Evaluation of Best Management Practices for Highway Runoff Control (2006)

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.

11.1 Motivation and Objectives Many forms of stormwater-related nonpoint source pollu- tion from highways are associated with detrimental water- quality characteristics of surface waters. Highways are the vital arteries of the nation, but prevention or mitigation of the discharge of pollutants from highways has become a primary goal for many jurisdictions, including state departments of transportation (DOTs). As vehicular traffic on highways has increased, vehicular-related pollution (oil and grease, heavy metals, nutrients, and sediment) has become an even greater problem. Furthermore, highways serve as a streamlined means of transport for other sources of pollution such as irri- gation run-on, pesticides and fertilizers from landscaped areas, and particulates from pavement breakdown. Finally, because of limited infiltration and expedited transport of runoff, paved surfaces promote a variety of indirect water- quality problems such as higher temperature of discharge and increased flooding hazards. In recognition of the urgent need by highway engineers and related environmental professionals for guidance, the National Cooperative Research Program (NCHRP) initiated NCHRP Project 25-20(1) with a goal of providing a means for evaluating best management practices (BMPs) and low- impact development (LID) for stormwater quantity and qual- ity. An evaluation scheme for implementation of BMP and LID facilities in the highway environment has been prepared. To the extent possible, fundamental principles of environ- mental engineering unit processes are applied to the analysis of the many BMP/LID options. In addition to considering hydrologic and water-quality issues, management practices are considered in the evaluation scheme, including issues of safety, operation and maintenance, practicability constraints, regional issues, costs, and other concerns. The evaluation methods developed during this project include methods applicable to both BMP and LID facilities. This report pres- ents findings related to the research that are moderately peripheral to the main goal of presenting screening guidance to highway engineers for selection and preliminary design of facilities for control of stormwater quality and quantity. That guidance is presented in CRP-CD-63 (affixed to the back cover of this book), which contains the User Guide for BMP/LID Selection, designated henceforth in this document by the briefer Guidelines Manual. While there are several reports and web sites with design guidance for BMPs, less guidance is available for LID, especially in the highway con- text. Hence, an additional project report, the Low-Impact Development Design Manual for Highway Runoff Control (des- ignated henceforth as the LID Design Manual and also avail- able on CRP-CD-63), presents detailed design guidelines for LID facilities in the highway environment. Two rainfall-runoff models were used extensively in the project to simulate regional hydrologic impacts on BMP per- formance. One model, the USEPA Storm Water Management Model (SWMM), may be obtained directly from that agency at http://www.epa.gov/ednnrmrl/models/swmm/index.htm. The second model is a spreadsheet model developed at the University of Florida, primarily for this project. That model and its documentation are also included in CRP-CD-63. Modeling results are presented in detail in the Guidelines Manual and its appendices. 1.2 Background The Clean Water Act was revised in 1987 in an attempt to address nonpoint source pollution via the National Pollutant Discharge Elimination System (NPDES). As a result, state agencies—such as DOTs—as well as cities, counties, and municipalities were required to meet discharge requirements for runoff originating within their jurisdictions. “Best man- agement practice” (BMP) became probably the three most common words in the stormwater management vocabulary and were used to describe everything from street sweeping to constructed wetlands, regardless of whether a particular C H A P T E R 1 Introduction and Objectives

management measure was the “best” management practice for the site conditions and constraints. In the late 1990s, decentralized hydrologic source control was formalized as low-impact development (LID) by the Prince George’s County Department of Environmental Regulation (Prince George’s County 1997, 2000) in Maryland as a viable alter- native to watershed outlet (“end-of-pipe”) treatment. Anal- ogous to the indiscriminate use of BMPs whether or not they are in fact the “best” management practices for a particular site is the broad interpretation of the term “LID,” especially with regard to what constitutes “low” impact. Nonetheless, the principle of decentralized, on-site retention of storm- water is of great value. 