Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
92 8.1 Introduction The objective of this chapter is to provide a brief summary of the results achieved during this research project. Various topics have been investigated including: ⢠Characterization of highway runoff ⢠Environmental chemistry of dissolved metals in natural surface waters ⢠Previous work on the removal of dissolved metals in runoff ⢠A laboratory evaluation of the dissolved metals removal for selected materials ⢠Development of conceptual BMP designs for the most promising materials ⢠Identification of future research needs Each of these topics is discussed in more detail below. 8.2 Characterization of Highway Runoff The concentrations of metals in highway runoff are gener- ally higher than observed in natural water bodies, which leads to concerns about potential toxicity. In fact, several studies have demonstrated acute and chronic toxicity to aquatic eco- systems, and metals are commonly listed as a cause of impair- ments in receiving waters in the United States. A fact that is often overlooked is that the distinction between dissolved and particulate associated metals is operational rather than chemically based. Laboratory methods for determining the fraction of a metal in the âdissolvedâ phase are based on the mass passing through a 0.45 mm filter. Clearly, this is a rather arbitrary division, since many particles can be smaller than this threshold. In addition, metal ions can be associated with organic and inorganic ligands rather than existing as free ions. Consequently, dissolved is not synonymous with ionic. A study of metal transport in the Sacramento River indicates that up to 70% of the metals characterized as âdissolvedâ according to standard protocols are actually associated with colloidal particles and may not be bioavailable. Organic constituents in highway runoff also have an important role in determining the bioavailability of dissolved metals and the effectiveness of various treatment processes. These constituents include oil, grease, humic acids, plastics, tire rubber, fecal material, PAHâs, phthalates, pesticides, and herbicides, which are derived from a variety of natural and anthropogenic sources. Stormwater runoff also transports significant loads of dis- solved, colloidal, and suspended particles in a complex mixture that includes metals and inorganic and organic compounds. Knowledge of the PSD and the association of metals with the various size fractions is required to design an effective system for reducing the concentrations of metals in highway runoff. Although small particles have a relatively high SA, the major- ity of the mass of sorbed metals is often found associated with larger particles because of their substantially greater mass. In order to accurately characterize dissolved metals charac- teristics in stormwater, a rigorous protocol is required for both sampling and analysis due to the very low concentrations at which these species impact aquatic life. A protocol was devel- oped as part of this effort and is included as a standalone doc- ument in Appendix A that can be used directly by DOTs and their consultants to assess dissolved metals concentrations. 8.3 Environmental Chemistry of Metals in Natural Waters Chemical speciation in natural waters has significant implications for controlling the mobility, toxicity, and bio- availability of metals in water. Metal ion properties and solution chemistry dictate the speciation and determine the extent of sorption to inorganic, organic, and biological particulate matter. The key processes that must be considered for predicting metal ion speciation are acid/base chemistry, complexation with simple inorganic ligands (CO3, SO4, etc.), C H A P T E R 8 Summary and Conclusions
93 complexation with NOM, oxidation/reduction reactions, precipitation, and sorption. This literature review presented a brief description of each of these processes and identified their impact on spe- ciation of commonly identified metal ions. The review has identified state-of-the-art approaches used to predict metal ion speciation for each of these processes, and has high- lighted interactions among these processes that affect metal ion speciation. The complexity of natural water chemistry was emphasized throughout the review, especially with respect to the heterogeneity of NOM and approaches for incorporating NOM into predictive speciation models. Finally, the impact of metal ion speciation on bioavail- ability was addressed, and a description of the BLM was pre- sented. The BLM represents an approach to incorporate all processes affecting metal ion speciation into a toxicity model for predicting the impact of metal ions on target organisms (the biotic ligand). The bioavailability of metal ions is depen- dent on the speciation in solution, the affinity of metal species for surfaces on organisms, the modes of transport through biological membranes, and metal ion toxicity. The BLM pre- dicts complexation of the metal ion with a specific receptor site within an organism where metal complexation leads to acute toxicity. This model represents the state-of-the-art with respect to predicting metal ion toxicity in natural waters and has significant potential as a regulatory tool. Water quality standards for dissolved metals in many receiv- ing waters are based solely on the hardness of the water, which is a function of calcium and magnesium concentrations. These other metal ions compete for sorption sites on aquatic organ- isms and reduce the toxicity of heavy metals that may be pres- ent. This review has indicated that there are many other factors that affect dissolved metals bioavailability and that need to be considered when developing water quality standards. Conse- quently, many current standards may be overly conservative and need to be reconsidered in light of this information. 