Journal of Environmental Engineering

Laboratory studies have shown phytoremediation to be a feasible method for remediating surface soils contaminated with organic compounds. Evaluation of this technology in the field is difficult because of the inherent spatial heterogeneity in the hydraulic and chemical properties of the soil. In this study, total petroleum hydrocarbon (TPH) degradation was monitored in a field site with three vegetative treatment plots, and one control plot undergoing natural attenuation. Within each treatment, TPH concentrations were monitored at 20 locations over time to study the phytoremediation potential of the different vegetative treatments. For comparing the performance of these treatments in a quantitative manner, first-order kinetics were assumed to be applicable at the local scale. The degradation rates and the initial contaminant concentrations were treated as spatially correlated random fields. Field-scale behavior was evaluated based on temporal variations of the means and variances of concentrations. Our results indicate the importance of spatial variability for an accurate assessment of phytoremediation in the field. From the degradation rate constants and mean reduction in TPH, rye grass and St. Augustine grass appear to be superior to sorghum and the unvegetated control in reducing contaminant concentrations in the field.

Dispersion in porous media has been postulated as a mechanism that can increase particle transport, but the magnitude of increased transport predicted using analytical models of column experiments is a function of the boundary conditions chosen for the models. To date, insufficient attention has been paid to the implications of these boundary conditions for column or field simulations of particle transport. In order to assess the potential effect of particle dispersion on bacterial transport during bioremediation using bioaugmentation, we review the most frequently applied boundary conditions and models used in particle transport modeling. Using a finite-difference model written in MATLAB to simulate a column experiment, we demonstrate that for boundary conditions typically employed in short (∼10 cm) laboratory columns, hydrodynamic dispersion in such systems does not appreciably alter the calculation of bacterial stickiness or overall transport. However, dispersion is known to increase with distance, resulting in dispersivities on the order of one to several hundred meters over distances of 100 m. Using boundary conditions appropriate for field tests (a semi-infinite domain), and increasing dispersivity in proportion to transport distance, it was similarly demonstrated that dispersion would not appreciably increase the distance of particle travel (defined as the distance prior to a 2-log reduction in bacteria concentration) in field situations. Thus, it is concluded that dispersion can be neglected in calculating particle collision efficiencies in both laboratory and field experiments.

The Surface Water Treatment Rule and the Interim Enhanced Surface Water Treatment Rule classify some ground-water sources as “ground water under the direct influence of surface water” (GWUDI), and ground waters classified as GWUDI must follow the same treatment guidelines as surface-water sources. The microscopic particulate analysis (MPA) method is the most widely used method to identify GWUDI; however, this method has several deficiencies. An improved protocol for classifying ground-water sources as GWUDI is presented in this paper. The procedure consists of calculating the absolute risk of illness from waterborne Giardia and Cryptosporidium in the pumped water, accounting for the uncertainty in the surface-water concentrations, hydrogeology, straining characteristics, pathogen viability in ground water, and observations of Giardia and Cryptosporidium in samples of the pumped water. In cases where the risk of illness exceeds the acceptable risk, the ground water should be classified as GWUDI. Two examples are presented to illustrate the application of the proposed method.

A methodology and assessment of the thermal balance of an on-stream storm-water pond are documented for a test facility in Kingston, Ontario, Canada. An energy balance model was used to estimate average pond water temperature as a function of thermal energy stored in the pond. Major thermal energy fluxes considered within the pond system included radiant heating and cooling of the pond, evaporation, and thermal energy inflow from the upstream catchment and outflow from the pond. A thermal energy balance was achieved between thermal energy input and output measured over the field season. During dry-weather periods, pond temperature increased as a result of solar heating, and thermal energy input exceeded output. Conversely, during wet-weather periods, pond temperature decreased as a result of limited solar radiation and replacement of warm pond water by cool inflow water from the upstream catchment, and thermal energy output exceeded input. A series of discrete temperature surveys in the pond revealed that the location of the stationary temperature probe, used to measure pond temperature, resulted in an underestimation of the average pond temperature during dry- and wet-weather periods. However, this discrepancy was comparable to the difference between measured and modeled values, which further confirmed the model validity.

A numerical model to compute the free-surface flow hydrodynamics and stream temperature dynamics by solving the depth-averaged, 1D unsteady flow and heat transport equations is presented. The hydrodynamics model considers the effects of arbitrary stream geometry, variable slopes, variable flow regimes, and unsteady boundary conditions. The thermal transport model accounts for the effects of solar radiation, air temperature, relative humidity, cloud cover, wind speed, heat conduction between water and streambed, subsurface flow, and shading by riparian vegetation. The model is verified with measurements in a stream in an upland agricultural watershed located in Indiana. Diurnal variations in the streamflow and stream temperatures are highly transient. The proposed model predicted well the streamflow and stream temperatures that were measured every 15 min over 25 days. The results of this study demonstrate that the solar (shortwave) radiation and subsurface inflow are the most significant contributors to the stream heat budget.

This research was carried out to evaluate feasibility of using an electrokinetic technique to remove hydrophobic organic pollutants from soils, with the assistance of a cosolvent (n-butylamine, tetrahydrofuran, or acetone) added to the conducting fluid. The experiments were carried out on glacial till clay with phenanthrene as the test compound. Desorption equilibrium was investigated by batch tests. The electrokinetic experiments were conducted using a 19.1 cm long × 6.2 cm inside diameter column under controlled voltage. Water or 20% (volume) cosolvent solution was constantly supplied at the anode. The concentration of phenanthrene in the effluent collected at the cathode was monitored. Each experiment lasted for 100 to 145 days. Results showed that the presence of n-butylamine significantly enhanced the desorption and electrokinetic transport of phenanthrene; about 43% of the phenanthrene was removed after 127 days or 9 pore volumes. The effect of acetone was not as significant as butylamine. The effluent flow in the tetrahydrofuran experiments was minimal, and phenanthrene was not detected in the effluent. The use of water as the conducting solution did not cause observable phenanthrene migration.

