See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228973275

Two-stage channel design procedures

Article · January 2004

CITATIONS READS
4 1,796

5 authors, including:

Andy D Ward Dan Mecklenburg

The Ohio State University The Ohio State University
118 PUBLICATIONS   2,225 CITATIONS    26 PUBLICATIONS   320 CITATIONS   

SEE PROFILE SEE PROFILE

Larry C Brown Anand D. Jayakaran

The Ohio State University Washington State University
104 PUBLICATIONS   2,002 CITATIONS    44 PUBLICATIONS   376 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Hydrological river engineering View project

Transforming Drainage Project (USDA-NIFA funded 2015-2019) View project

All content following this page was uploaded by Dan Mecklenburg on 31 May 2014.

The user has requested enhancement of the downloaded file.

 TWO-STAGE CHANNEL DESIGN PROCEDURES**

A. Ward1, D. Mecklenburg2, G. E. Powell3, L Brown4, A. Jayakaran3

ABSTRACT

Two-stage channel systems incorporate benches that function as floodplains. However, in

comparison to floodplains in natural streams, the width of the benches is often small due to the
confining geometry of the constructed channel. This manuscript outlines measurement and
analysis procedures that can used to size two-stage channel systems that are more self-sustaining
than conventional one-stage constructed channels. The analysis procedures are incorporated in
the STREAMS spreadsheet tools that have been developed by the Ohio Department of Natural
Resources and the Ohio State University. Channel design dimensions are determined by
measuring the effective discharge features at the project site and making regional curve
measurements. Bench elevations are dependent on correctly determining the size of the effective
discharge channel. The minimum total widths of the constructed benches are at least double the
width of the effective discharge channel and the overall conveyance capacities of these two-stage
systems can be sized based on the probability of out-of-ditch flooding into adjacent areas.
Construction of a two-stage channel system requires a significant capital investment to create a
wider surface width. However, it is anticipated that two-stage systems will have improved
conveyance capacity, will be more self-sustaining, will create and maintain better habitat, and
will improve water quality.

KEYWORDS. Drainage Ditch, Design Procedures, Ditch Stability, Fluvial Processes

ACKNOWLEDGEMENTS
Knowledge on fluvial processes in ditches and the development of a two-stage ditch design
approach were initiated through a grant from the Great Lakes Protection Fund. On-going
research to evaluate the water quality benefits of ditches is funded by a grant from the joint
USDA and EPA Nutrient Science for Improved Watershed Management Program.

**
This paper will be published in the Conference Proceedings for the ASAE Specialty Conference “Self-Sustaining
Solutions for Streams, Wetlands, and Watersheds”, September 12-15, 2004, St. Paul, Minnesota
1
Professor, Department of Food, Agricultural and Biological Engineering, The Ohio State University, 590 Woody
Hayes Drive, Columbus, OH 43210. Phone: 614-292-9354, Email: ward.2@osu.edu.
2
Ecological Engineer, Ohio Department of Natural Resources, 4383 Fountain Square Drive, Columbus, OH 43224.
Phone: 614-265-6639, Email: dan.mecklenburg@dnr.state.oh.us.
3
Graduate Associate, Department of Food, Agricultural and Biological Engineering, The Ohio State University, 590
Woody Hayes Drive, Columbus, OH 43210. Email: powell.354@osu.edu; jayakaran.1@osu.edu.
4
Professor, Department of Food, Agricultural and Biological Engineering, The Ohio State University, 590 Woody
Hayes Drive, Columbus, OH 43210. Phone: 614-292-3826, Email: brown.59@osu.edu.
 INTRODUCTION
The Problem

Highly modified channels drain extensive portions of productive agricultural land in the U.S.A.
In some areas virtually all of the natural channels have been deepened and straightened to
facilitate the flow of water from agricultural subsurface drainage outlets and to increase
conveyance. Work is done periodically to maintain the drainage function, typically including
removal of woody vegetation, removal of weeds, and “dipping out” deposited sediment.
Ancillary work includes stabilizing bank slope failures and toe scour.