8.4 Previous Studies of the Treatment of Dissolved Metals Previous studies concerning the removal of dissolved met- als, identification of candidate materials for evaluation, and how these materials could be included in conceptual designs for stormwater treatment systems were reviewed. Treatment of runoff transporting particulate-bound metals generally requires sedimentation, coagulation/flocculation, filtration, or combinations thereof. In contrast, dissolved metal complexes and metal ions can be more effectively removed using adsorp- tion, ion exchange, or surface complexation mechanisms. A variety of in situ structural BMPs that are designed for infiltration to surrounding soil have been developed that remove metals. Infiltration systems are adaptable and poten- tially can remove dissolved and particle-bound constituents. Benefits of these infiltration BMPs are increased mean resi- dence times relative to pre-BMP conditions, and promotion of high discharges to surrounding soils, even for soils with low saturated hydraulic conductivity. Infiltration BMPs in soils of lower hydraulic conductivity (typically clayey materials of higher SA and surface charge) have a reduced risk for soil and groundwater contamination resulting from increased loadings of pollution in urban and highway runoff, especially heavy metals. Metal adsorption onto natural and engineered media, such as sand, soils, activated carbon, and oxide-coated media, has been studied by many researchers. Results from this previ- ous work demonstrated that plain sand or gravel media has insignificant SA and is ineffective for removal of dissolved stormwater metals. Consequently, researchers have often aug- mented the sand filter media with organic matter, perlite, and zeolite, all of which have shown only modest effectiveness for dissolved metals removal. Oxide-coated metals have long been used for dissolved metal adsorption. These mineral surfaces are amphoteric (net surface charge is a function of pH). This leads to the idea that engineered amphoteric oxide-coated surfaces on stormwater filter media can have large SAs for adsorption and precipita- tion processes in a chosen pH range. Therefore, engineered media such as the oxide-coated filter substrates or cementi- tious media with high SA can be utilized to remove dissolved metals in a variety of treatment configurations. A common choice for sorption of dissolved metals is iron oxide coated media. Previous research with respect to per- formance of iron oxide-coated media for sorptive-filtration of stormwater has demonstrated proof-of-concept for oxide- coated media in stormwater treatment systems. Engineered media also include manganese oxide-coated polymeric and cementitious media. MO have a PZC below the typical pH range of runoff resulting, allowing their use as adsorbents without pH control. Cementitious media such as graded concrete, masonry, rubble, or brick, when coated, sprayed, or synthesized with oxides or oxide-admixture, can be effective for metal removal. The idea of using cementitious infrastructure material or recycled cementitious material is based on three advantages beyond the inherent economy of the substrate. First, cementi- tious material increases the alkalinity. This alkaline environ- ment will not only improve the metal removal efficiency by precipitation of dissolved metals, but also benefit from the electrostatic interaction between the oxide and dissolved cat- ionic metal, because MO have larger negative surface charge in higher pH solutions. A number of recent studies have evaluated low cost adsor- bents for metal ion removal from stormwater. While many of these have little potential for widespread application due
94 to impacts on pH, headloss, and clogging, one particular material that has demonstrated potential for application to stormwater treatment is crab-shell waste. The combination of chitin and calcium carbonate has been proposed as an ideal media combination because the calcium carbonate promotes the formation of metal carbonate microprecipitates and chitin is a good adsorbent for these microprecipitates. Based on this review of the literature, the following materi- als were recommended for further evaluation for the removal of dissolved metals in highway runoff. ⢠Iron and MO have demonstrated the highest potential for adsorbing the range of metals observed in highway run- off. Therefore, commercially available forms of these oxide metals should be evaluated as well as mixtures containing both types of media. The pH of the PZC is different for these two oxides. Therefore, both types of media should be tested over a range of pH values for a variety of metals that exhibit varying affinity for the oxides. Concentration ranges to be tested should be consistent with the concen- trations observed in stormwater. ⢠Concrete materials and calcite offer the possibility of pro- viding adsorptive capacity as well as pH buffering. Both of these materials should be assessed for their potential for adsorption of metal ions. While calcite represents a well- characterized and less variable material, concrete provides the opportunity to recycle waste materials. Preliminary experiments with concrete should include evaluating dif- ferences among different concrete sources including con- crete rubble, crushed concrete, or recycled concrete. ⢠Crab shell has also been shown to have potential for removal of metal ions from stormwater due to the presence of chi- tin and calcium carbonate. Thus, these sorbents should be evaluated for their potential to synergistically promote adsorption of metal ions. 8.5 Laboratory Evaluation of Dissolved Metals Removal The goal of this effort was to develop and verify labora- tory testing and modeling methods to determine the effec- tiveness of adsorption for reducing