See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/289293029

Biofilms in pharmaceutical waters

Article  in  Pharmaceutical Engineering · November 1997

CITATIONS READS
14 1,793

1 author:

Frank Riedewald
University College Cork
28 PUBLICATIONS   105 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Recycling of Endof-Life Products (PCB, ASR, LCD) View project

All content following this page was uploaded by Frank Riedewald on 29 September 2019.

The user has requested enhancement of the downloaded file.

NOVEMBER/DECEMBER 1997 PHARMACEUTICAL ENGINEERING 1

Reprinted from The Official Journal of ISPE

PHARMACEUTICAL ENGINEERING November/December 1997, Vol. 17 No. 6

BIOFILMS IN PHARMACEUTICAL WATERS

BY FRANK RIEDEWALD

THIS ARTICLE DISCUSSES CURRENT LITERATURE ON BIOFILM RESEARCH WITH RESPECT TO

PURIFIED WATER SYSTEMS AND GIVES STRATEGIES FOR THE PREVENTION AND REMOVAL OF
BIOFILMS IN THE PHARMACEUTICAL INDUSTRY.

Introduction different.8 Costerton et al.8 states, “We first noted that biofilm

B
iofilms are found in most natural and industrial cells are at least 500 times more resistant to antibacterial
aquatic systems.7,32 Single motile bacterial cells agents. Now we have discovered that adhesion triggers the
(known as planktonic cells) have been shown to expression of a σ factor that derepresses a large number of
prefer to attach themselves to and reproduce on any genes so that biofilm cells are clearly phenotypically distinct
available surface rather than remain in a planktonic state.7,32 from their planktonic counterparts.”
Attachment of bacteria to a surface is an important feature of
micro-organisms.7,24 In their attached form, biofilm forming Growth of a Biofilm
micro-organisms possess resistance properties far surpass- In the pharmaceutical industry, the successful manufacture
ing those of their planktonic counterparts.7,23 and distribution of purified and distilled water is of para-
In the pharmaceutical industry, the most deleterious effect mount importance. Bacterial growth within purified water
of a biofilm is the bacterial contamination of purified water systems presents potentially serious problems due to bio-
systems and possible contamination of product.11 The detec- deterioration of the purified water and product. Water puri-
tion, removal and prevention of a biofilm are therefore impor- fication or distribution systems are inherent sites for biofilm
tant considerations in the production of pharmaceutical wa- activity because the four conditions necessary for their growth,
ter. Production may be impeded if the purified water genera- namely water, nutrients, a support surface and the presence
tion or distribution systems are contaminated by biofilm of planktonic bacterial cells are generally present.
activity.
Beneficial uses of biofilms include their use in trickling Water
filters in waste water treatment plants and in the metal Since bacteria are the oldest and most successful life-form on
leaching processes of biohydrometallurgy. earth, it should not be surprising to find bacteria in almost all
forms and concentrations of water. Biofilms have been de-
What is a Biofilm? scribed in varied and surprising locations from potable and
A biofilm is defined as a bacterial population composed of cells domestic water distribution systems,5 to within copper pipes
which are firmly attached as micro-colonies to a solid sur- of a disinfection distribution system.13,14 In the pharmaceuti-
face.7,8 The bacterial cell colony secretes a substance, prima- cal industry, biofilms have been found in purified and distilled
rily polysaccharide, which has a slimy, tacky consistency.5 water distribution systems.11,14,25
This slime encourages the attachment of other organisms,5 is
a trapping device for nutrients and also protects the biofilm Nutrients
from the action of biocides e.g., chlorine.7,22 In general, bacteria growth is limited by the available nutri-
Biofilm bacteria are not just simply planktonic cells at- ents.8 Microbiologists usually grow and maintain bacteria on
tached to a surface. They are demonstrably and profoundly laboratory media containing high concentrations of organic
carbon i.e. > 2000 mg per liter.32 In nature; however, olig-
otrophic conditions i.e. starving conditions of between 1-15
mg carbon per liter are more common than one might think.32
In the deep waters of the oceans, dissolved organic matter
concentrations can be as low as 0.5 to 0.8 mg/liter of which
most is not bio-available.26 Some bacterial strains such as
Escherichia coli, Staphylococcus citreus, Bacillus megaterium
and Proteus vulgasis have been reported to grow and multiply
in oligothropic waters containing less than 0.4 mg organic
carbon per liter.17 However, organic carbon concentrations of
0.5 mg/liter are often detected in distilled water.
Favero10 investigated the growth of bacteria (Pseudomonas
aeruginosa) in distilled water, “When the organism is found in
distilled water, it is often assumed that it is simply surviving
or else growing very slowly.” Instead, he found that Pseudomo-
Figure 1. The three phases of biofilm growth.3,7 nas aeruginosa can grow at a rapid rate in distilled water. He

©Copyright ISPE 1997

2 PHARMACEUTICAL ENGINEERING NOVEMBER/DECEMBER 1997

ment since new planktonic cells can simply settle on the already
formed biofilm.

