Identification and control of filamentous micro-organisms in industrial wastewater treatment plants

The activated sludge process is widely used for treating domestic as well as industrial wastewater. Excellent treatment results can be obtained in properly designed and operated plants as long as the activated sludge flocs settle well in the final clarifier.

The settling properties depend upon the composition of the flocs. Well settling flocs are mainly composed of aggregated unicellular bacteria (Figure 1). A few filamentous species might be present, but a small filamentous population hardly affects the settling velocity of the more or less rounded flocs. However, when growth of filamentous species is favoured by the process conditions applied, they may start to predominate the mixed sludge population (Figure 2). Subsequently, the floc settling properties deteriorate and a problem known as bulking of activated sludge arises. Serious bulking will result in loss of solids with the final effluent. In addition, some filamentous species may also contribute to foaming of activated sludge, which frequently ends in a massive scum layer on the aeration tank and/or final clarifier. Thus, process stability and final effluent quality largely depends upon the biomass composition in the plant.

Dynamic nature requires microscopic process control

In a biological treatment plant, a large diversity of micro-organisms are fed with wastewater containing numerous (in)organic compounds. The nutrients entering the plant are used by the micro-organisms for their energy requirements and their growth. The amount of food supplied is not enough to allow unlimited growth of all micro-organisms present. This results in a strong competition between the various micro-organisms for the meagre nutrients. The winners will dominate the mixed sludge population. The competition is also affected by variable process conditions such as temperature, wastewater composition (COD:N:P ratio, compounds present, etc.) or oxygen supply. Consequently, the sludge floc composition continuously changes.

The population is not static, but dynamic in nature. Operational problems such as bulking and scum formation occur when the ‘wrong’ micro-organisms are dominating the sludge population. Microscopic sludge investigation is therefore a must for process control and stable plant operation. The knowledge concerning diagnosing and solving operational problems in activated sludge plants has greatly expanded during the last decades. As such information is not always easily accessible for those working in the field, a multimedia package about this subject was compiled a couple of years ago 1 .

This package includes a CD-ROM and a manual and summarises the available knowledge about the biology of the process (incl. trouble shooting). It’s a very valuable aid for those responsible for design and operation of biological wastewater treatment plants (WTPs).

Identification of filamentous micro-organisms

Filamentous micro-organisms, mainly bacteria and occasionally also fungi, are responsible for the major operational problems that might occur in WTPs. Filaments are formed because the cells remain attached to each other following cell division. Thus, a filament consists always of a chain of cells, even when the septa between the cells can only be observed by a very high magnification (electron microscopy). Growing as a filament is characteristic for certain bacterial species. These organisms form filaments under nearly all conditions. Two examples of filamentous organisms are given in the figures 3 and 4. Dozens of filamentous organisms have been observed in activated sludge, differing from each other in morphological features (‘appearance’), growth kinetics, physiology, etc.

Ascertaining their identity is mainly important for three reasons:

  • Selecting a strategy with the best perspectives for controlling growth of a certain species. A universal applicable control method does not exist
  • The presence of some species tells us something about the actual process conditions in that plant. So, they may act as indicator organisms
  • The various filamentous species do not contribute to the same extent to operational problems. For example, compared with short and thin filaments the negative effect on the settling velocity of the flocs of long robust filaments is much larger

Due to lack of better methods, the in situ identification of filamentous species in activated sludge has been based for a long time upon their morphological features such as size and shape of the filaments and the results of several staining techniques. Although this ‘conventional method’ has been a great help in treatment plant operation (and still is!), it does not provide reliable information about the exact identity of the filaments observed. Thus, it is better to designate the filaments observed by the terms ‘type’ or ‘morphotype’ instead of by species or strain if phase-contrast and bright-field microscopy are used for filament identification. It is known nowadays that a specific morphotype may include several species. The opposite may happen as well: several morphotypes representing only one species.

Since the 1990’s, a genetically based identification method has been developed. Identification through ‘fluorescence in situ hybridization’, with the acronym FISH, is based upon the very specific sequence of the building blocks (nucleotides) in bacterial RNA. As the acronym already suggests, RNA is very similar to DNA, which carries the hereditary properties of living organisms. Various RNA fractions can be distinguished, but the rRNA fraction which is used for identification through FISH, is located in so-called ribosomes. Each bacterial cell contains thousands of ribosomes, minuscule protein factories where the enzymes are produced. In contrast to DNA, RNA is single stranded which means that binding sites are present along the chain opposite to each nucleotide. Added synthetic oligonucleotides with a nucleotide sequence complementary to the sequence in rRNA can be bound to the strand, a process which is called hybridization. This phenomenon is used by applying Fluorescence In Situ Hybridization.

