Journal of Environmental Engineering

An SBR (sequencing batch reactor) system was evaluated for nutrient removal. The system is capable of removing 95% of influent ${\mathrm{PO}}^{\mathit{3}{\mathit{-}}_{\mathit{4}}}$, or from 6.7 to 0.4 mg P/L, with the addition of acetate of 120–150 mg COD/L in the feed solution (primary effluent). Nitrification was also achieved within the preset aeration cycle time in reducing the effluent ammonia level from 16.3–19.8 mg N/L to 0.2–0.3 mg N/L. However, denitrification was incomplete due to a slower endogenous nitrate respiration rate in the idle period, resulting in an effluent nitrate level of 7–8 mg N/L. A linear version of the ASM2 (Activated Sludge Model No. 2) was developed to model the performance of an SBR system for nutrient removal. The developed model appropriately predicts the dynamic behavior of the SBR system with respect to phosphate release/uptake, nitrification, ammonification, and denitrification. Compared with the full ASM2, the calibration of model parameters and model simulation require less computational time for practical implementation of the linear model into a process control system for the SBR.

Using a linear model, an optimization scheme for a sequencing batch reactor (SBR) system for phosphorus removal was investigated. The objective was to minimize energy consumption by reducing the aeration cycle time (t_air), while meeting the permit requirement (monthly average ${\mathrm{PO}}^{\mathit{3}{\mathit{-}}_{\mathit{4}}}$ of 0.5 mg P/L). Based on the model prediction and error feedback information, the proposed scheme controlled the SBR system well both in the simulation and the real application by adjusting the t_air to meet the effluent ${\mathrm{PO}}^{\mathit{3}{\mathit{-}}_{\mathit{4}}}$ constraint. Mismatch between the model prediction and the measured data was compensated for. In the simulation, the average aeration cycle time was calculated to be 2.8 h, while in the real system it was 3.5 h. The actual optimized system provided excellent removal of phosphorus, COD, and ammonia with efficiencies of 93% (7.4 to 0.5 mg P/L), 90% (420 to 43 mg COD/L), and 98% (22.1 to 0.4 mg N/L), respectively. However, the effluent nitrate concentrations were relatively high (10 mg N/L), due to a slower endogenous nitrate respiration rate.

The anaerobic-anoxic-nitrification (A₂N) process has been proved as a two-sludge system for denitrifying dephosphatation, based on denitrifying phosphorus-removing bacteria. Optimization of biological processes traditionally occurs by pilot-scale experiments. However, mathematical modeling may minimize pilot-scale experiments although it cannot completely predict or replace a reality. For this reason, an activated sludge model in AQUASIM 2.0 simulated the A₂N process. The used model originated from the Delft model for bio-P removal and activated sludge model No. 2 for COD and N removal. The model was first verified on its description of a laboratory A₂N system. The following simulations confirmed the experimental conclusion that an influent biodegradable COD:N = 3.4 mg COD/mg N (biodegradable COD:P = 26.7 mg COD/mg P) was an optimal ratio to realize full nitrification and complete denitrifying dephosphatation. Simulations also demonstrated the behavior of the A₂N process under the different COD/N, COD/P, solids retention time (days), temperature, and split ratios. Simulations indicated that the COD/${\mathrm{NO}}^{{\mathit{-}}_{\mathit{3}}}$ was an actual governing factor in denitrifying dephosphatation and that in practice a short aerobic period following the anoxic phase was needed to obtain a stable system.

A mathematical model was developed to describe phenol degradation with sulfate reduction in an anaerobic biofilm process. The model incorporates the mechanisms of diffusive mass transport and Monod kinetics. The model was solved using a combination of the orthogonal collocation method and Gear's method. A pilot-scale column reactor was conducted to verify the model. The batch kinetic tests were independently conducted to determine nine biokinetic coefficients used in the model while shear loss and initial thickness of the biofilm were assumed so that the model simulated the substrate concentration results very well. The removal efficiencies for phenol and sulfate are 98 and 88%, respectively. At a steady-state condition, the experimental data of phenol and sulfate concentrations were higher than those obtained from the model. The reason was that the effect of shear loss became significant as the biofilm grew thicker. The higher shear loss resulted in the more-suspended biomass. The suspended biomass decomposed and released soluble microbial products which increased the substrate effluent concentrations. The procedures presented in this paper could be employed for the design of anaerobic biofilm reactor systems for the biodegradation of multiple substrates.

