Journal of Environmental Engineering

Although bioretention has been shown to remove or sequester a wide range of pollutants, relatively little study has been performed to evaluate its ability to sequester indicator bacteria. Two adjacent bioretention areas in Wilmington, North Carolina, were studied. The primary difference in the design of the two systems was soil depth. One bioretention cell was constructed with 25 cm of fill soil (Bioretention-S) and one with 60 cm of fill soil (Bioretention-D). The systems performed differently for indicator bacteria on the basis of multiple performance evaluation metrics. Bioretention-D showed concentration reductions of 70% and 89% for E. coli and enterococci, respectively. Effluent concentrations from Bioretention-D compared well to US EPA target values and other studies in literature. Conversely, Bioretention-S showed concentration “reductions” of $-119\%$ and $-102\%$ for E. coli and enterococci, respectively. Effluent concentrations from Bioretention-S were substantially higher than USEPA target values and other studies in literature. Multiple factors were evaluated to determine the cause of performance differences between the two cells. The 25 cm of fill soil in Bioretention-S exhibited poorer runoff detention, likely because of higher soil water flux and decreased contact time relative to Bioretention-D. These differences seemingly led to diminished indicator bacteria sequestration. The results of this study suggest soil depth and hydraulic loading are important design parameters for bioretention implemented to sequester microbes.

In this study, laboratory experiments were carried out to investigate the environmental behavior of three selected phenolic compounds [2, 4-Dichlorophenol (DCP), 2, 4-Dinitrophenol (DNP) and 2, 4-Dimethylphenol (DMP)] in the aquatic environment. The results of batch sunlight photolysis experiments showed that the three phenolic compounds were photodegraded relatively easily (i.e., half-life $<151\mathrm{h}$) in June, whereas these compounds were relatively stable against sunlight in September (i.e., half-life $<330\mathrm{h}$). The results of batch biodegradation experiments using river water suggested relatively slow biodegradation (i.e., half-life $<2,310\mathrm{h}$ in June, half-life $<1,155\mathrm{h}$ in September, and half-life $<3,466\mathrm{h}$ in January) of three phenolic compounds, and the rate constant was dependent on sampling time. Freundlich model and the pseudo second-order model can describe the sorption behaviors of phenolic compounds on sediments very well. The intraparticle diffusion model for adsorption was also investigated and compared with the pseudomodel to identify sorption mechanism.

The removal of arsenic (As) by iron (Fe)-manganese (Mn) binary oxide–coated diatomite (FMBO-diatomite) and the transformation and transportation of As, Fe, and Mn in a fixed-bed reactor were investigated in this pilot-scale field study (arsenite $[\mathrm{As}(\mathrm{III})]=50–80\mu \mathrm{g}/\mathrm{L}$, arsenate $[\mathrm{As}(\mathrm{V})]=40–60\mu \mathrm{g}/\mathrm{L}$). Before regeneration, the effluent As(tot) was reduced to less than $10\mu \mathrm{g}/\mathrm{L}$ with no residual As(III). FMBO-diatomite exhibited elevated adsorptive capability of As after in situ regenerations with breakthrough bed volumes increasing from 550 to 760 bed volumes after three cycles of operation. The residual concentrations of As(III) and As(V) at different layers along the adsorption column indicated the oxidation of As(III) and the As transportation during the filter layer. Additionally, FMBO-diatomite could efficiently remove the turbidity, Fe and Mn in influents, without causing secondary pollution. Acid dilution analysis showed more Fe content and less Mn content after three cycles through the FMBO-diatomite upper layer. This was attributable to the oxidation of As(III) and Fe(II) by manganese dioxide (${\mathrm{MnO}}_2$) within FMBO and the transportation of Mn(II) thereafter. The backwashing procedure led to a size classification of adsorbents. However, the FMBO-diatomite also showed increased size diameter after three cycles of operation because of the in situ regeneration.

