PREDICTIVE MODEL FOR TREATMENT
OF PHENOLIC WASTES BY ACTIVATED SLUDGE
Alan Rozich, Research Assistant
Anthony F. Gaudy, Jr., Professor
Peter D'Adamo, Research Assistant
Department of Civil Engineering
University of Delaware
Newark, Delaware 19711
It is well know that certain toxic organic compounds, e.g., phenol or benzene, can serve
as sources of carbon for heterotrophic microorganisms and are thus susceptible to biological treatment. There is much interest in phenol because the production of phenolic wastes
can be expected to increase due to greater reliance on coal for energy and base organic
chemicals. Although fundamental metabolic and kinetic studies on removal of toxic organics
are not abundant, those which have been done are largely concerned with phenol. Thus,
phenol is becoming the most-used model compound for the study of inhibitory organics as
glucose has been for the study of metabolic mechanisms and kinetics for noninhibitory
organics.
As far back as 1929, Mohlmann reported that low levels of phenol could be successfully
treated by activated sludge [ 1]. Interest in the treatment of phenolic wastes has not waned
and there are several papers available which discuss general design and operational guidelines for biological treatment of wastes containing phenolic compounds. The general consensus is that the presence of phenolics enhances susceptibility of a biological process to periodic upsets [2-4].
Several interesting case histories are available. For example, Capestany et al. [5] reported
that an activated sludge plant fed phenol at 1000 mg/1 and operated with a mean hydraulic
retention time of 24 hr produced an effluent phenol concentration of 0.5 mg/1. These writers
also noted that sulfur was a key nutrient for success of the treatment process. Kostenbader
and Flecksteiner [6] also reported low effluent levels of phenol (less than 1 mg/1) for biological treatment of coke plant ammonia liquor. Volesky et al. [7] found that effluent
phenol concentrations of less than 0.2 mg/1 could be produced using a 24-hr mean hydraulic
retention time for a combined waste in which the influent phenol concentration was 120
mg/1. Adams [8] reported similar results using a 2-day detention time. In their pilot plant
studies, Holladay et al. [9] compared the performance of three types of biological reactors
degrading phenol—stirred tank, fluidized bed and packed bed. They concluded that activated
sludge was the least desirable since it exhibited the lowest degradation rate and was subject
to operational upset.
The results of several kinetic studies indicate conflicting conclusions regarding the need for
incorporation of an inhibition function in model equations for predicting effluent substrate
and biomass levels in fluidized (activated sludge) reactors treating inhibitory substrates.
Neufeld and Valiknac [10] reported that the Monod model for nonhibitory substrates
could suffice for expressing the relationship between specific growth rate and substrate
concentration for inhibitory substrates. Also, Beltrame et al. [ 11 ], as a result of batch
studies, reported finding no inhibition by phenol at high concentrations. However, a graph
depicting some of the kinetic experiments shows a lower substrate utilization rate with increasing initial phenol concentration. In a second study [12], these workers found no inhibition in a continuous stirred reactor employing cell recycle. They also felt that the normal
Monod expression was sufficient to relate specific growth rate to substrate concentration.
On the other hand, Sokol and Howell [13] have recently conducted batch growth studies
using a pure culture harvested from a continuous growth reactor and found that the sub-
619