A computer model is used to investigate the microbial uptake of oxygen and a carbon-source substrate for biofilms growing on gas-permeable, hollow-fiber membranes and impermeable solid supports of similar geometry. Substrate and oxygen fluxes are predicted for different biofilm thicknesses as a function of fluid velocity and substrate concentration. Under conditions of oxygen limitation, low water velocities, and moderate to high bulk liquid substrate concentration, the membranes have a clear advantage and outperform solid supports. This improvement in performance stems from the ability of the membrane to deliver high oxygen concentrations (8–20 mg/L) directly to the biofilm, whereas it is difficult to maintain bulk dissolved oxygen concentrations much above 4 mg/L in wastewater treatment. The growth of an active biofilm can actually increase the flux of oxygen across the membrane dramatically; however, the presence of a biofilm always reduces the ability of a membrane to oxygenate the surrounding wastewater. This drop in oxygen transfer performance is caused by the fact that the active biofilm consumes oxygen and impedes diffusion of the oxygen into the bulk water. In thick biofilms the oxygen flux can drop to zero so that the external regions of the biofilm and the external wastewater become anaerobic. This may cause some operating problems, but it may also facilitate nitrification-denitrification. Additional aeration of the external wastewater could improve biofilm performance and assist in controlling biofilm growth.

A 3.5-year pilot test of air sparging/soil vapor extraction (AS/SVE) was carried out to determine whether the heterogeneity of the Piedmont saprolite would allow adequate soil vapor velocities and effective vapor-phase extraction rates for petroleum hydrocarbon (PHC) remediation. The objectives were to compare: (1) the effectiveness of pulsed SVE versus pulsed AS/SVE operation; (2) benzene, toluene, ethylbenzene, and xylene (BTEX) versus PHC removal; and (3) biological versus physical removal of PHC. Stack exhaust gas, SVE wells, and soil vapor probes were monitored for total combustible hydrocarbons (TCH), BTEX, O₂, CO₂, temperature, and flow rate using handheld meters and gas chromatography. The majority of contaminant recovered was removed from the vadose zone via SVE. BTEX and TCH were both effectively removed from the more-permeable and highly contaminated unsaturated-saturated zone interface. Bioremediation accounted for 23% of total removal. Overall, the AS/SVE system physically removed an estimated 18,800 kg of PHC and 5,300 g of BTEX with an average rate of 70 kg d⁻¹ and 0.4 g d⁻¹, respectively, which was consistent with other AS/SVE studies in sandy media.

Lateral migration of landfill gases in soils surrounding old (closed) municipal landfills can lead to explosion hazards and damage to vegetation. Landfill gas production and migration is controlled by microbial activity and soil physical properties such as gas (air) permeability, gas diffusivity, and atmospheric pressure variations. Gas diffusivity and air permeability were measured on undisturbed samples collected at an agricultural field adjacent to an old Danish municipal landfill. Empirical expressions for predicting gas diffusivity and air permeability from soil-water content were fitted to the data. An empirical expression for calculating methane oxidation rate as a function of soil temperature was developed using site-specific measurements of methane oxidation rate. The transport and degradation parameter expressions were incorporated into a numerical model for simulating landfill gas transport, using soil-water content, temperature, and soil organic matter content as measured input parameters (functions of time and soil depth). Previous measurements of landfill gas (CO₂ and CH₄) concentrations and fluxes in the agricultural field soil conducted over a period of one year were used to calibrate the gas transport model using methane oxidation rate, landfill gas pressure, and wind-induced dispersion as fitting parameters. The model was subsequently tested against independent concentration and flux data (not used in the model calibration). This yielded a prediction accuracy similar to that found during the calibration. The model was used to evaluate the sensitivity of landfill gas concentration and flux with respect to a set of governing parameters. The results of the sensitivity analysis indicated that landfill gas migration in the field soil was most sensitive to air permeability and soil-water content. Methane oxidation rate and atmospheric pressure variations had some effects, especially on the gas flux, whereas gas diffusion was not important.

