Last updated on 1 November 2000
Food Standards Agency UK - Survey of Added Water in Chickens and Chicken Products (Number 08/00)
Food Survey Information Sheet
This survey was conducted as part of the Working Party on Food Authenticity programme of gathering information about the UK food supply.
1. The issue of excess added water in meat and fish products is known to be of concern to consumers and has been the subject of previous food authenticity surveys such as the Survey of Added Water in Cured Pork Products.1 The Working Party on Food Authenticity (WPFA) has now carried out a nationwide survey to investigate the amount of added water contained in whole chickens and chicken pieces. A total of 261 chilled or frozen samples were analysed (76 whole chickens samples and 185 samples of chicken pieces), where each whole chicken sample consisted of 7 birds of the same batch code (532 chickens), and each chicken part sample consisted of at least 5 chicken parts of the same type and batch (at least 925 pieces).
2. In the whole birds, added water was found in 39 of the 76 samples (51 per cent of those analysed), at levels up to 16 per cent. All 39 of these samples were frozen products. Of the chicken piece samples, 32 of the 185 (17 per cent of those analysed) had added water with levels ranging from less than 2 per cent to 37 per cent. Most of these were chicken breast samples (27) but there were also 5 thigh samples.
3. Consumers have indicated their concerns to the WPFA about the level of undeclared water in meat products. The WPFA is aware that there are at least 2 ongoing investigations being dealt with by UK trading standards departments concerning excess water in chickens. Also a recent report2 highlighted that added water levels as high as 40 per cent had been found in imported chicken meat. The purpose of this survey, therefore, is to try to determine the extent of the practice and inform consumers about the levels of added water present in whole chickens and chicken pieces. Within this survey, the term "added water" is used to denote any water over and above the physiological water content present in a whole or part chicken, which may be present for a number of technological reasons, and is not meant to imply deliberate additions to mislead consumers.
4. Commission Regulation (EEC) No. 2891/933 requires slaughterhouses to comply with specific requirements with respect to water content of frozen whole chickens. The slaughterhouse is required to carry out regular checks to determine water content, to keep records of the results of such tests for a year and to mark each batch of frozen or quick-frozen chickens in such a way that its date of production can be identified. Enforcement authorities are also required to carry out sampling checks in slaughterhouses, within their geographical remit, at intervals of at least every two months.
5. This legislation lays down limits for the "technically unavoidable water content", which arises during the processing of whole chickens. The limits vary depending on the chilling method used: 2 per cent for air chilling and 7 per cent for immersion chilling. If the difference between the total water less the theoretical physiological water content (added water level) in a particular batch is found to exceed these limits, then the competent authority may allow the batch to be marketed providing it carries a declaration in the form "Water content exceeds EEC limit".
6. The specific requirements of Regulation 2891/933 are applicable only to frozen and quick-frozen whole chickens and stipulate that the added water content should not exceed the technically unavoidable limits, dependant on the chilling method used (see para 11), though under EC Regulation 1538/91,4 the labelling of products with the method of chilling is optional. Commission Regulation No. 1072/20005 has now extended the water content controls to certain fresh and frozen chicken and turkey cuts and specifies a method for calculation of added water content.
7. Each of the chilling methods has associated limits for the maximum amount of water permissible and the Regulations specify that checks are carried out using either of two alternative tests, one for the determination of the total water content (Chemical Test) and one for the determination of thaw loss (Drip Test). For the Chemical Test (the method on which the analytical protocol for the survey is based) the associated water limits are: air chilling 2.0 per cent, air-spray chilling 4.5 per cent and immersion chilling 7.0 per cent. For the Drip Test the associated water limits are; air chilling 1.5 per cent, air-spray chilling 3.3 per cent and immersion chilling 5.1 per cent.
8. Within the context of the Regulation, both tests are only applicable to frozen and quick-frozen chickens. The Drip Test can only be used for frozen or quick-frozen whole chickens as it depends on measurements of the water lost during the thawing of frozen product. It is not suitable for use on chicken parts. Since the Chemical Test can equally well be applied to frozen or chilled whole chickens or chicken parts, it was the method selected for use in the survey.
9. For whole chickens or cuts of chicken where an ingredient, other than additives, water, self-basting preparations or seasonings, has been added then the product becomes subject to the provisions of the Meat Products and Spreadable Fish Products Regulations 1984, as amended (MPSFPR).6 This requires such products to declare the minimum meat content and a list of the ingredients.