Ditch maintenance typically restores the ditch to a trapezoidal shape designed to transport large
storm events (Figure 1). To facilitate drainage, and reduce the frequency of over bank flows,
trapezoidal ditches are designed to accommodate large flows (5 to 100+ year recurrence interval)
within the ditch. Also the width of the ditch bottom is constructed wider than the channel bottom
that would form by fluvial processes, thus making the effective discharge relatively wide and
shallow. Therefore, the constructed ditch channel is often oversized for small flows and provides
no floodplain for large flows. Despite the tradition of ditch maintenance and current majority
mind set, ditch dimensions are a result of fluvial (flowing water) processes and current practices
are antitheses to the natural processes.

Nature’s Solution

In contrast to trapezoidal agricultural drainage ditches, an integral part of natural stream channels
are the floodplains. The floodplains of high quality streams (except for those with steep bed
slopes) are characterized by frequent, extensive over-bank flow. In dynamic-equilibrium, a
stream system depends on both the ability of the floodplain to dissipate the energy of high flows
and concentrating the energy of low flows to effectively create a balance in sediment transport,
storage and supply. In natural streams, fluvial processes work to size and maintain the
dimensions of the main channel based on the effective discharge (Ward and Trimble, 2003).

The term effective discharge is based on concepts proposed by Wolman and Miller (1960) and is
the streamflow that transports the most sediment over the long term. This approach for
determining geomorphic work is known as the Wolman-Miller model (Andrews and Nankervis,
1995). The sediment transport rate is a function of the discharge (curve A in Figure 2). So, low
discharges are ineffective in transporting sediment and extreme events have very high sediment
transport rates (curve B). The extreme events however, occur infrequently so the total sediment
load they carry over a period of many decades is not large. When the event frequency (curve B)
is multiplied by the transport rate for that frequency we obtain curve C, which is a measure of the
total sediment load carried for that particular discharge and shows that the bulk of the sediment is
transported at near-effective flow.

In response to the construction of an oversized trapezoidal channel, fluvial processes work to

create a small effective discharge channel by building a floodplain or bench within the confines
of the ditch. If conditions allow, these benches can reach a stable size, thickly vegetated with
mostly grasses. The small effective discharge channel will often meander slightly within the
ditch. The effective discharge channel will usually have steep (1:1) sides and a bed consisting of
material courser than that of adjacent reaches where benches have not formed. Details on fluvial
processes in ditches are provided by Landwehr and Rhoads (2003) and Ward et al. (2002b).

DESIGN OBJECTIVE
This paper outlines a design strategy to construct a channel system (see Figure 3) that has a
compound cross section consisting of: (1) a channel that is sized to convey the effective
discharge, (2) a bench to serve as a floodplain for the smaller effective discharge channel, and (3)
a stage of adequate width to prevent flow overtopping the ditch banks and flooding surrounding
land. The design procedure will create a channel system that based on our qualitative experience
will then maintain the compound form. Currently, there is limited quantifiable scientific
evidence or theory on the size of bench necessary to establish a self-sustaining stable condition.
Research is being conducted to better determine the two-stage geometry where the benches
neither aggrade nor degrade and the average dimensions of the effective discharge channel do
not vary over time.

1
 DESIGN APPROACH
Design and construction of two-stage channels is different than that of traditional trapezoidal
channels. The following additional design procedures are proposed by the authors and have been
used for design in several applications in Ohio and Michigan.

The width of the candidate ditch or channel system is the key design characteristic. It will
determine the success in achieving the drainage and ecological benefits. The dimensions of the
effective discharge channel dictate the two-stage channel design. If properly sized, the effective
discharge channel will be maintained by fluvial processes and will eliminate large-scale channel
maintenance. The first step in developing a two-stage design is determining the probable
dimensions of the effective discharge channel. The effective discharge channel is sized by
detailed on-site investigation and by creating a watershed specific regional curve for the project.

