Dec. 5, 2002 -- As much of the organic material and the nutrients in municipal wastewater relate to particulate material, advanced particle removal by physical-chemical pre-treatment will lead to a lower pollutant load on subsequent treatment steps - so the total treatment system can be smaller and operated in a more energy efficient way.
Traditionally, the choice of treatment methods is based on effluent standards and practical experience. In particular, the choice of adding pre-treatment steps, customarily a primary sedimentation tank, ahead of activated sludge systems has not generally been based on technical knowledge and specific wastewater characteristics, but rather on traditional practice and plant size.
Earlier studies demonstrate that a more detailed characterisation of the particulate matter is required to better predict the removal efficiencies of physical-chemical treatment techniques and subsequent optimal chemical dosages. Such a characterisation should include the distribution of contaminants over various particle sizes. Intensive wastewater characterisations have been conducted world-wide to determine biological fractions for activated sludge models, but little data is available on the size distribution of wastewater pollutants.
This article describes experimental research aimed at determining how wastewater contaminants are distributed over different particle sizes. With this data, the effect of various particle removal techniques can be predicted, and the requirements and efficiencies of post-treatment determined. The results can also indicate whether a certain wastewater can be treated efficiently with physical-chemical pre-treatment methods, and to what extent.
Calculations were made to identify the effect of advanced particle removal in the first treatment step on the biological post-treatment and total wastewater treatment system. The environmental criteria evaluation model DEMAS was used to calculate energy and cost potentials of increasing particle removal.
The Experiment
To determine how contaminants in municipal wastewater are distributed over particle size, a size fractionisation was carried out on eight wastewaters from different sources. The different fractions were analysed for COD, BOD, nitrogen, phosphorus, suspended solids, turbidity and conductivity.
Sampling and fractionation
Wastewater grab samples were taken at WWTP in Apeldoorn, Arnhem, Bennekom, Berkel, Boxtel, Haarlem, Hoek van Holland, and Vlaardingen. The samples were collected after screening, just in front of the pre-treatment facility, since this provides the data required. Duplicate samples were taken on three different days at the same location at the same time, and where possible during the same weather conditions. In total, 48 samples were taken.
Grab sampling was preferred over flow proportional long term sampling since processes like adsorption, degradation or destruction could lead to changes in the wastewater composition regarding particle size distributions. However, it was acknowledged that grab samples would only give a 'snapshot' picture of the wastewater composition at those specific times.
The samples were immediately analysed for temperature and pH, stored in polyethylene containers of three to five litres, and sent directly to the laboratory for fractionation and further analysis. To minimise any effects caused by delay during transport and after arrival at the laboratory, the samples were kept at a constant temperature of 4C.
The samples were fractionated over clean, pre-flushed stainless steel sieves and membrane filters (cellulose nitrate composition) with five different pore sizes (63, 5, 1.2, 0.45 and 0.1µm). The filtrates were analysed for COD, BOD5, nitrogen, phosphorus, suspended solids, turbidity, conductivity and biodegradability via extensive BOD analyses on day 1, 2, 3, 5, and 10.
Results and Discussion
Average wastewater fractions
From the fractionation data an 'average' fractionated wastewater composition was determined. Forty four data sets per fraction were used for this calculation: of the original 48 samples, four from Hoek van Holland were eliminated following problems that influenced the composition. The average composition provides a general picture of the size distribution of the various contaminants in the different wastewater sources, but should not be taken as being generally valid.
In the achieved average wastewater fractionation, the percentage of oxygen consuming components related to the settleable particle fraction was low, with a maximum of 21 per cent for COD. A major part of the total BOD (44 per cent), COD (38 per cent) and phosphorus (35 per cent) was present in suspended and supra-colloidal particle fractions with particle diameters between 1.2 and 63 µm. For nitrogen, only 4 per cent can be related to settleable particles and 13 per cent to colloidal and suspended fractions; so 83 per cent of the nitrogen is present in the soluble form. The conductivity of the wastewater decreased insignificantly with increasing particle removal.
With the particle removal increased, the BOD/N ratio declined since more BOD than nitrogen is related to particles. The average BOD/N ratio decreases from 4.1 in the raw wastewater to 3.6 due to the removal of settleable particles down to 63µm. With total particle removal, the BOD/N ratio decreased to 2.4. In wastewater diluted by rain, the BOD/N ration was already low (2.4) in the influent and decreased to below 1 after complete particle removal. This will negatively influence the denitrification potential of the remaining wastewater.
