Wal.Sci.Tech. Vol. 32, No.8, pp. 257-261, 1995.Copyright e 1996 IAWQ
Printed in Great Britain. All rights reserved.0273-1223195 $9'50 + 0'00
LABORATORY BIOFILM REACTORS ANDON-LINE MONITORING: REPORT OF THEDISCUSSION SESSION
Aukje Gjaltema* and Thomas Griebe**
*KluyverLaboratory 0/Biotechnology. DelftUniversity ofTechnology,Julianalaan 67, 2628BC Delft. TheNetherlands**Institutfilr Siedtungswasserbau; Wassergute- undAb/allwirtschaft,Universiuu Stuttgart. Bandtale I, 70569 Stuttgart, Germany
At the dawn of this century. the secret life of attached microorganisms was discovered. Microbiologistsstarted realising that microorganisms not only flourished in suspensions but could also live attached tosurfaces. By now we know that the surface of a submerged or wetted substratum is quite a common habitatfor microorganisms. Biofilms can be found wherever we look for them, whether we like it or not, and theyare extensively studied. For this purpose. many biofilm reactors and monitoring systems have been designed.Their variety mirrors the variety of research questions, research aims and research groups involved inbiofilm studies. This has led to a need for information on and evaluation of biofilm systems.
A first division of laboratory biofilm reactors can be made on basis of the general research aim. On the onehand you have the reactors. which are used for optimizing technical processes. They are usually stratified,difficult to sample representatively. and with complicated hydrodynamics. Examples are the trickling filter,packed bed. moving bed. fluidized bed, airlift reactor and hybrid reactors. The second group consists of thedevices that have been developed for biofilm study. They should offer more defmed conditions and bettersample opportunities. Examples of this group are the flow cell, pipe reactor. rotating tube, Robbins device.membrane reactors, annular reactor and static devices.
To evaluate biofilm devices (and biofilm reactors), specific and quantifiable criteria need to be formulated.But different aims lead to different demands. Consequently, it is not possible to give general criteria. Thediscussion therefore concentrated on aspects that are important in choosing a biofilm device when theresearch question has been defined. An eight-point checklist was proposed and used to evaluate a number ofbiofilm devices (Table I):
Hydrodynamic characteristics like flow velocity, flow regime (laminar/turbulent) and shear stress areimportant factors influencing mass transfer, detachment rate and structure of the biofilm. Two- and threephase flows provide additional features like interactions with air bubbles and collisions of suspendedparticles (e.g. fluidized beds, airlift reactors).
Table 1. An example of biofilmdevices and their characteristics
Robbins' device rotating tube membrane reactor rotating annular airlift reactorreactor
flow size dependent low, ill-defined size dependent up to 3 m/s, 3-phase flow,turbulent turbulent
Mixing plug flow: limited not really plug flow ideal, slow good(recirculation) ?-
8Nutrient supply 1 point 1 point separated 1 point no problem ~Gradients yes yes designed vertical (mixing) no, biofilms mixed >
~Sampling not convenient destructive no / destructive convenient convenient, "":-i
not disturbing o~
Monitoring optical, pressure weight projecting optical, torque no (samples can be tilltrlreturned)
Septic operation no no yes yes yes
Substratum variable, studs variable membrane variable, slides variable, size anddensity limits
Remarks whole device rotates collisions!
Laboratory biofilmreactorsand on-linemonitoring 259
The mixing is closely linked with the type of biofilrn device and operational mode: batch or continuous. plugflow or stirred vessel. A batch reactor might be well mixed. but conditions vary strongly over time. A plugflow reactor has limited mixing. and concentration gradients in the direction of the flow are inherent to thesystem. Stirred vessels can be ideally mixed, but more important than checking whether a system is ideallymixed or not is to determine and compare the characteristic time constant for mixing. mass transfer andsubstrate uptake.
3. Nutrient supply.
How and where are the nutrients added? Is aeration possible?
Position-dependent unequal distributions of biofilm in a reactor can be a result of the combination of reactorhydrodynamics, mixing and nutrient supply, either designed for or unwanted. For example, in membranebioreactors, adverse gradients can be created by applying countercurrent separated nutrient flows (designed).In a rotating annular reactor with internal recycle, the total flow can be ideally mixed, but still gradients inbiofilm thickness will occur (and have been observed) if the mixing is too slow compared to the substrateconversion. Gradients can also be a result of local low flow areas (e.g, liquid sample ports) and wallirregularities (e.g, around biofilm sample studs). It is important always to check for gradients, to establishthe extent thereof. and to consider the consequences: They might turn reactor averages into meaninglessvalues. In some cases they can be avoided by small changes in the reactor design or fmishing.
