Journal of Environmental Engineering

Granular media are increasingly utilized for control of phosphorus (P), frequently indexed as total P (TP), which represents the sum of particulate (PP) and dissolved P (DP) fractions. The fate of PP notwithstanding, when such media are deployed for control of DP, the mass transfer kinetics from an aqueous matrix to the media is important for design and modeling tools. In this paper, the mass transfer of DP for a suite of common media are monitored and modeled with a focus on comparing the behavior of Al-oxide coated media (AOCM) subjected to synthetic and actual runoff (wet weather) as well as wastewater (dry weather) matrices. The media employed were ${\mathrm{AOCM}}_c$ (clay based), ${\mathrm{AOCM}}_p$ (pumice based), and ${\mathrm{AOCM}}_{\mathrm{pcc}}$ (Portland cement concrete based), and their corresponding media (UCM) substrates. A 2nd-order potential driving model is utilized for the overall total phosphorus (TP) transfer rate from aqueous solution to AOCM, whereas an intramedia diffusion model illustrates diffusion rates. In comparison with AOCM forms that demonstrate similar and rapid kinetics (lower for ${\mathrm{AOCM}}_p$), other media exhibited large variability, decreasing in order from Fe-coated perlite, activated alumina, expanded shale, and bioretention media. In contrast, Zeolite-perlite-GAC (granular activated carbon) and tire-crumb (Black and Gold, B&G) displayed net adsorption not significantly different from 0 ($p=0.05$). Comparing synthetic and actual runoff matrices demonstrated that actual runoff produced slower kinetics for ${\mathrm{AOCM}}_c$ and ${\mathrm{AOCM}}_p$, primarily attributable to the presence of competitive ions (${\mathrm{SO}}_4^{2-}$), whereas elevated ${\mathrm{Ca}}^{2+}$ at the alkaline surface of ${\mathrm{AOCM}}_{\mathrm{pcc}}$ offset the effect of ${\mathrm{SO}}_4^{2-}$ through surface precipitation.

A hydrodynamic separator (HS) is commonly used as a preliminary unit operation for separation of coarser particulate matter (PM) and PM-associated constituents transported by urban rainfall-runoff. Although the small spatial footprint of an HS is advantageous for integration into sewer (storm or combined) or drainage systems, the small footprint also concentrates flow energy. In many HS units, where PM sludge is not isolated or in units that are not maintained (cleaned), washout of previously separated PM sludge can result in a net export of PM. This study analyzes PM washout from a baffled HS as a function of steady-flow rates and particle-size distributions (PSDs) guided by a physical modeling framework utilized for scour certification. The HS hydraulic signature is determined through velocity measurements and a residence time distribution (RTD) analysis. As a function of surface overflow rate (SOR), the predeposited PM washout rate ranges from 0.4 to $13.3\mathrm{g}/\min$ for fine PM (sandy silt, SM $<75\mathrm{\mu m}$), from 0.3 to $4.9\mathrm{g}/\min$ for moderately coarser PM (SM II,$<100\mathrm{\mu m}$), and 0.2 to $3.1\mathrm{g}/\min$ for coarser PM (SM III,$<1000\mathrm{\mu m}$) as SOR ranges from $47\mathrm{L}/\min -{\mathrm{m}}^2$ ($0.91\mathrm{L}/\mathrm{s}$) to $581\mathrm{L}/\min -{\mathrm{m}}^2$ ($11.31\mathrm{L}/\mathrm{s}$). The washout rate decreases with time for each heterodisperse PM gradation, and the evolution of HS washout is modeled as a first-order exponential analogous to urban washoff models. Washout from the HS is modeled as a function of the densimetric Froude number and SOR. Results indicate that washout based on a Shields approach cannot be applied to HS units.

