Good Practices for Pilot Plants

Labs and pilot plants serve a wide range of purposes in universities, companies, and government agencies. The way in which a pilot plant is designed and how it is equipped and staffed are all affected by which mission is most important. Usually, flexibility is a dominant goal because missions can change, and many facilities have multiple missions.

University Pilot Plants
Universities have the simultaneous missions of education, research, and service. How do these affect a lab or pilot plant? In food science and chemical engineering, it is traditional to expose students to at least a few of the fundamental unit operations, such as drying, freezing, pasteurization (heat exchange), retorting, and pumping (fluid flow). If time, space, equipment, and staff allow, other experiences may be presented: membrane concentration, fermentation, milk separation, or chemical reactions.

Typically, a teaching lab or pilot plant uses small-scale equipment, often self-contained with supporting devices, such as feed tanks, instruments, and pumps. Suppliers of such units include MicroThermics (phone 919-878-8045, www.microthermics.com), Armfield (phone 732-928-3332, www.explorearmfield.com), and OMVE (phone 31 (0) 30 241 00 70, www.omve.com). Small-scale equipment always seems expensive for its capacity, but that is because there is as much, maybe even more, engineering and labor involved in designing and fabricating on a small scale as there is on a larger scale. Durability and quality of fabrication can be an issue at small scale because all the parts are smaller and possibly more fragile. In addition, intermittent operation by relatively inexperienced people is a challenge to the health of small equipment.

Many university pilot plants are out of date because they have been neglected as experienced faculty retire and newer faculty with other interests join departments. There may be a perception that the traditional unit operations are less fashionable than newer technologies, such as high pressure and pulsed electric fields, for which off-the-shelf small-scale equipment may not be available or is very expensive. Despite the challenges, industry recruiters want food science graduates who know and have experienced the basic unit operations that are common to most food processing. They do not expect food scientists to be engineers, but they do hope that food science graduates will have some sense of the capabilities and limitations of the equipment found in a typical food plant.

This hope, and the Institute of Food Technologists’ guidelines for food science academic programs, suggest that every food science department should have a well-equipped lab or pilot plant and a course that employs that facility to expose students to the fundamental unit operations, much as chemical engineering departments do. What should be in such a lab?

--- PAGE BREAK ---

At a minimum, there should be experiments or demonstrations of heat transfer, drying, mixing, fluid flow, evaporation, and biochemical reactions. Other possibilities include solids flow, humidification, solids mixing, emulsification, extraction, cooking, baking, and freeze-drying. Depending on a university’s history and resources, a course and lab might be built around a particular local industry, such as dairy, cereal grains, fruits, or meat processing. Sometimes such an approach becomes a substantial enterprise in its own right, possibly diverting resources and attention from the educational mission. Thus, some universities become known for their cheese, ice cream, jams, or sausage, often hiring staff to help keep the shelves of a store full. Whether this is a good use of resources in times of tight budgets is for each institution to decide. A new department, or one refreshing its facilities probably should avoid such an approach, but it is sometimes hard to abandon traditional practices. Further, teaching unit operations in an integrated and functional approach can be effective.

Research and Service
Research and service impose different requirements on a food pilot plant. For research, equipment might need special fabrication, especially for data collection or because a phenomenon of interest is not yet implemented commercially. Equipment constructed for research rarely lends itself to teaching, although there can be exceptions. If the same facility is used for teaching and research, because of the availability of space and utilities, there can be conflicts over schedules and resources. Storage of raw materials and ingredients needs to be organized, and labor may need to be shared and expenses allocated. None of these issues are insurmountable, but they need to be considered in the design and management of a facility.

The service function presents a distinct crop of new issues, depending on what service is rendered. Land grant universities and other public universities have an obligation to support taxpaying citizens, including food processors and entrepreneurs. Usually, this means providing advice, helping with regulations, or helping to identify co-manufacturers. However, some may want to manufacture small quantities of a new product or develop a new process. In such cases, the facility must conform to good manufacturing practices and be capable of performing all the steps in manufacturing, including packaging. Optimum efficiency may not be necessary, but complete performance is, which may require equipment that otherwise would not be present or needed.

Another instance where a more integrated system might be desirable is in teaching short courses or workshops. Some universities offer such courses in baking, winemaking, brewing, and cheese making, requiring equipment for all the steps in manufacturing. Such equipment may be specialized and only used a few times a year, but the courses can be popular and a source of income. One solution is portable equipment, which can be stored when not in use.

Industrial Pilot Plants
Industrial pilot plants have their own unique missions, including product and process development, simulation of process plant situations, and manufacture of test market products. Common to all of these is the need for equipment to reproduce realistic plant conditions. It is not usually necessary that pilot plant equipment be highly automated nor that it be completely integrated. Often, continuous processes can be simulated by batch operations, and automation or mechanization by manual labor.

The more critical elements are that pilot plant equipment applies shear and heat in amounts and rates that reproduce plant conditions. This requirement encounters the challenge of geometry. Equipment that preserves geometric similarity between two scales, meaning that ratios of dimensions remain constant at two different volumes, will depart in other ratios. For example, the surface-to-volume ratio of a jacketed kettle becomes smaller as the kettle volume increases. This means that heating or cooling rate in a cook kettle will change upon scale-up because heating is conducted through the jacket (proportional to the surface area).

--- PAGE BREAK ---

It is usually impractical to have full-scale equipment in the pilot plant, so what is good practice? One approach is to understand by testing whether heating and cooling rates are significant in a given case; sometimes they are not, but often they are. For instance, in hot filling, it is the time in the container that counts so the size of a kettle does not matter, at least so far as the adequacy of the thermal process is concerned. On the other hand, many foods can be damaged by overcooking so a product made successfully in a small kettle may taste different if made in a larger one. Adding external heating or cooling heat exchange area can compensate, but then the product is subjected to additional pumping, which could damage particulates. It is a challenge.

Extrusion is another unit operation in which scaling can be difficult. Likewise, mixing and emulsification may be different at different scales. Understanding the fundamental phenomena is critical, but is often overlooked in the eagerness to get a formula correct.

A good industrial pilot plant needs at a minimum small-scale equipment that enables people to make the company’s existing products realistically and with minimal waste and other costs. In addition, the pilot plant should have space and flexible utilities so that new types of equipment can be installed for testing. This may require some redundant support equipment, such as feed tanks, mixers, and pumps.

Companies should use their pilot plants for small production runs very selectively, if at all. If product is to be consumed, it must be made safely, and for some classes of product, the facility must be registered with the U.S. Dept. of Agriculture (USDA) or the Food and Drug Administration (FDA). USDA regulates foods with more than 2% meat or poultry and egg processing. FDA regulates all other foods and requires registration of all facilities that manufacture food in interstate commerce. State departments of public health may also regulate food manufacturing, for example, of dairy products. Such regulations may limit the flexibility of a food company to explore limits of a process.

The culture of manufacturing is different from that of research and development, meaning that research personnel may not be well suited for the routine of manufacturing and rarely are as efficient as well-trained plant people. On the other hand, plant trials with full-scale equipment are wasteful and expensive because they are usually done on weekends, with premium pay. One solution to that dilemma is a semi-works, designed for short runs, with small production equipment.

J. Peter Clark,
Contributing Editor,
Consultant to the Process Industries,
Oak Park, Ill.
[email protected]