A tubular membrane (MWCO = 200 k, PVDF, negative surface charge) and a hollow-fiber (HF) membrane (MWCO = 500 k, PS, neutral surface charge) underwent testing to determine their effectiveness in treating a sterilized wastewater from a dental clinic. The tubular membrane, which had a lower MWCO, produced a higher permeate flux due to the increased turbulence at the membrane surface that was possible with the tubular configuration (Reynolds numbers for the tubular module were over five times greater than the HF module). Both membranes had similar Hg rejections (tubular-97%, HF-98%) and permeate Hg concentrations (less than or equal to10 mug/L) due to particle size distribution of dental wastewater-the vast majority of the amalgam particles were larger than the pore size of either membrane. Permeate Hg was not a function of temperature, feed concentration, and intermittent operation. Because of their unique behavior and characteristics, both the HF and tubular modules will be tested at an actual dental clinic using non-sterilized dental wastewater.

Wastewater containing about 0.5% oil and grease (O/G) from a metal industry was treated by tubular ultrafiltration using membranes having a molecular weight cutoff (MWCO) of 120,000 and a negative surface charge (P membrane) and of 100,000 and no surface charge (M membrane), Permeate flux decreased dramatically during the first several hours of operation and then leveled-off for the remainder of semibatch operation. The average P membrane flux was significantly higher than the M membrane (38 versus 27 gal/ft(2).d) because of its higher MWCO and negative surface charge. Increasing the transmembrane pressure and crossflow velocity increased the permeate flux for both membranes. O/G concentrations less than 50 mg/L and total suspended solids (TSS) levels less than 25 mg/L were common for both membranes. O/G removal efficiencies (rejections) averaged 98% for the M membrane and 97% for the P membrane. TSS rejections were approximately 97% for both membranes. Effluent O/G concentration and turbidity from the P membrane were slightly higher than the M membrane because of the P membrane's higher MWCO and the larger flux. The average volume reduction and residual production were 97% and 32 gal/1000 gal, respectively, Acid cracking of the concentrate with sulfuric acid was marginally successful.

The SpinTek rotary ultrafiltration system (ST-II) uses membrane rotation to provide the turbulence required to minimize concentration polarization and flux decline. The ST-II system was effective in concentrating oily wastes from about 5% to as high as 75%. The decoupling of turbulence promotion from feed pressurization/recirculation by rotating the membrane is the primary reason for the improvement in performance over that observed with conventional UF systems. Flux increased by about 45% when the temperature was increased from 110 to 140 degrees F. A larger decrease in waste viscosity, over that predicted for water alone, was the primary reason for the stronger than expected flux-temperature relationship. The flux decreased with decreasing rotational speed (omega) and the gel layer did exhibit some sustainable stability with increases in omega. A ceramic membrane was superior to a polymeric membrane in regards to flux quantity and quality as well as membrane cleaning/durability.

Wastewater from metal industry hot milling operations contains oil primarily from coolant sprayed on metal strips to dissipate heat during the rolling of metals. The emulsified oil wastewater for this study was withdrawn from two holding ponds where free oil would rise to the surface and was removed periodically, and used as influent for the chemical addition (CA)-gravity separation (GS) process. The principal objective of this research was to evaluate the effectiveness of different polymer addition systems through a CA-GS process for oil/grease (O/G) and total suspended solids (TSS) removal. Polymers from two corporations were investigated. A dual polymer system was recommended and supplied by Calgon Corporation—a cationic polymer (W-2923) to break the emulsion and an anionic polymer (POL-Z-E 2706) to enhance coagulation. A single polymer system was recommended and supplied by Grace Dearborn (GD) Corporation—a cationic polymer (KLAR-AID 2400) as a coagulant. Two types of experiments were performed: jar tests and larger scale batch-mixing tank (BMT) experiments. O/G and TSS removal efficiencies were 99% for both the Calgon and GD polymer systems. The range of optimum dosage was broader for the Calgon polymers, but both systems were susceptible to overdosing. A longer settling time was required for the GD polymer, but the Calgon polymers were more sensitive to pH. Average residual production rates were 89 and 148 gallons per 1000 gallons of wastewater treated for Calgon and GD polymers, respectively.

