Journal of Environmental Engineering

A model of sediment oxygen demand (SOD) is presented that relates the SOD to flow velocity over the sediment. Previous analyses emphasize, almost exclusively, the relationship between SOD and sediment composition, i.e., their chemistry and biology. Herein a quantitative relationship is established between SOD and the velocity and dissolved oxygen concentration in the bulk water. Oxygen consumption in the sediment is expressed as the sum of biological consumption with Michaelis‐Menten kinetics, and the chemical consumption assumed to be a first order reaction of oxygen. At very low flow velocities, transport through the diffusive water boundary layer is the limiting factor of SOD, and SOD is expressed as a linear increasing function of velocity. On the other hand, when flow velocities are increased, SOD becomes independent of velocity, since the reactions in the sediment become rate‐limiting. The model also suggests that SOD is an increasing function of dissolved oxygen concentration in the water overlying the sediment. Combined with the linear theory of internal seiche motion, an average SOD in a rectangular, two‐layered lake is derived as a function of the wind velocity, aspect ratio of the lake and the depth of the thermocline. The average SOD has a minimum when the thermocline depth is one‐quarter of the total depth.

QUAL2E is a riverine water‐quality program that has been used to study a wide range of environmental problems. Through numerous updates and improvements, it now contains some of the most sophisticated water‐quality kinetics of any water‐quality code in common use. However, due to significant numerical dispersion in its advective routine, QUAL2E may be inappropriate for certain types of simulations, such as the pulse‐input loads for combined sewer overflows. The existing scheme is an implicit backward difference scheme that produces positive numerical dispersion for all conditions. To reduce this limitation, QUAL2E was modified and the existing solution scheme replaced with an explicit backward‐difference scheme, enabling it to more accurately simulate this class of problem under certain simplifying conditions and time‐step controls. Numerical tests showed that the changes achieved this objective, and an application to a combined sewer overflow study on the Charles and Mystic Rivers in Boston is then described.

Both self‐cleansing‐velocity and minimum‐shear‐stress approaches have been used to investigate the effect of the relative flow depth in a sewer as an important factor influencing the buildup of deposits in a sewer designed for full flow. Each of the approaches indicates favorable deposit‐free flow conditions for sewers with relative‐normal flow depths ranging between 0.5 and 1.0. For relative flow depths well below 0.5 the same sewers show a strong tendency to form deposits. Nomographs have been developed for circular sewers partially clogged by deposits for easier computations of flow depths. Shear stresses in partially clogged sewers have been computed and discussed in relation to deposit thickness; the study here is limited to initiation criterion only.

This article describes the results of a series of research tasks to develop a procedure to investigate non‐storm‐water (dry‐weather) entries into storm‐drainage systems (Pitt et al. 1993a, 1993b). Dry‐weather flows discharging from storm‐drainage systems contribute significant pollutant loadings to receiving waters, and although they can originate from many sources, the most significant include sanitary wastewater, industrial‐ and commercial‐pollutant entries, failing septic‐tank systems, and vehicle‐maintenance activities. Protocols are discussed to characterize the drainage area, locate and identify polluted outfalls, estimate the magnitudes of non‐storm‐water entries, and locate and correct the non‐storm‐water entries into the storm‐drainage system. If these loadings are ignored (e.g., by only considering wet‐weather storm‐water runoff), only limited improvement in receiving‐water conditions may occur with storm‐water pollution‐control programs.

A land‐application system's performance in limiting the leaching of nitrogen (N) and phosphorus (P) through the soil zone was evaluated. The system was located in Heber Valley, Utah, which has an average frost‐free growing season of only 100 days. The amount of N that leached through the soil zone was 6.3 kg N/ha or 24% of the N applied. The amount of P that leached through the soil zone was 0.84 kg P/ha or 0.2% of the P applied. Leachate quantities based on the consumptive water use of the alfalfa as determined by the modified Blaney‐Criddle formula were unsatisfactory. Instead, leachate quantities were estimated with a simple model based on the amount of the soil water collected by porous‐cup suction samplers. It was recommended that N leaching may be further reduced by growing a winter‐hardy grass, such as smooth bromegrass, with the alfalfa and by reducing leaching during the growing season by limiting irrigation to periods when soil‐water supplies are low. This study demonstrates that land‐application systems can be an effective method of wastewater management for mountain communities influenced by relatively harsh climates.

The feasibility, advantages, and problems of using linear‐programming methods in the preliminary analysis of central material‐recovery facilities are discussed and demonstrated. The design and operation of such facilities involves a wide variety of choices. Linear programming can help identify promising directions for design and operations from many feasible choices. Linear programming is likely to be most effective when used in conjunction with traditional engineering process flow diagrams, computerized in spreadsheet form. The proposed method is implemented using a commercial linear‐programming package and using Microsoft EXCEL spreadsheet software.

