Evaluation of Life Cycle Assessment (LCA) for Roadway Drainage

Jul 11, 2017 - Wang , G.-T.; Chen , S.; Barber , M. E.; Yonge , D. R. Modeling Flow and Pollutant Removal of Wet Detention Pond Treating Stormwater Ru...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/est

Evaluation of Life Cycle Assessment (LCA) for Roadway Drainage Systems Diana M. Byrne, Marta K. Grabowski, Amy C. B. Benitez, Arthur R. Schmidt, and Jeremy S. Guest* Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign, 205 North Mathews Avenue, 3221 Newmark Civil Engineering Laboratory, Urbana, Illinois 61801, United States S Supporting Information *

ABSTRACT: Roadway drainage design has traditionally focused on cost-effectively managing water quantity; however, runoff carries pollutants, posing risks to the local environment and public health. Additionally, construction and maintenance incur costs and contribute to global environmental impacts. While life cycle assessment (LCA) can potentially capture local and global environmental impacts of roadway drainage and other stormwater systems, LCA methodology must be evaluated because stormwater systems differ from wastewater and drinking water systems to which LCA is more frequently applied. To this end, this research developed a comprehensive model linking roadway drainage design parameters to LCA and life cycle costing (LCC) under uncertainty. This framework was applied to 10 highway drainage projects to evaluate LCA methodological choices by characterizing environmental and economic impacts of drainage projects and individual components (basin, bioswale, culvert, grass swale, storm sewer, and pipe underdrain). The relative impacts of drainage components varied based on functional unit choice. LCA inventory cutoff criteria evaluation showed the potential for cost-based criteria, which performed better than mass-based criteria. Finally, the local aquatic benefits of grass swales and bioswales offset global environmental impacts for four impact categories, highlighting the need to explicitly consider local impacts (i.e., direct emissions) when evaluating drainage technologies. water, seafood, and recreation.5 Recognition of the potentially dangerous effects of stormwater runoff has prompted drainage system components that were originally designed for flow attenuation (e.g., swales and basins) to now be considered for their water quality benefits;6 therefore, the fate and transport of pollutants through stormwater treatment devices such as swales7−11 and basins12−14 has been the subject of recent research. Globally, construction and maintenance of drainage system components require materials (e.g., concrete), equipment operation (e.g., excavator), transportation (e.g., hauling of materials and equipment), and disposal (e.g., landfilling), all of which contribute to global environmental impacts (e.g., climate change) throughout the system’s life cycle. Life cycle assessment (LCA) is useful for assessing environmental sustainability of water infrastructure because of its ability to collectively quantify impacts that occur at various phases of the infrastructure’s lifetime. While LCA has been applied extensively to wastewater,15 drinking water (e.g., refs 16 and 17), and integrated urban water systems,18 there are significantly fewer LCAs for stormwater systems. Of the

1. INTRODUCTION Conventional highway drainage systems are designed to costeffectively manage runoff for roadway safety and pavement protection while also supporting the broader goal of mitigating infrastructure impacts to local hydrologic conditions. To protect the safety of vehicles on the road and to protect the pavement from water damage, runoff must be rapidly removed from the pavement surface. To cost-effectively accomplish these goals, highway drainage systems have traditionally been designed to remove water from the driving lanes by directing it to the outside shoulders, after which water either infiltrates or discharges to surface water. This approach to roadway drainage design has traditionally prioritized water quantity and cost over local water quality considerations or broader global environmental impacts despite the fact that impervious surfaces cover more than 100 000 km2 in the contiguous United States and that stormwater pollution continues to grow in many watersheds throughout North America.1 Locally, highway runoff carries a variety of pollutants including particulate matter, metals (e.g., copper and zinc), and nutrients (e.g., nitrate and phosphate),2 making highways major contributors to nonpoint-source pollution.3 This contamination poses local environmental and health risks because many stormwater pollutants exhibit toxicity,4 and chronic and acute illnesses can be traced to runoff via exposure through drinking © 2017 American Chemical Society

Received: Revised: Accepted: Published: 9261

April 10, 2017 June 27, 2017 July 11, 2017 July 11, 2017 DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology

Information). For both the first (project level) and the second (individual drainage component level) tiers of the analysis, construction and maintenance (including material production, onsite equipment operation, and materials and equipment transportation to the site), and end of life phases are included (Figure 1). For the second tier of the analysis (individual

published stormwater LCAs, the majority are of green infrastructure (e.g., refs 19−26), four of which directly consider local water quality impacts associated with the use phase.19,22,25,26 For roadway systems, many LCAs have been conducted for pavement, mostly focusing on pavement materials;27,28 however, to the authors’ knowledge, there have not been any studies specifically focused on the environmental impacts of roadway drainage systems or the implications of LCA methodological choices in this context. For LCA to be useful to the roadway drainage and stormwater industry, LCA methodology choices need to be considered because roadway drainage systems differ from wastewater and drinking water systems to which LCA has more frequently been applied. A recent review of urban water systems highlighted the importance of considering direct aquatic emissions as well as expanding the life cycle inventory beyond raw materials to include construction and maintenance activities (e.g., excavation and mowing).18 However, roadway drainage and other stormwater systems have unique attributes within urban water systems and are influenced by local conditions such as rainfall and land imperviousness.19 Therefore, methodology considerations (e.g., functional unit and cutoff criteria) need to be evaluated specifically for roadway drainage as they have for wastewater15 and integrated urban water systems.18 Finally, LCA is of limited value without considering cost,29 which has always been emphasized in decision-making. A better understanding of the unique limitations of LCA applied to roadway drainage could help improve the tool’s usefulness. The over-arching goal of this work is to evaluate LCA methodological choices for roadway drainage to guide future applications of LCA to drainage design and operation. This is achieved by connecting detailed design parameters (e.g., grass swale dimensions and storm sewer pipe diameters) to direct aquatic and soil emissions (local environmental impacts), life cycle assessment (global environmental impacts), and life cycle costing (LCC; cost) in a Monte Carlo framework. To understand environmental impacts of drainage components in the context of actual projects as well as in terms of their individual life cycle phases and design parameters, this framework is applied in two tiers: (1) linking specific design parameters with representative cross-sections for 10 segments of a highway drainage system in the Midwest United States and (2) analyzing individual drainage components (basin, bioswale, culvert, grass swale, storm sewer, and pipe underdrain) across a landscape of observed design specifications. Ultimately, this quantitative sustainable design framework is leveraged (1) to elucidate the implications of methodological decisions (including the selection of functional unit, inclusion of life cycle phases, use of cutoff criteria, and inclusion of pollutant removal and discharge of direct emissions) on LCA results for roadway drainage and (2) to develop a path forward for effective application of LCA to roadway drainage and other stormwater systems to inform decision-making and to guide future research.