A company that produces an oily waste came under a consent order to upgrade their wastewater treatment facilities. WV University's Department of Civil and Environmental Engineering was asked to perform pilot-scale studies using the following treatment technologies: chemical-addition-dissolved air notation (CA/DAF); ultrafiltration (UF), biological filter, constructed wetlands (CW), and land application. The focus of this paper is the comparison of CA/DAF and UF treatment performances. The UF effluent had an O/G content that was consistently below 100 mg/L and a very low TSS. Effluent from CA/DAF was of similar quality but only after extensive jar testing followed by modification of the chemical(s) dosage. Because of the labor intensive nature of the CA/DAF system, the company chose the UF system as its secondary treatment system. UF effluent will be sent to a 15 acre hybrid CW/land application system for tertiary treatment, The UF residual (2600 gal/d, 5% oil) will be concentrated by a factor of 10 (260 gal/d, 50% oil) using the SpinTek high-shear rotary UF system. The highly concentrated oil waste will be disposed of via an off-site oil recycler.

Flushing of a Pb(II) synthetically contaminated sandy loam using 0.1 N HCl, 0.01 M EDTA, and 1 M CaCl2 was investigated in the continuous-flow (column) mode. Initial Pb soil concentrations ranged from 500 to 600 mg/kg. Pb-removal efficiencies (and final soil Pb concentrations) for HCl, EDTA, and CaCl2 were 85% (77 mg/kg Pb), 100% (approximate to 0 mg/kg Pb), and 78% (135 mg/kg Pb), respectively. Given that HCl and CaCl2 could not reduce the soil Pb to background levels (approximate to 25 mg/kg) for a synthetically contaminated soil brings into doubt their effectiveness in remediating actual contaminated soils where metal retention would be much stronger. While EDTA removed all lead (indigenous and ''artificial''), its treatment and reuse and possible adverse health effects makes its use difficult. The final soil pH for HCl was near 1, raising the concern of increased contaminant mobility, decreased soil productivity, and adverse changes in the soil's chemical and physical structure due to mineral dissolution. Final soil pHs for EDTA and CaCl2 ranged between 4.85 and 5.2.

Batch and column studies were performed to determine the effect of pH, empty bed contact time (EBCT), and hydraulic loading rate (HLR) on lead removal by granular activated carbon (GAC) columns. Lead removal increased with increasing pH, and for the majority of the adsorbate:adsorbent ratios investigated, was 100 % at pHs < pH prec . Column pH was extremely important to lead removal in GAC columns. A simple acid‐base regeneration procedure was found to be effective in desorbing/resolubilizing the carbon‐bound lead and raising the pH for the subsequent treatment run. Regeneration efficiencies were often less than 100%, but column performance was not adversely affected. For 1 mg/L lead, the optimum EBCT was between 6 and 7 minutes, whereas for 10 mg/L lead it was less than 10 minutes. The effect of HLR (4.9 and 9.8 m/hr) on column performance was minimal for 1 mg/L lead, whereas at 10 mg/L lead column removal was slightly better at the higher HLR. Carbon usage rates were higher than those observed for wastewaters containing organic contaminants, especially at 10 mg/L lead. However, given the relatively simple regeneration scheme, the applicability of GAC columns for metal‐bearing wastewaters appears to be technically feasible.

The in situ remediation of a lead-contaminated silt loam by electrokinetic (EK) soil flushing was studied. Two initial soil Pb concentrations (150 and 1000 mg/kg of Pb) and applied voltages (30 and 60 V) were investigated. The EK soil flushing process was less efficient for the 150 mg/kg of Pb soils despite these tests being operated for longer durations, having larger EO flows and energy inputs, and lower soil pHs. The decrease in effectiveness was attributed to a larger average metal-soil binding energy for the lower contaminated soil. Increasing the voltage increased the EO flow, current, energy input (kW-hr/kg of soil), and provided a more evolved low pH front, resulting in more soil being remediated. There appeared to be a correlation between the amount of EO flow and the desorption and transport of soil-bound lead. Because complete soil remediation did not occur in any of the tests, the final energy input per kilogram of soil could not be calculated.

