Journal of Environmental Engineering

The analysis of relative fluid motion by Camp and Stein, which led to their well‐known formulation of particle collision rate in flocculation, is examined using standard tools from continuum mechanics. As a result of certain apparent conceptual errors, including the notion that a three‐dimensional relative velocity field can be represented in general by a single velocity gradient, several recommendations are offered, including the abandonment of Camp and Stein's terms “absolute velocity gradient” and “root mean square (rms) velocity gradient.” Revisions in nomenclature and an alternate theory of particle collisions are discussed, as well as some implications of inhomogeneity in the turbulent flocculation problem.

Liquid film coefficients of ethylene gas, $K_L\text{,}$ are determined from data on ethylene and rhodamine WT dye concentration distributions, which were collected in shallow streams and rivers located in Southern Ontario, Canada. Relationships among $K_L\text{,}$ river channel hydraulic characteristics and chemical properties of organic compounds have been developed through dimensional analysis using the Buckingham π‐Theorem. A sensitivity analysis of various parameters on $K_L$ has been carried out. A statistical relationship developed by Rathbun and Tai (1982) has been found to underestimate $K_L$ values for the Southern Ontario streams and rivers. An evaluation of the theoretical relationships between $K_L$ and molecular properties (critical volumes and molecular weights) using experimental data reported in the literature has revealed that the relation between $K_L$ and critical volume is suitable to calculate $K_L$ values for highly volatile organic chemicals using the known $K_L$ values of ethylene gas or other compounds (e.g., propane, benzene); however, for moderate and low volatile compounds, no relationships could be obtained from the available data. The computation of $K_L$ for a given organic compound using the relationships developed in this study is outlined in a step‐by‐step procedure. An example illustrates the various computations involved. In general, the method is applicable to volatile compounds with Henry's law constants greater than ${10}^{-3}\mathrm{atm}‐m^3/\mathrm{mol}\text{.}$

Kriging is a technique which can be used to obtain minimum variance, unbiased estimates of the concentration of a pollutant at a point, or the average concentration in an area or volume. In this paper, the use of simple and universal kriging for analyzing the distribution of dioxin (2,3,7,8‐TCDD) in the sediments of a creek is demonstrated. The resulting point and block estimates obtained with both techniques were then compared. Although the universal kriging point estimates were slightly better than the simple kriging estimates, no clear advantage to using universal kriging was apparent in this example. In general, however, universal kriging should be considered whenever a significant drift is present in the data.

There is increasing interest in suspended sediment as a water quality parameter because of its potential as a transport mechanism for pollutants and because of its possible effects on fish habitat. The application of bivariate probability distribution functions to represent the suspended sediment concentration‐water discharge data for the San Juan River at Bluff, UT is presented. This approach has several advantages over more traditional methods, including preservation of sediment concentration information and convenience for computer implementation. A bivariate log‐normal density function was found to adequately represent the San Juan River data. Average sediment yields were calculated using the closed form solution of the moment generating function, and compared to the results from the flow duration‐sediment rating curve method. The bivariate log‐normal density function was found to be a convenient, accurate method for parameterizing the frequency distribution of sediment concentrations and water discharges for the San Juan River.

A stochastic optimization‐simulation method is presented for delineating least‐cost treatment sequences for a centralized liquid industrial waste treatment facility. A dynamic programming model performs the optimization. The function of the model is to delineate least‐cost treatment sequences that will produce an acceptable effluent stream qualify given a probabilistically‐generated influent waste regime. The model is structured to permit the following user‐determined options: waste types and respective volumes in the waste inventory; specific contaminants within each waste type; contaminant‐specific probability density functions for waste strength; unit treatment processes including performance efficiencies and related costs; and individual contaminant effluent standards. The stochastic dynamic programming model served as a screening device, identifying unit treatment processes and sequences of processes with favorable cost‐effectiveness attributes. The treatment paths thus identified were further analyzed and refined using stochastic simulation techniques.

The effects of documented acidification‐based increases in transparency, i.e., decreases in diffuse light attenuation, $k_d\left(m^{-1}\right)\text{,}$ on the occurrence and character of thermal stratification in lakes is evaluated with a mathematical mixed layer (integral energy) stratification model. Predicted changes in the character of stratification in deep lakes include: deeper epilimnia, reduced density gradients in metalimnia, increased hypolimnetic heating, and reduced stability. These changes, brought about as a result of a reduction in $k_d$ from $0.75-0.15m^{-1}\text{,}$ were generally as great as or greater than those associated with extremes in meteorological conditions in a north temperate climate. The changes in stratification may have important implications with respect to the vertical cycling of dissolved constituents and the oxygen resources of hypolimnia. Model predictions indicate that lakes of maximum depth less than 25 m located in the Adirondack Region of New York State may change in character from exhibiting strong summer stratification to stratifying only weakly or not at all as a result of similar decreases in $k_d\text{.}$ Shallower lakes are more susceptible to less extreme reductions in $k_d\text{.}$ A number of lakes in the Adirondack Region have probably been eliminated as cold‐water fisheries as a result of this effect.