Figure 1. Summary of life cycle phases, inventory categories, and life cycle impacts included for the first tier (project level) and second tier (individual component level) of the analysis (additional details in Table S2).

drainage component level), direct discharge of pollutants that travel from the road are considered for individual conveyance elements (grass swale, bioswale, and storm sewer) separately (Figure 1). All LCA and LCC results are calculated for a 60 year lifetime, after which major rehabilitation (e.g., storm sewer replacement) is likely required. LCA and LCC were executed for 10 sections of drainage systems along an interstate highway in the Midwest United States (annual average daily traffic of 23 220−45 350; section S2 of the Supporting Information). All projects were restricted to approximately 0.4 km (0.25 mi) stretches of roadway constructed from 2012 to 2013 and were chosen to provide a diversity of drainage components to include in the analysis and are not implicitly representative of all roadway drainage systems. Design drawings were used to extract specifications for each component’s design parameters (section S2 of the Supporting Information). The variability in design parameters across the 10 projects allows for drainage components to be analyzed concurrently to evaluate their relative significance within a given project while not directly comparing them to each other as equivalent alternatives. Within goal and scope definition, methodological choices for functional unit (sizebased and flow-capacity-based) and cutoff critiera (mass-based, energy-based, and cost-based) were compared. 2.1.2. Life Cycle Inventory. Material and equipment requirements for each drainage component (Table S2) were determined using design specifications from the Illinois Tollway’s Drainage Design Manual,30 Landscape Manual,31 and Standard Drawings32 as well as the Illinois Department of Transportation’s Standard Specifications,33 supplemented with manufacturer specifications or communication with the Illinois Tollway. The maintenance schedule (including mowing, herbicide and fertilizer application, seeding, cleaning, grading, and compacting) for each drainage component was based on Illinois Tollway standards of practice34 (section S3.4 of the Supporting Information). Inventory data associated with each material were acquired using the ecoinvent v3.135 database accessed via SimaPro v8.0.4 (section S3.3 of the Supporting Information). Due to the large volume of concrete required for certain drainage components,

2. METHODS 2.1. Life Cycle Assessment. 2.1.1. Goal and Scope Definition. The LCA and LCC system boundaries include construction, maintenance, use, and end of life. A total of six components were considered as potential parts of a roadway drainage system: basin, bioswale, culvert, grass swale, storm sewer, and pipe underdrain (section S1 of the Supporting 9262

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology

benzene, and xylenes (BTEX)) was applied (section S3.8 of the Supporting Information). 2.1.3. Life Cycle Impact Assessment. Life cycle inventory emissions were converted to ten midpoint impact categories using the U.S. EPA’s Tool for the Reduction and Assessment of Chemicals and Other Environmental Impacts (TRACI; 2.1 version 1.02 within SimaPro version 8.0.4) 45 and the cumulative energy-demand method. Equipment emissions from NONROADS were matched to emissions in TRACI using U.S. EPA conversion factors for hydrocarbons46 and particulate matter47 (section 3.7 of the Supporting Information). Direct emissions to water and soil included in the use phase were converted to impacts using TRACI characterization factors directly. For both TSS and TDS, characterization factors for both water and soil were zero, making assumptions regarding their fate inconsequential; future developments in impact assessment methodologies that address TSS and TDS could easily be incorporated with the current model structure. Given that the transportation industry accounted for 26% of U.S. greenhouse gas emissions in 201448 and climate change has the potential to greatly impact roadway costs,49 climate change impacts were used to characterize the implications of LCA methodological decisions on roadway drainage projects as a whole. To better understand the impacts of methodological decisions on LCA results for individual drainage components, all TRACI impact categories were characterized across life cycle phases, and the four categories (eutrophication, carcinogenics, noncarcinogenics, and ecotoxicity) offset by pollutant removal benefits are used to consider implications of direct emissions. 2.2. Life Cycle Costing. Construction and maintenance activities were matched to RSMeans data37−39 and used to obtain costs by materials, equipment, labor, and indirect expenses (section S3.6 of the Supporting Information). End of life costs were calculated using landfilling tipping fees, and costs of future maintenance activities and landfilling were converted to present-day prices using present worth analysis with a discount rate of 6.0%. 2.3. Uncertainty and Sensitivity Analyses. Uncertainty analysis was conducted using Monte Carlo with Latin Hypercube Sampling (a sampling technique to reduce computation time)50 implemented in Matlab. Uncertainty surrounding the use phase (e.g., imperviousness coefficient), costing (e.g., discount rate), transportation (e.g., distance), maintenance frequencies, concrete mix design (e.g., percent fly ash), and component specifications (e.g., concrete volume per catch basin) was quantified using uniform or triangular probability distributions (section S4.1 of the Supporting Information). Initial concentrations of pollutants in runoff were varied using empirical distributions based on NSQD data for entries with 100% land-use classification of highways and/or freeways (section S4.2 of the Supporting Information). Entries below the detection limit were set to the detection limit, and outliers were identified and removed (section S4.2 of the Supporting Information). Treatment efficacies for grass swales, bioswales, and basins were varied according to distributions presented in the National Pollutant Removal Performance Database44 (section S4.3 of the Supporting Information). Sensitivity analysis for climate change (representing global environmental impacts), eutrophication (representing local environmental impacts), and total cost (representing economic impacts) to all design and uncertainty analysis parameters (Table S7) was conducted using Spearman’s rank correlation coefficients, which determine the strength of a monotonic