A multiobjective optimization model is presented to determine the efficient aggregation of land parcels for use as a solid‐ or hazardous‐waste landfill. In this work, a new constraint for measuring compactness and contiguity of the selected subregion is introduced. This constraint measures compactness and contiguity as a function of the subregion's outside perimeter and area. The model optimally selects and sizes the landfill site from multiple bundles of available discrete and irregularly shaped land parcels. Each bundle consists of several individual and irregular land parcels and is disjoint from all other bundles. The model is multiobjective in nature, addressing land purchase cost, compactness, and contiguity considerations. The quantity of land required is determined by a projection of the flow of waste from the communities, commercial areas, and industries the landfill serves. To illustrate the model, a case study is presented using the landfill‐site search process in Orange County, North Carolina.

This paper reviews and extends understanding of evaporation/distillation as treatment for landfill leachate. Evaporation may produce condensate that is high quality and easier to dispose of than effluent from conventional leachate‐treatment processes with the volume of the concentrated residuals being a small fraction of the original leachate volume. Laboratory‐scale, one‐ and two‐step distillation experiments with pH adjustment were performed with ammonium acetate test solution and three leachate samples from older landfills. A single‐step, acidic distillation of the strongest leachate samples removed more than 95% of ionic impurities, except for volatile organic acids, which were removed at 85%. Two‐step acid‐base evaporation, or ammonia stripping coupled with evaporation, may effectively remove ammonia and volatile organic acids if present together in relatively high concentrations. A Monte Carlo analysis of the feasibility for energy‐self‐sufficient leachate evaporation via landfill methane combustion shows, theoretically, that there will be sufficient methane gas to vaporize leachate in a majority of modern landfill situations based on the limited published data and assuming uncorrelated input parameters. Balancing landfill leachate production to methane production may be tenuous for some landfill settings. Although the data developed here are limited, this investigation suggests that by‐product leachate and gas from landfill operations may be used beneficially in a novel, energy‐self‐sufficient treatment process.

Environmental decision making (EDM) analysis is developing into a recognized specialty of environmental engineering. One of the major challenges of EDM is to accommodate the high levels of uncertainty that exist in many environmental problems. EDM methods based on absolute or relative, probabilistic, multiattribute utility analysis are presented. Example EDM solutions are presented for the selection of remedial action alternatives for a Superfund site using the Environmental Protection Agency's decision criteria and New York's system for quantifying decision criteria. The advantages and limitations of both are discussed. Results illustrate that the major advantages of EDM analysis are in the need to formalize the decision criteria by which a decision is made and the value judgments that are applied to weigh multiple criteria. EDM analysis can also improve decision process documentation and can sharpen the quality of debate about controversial decisions.

Soil washing was evaluated in combination with biological treatment as a decontamination technology for petroleum‐contaminated soils. The bench‐scale soil washing system utilized to clean the soils also fractionated the bulk soil into sand, silt, and clay fractions. With tap water as the carrier, the petroleum removal efficiencies varied from 44% to 55% for three soils. The postwash hydrocarbon levels were in the range of 145–905 mg/kg for sands, 2,000–5,000 mg/kg for silts, and greater than 14,000 mg/kg for clays. Biological degradation was evaluated as a secondary treatment to reduce the contaminant levels on each fraction to desired levels. Simulated composting lowered hydrocarbon levels on sands to below 50 mg/kg. Slurry treatment of silt and clay fractions reduced hydrocarbon levels to near 100 mg/kg for silt and in the range of 500 to 1,000 mg/kg for clays. It was found that composting and slurry treatment effectively met the suggested target‐treatment level of 100 mg/kg for the sand and silt fractions. For the clays, slow desorption of the hydrocarbons and long treatment periods made slurry treatment an uneconomical alternative.

It is well known that residual NAPL (nonaqueous phase liquid) ganglia remain trapped in the saturated domain after pump‐and‐treat operations and create a long‐term source of contamination. Mobilization and extraction of these ganglia is a problem of paramount importance in ground‐water remediation. The present study consists of laboratory and theoretical investigations on the effect of vibrations on mobilization of these ganglia. Laboratory tests were conducted using a variable‐amplitude vibrating table on two sandy soils contaminated with hexadecane. Vibrations augmented by viscous forces due to flowing water yielded better recovery of ganglia than under buoyancy conditions alone. The mobilization of ganglia was found to depend on initial soil density and water flow rate under which the ganglia were trapped. In general, higher vibration amplitudes resulted in higher removal of NAPL ganglia. The existing mobilization criteria were modified to take into account the cyclic pore pressures and relative density changes observed during vibrations. The modified mobilization criteria when applied to the laboratory test conditions yielded information about stable lengths of ganglia in the two sands at each vibration amplitude. The percentages of trapped volume occupied by ganglia of various ranges of lengths in the two sand samples were deduced. The results indicate that localized vibrations in a saturated domain, when augmented by suitable flow conditions, can effectively mobilize and extract NAPL ganglia.