10. All other chicken products not covered by either the Poultrymeat Marketing Standards or the provisions of the MPSFPR 6 are covered by the provisions of the Food Labelling Regulations 1996,7 as amended, which require food to be labelled in a manner that accurately describes the product and does not mislead the consumer. Prepacked products are also subject to Quantitative Ingredient Declaration labelling (QUID provisions)8 and are therefore required to be labelled with a percentage declaration of the characterising meat species contained within them. However, non-prepacked products in this category are not at present covered by these rules.
Chilling methods for chickens
11. There are three methods of chilling used in the poultry industry:
- Air chilling, in which the carcase is chilled in cold air;
- Air-spray chilling, in which the carcase is chilled in cold air interspersed with a water haze or fine water spray;
- Immersion chilling, in which poultry carcases are chilled in tanks of cold water or ice and water. This method, used prior to freezing the poultry, is sometimes referred to in the industry as spin chilling.
12. Each of these methods will result in different amounts of water being incorporated into the chicken carcase, with air chilling giving the least and immersion chilling the most. Within the appropriate legislation, each of these methods also has associated limits for the maximum amount of water permissible (see para 7).
13. A total of 261 samples was purchased between December 1999 and January 2000 in 30 different regions of the UK from a range of commercial outlets including supermarkets, cash and carries, local butcher shops and other retailers. This was made up of 76 whole chicken samples and 185 chicken part samples (see paras 14 and 15). Although not collecting formal samples, the expertise and experience of trading standards officers was invaluable and samples were collected by officers from the following areas of the UK: Barnsley, Bedfordshire, Buckinghamshire, Gloucestershire, Hertfordshire, Kingston-Upon-Hull, Lancashire, Lincolnshire, London, Medway, Norfolk, Northern Ireland, Oxfordshire, Poole, Scotland, Somerset, Surrey, East and West Sussex, Wales, Worcestershire and West Yorkshire.
14. Samples included a range of frozen and fresh/chilled products, both pre-packaged and non-prepackaged (e.g. from butchers' shops). As decreed by the methodology protocol, each of the 76 whole chicken samples comprised 7 individual chickens, from the same retail outlet or supplier, of similar weight and carrying a common identifier e.g. the same "use by date" or lot number. Hence, 532 whole chickens were collected in total.
15. Chicken part samples were collected according to a range of category types. Types included: Half (approx. 10 per cent of samples); Quarter (approx. 10 per cent); boneless and skinless Breast (approx. 50 per cent); Thigh (approx. 20 per cent); Drumstick (approx. 5 per cent); and Wing (approx. 5 per cent). Each of the 185 chicken part samples comprised at least 5 parts of identical type, from the same retail outlet or supplier, either from the same pack or carrying a common identifier e.g. the same "use by date" or lot number. Thus, a total of nearly 1000 chicken parts was collected.
16. Product information was obtained from the product labels and recorded on a specially designed sample collection form by the collecting officers who were supplied with a written sampling protocol and detailed sampling plan outlining the information required, numbers and types of samples to collect and the details to be recorded for each sample.
17. Analysis of the samples by 9 separate laboratories, eight public analyst laboratories and one private laboratory, followed a protocol specified in European Commission legislation3 on poultrymeat marketing. Details of the results of the analyses for each sample can be found in the attached Tables 1 and 2 for whole chickens and chicken parts, respectively.
Method for Whole Chickens
18. The analytical method used to determine the level of added water in whole chickens for this survey was based on the Chemical Test specified in Commission Regulation (EEC) No. 2891/933. This is a well-established method widely used by UK public analysts and, of the two methods specified in this Regulation, is the only one that can equally well be applied to fresh chickens and to chicken parts (see para 6). Hence, for the purposes of this survey, the Chemical Test was used since both whole fresh and frozen chickens were sampled.
19. The Chemical Test involves the determination of water and protein in accordance with the recognised ISO (International Organisation for Standardisation) methods ISO 14429 and ISO 937,10 respectively. Determinations were made from the homogenised poultry carcases of each sample. The results were recorded for each of the 7 individual homogenised chickens within a sample and the average value calculated for each sample.
20. The total water content of the carcases was estimated, taking into account the physiological water content in relation to the protein content. The total water content is determined by comparison with a limit value dependent on the method of chilling used (see paras 6 and 7), to determine whether or not excess water had been taken up during processing. The results were calculated for all samples with respect to the air chill limit to give a comparison of the total added water, regardless of the chilling method.
