This paper describes experiments with domestic sewage using a 120 litre expanded granular sludge bed (EGSB) reactor and a 205 litre fluidised bed (FB) reactor. Presettled domestic sewage was used in the experiments, because these reactor systems are inadeguate in removing SS. Compared to conventional UASB reactors, the advantage of EGSB systems is the significantly better contact between sludge and wastewater.
A batch recirculation system was used to assess the maximum achievable removal of the different COD fractions under EGSB and UASB conditions. The results obtained with the EGSB reactor reveal a removal efficiency of 90% with respect to the maximum obtainable efficiency of the soluble COD fraction under dry weather conditions. This efficiency can be obtained at hydraulic retention times (HRT) exceeding 3 hours. Even at HRT's ranging from 2 - 1.5 hours, still 84% - 77% of the maximum possible removal efficiency as assessed in batch recirculation experiments can be achieved at temperatures exceeding 13°C.
The advantage of EGSB systems is the high volumetric loading rate that can be applied while maintaining a high removal of the dissolved COD fraction. Moreover, little if any accumulation of inert suspended solids occurs in the sludge bed.
On the other hand, the poor removal of suspended solids can be considered as a disadvantage of these systems in treating domestic sewage.
Conventional FB systems using sand as carrier material were shown to have little prospect for treating settled domestic sewage, because hardly any methanogenic activity will develop in such a system.
However, it was also found that a very satisfactory guality granular sludge developed on settled sewage when operating FB systems in a mode similar to EGSB systems. This granular sludge appeared to be of a guality egual or even better than that of granular seed sludge, cultivated in a UASB reactor treating papermill wastewater.