Journal of Environmental Engineering

In this paper, we propose a prediction method of the pollutant and a synchronous classification of the current state of ${\mathrm{SO}}_x$ emission in the power plant. We use the autoregressive with exogeneous (ARX) model as a predictor of ${\mathrm{SO}}_x$ emission and use the neural network (NN) as a pattern classifier. The ARX modeling scheme is implemented using the recursive least-squares method to update the model parameters adaptively. The capability of ${\mathrm{SO}}_x$ emission monitoring is utilized with the application of the NN classifier, where genetic algorithms are used to decide the structure such as the number of hidden nodes of a NN classifier. Experimental results show that the ARX model can predict the ${\mathrm{SO}}_x$ emission concentration well and the ARX modeling parameters can be a good feature for the state monitoring. In addition, its validity has been verified through the power spectrum analysis. Consequently, the NN classifier in combination with the ARX model is quite adequate for monitoring the state of ${\mathrm{SO}}_x$ emission.

The efficiency of cyclones varies widely with the size of particles and complex mathematic formulations are used for predicting the collection efficiency of a single particle size. Numerical integration of the latter, in conjunction with the appropriate in each case input particle size distribution, or direct use of nomographs if lognormal input particle size distribution can be assumed, yields the overall efficiency. In either case, no information is available regarding the size distribution of the penetrating particles. As shown in this paper, lognormal input particle size distributions yield lognormal output particle size distributions. Based on this and on the analysis of functional relations, a set of nomographs is developed directly yielding the mean diameter and standard deviation of penetrating particles as a function of the input particle size distribution parameters and key cyclone design and operating conditions. The above set the stage for the rigorous computation of the size-specific emissions from cyclones and the overall and size-specific efficiencies of secondary control systems that may follow. This analysis is in line with current trends, which, on the basis of health considerations, place emphasis on the size-specific rather than on the overall particle emissions.

Barrier systems have recently been employed for contaminant migration control and/or as an integral part of in situ remediation efforts. In this study, we investigate horizontal gel barrier systems constructed by injecting colloidal silica (CS) through horizontal pipes in unsaturated soils. The gel barrier systems are achieved from conversion of a gelling solution to a solid as gelation progresses. The gelation process is initiated in the CS with the addition of electrolytes such as NaCl. The system consisting of a set of horizontal wells is simulated by a vertical two-dimensional mathematical model and the effects of operating parameters and soil properties on emplacement of a CS barrier are evaluated. Laboratory-scale numerical experiments show that an increase in total CS release volume yields a better barrier formation by allowing two gel mixture plumes to merge halfway between each pair of injection pipes thus forming a continuous low-permeability layer between adjacent injection wells. In field-scale numerical experiments, direct correlations between the size of the horizontal gel layer and operating parameters such as the injection pressure head and gel point of the injected CS (a property relevant to NaCl concentration) are observed.

Leaching of trace elements including heavy metals, nonmetals, and radionuclides from surface impoundments of tailings generated during uranium mining and milling often leads to groundwater contamination. This paper reports results from coagulation and membrane filtration experiments designed to evaluate the capabilities of these processes to remove molybdenum, selenium, uranium, radium, thorium, and other mono- and divalent ions from contaminated groundwater in a shallow unconfined aquifer influenced by uranium tailings. A 10 mg ${\mathrm{Fe}}^{3+}/L$ dose at pH 4 and 10 was found to be very effective for removing radium and thorium that were associated with particles in the raw water. Molybdenum and uranium removals by coagulation are consistent with surface complexation and electrostatic interactions between major coagulant and contaminant complexes in solution. Poor selenium removal suggests that selenate [Se(VI)] was the dominant species in the surficial groundwater. Permeation coefficients (intrinsic transport parameters) for dissolved ions and complexes across three new generation thin-film composite nanofiltration and reverse osmosis membranes evaluated obeyed Gaussian distributions with ionic charge having a mean value of zero. Thus, dissolved solute rejection from brackish multicomponent solutions appears to be only a function of the magnitude of ionic charge and not its sign (positive or negative). Solute permeation coefficients decreased in a power-law fashion with increasing product of molecular weight and absolute ion charge suggesting that both properties determine their removal by nanofiltration and reverse osmosis membranes. Because nanofiltration and reverse osmosis were found to be highly effective for removing a variety of ionic solutes they can be employed in pump and treat operations for groundwater purification prior to reinjection.

A model for predicting biofiltration efficiency is developed and evaluated in this study. This model, based on quantitative structure-activity relationships (QSARs), relates the biofiltration maximum removal capacity of a pollutant to its physicochemical intrinsic descriptors. After a preliminary experimental determination of biofilter removal, for 11 volatile organic compounds, collected data allowed calibration of the model parameters. A screening of different possible molecule describing parameters determined that three of them play a significant role in biofiltration: the Henry’s law constant, the octanol–water partition coefficient, and the first-order connectivity index. The use of these parameters for evaluating biofiltration removal by a QSAR modeling approach allowed good agreement between experimental and model values, while keeping an excellent statistical significance.