1.3 What’s in a Name (Typology of Wet-Weather Control)? Wet-weather controls have various names around the world. BMPs are best known in North America (and perhaps worldwide). However, “sustainable urban drainage systems” (SUDS) is the terminology used in the United Kingdom, and “stormwater quality improvement devices” (SQIDs) is used in Australia. The name SUDS suggests that these systems rank higher on the sustainability scale (CIRIA 2000a, 2000b, 2000c). However, the term “sustainability” is not very well defined in the science of stormwater management. One gen- eral definition is that sustainability requires a balance among community development, economic development, and eco- logical protection (ICLEI 1996). The concept of sustainabil- ity can also mean calculating the cost of a wet-weather control based on compliance or resource protection over the life cycle of the project rather than just on the expenditures necessary to complete the project. LID has a more restricted meaning (decentralized hydro- logic source control) than does BMP, but it is also widely used in the United States. The LID approach is based on selecting “integrated management practices” (IMPs), which are dis- tributed small-scale controls that can closely maintain or replicate the hydrology of predevelopment conditions or achieve another identified regulatory requirement or other resource protection goals. Rather than working from a small range, or list, of BMPs, the goal is to achieve the highest effi- ciency in or effectiveness at approximating the predevelop- ment condition or other requirements. BMP and LID are most often associated with control of stormwater only, whereas runoff from urban areas can occur during dry weather as well (e.g., baseflow in channels from irri- gation and other common urban sources can cause runoff), and many of the same control principles (discussed below) apply to control of wet-weather phenomena such as combined sewer overflows (CSOs) and to dry-weather, sanitary sewer overflows (SSOs) (due to infiltration into the sewer system). With the exception of some baseflow in arid areas, most discharges do originate as a result of current or recent rain- fall. However, there appears to be no universally accepted ter- minology for control of such discharges. If a practice is shown to be the most cost-effective control, then it would be the “best” wet-weather control. This term combines and synthe- sizes all of the key characteristics of BMPs, IMPs, and SUDS. Wright and Heaney (2001) presented an overview of how dis- tributed BMPs (wet-weather controls) can be an integral and cost-effective component of stormwater management in urban areas. They argue that sustainability principles such as decentralized or distributed systems may provide better long- term solutions because the stormwater is managed close to its source in a distributed manner. The consensus within the highway engineering community is that whether or not BMP is the best terminology for a wet- weather control, almost every drainage engineer has a good idea of what is meant by BMP. BMP is generally an inclusive term, one that includes LID-like devices that emphasize infil- tration and evapotranspiration (ET) for retention of storm- water, but it often connotes an end-of-pipe treatment facility, such as a detention pond or wetland. Nonetheless, BMP is used throughout this report and the Guidelines Manual to mean control of discharges from highway and urban areas that typically originate from rainfall. BMP in this document will often, but not always, include LID within its meaning. BMP/LID will be used, despite its awkwardness, when it is necessary to emphasize application to both. 1.4 Environmental Engineering Principles Many organizations have developed their own stormwater BMP design manuals, and a large number are currently in existence. While many of these manuals are quite good and provide helpful recommendations on choosing and sizing structural stormwater BMPs, a number of them lack a con- ceptual framework for addressing specific stormwater-quality and -quantity issues occurring at a particular site. The general approach in most manuals that are currently available is to choose a BMP that has been shown to address the pollutants of concern and then apply “rules-of-thumb”sizing and design methods. While this is often an appropriate and valid approach, it does not adequately build upon more than a cen- tury of accumulated experience in the fields of environmen- tal process and wastewater engineering (e.g., Metcalf and Eddy 2003) and indeed the full suite of technical skills and experience available to professional civil and environmental engineers. Use of treatment trains (BMPs in series) and inte- gration of fundamental unit operations and processes (UOPs) are some of the most basic and profound concepts of environmental engineering. Unfortunately, these concepts 2

