104 - Hydrology and Hydraulics
104.1 Introduction
The primary objective in the design of a highway stream crossing is to avoid causing interruption of the traffic using the bridge or crossing and changes in the behavior of the stream. Other objectives of a hydraulic design are to determine the backwater and hydraulic capacity of the bridge or culvert; to identify the stream forces that may cause damage to the bridge, culvert or roadway system; and to provide a safe level of service acceptable to the traveling public without causing unreasonable effects on adjacent property or the environment.
104.1.1 Terms
ATON – Aids to Navigation
CBF – Channel bed fill
HEC-HMS – U.S. Army Corps of Engineers (USACE) Hydrologic Engineering Center (HEC) Hydrologic Modeling System
HEC-RAS – USACE HEC River Analysis System
HFAWG – Hydrologic Frequency Analysis Work Group
HY-8 – FHWA Culvert Hydraulics Computer Program
LiDAR – Light Detection and Ranging remote sensing method
NAVD 88 – North American Vertical Datum of 1988
PDM – DelDOT’s Project Development Manual (PDM; 2015)
PeakFQ – U. S. Geological Survey (USGS) computer program to estimate magnitude and frequency of floods
StreamStats – USGS web-based geographic information system (GIS) that provides analytical tools that are useful for engineering design applications, such as the design of bridges
TR-20 – Natural Resources Conservation Service’s (NRCS’s) hydrologic computer program
TR-55 and WinTR-55 – NRCS’s hydrologic method and computer program, respectively
UDC – New Castle County Unified Development Code
WATSTORE – National Water Data Storage and Retrieval System of the USGS
104.1.2 Coordination
Consideration of the effects of constructing a bridge or culvert across a waterway is key to ensuring the long-term stability of the structure. Confining the floodwater may cause excessive backwater or overtopping of the roadway, may impact structural stability when the water is impacting the superstructure of the bridge (i.e., causing a pressure flow situation), or may induce excessive scour. These effects may result in damage to upstream land and improvements or endanger the bridge. Conversely, an excessively long bridge does not create a backwater or any attenuation and may cost far more than can be justified by the benefits obtained. Somewhere between these extremes is the design that will be the most economical to the public over a long period of time, yet remain safe and stable during large storm events.
Standard DelDOT QA/QC procedures will be followed for development and review of hydrology and hydraulics submittals.
104.1.3 Design Responsibilities
Responsibilities for hydraulic design are divided between the Bridge Design Section and the Project Development Sections based primarily on the size of the drainage area. Bridge Design is responsible for all watersheds equal to or over 300 acres and existing structures with openings (bridge, culvert, pipes) that exceed 20 square feet. The Project Development Section is responsible for watersheds smaller than 300 acres. The Bridge Design Section is responsible for “bridge-only” projects where support from the Project Development Groups is not required. In those cases, the Bridge Design Section designs any pipe culverts, closed drainage and roadside ditches, and stormwater management systems affiliated with the bridge project. Typical projects include bridge replacement or rehabilitation projects.
When the Bridge Design Section collaborates on a project with the Project Development Section, the Project Development Section will develop the closed drainage and roadside ditches. A new alignment bridge is a typical project in which this type of coordination takes place: the Bridge Design Section designs the structure, while the Project Development Section designs the ramps, profiles, alignment, drainage, and all other aspects of the project.
Refer to Chapter 6 of the DelDOT Road Design Manual (2004) for the design and construction of adjacent drainage ditches, pipe culverts (less than 20 square feet), closed drainage systems, and erosion control near stream crossings.
104.1.4 Field Data Collection
One of the first and most important aspects of any hydraulic analysis is a field evaluation. This involves an in-depth inspection of the proposed bridge site and completion of the Field Hydraulic Assessment Checklist in Appendix 104-1. The designer is responsible for completing the checklist.
The purpose of field inspecting the proposed bridge site is to evaluate the stream characteristics and hydraulic properties, the performance of the existing bridge (if applicable), the channel and floodplain topography, and the adequacy and accuracy of the survey data. Any man-made dams located in the reach that will affect the bridge should also be investigated. Additionally an estimate of streambed particle size, including D50, can be made by visual inspection using field tools such as a sand gage card, gravelometer, or wire screen.