Nanoscale iron particles (NIPs) have been studied during the last few years for their potential for remediation of groundwater and soils contaminated by chlorinated hydrocarbons, pesticides, heavy metals, and some inorganic anions. However, the effectiveness of in situ remediation largely depends on quick and uniform distribution of NIPs into the contaminated subsurface zones. Previous studies have shown that transport of bare NIPs in subsurface soils is very limited and surface modification by dispersants such as polymers and surfactants is necessary. The objective of this study was to investigate eight different dispersants at different concentrations to determine their ability to modify the surface characteristics and increase the stability of the NIP suspension, thereby minimizing agglomeration and sedimentation of the NIPs. The studied dispersants and their tested concentration ranges [as % of NIP suspension, weight-to-weight ratio (w/w)] were as follows: aluminum lactate (Al-lactate; 2–15%), sodium lactate (Na-lactate; 6–12%), ethyl lactate (EL; 6–12%), aspartic acid (ASP; 2–8%), polyacrylic acid (PAA; 2–8%), 2-hydroxypropyl-$\beta$-cyclodextrin (1–4%), $\beta$-cyclodextrin ($\beta$-CD; 1–4%), and methyl-cyclodextrin (MCD; 1–4%). Zeta potential measurements and column experiments were performed on NIP-dispersant suspensions. Results showed that the zeta potential of bare NIPs was $41.7\pm 2.3\mathrm{mV}$. The influence of the dispersants was found to vary significantly depending on the chemical nature of the dispersant and the electrical charge of the ions in solution. Aluminum lactate released ${\mathrm{Al}}^{3+}$ into the solution, resulting in a reduction of the NIP zeta potential from $37.7\pm 1.8\mathrm{mV}$ at 2% concentration to $9.5\pm 0.7\mathrm{mV}$ at 15% concentration. Finally, cyclodextrins showed less influence on the zeta potential of NIPs with a slight increase (up to $49.3\pm 3.2\mathrm{mV}$). The results depict that a higher percentage of NIPs modified with aluminum lactate eluted from the soil. A 10% concentration of aluminum lactate exhibited the highest (93%) elution of the modified iron from the soil media. Aspartic acid and ethyl lactate showed less iron eluted (35–40%). The column tests demonstrated that aluminum or sodium lactate can increase the stability of the NIP suspensions and enhance their transport in subsurface soils. Furthermore, these dispersants are environmentally friendly, relatively inexpensive, and practical to use.

In this study, a mass balance model is developed to improve the understanding of phosphorus fluxes (P fluxes) between sediment and water in a river that is strongly influenced by low-head dams and discharges from wastewater-treatment plants (WWTPs). Before this study, recommendations had been made for dredging the river without a complete understanding of the causes and effects of sediment phosphorus flux. This study changed the way the river dynamics are understood and will provide a basis for watershed management and permitting decisions into the future. The analysis presented focuses on sorption processes at the sediment-water interface as a function of the diagenesis flux rate into the sediment, reactions within the sediments with partitioning between dissolved and particulate forms, and the inorganic phosphorus concentration in the overlying waters. The sediment is modeled as two layers with a thin (0.8 mm) surface aerobic layer and an underlying, thicker (10 cm) anaerobic layer. The phosphorus settling and diagenesis flux and the initial phosphorus concentrations in the sediment and water column are varied to simulate seasonal fluctuations and evaluate the effectiveness of different river management scenarios aimed at reducing phosphorus concentrations. This study shows the importance of including such a microanalysis within water-quality studies to enhance overall system understanding when making water management decisions.

Diagenesis of particulate organic matter in sediments can have important effects on the dissolved oxygen (DO) resources of lakes and delay potential benefits from reductions in phosphorus (P) loading. A sediment diagenesis model of intermediate complexity, $\mathrm{Sed}2K_s$, is developed and tested for Onondaga Lake, New York, and linked to a validated one-dimensional mechanistic hydrodynamic/stratification model (UFILS4). Modeling was supported by detailed characterization of the pelagic sediments and long-term monitoring of the lake’s water column that included transformation from hypereutrophy to upper mesotrophy and coupled improvements in hypolimnetic DO. $\mathrm{Sed}2K_s$ performed well in representing the observed decreasing trajectory of the downward flux of particulate organic carbon and in predicting the decreases in sediment effluxes of dissolved inorganic carbon and sediment oxygen demand over a 30-year period. $\mathrm{Sed}2K_s$ and the linked overall water-quality model, $\mathrm{UFILS}4/\mathrm{Sed}2K_s$, were applied to project the extent and timing of improvements in sediment fluxes and hypolimnetic DO resources for an array of P management alternatives. Substantial delays in response associated with the effects of diagenesis were predicted that extended $\ge 40\mathrm{years}$ to reach 95% of steady state. Long-term projections indicate that improvements from the most aggressive P control alternatives may be sufficient to maintain certain coldwater fish.