There is significant concern in water utilities regarding the permeation of petroleum-based hydrocarbons from contaminated sites through potable drinking-water pipes made of polyvinyl chloride (PVC). Previous studies have addressed the impacts of hydrocarbons on the pipe materials, but the impacts of the hydrocarbons on gasket materials joining and sealing the pipes have not been investigated. This study investigated the permeation of benzene, toluene, ethylbenzene, and xylene (BTEX) compounds through PVC pipe joints equipped with 50-mm styrene-butadiene rubber (SBR) and nitrile-butadiene rubber (NBR) Rieber gaskets exposed to various levels of gasoline contamination. Results of permeation experiments with premium gasoline indicated that permeation occurred only through the gaskets and not through the intact pipe materials. For experiments with premium gasoline, benzene was the first compound detected with a breakthrough time of approximately 21 days. However, for experiments with various gasoline concentrations in water, including water saturated with gasoline, no significant amounts of BTEX compounds were detected in the pipe water after 9 months of exposure. The steady-state permeation rates of benzene when exposed to premium gasoline were found to be $0.73\pm 0.29$ and $0.38\pm 0.32\mathrm{mg}/\mathrm{joint}/\mathrm{day}$ for SBR and NBR gaskets, respectively. The corresponding diffusion coefficients of benzene in SBR and NBR gaskets (which have not been reported before) were determined to be $1.1\times {10}^{-7}$ and $6.0\times {10}^{-8}{\mathrm{cm}}^2/\mathrm{s}$, respectively, by curve fitting the permeation data with a diffusion model. Although PVC pipe joints were found to be permeated by free-product gasoline, water flow in pipes would dilute the benzene in drinking water to below the maximum contamination level (MCL) of $5\mu \mathrm{g}/\mathrm{L}$.

The removal of various phenoxyalkanoic acid herbicides as single solutes (2,4-D; 2,4-DP; 2,4-DB; 2,4,5-T; MCPA) from aqueous solutions were studied by using biodegradable poly($\epsilon$-caprolactone) (PCL). A comparative study was achieved by using powdered and granular activated carbon as an adsorbent in a batch system. The equilibrium was studied at 25, 35, and 45°C. The results were correlated by using Freundlich Langmuir, Temkin, Redlich-Peterson, and Sips equations. To determine the best fit isotherm, three error analysis methods were used to evaluate the data: correlation coefficient ($R^2$), residual root-mean-square error, and normalized deviation ($\Delta Q$). The results show that all the isotherm models provide an appropriate fitting to the experimental data for the adsorption onto PCL. Adsorption of the studied herbicides onto granular activated carbon, powdered activated carbon, and poly($\epsilon$-caprolactone) decreased in the order of $2,4\mathrm{\text{-}}\mathrm{DB}>2,4,5\mathrm{\text{-}}\mathrm{T}>2,4\mathrm{\text{-}}\mathrm{DP}>2,4\mathrm{\text{-}}\mathrm{D}>\mathrm{MCPA}$. The thermodynamic parameters ($\Delta G°$, $\Delta H°$, $\Delta S°$) for the adsorption process were calculated, and the results suggest that the nature of adsorption is endothermic and the process is spontaneous and favorable. Overall the results show that biodegradable poly($\epsilon$-caprolactone) is a promising alternative adsorbent to be used in water treatment.

This paper reports the adsorptive removal of hydroquinone (HQ) from aqueous solution by granulated activated carbon (GAC). Physicochemical properties including surface area and surface texture of the GAC, before and after HQ adsorption onto GAC, were analysed by using X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX), and Fourier transform infrared spectroscopy (FTIR). The adsorption follows pseudo-second-order kinetics. An intraparticle diffusion study reveals that the pore diffusion is not the only rate-limiting step. The effective diffusion coefficient of HQ was of the order of $3.19\times {10}^{-10}{\mathrm{m}}^2/\mathrm{s}$. Equilibrium isotherm data were generated for HQ solutions over an initial concentration range of $0.18–9.08\mathrm{mmol}/\mathrm{L}$ for the GAC dosage of $10\mathrm{g}/\mathrm{L}$ at temperatures of 288, 303, and 318 K. Equilibrium data were well represented by the Temkin isotherm model. An increase in the temperature increases adsorption. The entropy change ($\Delta S_0$), heat of adsorption ($\Delta H_0$), and Gibb’s free energy ($\Delta G_0$) were determined. Elution of HQ from the loaded HQ showed that the thermal desorption may be a better option over that of solvent desorption. GAC worked well for at least five adsorption-desorption cycles, with continuous decrease in adsorption efficiency after each thermal desorption. Owing to its heating value, spent-GAC can be used as cofuel in the boiler-furnace.

Polyaniline (PANI) and humic acid (HA) composite sorbents were prepared by adding different amount of HA into the chemical oxidation process of PANI. The sorbents were characterized by scanning electron microscopy (SEM), Fourier transform infrared (FT-IR) spectroscopy, and ${\mathrm{N}}_2$ adsorption and desorption. Batch adsorption results showed that the sorbent PANI/HA-1 had the best affinity to Cr(VI) in aqueous solution. The adsorption kinetics results of Cr(VI) indicated that the adsorption reached equilibrium within 120 min, and adsorption rate could be described by pseudo second order kinetics. Sorption of Cr(VI) onto PANI/HA-1 fitted the Freundlich adsorption model better than the Langmuir adsorption model. The results of zeta potential measurement and batch experiments showed that the PANI/HA-1 was stable in the pH range of 4–7. The desorption results indicated NaOH eluent had better desorption rate than HCl eluent. The electrochemical-controlled desorption result showed that the loaded Cr(VI) on the PANI/HA-1 could be removed from the matrix, although the desorption rate was not as high as using the eluents. An adsorption mechanism was proposed on the basis of X-ray photoelectron spectroscopy (XPS) and zeta potential measurement results.