an ecoinvent process for concrete was modified to reflect the mix design specifications of the Illinois Department of Transportation33 (section S3.5 of the Supporting Information). Emissions from transportation of materials and equipment assumed an average transportation distance of 16 km (10 miles). Storm sewers, culverts, pipe underdrains, and basin outlet structures were conservatively assumed to be landfilled at the end of the 60 year lifetime (consistent with ref 19), whereas all parts of grass swales and bioswales were assumed to stay in place without disposal. Direct emissions from equipment were estimated using the United States Environmental Protection Agency’s (U.S. EPA) NONROADS model version 2008a.36 Productivity rates for each construction and maintenance activity or constituent element were obtained from RSMeans37−39 and used to calculate time of equipment operation. NONROADS was used to obtain fuel-efficiency data (to calculate fuel consumption) and equipment-specific emissions per gallon of fuel for total hydrocarbons, carbon monoxide, nitrogen oxides, carbon dioxide, sulfur dioxide, and particulate matter associated with combustion. These combustion emissions were combined with a US−EI process for diesel production40 to calculate total emissions associated with equipment operation for each equipment type (section S3.7 of the Supporting Information). Flow from the road into the drainage system was calculated using a modified rational method, consistent with water-quality volume design criteria41 using the design parameters of length and width to calculate the drainage area. Hourly precipitation data for Chicago from the Illinois State Water Survey was used to calculate precipitation depth and duration of each storm for 10 years of data (2004−2013).42 An imperviousness coefficient of 0.9 was used for the pavement surface. Pollutant load was assumed to come entirely from the road’s impervious area, neglecting adjacent pollutant sources that may have been present before roadway construction. Initial concentrations of highway runoff (as event mean concentrations, EMCs) were obtained from the National Stormwater Quality Database (NSQD) version 4.02 for database entries with a land use classification of 100% highways and/or freeways.43 Pollutants include total suspended solids (TSS), total dissolved solids (TDS), biochemical oxygen demand (BOD), total arsenic (As), total cadmium (Cd), total chromium (Cr), total copper (Cu), total iron (Fe), total nickel (Ni), total lead (Pb), total zinc (Zn), nitrate (NO3− as N), ammonium (NH4+ as N), and orthophosphate (PO43− as P). Use phase impacts of direct emissions from each drainage component (including emissions stemming from effluent, infiltration, sorption, and settling) were calculated individually as offsets relative to impacts incurred if no treatment occurred (i.e., as if pollutants from the road traveled through each drainage component without removal). Treatment efficacy for grass swales, bioswales, and wet and dry basins were quantified using percent removal data from the National Pollutant Removal Performance Database version 3 (aggregated from sampling studies).44 Removed TSS was assumed to settle out, and removed nutrients were assumed to settle or sorb and eventually be assimilated by plants; therefore, no additional impacts were included. Removed TDS and metals were assumed to deposit to soil. Removed BOD was assumed to become fugitive CO2 emissions from aerobic degradation, and a climate change characterization factor of 1.07 kg CO2 per kg BOD degraded (the average for benzene, toluene, ethyl9263

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology

Figure 2. Comparison of normalized climate change and normalized cost for 10 sample projects. (A) Values are normalized to length (m) for bioswales, culverts, grass swales, storm sewers, and underdrains and normalized to area (m2) for basins. (B) Values are normalized to length and flow (m and m3·s−1) for bioswales, culverts, grass swales, storm sewers, and underdrains and normalized to storage volume (m3) for basins. Whole project results stem from construction, maintenance, and end of life (the use phase was not included in this figure because its impacts cannot be attributed to individual drainage components), with points indicating medians and error bars extending to 10th and 90th percentiles from uncertainty analysis. Details for sample projects and a replicate of Figure 2 in U.S. units can be found in section S2 of the Supporting Information.

relationship between two variables (section S5.2 of the Supporting Information). Results were calculated across 1000 Monte Carlo simulations for the first-tier analysis (projects; varying uncertainty parameters with fixed design parameters) and across 10 000 simulations for the second-tier analysis (individual components; varying both uncertainty parameters and design parameters).

considerations are only encouraged when they are cost-effective and do not impact traffic safety.30 When drainage components are compared by length or area (Figure 2A), drainage components consistently clustered by technology, with the most material intensive components resulting in the largest normalized cost and climate change impacts. Out of the three conveyance element options managing runoff (grass swale, bioswale, and storm sewer), storm sewers were shown to be the greatest contributors to climate change, with up to 3 times the emissions of bioswales and 12 times the emissions of grass swales. In terms of cost over the system’s life cycle, bioswales did not provide a significant economic advantage over storm sewers (Figure 2A), with grass swales performing the best economically (up to 1/ 8th the cost of storm sewers and up to 1/7th the cost of bioswales). Pipe underdrains and basins had the lowest impacts normalized to length or area for both climate change (median values of 38−42 kg CO2 eq·m−1 for underdrains and 2−8 kg CO2 eq·m−2 for basins) and cost (median values of 38−59 USD·m−1 for underdrains and 3−17 USD·m−2 for basins). Alternatively, when costs and environmental impacts normalized to length and flow capacity or storage volume (Figure 2B) were compared, underdrains (made of high-density polyethylene, HDPE) resulted in larger normalized cost and climate change impacts than storm sewers or culverts (made of reinforced concrete). This is because, while smaller in size, underdrains carry notably less flow (0.01 to 0.02 m3·s−1 for sample projects) than storm sewers (0.1 to 0.4 m3·s−1 for sample projects). The larger impacts of HDPE pipe compared to reinforced concrete pipe when normalized to flow capacity are consistent with a previous LCA of pipe materials.51 The differences in conclusions based on normalization highlights the

3. RESULTS AND DISCUSSION 3.1. Selection of Functional Unit. Unless they serve a common function (e.g., storage and conveyance), stormwater alternatives cannot be directly compared.24 Although all drainage components serve the same roadway section, they may also manage water quantity and water quality unassociated with the roadway itself. For example, a culvert may manage the flow of a local stream while an underdrain manages infiltration through the pavement; therefore, both serve different functions despite managing water quantity. Additionally, drainage components are designed for either storage (in the case of basins) or conveyance (in the cases of grass swales, bioswales, and storm sewers). To highlight the variation of results based on functional unit for the 10 projects, costs and environmental impacts (as climate change) were normalized two different ways: by size (Figure 2A) or by size and capacity (Figure 2B). Conveyance elements were normalized to length or length and flow capacity (i.e., maximum flow rate; details are given in section S2 of the Supporting Information), whereas storage elements were normalized to area or storage volume. Water quality was not evaluated as part of the functional unit given that roadway drainage system design is focused on water quantity management, and, in the state of Illinois, water quality 9264

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology

Figure 3. Largest contributing life cycle phase by individual drainage component and impact category. Circle color and pattern identifies the phase that had the largest percent contribution to the corresponding impact category for the indicated drainage component. The size (by area) of each circle represents how large the indicated percent contribution was, with the outer boundary circle representing a 100% contribution. The results are median values from the uncertainty analysis, with design parameters chosen based on sample projects (full results with uncertainty can be found in section S5.1 of the Supporting Information). Gray column shows the results for cumulative energy demand impact assessment methodology.

