Yesterday’s article outlined Dr Morley’s top 10 tips for efficient troubleshooting. Today he gives us practical case studies of troubleshooting in action.
“At Leatherhead we receive all manner of trouble-shooting enquiries covering a diverse range of products and issues such as the following: Variable viscosity; Discolouration; Separation and release of oil or water; Sticky/clumping particles or pieces; High or low acidity; Chocolate confectionery bloom development.
These are in addition to the wide range of foreign bodies that come through our doors, including in recent times glass, plastics, insects and other small animals, larvae, pins and needles, and metal balls.
Of course the trouble-shooting work that we carry out is highly sensitive and confidential, and finding a solution if often urgent, but some case studies can be discussed in general terms in order to demonstrate how the top 10 tips can be used to diagnose the problems and propose solutions.Variable viscosity sauces
In this case it had been noted that two samples of sauce were of varying viscosity and we were asked to propose the changes that would be required to achieve a match. This type of enquiry could of course relate to a competitor benchmarking activity as much as straight-forward trouble-shooting resulting from a change in ingredient or process.
In this particular case we moved straight to tip 3 – analytical tests. In this case we made up the 2 samples according to the client’s instructions and then measured the viscosity as a function of increasing shear rate using the Bohlin controlled stress rheometer. From these data we were able to predict the increase in the concentration of sauce solids that would be required, and then test this out with further viscosity vs shear rate plots.
This ‘solution’ brings in tip 10 as the quick fix was to make a formulation change even though the root cause was not fully understood. The final solution for implementation could have ranged from an increase in the total solids or of one or more specific ingredients such as the gums or starches. We were unaware of the specifics of the formulation so this final aspect was left to the client.Separating mayonnaise
Mayonnaise can be a highly temperamental product. Full-fat mayonnaise, containing up to 80% fat, is likely to split and separate into an oily mess (try making it at home!), whilst lower fat ‘light’ mayonnaise contains starch which is susceptible to damage in the highly acidic product matrix and through the high shear colloid mills used in manufacturing.
One issue was the thinning of ‘light’ mayonnaise in a coleslaw application. In this case we started with tip 1 and kept it simple by visually inspecting the mayonnaise where it was in contact with the pieces of carrot or cabbage. The mayonnaise here appeared to be thinner than in the bulk so there was clearly an interaction with the vegetable pieces. The microscopy according to tip 2 demonstrated that some of the starch granules had been damaged which was consistent with a low viscosity.
The hypothesis therefore was that the mayonnaise had suffered enzyme damage to the starch granules, which could be resolved preferably by pre-treating the vegetable pieces or by replacing the mayonnaise with a higher fat version with less starch. The final solution for implementation was left to the client.
Another issue was separating mayonnaise following a transportation test. The so-called ‘bumpy road’ test, or preferably a simulation involving shaking the product, is an effective assessment of the potential abuse that the supply chain can inflict on a product.
However in this case the client’s transportation test was 2,000km at 40ºC! The mayonnaise that was separating had been imported by the client, whereas the client’s own product was much more stable.
The first step was to invoke tip 5 and interrogate the product design and specifications. The client’s own product was full fat, whilst the imported product was slightly lower in fat and contained a small amount of starch.
Furthermore the starch was a cold-swelling type for an ambient-manufactured product. This last point was key – mayonnaise manufacture using colloid mills can be a very high shear (energy) process, so highly robust starches are required.
The best starches tend to be highly modified cook-up versions, therefore the advice to the client was to minimise the usage of the cold-swelling starch either through switching to a cook-up version or by increasing the fat level. Again the final solution for implementation was left to the client.
The final separating mayonnaise case study required detailed investigations according to tip 6 – finding out what had changed. In this case the client was producing the same product for filling into glass jars and plastic bottles, but in the latter case a small amount of liquid was also present in the bottle.
The client believed this to be aqueous phase which was consistent with excessive shear (energy) forces through the filling nozzle resulting in water being ‘squeezed’ out of the product matrix. As the filling speed and nozzle design had been recently changed, the client wanted a quick fix, for example a formulation change, to solve the problem.
Our investigation started by keeping it simple (tip 1) and examining the separated liquid. In fact it was oil and not water, suggesting an emulsion stability issue. This was confirmed by microscopy which demonstrated a larger oil droplet size for the mayonnaise in the plastic bottle when compared to the glass jar.
It was therefore proposed that the emulsion was inherently slightly unstable and that a small amount of coalescence was occurring through the filling nozzle of the plastic bottle line. A potential course of action for increasing the emulsion stability is the addition of more egg yolk and this was proposed to the client.
Variable viscosity cream
Many ‘natural’ products, or those comprising predominantly ‘natural’ components, can be highly susceptible to quality changes through seasonality and animal feeding changes. Typical examples of this type include dairy products such as cream and butter, and an issue as reported to us was of double cream (48% fat) from 2 factories that were of varying viscosity.
Furthermore the cream from one of the factories was resulting in more consumer complaints of the ‘too thick’ type. Dairy cream, as for all emulsions, is inherently unstable and tends towards coalescence and the separation of the oil and aqueous phases.
As the emulsified oil phase droplets are semi-solid, the instability manifests itself as aggregation and an accompanying increase in viscosity. Hence the hypothesis in this case was that either the product itself or the manufacturing process was resulting in excessive instability at the point of filling.
The lengthy investigation of this issue utilised elements of tips 1-6, including a detailed comparison of the two manufacturing processes and it was eventually concluded that 2 factors were probably at work.
Firstly the homogenisation, a highly complex and poorly understood unit operation, appeared to result in emulsions that had similar droplet size distributions but differing stability. Secondly and probably more importantly, the temperature control post-homogenisation was less effective with higher temperatures for longer periods of time even though the process specifications were the same.
Therefore in summary one factory produced homogenised dairy cream droplets that were less stable and were then subjected to more temperature abuse than the other factory. So it is no surprise that the product was occasionally ‘too thick’.