We can’t go on like this!”
Almost every manufacturing operation eventually needs to change. The reasons for this can include expanding or relocating production, manufacturing cost reduction, equipment which is worn out, processes which are no longer acceptable from a hygiene or labour point of view, expansion into a new market place and many others
It is generally essential to make this transition without changing the product in any way and therein lies one of the most real and exciting challenges for the process engineer in the confectionery industry.
It is important to develop a solution which is appropriate, affordable and can be completed in an acceptable timescale. This aspect is important because the new project could be of a very different scale to the forgoing plant and may need a radically different approach to be economically viable. A 100,000 tonne per year plant may not be a good process model for a 5000 tonne per year one, and of course the reverse is also true.
Many (indeed most) products in the confectionery industry owe much of their identity to the manufacturing process by which they have been produced for many years. Those processes were rarely chosen to produce a specific flavour or texture, they were chosen because they were the best available way of making the product at the time it was developed. The relative longevity of plants and successful products in the confectionery industry has meant that there are many old processes still in use which have fundamental effects on the flavour and structure of the resulting products.
Drying is a good example of the ways in which processes have defined products. Many materials are dried as part of their manufacturing process, but the particular technology employed in a given case will have a significant effect on the flavour of the end product. In milk chocolate manufacture all the moisture has to be removed from the milk. In some companies this was done by producing milk powder (spray and roller drying bringing yet another variation), in others by drying a mixture of milk and sugar (“block milk”) and in yet others by drying a mixture of milk, sugar and cocoa mass (“crumb”). These different processes produce distinctly different product flavour profiles because the temperature profile, moisture profile and time taken is very different in each case and radically changes the very complex multi stage protein/lactose chemistry which generates a myriad of flavour components.
These different flavour profiles became ingrained as the characteristic flavour of different company’s products and core to their identities and success. However, when the manufacturing processes have to be changed, or new or larger factories built, getting the flavour from the new process to exactly match that from the old has proved to be extremely difficult and in many cases some change had to be accepted – tasting a single manufacturer’s product in different countries can be extremely instructive in this regard.
The precise ways in which the key flavour and texture characteristics are produced during manufacturing processes are generally poorly understood at best – simply because issues like in process flavour development are often extremely complex in the confectionery industry. This can make developing a new process or commissioning a new plant extremely difficult, particularly if time and cost limit the investigative work which can be carried out. Often it is simply not practical to perpetuate an old existing process in a new location due to cost, equipment availability or differences in scale, so the process engineer has to manipulate the operation of a new plant to produce the required product.
In addition to the effects of basic chemistry, physics and unit operations, there is the “mix” issue. Older processes are frequently batch based with individual batches varying significantly. Mixing these batches together reduces the variation and frequently gives a product which is extremely difficult to produce with a single, highly consistent new process.
Whilst most companies have excellent quality control departments with very capable scientists, few, particularly small and medium sized businesses, have specialist process engineering knowledge in house, simply because it is not a skill required routinely and people within a manufacturing orientated business would have little opportunity to gain broad experience in the field. This means they may need to rely on consultants skills for this type of input because whilst many equipment manufacturers also have resources they tend (quite reasonably) to be directed to selling their equipment rather than exactly meeting the client’s needs on product identity, which are frequently quite difficult to define objectively.
The first step in this matching process must be to try to define the existing product’s key characteristics and the degree of variation currently experienced. Formal or informal sensory methods will (hopefully) give an envelope for each key characteristic within which the product needs to lie. It is essential that this process involves the commitment and agreement of marketing, sales, technical and engineering people and that they all understand each other’s viewpoint. It highlights another problem – the new process may (indeed should) produce a consistent product. However if the product against which it is being compared varies significantly confusion can arise. I shall be forever grateful to a senior QC colleague who, when we were suffering just such a problem and making very little headway, took the position of making the product from the new process the standard – it was more consistent and delivered what the customer wanted – with the result that company sales increased steadily and it came to dominate its market because the improvements in the consistency and quality of the product eclipsed the very minor flavour changes.
