 September 2024. Part 2 contains two references to measurement. One refers to the length of a weapon’s blade and the other refers to the distance of the serrated cutting edge from a blade’s handle. And they are both in inches.

Part 2 contains the amendment of the Criminal Justice Act (Offensive Weapons) Order 1988. It defines the specific amendments to the Act. Here is the relevant text that uses inches in the amended legislation. This section inserts some text into the Act and includes the following text:

(iii) a blade of over 8 inches in length (the length of the blade being the straight-line distance from the top of the handle to the tip of the blade)

In another section in Part 2, the following text is also inserted:

(a) a serrated cutting edge (other than a serrated cutting edge of up to 2 inches next to the handle)

In July 2024, the Government also published the guidance for the Statutory Instrument that amends the Act. This is called the “Guidance for surrender of ‘zombie-style’ knives and ‘zombie-style’ machetes and claiming compensation”.

It seems that the Government is following the convention of using inch-based descriptions of knives and machetes. Why are they not leading by example and using metric units? It would cost nothing to banish all uses of imperial units from all legislation by replacing them with metric units through statutory instruments. Why is there a lack of commitment by British politicians to use metric units exclusively in legislation? It would help the UK to move to a single system of measurements that is used for all purposes in the UK and remove the need for Britons to cope with two competing systems of measurement.