Identification through FISH is only possible if the rRNA nucleotide sequence of that species is known. If this is the case, a short synthetic oligonucleotide, composed of 15-30 nucleotides, is synthesised with a base sequence exactly complementary to a specific rRNA region in that species. Subsequently, a fluorescent label is attached to the end of the chain. Such a synthetic oligonucleotide is called a probe. Probes are not very expensive and can be ordered from specialised companies. If a probe is added to activated sludge, binding will only occur in bacterial cells with an rRNA nucleotide sequence complementary to the nucleotide sequence in the probe. Thus, the attachment is very specific. Through applying fluorescence microscopy, red or green coloured filaments can be observed. An example is given in figure 5. The ultimate colour depends on the fluorescent dye applied. You are referred to reference 2 for more information about this method.

FISH versus conventional microscopy

FISH is an excellent identification tool. Due to the fluorescent signals obtained, it is relatively easy to ascertain the identity of filamentous species present in a mixed population of microorganisms. Compared with conventional microscopy, more accurate results are obtained with FISH. Through applying FISH it has also been ascertained that a specific morphotype may include several species. The opposite may happen as well: several morphotypes representing only one species. FISH also provides information about filamentous species more or less hidden inside sludge flocs. Finally, by applying FISH, it is possible to get information about the overall composition of activated sludge flocs.

However, just like any other method, FISH has its limitations and one should be well aware of these limitations when applying this method. First of all, identification through FISH is only possible if the probe required is available. Probes have been developed for a number of important filamentous species, but not yet for all species which might occur in activated sludge. A complete identification of all predominating filamentous morphotypes present by applying FISH was not possible in about 50% of the industrial samples screened during a research program carried out from 1998 until 2004. Thus, although numerous probes are already available, many probes are still missing. This means that relying completely upon FISH for filament identification in activated sludge is still not feasible.

FISH might affect the morphological features of filamentous species. As a consequence, it is not always easy to conclude which morphotype hybridised with the probe applied. Filamentous species, low in ribosomal content but still alive, will not give a clear fluorescent signal with the probe.

It might happen that a species specific probe also gives a fluorescent signal with other species, if their rRNA composition closely resembles each other. Such false-positive results can be prevented through applying so called competitor probes.

Sludges showing a large auto-fluorescence are not suitable for FISH. This will be the case, for example, with sludges rich in polycyclic aromatic compounds.

The identification of several Gram positive bacteria through FISH is only possible if an additional treatment with enzymes is applied to increase cell wall permeability for probes. A standard protocol describing such an enzyme treatment for different Gram positive bacteria is not available, which means that a trial and error approach is the only alternative.

Filament identification by conventional microscopy, and carried out by a well-trained person, takes less than two hours. FISH takes more time, particularly when several probes are required for a description of the filamentous population. Thus, if FISH is applied more accurate results can be obtained, but at higher costs.

Finally, FISH does not provide information about important sludge characteristics such as the protozoa population or the floc characteristics1. This means that FISH cannot replace conventional microscopy as an instrument for plant control.

Bulking in industrial WTPs

Filamentous bulking in domestic plants is nearly solved now. It is known which filamentous species are mainly involved and effective control methods have been developed. So, bulking is not any longer a serious problem in well-designed and operated, modern nutrient removal plants fed with domestic waste water.

On the contrary, up to recently hardly any reliable information was available about filamentous bulking in industrial waste water treatment plants. It was hardly known which filamentous species were causing operational problems and it turned out that the available bulking control strategies, mainly developed for application in domestic plants, were often ineffective when applied in industrial WTPs. To fill up this knowledge gap, two EU sponsored research projects were carried out between 1998 and 2004 by a consortium with partners from Italy, Denmark, Germany and The Netherlands. The research program included about 200 WTPs, covering a wide range of industrial sectors. Summarised, the following results were obtained:

  • Large filamentous population were observed in 74% of the samples investigated. So, bulking is still a serious problem in many industrial WTPs
  • 80 different filamentous morphotypes were observed, including about 50 completely unknown species. Consequently, the filamentous populations in domestic plants largely differs from those in industrial WTPs, which also implies that solving bulking in the latter will be more complicated (Figure 6)
  • Many new probes were developed to enable the fluorescence in situ identification of frequently occurring species. A complete identification of the filaments present by applying FISH was possible in 50% of the samples investigated
  • The nutritional requirements of several frequently occurring species were resolved, but more research about this aspect is still urgently needed
  • It was only rarely possible to correlate the occurrence of specific species with the wastewater characteristics
  • Identification keys, based upon FISH plus Conventional Microscopy, were developed to enable filament identification in industrial sludges

Detailed information about the results obtained is presented in reference 2.

Bulking control

Operational strategies aimed at bulking control in industrial waste water treatment plants are only summarised in this article. You are referred to the “Process Control” multimedia package 1 and other literature 3,4,5 for more detailed information about filamentous growth and bulking control options.