importance of functional unit selection. Normalizing to flow capacity or storage volume (Figure 2B) allows for comparisons in terms of a drainage component’s ability to manage water quantity. However, normalizing to length or area (Figure 2A) provides understanding of the relative impacts of drainage components in the context of a length of roadway. For example, length may be more appropriate than flow capacity for a project with storm sewers and underdrains, as these will have different functions (managing runoff versus infiltration) and, therefore, different flow capacities but will both serve a common function of managing water from a given length of roadway. 3.2. Life Cycle Phases. For all drainage components except grass swales, the LCA phase (construction materials, construction equipment, construction transportation, maintenance materials, maintenance equipment, maintenance transportation, and end of life) that had the largest contribution to each impact category was consistently related to construction activities (Figure 3; full results with uncertainty can be found in section S5.1 of the Supporting Information). For basins, construction equipment (specifically for excavation and grading) dominated 8 of 11 impact categories by at least 65% (black circles, Figure 3). For box culverts, construction materials (specifically concrete and steel) were responsible for an average of 82% of impacts across all categories (red circles, Figure 3), while pipe culverts and storm sewers were dominated by construction materials (45−99% median contribution) in 7 of 11 impact categories and by construction equipment in the other 4 (47− 91% median contribution). Unlike bioswales, which were governed by construction materials (sand and topsoil) for 10 of 11 categories, grass swales were the only drainage component not dominated by construction activities (Figure 3); however, this was due to the simplicity of grass swales, requiring minimal materials and equipment. This suggests that, in general, maintenance plays a minor role (less than 30% median contribution) in the life cycle environmental impacts of drainage systems as compared to construction unless initial

material requirements are very low (e.g., for grass swales). The significant role of construction materials is reflected in previous studies that showed the large effects of concrete and other materials on LCA results for green infrastructure.19,21,23,24,52 Unlike LCAs for wastewater and drinking water systems in which the construction phase is often neglected15 (cited as low as 10% of total impacts),53 these results suggest that the construction phase (both materials and equipment) can play a significant role in roadway drainage and potentially other stormwater LCAs, and therefore, it is necessary to develop a life cycle inventory including these construction activities. 3.3. Cutoff Criteria. The cutoff criterion for an LCA determines which inputs and outputs must be included within the system boundary. According to ISO standards, recommended cutoff criteria include mass, energy, or environmental significance.54 To understand the efficacy of ISO recommended cutoff criteria for roadway drainage systems, cumulative climate change versus cumulative mass (materials and fuel) was plotted for each of the 10 sample projects (Figure 4A), in which each point corresponds to a construction or maintenance activitity or constituent element (e.g., catch basin) included for that sample project. Percent contributions to total mass and total climate change impacts of each activity or constituent element were calculated, and activities were ranked from largest to smallest mass. The same procedure for mass was followed for cumulative energy (Figure 4B), with activities ranked from largest to smallest energy demand. Aside from environmental significance (a straight line correlation with environmental impacts), energy was shown to be a reliable cutoff criterion across all 10 projects, while mass proved to be ineffective. Depending upon the project, establishing an LCA system boundary that includes items contributing to 90% of energy demand captured anywhere from 87 to 92% of climate change impacts (Figure 4B), whereas establishing a boundary including items contributing to 90% of mass captured from 15 to 60% of climate change impacts 9265

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology

to 89% of climate change impacts (Figure 4C), superior to the mass-based criterion. This correlation between cost and climate change impacts suggests that cost may be a more reliable LCA cutoff criterion for roadway drainage systems than mass and demonstrates that reducing construction requirements (which govern global environmental impacts; Figure 3) can synergistically reduce both the cost and climate change impacts of a project. While environmental significance is the ideal LCA cutoff criterion, it is impractical because one would have to completely quantify the impacts of what can be excluded before justifying those exclusions.55 Moreover, there is no guarantee that small mass or energy contributions correspond to small environmental impact contributions.56 Using cost as a cutoff criterion would make environmental-based decisionmaking more accessible to the roadway industry as information regarding mass, energy, or environmental significance of materials, and processes may not be readily known or easily quantified. In contrast, cost is a critical criterion in roadway decision-making, with each activity’s or element’s cost recorded and easily accessible in a project’s bid tabs. 3.4. Use Phase Pollutant Removal and Direct Emissions. In addition to environmental impacts incurred during construction and maintenance (second-order processes),57 the use phase can result in discharge of direct emissions to soil and water (first-order processes).57 While it is generally accepted that vehicles are the main source of runoff contamination,28 other sources include the roadway itself (e.g., pavement leaching)58−60 and infrastructure installation and rehabilitation.61 These pollutant sources contribute to high variability of pollutant concentrations in highway runoff (Figure S3) due to temporal (e.g., first-flush phenomenon)62−65 or spatial (e.g., traffic volume,66 pavement type,67 winter maintenance,68 and local policies)69 differences, making predictive modeling such as regression analyses site-specific.62 Stormwater controls can mitigate the impacts of runoff on the receiving environment1 through sedimentation11,70 adsorption,70 filtration,2 biodegradation, and plant uptake.12 Treatment efficacy of swales and basins have been studied through experimentation8,10,11,13,14,70−75 and modeling,7,12,70,74,76,77 and results are variable, with TSS removal through grass swales ranging from negative removal to nearly 100%.78 Many factors contribute to removal efficacy of swales (e.g., infiltration rate and vegetation)78 and basins (e.g., surface area and storage volume)70 making removal efficacies site-specific9 with large uncertainty ranges (section S4.3 of the Suppporting Information). Recognizing these potential local environmental impacts of stormwater quality in parallel with the relatively low climate change impacts of drainage technologies in the context of an entire roadway system, LCA system boundaries should consider including direct emissions so that trade-offs can be evaluated. To this end, the impacts of direct aquatic and soil emissions were compared to global impacts incurred during construction, maintenance, and end of life for the three conveyance element options managing runoff (grass swale, bioswale, and storm sewer). Grass swales were shown to consistently incur the smallest global impacts, with bioswales second, and storm sewers incurring the most (Figure 5). Unlike grass swales and bioswales that provide local water and soil quality benefits, sewers provide no treatment of pollutants and, therefore, do not provide any offsets (Figure 5). Using LCA and LCC results for decision-making without including local benefits would

Figure 4. Whole project evaluation of LCA cutoff criteria of (A) mass, (B) energy, and (C) cost. Each line indicates one of the 10 sample projects (project descriptions and the top 5 most impactful activities of each project are provided in sections S2 and S5.3 of the Supporting Information). Each point corresponds to one of the applicable construction and maintenance activities included for that sample project. The percent of total mass, energy, or cost of each activity or constituent element was calculated, and activities were ranked from largest to smallest mass, energy, or cost for panels A, B, and C, respectively. Box and whisker plots to the right of the graphs indicate the percent of climate change impacts that are captured when a 90% cutoff criteria of mass, energy, or cost is applied.