21. The theoretical physiological water content in grams as determined by the method was calculated by the following formula:
WG=3.53 x RP^+ 23
Where RP^ = average protein weight of sample
WG = total water content in grams
22. The added water was calculated assuming that the minimum technically unavoidable water content absorbed during preparation amounts to 2 per cent (calculated on the basis of the carcase exclusive of absorbed extraneous water). The total added water content (WG) in grams for each sample was given by the following formula (including confidence interval):
WG = 3.65 x RP^ + 42
Method for Chicken Parts
23. The analytical method used to determine the level of added water in the chicken parts for this survey is similar to that proposed in Commission Regulation No. 1072/2000.5 This is also based on the Chemical Test specified in Commission Regulation (EEC) No. 2891/93 3 (see para 18).
24. The method involved the duplicate determination of water and protein in accordance with the recognised ISO standard methods (see para 19) of a composite of the 5 portions within each sample (or a representative selection of 5 portions, where a sample contained more than 5 portions). From the average composite sample analyses, water:protein ratios were calculated for each sample (Figure 1).
25. Authenticated reference material derived from Food Analysis Performance Assessment Scheme (FAPAS), with certified levels of nitrogen and moisture, was analysed in every batch of samples by all laboratories taking part in this survey, as a quality control check. The measurement of these standards had to fall within 2 standard deviations of their certified value, or the analysis was repeated. All laboratories participating in the analysis of samples were required to demonstrate adequate performance in appropriate rounds of FAPAS, or similar recognised measures of performance.
Interpretation of Results
26. The full results are presented in Tables 1 and 2. The survey found added water ranging from 2 per cent to 16 per cent in 39 of the whole chicken samples (51 per cent of those analysed), all of which were frozen chickens. Added water ranging from less than 2 per cent to 37 per cent was found in 32 of the chicken part samples (17 per cent of those analysed). Twenty-seven of these were chicken breast samples and five were thigh samples. Eleven of the samples were fresh and twenty-one were frozen. Overall, the results indicate that added water occurred predominantly in the frozen whole chickens and chicken breast samples.
27. Table 1 gives the results for the whole chicken samples using the Chemical Test (see para 7). This takes into account physiological water and compares all samples on the same basis, regardless of the method of chilling. Labelling of products with the method of chilling used is optional, and only one of the samples collected in the survey declared the method used. However, manufacturers were given the opportunity to comment (see Annex 1) on the method of chilling used for their products prior to publication of the results and some subsequently notified the chilling method used, which is reflected in Table 1. In Table 1, the amount of total added water is shown as a percentage and account is taken of any label declaration relating to added water. The inclusion of an entry for added water against a sample does not necessarily indicate that that product was mislabelled since account has to be taken of the method of chilling and the permissible limits for added water (see para 7).
28. All of the 39 whole chicken samples with added water were frozen chickens (see Figure 2), though 2 samples of the frozen chickens had no added water. Although the Food Labelling Regulations 1996 7 do not require added water applying to Commission Regulation 1538/914 to be declared in the ingredients list, 4 out of the 39 samples listed water as an ingredient on the product label. In the remaining samples which were above the limits of added water in frozen chickens, the extra water was neither declared nor listed anywhere on the product packaging.
29. The added water is not confined to any one country and of the frozen chickens with added water, 21 originated from UK processing factories, 10 from France, 6 from Denmark and 2 from Germany. Within the UK there are around 150 licensed poultry meat slaughter houses (10 currently producing frozen whole chickens).
30. It is evident from comments received from retailers and processors that most processors use the 'Drip Test' to check the added water in their process. The operators claim that the water content is within limits. Therefore it would appear from the results of this survey that the correlation between the two official test methods needs further investigation.
31. The results are presented for the chicken part samples in Table 2. The water:protein ratios were derived from data published by the European Commission on limits of physiological water content of frozen and quick frozen chicken and turkey parts,11 which is the same basis for the limits laid down in EC Commission Regulation No. 1072/2000.5 The non-physiological water was calculated and the results are presented in bands of 2 per cent added water. Calculations were based on the maximum tolerable values, which are designed to take into account: unavoidable water uptake during processing; the ratio between physiological water and protein and its variance; and the number of portions analysed in the sample (Figure 1). Figure 1 shows the result of extrapolating the data to zero added water. This indicates the water:protein ratio below which it is determined no added water is present (3.28 for breast meat and 3.93 for the other cuts).