On-Site Channel Geomorphology Study

Where a modification will be made to an existing channel detailed measurements should be

made of the channel profile and the dimensions of the effective discharge channel and benches
that have formed within the channel system. The measurement methods should be consistent
with the guidelines that are presented by Harrelson et al. (1994). Conducting an on-site
geomorphology study is a simple and reliable method, but is only adequate under the following
conditions: (1) if an effective discharge channel and benches have formed and (2) if the project
length is short such that the drainage area is relatively constant. Often, the effective discharge
channel will be over-wide and the benches will be intermittent and sloping.

A detailed survey along the reach of interest or a reference reach consists of measuring the
profile, pattern, and dimension of the channel. The profile is the slope of the bed surface
including all pools, riffles, and runs. The undulating elevations of the channel bed leads to
questions of the true channel profile, to compensate for the bed slope variability, the water
surface is also measured to represent the slope of the channel. The slope of the ditch will assist
in predicting the flow rate of the ditch and conveyance capacity.

The pattern of a reach measures the sinuosity of the fluvial channel. This is obtained using a
compass and measuring the azimuth angle from magnetic north. Most ditches do not have
significantly meandering channels, however if a ditch is reaching the 5:1 ditch width ratio the
ditch pattern is useful to see how the channel is meandering.

The dimensions of the channel are obtained by surveying cross sections, either at increments
along the reach or representative cross sections. A laser level and survey tape, or a total station,
are often used. The distance from the left channel bank and the change in elevation are measured
for each grade break across the channel cross-section. A pebble count should be performed to
estimate the mean bed particle size.

Regional Curve

In addition to on-site geomorphology studies, regional curves are developed to relate the
effective discharge channel dimensions to drainage area for a range of larger and smaller
drainage areas. Traditional regional curves are created by performing numerous profile and
cross section surveys at locations with different drainage areas, which often include USGS
stream gage sites. Using stream gage sites allow use of the recorded peak flow data for
recurrence interval determination. Typically, these measurements are only made in natural
streams at locations with well-defined channel dimensions. The regional curves are then
developed by plotting the measured dimensions against their respective drainage area and then
fitting regression lines through the plots. The measured data can be entered into one of the
STREAM (Spreadsheet Tools for River Evaluation, Assessment and Monitoring) spreadsheets
that have been developed by the Ohio Department of Natural Resources (Mecklenburg and
Ward, this proceedings). The STREAM tools can then be used to develop the regional curves
and to determine the dimensions of the effective discharge channel associated with these curves.

For two-stage channel design an abbreviated, “rapid regional curve” is typically adequate. The
method consists of finding ditches or streams with well-developed benches/floodplains and

2
measuring at least the width and depth of the naturally formed effective discharge channel. Sites
for any regional curve should also have similar characteristics to the project site. Several
measurements should be taken for each range of drainage areas to verify that the measured
feature is consistent with those across the watershed. Whenever possible precision surveying
instruments should be used to make the elevation and distance measurements. However, it has
been the authors’ experience that useful results can be obtained by just using a 100 ft tape and a
telescoping surveying rod. Typical results for the approximate approach are shown in Figure 6.
Using this regression equation, the drainage area for the project site can be used to determine the
dimensions of the effective discharge channel from the regional curve.

Potential sites are marked on a state Gazetteer or regional map that shows all of the ditch and
stream systems. Sites are selected to provide data for several size drainage areas within each log
cycle. More sites than might be needed are identified as some sites will probably not be used
because they lack adequate features or access to the site is not granted by the property owner. If
an established USGS gage is located down stream of the site, the regional curve analysis should
include data for the gage. Gage and streamflow data for USGS gages can be found at:
http://waterdata.usgs.gov/nwis.

Approximate drainage areas to each site are outlined in the Gazetteer by using the interfluve
method (Ward and Trimble, 2003). Once the field work has been completed the drainage areas
for the selected sites are determined using GIS tools data or by marking the drainage area
boundaries on a topographic map and using a planimeter.

Flooded Width

The flooded width is defined as the total width across the ditch at the stage elevation where
benches have formed and/or are anticipated to form (Figure 3). The two-stage width ratio is
defined as the flooded width divided by the top width of the effective discharge channel.
Typically, the ditch width ratio should satisfy or exceed 3:1 depending on project specific
feasibility and project objectives.