Controlling filamentous growth often resembles solving a puzzle with some missing pieces. The latter especially refers to the lack of information about the nutritional requirements and the growth kinetics of many filamentous species that might occur in industrial plants. Any systematic approach should always start with assessing the cause of the settling problems. Are filamentous species indeed present, which species or morphotypes are predominant in the sludge population and is any information available about the causative filamentous species?

‘Bad house keeping’ is a major cause for operational problems in industrial waste water treatment plants. Large, short-term fluctuations in flows, loads, temperature or pH should be prevented as they generally favour filamentous growth. So, buffer basins might be required for levelling peaks.

Employees in the production part of an industry should know which discharges to the sewer system will negatively affect a biological system. If an incident happens, they must immediately warn the plant operators.

Lack of oxygen reduces the activity of the aerobic floc forming bacteria, which means that the nutrients remain available longer for filamentous species. Maintaining complete aerobic conditions in the entire sludge floc requires an oxygen concentration of 2 mg/l in the bulk liquid in the aeration tank. Lower concentrations will often stimulate filamentous growth, especially in fully loaded plants.

There is a rule of thumb for the required BOD:Nitrogen:Phosphate ratio: 100 : 5 : 1. Lack of nitrogen or phosphate is responsible for bulking occurring in many industrial waste water treatment plants. More rarely, lack of micro-nutrients may play a role, which can be solved by dosing commercially available nutrient-mixtures.

Waste water rich in reduced sulphur compounds will nearly always stimulate growth of filamentous Thiothrix species. Solving bulking caused by Thiothrix requires the removal of these reduced sulphur compounds, for example through pre-aeraration, before the influent enters the aeration tank.

Three alternative routes remain when bulking cannot be solved by the measures mentioned so far. In contrast to domestic waste water, most industrial effluents contain a very large fraction of water soluble and often easily degradable compounds. It seems likely that many unknown filamentous morphotypes, just as it has been ascertained for most known species, use this fraction for their growth. When this fraction is largely removed before the flow enters the aeration tank, the chance on bulking occurring is greatly reduced. The soluble fraction can be removed largely through the incorporation of a highly loaded tank in the process line upstream of the aeration tank. Filamentous species do not grow fast enough to maintain themselves in such a highly loaded tank. Again, three options can be distinguished.

  • A two-stage anaerobic-aerobic configuration is a well known process. It prevents an excessive growth of many filamentous species, but will often not solve bulking problems caused by e.g. Haliscomenobacter hydrossis or Nocardioforms
  • A two-stage aerobic-aerobic activated sludge configuration with intermediate sludge settling and recycling is a known process as well. Aerobic-aerobic is also applied in the combination of a trickling filter plus a low loaded activated sludge tank as a second stage
  • Largely removing the soluble influent fraction can also be accomplished through incorporation of a highly loaded aerated tank without sludge settling and recycling as the first stage. The aerated liquor remains for about six hours in this tank and, subsequently, flows without settling directly into the aeration tank of the second stage. This relatively cheap solution has been successfully applied in several industrial waste water treatment plants in The Netherlands

The second main route concerns the application of so-called selectors. Aerobic, as well as anoxic and anaerobic selectors can be distinguished. In contrast to the previous options, the highly loaded selector is not a separate stage but more or less part of the aeration tank. Favouring floc-forming bacteria in their competition with filamentous bacteria for the available substrate is the aim of applying selectors. Selectors will not be effective with filamentous species characterized by their large substrate storage capacity. It is always recommended to start with a pilot plant before control methods, which require reconstruction of the plant, are applied on a full scale. The third main route for improving the settling velocity of activated sludge concerns the application of chemicals, either aimed at the destruction of the filaments or to improve the size or the weight of the flocs. This is only a temporary solution for settling problems, but might be required when bulking results in serious loss of biomass from the plant and immediate action is required.


1 1.Eikelboom, D. H. (2000) Process control of activated sludge plants by microscopic investigation. CD-ROM + Manual. IWA Publishing, London, UK.

2 Eikelboom, D. H. (2006) Identification and control of filamentous micro organisms in industrial wastewater treatment plants. CD-ROM. IWA Publishing, London, UK.

3 Jenkins, D., M. G. Richard and G. T. Daigger (2004) Manual on the causes and control of activated sludge bulking, foaming and other solids separation problems. IWA Publishing, London, UK

4 Lemmer, H und G. Lind (2000) Blähschlamm, Schaum und Schwimmschlamm – Mikrobiologie und Gegenmassnahmen. F. Hirthammer Verlag, München, Germany

5 Tandoi, V., D. Jenkins and J. Wanner (2005) Activated sludge separation problems – Theory, Control Measures, Practical Experiences. IWA Publishing, London, UK.

Published: 10th Sep 2006 in AWE International