The Stationary Phase or Plateau Phase

At this stage the biofilm is fully established. The extent of
biofilm is determined by the nutrient transport to and the
shear forces inflicted on the biofilm. Randomly cell clusters
may be sheared off and contaminate the bulk water-phase.

Survival Strategies of Bacteria in

Low Nutrient Environments
Purified waters are not ideal environments for microbial
growth. The oligothropic conditions present force bacteria to
adopt a variety of strategies in order to survive. These include
the following:

1. reducing their size

Figure 2. Survival rates of planktonic and biofilms cells in 1 ppm
chlorine.14 2. increasing their cell numbers

reported the growth of this bacteria in distilled water for 3. increasing their swimming speed
hospital use from 4,300 to 1,100,00 cells per milliliters at a
temperature of 25 °C within a 24-hour period. 4. slowing their metabolic rate

Surface 5. increasing their adhesion to surfaces

All surfaces used in water distribution are readily contami-
nated by a biofilm.5,31 Every 100m length of a 80mm diameter 1. Reducing Their Size
pipe has an internal surface area of 25m2. Such a large surface By becoming smaller, a cell decreases its food requirements.19
to volume ratio offers micro-organisms ample opportunity to Mueller27 found that Pseudomonas aeruginosa cells are able
attach to a surface. to decrease their average cell size from a normal average size
Rogers et al.31 investigated the microbial colonization of of 0.91 ±0.14 µm2 to 0.58 ±0.02 µm2 after 24 hours of starva-
different materials. They found that all materials were rap- tion. This represents a cell size reduction of almost 50%.
idly colonized by micro-organisms following their insertion The ability of bacteria to reduce their size has implications
into a biofilm model system containing a population of in the pharmaceutical industry. Flemming14 suggests that
Legionella pneumophila and other micro-organisms. Stain- some small, starved bacteria may be able to pass through a
less steel, polypropylene and PVC showed the lowest concen- 0.45 µm filter. Koch21 argues that even a 0.2 µm polycarbonate
tration of flora - Table A. Rogers et al.31 attribute the extensive filter may allow small, starved bacteria to pass through.
biofilms found on elastomeric surfaces to the leaching of
nutrients from these materials. This work agrees with the use 2. Increasing Their Cell Numbers
of stainless steel and polypropylene materials in purified Some bacteria respond to starvation by fragmenting. Frag-
water systems. mentation is defined as cell division without growth that leads
to an increase in numbers and a decrease in average cell size.20
Planktonic Micro-organisms Cell fragmentation can be a very fast and efficient method of
In a purified water distribution system, one of the main lowering nutrient concentrations within a cell, thus prevent-
sources of bacterial contamination is the raw water.12 Plank- ing nutrient loss by osmosis.21 Kjelleberg20 describes the
tonic bacteria entering the purified water system will move to bacterium Pseudomonas sp. S9 which has been shown to
and attach themselves onto a surface thereby starting the complete its fragmentation phase within one hour.
formation of a biofilm.
3. Slowing Their Metabolic Rate
Kinetics of Biofilm Growth in Pharmaceutical Waters In nature, bacterial growth is usually a slow process mainly
The growth characteristics of a biofilm on a solid, inert surface due to the lack of available nutrients. Morita26 points out,
can be divided into three distinct phases.3,7 “Where does one draw the line between growth and no growth,
especially when the doubling time of organisms in nature can
The Lag Phase be measured in terms of months and maybe years? ... the
Invading planktonic micro-organisms, transported by diffu- generation time for organisms in the deep sea was 210 days.
sion (which may be enhanced by turbulent transport) attach If you are dealing with laboratory culturing of organisms, this
themselves to an available surface. The number of planktonic situation would be termed ‘no growth’.” In general, bacteria
cells present and the surface roughness are important factors subjected to starving conditions require some time to grow
influencing cell attachment. Micro-organism numbers do not again if placed into a nutrient sufficient environment.26 The
increase by any significant amount during this period. time before growth occurs again is proportional to the time a
cell was subjected to starving conditions.26
The Exponential Phase
After some time, attached cells will begin to divide and logarith- 4. Increasing Their Adhesion to Surfaces
mic growth will occur. The biofilm is now formed. Planktonic cell With respect to the development of biofilms, the increased
numbers or surface roughness no longer influence cell attach- adhesion of starved cells to surfaces as a means of survival is

