Hydrology

Watershed Delineation & Runoff Analysis

Watershed Delineation and Runoff Analysis (Using HydroCAD)

Picture Source: Kim Roberts, Center for Watershed Protection

A watershed as defined by NOAA is “a land area that channels rainfall and snowmelt to creeks, streams, rivers, and eventually to outflow points such as reservoirs, bays, and the ocean”. (NOAA 2020) For design purposes, the ultimate downstream point (outfall) from where a watershed (or subwatershed) will start is usually referred to as either a Point Of Analysis (POA) or Point Of Interest (POI), either term is acceptable. This point is chosen by the designer and confirmed by the designer’s manager. The watershed area upstream and draining (runoff) to this point can best be imagined as a funnel. As rainfall falls onto the funnel’s rim, it either flows down inside to the eventual POA or down the outside and away from this area. To help accomplish the watershed delineation, a topographic map and in most cases along with a detailed survey of the area should be used.

The following is a series of steps on how to define the watershed.

Use a topographic map (plus survey information if available) of the project’s area and surrounding area to first determine where the POA will be located.
While still imagining a funnel feature, start marking out the rim / ridge features. This will approximate the outside boundary of the watershed where the runoff will go inside the funnel.
Next, determine the direction of drainage around the boundary in the watershed area by drawing arrows perpendicular to a series of contour lines that decrease in elevation. Stormwater runoff seeks the path of least resistance as it travels downslope. The path is the shortest distance between contours; hence, a perpendicular route.
The final boundary line will not be a straight line from ridge to ridge (in natural areas), but usually a curving line inside and outside of the original boundary line drawn between ridges. In unnatural areas, i.e. a roadway with curbing on both sides could potentially have a more straight-line boundary area.
To test the watershed boundaries on the map, imagine a drop of rain falling on the surface of the map. Imagine the runoff flowing down the slopes as it crosses contour lines at right angles. Follow its path to the POA being studied. Imagine this water drop starting at different points on the watershed boundary to verify that the boundary is correct.
Once the watershed boundary has been delineated on the map, it is best to verify the map drawing with a field visit. This should help in identifying any areas of concern and to confirm the boundary limit set. Note: at times there could be pipes transporting runoff from outside of the boundary into the watershed area (and vice versa) as well as swales sloping in opposite directions of the contours; hence, why a field visit is often necessary for confirmation. Revise the boundary as necessary and further research any contributing drainage runs.

Once the watershed area has been delineated, it’s time to determine the amount of runoff draining to the POA. There are various methodologies to achieve this result and the ones used on DelDOT projects will be briefly discussed herein.

Rational Method. This is used for drainage design and is mostly used by the Project Development sections. This is discussed in greater detail in the DelDOT Road Design Manual (Chapter 6) as well as FHWA Hydraulic Engineering Circular No. 22 (HEC 22).
Regression Equations. Used almost exclusively by the Bridge section. Regression equations are commonly used for estimating peak flows of larger drainage areas. The United States Geological Survey (USGS) has developed and compiled regional regression equations which are included in a computer program called StreamStats. StreamStats is a Web-based Geographic Information Systems (GIS) application that provides estimates of various streamflow statistics for user-selected sites. More information can be obtained about the regression equations and StreamStats in Section 104 of the DelDOT Bridge Design Manual.
NRCS Methods. This is mainly used for stormwater management purposes and calculates peak flow as a function of the drainage area, watershed storage (retention/detention), and the time of concentration. It is also used to produce runoff hydrographs to evaluate flow routing and flow peaking conditions in individual segments of complex drainage systems. This method will be the emphasis for the rest of this document.

It is presumed that the designer is familiar with the basic theory and methods of analysis and design in both hydrology and hydraulics. Also, the following information will rely heavily on the use of the HydroCAD program, and it is suggested that the designer be familiar with this program to better utilize this document. One of the biggest questions that always arises (after POA placement) is when to break down a watershed into subwatersheds. Ideally, a watershed should be a mostly homogeneous drainage area, i.e. the land cover / CN is mostly the same throughout. There are no hard and fast rules that apply on when exactly to break a bigger watershed into smaller pieces, but here are some helpful thoughts obtained from Dr. Carmine Balascio, Associate Professor of Water Resource Engineering at the University of Delaware:

“Weighted curve numbers are only representative of the total watershed characteristics if the different land uses are spread homogeneously throughout. In a situation when you have two very different land uses, impervious roadway and forest, both going to the same outfall, it would be best to represent them as two different subcatchments. Any time you have very different land uses (and, importantly, very different associated curve numbers), it’s best to split into different subcatchments – especially if the different land uses are highly segregated from one another. Also, if there is a complex drainage pattern with more than one main drainage pathway to the outfall, the watershed should be broken up into subcatchments associated with those different primary drainage pathways. The only exception is if times of concentration measured from the different subcatchments all the way to the common outfall are roughly equal. Curve numbers for the different subcatchments should also not differ greatly, of course. It is probably pretty safe when composite CNs for different subareas differ by 15 to 20% or less, but in instances where the differences in CN are much larger than that, further investigation would be warranted. If using a program such as HydroCAD or WIN-TR55, it’s pretty easy to model the watershed both ways and see if there’s much difference. The type of difference seen is that the model broken up into multiple subcatchments is almost certainly going to predict higher peak flow rates. Depending on how different the subcatchments are, the difference can be substantial. To be conservative, use the multiple subcatchment model if there’s much difference in peak flows compared to the model with a single catchment. With the ease of use with the computer models of today, there’s no reason not to use multiple subcatchments whenever it even looks like there could be an issue.” (Balascio, 2020)