In low-gradient natural stream system, typical of intensive agricultural drainage areas, a

streamway that is at least ten times the top width of a effective discharge channel sized by the
effective discharge will usually encompass meander migrations that occur over time (Ward et al.,
2002a). Based on bedload transport models, floodplains these sizes are very effective in
dissipating the energy of high flows. Based on visual observations and bedload transport models,
two other rules of thumb have been established for ditch width ratios. If the total width, when out
of channel flow is initiated, is less than three times the top width of the effective discharge
channel the benches might not fully develop, the benches are more likely to be unstable, and
shear stresses on the bed and banks of the ditch will be high during large events. If the total
width, when out of channel flow is initiated, is more than five times the top width of the effective
discharge channel the channel will begin to exhibit a natural meander pattern that at places in the
ditch is likely to cut into the banks of the ditch. Therefore, when out of channel flow is initiated
the total width should be between three to five channel widths if the objective is to only provide
adequate conveyance capacity and a more self-sustaining system while maintaining relatively
straight ditch geometry.

Discharge and Recurrence Interval Analysis

Ditch conveyance is based on the recurrence interval storm event that the ditch can transport.
Stream surveys at locations where there are long term records of streamflows that allows
determination of discharge associated with channel dimensions. However, in most parts of the
nations there are limited numbers of small watershed gages and typically these gages have short
records or have been discontinued. Therefore, measured streamflows at locations without gauges
are replaced by discharges calculated by using a resistance equation such as Manning’s equation
and corresponding recurrence intervals from hydrologic models.

The spreadsheet procedure used in this analysis also allows the user to input their own discharge
versus recurrence interval relationships. Commonly used methods for estimating discharge for
different recurrence intervals for ungaged streams in Ohio are: (1) the USGS Urban method for
Ohio (Sherwood, 1994); (2) calibrating the urban method based on discharge relations at the
USGS gage; and (3) the USGS method for rural watersheds in Ohio (Koltun and Roberts, 1990).

3
Similar methods exist for most states in the U.S.A. These equations are contained in the National
Flood Frequency (NFF) Program, which is a Windows program for estimating the magnitude and
frequency of peak discharges for unregulated rural and urban watersheds
(http://water.usgs.gov/software/).

STREAM Spreadsheet Tools

Details on the STREAM tools are presented by Mecklenburg and Ward and the tools can be
obtained by contacting Dan Mecklenburg (dan.mecklenburg@dnr.state.oh.us). Channel and
ditch survey data can be entered into either the two-stage channel design module or the reference
reach module of the STREAM tools. The two-stage channel module aids in designing a new
fluvial channel together with various size benches based on: (1) a regional curve for the area; (2)
cross-section data for a ditch or channel reach; (3) profile data for the reach that can include bed,
water elevation, bench, and top of ditch data; (4) the d50 of the bed material; and (5) user defined
adjustments to the channel, bench, and ditch geometry. The spreadsheet allows rapid
comparison of hydraulic characteristics of different channel configurations.

The channel width, depth and cross-sectional area associated with the effective discharge at each
location surveyed are entered into the spreadsheet to develop a regional curve. The calculated
drainage area and effective discharge at each location are also entered into the same spreadsheet.
A log-log plot is then made of each effective discharge dimension versus drainage area. A least
squares analysis is then used to fit a power regression line (a trend line) through each set of data
and calculate the coefficients of the regional curve.

In the spreadsheet, stage-discharge relationships for each site are obtained based on Manning’s
equation. A separate Manning’s n value is used for the effective discharge channel and the
vegetated benches and banks of the ditch. The roughness of the bed, banks, and benches vary
seasonally based on winter conditions, vegetation growth, maintenance, and scour or deposition
on these features. Discharge versus recurrence interval data are used together with Manning’s
equation to calculate the flow stage associated with each recurrence interval.