5. Biofilm sampling.
Biofilrn sampling for off-line analysis should be convenient and representative. In case of fixed biofilmsupports. a sufficient number of sampling possibilities (studs. slides, etc.) at different positions in the reactorshould allow for checking reactor gradients and following biofilm development over longer periods.Sampling.should not disturb the biofilm development in the system, e.g. by removing too much biofilm areaor by changing the local hydrodynamics.
6. Biofilm monitoring.
Possibilities for on-line biofilm monitoring are closely linked with reactor construction and the materialused. Often the reactor design will have to be adapted to allow for a certain monitoring method to beapplied. In general this will be easier in systems with removable wall sections (e.g. Robbins device). Ingradient systems an important question is where to monitor. Multiple monitoring at a combination ofdifferent well chosen positions is preferable here. In systems with a freely moving substratum. on-linebiofilm monitoring is not possible. This is partly compensated for by the possibility to return samples afternon-destructive off-line analysis.
7. Septic operation.
To allow for septic operation. apart from sealing and sterilisation of the reactor. sampling is probably themost critical issue. It should be rapid and convenient. and no uncontrolled flows in or out the reactor shouldbe possible.
With respect to the substratum several aspects may need to be considered. Can the substratum be easilyvaried. in different experiments or within one experiment? Can the biofilm be easily separated from thesubstratum for biochemical analysis? For suspended substrata there are limits to the size and density. For on-
260 A. GJALTEMAand T. GRIEBE
line monitoring, certain combinations of substratum and monitoring will not be possible. When differentsubstrata are applied in one reactor. or when the substratum is of a different material than the reactor. thepossibility of interactions with the biofilm development on the other surfaces needs to be checked. In allcases. all substrata need to be accurately described: type of material, surface treatments, surface roughnessand physico- chemical characteristics.
Numerous methods for qualitative and quantitative analysis are employed in the study of biofilms. Many ofthem require invasive sampling and removal of biofilm material from a surface for determination of biofilmthickness, total biomass, microbial composition or activity. In practise, biofilm sampling from a reactorsurface proves to be difficult because it requires easy access to the interior of the reactor without disturbingthe remaining biofilm and reactor operation. Furthermore a representative amount of biofilm has to beremoved which is often limited by the size of laboratory biofilm reactors. Therefore, non-destructive andnon-invasive detection techniques for monitoring of biofilm accumulation and activity are of interest. Toenable further research into the complex three- dimensional structure of biofilms and its dynamics. they alsoneed to have a high spatial resolution and be sufficiently fast.
Various on-line monitoring systems for non-destructive measurements of biofilm properties or systemsproperties, influenced by biofilm formation, are listed in Table 2. The use of more sophisticated on-linemonitoring systems is still limited, partly because most of these methods are still being developed andtested. Reactors and diagnostic tools are developed separately, and are therefore often incompatible. Inaddition. development often focuses on one aspect of biofilm kinetics or development As a result it isdifficult to combine different on-line monitoring methods to determine relevant parameters simultaneously.
Table 2. On-line monitoring methods
Projection, Microscopy, Fiber optics, Micro-electrodes, Pressure drop
Impedance, Heat Resistance, Micro Waves, Surface AcousticWaves, Crystal Balance
Fiber optics, FTIR-Spectroscopy
Bioluminescence, NMR Spectroscopy
NMR imaging, Laser Doppler Anemometry
Pressure drop, Torque
Micro Waves, Surface Acoustic Waves, Crystal Balance
In general, biofilm research can be characterized by a lot of different aims, different demands and a greatlack of standardisation. Therefore it is extremely important to accurately and extensively specify andquantify devices. experimental procedures and conditions in all publications. This also holds for observed
Laboratory biofilmreactorsand on-linemonitoring 261
biofilm structures and measured values. In this way the combined publications will provide a database whichwill help to further elucidate the mechanisms determining biofilm structure and development
There is a need for information exchange. especially on the use of monitoring systems. for training andnetworking. A list of people and institutions using advanced monitoring systems would be very helpful.
A number of sophisticated monitoring methods are in development They are not yet general applicable. norare they compatible with most existing biofilm devices or other monitoring methods. This definitely needsmore attention.