The designer should walk along the channel both upstream and downstream at a distance at least equal to the floodplain width, if possible. Any natural hydraulic controls such as rock shoals, or beaver dams as well as man-made controls such as bridges, dams, sewer or water lines suspended across the channel, or other constrictions that have taken place in the floodplain should be evaluated. If these controls have any effect on the high-water profile, they should be taken into account in the modeling. The stream alignment and relation to structure (e.g., outside of bend, bad angle of attack) should also be noted. Coordination is recommended with the Environmental Studies Section to determine if current environmental study, wetland delineation, and/or biological stream section forms are available that have any of the required information described above.
104.1.5 Topographic Survey and Extent of Hydraulic Study
Data for the project will be developed from available survey data and USGS, LiDAR, or other topographic mapping. If sufficient data are not available, additional survey data will have to be obtained. The channel and hydraulic controls should be surveyed so that their effects on the high-water profile can be defined. NAVD 88 is the required datum for hydraulic surveys and studies. Elevation contours at 2-foot intervals for the State of Delaware were produced for New Castle and Kent Counties (based on the 2007 LIDAR) and for Sussex County (based on the 2005 LIDAR.) Data are in line shapefile format. LiDAR data is typically useful for overbank elevation data; however, LiDAR data do not provide elevation data in the stream channel, so a survey is required. The LiDAR data and specifications with respect to the data may be accessed from the Delaware Geological Survey.
Data that will need to be gathered from a field survey include data on stream banks and the channel, any required dam data, and bridge/culvert data. If LiDAR data are available for data in the overbanks the survey of the channel and structures can be merged with the LiDAR data. If LiDAR data is not going to be used, the survey should include the overbank area with the lateral extents of the topographic data to contain the 100-year event within the hydraulic cross sections. A survey is required for all projects that require an H&H analysis, and it is the designer’s responsibility to request the survey. Any specific information needed for the Hydraulic Checklist or information in addition to that normally required must be included in the survey request. See Appendix 104-2 for a sample survey request.
For hydraulic studies, the downstream and upstream limits vary based on a number of factors, including tidal influences, other structures within the reach, backwater from other streams/rivers, and the slope of the channel. Streams with flatter slopes or with backwater conditions from a downstream river typically require a longer study reach to be able to balance energies and get an accurate analysis at the bridge.
The limits of the profile computation should be extended downstream to the point where a flow is not affected by the structure (i.e., the flow has fully expanded). This downstream limit can be determined by computing a sensitivity analysis. The HEC-RAS model can be executed starting at normal depth, and then subsequent runs can be started 1 foot below and above normal depth to see if the model converges before the location of the proposed bridge, as shown on Figure 104‑1. The expansion reach length is defined as the distance from the cross section placed immediately downstream of the bridge to the cross section where the flow is assumed to be fully expanded. Chapter 5 of the HEC-RAS River Analysis System Hydraulic Reference Manual (USACE HEC, 2010) provides additional guidance on determining the distance to the downstream end of the expansion reach.
The upstream limit should extend to where any increase from the new bridge or proposed modifications merges into the existing conditions profile (e.g., where the flow lines are approximately parallel and the cross section is fully effective). If the proposed conditions water surface elevation (WSE) is lower than the existing conditions profile, then the minimum distance upstream to be modeled shall be 500 feet. The model should be calibrated using known flood data if sufficient reliable data is available.
Note that for small in-kind pipe or culvert replacements with minimal changes to the hydraulic opening, width, and roadway profile, the upstream and downstream hydraulic limits may be shortened as appropriate. Also, for small projects that use HY-8 or a similar culvert modeling methodology and that do not require backwater calculations, a limited survey is required to define the downstream tailwater condition and the existing structure and roadway data.
104.2 Hydrology
104.2.1 Introduction
Hydrologic analysis is used to determine the rate of flow, runoff, or discharge that the drainage facility will be required to accommodate. The designer must evaluate existing upstream conditions in sizing a structure. If warranted, the designer may evaluate the potential effects of future land cover conditions on calculated flows by using procedures outlined in USGS Scientific Investigations Report (USGS SIR) 2006-5146, Magnitude and Frequency of Floods on Nontidal Streams in Delaware (2006) or by exercising engineering judgment.