Time series of total suspended solids, and total and dissolved copper, zinc, and lead concentrations were compared between milled hot-mix asphalt (HMA) roadway sections ($n=6$) and unaltered HMA roadway sections ($n=6$). Milled and unaltered road surface runoff water quality parameter values were within the ranges reported previously for unaltered rural and urban roadways. Event mean concentrations of suspended solids and lead were greater for milled roadway surfaces than for unaltered surfaces. No differences in copper and zinc concentrations were observed between road treatments. Contaminant mobilization was flow-driven, with incomplete wash-off of roadway constituents during the short-duration convective storm events. Storm precipitation intensity was not an indicator of storm-water quality. The presence of lead in runoff from milled roadway sections in this study indicated a road side soil source of the solids resulting from overland flow from adjacent land surfaces onto the roadway. Future milling operations should proceed with consideration not to create new hydraulic connections with the adjacent roadside.

Environmental systems modeling has always been at the core of gaining insight into the world. Most environmental systems behave very differently from fairly predictable systems because of nonlinearity in approach and behavior. This paper discusses the problem that exists in effectively modeling a multicriterion and nonlinear parametric system. A cyclical hierarchical model was proposed that allows the user to model the effects of various environmental parameters, which not only propagate as forward processes, but also may have dependencies and carryover effects on other primary set of parameters. This is an adaptation from ecological hierarchies for development of parametric modules in an environmental system. This effect was captured by a set of forward and backward carryover functions that are to be generated as process-specific and time-specific functions. The user expresses control over the effect of each parameter on the other and in terms of iterations that are performed within the cyclical hierarchies and the value of the forward and backward functions that need to be developed. This was integrated into the core of an environmental decision support system (EDSS), which prioritizes a parameter set on the basis of analytical hierarchy process (AHP) to establish a global hierarchy that is then studied in perspective to cyclical hierarchies and nodal relations that exist in such hierarchical forms. The study concerns itself with an urban river system, which involves a multitude of such multivariate parameters. Iterations were performed to see the variations in the global hierarchy. The parameterized form can be used to generate criterion-based output for the cyclical hierarchy sets used, and the change may be adapted pertaining to EDSS output.

To understand wastewater quality transformations in the sewer system, a full-scale study focused on the characteristics of the influent wastewater in a middle-sized wastewater-treatment plant receiving only municipal sewage. The objective was to identify conditions in which prefermentation prior to the biological nutrient removal (BNR) process would be useful. Prefermentation aims at increasing the fraction of readily biodegradable organic matter in the wastewater. A correlation between weather conditions and the presence of readily biodegradable organic matter in the feed wastewater was found, but contrary to the expectations, wet weather periods deteriorated the quality of the feed water more than the cold period. Moreover, influent load showed seasonal variation which changes in discharged wastewater could not explain. During the warm water period, significant biological oxidation and nitrogen removal took place in the sewer system, but the plant influent water contained readily biodegradable organic matter and potential for its formation during the warmest period. It was concluded that prefermentation would be mostly needed during the wet weather period. In addition, the study offered valuable information about the seasonal behavior of the sewer network.