©Copyright ISPE 1997

NOVEMBER/DECEMBER 1997 PHARMACEUTICAL ENGINEERING 3

of most interest.7,8,24 Small, starved bacteria are the primary sanitation strategy.13,14 A dead biofilm remaining in a system
colonizers of surfaces.24 Mueller27 showed that Pseudomonas provides an easy surface for planktonic micro-organisms to
fluorescens cells increased their rate of adsorption onto sur- attach to and feed on.13,14 If a dead biofilm remains in a system,
faces by a factor of four when forced to starve in low nutrient a new biofilm may form faster than the old one.29,37
waters. Marshall24 observed the regrowth of starved bacteria
to their normal size while attached to a surface in low nutrient Removal of a Biofilm by the Shear Forces
waters. The bacteria multiplied every 60 minutes in nutrient- of a Flowing Fluid
poor waters, but no growth was observed for the same type of Jennings18 investigated the role of turbulence, time, tempera-
bacteria while still suspended in the water phase as plank- ture and cleaning solution strength on the removal of soil
tonic cells. deposits from pipelines in the dairy industry. He suggests a
The formation of a biofilm offers the following advantages minimum Reynolds number of 25,000 as, “There was little
for bacteria to grow.7,8 cleaning action at any lower Reynolds number.”
The dimensionless Reynolds number is defined9 as the
1. Fatty acids, alcohols and other organic molecules consum- quotient of the velocity w of the fluid, its density 'ρ' and the
able by micro-organisms tend to accumulate on any avail- pipe diameter 'd' divided by the dynamic viscosity 'η' of the
able interface or surface. It is energetically more favorable fluid:
for such organic molecules to be at an interface, than to be
in solution.1 wρd
Re = ________
2. Attached cells which have died can be used as nutrients by η
neighboring cells.7
Dimensionless numbers such as the Reynolds number can be
3. The exopolysaccharide matrix formed at the biofilm-water misleading. Pipe diameter influences the Reynolds number,
interface acts as a nutrient trapping device.7 so a high Reynolds number in one occasion can mean a high
fluid velocity, and in another a low fluid velocity (assuming
Some bacteria can colonize a surface by moving along it and constant density and viscosity). The fluid velocity in a pipe is
feeding off accumulated nutrients. Marshall24 observed the directly related to the shear stress which a biofilm experi-
migration of a bacterium along a surface in oligotrophic ences. It is the shear stress induced by the fluid onto a biofilm
waters. “The small, starved bacteria obviously scavenged the that is of interest; therefore, in the discussion of biofim
surface-bound fatty acid, grew to normal size, began to divide, removal, fluid velocity is a more meaningful parameter to use
and surprisingly, the newly divided cells began to slowly than the Reynolds number.
migrate from each other across the membrane surface. The Powell30 conducted experiments on the removal of bacterial
migration rate was very variable and extremely slow (0.04- cells from glass surfaces using fluid shear. He used various
0.19 µm/min, i.e., from about 5 to 25 min to move 1 µm, the water flow rates through rectangular capillaries containing
length of a cell)”.24 biofilms. Cell detachment was measured visually via micro-
scopic examination of the glass surface while a certain shear
Biofilm Detection stress was employed. Critical sheer stresses for various
The presence of a biofilm in a purified water system can be bacteria studied are summarized in Table B.
difficult to detect. Water quality is normally monitored by The results indicate that the type of bacteria has a high
sampling the bulk phase of the water. Such samples do not degree of influence on the critical shear stress 'τc'. The critical
reflect the location or extent of a biofilm.13,14 No correlation shear stress 'τc' is defined as the stress necessary to remove all
exists between the bacterial count of a bulk phase sample and but 2% of the cells.30 Power30 speculates that the critical shear
the biofilm thickness.13,14 A low bacterial count of a bulk phase stresses 'τc' for glass may be equally applicable to other
sample could in fact mean extensive biofilm activity, since materials since detachment of bacteria by fluid shear is not
sheared off bacteria from a biofilm may be diluted by the great influenced by surface characteristics.
volume of the water.14 Alternatively, particles of biofilm may In turbulent flow situations, common in technical systems,
randomly contaminate the bulk water and this might result the following equation allows the calculation of the average
in wildly varying bacterial counts, a bacterial count of 1000 velocity ' w ' from the shear stress at the pipe wall 'τw'.33 The
CFU/ml in one bulk water sample and 10 CFU/ml in another
sample is thus possible.11
Detection of a biofilm therefore requires sampling from
internal surfaces.13,14 Internal surface sampling can be prob-
lematic as a biofilm can be patchy,3,17 especially in nutrient-
poor conditions such as in water for pharmaceutical use or in
situation where the fluid velocity is relatively high. Areas at
risk for biofilm growth include gaskets, pipe bends and
locations were the fluid velocity is lower, such as short dead
legs, heat exchangers and holding tanks.