Conventional clay liners, which are widely used as hydraulic barriers to water, have been shown to be adversely affected by organic fluids. However, the addition of quaternary amines into bentonite greatly enhances its compatibility with organic fluids and thus allows the clay barrier technology to be extended to the treatment of organic contaminants. In this study, an organically modified clay was studied for use as a secondary containment for gasoline underground storage tanks. Free swelling and hydraulic conductivity tests were performed on organoclay, pure bentonite, and natural soils. Results show that organoclay has a large swelling capacity in gasoline, whereas bentonite and natural soils shrink when immersed in gasoline. The hydraulic conductivities of bentonite and natural soils to gasoline are 2–5 orders of magnitude higher than that to water. In contrast, the hydraulic conductivity decreases by 2 orders of magnitude for organoclay and this low value can be maintained even under freeze-thaw and dry-wet cycles.

Horizontal infiltration experiments were performed to validate a plug flow model that minimizes the number of parameters that must be measured. Water and silicone oil at three different viscosities were infiltrated into glass beads, desert alluvium, and silica powder. Experiments were also performed with negative inlet heads on air-dried silica powder, and with water and oil infiltrating into initially water moist silica powder. Comparisons between the data and model were favorable in most cases, with predictions usually within 40% of the measured data. The model is extended to a line source and small areal source at the ground surface to analytically predict the shape of two-dimensional wetting fronts. Furthermore, a plug flow model for constant flux infiltration agrees well with field data and suggests that the proposed model for a constant-head boundary condition can be effectively used to predict wetting front movement at heterogeneous field sites if averaged parameter values are used.

Sulfur hexafluoride (SF₆) was injected into a 9 km reach of the Santa Ana River in Orange County, CA, over a period of two weeks. The entire flow of this river, which averaged 2.8 m³ s⁻¹, percolated into the ground in the field area. The tracer was monitored at wells near the river to determine subsurface flow patterns and flow times with an accuracy much greater than could be achieved using numerical simulations of ground-water flow. During the experiment, SF₆ effectively tagged 3.7 × 10⁶ m³ of water. The tracer plume was mapped in the subsurface for 18 months and indicates that linear ground-water velocities averaged about 2 km year⁻¹. The tracer reached two wells adjacent to the river (about 200 m away) within three weeks, giving evidence that SF₆ was not retarded significantly relative to the ground-water flow. This is in agreement with previous laboratory experiments.

This study presents the results of batch experiments investigating selenite reduction by an enriched anaerobic mixed culture as a function of several environmental factors, including pH, temperature, different electron donors and acceptors, as well as initial selenite concentrations. The initial selenite reduction was a zero-order reaction and was inhibited at higher selenite concentrations (>33 mg Se/L). The optimal temperature/pH for microbial reduction occurred at 30°C and pH 7.2. Selenite reduction was affected in the presence of high concentrations of sulfate [45 times the Se(IV) molar concentration], significantly affected by nitrate [105 times the Se(IV) molar concentration], and completely inhibited by chromate and oxygen. Ethanol was the preferred carbon source for selenite removal, followed by acetate, citrate, lactate, and glucose. The selenite-acclimated culture also reduces selenate without any lag period.

The dynamics of a self-inoculated upflow anaerobic sludge blanket (UASB) reactor treating raw sewage has been studied. The results suggest two types of start-up periods. During the first 6 weeks of operation, the reactor achieved a removal efficiency of up to 65% of total chemical oxygen demand (COD) and 73% of suspended solids. During this primary start-up period, the removal of organic matter was mainly due to interception in the reactor. The second start-up period expressed by biogas release was reached after 14 weeks. An intermediary period between these two start-up periods was characterized by a slight drop of pH and increase of volatile fatty acid in the effluent. The long time to reach the second start-up period seemed to be responsible for the large discrepancy between removed COD and recovered biogas (107 mL/g COD removed). It is thus suggested that process optimization should focus on shortening the time needed to reach the second start-up period. A model that represents a possible description of the process dynamics of self-inoculated UASB reactors is proposed.