Journal of Environmental Engineering

The goal of this study was to examine the effect of an organic shock load on the performance and stability of a laboratory-scale anaerobic migrating blanket reactor (AMBR). To accomplish an organic shock load, nonacidified sucrose solution was almost doubled in concentration, while maintaining a constant hydraulic retention time. The volumetric loading rate (VLR) was increased from 27 to 50 g chemical oxygen demand (COD) L⁻¹ day⁻¹ for a period of six hydraulic retention times (42 h). This resulted in an increase in the standard methane production rate (liters of methane at standard temperature and pressure per liter reactor volume per day) from 7 to 12 L L⁻¹ day⁻¹. The pH levels stayed favorable and biomass washout was limited during the shock load due to the damping effects of a compartmentalized configuration. During the shock load, the propionate production in the initial compartments of the AMBR remained at the same level as before the shock load, while the acetate production rose sharply. Because propionate is the most difficult volatile fatty acid to be removed, unstable conditions due to excessive propionate accumulation during the shock load were prevented. Meanwhile, the acetate concentration in the liquid phase and hydrogen content in the headspace of the final compartments remained low, which ensured propionate degradation. Due to these intrinsic characteristics of the AMBR, the soluble COD removal efficiency stayed above 87% under these stressed conditions. Moreover, the performance of the AMBR reached pre-shock-load levels immediately after the VLR was restored to 25 g COD L⁻¹ day⁻¹.

In Taiwan, surfactants are frequently used in the flotation process to aid in solid/liquid separation. Their effect on the dewatering of alum sludge was investigated. Various amounts of cationic and anionic surfactants were added to sludge samples, and the dewatering characteristics of the sludge and the water content of sludge cakes were evaluated. Both surfactants improved the dewatering of the sludge by lowering the specific resistance to filtration, decreasing the bound water content, and increasing the dewatering rate of the sludge. Different combinations of anionic and cationic surfactants and polyelectrolytes were also experimented on to study the effect of surfactant addition on the dewatering characteristics of polyelectrolyte-conditioned sludge. Experimental results indicated that both cationic and anionic surfactants adversely affected the dewatering of the conditioned sludge. The addition of surfactant to the oppositely charged polyelectrolyte proved to be most detrimental to sludge dewatering due to the precipitation between surfactant and polyelectrolyte. The addition of cationic surfactant to the cationic polyelectrolyte-conditioned sludge had the least effect.

A one-dimensional mathematical model was developed to simulate pH control using an encapsulated phosphate buffer during denitrification in a sand column. The parameters required for the model were obtained from direct physical measurement, from a tracer study to characterize the dispersion coefficient in the column, and from batch experiments designed to obtain an empirical expression describing the variation of the first-order rate constant for the encapsulated buffer core release with pH. First-order kinetic constants describing the rates of denitrification and ethanol biodegradation were obtained by fitting the model to column runs without the encapsulated buffer. With these parameters, the model was subsequently used to predict the performance of column runs containing the encapsulated buffer. Since denitrification was essentially complete in the sand columns, an increase in the effluent pH was observed. This pH increase was counteracted by the controlled release of the acidic core of the encapsulated buffers added in the columns. The model reasonably predicted the release of the encapsulated buffer core and the performance of the encapsulated buffer for controlling pH in the column.

A nonintrusive method, based on the planar laser imagery technique, was developed to capture images of droplets dispersed in turbulent flow without disturbing the flow field. The methodology and technique of capturing the images, processing, recognition, and measurement were presented. Experimental findings show that this is a workable method in laboratory investigations of droplet dispersion. Results of droplet size distribution obtained and their comparison with those in the literature indicated that the proposed nonintrusive method was successful at capturing fast-moving droplets in turbulent flow without disturbing the flow field. The image processing methodology adopted to recognize and measure the droplets was found to be efficient and effective. The results of the normalized droplet size distribution (number frequency) depicts clearly discernible self-similarity characteristics.

The prediction of air quality parameters is of great interest in environmental studies today due to the health impact caused by airborne pollutants [e.g., sulfur dioxide $\left({\mathrm{SO}}_2\right);$ nitrogen oxide $\left({\mathrm{NO}}_x\right);$ nitric oxide (NO); nitrogen dioxide $\left({\mathrm{NO}}_2\right);$ carbon monoxide (CO); respirable suspended particulates (RSPs), etc.] in urban areas. Artificial neural networks are regarded as a reliable and cost-effective method for prediction tasks. The work reported here develops an improved neural network model which combines both the principal component analysis (PCA) technique and the radial basis function (RBF) network to analyze and predict the pollutant data recorded. In the study, PCA is used to reduce and orthogonalize the original variables. The variables treated are then used as input vectors in a RBF neural network model to forecast the pollutant levels, e.g., the RSP level in the downtown area of Hong Kong. This improved method is evaluated based on hourly time series RSP concentrations collected at the Causeway Bay roadside gaseous monitoring station in Hong Kong during 1999. The simulation results show the effectiveness of the model. For high-dimensional input vectors including simpler network architecture and faster learning speed without compromising the generalization capability of the network, the proposed algorithm has advantages over traditional RBF network learning.