have only recently been advocated as design approaches for stormwater treatment. As stormwater regulations continue to become more stringent, the need for more advanced treat- ment technologies will grow. However, in order to meet this need, the collective knowledge and understanding of stormwater treatment must be reduced to a more fundamen- tal level, which may require drainage engineers to be open to new ideas, as well as the tried-and-true treatment technolo- gies of the wastewater industry. Because of the complexity of the fundamental unit processes for treating many stormwater constituents, available information on field-verified treatabil- ity is currently limited. Thus, until the knowledge base of UOPs for stormwater treatment is expanded, reliance on the- oretical principles and laboratory analyses will be needed. The Guidelines Manual of this project provides a frame- work for applying fundamental principles of UOPs, such as those commonly applied in water and wastewater engineer- ing, to aid in the evaluation and selection of runoff manage- ment and treatment control systems for highway and urban areas. As opposed to other design approaches that recom- mend the selection of BMPs based solely on documented per- formance factors such as percent removal and effluent quality and/or percent capture, the design approach presented in this project is to first select the UOPs that address the pollutants of concern and then to individually select treatment system components (TSCs) based on those UOPs. However, in accordance with the earlier discussion of terminology, BMP is used generically instead of TSC. Within this research report, UOPs are discussed in the context of their incorporation into common BMPs. This information serves as a technical back- ground for the BMP/LID evaluation strategy presented in the Guidelines Manual. LID facilities treat stormwater according to exactly the same UOP principles, and this research report also provides background on the design principles described in the LID Design Manual. Available stormwater treatability options as a function of complexity and scope are presented schematically in Figure 1-1. Within this project, the Guidelines Manual and the LID Design Manual refer to the material presented in the lower part of the figure, whereas this research report deals more with the principles embodied in the middle and top parts of the figure. 1.5 Taxonomy of Road and Drainage Systems It is useful to place stormwater issues for highways in con- text. FHWA statistics classify urban and rural roads according to population density rather than design capacity or other functional characteristics. The FHWA provides statistical data concerning highway planning, development, financing, con- struction, operation, modernization, maintenance, safety, and traffic conditions (http://www.fhwa.dot.gov/policy/ ohim/hs04/index.htm). These data are needed to meet DOT responsibilities to Congress and the general public. The data are not used for roadway design purposes and/or BMP design, but rather as an accounting and planning tool. Nonetheless, the data are useful in conveying the enormous magnitude of highways in the United States. Nearly four million miles of road exist in the United States, as shown in Table 1-1. Over 77% of these roads are in rural areas. In keeping with this report’s focus on the handling and treatment of stormwater,“urban”and “rural”are defined here according to the type of runoff conveyance system used. The 2006 Florida DOT Drainage Manual (State of Florida DOT 2006) differentiates between urban and rural as follows, with respect to major culvert installations:“Urban facilities include any typical section with a fixed roadside traffic barrier such as curb or barrier wall. Additionally, rural typical sections with greater than 1,600 ADT are also included in this [major cul- vert installation] category.”The implication is that rural roads generally have open drainage, such as ditches, whereas urban roads may have piped systems to accommodate curbs and gutters. The focus of this project is on how BMP and LID facili- ties can be incorporated into both types of DOT road sys- tems. Typically, public transportation agencies own and manage about 20% of these roads. (Some states, such as Vir- ginia, maintain almost all of the highway and local roadway systems.) In this project, attention has been restricted to the typical DOT highways, to focus efforts on the 20% of the transportation network that carries the bulk of the traffic. The remaining 80% of the roads are used less intensively, and many of them are primarily access roads; however, many stormwater control practices can be applied to these situations as well. DOT highways are those that are consid- ered to be “large-scale” facilities with extensive infrastruc- ture designed either to convey water generated from the roadway system and/or to convey off-site stormwater through the system. The mix of roads owned by state DOTs is shown in Table 1-2. Nearly 86% of these roads are rural. Interstate and other expressways account for about 19.4% of the total mileage. Minor arterial roads comprise 28.5% of the mileage, other principal arterials comprise 21.5%, and local roads comprise 20.4%. Interstates, other expressways, and other principal arterial roads typically have four or more lanes and carry a disproportionate amount of the traffic flow. Thus, their rela- tive importance would increase substantially if lane-miles and/or traffic flow were used as the measure of activity. Heaney (2000) reviewed opportunities for using decentral- ized or distributed wet-weather controls associated with trans- portation activities and concluded that the most promising stormwater control opportunities are associated with smaller 3