104.2.2 Documentation
The design of highway facilities should be adequately documented. It is frequently necessary to refer to plans, specifications, and hydrologic analyses long after the actual construction has been completed. One of the primary reasons for documentation is to evaluate the hydraulic performance of structures after large floods to determine whether the structures performed as anticipated or to establish the cause of unexpected behavior. In the event of a failure, it is essential that contributing factors be identified to avoid recurring damage and help improve future hydraulic designs.
The documentation of a hydrologic analysis is the compilation and preservation of all pertinent information on which the hydrologic decision was based. This might include drainage areas and other maps, field survey information, source references, photographs, hydrologic calculations, flood-frequency analyses, stage-discharge data, and flood history, including narratives from highway maintenance personnel and local residents who witnessed or had knowledge of an unusual event.
Hydrologic data shown on project plans ensure a permanent record, serve as a reference in developing plan reviews, and aid field engineers during construction. Required plan and H&H Report presentation data are provided on Figure 104‑2 and in Section 104.6 – Hydrologic and Hydraulic Report.
104.2.3 Precipitation
Several hydrologic methods that can be used to estimate flows will require precipitation data in the form of total precipitation or as rainfall intensity as part of the hydrologic input for the method. As precipitation data are regional, the National Oceanic and Atmospheric Administration (NOAA) Atlas 14 has published rainfall intensity-duration data for Delaware’s 12 rainfall gages located throughout the state. Precipitation values in Figures 6-5 through 6-7 of the Road Design Manual correlate well with the rainfall data in NOAA Atlas 14. Precipitation intensity values for use in the Rational Method may be obtained from Figures 6-5 through 6-7. Figure 104‑3 provides 24-hour rainfall totals for use in methods requiring a 24-hour duration, such as the NRCS Curve Number method. Figure 6.11 in the Road Design Manual provides these values.
104.2.3.1 The Rational Method
The rational method is an empirical formula relating rainfall to runoff. It is the method used almost universally for computing urban runoff. It is also used to estimate bridge deck drainage for the design of scuppers.
Discharge, as computed by this method, is related to frequency by assuming the discharge has the same frequency as the rainfall used. The storm duration is set equal to the time of concentration of the drainage area. Because of the assumption that the rainfall is of equal intensity over the entire watershed, it is recommended that this formula should be used only for estimating runoff from small areas. Although the rational method is typically only applied to a maximum watershed size of 200 acres, with caution and consideration for watershed characteristics, larger watersheds up to 326 acres (the lower limit of the regression method) may be applicable. The rational method is most frequently used for estimating small, homogenous, or highly impervious drainage areas.
Section 6.6.3.1 of the Road Design Manual provides more specifics on use of the rational method, including the procedure, time-of-concentration (Tc) calculations, acceptable “C” value sources, and determination of rainfall intensity. It should be noted that the Road Design Manual provides equations to calculate Tc for the rational method (Section 6.6.3.1) that are different from those that it provides for the NRCS curve number method (Section 6.6.3.2). The rational “C” values should be obtained from Figure 6-8 of the Road Design Manual.
104.2.3.2 Delaware Regression Method (SIR 2006-5146)
DelDOT uses the equations in the current version of the SIR 2006-5146 to estimate flood runoff. These equations are based on specific studies of the nontidal watersheds in Delaware and adjacent states. This method relies on data from streamflow gaging station records combined statistically within a hydrologically homogenous region to produce flood-frequency relationships applicable throughout the region. If the designer is using gaging station records and wishes to evaluate these values for upstream or downstream sites, the procedures in the USGS publication should be followed.
From the study, it was concluded that reasonable estimates of flood runoff can be made by dividing the state into two hydrologic regions, which correspond to the Coastal Plain and Piedmont physiographic regions as shown on Figure 104‑4. In the Piedmont region, the size of the drainage area, percent of forest, percent of hydrologic soil group “A” and percent of storage from the National Hydrography Dataset (NHD) are considered in the equations. The variables used vary based on the design event. In the Coastal Plain region, the mean basin slope (in percent) determined from a 10-meter digital elevation model (DEM) must be considered in addition to the drainage area and percent hydrologic soil group A. Each of these parameters is discussed in the SIR 2006-5146 publication. Land use is not considered in the runoff equations for the Coastal Plain region.
