Metals Treatment and Removal Mechanisms

1. Effluent Influent Glass aquarium S2- Aerated Tap Water 18” Substrate 12” 12” Dialysis Tubing Mixer 3” 6” 26” PbS Gravel Chemicals Plexiglas screen Wetland Media Acknowledgements This work was funded by the Midwest HSRC (National Center for Environmental Research STAR Program, EPA) and Doe Run Mining Company. The authors acknowledge Bill Duncan, Teck Cominco Mining, Trail BC, Canada for data and support. We also thank Cem Selman, Rebecca Leerssen, Ying Song, and Chad Ross for assistance. Influent (DOM) M-DOM M - DOM 40-100 ppb 5-10% M2+ M+2 M+2 1-5% MCO3 M - FemOHn S-2 Fe+2 53% S 47% Pb M M-Particles 30-70% ? reversibly 1-5% adsorbed MS SO4-2 SRB Effluent 20-40% 10-15% Wetland substrate Gravel lens Neutral Mine Drainage Treatment with Constructed Wetlands Mark W. Fitch, Joel G. Burken, Chang Ye Department of Civil, Architectural and Environmental Engineering; University of Missouri-Rolla, Rolla, Missouri Abstract: Heavy metals at low levels in mine drainage, mine tailings leachate, and industrial wastewater is a national issue. In Missouri, the legacy of lead mining during the last 130 years includes several Superfund sites. Constructed wetlands pose a passive treatment system that may provide efficient, sustainable treatment for long periods at such sites. Laboratory scale wetlands have been operating for 6 years, sustaining effective lead and zinc removal and reducing toxicity to acceptable levels. Removal mechanisms include: sorption to the wetlands media, precipitation as metal sulfides, and co-precipitation with iron oxy/hydroxides. Any of the three mechanisms can dominate under different design and operating conditions. Tests on the wetland media following years of metals accumulation show that small fractions of the metals can be remobilized under certain conditions. Physical disturbance of the media resulted in elevated metals being released in a colloidal fraction, whereas a drying period resulted in an increase in the soluble metals, likely do to oxidation of metals sulfides. Findings overall show that constructed wetlands can efficiently and effectively treat alkaline mine drainage and process waters, and design and operations are important in optimizing treatment and long-term stability. Metals Treatment and Removal Mechanisms Overview: Treatment wetlands were prepared in the laboratory, Figure 1. Media used in the wetland were partially based on local availability and low cost and included: chip bark, alfalfa, composted manure, silica sand, and peat moss. Removal has averaged 90% for Pb and 75% for Zn overall and been steady over 6 years. Individual removal mechanisms were investigated in laboratory experiments, including sulfide generation and precipitation of metal-sulfide precipitates, adsorption isotherm determinations and sequential extraction and microscopic analyses of media from operating wetlands. Adsorption: Greater sorption potential than previously thought was discovered. Individually, wetland components had higher sorption capacities than had been shown, Figure 2. Recent kinetics testing suggests that sorption may be more rapid, reaching apparent equilibrium in as little as 12 hours. Sulfide Precipitation: Sulfide generation and metal-sulfide precipitation potential was tested using a dialysis membrane to isolate media and surface phenomenon. Sulfide was actively generated by the wetland media, and lead was repeatedly removed from solution, Figure 3. Sulfate reducing bacteria (SRB) were identified in the laboratory wetlands. Using Scanning Electron Microscope-Energy Diffraction Spectrometry (SEM-EDS) zinc sulfide and lead sulfide, Figure 4, were positively identified in wetland media. Iron Coprecipitation: Lead and zinc rapidly coprecipitated with iron oxy/hydroxides. Iron oxides were clearly present in the laboratory wetlands. Sequential extraction showed that up to 30 to 70% of Pb and Zn were removed in the gravel lens and the initial portion of the wetland media. Coprecipitation was highly variable on iron content and aeration time. Summary: Overall, removal of zinc and lead was consistent over 6 years in laboratory wetlands. The effluent had a negligible toxicity to fathead minnows and C. dubia. Estimates on relative fates were generated, Figure 5. One remaining question is the potential precipitation of metals following desorption from the wetland media if the local metals concentration is decreased in the presence of sulfide. Kinetics suggest that sorption is much more rapid, yet sulfides were readily located. Figure 1Schematic of laboratory wetlands • Objectives: • Investigate the relative fates of Pb and Zn in constructed wetlands.. • Elucidate removal capacity and failure modes of constructed wetlands treating Pb and Zn. • Determine the impact of disturbances on the sequestered metals. • Develop design and operational information on wetlands treatment of mine drainage and process waters. Figure 2 Sorption to wetland media constituents Disturbance Testing: Release of Metals Rationale: Various forms have differing stability and are stability is environmentally dependent. When environmental parameters such as pH or redox change, different forms may not remain sequestered. Some of these environmental changes such as reconstruction work, bioturbation by rooting animals, or drying from drought conditions or drainage can be abrupt and could potentially result in a high concentration, short duration release of the sequestered metals. The uncertainty of the effects of disturbances on the fates of fixed metals in constructed wetlands could stifle application. Testing: Wetland media was taken from the 6 year old wetland, separated and placed into two new wetlands, similar to Figure 1. The two were subjected to different tests; 1. Physical disturbance (i.e. repeated mixing) and 2. Drying. Following the perturbations, influent was returned and the effluent was tested for both filtered and unfiltered Pb and Zn levels. Findings: Effluent metals concentrations in the physical disturbance test were elevated in the unfiltered samples, Figure 6. The particulate form of the metals suggests that sequestered metals were not chemically altered, but rather the release of colloidal sized particulates, as is likely as colloidal lead sulfides and lead-DOM complexes have been identified, Figures 4 & 5. The release was very brief and repeated mixing did not result in continued release, showing that the sequestered metals were stable. Effluent from the drying test was elevated in dissolved metals, reaching up to 450 ppb. The release was quite brief as in just a few pore volumes, the lead levels dropped back below the average influent concentration, Figure 7. The brief duration of the impacts suggest that wetlands are robust in their functionality and can quickly recover after anticipated perturbations. Figure 6 Effluent from physical disturbance test. Closed symbols are unfiltered samples, open symbols are filtered Figure 4 Lead-sulfide Via SEM - EDS Figure 3 Sulfide precipitation Figure 5 Composite of metals’ fate in lab scale wetlands • Conclusions & Recommendations • Constructed wetlands can provide sustained removal of Pb and Zn and reduce whole effluent toxicity (WET) with minimal operations and maintenance. • Metals’ sequestration occurs via three primary mechanisms: Adsorption, Sulfide precipitation & Iron oxy/hydroxide coprecipitation. Multiple, complimentary mechanisms result in the observed reliability. • Adsorption capacity is higher than previously thought, related to wetlands’ composition. • Sulfide precipitation was observed, but was slower than thought and appears to be of lesser importance than previously published. • Remobilization is a minimal concern after the wetlands have sequestered the metals and can be mitigated by avoiding dry conditions and major disturbances of the wetland. Figure 7 Effluent from drying test for filtered and unfiltered samples