Biofilm Removal
Mechanical cleaning techniques, or the shear forces induced
by fluids may be used to detach biofilms from their support
surface. Biocides such as chloride or ozone may be used to kill
biofilm bacteria. Removal of a biofilm from a system is much Figure 3. Effect of different fluid velocities on the removal of an
more important than its disinfection and should be part of a ozone weakened biofilm.14

©Copyright ISPE 1997

4 PHARMACEUTICAL ENGINEERING NOVEMBER/DECEMBER 1997

patients infected with the same organism, “When planktonic

cells within this system were treated with 50 µg ml-1 of
tobramycin, 8 hours of contact was sufficient to yield a
complete kill, but 12 hours contact with 1000 µg ml-1 of
tobramycin did not kill the biofilm cells. When these resistant
biofilms were washed free of tobramycin and dispersed to
yield single cells, these essentially planktonic cells were
susceptible to 50 cells µg ml-1 of the antibiotic.”
Other researchers confirm this observation.14,23,35 Costerton8
showed that biofilm cells are at least 500 times more resistant
to anti-microbial agents than planktonic bacteria of the same
type. LeChevallier23 found that biofilms can be up to 3000
times more resistant to free chlorine than its planktonic
counterparts. The resistance of biofilms to biocides seems to
be primarily due to the protective nature of the exopolysaccharide
Figure 4. Biofilm thickness as a function of fluid velocity adapted matrix which is formed on top of the biofilm.8 Costerton7 points
from Powell.30 out that, “treatment with traditional concentrations of biocides
kills the planktonic organisms and gives the appearance of
term 'f' is the dimensionless friction factor and can be calculated
success, but leaves the biofilm virtually unaffected.”
using the Colebrook-White equation which is used in pressure
As discussed earlier, Powell’s work30 suggests that water
drop calculations.9
velocities of ca. 5 m/s are required to remove a biofilm by fluid
shear stress alone - Table B. Biocides weaken the binding
1
forces of a biofilm to a surface, thereby reducing the critical
τw = __ f ρ 2
shear stress required to remove the biofilm14 - Figure 3.
8
Time also plays an important role in disinfection and
detachment of a biofilm. Detachment of a biofilm can be a slow
Applying this equation to Powell's results, equivalent water
process as Srinivasan35 found, “Detachment and disinfection
velocities and steam velocities for electropolished steel may
are independent processes. It is quite possible to disinfect a
be calculated - Table B.
biofilm without removing it. In our experimental system
Purified water systems are usually designed for water
detachment was found to be a significantly slower process
velocities of about 1.5 m/s to 2.5 m/s, higher velocities will
than disinfection.”
dramatically increase the pressure drop, rendering the system
Biocides are not without their disadvantages. The biocide
uneconomical. (e.g. water flowing with a velocity of 1.5 m/s
itself is a contaminant in purified water and must be removed
through a 100 meter straight length of 50 mm diameter
prior to use. Ozone and chlorine can increase the available
electropolished pipe experiences a pressure drop of 0.7 bar.
nutrient concentration by breaking down non-biodegradable
Increasing the fluid velocity to 5 m/s would increase the
molecules into bio-degradable ones.15
pressure drop to 6 bar.)
Bacteria can become resistant to biocides. Therefore, chemi-
Therefore, Powell’s30 data indicates that an existing biofilm
cal biocides should be changed frequently to avoid immuniza-
in a technical water distribution system will not be removed
tion.37
by fluid shear stress alone.
Removal of a Biofilm by a Freezing Technique
Removal by Mechanical Forces
The freezing method of biofim removal takes advantage of the
Mechanical means can be very effective in the removal of a
high water content of a biofilm and involves slow freezing
biofilm and should be used wherever practical.14,16,29 A smooth
cycles using a coolant liquid to produce large ice crystals
surface will allow mechanical forces to be even more effec-
within the biofilm. Subsequent thawing of the crystals leads to
tive.16 In tanks and other accessible areas, removal of biofilms
complete removal of the biofilm.7
is achieved by brushing or use of high pressure water jets
etc.14,29 In inaccessible areas, such as within distribution
pipes, heat exchangers etc., sponge rubber balls, suspended in
water, can be pumped through to remove the biofilm.37
Although very effective, mechanical cleaning techniques Material Biofilm Planktonic Phase
can not remove the entire biofilm.29 A residual biofilm will (CFU cm-2) (CFU ml-1)
remain in the system, which will result in faster growth of a Total flora Stainless steel 2.13 x 105 2.23 x 105
new biofilm.29 Successive mechanical cleaning can cause the Polypropylene 4.54 x 105 4.48 x 105
build-up of an increasingly resilient biofilm.29 PVCe 5.14 x 105 6.20 x 105
PVCu 6.23 x 105 3.22 x 104
Removal of a Biofilm by Biocides (Disinfectants) Mild steel 1.69 x 106 2.23 x 105
w
Chlorine is one of the most widely used biocides in domestic Polyethylene 2.75 x 106 1.62 x 106
water supply systems.5,14 Ozone is frequently used in purified Ethylene- 1.08 x 107 7.45 x 105
water systems.11 propylene
Planktonic bacteria are relatively susceptible to biocides; Latex 5.50 x 107 1.87 x 106
however, biofilms are not. The polysaccharide layer covering
a biofilm and other cell modifications have been shown to
Table A. Microbial colonization of different materials showing mean
partially protect the biofilm against disinfectants7 - Figure 2
number of microorganisms occurring in the biofilm and the
Costerton7 examined biofilms on catheters recovered from planktonic phases according to Rogers et al.31