The next stage in the change process must be to try to understand the manufacturing process and how various components in it affect the flavour and texture of the product. This often requires not just considerable insight and experience of the product and the process but often sophisticated analytical and sensory analysis of intermediate stages of the process which is rarely easy. As a chemical engineer working in the food industry I often find that the process technology is poorly understood, particularly in smaller businesses, especially in respect of some of the mechanisms of diffusion, heat and mass transfer. A good understanding of the process technology – sometimes on a micro scale – with some creative engineering science can often solve difficult problems. The food industry is particularly interesting in that exact data on the characteristics of many materials and processes are simply not available or vary significantly due to the natural variation of natural materials. This means that broad experience is extremely valuable and often information from quite different industries may be needed for interpolation.
In moving to design new facility using modern equipment and technologies to conduct the manufacturing process many of the same questions arise. Although a unit operation may be conducted much more efficiently it may embody subtle changes which affect the process. Even where flavour may be unchanged small and difficult to detect changes in structure can affect subsequent processes. For example, on one occasion we found a product from a changed upstream process would not make a particular product efficiently because of small changes in the way fat was encapsulated in the structure. We had to use electron micrography as well as insight into the product to understand the changes and to understand how we had to alter the upstream process to correct this.
As a consultant it is important to go into this type of project with a completely open mind and a willingness to learn. Often you will be faced with traditional practices which are poorly understood from a technical point of view, and even the explanations given may be incomplete. Not infrequently there is important “custom and practice” in the operation of a plant which operators are aware of but which managers either do not know about or have dismissed (wrongly or rightly) as being of no consequence. Time on the factory floor is invaluable in actually sorting out a process with plenty of ad hoc tasting. It is also vital to develop a working relationship with operators as they may be defensive or even feel threatened by changes and be unwilling to share information (even more so if you have a language or cultural barrier, so learn enough of their language to work directly with the operators rather than having to rely on interpreters)
Measurement of process conditions and sampling can be extremely difficult as most processes are enclosed and potentially “fugitive” – almost as soon as samples are taken they change and become difficult to interpret. This can apply to both the old and the new processes and makes flavour optimising very difficult even if you have taken the trouble to understand the process, technology and science very thoroughly. Sometimes the opportunity does exist to undertake tests and study the effects of deliberate changes to process conditions, but often output requirements and costs mitigate against this. More frequently the final product matching has to be done during process commissioning and this can require some original thought and result in quite significant changes to the process operation as well.
The detailed process design of a new plant for most confectionery products almost always has to go ahead with less than perfect information available and so inevitably has to rely on experience to a degree. It is therefore essential to examine the implications of inaccuracy of data and to test at least mathematically the effect of this on the design, operation and output of a plant. Sufficient margin must always be built in to calculations to allow for the unexpected, but not so much that a plant is hopelessly oversize and so expensive the project has to be abandoned.
It is absolutely essential and too often forgotten to allow adequate time and budget for commissioning of a new plant and matching the required product characteristics and output. Commissioning time and resources are almost always eroded by delayed project starts, delays in construction, overruns on costs etc. and whilst delivery on time is essential to any project, excessive optimism over commissioning time has led to many a missed marketing deadline.
Commissioning needs extremely careful planning and execution and there will always be scrap material – dispose of it sooner rather than later because storing a mountain of scrap and disposing of it later is more expensive (and a lot more troublesome) than the rather brutal approach of instant disposal – finance and operations will be unhappy when you do it, but even more unhappy when you don’t and it has accumulated notional “value” due to storage and interest charges. And make sure you resist the “suggestion” to increase rework levels in order to get rid of the rework mountain, this will almost always degrade the product and make the task of matching even more difficult.
And finally ………
How to “prove” that the product hasn’t changed? Methods vary from “one man’s opinion” to the use of a bewildering variety of sensory evaluation techniques. Although these sensory techniques are undoubtedly soundly based on statistics, they are, at least in my experience, occasionally capable of throwing up extremely strange results. Rather like the tendency to follow a satellite navigation system up a stream bed, always apply a coating of good sense to the results!
And even more finally ……
Don’t walk away at the end of commissioning!. As you hand over for “Beneficial Operation” the plant is at its lowest efficiency and reliability, it works, but not much more. Much remains to be done to make it realise its true potential and every hour of your time should be handsomely repaid in improved costs and outputs.