In a subtropical area, a downflow-upflow biological aerated filter (D-UBAF) was used in the pretreatment of raw water containing seasonal high levels of ammonia nitrogen (${\mathrm{NH}}_3\mathrm{\text{-}}\mathrm{N}$) and organic pollutants. Operating parameters for low-flow periods were as follows: filtration rate of $8\mathrm{m}/\mathrm{h}$ and gas-water ratio of $0.5:1.5$. In terms of water quality, the average concentrations of ${\mathrm{NH}}_3\mathrm{\text{-}}\mathrm{N}$ and permanganate index (${\mathrm{COD}}_{\mathrm{Mn}}$) and the turbidity of the raw water were $5.02–9.45\mathrm{mg}/\mathrm{L}$, $5.79–10.1\mathrm{mg}/\mathrm{L}$, and 19.7–63.1 NTU, respectively. The corresponding indexes of the D-UBAF effluent were $0.30–0.96\mathrm{mg}/\mathrm{L}$, $3.24–5.85\mathrm{mg}/\mathrm{L}$, and 3.83–19.9 NTU, respectively. The removal efficiencies were 90.3%, 42.7%, and 66.3%, respectively. Operating parameters for high-flow periods were as follows: filtration rate of $12\mathrm{m}/\mathrm{h}$ and gas-water ratio of $0.3:0.5$. In terms of water quality, the average concentrations of ${\mathrm{NH}}_3\mathrm{\text{-}}\mathrm{N}$ and ${\mathrm{COD}}_{\mathrm{Mn}}$ and the turbidity of the raw water were $2.19–3.41\mathrm{mg}/\mathrm{L}$, $5.30–7.56\mathrm{mg}/\mathrm{L}$, and 27.3–40.1 NTU, respectively. The corresponding indexes of the D-UBAF effluent were $0.18–0.41\mathrm{mg}/\mathrm{L}$, $2.87–4.50\mathrm{mg}/\mathrm{L}$, and 5.43–16.8 NTU, respectively. The removal efficiencies were 89.3%, 40.2%, and 65.0%, respectively. The D-UBAF with one DBAF and two UBAFs constituted a “

A model was developed that simulates the operation and process costs of recycled cooling water treatment using an advanced softening process called ultrahigh lime with aluminum softening (UHLA). The model was coded in MATLAB and can describe multicomponent removal and predict the chemical doses and process costs for any influent and effluent water quality. PHREEQC was integrated with MATLAB to calculate effluent concentrations from treatment processes. Material balance equations for the constituents in the water were constructed to relate the concentrations and flow rates in the system. Using only the makeup water chemical composition, flow rate in the cooling system, and required cooling water quality, the model is able to predict the chemical doses, treatment system flows, and total process costs. Alternatively, the model can be used to calculate cooling water quality given flow through the treatment system and its chemical doses. The model has the ability to simulate other applications that use UHLA such as brine treatment and pretreatment before membranes.

Waterborne disease outbreaks in U.S. swimming pools have been increasing in recent years, and the majority of the outbreaks are caused by Cryptosporidium oocysts. This research project evaluates sand filtration for swimming pools without and with an amendment, perlite. The evaluation was formed on the basis of removing 5-μm polystyrene microspheres, as a surrogate for Cryptosporidium oocyst, from simulated pool water stored in a 757 L (200 gal) tank. The results showed that a sand filter, without perlite, was not efficient at removing Cryptosporidium-sized particles (i.e., the 5-μm microspheres), with removals averaging 19% (or 0.09 log). The sand filter with a thin layer ($1.2\mathrm{kg}/{\mathrm{m}}^2$ or $0.25\mathrm{lbs}/{\mathrm{ft}}^2$) of perlite media on top demonstrated removals of 98% (or 1.8 log). The filtration rate was maintained at $49\mathrm{m}/\mathrm{h}$ ($20\mathrm{gpm}/{\mathrm{ft}}^2$) in all experiments. These results indicate that perlite may hold promise in reducing the likelihood of outbreaks of cryptosporidiosis associated with swimming pools.

The coagulation process, which involves many complex physical and chemical phenomena, is one of the most important stages in water-treatment plants. The coagulant dosage rate is nonlinearly correlated to raw water characteristics such as turbidity, conductivity, and pH. The coagulation reaction is hard or even impossible to control satisfactorily by conventional methods. The coagulant dosage has typically been determined through the jar test, which requires a long experiment time in a field-water-treatment plant. Modeling can be used to overcome these limitations. In this study, a model for the approximation of coagulant dosage rates in water-treatment plants in Algeria has been developed using artificial neural network (ANN) techniques. Two different ANN techniques, the generalized regression neural network (GRNN) and the radial-basis function neural network (RBFNN), were tested for this purpose. The trained GRNN model outperforms the corresponding RBFNN model.

Disinfection of bacterial viruses MS2, $\mathrm{Q}\beta$, and $\phi \mathrm{X}174$ in water was successfully carried out using deep ultraviolet light-emitting diodes (DUV-LEDs) operated at 280 nm and 255 nm. It was shown that the 280-nm DUV-LEDs are much more suitable than the 255-nm DUV-LEDs for the disinfection of water because, even though the disinfection efficiency of the former DUV-LEDs is only half that of the latter ones, the production of high-power 280-nm DUV-LEDs is much easier than that of the 255-nm ones because the external quantum efficiency of the 280-nm DUV-LEDs is more than ten times that of the 255-nm DUV-LEDs. We propose the use of vertical emitting DUV-LEDs for the design of an efficient DUV-LED disinfection chamber. This disinfection technique using DUV-LEDs will open new avenues for water purification.