©Copyright ISPE 1997

NOVEMBER/DECEMBER 1997 PHARMACEUTICAL ENGINEERING 5

Bacteria Critical Calculated Equivalent Calculated Equivalent

Shear Stress Water Velocity [m/s] Steam Velocity [m/s] to
[N/m2] to to remove bacteria remove bacteria from
remove from an 80mm an 80mm
bacteria electropolished pipe at electropolished pipe at
from glass 20°C 150°C

B. cereus 44.0 5.0 230

B. cereus 28.0 4.0 180
S. thyphimurium 53.0 5.6 254
B. licheniformis 8.4 2.0 93
E. coli K12 50.0 5.4 247
B. stearothermophilus 1.1 0.6 30

Table B. Critical shear stresses 'τ'c required to detach selected bacteria from glass surfaces,30 and the equivalent water and steam velocities
required for electropolished steel, calculated using equation.2

between 40009 - 8000.37 In turbulent flowing waters, mass

Nomenclature transport of nutrients to the biofilm is greatly improved due
to additional influence of turbulence.2,33,34 Bacterial growth in
d = Diameter of pipe [m] turbulent waters should therefore be greater than in laminar or
standing waters - Figure 4.
f = Friction factor, dimensionless Fleming12 investigated the influence of recirculation on
bacterial growth on ion exchangers. He found that bacterial
Re = Reynolds number, dimensionless growth was not less when the water was recirculating than
when it was standing still. According to Fleming,12 “It can be
w = Fluid velocity [m s-1] said that the recirculated system is similar to a fermenter
operating at a very low nutrient level. Without additional
ρ = Fluid density [kg m-3] techniques, e.g. filtration, recirculation does not seem to be a
useful method for preventing bacterial growth during off-
τ = Shear stress [N m-2] periods.” Fleming12 points out that, “The influence of recircu-
lation on bacterial growth has not been investigated: maybe
η = Fluid viscosity [Pa s] its effect has been taken for granted.”

Design Parameters of Purified Water Systems

Limitations of Some Sanitation Techniques McFeters et al.25 compared biofilm growth in two Milli-Q Plus
Sanitation using UV-lights, heating or recirculation of the laboratory water purification systems after just three months
purified water is limited in its ability to curtail bacterial of operation. A UV-Sterilizing lamp was installed in the
growth. recycling loop of one of the Milli-Q Plus systems. Planktonic
The FDA11 recognizes that,“hot (65 - 80°C) systems are self bacteria concentrations and biofilm densities were found to be
sanitizing,” as most bacteria adapted to water environments lower in the unit which contained the UV lamp, surprisingly
can not survive water temperatures very much greater than “some evidence of a biofilm was found on the UV lamp
30°C.32 However, some bacteria (i.e. Pseudomonas aeruginosa) housing.” Both systems; however, delivered water of accept-
although killed by heat at 71°C are still capable of attachment able quality with regard to bacterial numbers, resistivity and
to surfaces,36 and these attached dead cells are then an endotoxin concentration despite the presence of the biofilms.
available source of nutrients for other bacteria. Constant high McFeter’s25 study shows that biofilms can exist in purified
water temperatures might favor the selection of thermophilic water systems, their activity; however, may be small enough
bacteria.14 not to cause excessive bacterial contamination. A sudden
UV light is often used in purified water systems to control problem with bacterial contamination of purified water is not
planktonic bacterial numbers. Some bacteria although killed due to a newly formed biofilm, but to an increased thickness
with UV radiation may attach as readily as living bacteria to and activity of one which has already formed.14 The reason for
solid surfaces.5 Fewer cells attach when killed by heat.5 The this increased activity should be investigated and counter-
higher attachment rate of bacteria killed by UV rather than measures taken.14
by heat is thought to be a feature of the bacterial membrane Biofilms can form in purified water systems after rela-
which remains largely intact when cells are killed by UV tively short periods of operation. Fleming14 notes that the
radiation.5 Dead attached cells may provide nutrients to a initial attachment of planktonic bacteria to surfaces occurs in
biofilm. the first couple of hours of operation. Stanley36 showed that
The FDA11 states that, “obviously water in constant motion some cells begin to attach onto stainless steel surfaces in less
is less liable to have high levels of contaminant.” In standing than one minute.
or in laminar flowing waters, nutrient transport to a biofilm Since biofilms can form after a short period of operation
is dependant upon molecular diffusion. This curtails bacterial and are so difficult to remove, purified water systems should be
growth as molecular diffusion is the slowest mass transport designed and operated with this in mind. Each different com-
process.2,34 Most purified water systems operate in the turbu- ponent therefore must be considered in an overall strategy to
lent flow region well beyond the critical Reynolds number of keep biofouling within an acceptable level.13,14