In areas where land use may change, the empirical methods using lump parameters or models such as WinTR-55, HEC-HMS, or HEC-1 are recommended. If the Delaware regression method is used, based on engineering judgment, the designer may consider the effects of possible changes in land use.
The SIR 2006-5146 method is incorporated into the USGS online StreamStats program. StreamStats is a web-based GIS that provides users with access to an assortment of analytical tools that are useful for water-resources planning and management and for engineering design applications, such as the design of bridges. StreamStats allows users to easily obtain streamflow statistics, drainage-basin characteristics, and other information for user-selected sites on streams.
The best estimates of flood frequencies for a site are often obtained through a weighted combination of estimates produced from the regression results and the results from a statistical analysis of stream gage data. The U.S. Department of the Interior, Interagency Advisory Committee on Water Data (1982) recommends, and Tasker (1975) demonstrated, that if two independent estimates of a streamflow statistic are available, a weighted average will provide an estimate that is more accurate than either of the independent estimates. Improved flood-frequency estimates can be determined for Delaware stream gaging stations by weighting the systematic peak-flow record estimates at the station with the regression peak flow estimates. SIR 2006-5146 provides guidelines for the weighting process as well as procedures and equations to estimate flows for a site upstream or downstream of a gaged location and for sites between gaged locations.
104.2.3.3 Published Reports
Published reports may be used for comparison with the calculated runoff. Delaware Department of Natural Resources and Environmental Control (DNREC) has developed or is in various stages of developing watershed stormwater management plans for the Appoquinimink, Upper Nanticoke, and Murderkill watersheds. These plans include a detailed hydrologic model that has been calibrated against stream gage data, data obtained by regression methods, or other reliable hydrologic data. The watersheds were divided into subwatersheds; therefore, flow values at road crossings may be available that are not available from any other source. For projects located in these watersheds, these reports should be reviewed and flows used as appropriate. The calibrated HEC-HMS files may be available from DNREC.
The Federal Emergency Management Agency (FEMA) Flood Insurance Study (FIS) contains runoff information for many streams in Delaware. The report documents the methods used to determine runoff for each stream. The FIS reports were prepared by a variety of sources (e.g., the USACE, private consultants, the Delaware River Basin Commission). These reports contain floodplain information for many streams in Delaware. The reports include historical runoff data as well as calculated runoff data. However, due to the variety of preparers, the flows should be checked against other reliable methods. The flow values reported in these other reports should be verified as to consistency with the standards presented herein, and checked for validity of data utilized, and methodology. Published dam reports should also be referenced as contained in Section 104.3.3 – Hydraulics for Dam Safety Projects.
The USACE and FEMA have developed and applied a state-of-the-art storm surge risk assessment capability for Region III, which includes the Delaware Bay, the Delaware-Maryland-Virginia Eastern Shore, and all the waterways connected to these systems. This information, some of which is contained in ERDC/CHL TR-11-1 Coastal Storm Surge Analysis: Modeling System Validation Report 4 (USACE ERDC, 2013), would be helpful for any tidal bridges.
104.2.3.4 Flood-Frequency Analysis of Recorded Stream Gage Data
The method of analyzing flood-frequency relationships from actual streamflow data for a single gaging station enables the use of records of past events to predict future occurrences. The procedures described in Guidelines for Determining Flood Flow Frequency, Bulletin 17B (U.S. Department of the Interior, Interagency Advisory Committee on Water Data, 1982) should be followed. This method is often referred to as the Bulletin 17B method, and uses the Log Pearson Type III distribution. The Log-Pearson Type III distribution is a statistical technique for fitting frequency distribution data to predict the design flood for a river at some site and is performed on records of annual maximum instantaneous peak discharges collected systematically at streamflow gaging stations.
104.2.3.4.1 Flood-Frequency Analysis Guidelines
The HFWAG, consisting of representatives from Federal agencies, private consultants, academia, and water management agencies, has recommended procedures to increase the usefulness of the current guidelines for hydrologic frequency analysis and to evaluate other procedures for hydrologic frequency analysis (HFAWG, 2013). The HFAWG will be incorporating their findings into Bulletin 17C. When Bulletin 17C is adopted by the FHWA, the procedures in that publication should supersede those in Bulletin 17B.
