Taking a Broader Approach to Process Optimization

Sophisticated mathematical approaches to process optimization, in which some objective function is maximized or minimized subject to chosen constraints, are widely published in the literature. We suggest that such optimization problems are often restricted to an unnecessarily narrow domain and that a broader perspective may reap greater rewards.

Questions as to just what should be maximized or minimized, or what are the real constraints as opposed to only those that are immediately apparent, are often posed without a broad enough view of the "big picture." The following are some examples.

• In the manufacture of tomato juice concentrate, water removal by evaporation can be a very energy-intensive and costly unit operation. The objective in optimizing such a process is normally to minimize energy cost and consequently the cost to manufacture. This is often accomplished by finding the optimum number of evaporator effects by an economic balance between the energy savings obtained by multiple effects and the added capital investment.

However, the relatively long residence time as the product flows through the multiple stages in the evaporator can contribute to product quality degradation. Alternatively, a process optimization directed toward maximizing product quality may produce a superior product that could command a higher price and competitive edge in the marketplace, although at a higher cost to manufacture.

• Maximizing quality retention without compromising microbial lethality has always been important in the design of optimum retort process conditions.

Mathematical approaches to this problem have been applied to find the optimum variable-temperature profile that would maximize thiamin retention without compromising lethality, but no more than 2–3% improvement could be realized over using an optimum constant-temperature process. This degree of improvement would be imperceptible in the marketplace and could not justify the complexity of dynamic retort control.

However, consideration of an alternative quality attribute could make such modest improvements extremely valuable. For example, if the water-holding capacity in canned oyster and mussel meats could be improved to increase drained weight by just 1–2%, annual profits to the company could increase significantly enough to justify adoption of an optimum process alternative.

• Significant process improvements can sometimes be realized by simply taking a fresh look at an existing process to identify unnecessary constraints. For example, during the Arab oil embargo in the early 1970s, when energy costs increased dramatically, an engineer was hired by a food company to review the process for manufacturing a spray-dried powdered formula that the company had identified as its most energy-intensive product.

The starting ingredient was nonfat milk with 9% solids, to which soluble carbohydrates were added along with fats and oils and emulsifiers to reach a liquid formula mix with 20% solids. This mix had to be pre-concentrated by evaporation to reach 55% solids prior to entering the spray dryer for maximum spray drying efficiency.

In the standard procedure for making the powdered formula, the nonfat milk was supplied as a spray-dried powder and water was added to produce the 9% solids starting ingredient. This same water was then subsequently removed by evaporation prior to spray drying.

The engineer discovered that the nonfat dry milk powder, along with all the other ingredients, could be mixed directly with just enough water to reach the 55% solids feed needed to go directly to the spray dryer, thus eliminating any need for evaporation, with appreciable energy cost savings.

A potential explanation for our relative blindness to a broader approach to optimization is that we have often used traditional engineering approaches to understand and analyze most food processes. That is certainly a good start, but it may be somewhat limited and might have inhibited us from seeing more broadly. For example, if we consider quality more frequently as an intrinsic and integral part of process design, it will be possible to identify several potential improvements. In some cases, it is necessary to start with an in-depth analysis and synthesis of food processes to answer basic questions like how should the equipment be interconnected, what are the appropriate operating conditions, or should there be recycle streams or purge streams.

The background knowledge and logic that we carry often limit our vision to design optimum food processes more broadly. What one person sees as process optimization may not necessarily be what others could see.

Ricardo Simpson ([email protected]) and Sergio Almonacid ([email protected]), Professional Member and Member, respectively, of IFT, are Professors, Dept. of Chemical, Biotechnological, and Environmental Processes, Universidad Técnica Federico Santa María, Valparaíso, Chile. Art Teixeira ([email protected]), a Professional Member of IFT, is Professor, Dept. of Agricultural and Biological Engineering, University of Florida, Gainesville.