Granular activated carbon (GAC) columns that were used for Pb removal were regenerated using several regeneration schemes. Acid type (HCl and HNO3), regenerant concentration (0.1 N or 1 N), acid reuse, and a NaOH-only rinse were studied. Column performance was assessed using the Pb surface loading (X/M) at exhaustion, and the number of bed volumes (BVs) treated at breakthrough and exhaustion. The type of acid used did not affect column performance. Increasing the regenerant concentrations from 0.1 N to 1 N increased the surface loading at exhaustion and the BVs treated at breakthrough by 20% and 25%, respectively. Acid reuse did not significantly affect column performance, although desorption efficiencies decreased with the run number. The insensitivity of Pb removal to desorption efficiency indicates that the dominant removal mechanism is precipitation on the carbon surface or in the carbon pore, not adsorption. All measures of column performance deteriorated slightly when base-only regeneration was employed. However, column performance for the 1 N NaOH experiments were similar to the 0.1 N HCl-0.1 N NaOH studies.

In an earlier study the removal of lead by a granular activated carbon (GAC) column was increased by over 600% when the carbon ]Hydrodarco 4000 (HD4000), Norit Americas, Inc.] was contacted with a 0.1 N HNO3-0.1 N NaOH rinse. Hypothesized removal mechanisms were adsorption, surface preceipitation, and pore precipitation. In this work a series of experiments were conducted on samples of the virgin and acid-base rinsed carbon to determine their acid-base behavior, pH(zpc), and Pb removal ability. If adsorption was a dominant removal mechanism, then significant differences in these parameters for the virgin and acid-base rinsed carbons would be expected. The strong acid-base rinse did not significantly alter the acid-base behavior, pH(zpc), or the Pb removal ability compared to virgin HD4000. Thus, it appears that the dramatic increase in metal removal by the regenerated GAC columns was not caused by an increase in the number or type of adsorption sites but was due to the precipitation of Pb on the carbon surface or in the carbon pore liquid. Future research efforts will focus on the modeling of Pb removal in GAC columns using precipitation as the primary removal mechanism.

The in-situ remediation of a lead-contaminated soil (silt loam, K-H = 5 x 10(-8) cm/s, soil Pb = 1,000 mg/kg) by electrokinetic (EK) soil flushing [60 V (DC)] was studied. Research focused on the chemical conditioning of the electrode reservoirs with either 500 mu S/cm (as NaNO3, baseline behavior), acetic acid (HAc), HCl, or EDTA. For baseline tests there were significant amounts of lead transported through the soil, but the Pb precipitated or was readsorbed on the soil adjacent to the cathode because of the high soil pH in that region. The addition of 1 M HAc to the cathode reservoir prevented the formation of the basic conditions in the soil, and about 65% of the Pb was transported into the cathode. When HCl was added to the anode and HAc was added to the cathode, more than 75% of the lead resided in the cathode. Pb removals in the EDTA-experiments were greater than those observed in the baseline experiments and were similar to those observed in the HCl-HAc experiments. A low anode reservoir pH resulting from a high current was the most likely reason.

The use of granular activated carbon (GAC) columns to treat lead wastewaters was investigated. Synthetic wastewaters containing Pb (10 or 50 mg/L), acetic acid (0.001 N) or EDTA (1:0.1 or 1:1 Pb:EDTA molar ratios) were studied. For lead-only and lead-acetic acid experiments significant quantities [70 to 325 bed volumes (BV)] of wastewater were treated prior to breakthrough (C(e) = 0.03C(o)); X/M-values were as high as about 30 mg Pb/g carbon. For EDTA experiments, C(e) was always >0.03C(o). The amount of lead not removed corresponded to the amount that was complexed by EDTA. Column pH is the critical parameter influencing column performance. The increase in effluent Pb concentration corresponded with the decrease in column pH. GAC columns were successfully regenerated using a 1 L (almost-equal-to 8 BV) 0.1 N HNO3 rinse followed by a 1 L 0.1 N NaOH rinse. Column performance was not adversely affected by regeneration. When the regeneration step was used on virgin carbon, a dramatic improvement in column performance was observed and was attributed to the increase in carbon surface pH (pH almost-equal-to 11) and the deposition of OH - in the pore liquid. Possible removal mechanisms are precipitation of lead on the carbon surface, precipitation in the pore liquid, and adsorption (surface complexation).

The objective of this research was to investigate lead removal efficiencies from various soils using a variety of washing solutions. Most soil types have a strong affinity for lead. Thus, it is plausible to expect washing solutions that are capable of removing lead could also remove other divalent heavy metals. Four soil samples from the eastern US were collected and characterized for this study. The study soils were then spiked to approximate lead concentrations of 1,000 and 10,000 mg Pb/kg soil. The efficiencies of six washing solutions in removing lead from the contaminated soils were then investigated via lab-scale batch washing experiments. Unlike current field-scale soil washing practices, all particle size fractions were washed and recovered in these experiments. (Solutions investigated include: tap water, HCl, EDTA, HNO{sub 3}, CH{sub 3}COOH, and CaCl{sub 2}.) In order to examine the effect of aging upon soil washing efficiencies, some of the spiked soils were washed a second time after an aging period of nearly 2 years.

