Foreign matter contamination is the main source of recalls and rejections and may lead to injury to customers, loss of brand loyalty, and large recall expenses. These undesirable additions differ from food groups and, depending on the type of food product, can be anything from stem stalks to bone fragments.
Numerous emerging technologies and identification methods, ranging from physical examination to ultrasound to sophisticated nuclear magnetic resonance techniques, can detect foreign matter. PacMoore Products (Hammond, Ind.), a dry repack and blending contract manufacturer, meets various client needs that involve repackaging or blending products that include dry ingredients for baking, dairy, meat, or ready-to-eat, micro-sensitive applications. PacMoore utilizes various detection methods to control and detect foreign matter, including visual examinations, screens, magnets, and metal detectors.
A Combination of Characteristics
Product type, flow characteristics, particle size, density, blended components, metal detector capabilities, screen size, and magnet strength all play roles in foreign matter detection, says Scott Reid, PacMoore’s general manager.
The food industry chooses the mesh sizes of its screens based on the product. Figure 1 (p. 51) compares the approximate size of particles that can be seen with the naked eye with different types of detection tools used in the food industry. Mesh sizes denote the number of small squares in a one-square-inch area. The larger the mesh sizes, the finer the screen and detection capability. There are no standards designating which screens are to be used for different food applications; this is dictated instead by the flow characteristics of the product. For instance, it is common to use a 20-mesh screen (840 µm in opening) for dry food applications.
Particles of foreign matter in processing utilities—water; gas; heating, ventilating and air conditioning (HVAC) systems; manufacturing equipment; container closures; and packaging—are intrinsic. And in a dry ingredient process, a food manufacturer can expect some metal-to-metal contact, which may produce a fine dust.
Is Zero Defect a Myth?
Preventive measures like filtration, in-line filtration, at-the-point use filters, screens, and magnets can prevent these issues, but it may still be difficult for a food manufacturer to nullify or control all intrinsic foreign matter. Practical limits based on risk assessment and process capabilities need to be set specific to a food process, and the risk assessment should consider the type of process being used (e.g., grinding versus milling versus just repacking).
“In a company where there are few metal-to-metal moving parts, where equipment handling is minimal, and where gravity and pneumatics are used for material transport, the metal contamination is very low,” says Bill Moore, PacMoore’s president.
But as a proactive approach, any amount of metal dust collected is measured, and product history is developed. “This helps us identify client issues when we handle their products,” says Bob Lyman, vice president for sales at PacMoore. Knowing this type of history helps clients tweak their processes. Magnets are placed prior to introducing the product into repacking/blending operations, something that helps identify the source of metal contamination.
It is common to find metal dust when there is new machinery or when a new supplier is qualified. “We ask if preventive maintenance was done on the equipment recently or if any repair was conducted. If the answer is no, most often the contamination is external to our systems,” Lyman says.
Establishing a Baseline
Inherent process assessments need to be different for different food manufacturers. Most hazard analysis and critical control point (HACCP) plants and clients require metal detection as a critical control point or set criteria for it without establishing a baseline. Even though the Food and Drug Administration (FDA) has established defect action levels, these have mostly been for extrinsic foreign matter such as insect parts and fragments, and the limits are vague.