For the past couple years, people coming to work at the Post Cereal Brands plant in Jonesboro, Ark., have changed out of street clothes into uniforms and shoes worn only at the factory.
Other cleanliness refinements in recent years include significant investment in dry cleaning methods, increased vector swabbing, efficiencies from more sensitive instant-read tests and Zone 3 area sanitizing updates.
The plant’s approach to cleaning and sanitation illustrates how vast quality teams’ responsibilities are. We’ve gathered experts’ best practice tips for how to make those jobs easier and avoid common problems. And they shared insights about the latest in cleaning technology and improvements to scientific methods for validating sanitation.
At the Post plant, which makes 19 varieties of cereal, people who clean the machines sign documents stating that the equipment is food safe, Plant Manager John Kompa says.
“In terms of building the culture of safety, we recognize that we serve food to men, women and children all over the United States. We take that very seriously,” he says.
Keeping the work standard
The Post plant worked with Myrtle Consulting Group to implement a new cleaning operations procedure, and QFS Scientist Bob Storey says one of the “biggest wins” was implementing short interval control meetings throughout the sanitation cycle to discuss handling obstacles effectively and efficiently.
The planning helps sanitation windows run smoothly, alleviating pressure from multiple teams, including maintenance and engineering, competing to access machines during downtimes, he says.
Rigo Frias, an El Paso, Texas-based director with Myrtle, says food processors can improve cleaning and sanitation by ensuring work is done the same way every time.
“If how you’re performing the job is not consistent, it will lead to potential contamination or to your equipment not being fully sanitized,” he says.
He regularly sees companies with less robust programs than the Post plant neglect parts of sanitation and cleaning. “Organizations often fail to truly understand or to truly document the end-to-end sanitation process,” he says.
Myrtle helps implement plans using SMED (Single-Minute Exchange of Dies) time-saving techniques, and Frias gives examples of deficiencies solved and efficiencies gained:
- Employees have the right cleaning tools, which can be as simple as buying a second hose to wash two lines at once
- Supplies are easy to find and close to the work area
- Pre-startup checklists clarify cleaning tasks and products to use
- Cleaning instructions use visuals and are easy to comprehend
- Tasks that can be done before the line goes down are identified; other steps might be eliminated or done simultaneously to speed up the process
- Staffing is addressed to counter a common problem of insufficient supervision of cleaning processes
- Planning meetings are held for each full or partial sanitation cycle; review sessions identify needed improvements
“It’s really about the race car mentality,” Frias says. “Once the production line goes down, everyone’s ready to go with everything we need. Sometimes they can’t find the tools or the chemicals or they’re walking around. That’s a lot of lost time.”
Errors can happen when sanitation takes longer than expected and the cleaning team is under pressure to start up the line, he says. “It’s just human nature to have people take shortcuts to complete sanitation in the timeframe.”
Rooting out inefficiencies
Steven Landers, category manager for chemistry and hand hygiene at Veritiv Corporation, uses lean processes and 5S methodology to search for waste in cleaning and sanitation practices.
“Quite often, I’ll see an organization that’s very buttoned-up, and they have the right plan, the procedures and they’re doing a good job of protecting the consumers, but their supplies are inefficiently arranged and … just stuffed into an area.”
Landers looks for defects, including:
- Soil leftover in the work area, which could prompt re-examination of how results are verified
- Procedure lists not located in sight of workers
- Supplies that aren’t kept in standard, accessible spots on a wall or on the ground, such as inconveniently located hand-cleaning stations
- Testing methods that aren’t the most accurate available
- Chemistry problems: weak dilution, too much of a chemical causing a film, too many products, products that are too harsh and could cause pitting or rust
One chemistry defect example is using a product that requires many minutes to disinfect when one that works in seconds is available, he says. “That’s a big deal. That’s downtime. That’s labor that is not being utilized. It is about chemical selection.”
Like the Post cereal plant, companies must pay more attention to nonfood areas—from office spaces to the loading area to the often forgotten, dirty toolshed out back, Landers says.
“They’re putting so much effort into cleaning the meat slicer, and they’re not remembering the area where all the people congregate in the breakroom has to be just as clean.”
If a control point fails, Veritiv can help companies figure out why, Landers says. A consultation is “from the whole facility standpoint, and we’re able to help them see these ways that they didn’t know were there.”
Ins and Outs of Validation
Sound scientific backing is key to validating processes, Landers says. “Validation is making sure that you’re doing things in an authoritative way. You’re not using junk science, and you’re actually doing something that is scientifically proven and factual.”
Tony Lupo, director of technical services for Neogen Corporation food safety company, explains that preventive controls have formalized the differentiation between validation and verification, terms that many people in the industry used interchangeably in the past.
According to Lupo, the proper way to describe the terms is:
The more rigorous of the two, with workers likely climbing into areas they don’t regularly access.
- Put best practices in place to ensure that allergen residue or pathogens can be removed consistently
- Add different interventions if results are inconsistent
- Before declaring a process validated, reproduce successful results at least three independent times, typically using the most sensitive, quantitative methods available
- Establish how often validation will be repeated in the future
- Identify changes that would prompt validation outside the regular schedule, such as updates to formulas, plant engineering or chemical suppliers
More daily, routine testing that doesn’t necessarily involve every nook and cranny.
- Monitor to make sure that operations haven’t drifted outside parameters of validated processes, typically using a qualitative approach
- Rely more on binary types of tests, using methods such as lateral-flow monitoring or ATP rapid testing (instead of total-count testing)
While regulations outline standards that companies will be judged by, the specifics of how to test and how to clean are up to them, Lupo says. “The only real requirement is that you are in a sound, scientifically defendable position, with data and some logic behind how you’ve chosen what to do and why to do it.”