The computer programs PeakFQ, developed by the USGS, and Hydrologic Engineering Center Statistical Software Package (HEC-SSP), developed by the USACE HEC, provide estimates of instantaneous annual-maximum peak flows for a range of recurrence intervals. The Pearson Type III frequency distribution is fitted to the logarithms of instantaneous annual peak flows following the Bulletin 17B guidelines of the U.S. Department of the Interior Interagency Advisory Committee on Water Data. The parameters of the Pearson Type III frequency curve are estimated by the logarithmic sample moments (mean, standard deviation, and coefficient of skewness) with adjustments for low outliers, high outliers, historic peaks, and generalized skew.
PeakFQ reads annual peaks in the WATSTORE standard format and in the Watershed Data Management (WDM) format. Annual peak flows are available from NWISWeb (http://nwis.waterdata.usgs.gov/usa/nwis/peak). (Data should be retrieved in the WATSTORE standard format, not the tab-separated format.)
This method assumes that there are no changes during the period of record in the nature of the factors causing the peak magnitudes. The ramifications of this assumption can be minimized by making every effort to determine the past conditions of the drainage area and, if possible, making allowances for changes. The most common changes are man-made and consist of such modifications as storage and land development. The user of hydrologic data must be acquainted with the procedures for evaluating streamflow data, the techniques for preparing a flood-frequency curve, and the proper interpretation of the curve.
Since most of the stream records in Delaware are sufficiently long to give good flood-frequency relationships, considerable weight should be given to the stream record in estimating design floods. When a gage record is of short duration or poor quality, or when the results are judged to be inconsistent with field observations or sound engineering judgment, the analysis of the gage record should be supplemented with other methods. The validity of a gage record should be demonstrated and documented. Gage records should contain at least 10 years of consecutive peak flow data, and they should span at least one wet year and one dry year. If the runoff characteristics of a watershed are changing (e.g., from urbanization), then a portion of the record may not be valid.
Where there is a stream gage at a bridge or culvert, the USGS has developed the flood flow frequency for the 50- , 20-, 10-, 4-, 2-, 1-, 0.5-, and 0.2-percent-chance of occurrence (2-, 5-, 10-, 25-, 50-, 100-, 200-, and 500-year) events; these frequencies are reported in Table 6 of USGS SIR 2006-5146. In addition, the publication reports calculated SIR 2006-5146 values as well as a weighted value based on the statistical analysis value and the regression method in Table 8. As time elapses and more annual rainfall events are recorded, the data in these tables will become outdated.
There will be times when estimates made from a flood-frequency analysis of a gaging station on the stream being studied will not agree with a regional analysis, such as the SIR 2006-5146 method. Various factors such as length of runoff records, storm distribution, and parameters used in the regional analysis could account for some of the discrepancies. When gaging station records are used, the designer should consult SIR 2006-5146 and current USGS data.
104.2.3.4.2 Transposition of Flows
When a project site falls between 0.5 and 1.5 times the drainage area of a stream gaging station on the same stream, the flow may be transposed to the project site using the methodology presented in SIR 2006-5146 (equation 22, page 31).
104.2.3.5 Other Methods/Models
104.2.3.5.1 NRCS TR-55 Curve Number Method (WinTR-55 Program)
Technical Release 55 (TR-55) Urban Hydrology for Small Watersheds is an NRCS, formerly the Soil Conservation Service (SCS), curve number method and is applicable to small urban watersheds (NRCS, 1986). The report provides a graphical and a tabular method for computing peak discharges of drainage basins with areas ranging from 10 acres up to 2,000 acres (3.1 square miles); however, the Road Design Manual states that TR-55 can be used for complex watersheds up to 300 acres. The required input data are drainage area, curve number (which is a function of land cover and hydrologic soil group), and a Tc. The Delaware soils and their assigned hydrologic soil group are shown on Figure 6-10 of the Road Design Manual.
TR-55 uses a segmental method to compute Tc (i.e., flow time is computed by adding the times for the overland, shallow concentrated, and channel segments). Chapter 6 of the Road Design Manual provides the methodology and equations to compute Tc. Travel time (Tt) is the ratio of flow length to flow velocity.