Performance of a two-stage full-scale reverse osmosis (RO) process for a desalination plant in Florida was simulated with a mathematical model based on the principles of membrane transport and mass conservation. In this model, water flux at any point along the filtration channel is calculated locally according to the basic transport theory of RO membranes. The changes in cross-flow velocity and salt concentration along the filtration channel were determined using mass balance principles of water and salt. Simulations of the plant performance were compared with the in-plant observation data over a period of more than 300 days. The results showed that the model could adequately describe the performance of the full-scale RO process based on a few module and operating parameters. The study also revealed that salt rejection of a RO membrane changed with feed salt concentration. The osmotic pressure coefficient that fits best with performance of this plant was substantially lower than the value determined with the “rule of thumb” (i.e., osmotic pressure in $\mathrm{psi}\approx 0.01\times \mathrm{total}$ dissolved solids in mg/L) and had to be determined specifically for the particular feed water being processed.

An artificial neural network and genetic algorithm routine has been developed for predicting and optimizing membrane system performance. The model predicted system behavior in response to operating conditions of applied pressure and crossflow velocity. Artificial neural networks accurately modeled mechanisms involved in fouling of membranes by natural organic matter. The model correctly predicted the effects of calcium within the solution in exacerbating fouling, binding of the divalent calcium ions to the natural organic matter macromolecules, and the formation of complexes. The model also correctly predicted the role of increased pressure in inducing fouling and the reverse scenario of mitigating fouling with increased crossflow velocity. The model was applied to membrane plant design for determining cost-effective operations. The genetic algorithm routine searched the predictions of the system model to determine the optimal operating conditions. Fouling conditions induced by the presence of calcium resulted in escalating costs with increases in calcium concentration. Membrane-related cost components were shown to be a significant cost factor that is sensitive to operating conditions and represents a prime target for optimization.

Treatment of fluoride-containing wastewater from the semiconductor industry by dissolved air flotation was investigated. Calcium chloride was added to the wastewater to form calcium fluoride precipitate $\left({\mathrm{CaF}}_2\right)\text{,}$ which was then removed from the aqueous solution by flotation. When the molar ratio, $\left[{\mathrm{Ca}}^{2+}\right]/\left[F^{-}\right]\text{,}$ was equal to 0.5, residual fluoride concentration was lower than 15 mg/L. All flotation reactions were completed in 5 min. Experimental results indicated that the sodium oleate (SOl) dosage was the major factor in determining the removal efficiency of ${\mathrm{CaF}}_2\text{.}$ Effective flotation efficiency (>95%) of ${\mathrm{CaF}}_2$ was found in the pH range of 5.5–9.5. Flotation efficiency increased slightly with increasing saturation pressure (3 to 5 $\mathrm{kg}/{\mathrm{cm}}^2)$ and recycle ratio (20–40%). The presence of sulfate and phosphate resulted in higher residual fluoride concentration and lower flotation efficiency. The interference could be overcome by increasing $\left[{\mathrm{Ca}}^{2+}\right]\left[F^{-}\right]$ to 1.0. However, a higher dosage of SOl was needed, An ionic strength higher than 0.1 M did not affect the removal of ${\mathrm{CaF}}_2\text{,}$ but rendered residual fluoride concentrations higher.

This paper describes an integrated solid waste management (ISWM) model to assist in identifying alternative SWM strategies that meet cost, energy, and environmental emissions objectives. An SWM system consisting of over 40 unit processes for collection, transfer, separation, treatment (e.g., combustion, composting), and disposal of waste as well as remanufacturing facilities for processing recycled material is defined. Waste is categorized into 48 items and their generation rates are defined for three types of sectors: single-family dwelling, multifamily dwelling, and commercial. The mass flow of each item through all possible combinations of unit processes is represented in a linear programming model using a unique modeling approach. Cost, energy consumption, and environmental emissions associated with waste processing at each unit process are computed in a set of specially implemented unit process models. A life-cycle approach is used to compute energy consumption and emissions of CO, fossil- and biomass-derived ${\mathrm{CO}}_2\text{,}{\mathrm{NO}}_x\text{,}{\mathrm{SO}}_x\text{,}$ particulate matter, ${\mathrm{PM}}_{10}$ and greenhouse gases. The model is flexible to allow representation of site-specific issues, including waste diversion targets, mass flow restrictions and requirements, and targets for the values of cost, energy, and each emission. A companion paper describes the application of this model to examine several SWM scenarios for a hypothetical, but realistic, case study.

A companion paper described the development of the integrated solid waste management (ISWM) model that considers cost, energy, and environmental releases associated with management of municipal solid waste. This paper demonstrates the application of the ISWM model to a hypothetical, but realistic, case study. Several solid waste management (SWM) scenarios are studied, including the variation in energy and environmental emissions among alternate SWM strategies; the effect of mandated waste diversion (through recycling and other beneficial uses of waste such as combustion to recover energy) on environmental releases and cost; the tradeoff between cost and the level of waste diversion; and the tradeoff between cost and greenhouse gas emissions. In addition, the flexibility of the model is illustrated by the identification of alternate SWM strategies that meet approximately the same objectives using distinctly different combinations of unit processes. This flexibility may be of importance to local solid waste management planners who must implement new SWM programs. Use of the model illustrates the potential impact of solid waste management policies and regulations on global environmental emissions.

We appreciate the comments of Wahlberg et al. on the discussion of our manuscript and regret there were some errors and omissions in the final manuscript that was published. Yet, we firmly stand by Eq. (7), which was developed during our study and which corroborates Eq. (3) proposed by Daigger (1995). We strongly disagree with some of the interpretations and conclusions that were advocated in their discussion and therefore, we will address each of the seven issues individually.