©Copyright ISPE 1997

6 PHARMACEUTICAL ENGINEERING NOVEMBER/DECEMBER 1997

Fleming suggests that by extending the lag phase of biofilm 4. Frequent Sanitation and Cleaning
development, the growth of a biofilm may be delayed, sanitation Frequent sanitation and cleaning of the water system should
and cleaning intervals can then be extended.14 To achieve this, be performed as a preventative measure.14
the design of a purified water system should consider incorpo-
rating the following suggestions: 5. Checking Effectiveness of Sanitation and Cleaning
The effectiveness of sanitation and cleaning should be checked
1. high water velocities on representative surfaces.14 Bacterial counts of water samples
are only indirectly - if at all - related to biofilm activity. Short
2. biocides pieces of removable pipe may be installed into a very long
distribution system to act as representative surfaces. These
3. smooth surfaces surfaces should be checked before and after any sanitation is
performed, to evaluate its effectiveness.
4. frequent sanitation and cleaning
6. Monitoring
5. checking effectiveness of sanitation and cleaning Monitoring representative surfaces for evidence of biofilm
growth will ensure that early intervention is possible. Prelimi-
6. monitoring nary research indicates that younger biofilms are easier to
remove than older ones.14
7. minimizing nutrients
7. Minimizing Nutrients
1. High Water Velocities Maintaining nutrient concentrations at organic carbon con-
High water velocities hamper the attachment of bacteria to a centrations well below 0.5 mg per liter might be the best
surface, since a discrete time interval is required for cell method of biofilm prevention.13 In extremely nutrient defi-
attachment.36 Biofilm thickness also will be reduced as a cient environments, bacteria will not attach to surfaces.8 If
result of the increased shear forces of the fluid.3,30 bacteria do attach in these very low nutrient waters, they will
The maximum growth rate of a biofilm occurs at fluid not be able to form a biofilm since they lack “the metabolic
velocities of ca. 1m/s varying with the bacterial type. At very capability for exopolysaccaride production to accomplish cel-
low fluid velocities (laminar flow), biofilm thickness is limited lular adhesion and biofilm formation.”8
by nutrient diffusion. At very high velocities (turbulent flow
with velocities >1 m/s), thickness is limited by the shear stress Conclusion
of the flowing fluid2,33,34 - Figure 4. Biofilms consist of bacterial cells which display survival
Bryers3 studied biofilm formation in turbulent flow condi- capabilities far surpassing their planktonic counterparts.
tions, and he observed biofilm growth up to a Reynolds number Biofilms therefore are highly resilient structures. If a biofilm
of 28,800. i.e. approximately 2.2 m/s. Therefore, fluid velocities has been identified as the source of bacterial contamination in
lower than 2.2 m/s should not prevent biofilm formation. As a purified water system, its complete removal is crucial.
many factors may influence biofilm formation, it is difficult to However, the complete removal of a biofilm from an enclosed
pinpoint the critical water velocity required to prevent a biofilm system can be difficult to achieve.
forming. Nevertheless, it can be suggested that the water Prevention of biofilm growth is the ideal control strategy;
velocity within distribution pipelines should be as high as is however, it may be easier to achieve in theory than in practice.
practicable (2m/s or higher). Nevertheless, purified water systems should be designed
with the prevention of biofilms as an important consideration.
2. Biocides Constant vigilance is required in the successful manage-
Biocides may be used to kill planktonic micro-organisms. ment of purified water systems. Biofilm control strategies
However, the thickness of a biofilm can not be reduced by should include the use of high water velocities, biocides to kill
simply reducing the number of planktonic cells in the bulk planktonic organisms and to weaken a biofilm. Smooth sur-
water.13 Biocides may be employed to weaken a biofilm faces should be specified and available nutrients must be kept
structure, and fluid shear forces will then be more effective in to a minimum. Frequent sanitation and cleaning routines
the prevention and removal of the biofilm. Flemming14 suggests should be incorporated and their effectiveness checked. Rep-
using an ozone concentration of between 0.5 to 1.0 ppm to control resentative surfaces should be monitored for early detection of
biofouling. biofilm growth.

