Journal of Environmental Engineering

The trickling filter solids contact (TF/SC) process is now over 20 years old, having been first demonstrated successfully in 1979 at Corvallis, Oreg. Design and operating concepts have evolved significantly since the first generation of plants. The purpose of this paper is to document the changes that have occurred since prior reviews. Data from four large plants (Santa Cruz, Littleton/Englewood, Central Valley, and Annacis Island) are used for illustration. The basis of the original peer reviewed U.S. EPA definition of the TF/SC process is reviewed. Pilot-scale and plant-scale research coupled with advanced analysis tools had allowed the overall horizons of the TF/SC process to be expanded, such that higher loadings have been obtained. For instance, use of modern plastic media has allowed the criterion that the majority of the biochemical oxygen demand be removed on the trickling filter while sustaining high total organic loadings. Owing to the features of the flocculator-clarifier and the superior thickening and compaction qualities of TF/SC solids, high solids loading rates can be obtained. As a result, current advanced designs occupy less footprint, and it is axiomatic that costs have been reduced. Furthermore, there is considerable potential to rerate, optimize, and debottleneck existing TF/SC plants to release low cost capacity.

The effects of biomass accumulation and distribution on air pressure losses in biofilters were experimentally studied. Two bench-scale biofilters, one packed with inert porous pellets (Nova Inert) and the other with wood chips, were operated under similar conditions with excess nutrients to treat an airstream containing methanol, at loading rates of 100–150 g methanol/m³ bed/h. Localized biomass accumulation in the biofilter beds was the key factor increasing the pressure drop, which was caused by local bed clogging due to biomass growth. The highest pressure drops in the beds (wood chips: 2,600 Pa/m; Nova Inert: 550 Pa/m) occurred in sections where there were high biomass levels with high water content. The pressure drop varied nonlinearly with the amount of accumulated biomass and the amount of methanol consumed. Sixfold higher pressure drops were measured in the wood chip biofilter than in the Nova Inert biofilter because of more biomass growth and bed compaction. A model, based on the Ergun equation, was developed to predict biomass-affected porosity and pressure drop as a function of the biomass concentration in a bed packed with spherical pellets. A comparison of the experimental and the predicted pressure drops showed that the model provided good estimates of biomass-affected porosity and pressure drop in the biofilter packed with spherical porous pellets with even biomass distribution.

A control system to automate the start-up of anaerobic wastewater treatment reactors is presented. The system controls the feedflow rate, using the biogas production or the biogas flow rate as the only on-line variable. Furthermore, other off-line parameters, such as methane content in the biogas and COD of influent are also necessary to calculate the COD balance in the reactor. Two different start-up strategies were applied: fed batch and continuous operation. In the fed-batch operation the frequency of feeding is automatically set depending on the efficiency in COD removal. The residual organic load fraction (the fraction of influent COD not degraded to methane), is the key parameter enabling the controller the modification of feeding frequency. In order to improve reliability of the system, a second parameter k representing the gas flow rate, was introduced. By using this strategy, starting from an organic loading rate (OLR) lower than 0.5 kg COD/(m³⋅day), a load higher than 8 kg COD/(m³⋅day) was achieved in only 33 days, with an efficiency, in terms of COD removal, higher than 90%. When the system was operated in continuous mode, the key parameter applied is again the residual organic loading rate fraction, expressed as the percentage of COD that remained undegraded in methane. Two extreme values—“minimum” and “maximum”—have to be defined. Once the system reaches the minimum, the controller increases the feeding pump speed automatically, and when it attains the maximum value, the feeding flow is decreased proportionally. In order to ensure stable operation, a further parameter, the waiting time between the moment at which the threshold value is reached and the modification of the feeding flow rate, was introduced. 24 h has proven to be an excellent value for this purpose. By means of this strategy, starting from an influent OLR lower than 0.5 kg COD/(m³⋅day), an OLR of 9–12 kg COD/(m³⋅day) was achieved in 40 days, with COD removal efficiency higher than 95%.

