Report on hygiene experiments with Flat-Flex^® and Plastic Modular belting

Executive Summary

The extensive research done by Mark Vickery at South Bank University clearly shows that stainless steel Flat Flex^® belting is far superior to plastic modular belting, both in terms of hygienic design and cleanability.  Across the wide range of products tested the Flat Flex belt was at least 10 times cleaner than the plastic modular belt and in some cases more than 100 times cleaner.

The increased openness of the Flat Flex^® design over plastic modular allows less build-up and enables easier cleaning with visual inspection of drive shafts without dismantling.  In contrast plastic modular is almost impossible to clean without dismantling and it is also impossible to do visual checks for cleanliness internally.

Stainless steel’s superior qualities in terms of resistance to damage, ease of cleaning and efficacy of cleaners used make it the material of choice over materials such a Polypropylene, Polyethylene, Polyurethane, Polyvinylchloride, Polypropylene and Mineral Resins.  Stainless steel should be used in preference to these above-mentioned materials, particularly in areas were cleaning is a problem in terms of accessibility and extended production runs and where hygiene is of paramount importance.  Stainless steel Flat Flex^® should therefore be the belt of choice where hygiene and ease of sanitation are important. 

Experiments 

The aim of the project was to establish whether stainless steel Flat Flex^® belting (FF) was more hygienic than Polypropylene modular belting (PMB).  A series of experiments using carrots, fish and chicken were undertaken at Chiswick Foods and South Bank University on a number of custom made conveyors with stainless steel Flat Flex^® 7.26 x 1.57 and generic Polypropylene modular belt.

The results showed bacterial counts for carrots on FF to be on average 1.0 log/cm² lower than PMB, and the effectiveness of cleaners to be on average 0.5 log lower than on FF.  Furthermore acetic acid (AA) was shown to be as effective as a standard cleaning agent Multikleen^® (MK) containing Caustic Soda at concentrations of 2%.  Chlorofoam^® (CF), a standard foaming agent with sodium hypochlorite and potassium hydroxide, was shown to be marginally more effective than AA and MK at removing bacteria initially at concentrations of 2%.

For chicken meat the bacterial counts showed FF to be an average 0.6 log lower than PMB, with the effectiveness of AA being 0.3 log lower on PMB and 0.4 log lower for MK on PMB.  For the fish experiments the bacterial counts showed FF to be on average 0.9 log lower than PMB, with the effectiveness of AA being 0.4 log lower on PMB and 0.2 log lower for MK on PMB.

In addition to this observations made during the experiments with carrots showed that FF could usually be cleaned to a satisfactory ATP level (below 100 units) with just one clean, while PMB often required a 2^nd or even 3^rd clean to reach ATP levels acceptable for production to start.  This was due to product getting lodged in the belt making surfaces inaccessible to normal sanitising and rinsing routines.

Similar observations were made for the experiments with meat and fish, with the FF belt easier to clean, while PMB tended to contain trapped debris even after thorough sanitising and rinsing.

Conclusions and Recommendations 

Biofilms as described by Poulsen are one of the biggest problems facing the food processing industry.  A Biofilm consists of both microbes and their extracellular polymeric substances (EPS).  Once established on production surfaces they are almost impossible to remove completely and can harbour pathogens such as Salmonella, Listeria and Staphylococcus as reported by Genigeorgis.

The 0.95 log difference in contamination between FF (7.85±0.07) and PMB (8.80±0.09) belting in the pre/post tests for carrots and the 4 hour sanitiser tests should therefore be of considerable concern, as it leaves a higher level of bacteria on the belt after cleaning, enabling critical levels of bacteria on the production surfaces to be reached quicker.  In addition to this the reduced level in cleanability of PMB belting by 0.5 log for all cleaners increases this margin further, causing a cumulative effect as the critical mass required for biofilm formation is reached quicker, making effective removal more costly in terms of both chemicals and time employed.

The results of the 8-hour tests with carrots are even more definitive.  The initial reduction on PMB from 10.22±0.07 to 8.35±0.03 is almost a 2-log reduction while it is almost a 3-log reduction on FF at 9.41±0.05 to 6.58±0.04.  Even when the difference in starting values of 0.81 log is factored into the equation the log margin in higher cleanability of FF over PMB at time 0 is 0.96 and increases from 1.59 log at time 240 to 1.98 at time 480.  This clearly shows that PMB not only reaches a higher level of contamination pre sanitation but also is harder to clean and re-contaminates quicker.