The science of testing
Lupo sometimes encounters a lack of understanding about how many sample points are needed. Testers might take one small sampling that comes back negative and erroneously think they’re in the clear.
“You may not necessarily have a statistically valid sampling that is representative of an entire batch or your entire process.”
Newer companies that are still understanding the technology sometimes choose tests that don’t give them the exact answer that they think they’re getting, such as failing to detect a particular ingredient. They can end up with inappropriate products when they buy off-the-shelf or from an internet search without asking questions, he says.
He recommends that companies ask test suppliers for data to verify that they selected the right technology. “Understand how it’s been validated, and make sure that it’s fit for purpose.” When audited, companies will need that data, he points out. “Being able to defend the selection of that technology is important.”
Companies need to test “anything and everything” at first to understand where risks are, Lupo says. Based on the frequency that those risks occur, perhaps when running a specific product, safety teams can determine the chances of problems if they tested less often. Then they can decide how to pare down to the appropriate testing level, he explains.
Plants typically are testing every changeover for high-risk products or ones that are inherently difficult to clean up, such as products with flecks that disperse. For more uniform, homogenous products that are simpler to clean up after, producers might test less frequently.
Old or poorly designed machines also will need more frequent testing because they harbor more food residue, he notes.
Starting clean with design
Ken Lento, senior business strategist hygienic development for North America at FlexLink Systems, says he’s told his family to stop eating products that he’s seen made on unhygienic equipment.
“With some pieces of equipment … I’m convinced: I don’t care how much cleaning you do, how much brushing you do, how much foaming, how much rinsing, you’re not ever going to get them clean,” he says.
Old machines have the most problems, but new machines can still have hygienic design flaws, he says.
Lento gives two examples of design problems:
- Hollow tube legs: In theory, all joints are sealed well, but if a pinhole leak breaches a weld, moisture and microbes can get inside. “You basically have an incubator for bacteria.”
- Metal-to-metal contact: When bolting parts together, FlexLink avoids metals touching by using a sealant and washers. When metals pieces touch, the material corrodes a bit and forms niches where bacteria can grow.
FlexLink third-party tests its conveyors, covering them in a paste of soured milk and fats with a dye, used to spot contaminated parts after cleaning. Imagine trying to get that paste out of corroded metal, he points out. Early designs are improved based on several rounds of testing.
Such extensive testing is not usual, and conveyors that cost a little less might not come with assurance that they’ll wash up to the highest standards, he says. “We’ve been going into companies and saying, ‘Do you want to buy a compliance tool, or do you want to buy a conveyor?’”
Lento suggests asking machine companies if they can back up hygienic design features with testing, and he warns to watch for claims that companies follow 3-A Sanitary Standards instead of completing more stringent measures to get 3-A Certification.
“If I was a customer buying equipment, I would demand that there be some validation testing of the equipment I’m buying,” he says. “I don’t want to be the validation for your design when I’m using it for production.”
With FSMA, companies not only must show their cleaning processes, but also must prove that they’re confident the equipment is coming out clean, Lento says. While many companies do all they can to keep clean and safe, he doubts some processors are willing or able to invest in upgrading outdated equipment that’s hard to clean.
“They’re going to try to get by,” Lento says. “They’re going to try to clean as much as they can. They’re going to swab conveniently so that they don’t get positives where they have to take action. One of these days the FDA shows up, and that’s probably when things might have to change.”
Sensors as validation helpers
One way to get the proof that cleaning processes are sound is to monitor continuously with sensors and transmitters. Emerson Automation Solutions is launching an equipment line for the food and beverage industry, and its sensors can ensure cleaning processes are doing what they’re supposed to do.
Temperature gauges show whether cleaning solutions are hot enough for long enough. Pressure controls ensure tank levels are correct, potentially preventing cleaning solution overflow. Conductivity meters help monitor for detergent strength declines that might go undetected with manual checks, says Josh Friesz, global product manager for Emerson’s Rosemount Hygienic Products.
“You are documenting that, over the course of your entire cleaning procedure, you have the proper cleaning concentration and you have a calibrated instrument that is proving that for you, so if you ever need to submit any documentation for process validation, you have full coverage,” he says.
Most clean-in-place (CIP) equipment is set up to run cleaning steps for standard amounts of time. But sometimes those steps are set to run longer than needed, and clean water is flushed down the drain, Friesz says.
“If you’re utilizing conductivity sensors, you can know exactly when your water is running clean,” he says. “You’re cleaning only as long as you need to, and you’re more efficiently using all of your resources: your chemicals, your heating utility system and water.”
More and more testing
At the Post plant, Quality Manager Tom Gleason says increased swabbing for Listeria and Salmonella fits well with FSMA requirements, but the change was driven more by a facility re-evaluation in recent years, with the help of a corporate microbiologist. The review covered Zone 3 areas and resulted in more tests during any construction to make sure contaminants weren’t dragged in.
Better tests with quick, in-house results have improved verification, and processes are revalidated regularly. “Definitely the capabilities are much better than they used to be, and that helps us out, especially with the instant techniques that are available now,” he says.
Lupo says that sales of Neogen’s testing supplies are increasing because of FSMA and swabathons, although not typically because companies need corrective action after inspections. Instead, they want to be prepared to defend their processes in detail.
“If they have good, strong practices in place to have data accumulated over time, it will put them in a defendable position to say, ‘We’ve done our due diligence: Here, come and look. Here’s all of our historical data that shows that we feel really good about what we’ve done.’”