(Figure 4A). For roadway drainage systems, the unreliability of mass as a cutoff criterion is primarily caused by construction and maintenance activities composed mainly of equipment operation (e.g., excavation and mowing), which requires fuel but no other raw materials. Energy better reflects the climate change impacts of equipment operation than mass, making energy a more reliable cutoff criterion. The results for each drainage component within each project revealed a correlation between climate change and economic impacts (Figure 2). Therefore, this relationship was explored to evaluate cost as a cutoff criterion (Figure 4C) by following the same procedure for mass and energy, with activities ranked from most to least expensive. Across all projects, cost has a more consistent correlation with climate change impacts than mass but with less precision than energy. Although not ISO recommended, establishing a system boundary that includes items contributing to 90% of cost captured anywhere from 72 9266

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology

partial replacement of clinker in cement.79,80 However, beyond these measures, it may not be meaningful to make design decisions based on climate change impacts given the contribution of drainage systems to greenhouse gas emissions relative to vehicles on the road. Specifically, the four projects analyzed that included storm sewers ranged from 450 to 710 t CO2 eq/km of storm sewer across the drainage systems’ 60 year lifetime. Assuming 300 000 vehicles per day,81 vehicles on the same 1 km stretch of roadway will be responsible for the same magnitude of CO2 equivalents in only 4−7 days. Thus, global impacts will be driven by other roadway activities (i.e., not drainage), but drainage systems will continue to have a disproportionately high impact on the local aquatic environment. 4.3. Research Needs for Inclusion of Direct Emissions. Recognizing the sensitivity of environmental impacts to use phase data (section S5.2 of the Supporting Information) and the disproportionate impact of roadway drainage systems on the local aquatic environment, addressing current limitations in quantifying water and soil emissions, and their associated environmental impacts could enhance the LCA’s applicability to roadway drainage. At the inventory level, only pollutants with ample data for both initial concentrations and removal efficacies were able to be considered;43,44 however, there are other pollutants worth tracking (e.g., oil and grease)2 if more robust data sets become available. Furthermore, while nutrients may be taken up by plants or transformed (e.g., via nitrification and denitrification), metals have the potential to be released again (e.g., while adsorbed to particulate matter);10 therefore, more complexity for pollutant fate could offer additional insights for consideration of local impacts. Additionally, differentiating between emissions to surface water and groundwater would be advantageous;82 however, current TRACI characterization factors do not leverage this level of detail. Furthermore, all characterization factors for TDS (represented as chloride in TRACI) are zero despite the effects of road salt on local aquatic environments.83 Research is needed to determine how to best account for TDS and other pollutants not adequately addressed during impact assessment. Furthermore, spatially resolved characterization factors could help to harmonize local and global impact assessments.84−88 Finally, incorporating hydraulic and hydrologic models (e.g., U.S. EPA SWMM) could provide more accurate estimates (compared to generic percent removals) of direct aquatic emissions but would introduce uncertainty surrounding required site-specific modeling data (e.g., build-up and washoff parameters) that could be potentially unknown or variable. Future research to better account for local aquatic impacts will further strengthen LCA for more comprehensive decisionmaking regarding roadway drainage systems.

Figure 5. Global and local environmental trade-offs for three conveyance elements managing runoff (grass swale, bioswale, and storm sewer) with impacts and offsets normalized to storm sewer global impacts. Impacts incurred from construction (Con), maintenance (Main), and end of life (EOL) are above zero (equal to 100% for storm sewers), and offsets from the benefits of pollutant removal are below zero. Individual conveyance elements were sized for a design flow of 0.03 m3·s−1 (1 ft3·s−1) and a length of 0.30 km (1000 ft). Relative differences in impacts are sensitive to the choice of length, but trends are consistent (section S5.4 of the Supporting Information).

suggest that grass swales would be preferable to bioswales; however, these relative differences caused by additional material requirements of bioswales (e.g., sand and topsoil) become lesssignificant when one is also comparing the global impacts incurred by storm sewers (Figure 5). In this context, focus should instead shift to local water quantity and quality impacts. Choosing one green infrastructure technology over another may modestly increase global environmental impacts; however, these technologies can provide meaningful improvements for the local aquatic environment by filtering pollutants and promoting additional infiltration.

4. PATH FORWARD 4.1. LCA Method Implications. The choice of functional unit greatly influenced results causing climate change impacts of culverts to change an order of magnitude relative to pipe underdrains. In general, construction (materials and equipment) played the largest role in total environmental impacts and, consequently, should not be neglected in future roadway drainage and stormwater LCAs. Cost-based cutoff criteria performed better than mass-based criteria, and therefore, continued investigation of its validity is recommended for additional drainage projects and other parts of roadway systems. Finally, local aquatic benefits offset global environmental impacts of grass swales and bioswales for four impact categories; thus, local impacts (i.e., direct emissions) need to be explicitly considered when using LCA to evaluate drainage technologies. 4.2. Local versus Global Inventory, Impacts, and Decision-Making. While some flexibility exists for overall drainage system design (e.g., constructing bioswales rather than grass swales), many drainage components serve different functions, often limiting choices among them. Total environmental impacts mainly stem from construction materials and equipment (Figure 3) and are consequently sensitive to design parameters affecting drainage component size (section S5.2 of the SI); however, design and material needs are strongly impacted by site-specific details regarding space availability. Roadway drainage designers can emphasize the construction of less material-intensive drainage components, use alternative materials to replace concrete (the most common material used), or leverage supplementary cementitious materials for



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.7b01856. Details on roadway drainage components; sample projects; inventory and impact assessment assumptions; uncertainty characterizations; sensitivity analyses; and additional results for life cycle phases, sensitivity analysis, and conveyance element comparison. (PDF) 9267