Energy calculations based on wastewater fractionations
Calculating the potential energy consumption of the remaining wastewater after removal of certain particle fractions can show how advanced particle removal can effect the energy efficiency of a WWTP.
First, the energy consumption in a theoretical activated sludge system with COD and N removal was calculated based on the wastewater fractionation. With the evaluation model DEMAS, the total energy balance of the water treatment and sludge handling system for 100000 p.e. and a flow of 20000 m3/day was calculated. The total energy balance shows a steeper decline in energy consumption with more advanced particle removal, due to the fact that sludge production and handling depend strongly on particles and may lead to a positive energy input. The highest energy saving is possible by the removal of the settleable fraction down to 63 µm (energy savings of 364 MWh/y, 34 per cent of total energy savings) and the suspended particle fraction between 63 µm and 5µm (458 MWh/y, 46 per cent of total energy savings). Due to the removal of the remaining colloidal particles between 5 and 0.1 µm a further 20 per cent (equivalent to 284 MWh/y) of the energy can be saved.
The overall consumption of each particle removal technique has to be compared with the overall energy saved by removing specific particle sizes. Thus the application of a certain pre-treatment technique can be counterbalanced with the resulting overall energy savings in the WWTP.
Cost calculations based on wastewater fractionation
In addition to energy calculations, space requirements, chemical use, final sludge treatment and even effluent quality can be calculated based on wastewater characterisation by particle size. This can be summed up in the total treatment costs calculated as net present values. For example, to achieve break-even, a technique that removes particles down to 63 µm (for example, primary sedimentation) may cost about Euros 5 million. For a pre-treatment technique that removes particles down to 5 µm, Euros 13 million may be invested, while an additional removal to 0.1 µm calls for an additional investment of less than Euros 4 million.
Pre-treatment potential and optimal chemical dosage
As described above, wastewater fractionation based on particle size makes it possible to predict the suitability and effect of specific pre-treatment techniques on a specific wastewater - the so-called pre-treatment potential. From the fractionation test not only the primary effluent composition but also the quantity and quality of the primary sludge can be derived. This provides an opportunity to create (for example) specific polluted and non-polluted sludge streams.
Another optimisation step using wastewater fractionation based on particle size could be the determination and control of chemical dosages. By using online fractionation data, an optimal coagulant or flocculant dosage could be applied to remove the desired number of particles or specific components from the wastewater, or online turbidity-controlled polymer dosing could be used to set particle removal to the level that allows biological treatment to work optimally.
CONCLUSIONS AND RECOMMENDATIONS
In the 'average' wastewater fractionation, the percentage of oxygen consuming components related to the theoretically determined settleable particle fraction is low, with a maximum of 21 per cent for COD. A major part of BOD (44 per cent), COD (38 per cent), and phosphorus (35 per cent) is present in suspended and supra-colloidal particle fractions with particle diameters between 1.2 and 63µm. For nitrogen, only 4 per cent can be related to settleable particles and 13 per cent to colloidal and suspended fractions; that leaves 83 per cent of the nitrogen present in soluble form.
Through wastewater fractionation, the 'pre-treatment potential' of a specific wastewater can be established. In addition, calculations show the importance of particle removal in the pre-treatment down to approximately 5 to 1 µm, depending on the wastewater, to save energy and costs in the total wastewater treatment system.
From wastewater fractionation experiments, the primary effluent composition and the amount and composition of the produced primary sludge for a specific wastewater can be established. The specific coagulant or flocculant dosage to gain the most efficient particle removal in the pre-treatment can be determined, and controlled by turbidity-related dosing of organic polymers and online particle size counting.
The results derived from the experiment were promising and merit testing on a broader scale. More advanced (i.e. online) sampling, separation, fractionation and analysing techniques should be implemented to optimise the tests.
The authors of the original paper are A. F. van Nieuwenhuijzen and A. R. Mels. It describes how the characterisation of particulate matter in municipal wastewater can lead to improved efficiencies in treatment. A. F. van Nieuwenhuijzen works with Witteveen+Bos Consulting Engineers and Delft University of Technology's Department of Sanitary Engineering in The Netherlands. He can be contacted by email at: [email protected].
