UNDERSTANDING THE RESPONSE OF ENHANCED BIOLOGICAL PHOSPHORUS REMOVAL BIOMASS TO A RANGE OF ORGANIC SUBSTRATES

Anaerobic/aerobic batch experiments were conducted with a variety of volatile fatty (VFAs) on two sequencing batch reactor (SBR) populations displaying Enhanced Biological Phosphorus Removal (EBPR). The batch experiments were fairly consistent between the two systems and with the past literature: acetic and isovaleric acid were the most efficient substrates with respect to increasing net phosphorus (P) removals, and valeric and propionic acid were less efficient than these two VFAs. The fundamental reason was because both isovaleric and acetic acid resulted in predominantly 3HB storage as opposed to 3HV. 3HB content correlated with higher aerobic P uptake per unit PHA (i.e. ƒP,UPT using the nomenclature of Barker and Dold, 1997) than 3HV. It was also noted that 3HB content had an inverse relationship with carbohydrate/glycogen biosynthesis during aerobiosis. Thus it may be that ƒP,UPT was not different for 3HB vs. 3HV, but that ƒP,UPT is different depending on the fate of PHA carbon (i.e. does it go to glycogen, CO2, or is it used for maintenance). The model of Smolders et al. (1995) for acetic acid transformation to 3HB implies that ƒP,UPT would be lower if the fraction of PHA carbon shunted to glycogen increased relative to that dedicated to polyphosphate biosynthesis. By inference this data also indicated that a simple correlation between negative redox balance of VFA→PHA transformation and high net P removals was probably coincidence and the negative redox balance was not the cause. This was confirmed because low PHA biosynthesis for substrates with positive redox balances did not occur. SBR data implied that acetate/propionate ratio is potentially a significant parameter with respect to EBPR and biological nutrient removal (BNR) system performance, and by implication design and modeling.