The concentrations of dissolved oxygen and chlorophyll have declined in the Seneca River, New York since 1991 as a result of zebra mussel respiration and filtering. Recently measured data were used to approximate flux rates of zebra mussel respiration, filtering, and phosphorus and ammonia mobilization. A model is developed that for the first time links dissolved oxygen, phytoplankton growth, and zebra mussel respiration and filtering. Good agreement is attained between the model calculations and longitudinal profiles of both dissolved oxygen and chlorophyll measured at low flow during 1997–1999. Flux rates derived from these analyses are consistent with results from other studies when normalized on the basis of zebra mussel number and tissue dry weight. Kinetic formulations for the zebra mussels are developed that describe respiration and filtering rates as a function of density, size distribution, and dissolved oxygen concentration. Several plausible test scenarios are examined using the model that indicate that both zebra mussel numbers and size distributions have a profound effect on the water quality of the Seneca River and therefore have important management and planning ramifications.

Electric arc furnace (EAF) dust is a hazardous waste product of the steelmaking industry with a high concentration of heavy metals, especially Zn and Pb. Existing treatment processes are industrially viable only when the zinc content is sufficiently high. In this research we present an industrial-scale process to enhance EAF dust by recycling it to the same furnace in which it was generated. This process, called RECUMET^®, involves two parts: (1) the manufacturing of self-reducing briquettes composed of dust from the steelmaking process, reducing agents and agglutinant agents; and (2) the loading of the briquettes into the EAF. Several tests have been performed to improve the chemical, mechanical and thermal properties of the briquettes. These include determinations of compressive, breaking and abrasion strength, softening and melting points, and thermal behavior. The theoretical energy balance calculated from the process of loading briquettes into the EAF has yielded a positive value of $\sim 5\times {10}^5\mathrm{kcal}/t$ of briquettes, showing that extra energy is clearly supplied to the furnace. Results corresponding to trials in the EAF are also shown. These include data on taps, slags, energy consumption and the quality of new dust.

Injecting aqueous solutions of hydrogen peroxide $\left(H_2{O}_2\right)$ into hot flue gases can split the peroxide into OH and ${\mathrm{HO}}_2$ radicals. These reactive radicals readily oxidize air pollutants such as CO, VOCs, NO, mercury, and others. $H_2{O}_2$ is thermally “activated” (split into free radicals) rapidly at temperatures of 500°C and above. At lower temperatures, such as found in boiler exhaust flue gases, ultraviolet (UV) light can be used to activate the peroxide molecules. However, placing the UV lamps directly in the flue gases can lead to operating and maintenance problems, and “dilutes” the UV energy due to absorption by other gases. A “UV nozzle” has been developed that produces $H_2{O}_2$ radicals and delivers them into a flowing stream of boiler flue gases. Using a previously constructed pilot scale system at NASA's Kennedy Space Center, experiments were run to prove the concept of the nozzle, measuring the oxidation of NO as an indicator of radical formation and delivery. Data were taken at three temperatures, with none, one, or two UV lamps on, and with various injection rates of peroxide. Flue gas temperatures ranged from 85 to 304°C (186 to 580°F), and the molar ratios (inlet peroxide to inlet ${\mathrm{NO}}_x)$ ranged from about 1.5 to over 15. Conversions of NO varied from 0% (at the lowest temperature tested) to above 50% (at highest temperature). Although increasing temperature had a marked effect on conversion, the activation of hydrogen peroxide by UV light was demonstrated in the temperature range of final flue gas exhaust gases (290–350°F). These results indicate that radicals can be created from hydrogen peroxide at reasonable temperatures using UV light, and that the radicals can be delivered into a flue gas stream where they can oxidize pollutants.

Characteristics of chlorine and carbon flow at two municipal solid waste incinerators (MWIs) equipped with different air pollution control devices in Taiwan are investigated based on sampling/analysis and system operating data. The distributions of chlorine and carbon at various locations were evaluated based on the mass concentrations and emission rates. The removal efficiencies of HCl are 88.3 and 94.1% at MWI-A and MWI-B, respectively. Approximately 60% of the chlorinated compounds (expressed as chlorine) from municipal solid waste (MSW) at both MWIs exists in particulate form after incineration and is captured by the baghouse or electrostatic precipitator. The rest, 30–40% of the chlorine, is removed with bottom ash and wet scrubbing liquid or cyclone ash discharge at both MWIs. The results indicate that approximately 98% of the carbonaceous compounds contained in MSW is converted to ${\mathrm{CO}}_2$ and emitted from the stack at both MWIs due to good incineration efficiency. ${\mathrm{CO}}_2$ emission factors are 845 and 816 kg/ton waste for MWI-A and MWI-B, respectively. The mass percentages of chlorine and carbon in the municipal solid waste are then back-computed. The chlorine contents are 0.53 and 0.72% whereas the carbon contents of MSW are back-calculated as 23.5 and 22.7% for MWI-A and MWI-B, respectively.