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology



open stormwater retention basin − Loads, efficiency and importance of uncertainties. Water Res. 2015, 72, 239−250. (15) Corominas, L.; Foley, J.; Guest, J. S.; Hospido, A.; Larsen, H. F.; Morera, S.; Shaw, A. Life cycle assessment applied to wastewater treatment: State of the art. Water Res. 2013, 47 (15), 5480−5492. (16) Igos, E.; Dalle, A.; Tiruta-Barna, L.; Benetto, E.; Baudin, I.; Mery, Y. Life Cycle Assessment of water treatment: what is the contribution of infrastructure and operation at unit process level? J. Cleaner Prod. 2014, 65, 424−431. (17) Del Borghi, A.; Strazza, C.; Gallo, M.; Messineo, S.; Naso, M. Water supply and sustainability: life cycle assessment of water collection, treatment and distribution service. Int. J. Life Cycle Assess. 2013, 18 (5), 1158−1168. (18) Loubet, P.; Roux, P.; Loiseau, E.; Bellon-Maurel, V. Life cycle assessments of urban water systems: A comparative analysis of selected peer-reviewed literature. Water Res. 2014, 67, 187−202. (19) Wang, R.; Eckelman, M. J.; Zimmerman, J. B. Consequential Environmental and Economic Life Cycle Assessment of Green and Gray Stormwater Infrastructures for Combined Sewer Systems. Environ. Sci. Technol. 2013, 47 (19), 11189−11198. (20) Wang, M.; Zhang, D.; Adhityan, A.; Ng, W. J.; Dong, J.; Tan, S. K. Assessing cost-effectiveness of bioretention on stormwater in response to climate change and urbanization for future scenarios. J. Hydrol. 2016, 543, 423−432. (21) Spatari, S.; Yu, Z.; Montalto, F. A. Life cycle implications of urban green infrastructure. Environ. Pollut. 2011, 159 (8−9), 2174− 2179. (22) Kosareo, L.; Ries, R. Comparative environmental life cycle assessment of green roofs. Build. Environ. 2007, 42 (7), 2606−2613. (23) De Sousa, M. R. C.; Montalto, F. A.; Spatari, S. Using Life Cycle Assessment to Evaluate Green and Grey Combined Sewer Overflow Control Strategies: Evaluating Watershed-Scale CSO Strategies with LCA. J. Ind. Ecol. 2012, 16 (6), 901−913. (24) Brudler, S.; Arnbjerg-Nielsen, K.; Hauschild, M. Z.; Rygaard, M. Life cycle assessment of stormwater management in the context of climate change adaptation. Water Res. 2016, 106, 394−404. (25) Flynn, K. M.; Traver, R. G. Green infrastructure life cycle assessment: A bio-infiltration case study. Ecol. Eng. 2013, 55, 9−22. (26) Xu, C.; Hong, J.; Jia, H.; Liang, S.; Xu, T. Life cycle environmental and economic assessment of a LID-BMP treatment train system: A case study in China. J. Cleaner Prod. 2017, 149, 227− 237. (27) Santero, N. J.; Masanet, E.; Horvath, A. Life-cycle assessment of pavements. Part I: Critical review. Resour. Conserv. Recycl. 2011, 55 (9−10), 801−809. (28) Santero, N. J.; Masanet, E.; Horvath, A. Life-cycle assessment of pavements Part II: Filling the research gaps. Resour. Conserv. Recycl. 2011, 55 (9−10), 810−818. (29) Norris, G. A. Integrating life cycle cost analysis and LCA. Int. J. Life Cycle Assess. 2001, 6 (2), 118−120. (30) Illinois State Toll Highway Authority. Drainage Design Manual; Illinois State Toll Highway Authority: Springfield, IL, 2012. (31) Illinois State Toll Highway Authority. Erosion and Sediment Control, Landscape Design Criteria; Illinois State Toll Highway Authority: Springfield, IL, 2013. (32) Illinois State Toll Highway Authority. Standard Drawing Revisions: Section B - Drainage Structures, Curbs, Curbs & Gutter and Ditches; Illinois State Toll Highway Authority: Springfield, IL, 2013. (33) Illinois Department of Transportation. Standard Specifications for Road and Bridge Construction; Illinois Department of Transportation: Springfield, IL, 2012. (34) Smith, N. AECOM, Lisle, IL. Personal communication, 2014. (35) Swiss Centre for Life Cycle Inventories. ecoinvent 3.1 database; Swiss Centre for Life Cycle Inventories: Zurich, Switzerland, 2015. (36) U.S. EPA. NONROAD2008a Model; Office of Transportation and Air Quality, U.S. Environmental Protection Agency: Washington, DC, 2008. (37) RSMeans Heavy Construction Cost Data 2015, 29th annual ed.; Fortier, R., Ed.; RSMeans: Rockland, MA, 2014.

AUTHOR INFORMATION

Corresponding Author

*Phone: (217) 244-9247; e-mail: [email protected]. ORCID

Diana M. Byrne: 0000-0002-5793-6024 Jeremy S. Guest: 0000-0003-2489-2579 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors thank Rebekah Yang for many helpful contributions, especially regarding equipment impacts and concrete mix design, as well as Hasan Ozer, Imad Al-Qadi, Nick Smith, Daniel Tobias, and Steven Gillen for assistance regarding many aspects of roadway drainage design and maintenance. This project was funded by the Illinois State Toll Highway Authority through the Illinois Center for Transportation (ICT). The contents of this report reflect the views of the authors, who are responsible for the facts and accuracy of the data presented herein. The contents do not necessarily reflect the official view or policies of the Illinois Tollway or ICT. This paper does not constitute a standard, specification, or regulation.



REFERENCES

(1) Rainfall to Results: The Future of Stormwater; Water Environment Federation (WEF): Alexandria, VA, 2015. (2) Kayhanian, M.; Fruchtman, B. D.; Gulliver, J. S.; Montanaro, C.; Ranieri, E.; Wuertz, S. Review of highway runoff characteristics: Comparative analysis and universal implications. Water Res. 2012, 46 (20), 6609−6624. (3) Ferreira, M.; Lau, S.-L.; Stenstrom, M. K. Size Fractionation of Metals Present in Highway Runoff: Beyond the Six Commonly Reported Species. Water Environ. Res. 2013, 85 (9), 793−805. (4) Pitt, R.; Field, R.; Lalor, M.; Brown, M. Urban stormwater toxic pollutants: assessment, sources, and treatability. Water Environ. Res. 1995, 67 (3), 260−275. (5) Gaffield, S. J.; Goo, R. L.; Richards, L. A.; Jackson, R. J. Public health effects of inadequately managed stormwater runoff. Am. J. Public Health 2003, 93 (9), 1527−1533. (6) Ferreira, M.; Stenstrom, M. K. The Importance of Particle Characterization in Stormwater Runoff. Water Environ. Res. 2013, 85 (9), 833−842. (7) Deletic, A.; Fletcher, T. D. Performance of grass filters used for stormwater treatmenta field and modelling study. J. Hydrol. 2006, 317 (3−4), 261−275. (8) Yu, S. L.; Kuo, J.-T.; Fassman, E. A.; Pan, H. Field test of grassedswale performance in removing runoff pollution. J. Water Resour. Plan. Manag. 2001, 127 (3), 168−171. (9) Backstrom, M. Grassed swales for stormwater pollution control during rain and snowmelt. Water Sci. Technol. 2003, 48 (9), 123−124. (10) Stagge, J. H.; Davis, A. P.; Jamil, E.; Kim, H. Performance of grass swales for improving water quality from highway runoff. Water Res. 2012, 46 (20), 6731−6742. (11) Winston, R. J.; Hunt, W. F.; Kennedy, S. G.; Wright, J. D.; Lauffer, M. S. Field Evaluation of Storm-Water Control Measures for Highway Runoff Treatment. J. Environ. Eng. 2012, 138 (1), 101−111. (12) Wang, G.-T.; Chen, S.; Barber, M. E.; Yonge, D. R. Modeling Flow and Pollutant Removal of Wet Detention Pond Treating Stormwater Runoff. J. Environ. Eng. 2004, 130 (11), 1315−1321. (13) Bentzen, T. R.; Larsen, T. Heavy Metal and PAH Concentrations in Highway Runoff Deposits Fractionated on Settling Velocities. J. Environ. Eng. 2009, 135 (11), 1244−1247. (14) Sébastian, C.; Becouze-Lareure, C.; Lipeme Kouyi, G.; Barraud, S. Event-based quantification of emerging pollutant removal for an 9268

