My last article, “Weasel Words One”, was about two words whose definitions included the word “clean”, so what does the word “clean” mean to process engineers?

Let’s start with what I wrote on the subject in my 2018 book, An Applied Guide to Water and Effluent Treatment Plant Design (9780128113103). This was intended to disambiguate the word “disinfection” rather than the word “clean”, but it serves this purpose just as well:

“Three different processes (cleaning, disinfection, and sterilization) are commonly referred to as “disinfection”, so it would be useful to start by defining our terms. Cleaning simply reduces the number of contaminants present and, in doing so, removes a proportion of organisms present. Disinfection removes most pathogenic organisms. Sterilization is the killing or removal of all organisms.”

Drinking water treatment only involves cleaning and disinfection, but the production of water for higher grade uses, such as the water for injection (WFI) used in pharmaceutical formulation requires sterilization.

The analysis of these processes from the point of view of approaching sterilization often involves the use of kill curves, which display a logarithmic relationship between the “dose” of sterilizing agent applied and the percentage of the initial number of organism present.

These curves tell us that three things matter in the removal of all life from a system: how much life you started with, plus both the intensity and duration of your killing agent. This holds true whether we are using electromagnetic radiation, heat, or chemicals.

Therefore, we tend to analyze all three types of process with respect to log reductions, whether that be in organisms in general or pathogens in particular. We frequently use indicator organisms such as colony forming units or coliforms to stand in for the harder-to-test-for organisms which also tend to be markers of fecal contamination.

Consequently, regulators tend to specify a certain number of colony forming units and coliforms per unit volume of water. We don’t know which microorganisms these proxies represent, but we assume that if these easy-to-culture organisms made it through our process, many pathogens did too.

Conventional drinking water treatment processes prior to the disinfection stage give a 4-log reduction in CFUs, a 2-log reduction in coliforms, a 2-log reduction in (far smaller) Cryptosporidium and Giardia cysts and a 2-log reduction in (smaller still) viruses.

Conventional disinfection such as chlorination after such treatment gives a 2-log reduction in CFU, coliforms, and viruses, but no reduction in (far smaller) Cryptosporidium cysts. This last is a problem because at least a 3-log reduction in cryptosporidium is required in many regulatory regimes. Conventional treatment and chlorination may not be enough to ensure safety.

For comparison, sterilization processes are normally specified as providing at least a 12-log reduction in all types of organisms.

So, my starting position is that “cleaning” is just wiping the dirt off. If we want to make something that people are going to take into their bodies “safe” (that’s a weasel word for another day, one which underpins the WHO guidelines for drinking water quality, as well as featuring in the often-confused “process safety” and “personal safety”), I guess we need to look at the nature of that “dirt” we are wiping off, or more generally those “contaminants”. as I call dirt above. Though dirty is the opposite of clean, “contaminants” is the better word here, evocative of the US FDA’s requirements for cleanliness of food and drug processing equipment.*

What then are “contaminants”? To me, “contaminants” are like weeds, things found where they are not wanted. For example, the presence of geosmin, (the stuff which makes beetroot taste “earthy”) is no problem in products containing beetroot, but mere nanograms per liter are unacceptable contamination in drinking water.

Let’s consider contaminants by scientific discipline:

Physical contaminants, whether they are living or dead organisms, organic or inorganic particles, fats, oils, greases, slimes and so on, tend to be dealt with by “cleaning” as I define it above. (So, if that’s what “cleaning” is, what is “cleaning in place”? In a potable water context, it usually just involves cleaning things without taking them to bits, but as commonly used in the food and beverage industry it includes a disinfection (sanitization in the US) step, and is therefore a misnomer for what is actually “cleaning and disinfection in place”.)

Then there are chemical contaminants. No-one wants any botulism toxin in their food and drinks, but some other aspects are more flexible than you might think. I was once surprised by a soapy-tasting Portuguese mineral water which was pH 9 or 10 and tasted like a naturally occurring dilute caustic solution. So, is a tiny trace of the “caustic” used as a clean in place (CIP) fluid in your food factory getting into your product a disaster?

Microbiologically speaking, a food can be fit for purpose when it is merely clean, even only by the five second rule (a rule which does not find favor with FDA inspectors). Most foods and drinks are supplied disinfected, or even sterilized, because they have to be fit for use by a wide range of consumers, which may include people whose immune systems do not work well. Many cheeses are however not even clean, which is why the immunocompromised are warned off.

Are “contamination”, “purity” and “hygiene” absolute terms then, as was suggested to me in a recent design review? I think not. None of the definitions above support such an idea. Engineers need to define and where possible quantify such terms so that they can validate (another important word, in this context again most useful as it is defined by the FDA) compliance with a given specification.

The product of any process which meets its specification has a non-zero, rationally predetermined acceptable level of contamination. In my experience, people can be a bit precious about this issue in the food and drink industry, just as people can in the safety business. Each incremental step we take towards perfection in either respect costs more than the one before it. Even if there were such a thing as perfect cleanliness, it would be neither practical nor affordable, and engineering is a practical, commercially minded business.

How clean is “clean”? Verifiably clean enough, (exactly as safe as “safe” is), I would argue.

“Fitness for purpose” is what matters here, though I’m leaving defining that phrase for another day as well. It’s a form of words which needs tight definition in an engineering context, especially when it is found in a contract, the place where words matter most of all, as ultimately such words are interpreted by lawyers, not engineers.

*“Equipment and utensils shall be cleaned, maintained, and sanitized at appropriate intervals to prevent malfunctions or contamination that would alter the safety, identity, strength, quality or purity of the drug product beyond the official or other established requirements.” Source.

**I don’t mean zero to one decimal place