ABSTRACT: A contaminated, iron oxide coated (15.5% Fe203), sandy soil was studied to evaluate the effectiveness of conventional soil‐washing extractants for metal removal. Metals of interest included As, Cu, Pb, Hg, and Zn. The extractant solutions were HC1, HN03, H2S04, EDTA and NH2OH · HC1. Experiments assessed effects of extractant strength and contact time. Individual soil size fractions were studied for the various extractants. Metal‐binding mechanisms were evaluated using a sequential extraction procedure. Metals in all size classes were bound strongly, presumably by an Fe oxide coating observed on the sandy soil. From an analysis of particle size fractions, contaminated soil metal concentration correlated approximately with the surface area/volume ratio of soil particles, suggesting that the metals were associated with soil surfaces. During soil washing, a typical rapid metal release in HC1 occurred initially, followed by a much slower step. The fraction associated with the slow metal release correlated reasonably well with the residual metal fraction. Although sandy soils often are good candidates for soil washing, surface coatings may make metal extraction for even sandy soils difficult.

The use of granular activated carbon (GAC) columns to treat metal-bearing wastewaters was investigated. Synthetic wastewaters containing Pb and Cd (10 or 50 mg/L), acetic acid (0.001 N) or EDTA (1:0.1 or 1:1 Me.EDTA molar ratios) were studied. For metal-only and metal-acetic acid experiments, significant quantities (as high as 325 bed volumes (BV)) of wastewater were treated prior to breakthrough (C(e) = 0. 03 C(o)). X/M values were as high as about 30 mg Pb/g carbon. For EDTA experiments, C(e) was always > 0.03 C(o). The amount of metal not removed corresponded to the amount that was complexed by EDTA. Column pH is the critical parameter influencing column performance. The increase in effluent metal concentration corresponded with the decrease in column pH. GAC columns were successfully regenerated using a 1 L (almost-equal-to 8 BV) 0.1 N HNO3 rinse followed by a 1 L 0.1 N NaOH rinse. Column performance was not adversely affected by regeneration. When the regeneration step was used on virgin carbon, a dramatic improvement in column performance was observed and was attributed to the increase in carbon surface pH (pH almost-equal-to 11) and the deposition of OH- in the pore liquid. Possible removal mechanism are precipitation on the carbon surface and in the pore liquid, and adsorption.

Adsorption of cadmium by two powdered activated carbons (PAC) is investigated in single and binary adsorbent systems. Both PACs are effective in removing cadmium from solution. Cadmium adsorption increased with increased solution pH. The surface complex formation (SCF) model, a surface-solution chemical equilibrium model, successfully simulated the individual pH-adsorption edges. Cadmium adsorption decreased with increased cadmium surface loading. This phenomenon was attributed to the surface sites having a range of binding energies. Reflecting this phenomenon, several of the complexation constants decreased with the cadmium/carbon ratio. The SCF model successfully simulated the cadmium surface loading versus soluble cadmium curves for the entire range of cadmium and carbon concentrations studied when the cadmium-carbon complexation constants are modified to account for their variation with the cadmium/carbon ratio. Cadmium adsorption in the binary PAC system was observed to follow ideal behavior. The SCF model, using complexation constants determined from single PAC experiments, successfully predicted cadmium removal from solution.

Powdered activated carbon (PAC) can behave as a weak acid in solution. Historically, the amphoteric nature of hydrous solids has been modeled as a single weak diprotic acid. To determine if the single diprotic representation was the most appropriate description of the PAC surface, two commercially available PACs underwent acidimetric-alkalimetric and NaNO3 titrations. The surface complex formation (SCF) model, a variation of a soluble equilibrium chemical model, was used to simulate experimental titration curves using several surface representations. A three-monoprotic site model was chosen to represent the amphoteric nature of both PACs studied, based on the following criteria: the goodness of fit of model results; the coherence of the representations to previously published results; and the agreement of surface site representations at different ionic strengths for the same PAC. Surface acidity parameters (pH(zpc), acidity coefficients and number of surface sites) and surface speciation diagrams are presented.