Lesson 8 - Quality Management End of Unit SAQ

***A new brand is being launched into the marketplace and you have been tasked with compiling a production manual for its production across multiple sites. The beer is a citrus-accented pale ale with no sulphur notes or diacetyl present. The manual would include analytical specifications, a flavour profile, and guidelines on sensory assessment of the beer.

a)    Explain how you would develop the specifications for the beer.

b)    Explain how you would develop the model flavour profile for regular trueness-to-type testing of the beer.

c)    Which flavour attributes would you expect to see in the standard? Write a short procedure for assessors which guides them in how to undertake the trueness-to-type test and calculate the scores.

Response:

a) The specification for the beer should reflect the needs and wants of the consumer. It should also be within the capability of the process and reproducibility of the method of analysis for the parameter being specified. It is represented by a target and then a band width around the target. The target value for each parameter would have been generated during prototype development so will be known. All that is left to calculate is the band width of the specification. To develop bandwidths, we can use process data and analytical data. There are two approaches to using these, using SPC principles or using ISO 4259. In the SPC method the following equation is used.

 Cp = A1 – A2

          6 x s

Cp is the capability of the analytical or production process. The higher the Cp the fewer defects there are. A Cp of 2 is normally used.

A1 – A2 is the band between the upper and lower limits of specification. 

6 x s is six times the standard deviation of the data used.

As it is a new product it is unlikely that sufficient data from the brewing process is available so using the R₉₅ for the analysis method would be a more robust means to calculate the band width. ISO 5725 defines R₉₅ as 2.8 x standard deviation so to calculate band width using R₉₅ we use this equation with 2 as the process capability.  

It could be argued that the process used to brew the new beer is so similar to others brewed in the brewery before, that data from the process for a similar beer could be used. If that approach is used the standard deviation of a set of process data could be inserted into the equation above in place of to give the band width.

ISO 4259 recommends that the specified band width should not be less than 4 x R₉₅ and specifies the upper and lower limits of the band width as ‘not greater than’ and ‘not less than’. As ISO 4259 relates to petroleum products it should only be used to set specifications where the homogeneity of the samples is very high.

As all specifications need to reflect the needs and wants of the consumer, if the band width allows variation which would be unacceptable to the consumer, either the analytical process or production process needs to be improved.

b) In order to develop a flavour profile, descriptive flavour profiling would be used. The panel undertaking the flavour profiling would have been screened for sensory blindness for key elements and trained for around a year in sensory terms. They would have a working knowledge of the flavour compounds of interest, developed through the use of samples spiked with flavour standards. They would also have a working knowledge of the intensity scales used in flavour profiling.  

The process would establish the key flavour attributes of the beer through evaluation of around 50 before narrowing down to 15 to make the truenes- to-type test workable without sensory fatigue becoming a factor. 

Evaluation of the beer would be undertaken by the assessors in sensory booths in isolation. They would not know what the beer was or how it was designed to taste. After tasting and the collection of the tasting sheets the beer would be discussed and a consensus reached on the key elements of the flavour profile. A spider plot would then be produced from the data to help with the training and familiarisation of the trueness-to-type flavour panel. If possible, the same panel would undertake the flavour profiling and the trueness-to-type tasting.  

c) The beer is a citrus-flavoured pale ale so fruity and citrus would be in the standard attributes along with the usual bitter/sweet/hop notes. It should be free of sulphur notes and diacetyl so drains, rotten egg, baked beans, struck match, and butterscotch would be in the negative attributes section.

Standard attributes

• Fruity

• Hoppy

• Citrus

• Caramel

• Malt

• Sweet

• Bitter

• Body

• Dry

• Linger

Negative attributes

• Bad eggs

• Drains

• Struck match

• Butterscotch

• Baked beans

Test procedure:

1. Taste the beer and rate each standard attribute with reference to flavour profile.

2. Score the closest match for each attribute with a 3.

3. Score a near match with a 2.

4. Score an approximate match with a 1.

5. Score no match with a 0.

6. Taste the beer and rate each negative attribute.

7. Score the negative attributes on the 1-3 intensity scale only if you consider them to be present.

8. If they are not detectable score 0.

9. If they are subtle score 1.

10. If they are obvious score 2.

11. If they are pronounced score 3.

12. Sum the standard attributes scores.

13. Sum the negative attributes scores.

14. Subtract the negative attribute score from the standard attribute score and divide the total by the maximum possible score from the standard attributes.

15. Multiply the answer by 100 to give % trueness to type.