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology (38) RSMeans Site Work & Landscape Cost Data 2015, 34th annual ed.; Fortier, R., Ed.; RSMeans: Rockland, MA, 2014. (39) RSMeans Building Construction Cost Data 2015, 73rd ed.; Plotner, S., Ed.; RSMeans: Rockland, MA, 2014. (40) US-EI 2.2 Database. EarthShift: Huntington, VT, 2010. (41) U.S. EPA. The Use of Best Management Practices (BMPs) in Urban Watersheds; EPA/600/R-04/184; Office of Research and Development: Washington, DC, 2004. (42) Illinois State Water Survey. Cook County Precipitation Network Daily Data Archive; Illinois State Water Survey: Champaign, IL, 2004−2013. (43) Pitt, R.; Maestre, A. National Stormwater Quality Database (NSQD), version 4.02; University of Alabama, Center for Watershed Protection: Tuscaloosa, AL, 2015. (44) Center for Watershed Protection. National Pollutant Removal Performance Database, version 3; Center for Watershed Protection: Ellicott City, MD, 2007. (45) Bare, J. TRACI 2.0: the tool for the reduction and assessment of chemical and other environmental impacts 2.0. Clean Technol. Environ. Policy 2011, 13 (5), 687−696. (46) U.S. EPA. Conversion Factors for Hydrocarbon Emission Components; EPA-420-R-10-015; Assessment and Standards Division Office of Transportation and Air Quality; U.S. Environmental Protection Agency: Washington, DC, 2010. (47) U.S. EPA. Exhaust and Crankcase Emission Factors for Nonroad Engine Modeling - Compression-Ignition; EPA-420-R-10-018; Assessment and Standards Division Office of Transportation and Air Quality; U.S. Environmental Protection Agency: Washington, DC, 2010. (48) US EPA. Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990−2014; EPA 430-R-16-002; U.S. Environmental Protection Agency: Washington, DC, 2016. (49) Chinowsky, P. S.; Price, J. C.; Neumann, J. E. Assessment of climate change adaptation costs for the U.S. road network. Glob. Environ. Change 2013, 23 (4), 764−773. (50) Mckay, M. D.; Beckman, R. J.; Conover, W. J. A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code. Technometrics 2000, 42 (1), 55−61. (51) Du, F.; Woods, G. J.; Kang, D.; Lansey, K. E.; Arnold, R. G. Life Cycle Analysis for Water and Wastewater Pipe Materials. J. Environ. Eng. 2013, 139 (5), 703−711. (52) O’Sullivan, A. D.; Wicke, D.; Hengen, T. J.; Sieverding, H. L.; Stone, J. J. Life Cycle Assessment modelling of stormwater treatment systems. J. Environ. Manage. 2015, 149, 236−244. (53) Stokes, J.; Horvath, A. Life Cycle Energy Assessment of Alternative Water Supply Systems. Int. J. Life Cycle Assess. 2006, 11 (5), 335−343. (54) International Organization for Standardization. ISO 14044: Environmental Management - Life Cycle Assessment - Requirements and Guidelines; ISO 14044:2006; International Organization for Standardization: Geneva, Switzerland, 2006. (55) Raynolds, M.; Fraser, R.; Checkel, D. The relative mass-energyeconomic (RMEE) method for system boundary selection Part 1: A means to systematically and quantitatively select LCA boundaries. Int. J. Life Cycle Assess. 2000, 5 (1), 37−46. (56) Suh, S.; Lenzen, M.; Treloar, G. J.; Hondo, H.; Horvath, A.; Huppes, G.; Jolliet, O.; Klann, U.; Krewitt, W.; Moriguchi, Y.; et al. System Boundary Selection in Life-Cycle Inventories Using Hybrid Approaches. Environ. Sci. Technol. 2004, 38 (3), 657−664. (57) Foley, J.; de Haas, D.; Hartley, K.; Lant, P. Comprehensive life cycle inventories of alternative wastewater treatment systems. Water Res. 2010, 44 (5), 1654−1666. (58) Birgisdóttir, H.; Bhander, G.; Hauschild, M. Z.; Christensen, T. H. Life cycle assessment of disposal of residues from municipal solid waste incineration: Recycling of bottom ash in road construction or landfilling in Denmark evaluated in the ROAD-RES model. Waste Manage. 2007, 27 (8), S75−S84. (59) Birgisdóttir, H.; Pihl, K. A.; Bhander, G.; Hauschild, M. Z.; Christensen, T. H. Environmental assessment of roads constructed