access roads and parking facilities that are infrequently used. The key factor in reducing the impact of transportation- related wet-weather flows was the ability to avoid curbs and gutters, which result in closed drainage systems. Much of the general literature about LID has been devoted to promoting distributed open drainage systems in lower- activity areas. However, the objective of this research project is to identify opportunities and develop strategies for using BMP/LID on high-volume roads with open and closed drainage systems. Much of the work is based on the success of, and lessons learned from, the use of stormwater controls in less-intense roadway systems and in land development projects. To summarize: the focus of this project is on large, linear highways as opposed to the vehicular transportation system of a typical urban area. That is, the focus is away from 4 Figure 1-1. Available stormwater treatment options as a function of complexity. Miles Organization Rural Urban Total % of Total State Highway Agency 662,855 110,434 773,289 19.5 County 1,628,510 144,615 1,773,125 44.7 Town or City 606,389 624,163 1,230,552 31.0 Other Jurisdictions 56,254 12,695 68,949 1.7 Federal Agency 117,751 2,819 120,570 3.0 Total 3,071,759 894,726 3,966,485 100.0 Percent of Total 77.4 22.6 100.0 Source: Office of Highway Policy Information 2002. Miles Type of Road Rural Urban Total % of Total Interstate 31,445 12,528 43,973 5.7 Other Expressways 97,784 8,447 106,231 13.7 Other Principal Arterials 130,362 35,787 166,149 21.5 Minor Arterials 195,939 24,535 220,474 28.5 Collector 67,092 11,726 78,818 10.2 Local 140,233 17,386 157,619 20.4 Total 662,855 110,409 773,264 100.0 Percent of Total 85.7 14.3 100.0 Source: Office of Highway Policy Information 2002. Table 1-1. Ownership of U.S. highways, 2002. Table 1-2. Roads owned by state highway agencies, 2002.

neighborhoods and arterials, even though virtually the same guidance methodology could be applied in those set- tings as well. A parallel research effort for the Water Envi- ronment Research Federation (WERF) (Strecker et al. 2005) extends the BMP selection methodology to the more general urban situation. 1.6 This Document This research report mainly presents information in sup- port of the BMP selection methodology provided in the Guidelines Manual and of the design methods presented in the LID Design Manual. This approach allows the other two documents to have a more practice-oriented focus. This research report also presents other general background infor- mation as well as recommendations applicable to the tech- nology presented in this project. The performance and effectiveness of a BMP for treating and/or controlling stormwater runoff depend upon numerous variables, related not only to the design and operation of the system, but also to the conditions of the site, techniques related to sampling, and constituents found in the water (Strecker et al. 2001). These variables, when combined with the discrepancies and inconsistencies regarding analysis and reporting of data, prevent many performance assessments from being used on a widespread basis (Strecker et al. 2001). Thus, the selection, design, and in some instances, approval status, of a particular system for use on a site may not be consistent from jurisdiction to jurisdiction. Background on typical stormwater BMPs and on application of LID in the highway environment is presented in Chapters 2 and 3, respectively. Stormwater and fundamen- tal processes for its treatment are characterized in Chapter 4. The influence of highway and hydrologic characteristics is described in Chapters 5 and 6, while institutional and many other regional influences are described in Chapter 7. Chapter 8 treats the topic of performance evaluation. In order to ensure that BMP/LID facility performance data may be transferable and comparable between locations and types of BMP systems and thus ensure that the evaluation and selection of a facility is consistent, the overall evaluation methodology allows for both performance- and practicability-based assess- ment by evaluating BMP performance data from the Interna- tional BMP Database (see Section 8.3) (Strecker 1994). Principles of environmental engineering unit operations as well as BMP/LID performance data have been integrated into an overall evaluation strategy that is outlined in Chapter 9 and pre- sented in detail in the Guidelines Manual. A major modeling effort has been conducted to evaluate regional hydrologic influ- ences on BMP/LID performance. Extensive results, based on continuous simulation of highway runoff quantity and quality, are presented for both flow-limited (meaning minimal storage, such as filters and some proprietary devices) BMPs and off-line and on-line volume-limited BMPs (e.g., ponds and extended detention basins). The basis for these results is presented in Chapter 10. A summary, conclusions, and recommendations are presented in Chapter 11. 5