For the example below, it will start with two subwatersheds that combine into a bigger watershed that leads to the eventual POA. A rough sketch of the areas in question are shown here:

Example Problem

For the existing condition shown above find the runoff value at POA 1 utilizing the HydroCAD program for the 10yr storm event. This project is in Kent County.

The first thing to do in HydroCAD is to setup the overall parameters in the ‘Calculation Settings’. The only two tabs that would need modification at this time are the ‘Rainfall’ and ‘Unit Hydro’. Under the ‘Rainfall’ tab, change the ‘Storm Type’ to ‘NOAA 24-hr’, the ‘Storm Curve’ to ‘C’ and under ‘Rainfall Event’, click the ‘Import Events From’, then click the ‘Lookup Table’. A new window will pop up, then scroll down to highlight the row of ‘DE Kent…’, then click ‘OK’. A new window will pop up asking which storm distribution to select. Select ‘NOAA_C’, then click ‘OK’. Screen shots are shown below.

Now go to the ‘Unit Hydro’ tab and change the ‘Unit Hydrograph’ to ‘Delmarva’, then click ‘OK’. The ‘Calculation Settings’ are now complete.

Setup the above rough sketch into a HydroCAD routing diagram. Refer to the HydroCAD documentation on how to change the node appearance i.e. name and number.

Input the parameters needed for each drainage area. HydroCAD refers to these as subcatchments. The information needed will be the ‘Drainage Area’ + ‘CN’ under the ‘Area’ tab and the ‘Tc’ time under the ‘Tc’ tab.

The Drainage Area is computed from survey and topographic data that was explained above in watershed delineation. CAD programs can compute this area rather quickly after delineation. Refer to the CAD computer program documentation on how to calculate the area.
The CN (curve number) indicates the runoff potential of a given area as determined by a soil type and cover condition. Refer to TR 55 documentation on the procedure for a CN determination. Sometimes this will be referred to as RCN (runoff curve number).
Tc (time of concentration) represents the time it takes for water to travel from one location to another within a drainage area. The Tc is the sum of all travel times for consecutive segments of the drainage conveyance system and shall start from the hydraulically most distant point. This could take multiple iterations to find the hydraulically most distant point. Again, this procedure can be found in the TR 55 documentation.

For this example, use the following given information in the table below that was already calculated and input these parameters into each subcatchment. Note: this is information that the designer will have to figure out on their own for each individual watershed.

	Area (ac)	CN	Tc (min)
Subwatershed 1	2	75	12
Subwatershed 2	3	81	15

After the information is entered, double click on each subcatchment to view the hydrograph for each individual subcatchment.

For the culvert where both drainage areas enter, the below information was obtained from field survey data. Input these parameters into the reach node.

21” Reinforced Concrete Pipe (RCP)
48’ total length
Entrance invert = 16’
Exit Invert = 15.5’
Free discharge, i.e. no tailwater conditions

After the above information is entered, double click on the ‘Culvert’ reach node to view the hydrograph and the ‘Outflow’ for POA 1, which is the answer needed (10.23 cfs).

This is the same basic process for comparing an existing / predeveloped flow versus a post-developed (after project completed) flow comparison. The input parameters will most likely change from pre to post, which is what gives the outflow different values.

For the ‘Calculation Settings’ aspect, depending on where the project is located in the state, the rainfall and unit hydrograph information will change. Below, is a table to help determine which input parameters should be used.

Location (by county)*	Rainfall Storm Curve	Unit Hydrograph
New Castle - above the C&D Canal	C	SCS
New Castle - below the C&D Canal	C	Delmarva
Kent	C	Delmarva
Sussex	D	Delmarva

*The rainfall storm type is the same for all, “NOAA 24-hr”

For any questions concerning this guidance or HydroCAD, please consult with the Stormwater section.

Cv & Fv Compliance

Compliance Flowchart

Compliance Explanation

DNREC / DelDOT Forum for Cv and Fv Compliance

(August 22, 2019)

In accordance with the below excerpts from the DSSR here is the clarification for 5.3.3.5 and 5.4.3.5:

The biggest factor for the language is that it states, ‘discharges directly to a natural stream, waterbody, or tax ditch’. If conveyance is required to reach said stream or water body, then that is not a direct discharge as well as proving that you have adequate conveyance to that point also does not count. Adequate conveyance to a direct discharge is only applicable for a tidal outfall. If outflowing directly into a tax ditch, you need to get approval from the DNREC Drainage Program first, especially if you are increasing any flows and/or changing the land use within that tax ditch boundary or modifying the tax ditch watershed in any way. Any discharge that goes into a closed drainage system does not qualify.