The distribution of natural organic carbon was investigated at a chloroethene-contaminated site where complete reductive dechlorination of tetrachloroethene (PCE) to vinyl chloride and ethene was observed. In this study, operationally defined potentially bioavailable organic carbon (PBOC) was measured in surficial aquifer sediment samples collected at varying depths and locations in the vicinity of a dense nonaqueous phase liquid (DNAPL) source and aqueous phase plume. The relationship between chloroethene concentrations and PBOC levels was examined by comparing differences in extractable organic carbon in aquifer sediments with minimal chloroethene exposure relative to samples collected in the source zone. Using performance-monitoring data, direct correlations with PBOC were also developed with chloroethene concentrations in groundwater. Results show a logarithm-normal distribution for PBOC in aquifer sediments with a mean concentration of $187\mathrm{mg}/\mathrm{kg}$. PBOC levels in sediments obtained from the underlying confining unit were generally greater when compared to sediments collected in the sandy surficial aquifer. Results demonstrated a statistically significant inverse correlation ($p=0.007$) between PBOC levels in aquifer sediments and chloroethene concentrations for selected monitoring wells in which chloroethene exposure was the highest. Results from laboratory exposure assays also demonstrated that sediment samples exhibited a reduction in PBOC levels of 35% and 73%, respectively, after a 72-h exposure period to PCE ($20,000\mathrm{\mu g}/\mathrm{L}$). These results support the notion that PBOC depletion in sediments may be expected in chloroethene-contaminated aquifers, which has potential implications for the long-term sustainability of monitored natural attenuation.

The use of computational fluid dynamics (CFD) as an engineering tool for the design of storm water retention ponds is a rapidly growing area of interest, but there is a large gap in the literature with regard to validating the CFD models against experimental data for investigation of flow patterns and velocity distributions in storm water retention ponds. This paper assesses a CFD model against experimental flow data from a laboratory-scale physical model of an existing field retention pond. The simulated results were compared to each other and also to the experimental data to test the ability of numerical simulations for this type of problem. A representative and realistic range of flow rates from 0.16 to $1.5\mathrm{L}/\mathrm{s}$ was tested in the physical model for comparison with the CFD model. Also, the vorticity from the physical model tests was compared to that from the numerical model to validate the CFD model. The results confirm previous findings that CFD modeling is a potential engineering tool to simulate hydraulics of storm water retention ponds and can reliably be used in pond design even at moderate computational cost. Also, it was found that CFD is relatively insensitive to the turbulence model used and grid density within a wide range of grid densities for observing general flow patterns. However, it is sensitive to the advection schemes for this particular problem. Higher order differencing schemes (high-resolution scheme) worked better than simple differencing schemes like the upwind differencing scheme (UDS). It was also found that the strength of the vorticity increases with increasing flow rate for both models, and at higher flow rates CFD is more consistent in predicting the vorticity than that of the particle tracking velocimetry (PTV) technique used in physical models.

The City of Atlanta plans to convert a recently acquired urban quarry into Bellwood Quarry Reservoir (BQR), an off-stream supplemental water supply. CE-QUAL-W2, Version 3.5, was used to perform five-year water quality simulations of the reservoir. Nine inflow/outflow design and operation scenarios were evaluated for thermal stratification, dissolved oxygen (DO), algae, and iron (Fe) and manganese (Mn). For a perpetual storage scenario with no inflow and outflow and an operating scenario with monthly pump exercising flows, the model predicts ample reservoir DO to maintain an oxidizing environment in the hypolimnion and low dissolved Fe and Mn. A single portal near the bottom of the reservoir is recommended for alternating pump exercise inflows and outflows. Continuous flow-through operating scenarios do not improve predicted water quality. However, if selected, a mid-depth inflow and bottom outflow design configuration is recommended. A natural monomictic cycle is essential to maintain hypolimnetic DO. BQR’s relatively large mean depth (60 m) and low sediment surface area to volume ratio is consistent with other mesotrophic lakes of similar water quality.

Dissolved organic matter (DOM) and bromide, two primary precursors of disinfection by-products, cannot be effectively removed by conventional coagulation processes. Alternative methods are required to ensure the safety of drinking water. The effectiveness of using magnetic ion exchange resin (MIEX) and high-performance coagulant (HPAC) in a hybrid process for the removal of turbidity, DOM, and bromide in raw water supplies was investigated. Results showed that a wide range of DOM, with different molecular weights (MWs) containing soluble microbial byproduct-like organic matters, aromatic protein, and fulvic/humic acid, was removed by MIEX in the two studied water sources. In addition, up to 91% of bromide was removed in bromide spiked raw waters. In comparison, HPAC coagulation did not effectively eliminate low MW or soluble microbial byproduct-like organic matters in raw water. The combined MIEX and HPAC process more significantly improved the performance of turbidity, DOM, and bromide removal than coagulation alone. With MIEX pretreatment, the requisite HPAC dosage was reduced by at least 67%.