the dissolved metals concentration in stormwater. One of the major challenges of this project was in selecting the composition and sources of the synthetic stormwater to be used during the experimental testing phase. Previous research has employed both natural stormwater that has been obtained from actual field highway runoff sites (natural stormwater) as well as synthetic storm- water compositions that have attempted to mimic the major components of natural stormwater. A compromise between using natural and synthetic storm- water was made by recognizing the importance of capturing the organic matter from natural stormwater, but adding the flexibility of using synthetic stormwater to provide the ionic composition. To alleviate concerns associated with storing large volumes and aging of organic solutions, the stormwater was concentrated within 24 hours of collection using reverse osmosis and then the organic matter freeze-dried. The freeze- dried organic matter was reconstituted as needed at concen- trations that mimic the initial total organic concentration of the stormwater when it was collected. Two types of experiments were conducted to evaluate dis- solved metal behavior: batch experiments and column experi- ments. Batch experiments were used to evaluate sorption equilibria over several orders of magnitude in initial metal ion concentration at a fixed pH. Experiments were conducted with each metal ion individually in single solute systems as well as in multi-solute systems. The data from these models were then used to calibrate and verify surface complexation model parameters that can be used to predict the extent of sorption over a range of solution conditions such as pH, ionic strength, and solution compositions. Upon completion of the batch equilibrium experiments, bench scale flow-through columns packed with the selected media were used to assess the rates of adsorption and to verify the equilibrium studies. These data will be used to calibrate dynamic models or used in scaling approaches to predict performance at field scale. A number of issues were identified and addressed while con- ducting the laboratory experiments that will be incorporated in the recommended protocol for future testing. An example of one of these issues is the difficulty in controlling pH during each experiment. pH is by far the largest factor affecting the solubility, adsorption, and precipitation of dissolved metals in runoff, so it is of paramount importance that it be main- tained at a constant level during the column experiments. It was observed that even exposure to the atmosphere will cause the pH to drift as carbon dioxide is absorbed. Consequently, a nitrogen blanket over the synthetic stormwater to manage the pH level had to be provided. The experimental work evaluated two metal oxides, iron and manganese, both of which are available commercially. The results indicate that the iron oxide provided substantially better removal of dissolved metals under the aerobic condi- tions of this study than manganese oxide. The high removals reported for manganese dioxide in previous experiments were likely the result of the cementitious substrate on which the manganese was deposited. The two metal oxides in conjunction with crab shell waste and recycled concrete were also evaluated. Both materials raised the pH at least briefly at the start of an experimental run, which increased zinc removal, but had little effect on copper adsorption. The crab shell and recycled concrete also tended to stabilize the pH, which would be beneficial in the design of a passive treatment system for metal ion removal,
95 since the pH of rainfall can vary substantially. Another impor- tant point of interest is that the pH of the system increased for a short period of time after the flow was interrupted for more than 24 hours. This type of behavior may result in signifi- cantly better removal in field scenarios and should be evalu- ated further. The research also found that DOM significantly reduces dissolved metal sorption and should be included in any syn- thetic stormwater cocktail. In this study, we first concentrated DOM in highway runoff and then had it freeze-dried to reduce the potential changes in form that might occur over time. Re-dissolving the organic matter was not as straightforward as just adding water, however. This difficulty was addressed by including a sonication step, but in the future, it is recom- mended that freeze drying this material be avoided, since the changes that might occur during this process could easily be as large as those caused by aging. It was also demonstrated that the laboratory results were in good agreement with model predictions. There are com- plex models such as SCMs and ligand binding models that can be incorporated into transport codes and are useful for evaluating the impact of various parameters on adsorption; however, the results of this research suggested that pH control is the most important operating parameter for stable opera- tion of a treatment system. Consequently, a relatively sim- plistic and user-friendly model, CXTFIT/EXCEL, was used to model these experimental results. This model can be used in a predictive mode by adjusting the hydrodynamic param- eters associated with the field scale conditions that should be evaluated in future phases of research. Runs were also conducted using actual stormwater runoff collected from a highway in the Austin, TX area. Removal of dissolved copper was much worse than expected. This is likely attributable to the high alkalinity and pH of the runoff, since in this environment much of the copper is present in a relatively inert form. By simulation with Visual MINTEQ, it was determined that under the conditions found in our actual stormwater, over 80% of total copper exists as dissolved copper carbonate, which