«February 2014 Final United States Environmental Protection Agency Office of Solid Waste and Emergency Response Office of Resource Conservation and ...»
The evaluation retained all other COPCs for further consideration. Although only a small amount of data on concentrations of some COPCs in portland cement was available for comparison (e.g., aluminum, boron, molybdenum), the evaluation found these COPCs to be higher in fly ash and retained them for further consideration. The current evaluation considers the consequences of incorrectly accepting the null hypothesis (H0) and removing a COPC from consideration (i.e., Type II Error) more severe than incorrectly rejecting H0 and retaining the COPC for further consideration in subsequent stages of the evaluation (i.e., Type I Error). Therefore, the uncertainty associated with the potential error due to incorrectly retaining these constituents at this stage of the evaluation is considered acceptable.
Dust Generation Rate Studies show that generation of particulate matter from concrete is possible in high abrasion environments, such as roadways exposed to studded tires. However, the evaluation did not identify any studies that evaluated the rate at which concrete dust is generated under these conditions. To address this uncertainty, the current evaluation assumed that 10 percent of the soil was composed of fly ash concrete dust. This value originates from the US EPA (2010a) screening assessment, which evaluated the potential for overland transport of fly ash from uncovered CCR landfills through wind dispersion and runoff. Levels this high are unlikely for encapsulated concrete because the tire studs on passing cars must grind the concrete for dust to be released, compared to granular fly ash, which is available for release at any time. The Washington State Department of Transportation estimated that about 0.25 mm of concrete pavement wear away per one million studded tire vehicle passes, with measured erosion rates between 0.04 and 0.5 mm/yr (WSDOT, 2010). Furthermore, the contributing area of even a moderate-sized CCR landfill is much greater than that of even the most heavily traveled roadways.
Therefore, the assumptions used in the current evaluation are likely to overestimate releases. While the magnitude of this overestimation is unknown, it is anticipated to be considerable.
Bioavailability Bioavailability is the fraction of the total contaminant mass that is available to interact with and potentially cause harm to a receptor’s body. The remaining fraction of the compound that is not bioavailable will pass through or over the body with no effect to the receptor. A host of different factors, such as the pH, particle size, moisture, and redox potential of the environment influence bioavailability.
In addition, the receptor’s age, sex, nutritional state, and physiological state affect bioavailability (US EPA, 2007). Given the numerous variables involved in determining bioavailability, this remains a source of uncertainty. The current evaluation assumed that the total mass of each COPC ingested was entirely bioavailable. This assumption can only overestimate exposure. The magnitude of this overestimation will vary on a case-by-case basis, depending both on the characteristics of the contaminated media and the individual receptor.
5-9 Roadway Composition The current evaluation assumes that the uppermost layer of the roadways that is exposed to high abrasion is composed of only concrete. In reality, many roads are either entirely composed of or overlain by other materials, such as asphalt. Therefore, as Zubek et al. (2004) points out, generated dust may contain no concrete at all. While the assumption of an exposed concrete surface used in the current evaluation may represent high-end exposure scenario, it likely overestimates typical exposures. The magnitude of this overestimation is unknown.
Receptor Habitats near Roadways It is unknown to what extent receptors may be exposed to any dust that accumulates near roadways and other concrete surfaces subjected to high levels of abrasion. Both child and adult residents are unlikely to spend extended periods of time alongside major roadways. Even under the theoretical scenario that an abandoned highway is converted to residential property, the construction would disturb the surface soil and either dilute the concrete dust present with subsurface soils or remove it entirely. As a result, the high concentrations evaluated likely overestimate exposures. However, the magnitude of this overestimation is unknown.
Wildlife is more likely to spend appreciable time near roadways. The level of exposure depends on a number of factors, including the foraging range of the species, the quality of available food sources, and the season, as well as intra-species and inter-species competition. Furthermore, fragmented land near a roadway is often poor habitat for permanent ecological populations. Consequently, this uncertainty is likely to overestimate exposures. However, the magnitude of this overestimation is unknown.
5.2.2 Uncertainties for Ground and Surface Water Exposures The uncertainties addressed in this subsection pertain to the evaluation of leachate released from fly ash concrete during use, the transport to ground and surface waters, and the resulting receptor exposures.
This document does not discuss uncertainties introduced through the use of the Industrial Waste Evaluation Model (IWEM). IWEM is a peer-reviewed model, and IWEM Technical Documentation and User’s Guide (US EPA, 2010d) discusses the various uncertainties associated with this model.
Selection of COPCs The current evaluation selected COPCs based on the findings of the 2010 CCR Risk Assessment (US EPA, 2010a). This risk assessment aggregated modeling results for all types of CCRs when drawing conclusions. It is possible that reliance on these findings to select COPCs may result in retaining COPCs related to CCRs other than fly ash. Conversely, it is possible that consideration of data on CCRs other than fly ash may dilute modeling results enough to eliminate COPCs that would have otherwise been retained. However, Kosson et al. (2013) demonstrated that fly ash concrete leach at rates lower than pure fly ash. As a result, any constituents that may have been screened out in US EPA (2010a) as a result of the dilution of fly ash data are likely to have also been screened out in the current evaluation due to the lower leaching rates. Therefore, the magnitude of this uncertainty is considered small.