From these experiments it emerges that AA, MK and CF do not have significant differences directly post cleaning on FF, all having reached about a 4 log reduction.  On PMB they reach on average a 3.5 log reduction.  The main differences appear after an initial lag phase.

On PMB the log values begin to rise rapidly for CF from time 0 onwards and for AA and MK after an initial stagnant phase.  In contrast CF rises slowly on FF while AA and MK fall by almost a log before rising.  It would seem that the efficacy of the sanitisers is more effective and longer lasting on FF than on PMB.

The results of the meat and fish experiments follow a similar trend.  On average the difference in cleanliness between FF belting and PMB pre cleaning is 0.6 log for chicken and between 0.6 to 1.2 log for fish in favour of FF belting.  The difference increases to 1 log for belts contaminated with chicken and cleaned with MK and AA post sanitation, and by 0.9 to 1.5 log for fish respectively.

Over the 5-hour period the difference between the two belts widens to between 0.7 to 1.1 log for chicken with MK and AA and around 1.5 log for belts exposed to fish respectively.  This indicates that FF not only picks up less bacteria and allows growth but is also easier to sanitise and maintains its lower level of contamination over time.

Another factor could be that the gaps in the PMB belt, which cannot be as readily cleaned as FF, harbour bacteria and so lead to a quicker recontamination of the belt.  It is also possible that the 10 minutes of contact time were insufficient for the foaming agent to react sufficiently with the biofilm.  Particularly the drive shafts and underside of the PMB belt are difficult to clean compared to FF belting.  (See pictures of two belt surfaces) 

Previous work done in this area by Mafu et al, LeChevallier et al (a, b), Frank and Kofi, Lee and Frank and Sommer et al all established that attached cells were significantly more resistant to sanitisers that unattached cells.  In some cases resistance was 10 – 3000 times higher for attached cells than for planctonic cells, dependent upon species, chemicals and surface material used. 

Mosteller and Bishop state that an effective sanitiser should reduce sessile bacterial counts by ≥ 3.0 log and planctonic bacterial counts by ≥ 5.0 log.  This reflects on the reported fact that bacteria, which have established themselves as a biofilm attached to a surface, are considerably more difficult to remove that those that have not established themselves.

In relation to Mosteller et al’s test all 3 sanitisers used in the carrot experiment passed the criteria for reduction of sessile bacteria on FF at 3.87 (AA), 3.76 (MK) and 4.10 (CF) for recommended dosage levels.  While for PMB the log reductions in the pre/post test also all passed this requirement with 3.47 (AA), 3.29 (MK) and 3.56 (CF) there was little margin for error in terms of one of the four key elements of cleaning (temperature, concentration, time and kinetic energy) being compromised or belt damage due to ageing and wear.

For the experiments with meat and chicken the levels of reduction reached for MK and AA on FF and PMB were all less than the accepted levels according to Mosteller et al.  They were 1.42 log for AA and chicken on PMB and 1.87 log for FF.  For MK the values for chicken were 0.97 log for PMB and 1.30 for FF. With fish and AA the reduction was 1.48 log on PMB and 1.70 log on FF.  For MK the values for fish were 1.1 log on PMB and 1.49 log on FF.

While none of the sanitisers used on meat or fish reached acceptable levels of reduction it can be noted that on average FF belt surfaces were reduced by significantly higher levels than PMB belts.  A consideration for further experiments would be the use of sanitisers more suited to the removal of protein, fat and blood.

These findings further validate work done by Frank and Chmielewski, Holah et al and Leclerqe-Perlat et al on surface resistance to damage having a significant impact on the cleanability of food surfaces.  It also reaffirms work done by Krysinski et al, Mafu et al and Boulange-Petermann on the superior hygiene properties of FF over PP, PE, PU, PMB and PVC as well as a number of other materials.

Stainless steel Flat Flex^® superior qualities in terms of resistance to damage, ease of cleaning and efficacy of cleaners used make it the material of choice over materials such a Polypropylene, Polyethylene, Polyacetal, Polyurethane, Polyvinylchloride and Mineral Resins. 

This is a précis report and a full copy of the report or further assistance on this subject is available from Mark Vickery at Wire Belt Company Limited

No part of this document, be it text, picture or graphic, may be published, copied or changed in any way without prior written consent by the author.
Mark L G Vickery 2001