A silica–titania composite (STC) material has been commercialized for mercury removal from caustic exhaust at a U.S. chlor-alkali facility employing a mercury-cell process. A second-generation STC, applicable to larger flow rates such as those associated with coal-fired boiler flue gas, was developed and evaluated for this study. This new material, consisting of high void space chemical tower packing dip coated with an STC formula, was named Silica-Titania–Coated Packing (STCP). The STCP was capable of 90% oxidation and removal of mercury from a simulated flue gas under conditions where other technologies would fail. Additionally, a preliminary cost estimate determined that the STCP would have much lower average annual costs than two commercial technologies that are widely accepted as the best available for this application.

Dynamic experiments on the adsorption of copper ions were conducted in a laboratory packed bed containing amine-modified SBA-15 pellets. Breakthrough curves were analyzed at different flow rates and after two adsorption–desorption cycles. A model based on mass balance was developed and tested to predict the breakthrough curves under the different experimental conditions. The model was found to agree with the experimental data. Bed regeneration was performed by circulating a ethylenediaminetetraacetic acid solution through the column.

The U.S. Army is seeking to produce safer, less sensitive munitions through the addition of two new energetics, 2,4-dinitroanisole (DNAN) and $N$-methyl paranitro aniline (MNA), to the munitions’ formula. Production of these munitions would add them to the waste stream. The use of an anaerobic fluidized-bed bioreactor (AFBB) was studied for treating these compounds in a simulated wastewater with ethanol as the electron donor. The reactor degraded both of the compounds to below detection limits over a wide range of ethanol concentrations. The degradation was found to be a transformation into secondary products for both energetics: diaminoanisole (for DNAN) and $N$-methyl-$p$-phenylenediamine (for MNA). Both of these by-products reacted upon exposure to air, forming azobond dimers. Potassium perchlorate was added to the feed stream to test if additional energetics would disrupt the transformation. The AFBB continued transformation and was able to remove the perchlorate after a two week acclimation period. The AFBB was an effective treatment method for DNAN and MNA, but further study is required to investigate the transformation products more thoroughly.

This paper reports the application of flocculation, Fenton’s process, and ozonation on the depuration of distillery wastewaters. The treatment processes were analyzed separately; nevertheless, integration schemes were also considered. The commercial flocculant 923PWG led to up to 84% of total suspended solids (TSS) removal even if a very low chemical oxygen demand (COD) abatement was observed ($<2\%$). The effect of hydrogen peroxide concentration and $[{\mathrm{H}}_2{\mathrm{O}}_2]/[{\mathrm{Fe}}^{2+}]$ molar ratio were tested on the Fenton’s process efficiency. Moreover, the ${\mathrm{H}}_2{\mathrm{O}}_2$ injection mode was still analyzed. Gathering up both COD degradation and biodegradability enhancement, it was possible to select the following operating conditions: $[{\mathrm{H}}_2{\mathrm{O}}_2]=0.5\mathrm{M}$ and $[{\mathrm{H}}_2{\mathrm{O}}_2]/[{\mathrm{Fe}}^{2+}]=10(\mathrm{mol}/\mathrm{mol})$, being the oxidant totally introduced at the beginning of the reaction. Ozonation showed poor results regarding organic matter removal; however, it seems to be an interesting methodology raising the effluent’s biochemical oxygen demand $({\mathrm{BOD}}_5)/\mathrm{COD}$ ratio. In fact, in the strategies involving the integration of ozonation with Fenton’s process, it was possible to conclude that ozone oxidation did not improve COD degradation; nonetheless, it was essential to improve biodegradability, particularly when conducting ozonation (at $\mathrm{pH}=3$) after Fenton’s process was revealed to be an interesting approach leading to an effluent amenable to further biodegradation.

