I have been to hundreds of food processing plants in the U.S. from canning and bottling, vegetable and fruit processing, kill plants on through to further processing. All have their rules and regulations assuring consumer safety. Food-processing facilities operate under an entirely reasonable expectation that lubricants used around machinery where incidental food contact could occur should present as little risk as practical, which is why National Sanitation Foundation (NSF) H1 lubricants have become such an important part of lubrication programs throughout the food and beverage industry.

Lubricants should be formulated from substances acceptable for incidental food contact because machinery requires lubrication, seals occasionally fail, grease can migrate, oil can leak. An unexpected encounter between lubricant and food does not automatically create an unacceptable toxicological hazard.
There is nothing particularly wrong with that reasoning, but there may be something incomplete about where we have placed our attention.

The regulatory foundation behind H1 lubricants is largely found in the U.S. Drug and Food Administration (FDA) 21 CFR 178.3570, which establishes requirements for lubricants with incidental food contact and identifies acceptable ingredients along with conditions governing their use. Certain lubricant constituents are associated with limits such as 10 parts per million in food, while others have different restrictions, so it would be incorrect to characterize H1 as simply establishing a universal 10-ppm limit for every lubricant. Nevertheless, the existence of these concentration limits raises an interesting practical question that deserves considerably more attention from reliability professionals.

How exactly do we know how much lubricant contacted the food?

That question is surprisingly difficult to answer once we leave the regulatory document and walk onto the production floor.
Take for instance a bearing mounted above a conveyor carrying a food product, where a deteriorating seal allows an occasional droplet of oil to escape onto the moving product below. One droplet contacts one piece of food, another falls harmlessly between products, several hundred pieces receive nothing at all, and another receives considerably more contamination than the average concentration calculated across the production lot would suggest.

We now have a problem that is not simply analytical chemistry, because before we can accurately measure the contaminant, we must first collect a sample that actually represents the contamination event.

I have worked in analytical laboratories and authored books on chemical analysis, and I know modern analytical instrumentation is extraordinarily capable. Techniques such as gas chromatography, mass spectrometry, liquid chromatography and infrared spectroscopy can detect remarkably small concentrations of particular lubricant constituents when appropriate analytical methods have been developed. What these instruments cannot magically solve, however, is the fundamental problem created by random, heterogeneous and intermittent contamination, because an analytical instrument can characterize only the sample placed before it, and the most sophisticated instrument in the world cannot measure contamination that happened to land on a product that was never sampled.

This is where the discussion becomes a reliability problem rather than merely a lubricant-selection problem.

If lubricant is escaping from a gearbox, bearing housing, hydraulic system or lubricated chain positioned near exposed food, we should be asking why the lubricant escaped from the machine at all.

• Was the seal damaged?
• Was the bearing housing overfilled?
• Has shaft misalignment accelerated seal wear?
• Has the sealing surface deteriorated?
• Is internal pressure forcing lubricant through the seal?
• Is the lubricant chemically incompatible with the elastomer?
• Is an automatic lubricator delivering excessive grease?
• Is a breather blocked? Has a gasket hardened, cracked or lost compression?
• Has lubricant consumption changed gradually without anyone noticing?

Those questions move us upstream from consequence management toward failure prevention.

H1 is necessary but only the foundation of lubrication strategy, not the whole house

The use of an H1 lubricant remains entirely sensible because no mechanical system can be guaranteed never to leak, and reducing the potential consequence of an unforeseen contamination event is sound engineering practice. The problem begins when the designation unintentionally creates the perception that leakage has somehow become acceptable simply because the material leaking from the machine is food grade. 

A lubricant carrying an H1 registration should be considered one layer in a larger food-safety and reliability architecture, much as secondary containment, alarms, protective devices and emergency shutdown systems provide additional layers of protection in other industrial environments. These controls matter tremendously when something goes wrong, but their existence does not eliminate the responsibility to prevent the initiating failure.

The more useful hierarchy would therefore begin with:

• equipment integrity and containment
• leakage prevention
• condition monitoring
• inspection and rapid maintenance response.

H1 lubricant selection provides an additional consequence-reduction layer should those controls fail.

This is important because while H1 addresses the consequence of lubricant escaping, seal integrity addresses the probability of lubricant escaping in the first place. Reliability engineering has traditionally been strongest when it works on the probability side of that equation.

Contamination prevention starts with better inspection at the barrier
A more mature food-processing lubrication program would therefore place considerably greater emphasis on the components separating lubricants from food products and physical barriers intended to keep lubricant where it belongs, including:

• seals
• gaskets
• shaft surfaces
• bearing isolators
• breathers
• lubricant reservoirs
• automatic lubrication systems
• drip shields.

The condition of those barriers could become a measurable component of the plant’s food-safety and reliability program.

Lubricant consumption itself can provide useful information because a gearbox that historically requires little or no makeup oil but suddenly begins consuming lubricant is telling us something about the integrity of the system. Likewise, an automatic grease lubricator whose consumption rate unexpectedly increases, a bearing housing repeatedly found wet around the seal, or a gearbox developing oil residue along a shaft should be regarded as early indications that one of the barriers protecting the product may be deteriorating. Condition monitoring could therefore extend beyond vibration, temperature and oil analysis to include lubricant integrity as an asset-health parameter. 

Instead of asking whether a lubricant is safe enough to contact food, we would ask whether the machine is reliable enough to prevent that contact from occurring.

Metrics for predicting lubrication deterioration: Measure what we can actually control

There is an additional advantage to this approach because seals, gaskets and lubricant consumption are things that can actually be inspected, trended and managed systematically, by:

• establishing acceptable leakage criteria
• documenting seal condition 
• trending lubricant makeup volume
• inspecting critical lubrication points positioned above exposed food streams 
• identifying machines where a lubrication failure creates a direct contamination pathway
• improving sealing arrangements
• installing shields
• correcting overgreasing
• selecting more compatible elastomers
• redesigning equipment where necessary.

Most importantly, we can detect deterioration before the lubricant reaches the food. Trying to determine afterward whether an intermittent contamination event resulted in 3 ppm, 8 ppm or 15 ppm somewhere within thousands of kilograms of heterogeneous product is a fundamentally different proposition, because the uncertainty associated with sampling may overwhelm the apparent precision of the analytical measurement itself. A very uncomfortable truth is that sometimes we forget that measurement precision does not compensate for sampling uncertainty. A laboratory might report a beautifully precise analytical result to several significant figures, but if the contaminated portion of the production lot was never included in the sample, that precision tells us very little about what actually occurred on the production line.

The disciplined management of uncertainty: Intervening before deterioration becomes contamination

None of this diminishes the importance of H1 lubricants, because selecting a lubricant appropriate for incidental food contact remains a necessary risk-control measure wherever such contact could reasonably occur. What it does suggest is that the lubricant itself should not become the centerpiece of the contamination-prevention strategy when the more fundamental engineering issue is the integrity of the machine containing it.

Reliability has always been, at its core, the disciplined management of uncertainty, and food-processing equipment presents precisely that challenge because we cannot know exactly when every seal will begin to deteriorate, when every gasket will lose integrity or where every escaping droplet might eventually land. What we can do is systematically reduce that uncertainty by inspecting the barriers, monitoring the equipment, understanding the failure mechanisms and intervening before deterioration becomes contamination. Perhaps, then, the food industry could spend a little less time asking how much lubricant is permitted to reach the food and considerably more time asking why the lubricant was able to reach the food at all.