5-10 Available Data Based on the available studies of the leaching from fly ash alone and fly ash in cement materials, the leaching of COPCs from fly ash was not increased by incorporation of the fly ash in cement materials (US EPA, 2012b; Kosson et al., 2013). Therefore, the current evaluation assumed that, if the leaching behavior of fly ash was adequately captured in the available samples, so was the contribution of fly ash to leaching from fly ash concrete. Kosson et al. (2013) analyzed the leaching behavior of the pure fly ash prior to mixing in concrete. One or more of the fly ashes used in this study exhibited arsenic, antimony, boron, chromium, lead, and molybdenum leaching close to the upper bounds identified as part of a broader sampling effort for US EPA (2009a). However, these fly ashes exhibited cadmium, selenium, and thallium leaching closer to the median identified in US EPA (2009a). Therefore, the current evaluation may underestimate high-end leaching of these three COPCs from fly ash concrete.
The fly ashes used in Kosson et al. (2013) and Garrabrants et al. (2013) are from the subset of ashes known to meet the specifications for use in concrete. It is unknown whether this subset has the same upper bounds of cadmium, selenium, and thallium leaching as the fly ashes in US EPA (2009a). Yet, even if these higher leaching fly ashes are appropriate for use in concrete, the potential increase in fly ash concrete leachate concentrations is unlikely to alter the conclusions of the evaluation. The current evaluation identified cadmium, selenium, and thallium as COPCs based on the findings of the 2010 CCR Risk Assessment (US EPA, 2010a), which identified the 90th percentile hazard quotients of 5.0 for cadmium (ecological sediment exposure), 2.0 for selenium (ecological surface water exposure), and 3.0 for thallium (human ground water exposure). Kosson et al. (2013) showed that leaching from fly ash concrete was consistently lower than leaching from pure fly ash by at least a factor of two for selenium and an order of magnitude for both cadmium and thallium. The magnitude of these decreases alone would be sufficient to eliminate these constituents as COPCs in US EPA (2010a). Furthermore, the LEAF Method 1313 data used in Kosson et al. (2013) represent saturated leachate from ground up concrete. Consideration of time-dependent leaching of intact concrete would further reduce leachate concentrations. Based on these different lines of evidence, the magnitude of this uncertainty is likely to be small.
Non-Detect Values The current evaluation eliminated arsenic, cadmium, lead, molybdenum, and thallium as COPCs at Step 2 (Comparison of Available Data) because Garrabrants et al. (2013) did not detect these COPCs in leachate collected from fly ash concrete or portland cement concrete with EPA Method 1315. Based on the available data, the evaluation concluded that the rate at which these COPCs leach from fly ash concrete is comparable to that of portland cement concrete. These constituents may still be present in the samples at concentrations below the MDL. However, they are unlikely to be present near the MDL. The concrete matrix is known to become denser as it cures (Garboczi, 1995). As the internal pore spaces decrease in size, the rate at which COPCs can leach out decreases. Selenium provides the best example for this: it was not detected in any of the leachate from three month cure samples, but this evaluation retained selenium for further consideration because this COPC was detected in leachate from a few samples of fly ash concrete cured for only 28 days. Therefore, the other COPCs that were not detected even at 28 days are likely to be even lower after a three month cure time. In addition, Garrabrants et al.
(2013) did not detect arsenic, cadmium, lead, molybdenum, and thallium during the leaching time step 5-11 of 14 days. Therefore, the cumulative mass leached for a duration of fewer than 2 days will be even further below the MDL. The potential for these constituents to be released at higher levels from fly ash concrete than from portland cement concrete remains. However, the incremental increase in exposures at these already low levels is unlikely to result in appreciable risk to downgradient receptors. Therefore, the impact of this uncertainty on the findings of the current evaluation is likely to be minimal.
Incremental Releases The current evaluation does not consider the incremental fly ash contribution to releases from concrete. Leaching is not necessarily an additive process, and the available dataset makes it difficult to empirically parse out the exact contribution of fly ash to releases from concrete. Therefore, to remain conservative, the evaluation used total releases from fly ash concrete to estimate potential exposures. As a result, the COPC releases that result from incorporation of fly ash into concrete may be lower than those presented in Step 4 (Screening Assessment). This assumption will likely overestimate the COPC exposures associated with the beneficial use of fly ash in concrete. In particular, chromium was shown in Step 2 to be present in and released from portland cement concrete at levels approaching those of fly ash concrete. The magnitude of this overestimation is not known.
Chromium Speciation Chromium was the one COPC in this evaluation that warranted fate and transport modeling. The speciation of chromium is an important consideration because its toxicity is dependent on speciation.
The evaluation did not identify any studies evaluating the speciation of constituents leaching from concrete. Even if these data were available, receptors are unlikely to ingest leachate directly from the concrete. Instead, the leachate will migrate through the soil and ground water table prior to contact with any receptors. During transport, the leachate will change in pH and be exposed to a different set of redox conditions, which can dramatically change speciation. Therefore, even when one species leaches from concrete, the amount released and the amount that reaches the receptor may not be the same. Because of the high uncertainty surrounding the exact speciation of chromium at the point of exposure, the evaluation conducted fate and transport modeling under the assumption that chromium was released in the form most mobile in the environment. Therefore, this assumption may overestimate potential exposures. However, the magnitude of this overestimation is unknown.
Roadway Dimensions The width and thickness of concrete roadways used to model concrete leaching in the current evaluation are based on the upper bound of specifications from the American Association of State Highway and Transportation Officials (AASHTO, 1993; 2004). The use of the upper bound values for these variables likely overestimates potential exposures. The length of the concrete roadways is uncertain. The needs of the individual project, rather than best design practices define this variable.
However, the exact length of the concrete is unlikely to have an impact on the current evaluation of HEIs. The evaluation assumes that human and ecological receptors are located midway along the length of the concrete source where they are exposed to the highest ground water concentrations. Extending the ends of the concrete out even farther will not change the concentration at the centerline of the plume.
5-12 Therefore, the assumed roadway dimensions may overestimate releases and subsequent exposures.
However, the magnitude of this overestimation is unknown.