Electrokinetic remediation is an emerging technique that can be used to remove metals from saturated or unsaturated soils. In unsaturated soils, control of the medium's water content is essential. Previously used electrode designs have caused detrimental soil wetting due to excess electroosmotic flow out of ceramic-encased anodes. We tested a method to reverse the electroosmotic flow at the anode by treating the ceramic casing with the cationic surfactant hexadecyltrimethylammonium (HDTMA). Laboratory tests showed the untreated ceramic had an electroosmotic permeability of 2.4 × 10⁻⁵ cm² V⁻¹ s⁻¹. Ceramic treated with HDTMA had an electroosmotic permeability of −1.3 × 10⁻⁵ cm² V⁻¹ s⁻¹. Under an applied electric potential, electroosmotic flow was reversed in the HDTMA-treated ceramic, indicating a reversed zeta potential due to formation of an HDTMA bilayer on the ceramic surface. Field tests conducted over a 6-month period showed negligible water loss from HDTMA-treated ceramic (0.03 L h⁻¹) compared to untreated ceramics (up to 6 L h⁻¹). The results indicated that a surfactant treatment to the anode ceramic casing can greatly improve the application of electrokinetics in unsaturated environments.

A bioelectrochemical method was applied to a submerged biofilter process to improve its nitrogen removal performance. Packed beds of activated carbon submerged into the aerobic tank of the submerged biofilter process were used as the electrodes and support for attached microbial growth. The experiments were conducted under different temperatures, dissolved oxygen (DO) concentrations and electric currents. The results showed that nitrification and denitrification rates were enhanced by supplying oxygen and hydrogen, respectively, from the substratum through electrolysis of water. The nitrification and denitrification rates were increased with increasing electric current. The effects of electric current on nitrification and denitrification rates were clearly shown under lower bulk liquid DO concentration. There was an optimum DO concentration to give the largest nitrogen removal rate in the bioelectrochemical compartment. The optimum DO concentration was in the range of 1.5–2.0 g/m³.

This paper describes the results of side-by-side full-scale aeration testing of a plug-flow process and a modified contact stabilization process incorporating an anaerobic selector at the wastewater treatment facility in Fredonia, N.Y. Over 40 tests were completed utilizing the off-gas technique during the 2-month investigation period (summer of 1995). Compared to the plug-flow process, the modified contact stabilization process with internal sludge recycle was shown to have higher α values and to require less blower energy consumption when the selector operation was properly controlled. Dissolved oxygen concentration, selector COD concentration, and internal recycle sludge levels were found to be critical parameters in the successful operation of the modified process. Higher internal recycle sludge levels allowed the plant to run at more stable operating conditions in terms of the oxygen transfer efficiency, α, and sludge volume index.

Drinking water utilities are facing increasing regulatory pressure to optimize treatment processes, particularly the filtration process that is relied on as a physical barrier to pathogen contamination. In the presence of high concentrations of natural organic matter, traditional coagulation processes can be inadequate to meet these future standards. In this study, a natural water with a high total organic carbon concentration (>6 mg/L) was used to compare one- and two-stage alum coagulation processes. The two-stage coagulation process involved the application of alum at two locations separated by approximately 60 s of detention time. Bench-scale experiments were conducted with various water quality conditions and verified at the pilot scale. The settled-water turbidity was 25% less using the two-stage process, and the steady-state filter effluent particle counts for the two-stage process were approximately 50% lower than for the single-stage process. More importantly, the two-stage process did not exhibit particle breakthrough during a typical filter run, whereas the one-stage process broke through 3 h before the hydraulic completion of the run. Additional bench-scale experiments allowed a mechanism for the improved process performance to be proposed.

Activated carbon cloths show interesting properties for the treatment of air that is loaded with volatile organic compounds. The industrial implementation of these new adsorbent materials requires the design of new reactors. Cylindrical pleated filters made from activated carbon cloths and polypropylene layers are studied using a laboratory pilot unit. Experimental pressure drops are measured as a function of the number of pleats. The effect of the overlapping of the pleats, due to the significant thickness of the medium, is shown to be very important. The use of two different parameters (the equivalent section and the equivalent thickness), corresponding to different approaches, is proposed to explain the evolution of pressure drops in such filters. Finally, a threshold criterion for determining a relevant number of pleats is proposed.

A simple predictive correlation for bed voidage in a three-phase fluidized bed biofilm reactor (FBBR) has been proposed in this study, based on a relationship between the effective up-flow superficial liquid velocity (U_L) and the corresponding up-flow superficial liquid velocity in a three-phase FBBR $\left({\mathit{U}}_{\mathit{L}}^{\mathit{\prime }}\right)\text{.}$ The experimental observations of bed voidage both for this study as well as for experimental data from a study published by Trinet et al. for a three-phase FBBR compared reasonably well with the model predictions developed in this study, with mean deviations of 0.26% and 7%, respectively. Further, published experimental observations of bed voidage in a two-phase FBBR agreed well with the model predictions, with mean deviations of 26.4%, 5.2%, and 9.7% for studies conducted by Mulcahy and Shieh, Ngian and Martin, and Setiadi, respectively, and with a mean deviation of 18.7% for the total published experimental data (297 points) of the above researchers.