5101 Sediment and Stormwater Regulations

5.0 Performance Criteria for Post Construction Stormwater Management

5.3 Conveyance Event Criteria

5.3.3 Compliance with subsection 5.3 shall be accomplished through the following provisions:

5.3.3.1 The Cv shall be managed using BMPs as set forth in Section 11.0 such that there is no adverse impact by limiting the increase in the downstream post-developed water surface elevation to no more than 0.05 feet; or

5.3.3.2 Improving the existing downstream conveyance system to the point where the downstream condition meets the "no adverse impact" criteria of subsection 5.3.3.1; but no farther than the point where the LOD is less than 10% of the contributing drainage area; or

5.3.3.3 Provisions will be made or exist for a non-erosive conveyance system to tidal waters by either a closed drainage system or by open channel flow that has adequate conveyance for the Cv; or

5.3.3.4 Demonstration that the location of a project within a watershed would aggravate flooding or channel erosion by the imposition of peak control requirements, as evidenced by a downstream analysis that shows the inflection point of the site hydrograph occurs prior to and is less than the peak of the upstream hydrograph; or

5.3.3.5 The site LOD comprises 10% or less of the total upstream contributing drainage area at the point of discharge for sites that discharge directly to a natural stream, waterbody, or tax ditch; or

5.3.3.6 The proposed project will generate only a de minimis discharge and will have no adverse impact on the receiving wetland, watercourse or downstream property as determined on a case-by-case basis.

5.4 Flooding Event Criteria

5.4.3 Compliance with subsection 5.4 shall be accomplished through the following provisions:

5.4.3.1 The Fv shall be managed using BMPs as set forth in Section 11.0 such that there is no adverse impact by limiting the increase in the downstream post-developed water surface elevation to no more than 0.05 feet; or

5.4.3.2 Improving the existing downstream conveyance system so that the downstream condition meets the "no adverse impact" criteria of subsection 5.4.3.1; or

5.4.3.3 Provisions will be made or exist for a non-erosive conveyance system to tidal waters by either a closed drainage system or by open channel flow that has adequate conveyance for the Fv; or

5.4.3.4 Demonstration that the location of a project within a watershed would aggravate downstream flooding or channel erosion by the imposition of peak control requirements, as evidenced by a downstream analysis that shows the inflection point of the site hydrograph occurs prior to the peak of the upstream hydrograph; or

5.4.3.5 The site LOD comprises 10% or less of the total upstream contributing drainage area at the point of discharge for sites that discharge directly to a natural stream, waterbody, or tax ditch; or

5.4.3.6 The proposed project will generate only a de minimis discharge and will have no adverse impact on the receiving wetland, watercourse, or downstream property as determined on a case-by-case basis.

Water Quality Stormwater Management Approach for Infield Areas

The infield area would be defined as contained within an access road / ramp or an area contained within bordering roads.

Below are four different potential scenarios with a SWM approach for each one that is acceptable to DNREC in achieving water quality (RPv) compliance as based on the LOD runoff: (NOTE: jurisdictional wetlands for these matters are defined as any areas marked on the plans within wetland delineation lines.)

Non-jurisdictional wetland with no outlet

This is considered as a retention/infiltration basin, which in turn would account for 100% RPv credit.
The size of the basin is already set for a project such as this, but a check should be done of the current geometry to help insure that the ponding depth should be spread over as large an area as possible. The maximum ponding depth should be 24”. It may be necessary to do some grading within the infield area to help insure that ponding is not occurring in one small section. And depending on the geometrics of the site, drainage/flooding considerations should be checked to help insure that the surrounding roadway will not be overtopped in a 100-yr rain (Fv / 1% storm) event.

Jurisdictional wetland with no outlet
1. The normal pool elevation has to be maintained, meaning that the RPv volume cannot inundate the wetland above the normal pool elevation for more than 48 hours.
Non-jurisdictional wetland with an outlet

Depending on the particular site configuration or how the site can be designed and constructed, 100% RPv reduction can be achieved. This is done by providing storage for the RPv volume below the invert of the outlet.
If a site cannot be constructed to this requirement and the site could be defined as a wet pond, then a RPv reduction can be achieved as per current DNREC regulations concerning extended detention for wet ponds.

Jurisdictional wetland with an outlet

The RPv volume must be detained the same as the extended detention wet pond criteria, but also has to be fully released within 48 hours, so as to not impact the surrounding wetland vegetation.

For all of the above options, 100% of the inflow must have pretreatment and can consist of one of the below options or a combination thereof. Each of these options is referenced in the DNREC Post Construction Stormwater BMP Standards and Specifications.

Vegetated channel
Grass filter strip (Sheet Flow to Open Space)
Forebay (minimum size of 10% of RPv volume)
Sand filter (Stormwater Filtering Systems)
Other practice as approved by DelDOT (Proprietary Practices)

If the DURMM analysis shows the runoff reduction requirement is more than 1”, then the option of using the alternative methodology of compliance based on 1” of runoff can be used. If including OLOD areas in the total runoff, they can be analyzed as separate subareas in HydroCAD and sum them with the LOD area to account for the total contributing runoff to the BMP.