does not directly sorb to GFO, and less than 3% of total copper is present in the form of cupric ion. Consequently, adsorption of dissolved copper will not be a very effective process in runoff with substantial carbonate concentrations. Conversely, zinc removal was relatively more successful as compared to the tests utilizing synthetic storm- water. This can be explained by the high pH of the stormwater, which increases the tendency of zinc ion to sorb on to the surface of the oxide media. Processes that increase precipita- tion of carbonates could reduce the carbonate concentration to levels that would substantially increase Cu sorption. Finally, some calculations to determine the amount of iron oxide media that would be required for a system with an operational life (the period over which a dissolved cop- per concentration of less than 3 mg/L is achieved) of 10 years were made. The calculations indicate that the footprint of the system would not be excessive (approximately 30 square feet) and the cost of media to provide 10 years of treatment for the runoff from one acre of pavement getting 30 inches of rain/ year could be as little as $6,400/acre. 8.6 Stormwater Treatment Conceptual Designs The conceptual designs for the treatment devices were based on the opportunities and constraints identified in the literature review for candidate material selection. The can- didate materials have a variety of preferred operating condi- tions that were accommodated in the prototype designs. The designs targeted compliance with the most restrictive receiv- ing water standards for dissolved metals. The prototype designs are passive, practical, and able to operate with deferred maintenance within the constraints of a highway environment. Some of the practical constraints identified in previous treatment BMP studies for highways (Caltrans 2004) that were used include: ⢠No obstructions (above-ground obstacles and slopes greater than 4:1) within the clear recovery zone (30 feet from edge of traveled way) should be avoided. ⢠Maintenance access with adequate room for maintenance vehicles and equipment must be available outside of the traveled way and shoulder areas. ⢠Passive operation required (a minimum of mechanical and powered components). ⢠Inspection should be non-destructive and require mini- mum time and training of personnel. ⢠Maintenance intervals should be reasonable (annual is a preferred minimum). ⢠The prototype design should be compatible with high- way drainage systems and require a minimum amount of physical space. Three configurations have been developed to accommo- date the majority of the design scenarios that DOTs will need: highly urban areas, more rural locations with vegetated shoul- ders, and on bridges. The urban area design was developed for highway cross sections that are largely impervious within the ROW. In the urban configuration, runoff is conveyed along the curb to a drain inlet that discharges to a longitudinal drainage system. The drainage system will typically discharge to a cross culvert (transverse to the highway centerline) at a location joining the adjacent municipal storm drain system. A below- grade vault design has been developed for urban locations that are compatible with the longitudinal drainage system, to pre- serve surface use and avoid above-ground obstructions.
96 Locations in rural areas with vegetated shoulders may also have a curbed section and a longitudinal drainage system, but in many cases the paved shoulder abuts a vegetated buffer strip that discharges to a longitudinal vegetated swale. The advan- tage of this type of design is that runoff may be intercepted in a sheet flow condition and vegetated strips are effective for pre- treatment to remove solids in runoff. The prototype design is also configured to work with the MFD developed by WSDOT. A third configuration has been developed for bridge decks. There may be some instances where metals removal from bridge deck runoff is desirable. Piping water to the bridge abutment is costly and introduces aesthetic and maintenance issues. The conceptual design integrates the media into the bridge scupper to treat bridge deck runoff. One of the major needs when installing systems for dis- solved metals removal was identified as the need to provide pretreatment to reduce solids loadings on the systems. Poten- tial pretreatment options were identified as vegetated filter strips, MFDs, and the PFC. The feasibility of each of these options is to a large extent a function of the available ROW. A sizing spreadsheet was also developed that provides capi- tal, O&M, and whole life cost estimates using userâprovided unit prices. The user can overwrite the default unit prices for GFO, poured in place concrete, and the cost to remove and dispose of spent filter media on a per vault basis. Locations with high metals loading rates, and high average annual rain- fall will require larger amounts of media and more frequent maintenance intervals than those areas with lower metals loading and lower annual rainfall. Annual loading rates for TSS and metals will be the most important considerations when sizing the vault or bridge scupper. 8.7 Recommendations for Future Research Finally, a number of items were identified for future research. These include: ⢠Field testing of GFO conceptual BMP designs ⢠Characterization of highway runoff ⢠Modification of PFC The most obvious need is a field testing program for the conceptual BMP designs. There are two primary objectives. One is to validate the laboratory evaluation protocol to verify that the results are applicable in the field setting. The second objective is to determine the efficiency of pretreatment pro- cess for reducing the solids load on the adsorptive media and how the presence of the remaining solids impacts the adsorp- tion of dissolved metals.