The factors affecting the dissolution kinetics of 1,1,2‐trichloroethene (TCE) and 1,1,2‐trichloroethane (TCA) pools in saturated porous media were investigated, using an experimental aquifer and a pool formation method specially developed for dense nonaqueous‐phase liquids (DNAPLs). Under conditions of steady and uniform flow, the dissolved DNAPL concentrations decreased with increasing vertical distance from the pool resulting in steep concentration gradients. The concentration gradient increased with increasing ground‐water velocity. The vertical extent (thickness) of the dissolved plume increased with increasing distance along the pool and decreasing ground‐water velocity. In all cases, measured concentrations were a small fraction of the respective DNAPL solubility. A two‐dimensional steady‐state transport model successfully predicted the TCA experimental data, but the prediction for TCE was rather poor. This was attributed to heterogeneities in the TCE pool. Depending on ground‐water velocity and DNAPL solubility, pool removal times by pumping were determined to be between 0.4 and five years, which is an extremely long period for the small volumes (200 mL) of DNAPL considered.

A general composite sorption isotherm is developed to predict experimental isotherms of the types I, II, and III. The isotherm, based on heterogeneous chemical thermodynamic equilibrium considerations, is used to represent single vapor sorption isotherm data at high concentrations. Sorbents investigated include polybutylene, low‐density polyethylene, isopolypropylene, polytetrafluoroethylene, polyvinyl chloride, granular activated carbon, graphon, cabosil, Iceland spar, and quartz. Sorbates include ethane, n‐butane, n‐pentane, n‐hexane, n‐octane, trichloroethylene, tetrachloroethylene, methyl bromide, dichloromethane, vinyl chloride, benzene, ethanol, and ethyl ethanoate. The isotherm has two terms with two parameters each. One term represents the absorption of individual vapor molecules, or monolayer adsorption. The second term represents the absorption of vapor molecule clusters, or multilayer adsorption. Contributions of terms 1 and 2 to overall sorption are generally predominant at relatively low and high sorbate concentrations, respectively. In contrast with the classic concept of infinite sorption for the BET isotherm, composite isotherm term 2 sorption capacity is finite.

In unsaturated soil, methane and volatile organic compounds can significantly alter the density of soil gas and induce buoyant gas flow. A series of laboratory experiments was conducted in a two‐dimensional, homogeneous sand pack with gas permeabilities ranging from 110 to 3,000 darcy. Pure methane gas was injected horizontally into the sand and steady‐state methane profiles were measured. Experimental results are in close agreement with a numerical model that represents the advective and diffusive components of methane transport. Comparison of simulations with and without gravitational acceleration permits identification of conditions where buoyancy dominates methane transport. Significant buoyant flow requires a Rayleigh number greater than 10 and an injected gas velocity sufficient to overcome dilution by molecular diffusion near the source. These criteria allow the extension of laboratory results to idealized field conditions for methane as well as denser‐than‐air vapors produced by volatilizing nonaqueous phase liquids trapped in unsaturated soil.

Undisturbed urban snowpack effectively collects various airborne pollutants (including PAHs), which are deposited within the snowpack through the process of wet and dry deposition. The deposition‐capture and long‐term‐retention properties of snowpacks are utilized in developing a source‐receptor model for PAH source‐apportionment for various PAH species for the city of Sault Ste. Marie, Ontario, Canada. The source‐receptor modeling approach uses the source‐contribution results of a vector of PAHs as a means of characterizing the industrial and urban deposition contributions. The method described in this paper relates air concentration to deposition level without violating source‐receptor modeling's basic premise of linearity in the system response to system emission, thus providing opportunities for source apportionment in air concentrations and deposition levels. The results of source contributions match well the observed deposition levels in a qualitative sense for individual species. Statistical interpretation techniques are used to quantify the source contributions.

A model was developed for predicting the fate of metals in municipal wastewater treatment. The model, which incorporates metal solubility and sorption onto solids, was calibrated with experimental data obtained by a titration technique. The calibrated model was evaluated using dynamic data collected in a separate study from three Ontario wastewater‐treatment plants (WWTPs). A goodness‐of‐fit test was employed to determine if the model predictions were statistically different from the observed effluent concentrations. Predictions of copper and zinc effluent concentrations at two WWTPs were found to lie within the confidence limits defined by the variability of field replicates. Lack of fit of some model predictions was attributed to biasing of the data at low concentrations and lack of model calibration. The configuration of the model was concluded to be correct since the behavior of metals with significantly different physical properties were well predicted. It is concluded that the most accurate model predictions can be made if the model is calibrated at the specific WWTP.