with and without bottom ash from municipal solid waste incineration. Transp. Res. Part Transp. Environ. 2006, 11 (5), 358−368. (60) Schwab, O.; Bayer, P.; Juraske, R.; Verones, F.; Hellweg, S. Beyond the material grave: Life Cycle Impact Assessment of leaching from secondary materials in road and earth constructions. Waste Manage. 2014, 34 (10), 1884−1896. (61) Tabor, M. L.; Newman, D.; Whelton, A. J. Stormwater Chemical Contamination Caused by Cured-in-Place Pipe (CIPP) Infrastructure Rehabilitation Activities. Environ. Sci. Technol. 2014, 48 (18), 10938− 10947. (62) Murphy, L. U.; Cochrane, T. A.; O’Sullivan, A. Build-up and wash-off dynamics of atmospherically derived Cu, Pb, Zn and TSS in stormwater runoff as a function of meteorological characteristics. Sci. Total Environ. 2015, 508, 206−213. (63) Flint, K. R.; Davis, A. P. Pollutant Mass Flushing Characterization of Highway Stormwater Runoff from an Ultra-Urban Area. J. Environ. Eng. 2007, 133 (6), 616−626. (64) Han, Y.; Lau, S.-L.; Kayhanian, M.; Stenstrom, M. K. Characteristics of Highway Stormwater Runoff. Water Environ. Res. 2006, 78 (12), 2377−2388. (65) Li, M.-H.; Barrett, M. E. Relationship Between Antecedent Dry Period and Highway Pollutant: Conceptual Models of Buildup and Removal Processes. Water Environ. Res. 2008, 80 (8), 740−747. (66) Kayhanian, M.; Singh, A.; Suverkropp, C.; Borroum, S. Impact of Annual Daily Traffic on Highway Runoff Pollutant Concentrations. J. Environ. Eng. 2003, 129 (11), 975−990. (67) Pagotto, C.; Legret, M.; Le Cloirec, P. Comparison of the Hydraulic Behaviour and the Quality of Highway Runoff Water According to the Type of Pavement. Water Res. 2000, 34 (18), 4446− 4454. (68) Michael Fitch, G.; Smith, J. A.; Clarens, A. F. Environmental Life-Cycle Assessment of Winter Maintenance Treatments for Roadways. J. Transp. Eng. 2013, 139 (2), 138−146. (69) Ozaki, H.; Watanabe, I.; Kuno, K. Investigation of the heavy metal sources in relation to automobiles. Water, Air, Soil Pollut. 2004, 157 (1−4), 209−223. (70) Maniquiz-Redillas, M. C.; Geronimo, F. K. F.; Kim, L.-H. Investigation on the effectiveness of pretreatment in stormwater management technologies. J. Environ. Sci. 2014, 26 (9), 1824−1830. (71) Andrés-Valeri, V. C.; Castro-Fresno, D.; Sañudo-Fontaneda, L. A.; Rodriguez-Hernandez, J. Comparative analysis of the outflow water quality of two sustainable linear drainage systems. Water Sci. Technol. 2014, 70 (8), 1341. (72) Barrett, M. E.; Walsh, P. M., Jr; Malina, J. F.; Charbeneau, R. J. Performance of vegetative controls for treating highway runoff. J. Environ. Eng. 1998, 124 (11), 1121−1128. (73) Lucke, T.; Mohamed, M.; Tindale, N. Pollutant Removal and Hydraulic Reduction Performance of Field Grassed Swales during Runoff Simulation Experiments. Water 2014, 6 (7), 1887−1904. (74) Fletcher, T. D.; Peljo, L.; Fielding, J.; Wong, T. H.; Weber, T. The performance of vegetated swales for urban stormwater pollution control. Bridges 2002, 10 (40644), 51. (75) Yousef, Y. A.; Hvitved-Jacobsen, T.; Wanielista, M. P.; Harper, H. H. Removal of contaminants in highway runoff flowing through swales. Sci. Total Environ. 1987, 59, 391−399. (76) Deletic, A. Modelling of water and sediment transport over grassed areas. J. Hydrol. 2001, 248 (1), 168−182. (77) Wong, T. H. F.; Fletcher, T. D.; Duncan, H. P.; Jenkins, G. A. Modelling urban stormwater treatmentA unified approach. Ecol. Eng. 2006, 27 (1), 58−70. (78) Backstrom, M. Sediment transport in grassed swales during simulated runoff events. Water Sci. Technol. 2002, 45 (7), 41−49. (79) Gartner, E. Industrially interesting approaches to “low-CO2” cements. Cem. Concr. Res. 2004, 34 (9), 1489−1498. (80) Juenger, M. C. G.; Siddique, R. Recent advances in understanding the role of supplementary cementitious materials in concrete. Cem. Concr. Res. 2015, 78, 71−80. (81) Office of Highway Policy Information, Federal Highway Administration (FHWA). 2008 Highway performance Monitoring 9269

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270

Article

Environmental Science & Technology System (HPMS); http://www.fhwa.dot.gov/policyinformation/tables/ 02.cfm (accessed Jun 9, 2016). (82) Kounina, A.; Margni, M.; Bayart, J.-B.; Boulay, A.-M.; Berger, M.; Bulle, C.; Frischknecht, R.; Koehler, A.; Milà i Canals, L.; Motoshita, M.; et al. Review of methods addressing freshwater use in life cycle inventory and impact assessment. Int. J. Life Cycle Assess. 2013, 18 (3), 707−721. (83) Corsi, S. R.; Graczyk, D. J.; Geis, S. W.; Booth, N. L.; Richards, K. D. A Fresh Look at Road Salt: Aquatic Toxicity and Water-Quality Impacts on Local, Regional, and National Scales. Environ. Sci. Technol. 2010, 44 (19), 7376−7382. (84) Gallego, A.; Rodríguez, L.; Hospido, A.; Moreira, M. T.; Feijoo, G. Development of regional characterization factors for aquatic eutrophication. Int. J. Life Cycle Assess. 2010, 15 (1), 32. (85) Struijs, J.; Beusen, A.; de Zwart, D.; Huijbregts, M. Characterization factors for inland water eutrophication at the damage level in life cycle impact assessment. Int. J. Life Cycle Assess. 2011, 16 (1), 59−64. (86) Azevedo, L. B.; Henderson, A. D.; van Zelm, R.; Jolliet, O.; Huijbregts, M. A. J. Assessing the Importance of Spatial Variability versus Model Choices in Life Cycle Impact Assessment: The Case of Freshwater Eutrophication in Europe. Environ. Sci. Technol. 2013, 47 (23), 13565−13570. (87) Basset-Mens, C.; Anibar, L.; Durand, P.; van der Werf, H. M. G. Spatialised fate factors for nitrate in catchments: Modelling approach and implication for LCA results. Sci. Total Environ. 2006, 367 (1), 367−382. (88) Helmes, R. J. K.; Huijbregts, M. A. J.; Henderson, A. D.; Jolliet, O. Spatially explicit fate factors of phosphorous emissions to freshwater at the global scale. Int. J. Life Cycle Assess. 2012, 17 (5), 646−654.

9270

DOI: 10.1021/acs.est.7b01856 Environ. Sci. Technol. 2017, 51, 9261−9270