In the spreadsheet, bedload transport in the fluvial channel is calculated based on the probable
discharges that will occur during a 100 year time period, the d50 of the bed material, and based on
estimates obtained from the Meyer-Peter-Muller bedload equation (Ward and Trimble, 2003).
Estimates are obtained of the bedload, recurrence interval of the effective discharge, and the
probable stage of the effective discharge. Bedload transport methods are notoriously inexact so
the provided information should only be used to provide an indicator of the relative benefits of
different ditch geometries.

DISCUSSION

The design procedures outlined earlier have been used for several applications in Ohio and
Michigan. While the procedures were initially developed for application in rural settings they are
starting to see application as a storm water management strategy in urbanizing watersheds.

Typical, results for a ditch in Hillsdale County, Michigan are illustrated in Figure 5 and 6. Figure
5 displays a plot from a profile survey that includes the bed elevations, water surface, bench
features, and top of the ditch banks. Figure 6a illustrates a surveyed cross section overlaid with a
two-stage ditch designed. There was a grade break but weak bench formation at the elevation
that it was determine that a bench would naturally form. The existing main channel has a similar
geometry to the projected geometry. Figure 6b, shows construction of the two-stage geometry.
While the channel dimensions are the basis of the design, the channel itself will not usually need
to be constructed. Construction consists of excavating down to the bench elevation and out to
the desired bench width. The design ditch width can be created by excavating a portion of the
width on each side of the channel or entirely to one side of the channel. Idealistically, the
vegetation at the bench elevation and the channel should be left intact to help stabilize the
effective discharge channel during the construction period (see Figure 6b). The effective
discharge channel geometry is not constructed because over time fluvial processes will shape the
channel dimensions.

In the approach we have proposed only the second stage (geometry between the benches and the
top of the ditch) might be designed based on the recurrence interval of the discharge it can

4
convey. Attempting to size the effective discharge channel based on a recurrence interval will
probably result in an oversized channel. Many studies have been made on bank forming
discharges and reference should be made to Simon et al. (2004). They determined that for most
ecoregions in the United States the median recurrence interval of the effective discharge, based
on suspended sediment data, ranged from 1.1 to 1.7 years. However, in wetland streams, with
predominantly sand and silt bed material, this interval may be much smaller (Jurmu and Andrle,
1997). Our experience with low gradient ditches in the Midwest region of the United States
suggest that the effective discharge is associated with high flows from subsurface drainage
systems and that these flows occur many times annually.

Apparent benefits of a two-stage ditch over a conventional ditch are improved drainage function
and ecological function. Drainage benefits may include increased ditch stability, increased
capacity and reduced maintenance. Evidence and theory both suggest that ditches prone to
filling with accumulated sediment may require less frequent “dipping out” if constructed in a
two-stage form. Second, channel stability may be improved by a reduction in the erosive
potential of larger flows, as they are shallower and spread out across the bench. Stability of the
ditch bank should be improved where the toe of the ditch bank meets the bench rather than the
ditch bottom. Here the water depth is effectively reduced and the shear stress (erosive force) on
the toe of the bank is less. The bank material will be dryer because it is not in contact under low
flow conditions. There is a danger of continued bank erosion where the fluvial channel
meanders to the toe of the ditch bank. Additional research on the bench width required for
stability will eliminate this threat.

A two-stage channel system has the potential to create and maintain better habitat. The narrow
deep fluvial channel provides better water depth during periods of low flow. Grass on the
benches can provide quality in-stream cover and shade. The substrate in the fluvial channel is
improved as the two-stage form increases sediment conveyance and sorting, with fines deposited
on the benches and courser material forming the bed. Two-stage ditches might also be useful in
improving water quality particularly for nutrient assimilation (Ward et al., 2004). Work has been
initiated on the ecology of these ditches and the role of the channel and benches in improving
water quality and habitat.

The primary costs of two-stage ditches are associated with the increased ditch width required.
Wider ditch top width requires surrendering of surrounding agricultural production land. The
increased width requires additional initial earthwork. Costs for construction increase with both
watershed size and ditch depth and might range from $5 to $20/linear foot.