This method must also meet the following conditions:
- Assumes that rainfall is uniformly distributed over the entire basin.
- Basin is drained by a single main channel or by multiple channels with times of concentration (Tc) within 10 percent of each other.
- Tc is between 0.1 and 10 hours.
- Storage in the drainage area is ≤ 5 percent and does not affect the time of concentration.
- Watershed can be accurately represented by a single composite curve number.
The TR-55 method greatly overestimates runoff for very flat watersheds in the Delaware coastal plain. If TR-55 is used, the Curve Number must be calibrated by comparing the flows generated by TR-55 against results from another method in this section, such as the Delaware regression method. Curve Numbers must be adjusted to match the desired peak of the design event within the 90-percent confidence interval, upper limit.
Limitations of TR-55 are described on page 5-3 of the Small Watershed Hydrology: WinTR-55 User Guide (USDA NRCS, 2009). The tabular method is used to determine peak flows and hydrographs within a watershed. However, its accuracy decreases as the complexity of the watershed increases. NRCS recommends that the computer program TR-20 be used instead of the tabular method if any of the following conditions apply:
- Tt is greater than 3 hours.
- Tc is greater than 2 hours.
- Drainage areas of individual subareas differ by a factor of 5 or more.
- The entire flood hydrograph is needed for flood routing.
- The time to peak discharge needs to be more accurate than that obtained by the tabular method.
WinTR-55 is a single-event, rainfall-runoff, small-watershed hydrologic model based on the TR-55 methodology. The WinTR-55 program can generate and plot hydrographs, compute peak discharges, and perform detention pond storage estimates. It can account for hydrograph shift and attenuation due to reach routing. WinTR-55 has limitations that assume a less complex watershed (e.g., 10 subwatersheds or less, 25-square-mile drainage area maximum, trapezoidal-shaped channel, and 2-point stage-storage curve for a reservoir). Refer to the TR-55 manual and WinTR-55 User Guide for additional limitations.
104.2.3.5.2 WinTR-20
The WinTR-20 computer program, developed by the NRCS, computes flood hydrographs from runoff and routes the flow through stream channels and reservoirs; WinTR-20 is preferred over TR-55 and the DOS version of TR-20.
In the WinTR-20 program, routed hydrographs are combined with those of tributaries. The program provides procedures for hydrograph separation by branching or diversion of flow and for adding baseflow. Peak discharges, their times of occurrence, WSEs and duration of flows can be computed at any desired cross section or structure. Complete discharge hydrographs as well as discharge hydrograph elevations can be obtained if requested. The program provides for the analysis of up to nine different rainstorm distributions over a watershed under various combinations of land treatment, floodwater retarding structures, diversions, and channel modifications. Such analyses can be performed on as many as 200 subwatersheds or reaches and 99 structures in any one continuous run.
104.2.3.5.3 HEC-HMS and HEC-1
HEC-HMS is a program that is a generalized modeling system capable of representing many different watersheds. It is designed to simulate the precipitation-runoff processes of dendritic watershed systems and is applicable in a wide range of geographic areas for solving the widest- possible range of problems. HEC-HMS, like its predecessor, HEC-1, is extremely flexible in that its hydrologic elements include subbasins, reaches, junctions, reservoirs, and diversions. Hydrograph computations should be performed using the Delmarva Unit Hydrograph (UH) for all projects south of the Chesapeake and Delaware Canal, and the NRCS standard UH or Snyder UH north of the Chesapeake and Delaware Canal. User-specified s-graphs and UHs are allowed if more specific data are available. The Snyder method allows for variable peak flow rate factor, and calibrating the Snyder method makes it more versatile. HEC-HMS is preferred over HEC-1, which is Fortran based; HEC-HMS is Windows-compatible with a graphical user interface.
104.2.3.5.4 GIS Preprocessing Models
There are various GIS software packages and/or extensions to GIS software that allow preprocessing of digital terrain (DEMs, triangulated irregular networks [TINs]), land use, and soil data to develop the parameters (time-of-concentration values, rational “C” values, NRCS CN, etc.) required for hydrologic methods or models. These packages can save valuable time and provide accurate data and parameters that can be modified for various scenarios. Two such packages are discussed below.