In‐situ oxygen consumption rate (OCR) observations for trickling filter biofilms under different hydraulic, substrate loading, and oxygen availability operating conditions are presented. OCRs were determined for sucrose‐ and dextrin‐based synthetic feedstocks by growing biofilms on a 1.2‐m deep section of 60° cross‐flow media that was enclosed within a reactor vessel, sealing the reactor vessel to outside air entry, and observing the time rate of decrease in the reactor's oxygen partial pressure. The results indicate that OCRs increased proportionately with increasing influent substrate concentrations for the 40 to 120 mg/L SCOD range, increased slightly or remained relatively constant with increasing influent substrate concentrations for the 120 to 200 mg/L SCOD range, increased slightly with increasing DO saturation condition for the 7.4 to 10.1 mg/L range, were affected by oxygen availability for influent substrate concentrations throughout the 70 to 175 mg/L SCOD range, increased with increasing hydraulic application rate from 2.4 to 4.9 m³/m²‐h, and were slightly greater than predicted by currently acknowledged models. Other observations include the following: measured OCRs were at least three times greater than those estimated from clean media oxygen‐transfer testing; without liquid and feedstock application, biofilms consumed oxygen at 80–104% of their usual rate; and maximum biofilm oxygen consumption was ∼ 680 mg/m²‐h.

The accumulation of the nitrification intermediate, nitrite, is undesirable because of its toxicity to fish. The elevated nitrite concentration in wastewater‐treatment processes significantly increases the amount of chlorine required for disinfection. Consequently, environmental and operational factors of hydraulic and ammonium loading rates as well as pH and alkalinity were evaluated for their effects on the accumulation of nitrite in a fixed‐film submerged filter system. Elevated nitrite levels were observed at higher hydraulic (5 L/d·cm²) and ammonium loading rates (0.7 kg N/d·m²), $\mathrm{pH}=8.8$ and low alkalinity/ammonium ratio $(6\mathrm{mg}{\mathrm{CaCO}}_3/\mathrm{mg}{\mathrm{NH}}_4^{+}‐N)$. Addition of 2.5–5 mg N/L hydroxylamine, an intermediate in the ammonium oxidation to nitrite, also significantly enhanced the accumulation of nitrite concentration. The inhibitory effect of hydroxylamine on Nitrobacter was irreversible.

Formate in wastewater was effectively removed at 37°C by the upflow anaerobic sludge blanket (UASB) process. This was accomplished by keeping the recycle ratio at 3.0 and by lowering the pH of incoming wastewater to pH 3.8. The COD removal efficiencies were 97–98% at loading rates of 10–20 g‐COD·L⁻¹·day⁻¹, but were reduced to 86–90% at 29–65 g‐COD· L⁻¹·day⁻¹. When the loading was further increased to 77 g‐COD·L⁻¹·day⁻¹, the reactor abruptly failed due to the sudden decrease of the pH. About 94% of COD of the formate removed was converted to methane with a sludge yield of 0.05 g‐VSS·g‐COD⁻¹. The small sludge granules settled satisfactorily and had a maximum specific methane production rate (SMPR) of 2.06 g‐COD·VSS⁻¹·day⁻¹. The specific methane activity (SMA) of granules using formate as substrate was 2.90 g‐COD· g‐VSS⁻¹· day ⁻¹; but the granules were unable to degrade other fatty acids. Microscopic examinations indicated that the granules mainly comprised Methanobacterium formicicum‐like bacteria, plus a small amount of Methanococcus vannielii‐like bacteria, and an unknown species of bacterium.

A histological method was demonstrated for the microscopic analysis of granules from upflow‐anaerobic‐sludge‐blanket (UASB) reactors. Granules were first fixed with formaldehyde and embedded in paraffin, before being sectioned by a microtome to a thickness of 3–5 μm. The section was then mounted on a glass slide for microscopic examination using phase contrast and epi‐fluorescent excitation. Using this simple method, the microstructure of the granules and the distribution of methanogenic bacteria could be analyzed. Over 100 granules from UASB reactors treating sucrose wastewaters were examined. Microscopic photos showed that the granules had a surface layer comprised of hydrogen‐consuming methanogenic cocci and bacilli, and a loosely packed interior, mainly comprised of acetotrophic Methanothrix. This simple method provides engineers with a practical tool for studying the microstructure of the granules, the mechanisms of their formation, and kinetic modeling.