The concept and development of a new type of air classifier, the pulsed‐flow air classifier, are described. Field experience with air classifiers has indicated the need for re‐design. Careful examination of shredded material in light of the literature enables a coherent description of solid waste components. Analyses of fall times show why conventional air classifiers fail to separate adequately. The objective is to achieve separation based more on density and less on aerodynamic characteristics than is possible with current classifier technology. Based on these results and drawing from separation technology, the concept of pulsed‐flow air classification is developed. The theory is briefly examined and compared with previous work. Carefully controlled experimentation with specially constructed equipment shows that pulsed flow air classification is capable of separations by density of which conventional classifiers are not capable. The model is shown to aid design of such classifiers.

The performance of several types of water treatment plants in removing various molecular weight (MW) fractions of naturally occurring aquatic organic matter and humic substances is described. An assessment was made of the performance of direct filtration, conventional treatment, and softening in removing trihalomethane (THM) precursors from a diverse array of water sources. In addition, a comparison was made between conventional treatment and direct filtration in removing THM precursors from a common water source, the Colorado River. As a general rule, THM reactivity (μg THM/mg C) increased with MW although the $<10\text{,}000\mathrm{MW}$ range was found to be the most consistently reactive fraction of aquatic organic matter. All of the various treatments preferentially removed the most reactive fraction of precursor present in each molecular weight range. None of the various treatments proved to be very effective in removing precursor material below a MW of $<500\text{.}$ The ability to remove THM precursors appears to be related to both the source of humic substances as well as the type of treatment employed.

Rapid infiltration systems for wastewater treatment and disposal can be designed rationally based on measured infiltration rates and treatment requirements. The hydraulic loading rate is determined using a percentage of the design infiltration rate. The percentage varies from 2–10%, depending on the nature of the field test for infiltration rate, the variability of the site soils, and the ratio of application period to drying period. The hydraulic loading also affects the treatment efficiency for nitrogen and phosphorus. The basis for designing rapid infiltration systems for nitrogen and phosphorus removal and treatment performance data are provided.

Relationships among raw water quality, pretreatment facilities, and the design of packed bed filters are presented and applied. The particle size, particle concentration, particle surface characteristics, and solution chemistry in the raw water supply have important and predictable effects on filter design. An integrative approach to water treatment plant design, from raw water quality to filter bed performance, will facilitate process evaluation and has the potential for providing a basis for optimal plant design.

External mass‐transfer coefficients are measured with organic compounds, p‐nitrophenol and 2,4‐dichlorophenol, under conditions representative of fixed‐bed activated‐carbon adsorption for water treatment. The superficial velocity was in the range 1.3–8.0 mm/s and the Reynolds number range was 0.8–5. The observed external transfer coefficients are compared with predicted values from four mass‐transfer models; the relationship of Gnielinski best predicts the velocity dependence. The observed transfer coefficients in all cases exceeded the predictions based on ideal spherical geometry. Depending on the conditions and the choice of predictive model, the particle size correction factors ranged from 1.44–2.04. The data from previously published studies show deviations that are similar in direction, but smaller in magnitude.

Adequate control of wastewater treatment processes with variable input parameters can be achieved through the employment of equalization basins. Variability of input and output parameters are analyzed, and mathematical models of conventional and new configurations of equalization basins describing the propagation of regular and random input signals are presented. Comparison of various designs of equalization basins show that new configurations of equalization basins are more efficient than one‐channel tanks (complete mix, m‐tanks in series, or diffused flow). A substantial advantage of new basin types is in that they can be tuned to the particular spectrum of incoming fluctuations.

An equation is developed that allows direct calculation of the minimum slope required to maintain a given self-cleansing velocity in partially full sewers. The self-cleansing slope is a function of the Manning constant, pipe diameter, ratio of actual flowrate to full flowrate, and self-cleansing velocity. It is found that current recommended minimum slopes correspond to an actual velocity of 0.45 m/s when the actual flowrate is approximately 15% of the full flowrate.