Natural organic material (NOM) has been the focus of many studies because of its ability to compromise water treatment processes. This case study utilized ultraviolet (UV) spectroscopy and molecular weight distributions to investigate the impact of six water treatment processes (alum coagulation, magnetic ion exchange (MIEX) resin treatment, chlorination, ozonation, powdered activated carbon (PAC) adsorption, and biological sand filtration) on the removal of NOM from an Australia water source, Myponga Reservoir. Each of these processes displayed different effects on the concentration and character of NOM. The removal of dissolved organic carbon (DOC) and UV absorbance at 254 nm (${\mathrm{UV}}_{254}$) by MIEX and the biological sand filter was shown to follow first-order kinetics with rate constants ranging from $9.0\times {10}^{-8}{\mathrm{s}}^{-1}$ (biological sand filter) to $6.3\times {10}^{-5}{\mathrm{s}}^{-1}$ (MIEX treatment). UV spectroscopic investigations showed the potential to predict the formation of disinfection by-products from chlorination with strong correlations ($R^2$ of 0.96) observed between the formation of trihalomethanes and the differential UV absorbance at 265 nm. Ozonation and biological sand filtration also appeared to target NOM absorbing at 265 nm. Molecular weight distribution analyses showed MIEX treatment to be the most effective single process in achieving high removals of a wide molecular weight range of NOM, consistent with the high removals of DOC and ${\mathrm{UV}}_{254}$ (up to 90% and 80% removal, respectively). The combination of alum and PAC, treatment options which exist at the majority of Australian water treatment plants, also proved effective for high NOM removal over a wide range of molecular weights, where alum effectively removed high molecular weight compounds, and PAC effectively removed low molecular weight compounds. This study has demonstrated that valuable information can be gained through simple manipulation of UV absorbance and molecular weight distribution data, which could be beneficial to water utilities in not only facilitating the selection of treatment processes when commissioning WTPs, but also optimizing existing treatment processes for effective NOM removal.

In this study, two types of commercial nanofiltration (NF) membranes (NF70 and NF270) were chosen to remove four model disinfection by-product (DBP) precursors (resorcinol, phloroglucinol, 3-hydroxybenzoic acid, and tannic acid) with different functional groups. The clean NF membranes were also characterized by physicochemical properties of membrane surface charge (determined by streaming potential measurement), hydrophobicity [determined by contact angle measurement), membrane surface morphology (determined by scanning electron microscopy and atomic force microscope (AFM)], surface roughness (determined by AFM), and surface composition (X-ray photoelectron spectroscopy) to validate the NF rejection mechanisms of the DBP precursors. The filtration experiments of the model DBP precursors were assessed under various pH values (3–10), and the removal efficiencies were good at high pH values for both membranes (84–99% at pH 9 and 89–100% at pH 10). Electrostatic repulsion is the prevailing mechanism for the model compounds with negatively ionizable functional groups rejected by the negatively charged NF membranes at high pH values, whereas steric hindrance exclusion and adsorption are controlling factors for the rejection of unionized small organic molecules. For the examination of membrane surface roughness, it is essential to use the same scan area when comparing the surface roughness of different membranes. Furthermore, NF70, with a rougher surface, reasonably posed a higher fouling potential, which makes the flux decline more serious than that of NF270 under the same operation conditions.

Free chlorine, bovine catalase, sulfite, and thiosulfate were examined as methods for quenching residual hydrogen peroxide after advanced oxidation treatment of water for subsequent laboratory analysis. The challenges and advantages of each method are presented to help inform researchers and practitioners. Quenching hydrogen peroxide with free chlorine is determined to be a better alternative for disinfection byproduct testing, while bovine catalase is a preferred quenching agent for toxicity testing and mass spectrometry analysis.

In this study, a relatively high-rate treatment (hydraulic retention time $\le 2.5\mathrm{days}$) of latex-processing wastewater was achieved in a two-stage upflow anaerobic sludge blanket reactor. A maximum acid production rate of $1.8\mathrm{g}{\mathrm{L}}^{-1}{\mathrm{d}}^{-1}$ acetic acid was obtained when the acidogenic reactor was operated over a hydraulic retention time (HRT) of 4 hr. Total chemical oxygen demand (COD) and sulfate removal efficiencies of the system were 51.6 and 75.9%, respectively, when the methanogenic reactor was operated over an HRT of 24 h. The specific acidogenic activity of granular sludge in the acidogenic and methanogenic reactor varied between 1.33–2.02 g COD g volatile suspended solids ${(\mathrm{VSS})}^{-1}{\mathrm{d}}^{-1}$, and the specific methanogenic activity ranged from 0.0078 to $0.0138\mathrm{g}\mathrm{COD}\text{-}{\mathrm{CH}}_4\mathrm{g}{\mathrm{VSS}}^{-1}{\mathrm{d}}^{-1}$. Microbial diversity monitoring using the denaturing gradient gel electrophoresis technique revealed that there was more sulfate-reducing bacteria in both reactors than archea bacteria.