32. The results are given as the percentage of water determined from the average composite sample water:protein ratios. Samples containing less than the minimum physiological limit were interpreted as containing no added water. It should be noted that only breast samples without skin were sampled, while the other types of parts were with skin attached. Account is taken, in Table 2, of any label declaration relating to added water. Figure 3 shows the water:protein ratios for the chicken breast samples with the 3.4 limit for 2 per cent added water marked by a dotted line. It can be seen that most of the breast samples fell below this line.
33. Of the 32 chicken part samples with added water, 27 were breast portions and 5 were thigh portions. In 17 cases, the added water was neither declared nor listed anywhere on the product packaging. Four of the products had water listed as an ingredient and eleven had a "with added water" declaration, or a minimum meat content percentage declaration (or both) clearly visible on the product label. However, there were only 3 products (from a total of 7 where the percentage amount of added water was declared either directly or indirectly) which accurately quantified the amount of added water. Hence we found that even where declarations about water content are given on product labels, these are not always accurate.
34. Although fewer problems were found with the chicken part samples than with the whole chickens, there is generally a wider range of potential legislation that could apply, dependent on the type of product. Not all of this legislation was in force at the time the samples were collected (see paras 4 - 10). In the case of non-prepackaged samples, e.g. many of the samples obtained from butchers', although some had added water of over 30 per cent, they are only regulated generally by the Food Safety Act12 and Food Labelling Regulations, 7and may not, in fact, contravene current legislation. Also in some of the non-prepackaged samples, the bulk batches were correctly labelled by the wholesaler but the relevant information was not then displayed or made available to the ultimate consumer in the individual shops at the point of retail sale.
35. Most of the chicken portion samples found to contain added water were frozen (21 of the 32). As with the whole chickens, it is possible that immersion chilling methods were used and that this has led to increased levels of water under the skin of the thigh pieces (the breast samples were all skinless). For the breast samples the reason for the added water may be levels of ice-glaze, used for protection of the product during cold storage. In other cases, water has been deliberately added, since it is claimed on the label, and in cases where the claimed level was exceeded this may be accounted for by poor process control.
36. The survey found added water in a total of 71 products (27 per cent) sampled in the survey. However, not all of these products would be considered mislabelled or in contravention of any regulations. Even a product containing undeclared significant amounts of added water may not be in contravention of legislation. However, unless a clear indication of the amount of added water is given consumers may find they might not be informed about the true nature of the products they were buying.
37. Even when correctly labelled with respect to the appropriate legislation, the survey has revealed that chicken products bearing declarations about added water content may still be a source of confusion, as different wording and/or quantification is required depending on which Regulation applies. Additionally, not all types of chicken products are covered by legislation which requires a declaration of added water or meat content.
38. The respective results from the two official EC test methods, the Chemical Test and the Drip Test, would appear to lack correlation and therefore requires further investigation.
39. The results of the survey have been circulated to the Food Advisory Committee for information.
1. Joint Food Safety and Standards Group. Survey of Added Water in Cured Pork Products. Food Surveillance Information Sheet No 132, 1997.
2. The Food Magazine, 2000, 50(Jul/Sep), 8.
3. Commission Regulation 2891/93 amendment to Commission Regulation 1538/91. Official Journal of the European Communities, 1993, L263.
4. Commission Regulation 1538/91 introducing detailed rules for implementing Council Regulation (EEC) 1906/90 on Certain Marketing Standards for Poultry. Official Journal of the European Communities, 1991, L143.
5. Commission Regulation (EC) No. 1072/2000 amending Regulation (EEC) No. 1538/91 introducing detailed rules for implementing Regulation (EEC) No. 1906/90 on certain marketing Standards for poultrymeat. Official Journal of the European Communities, 2000, L119/21.
6. Meat Products and Spreadable Fish Products Regulations 1984, as amended. Statutory Instrument 1984 No. 1566. 1984, London: HMSO, 1984.
7. The Food Labelling Regulations 1996, as amended. Statutory Instrument 1996 No. 1499. London: HMSO, 1996.
8. The Food Labelling (Amendment) Regulations 1998. Statutory Instrument 1998 No. 1398. London: HMSO, 1998.
9. Determination of moisture content. ISO Method 1442.
10. Determination of protein content. ISO Method 937.
11. Commission of the European Communities Directorate-General for Agriculture. Physiological water content of frozen and quick frozen chicken and turkey parts. ISBN 92-826-6913-0. 1993.
12. Food Safety Act 1990 Chapter 16. London: HMSO, 1990.