A biofilm based, annular reactor method was developed and used to measure the biological regrowth potential of effluent water from various pilot treatment processes at the New York City Croton Lake Pilot Plant. A series of studies were carried out over the year-long study to collect bacterial growth and organic carbon biodegradation data for waters from six treatment options, including the raw source water. Quantitative and qualitative evaluations were made to determine the effects of filter media type, direct filtration, preozonation, and primary chlorination on the relative biostability of the produced waters compared to that of the original source water and water currently being distributed to consumers. In addition, results were compared to those obtained using traditional biodegradable organic material measuring methods such as assimilable organic carbon and biodegradable organic carbon. Quantitative biostability factors were developed that take into account both biological growth potential and biodegradability of the tested waters. Results from these studies were used to compare various piloted treatment processes and to assess pilot plant operation, design parameters, and seasonal source water quality.

This research investigated whether iron oxyhydroxides used in conjunction with microfiltration could improve the removal of metals from a waste while maintaining adequate flux. Filtration of individual metals, a mixture of metals, and the mixture with two iron oxides were investigated. The research indicated that use of a coagulant (an iron-containing salt) might not be as important in microfiltration systems as in conventional gravity separation systems, because of the small pore size of microfilters. In some cases, filtration efficiency was relatively poor at the beginning of a treatment cycle, but in almost all cases it became excellent once a layer had built up on the membrane surface. The greatest benefit provided by iron oxides might be to reduce membrane fouling. A crystalline iron oxide such as goethite is more attractive than ferrihydrite. The flux improvement with goethite is greatest if a thin layer is deposited on the membrane surface before the contaminant metals are injected into the system, so that the goethite can trap the potentially foulant metal hydroxide particles away from the membrane surface.

Bentonite, forest soil, and spruce bark were submitted to batch adsorption testing, leaching cell testing, and selective sequential extractions (SSEs) to investigate the heavy-metal compatibility of clay barriers and the potential of forest soil and spruce bark as clay barrier materials. The materials ranked as follows according to sorption capacity: forest soil > bentonite = spruce bark. The hydraulic conductivity values of heavy-metal leachates were two orders of magnitude greater than those of the blank (0.01 mol calcium nitrate) leachate. The forest soil admix ranked first in terms of heavy-metal retention capacity and breakthrough points. The mobility of Cd was 4.5 times higher than that of Pb, and Cu was 2.5 times more mobile than Pb. The leaching cell and SSE results indicate that heavy metals cause significant preferential channeling. The SSE results show that the addition of forest soil and spruce bark to clay barrier mixes promotes heavy-metal fixation.

Municipal and industrial sewers may be localized sources of volatile organic compound (VOC) emissions to the ambient atmosphere. Previous studies of VOC emissions from sewers have focused on sewers with large diameters that are often characterized as having mild channel slopes and as conveying relatively large wastewater flow rates. The study described in this paper was completed to better understand VOC emissions from sewer reaches with small diameters, steep channel slopes, and relatively low wastewater flow rates (e.g., as might be typical for building laterals, street sewers, and on-site industrial sewers). Mathematical models were developed to investigate the nature of mass transfer kinetics and equilibrium conditions in such sewers. A series of 20 experiments were then completed to determine liquid-phase and gas-phase mass transfer coefficients for a range of sewer operating conditions and chemical properties. Experiments were completed in an experimental sewer reach (60 m length, 0.2 m diameter) using five volatile chemicals (acetone, ethyl acetate, toluene, ethylbenzene, and cyclohexane, listed in order of increasing Henry's law constants). Experimental stripping efficiencies were as high as 47% for cyclohexane and as low as 0.3% for acetone. Experimental and mathematical results indicate that VOCs with low Henry's law constants (e.g., acetone) can reach equilibrium conditions rapidly in sewers. However, emissions of VOCs with high Henry's law constants (e.g., cyclohexane) are kinetically limited, allowing for the sewer to be treated as an “open” system. The findings described herein suggest that a large fraction of VOCs with high Henry's law constants may be emitted to the ambient atmosphere in the near vicinity to the point of discharge.

Catalytic incineration is one of the cost-effective technologies to solve troublesome volatile organic compounds. However, some sulfur containing volatile organic compounds, such as ethyl mercaptan, may deactivate the Pt catalyst that is commonly used in the catalytic incineration process. This paper provides information on the poisoning effect of ethyl mercaptan. The catalytic incineration of ethyl mercaptan, typically emitted from the petrochemical industry, over a Pt/Al₂O₃ fixed-bed catalytic reactor was studied. The effects of operating parameters including inlet temperature, space velocity, C₂H₅SH concentration, O₂ concentration, and catalyst size were characterized. Catalytic incineration on a mixture of C₂H₅SH with CH₃SH was also tested. The results show that the conversions of C₂H₅SH increase as the inlet temperature increases and the space velocity decreases. For the temperature from 200–260°C, the higher the C₂H₅SH concentration is, the lower its conversion. The O₂ concentration has a positive effect on the conversion of C₂H₅SH. C₂H₅SH has a poisoning effect on the Pt/Al₂O₃ catalyst, especially at lower temperature. The existence of CH₃SH has no effects on the conversion of C₂H₅SH.

