Peter Fox

A comparison of media types in acetate fed expanded-bed anaerobic reactors.

Abstract A synthetic feed, containing acetate as the only carbon source, was used to start-up four different anaerobic expanded-bed reactors containing three different types of microbial attachment media. The media types used were low-density anthracite, granular activated carbon (GAC) and two sizes of sand. All media types were of the same average diameter, 0.7 mm, except for a smaller sand, 0.35 mm. These media types were chosen to compare surface roughness, macroscopic shear stresses due to upflow velocity and sphericity. The 0.7 mm sand required the greatest upflow velocity, 16 cm/s, while the other reactors had upflow velocities of 5.5–6.0 cm/s. Sand had the least surface roughness and GAC had the roughest surface, while anthracite had the most angular shape. At steady-state, the GAC reactor retained 3.75–10 times the attached biomass retained on the other media tested and the GAC reactor accumulated biomass at a faster rate during start-up. Shear losses reflected the biomass accumulation with the two sand and anthracite media having shear loss coefficients 6–20 times greater than that of the GAC medium. Sand induced the formation of sludge granules in both sand reactors with two species of methanogens and stability of the sludge blankets was critical to reactor performance. Scanning electron microscopy demonstrated that attached growth developed in crevices where biomass was protected from shear forces. Attached growth on the sand and anthracite media was located only in crevices, while the GAC medium is completely covered with crevices and biofilm developed on the entire GAC particle. Surface roughness was critical to biofilm development with the rougher surface providing the better attachment medium.

Surface area and travel time relationships in aquifer treatment systems.

Soil aquifer treatment (SAT) and bank filtration use natural attenuation processes to purify water for subsequent use. Soil aquifer treatment may constitute both unsaturated and saturated flow conditions, while bank filtration systems are primarily saturated flow. This analysis focuses on the saturated zone, where the majority of residence time occurs, in both SAT and bank filtration systems. Sustainable removal mechanisms during subsurface flow are primarily surface-mediated and therefore depend on surface area. By analyzing saturated subsurface flow hydraulics in granular media, a relationship between surface area and travel time was developed. For saturated subsurface flow, the ratio of surface area-to-travel time varied by approximately a factor of 3, for common aquifer materials subject to identical hydraulic gradients. Because travel time criteria often are used to regulate SAT and bank filtration systems, these criteria also may determine the surface area and associated surface-mediated reactions for water purification. The ratio of surface area-to-travel time increases with increasing hydraulic gradient, implying that surface area is relatively constant for specific travel times, even if the hydraulic gradient changes; however, the increasing hydraulic gradient will increase the distance from the recharge zone to the recovery well. Therefore, travel time assessments based on maximum possible hydraulic gradients increase surface area and could provide a conservative limit for surface-mediated reactions. This analysis demonstrates that travel time criteria for SAT and bank filtration systems indirectly provide a minimum surface area that may support sustainable removal mechanisms.

Effects of primary substrate concentration on NDMA transport during simulated aquifer recharge.

N-nitrosodimethylamine (NDMA) is known to be highly carcinogenic and is present in drinking water, wastewater, and a variety of foods. Because of its presence in chloraminated water at nanogram per liter concentrations, NDMA has become an emerging issue for reclaimed water which may be used for aquifer recharge or irrigation. This research investigated the fate of NDMA in two soil column systems used to simulate subsurface transport. One column system was operated under aerobic conditions with increasing primary substrate concentration where the biodegradable organic carbon (BDOC) in reclaimed water was used as the primary substrate. The reclaimed water content in the influent was increased from 0 to 25% in the column to increase the BDOC concentration. Negligible NDMA removal was observed at 0% reclaimed water and increasing the primary substrate in the influent resulted in NDMA removal suggesting that biodegradation of NDMA might be a cometabolic process. The effects of redox conditions on NDMA fate was studied by operating a second column system with 100% reclaimed water under anoxic conditions and then changing the conditions to aerobic. It was observed that NDMA removal was similar under both aerobic and anoxic condition, however, much lower effluent concentrations were observed under aerobic conditions. Under anoxic condition, a normalized mass removal rate of 254 ng NDMA/mg DOC was observed which increased to 273-ng NDMA/mg DOC under aerobic conditions. The majority of NDMA and substrate removal occurred in the first of three columns in series column under both aerobic and anoxic conditions. Normalized mass removal rates of NDMA after the first, second, and third columns were 372, 30, and 20 ng NDMA/mg DOC, respectively. Since the majority of dissolved organic carbon was also removed in the first column, NDMA biodegradation was consistent with cometabolic activity. Batch tests verified the biodegradation removal potential of NDMA. Addition of a methylotrophic substrate, methanol and an aromatic substrate, toluene, did not increase NDMA removal.