3. Smooth Surfaces References

Electropolished steel is commonly used in purified water 1. Adamson, A.W., (1976), Physical Chemistry of Surfaces,
systems. Decreasing surface roughness increases the prob- 3.ed., John Wiley & Sons, New York.
ability of spontaneous desorbtion of bacteria.28 However, 2. Bird, Steward, Lightfoot, (1966), Transport Phenomena,
surface roughness is not a significant factor in the total John Wiley & Sons, New York.
amount of biofilm formed,5 and electropolishing alone can not 3. Bryers, J., W. Charkalis, (1981), Early Fouling Biofilm
prevent or even reduce the thickness of a biofilm.13 However, Formation In A Turbulent Flow System: Overall Kinetics,
cleaning is easier.14 It is important to use sanitary fittings as Water Research, Vol. 15, pp. 483-491.
they are of a cleaning-friendly design.14 Non-sanitary fittings 4. Carson, L.A., M.S. Favero, W.W. Bond, N.J. Petersen, (1973),
have cavities which may provide a hideaway for micro- Morphological, Biochemical, and Growth Characteristics of
organisms. It is nearly impossible to remove micro-organisms Pseudomonas Cepacia from Distilled Water, Applied and
from such places.16 Environmental Microbiology, Vol. 35, pp.476-483.
5. Charkalis, W.G., (1973), Attached Microbial Growths - I.