Magnetic separation technology was applied to remove dispersants and crude oil from water with magnetite and maghemite. Maghemite exhibited rather constant removal efficiency for dispersants regardless of surfactant types, while magnetite exhibited higher removal efficiency for anionic surfactant, and the efficiency was higher in deionized water than in salty water that contains more ions. Sorption of the dispersants to magnetite can be explained with electrostatic attraction, while binding of the dispersants to maghemite can be described with electrostatic attraction as well as with the structural characteristics that provide high sorption capacity. The result from a water bath experiment, which was to test the collection efficiency of magnetic particles from water, indicated that the recovery efficiency of magnetic particles was nearly 100% after the dispersants had been sorbed. More than 80% of the oil was collected when the magnetite-to-oil ratio was more than 0.89, while the same percentage of oil was harvested when the maghemite-to-oil ratio was more than 0.46 in the oil removal experiment. Sorption of crude oil to magnetic particles can be explained with the fine particle–oil flocculation, which is associated with an electrostatic attraction between the magnetic particles with charged surface and polar compounds in the crude oil.

In the transportation planning for some industrial wastes, in addition to hauling cost, environmental impact must frequently be considered. A notable example is transporting waste soil generated by major construction projects. Adequate transportation planning is particularly important for construction in a metropolitan area. In this study, we present a novel two-phase approach to address the multiple-criteria decision problem. The first phase applies the fuzzy analytic hierarchy process to obtain a “composite impedance” for each road sector where transportation costs, environmental impact, and traffic congestion are considered in the evaluation. The second phase employs fuzzy mathematical programming to find the optimal transportation network based on the fuzzy impedance. An illustrative example is provided for the transportation planning for waste soil of the Kaohsiung mass rapid transit system construction project. The optimal solutions using the proposed approach are compared with the solutions using the conventional shortest-path approach where minimizing the transportation cost is the only objective.

Ozone inactivation of Cryptosporidium parvum oocysts was studied at bench-scale in 0.05 M phosphate buffer at 1 to 37°C, pH 6–8. Animal infectivity using neonatal CD-1 mice was used for evaluation of oocyst infectiousness following treatment. Survival curves of ozone inactivation were characterized by a tail-off effect, with an initial shoulder most evident at low temperature. Temperature was a critical factor for ozone inactivation kinetics with a significant decrease of ozone efficacy at low temperature. Accounting for ozone residual stability at different pH conditions, pH was found to have no significant effect on the activation of C. parvum by ozone. Inactivation kinetics at different temperatures were expressed as an Incomplete gamma Hom model with different reaction rate constants, adjusted for water temperature using the van't Hoff-Arrhenius relationship. Between 1 and 37°C, for every 10°C decrease in the water temperature, the inactivation rate constant decreased by a factor of 2.2, corresponding to activation energy of 51.7 kJ/mol. Ozone disinfection design criteria for 1.0 and 2.0 log-units of inactivation of Cryptosporidium were developed for various water temperatures, and 90% confidence intervals are also provided.

Perchlorate $({\mathrm{ClO}}^{{\mathit{-}}_{\mathit{4}})}$ can be used as a terminal electron acceptor by some bacteria for cell respiration during the oxidation of substrates. Packed-bed biofilm reactors have been used to treat perchlorate-contaminated waters with rates varying over six orders of magnitude (0.0007 to 20 mg/L-min). While this range implies large differences in treatment efficiencies, it is demonstrated in this report that these rates are correlated to concentrations of perchlorate in the reactor. Overall perchlorate reduction rates obeyed first-order kinetics for reactors using organic substrates as electron donors (acetate or a complex high-protein medium). Higher reaction rates appear possible in hydrogen-gas feed reactors rather than in organic-feed reactors, although additional work is necessary to confirm this preliminary result. Findings of first-order perchlorate kinetics will help engineers design packed-bed reactors for the remediation of perchlorate-contaminated waters.