Calculation of residence time distribution from tracer recycle experiments

This note develops a procedure that can be used to calculate the single pass residence time distribution in a reactor system, such as a landfill, operated with recycle, from pulse tracer studies. An equation relating the effluent tracer concentration and residence time distribution (“E curve”) is derived, and a numerical technique to solve such an equation is presented and applied to data obtained from experiments on a laboratory-scale reactor. The significance of the procedure developed to biological reactors is discussed.

Characteristics of biotic and abiotic removals of dissolved organic carbon in wastewater effluents using soil batch reactors.

Biodegradable dissolved organic carbon (BDOC) analyses and abiotic adsorption of dissolved organic carbon (DOC) from different wastewater effluent were conducted to evaluate biotic and abiotic removal mechanisms as a function of the initial DOC concentration and source of DOC using soil batch reactors. To obtain high DOC concentrations, a laboratory-scale reverse osmosis unit was used. It was found that BDOC fraction was independent of the initial DOC concentration and was dependent on the source of wastewater and/or the types of wastewater treatment. The BDOC fractions varied from 9 to 73%. Trickling filter effluent (Tucson, Arizona) showed the highest BDOC, ranging from 65 to 73% biodegradable, while wastewater treated by the soil aquifer treatment (SAT) (NW-4) was found to be most refractory, with DOC removals of 9 to 14%. For nitrified/denitrified tertiary effluent (Mesa, Arizona) and secondary effluent (Scottsdale, Arizona), 36 to 42% removal of DOC was observed during the BDOC test. The amount of BDOC in the wastewater depended not on the concentration of DOC, but on the effectiveness of pretreatment. Abiotic adsorption capacity of wastewater effluent varied from 6 to 18%. Molecular weight distribution analyses showed that more than 50% of DOC in the Scottsdale concentrate had a molecular weight of less than 1000 Da, and no significant change in distribution profiles occurred after approximately 12% abiotic adsorption with both soils with acclimated microorganisms (SAT soil) and soils without acclimated microorganisms (non-SAT soils). Hence, preferential adsorption was not observed and the presence of acclimated microbes did not influence adsorption.

A comparison of expanded-bed GAC reactor designs for the treatment of refractory/inhibitory wastewaters

Abstract A hybrid granular activated carbon (GAC) expanded-bed reactor designed to decouple biological removal from physical removal and a GAC expanded-bed reactor were studied under similar operating conditions. The hybrid GAC reactor consisted of a GAC expanded-bed reactor which was retrofitted with a GAC adsorber containing 11.1% of the total GAC mass in the system. A synthetic wastewater composed of acetate and 3-ethylphenol (3-EP) at influent concentrations of 5 g acetate/l and 1.5–2.5 g 3-EP/l was used at COD loadings of 18.7–23.8 kg COD/m 3 . Partial replacement of GAC in both reactors facilitated the physical removal of 3-EP. GAC was replaced from only the side adsorber in the hybrid GAC reactor while GAC was replaced directly from the GAC reactor. Previous studies indicated that increasing GAC replacement rates washed biomass out of the GAC reactor while the hybrid reactor was resilient to changes in operating conditions. This study compared the reactors while maintaining a constant GAC replacement rate with an increasing influent loading of the inhibitory/refractory compound, 3-EP. The performance of the GAC reactor deteriorated with increasing 3-EP loading rates until reactor failure occurred due to inhibition. The hybrid reactor responded to increases in the influent 3-EP loading rate with greater than 60% biological removal of 3-EP and only minor changes in effluent quality. Termination of GAC replacement in the hybrid reactor allowed for almost complete biological removal of 3-EP in the hybrid reactor. The resilience of the hybrid GAC reactor was attributed to the accumulation and maintenance of a stable microbial population capable of degrading an inhibitory substrate.

Adsorption of biologically inhibitory compounds as a process control mechanism in biological reactors

Abstract An innovative reactor design that decouples biological removal mechanisms from physical removal mechanisms has demonstrated promise for the treatment of wastewaters containing high concentrations of inhibitory compounds. Inhibition and toxicity prevent treatment of such wastewaters in conventional biological reactors. The reactor design consists of a high-rate biological reactor with a granular activated carbon (GAC) adsorber inserted into the recycle line of the biological reactor. Partial replacement of GAC from the GAC adsorber provides a mechanism for controlling the concentration of inhibitory compounds in the biological reactor. As a process control parameter, GAC replacement can be used to maintain the concentration of inhibitory compounds in a range optimal for growth and acclimation. GAC replacement is also varied in response to changes in the influent loading and can be used to provide rapid recovery from shock loadings. Agreement between isotherm studies and experimental data from pilot-scale systems was observed when the average GAC particle residence was greater than 3.75 days. Isotherm studies may be used to design the GAC adsorber and predict optimal operating conditions. The concept of using an adsorption process to optimize biological removal provides an environmentally sound treatment alternative for many high-strength wastewaters.