©Copyright ISPE 1997

NOVEMBER/DECEMBER 1997 PHARMACEUTICAL ENGINEERING 7

Attachment And Growth, Water Research, Vol. 7, pp. 25. McFeters, G.A., S.C. Broadaway, B.H. Pyle, Y. Egozy,
1113-1127. (1993), Distribution of Bacteria within Operating
6. Charkalis, W.G., (1973), Attached Microbial Growths - II. Laboratory Water Purification Systems, Applied and
Frictional Resistance Due to Microbial Slimes, Water Environmental Microbiology, Vol. 59, pp. 1410-
Research, Vol.7, pp. 1249-1258. 1415.
7. Costerton, J.W, K.J. Cheng, G.G. Geesey, T.I. Ladd, J.C. 26. Morita, R.Y., (1988), Bioavailability of Energy and its
Nickel, M. Dasgupta, T.J. Marrie,(1987), Bacterial Biofilms Relationship to Growth and Starvation Survival in Na-
in Nature And Disease, Annual Review of Microbiology, ture, Canadian Journal of Microbiology, Vol. 34, pp.
Vol. 41, pp. 435-464. 436-441.
8. Costerton, J.W., Z. Lewandowski, D.E. Caldwell, D.E. 27. Mueller, R.F., (1996), Bacterial Transport and
Korber, H.M. Lappin-Scott, (1995), Microbial Biofilms, An- Colonisation In Low Nutrient Environments, Water
nual Review of Microbiology, Vol. 49, pp. 711-745. Research, Vol. 30, pp. 2681-2690.
9. Douglas, J.F., J.M. Gasiorek, J.A. Swaffield, (1986), Fluid 28. Mueller, R.F., W.G. Charkalis, W.L. Jones, J.T. Sears,
Mechanics, 2.ed., New York. (1991), Characterisation of Initial Events in Bacterial
10. Favero, M.S., L.A. Carson, W.W. Bond, N.J. Petersen, Surface Colonization by Two Pseudomonas Species Using
(1971), Pseudomonas aeruginosa; Growth in Distilled Water Image Analysis, Biotechnology and Bioengineering,
from Hospitals, Science , Vol. 173, pp. 836-838. Vol. 39, pp. 1161-1170.
11. FDA, (July 1993), Guide to Inspection of High Purity Water 29. Nickels, J.S., R.J. Bobbie, D.F. Lott, R.F. Martz, P.H.
Systems Benson, D.C. White, (1981), Effect of Manual Brush
12. Flemming, H.C., (1987), Microbial Growth on Ion Exchang- Cleaning on Biomass and Community Structure of
ers, Water Research, Vol. 21, No. 7, pp. 745-756. Microfouling Film Formed on Aluminium and Titanium
13. Flemming, H.C., (1992), Biofouling bei Mem- Surfaces Exposed to Rapid Flowing Seawater, Applied
branprozessen, Pharma-Technologie Journal, CON- and Environmental Microbiology, Vol. 41, pp. 1442-
CEPT Heidelberg, Symposion Frankfurt, 1992 1453.
14. Flemming, H.C., G. Schaule, (1994), Bekämpfung von 30. Powell, M.S., N.K. Slater, (1982), Removal Rates of
Biofouling in wäßrigen Systemen, Werkstoffe und Bacterial Cells from Glass Surfaces by Fluid Shear,
Korrosion, Vol. 45, pp. 40-53. Biotechnology and Bioengineering, Vol. 24, pp. 2527-
15. Gilbert, E., (1987), Biodegradability of Ozonation Products 2537.
as a Function of COD and DOC Elimination by the Example 31. Rogers, J., A.B. Dowsett, P.J. Dennis, J.V. Lee, C.W.
of Humic Acids, Water Research, Vol. 22, pp. 123-126. Keevil, (1994), Influence of Plumbing Materials on Biofilm
16. Holah, J.T., R.H. Thorpe, (1990), Cleanability in Relation Formation and Growth of Legionella pneumophila in
to Bacterial Retention on Unused and Abraded Domestic Potable Water Systems, Applied and Environmental
Sink Materials, Journal of Applied Bacteriology, Vol. Microbiology, Vol. 60, pp. 1842-1850.
69, pp. 599-608. 32. Roznak, D.B., R.R. Colwell, (1987), Survival Strategies
17. Jain, D.K., (1995), Microbial Colonisation of the Surface of of Bacteria in the Natural Environment, Microbiologi-
Stainless Steel Coupons in a Deionized Water System, cal Review, Vol. 51, pp. 365-379.
Water Research, Vol. 29, pp. 1869-1876. 33. Schlichting, H., (1960), Boundary Layer Theory, 4 ed.,
18. Jennings, W.G, A.A. McKillop, J.R. Luick, (1957), Circula- McGraw Hill, New York.
tion Cleaning, Journal of Dairy Science, Vol. 40, pp. 1471- 34. Sherwood, T.K., R.L. Pigford, C.R. Wilke, (1975), Mass
1479. Transfer, McGraw Hill, New York
19. Kjelleberg, S. M., B.A. Humphrey, K.C. Marshall, (1982), 35. Srinivasan, R., P.S. Stewart, T. Griebe, C. Chen, X. Xu,
Effect of Interfaces on Small, Starved Bacteria, Applied (1995), Biofilm Parameters Influencing Biocide Effi-
and Environmental Microbiology, Vol. 43, pp. 1166- cacy, Biotechnology and Bioengineering, Vol. 46, pp.
1172. 553-560.
20. Kjelleberg, S.M., Hermansson, (1984), Starvation-Induced 36. Stanley, M.S., (1983), Factors Affecting the Irreversible
Effects on Bacterial Surface Characteristics, Applied and Attachment of Pseudomonas Aeruginosa to Stainless
Environmental Microbiology, Vol. 48, pp. 497-503. Steel, Canadian Journal of Microbiology, Vol. 29, pp.
21. Koch, A.L., (1996), What Size Should A Bacterium Be? A 1493-1499.
Question of Scale, Annual Review of Microbiology, Vol. 37. VDI Heat Atlas, (1993), Verein Deutscher Ingenieure, 6.
50, pp. 317-348. ed., Düsseldorf, VDI-Verlag.
22. Ladd, T.I., J.W. Costerton (1990), Methods for Studying
Biofilm Bacteria, Methods in Microbiology, Vol. 22. About the Author
23. LeChevallier, M.W., C.D. Cheryl, R.G. Lee, (1988), Factors Frank Riedewald is Process Engineering Manager with
Promoting Survival of Bacteria in Chlorinated Water Sup- Process & Industrial Design Consultants Ltd. Ireland.
plies, Applied and Environmental Microbiology, Vol. He holds a BE in mechanical engineering from the
54, pp. 649-654. Technische Universität Clausthal and a ME in chemical
24. Marshall,K.C., (1988), Adhesion and Growth of Bacte- engineering from the Technische Universität Berlin. He
ria at Surfaces in Oligotrophic Habitats, Canadian is a chartered engineer and a member of the IChemE and
Journal of Microbiology, Vol. 34, pp. 503-505. the AIChE.
He can be contacted at Tel: 353-21-613521.

©Copyright ISPE 1997

View publication stats