To offset landowner costs the potential for including the bench width in buffer programs is being
explored. Buffers have typically been measured from the top of the ditch. Alternatively
measuring from the top of the small channel thus including the bench and the main side slope of
the ditch is preferable from a water-quality perspective and profitability perspective.

REFERENCES

Harrelson, C.C., C.L. Rawlins, and J.P. Potyondy. 1994. Stream Channel Reference Sites: An
Illustrated Guide to Field Technique. USDA Forest Service, General Technical Report RM-
245, Fort Collins, Colorado, 62 pp.
Jurmu, M.C. and Andrle, R. 1997. Morphology of a Wetland Stream. Environmental
Management 21, No. 6, pp. 921-941.
Koltun, G.F. and J.W.Roberts. 1990. Techniques for Estimating Flood-Peak Discharges of Rural,
Unregulated Streams in Ohio. U.S. Geological Survey, Water Resources Investigations
Report 89-4126, USGS Denver, CO.
Landwehr, K. and B. Rhoads. 2003. Depositional Response of a Headwater Stream to
Channelization, East Central Illinois, USA. River Research and Applications, 19:77-100.
Simon A, W. Dickerson , and A. Heins. 2004. Suspended-sediment transport rates at the 1.5-year
recurrence interval for ecoregions of the United States: transport conditions at the bankfull
and effective discharge? Geomorphology 58: 243-246.
Sherwood, J.M. 1994. Estimation of Volume-Duration-Frequency Relations of Ungaged Small
Urban Streams in Ohio. AWRA Water Resources Bulletin 30(2):261-269.

5
USGS. 2002. NWISWeb Data for the Nation. U.S. Geological Society, Department of the
Interior. 2002. <http://waterdata.usgs.gov/nwis?>.

Ward, A., D. Mecklenburg, J. Mathews, D. Farver. 2002a. Sizing Stream Setback to Help
Maintain Stream Stability. ASAE paper number 022239. Annual ASAE International
meeting 2002.
Ward, A., D. Mecklenburg, L. Brown. 2002b. Using knowledge of fluvial processes to design
self-maintaining agricultural ditches in The Midwestern Region of the USA. Proceedings of
the International Conference on Environmental Flows for River Systems, Cape Town, South
Africa, March.
Ward, A. and S. Trimble. 2003. Environmental Hydrology 2nd Edition. CRC Press.
Ward A.D., M. T. Moore, V. L. Bouchard, K. Powell, D. Mecklenburg, C. M. Cooper and S.
Smith, Jr. Water quality benefits of grassed fluvial features in drainage ditches. Electronic
Proceedings, The Sixth International Conference on Hydro-science and Engineering
Brisbane, Australia May 30 - June 3.
Wolman, M. G. and Miller, J.P. 1960. Magnitude and frequency of forces in geomorphic
processes, Journal of Geology 68: 54-74.

6
 Figure 1: A Minnesota ditch (left) before maintenance, and (right) after maintenance.

Loramie Creek, Ohio

A: sediment function (x 20) C: geomorphic work B: days

50000 250

Effective Discharge
sediment discharge (tons x 10)

40000 200
geomorphic work

B: Days Occurring C: Geomorphic Work

days of occurrence
30000 150

20000 100
A: Sediment Transport Rate

10000 50

0 0
0 500 1000 1500 2000 2500 3000
discharge rate (cfs)

Figure 2. Illustration (for Loramie Creek, Ohio) of how the effective discharge is estimated as the
flow that carries the largest total sediment load (1 m3/s = 35.3 ft3/s or cfs).

Figure 3: A two-stage channel geometry with minimum sized benches.

Regional Curve A: Loramie Creek Watershed Regional Curve B: St. Joseph River Watershed

100 100
Effective Discharge Width & Depth

Effective Discharge Width & Depth

W = 6.27(DA)0.47 W = 4.67(DA)0.47
R2 = 0.79 R2 = 0.91
10 10
(ft.)

(ft.)

1 1
D = 1.06(DA)0.29 D = 1.35(DA)0.24
R2 = 0.62 R 2 = 0.78

0.1 0.1
0.1 1 10 100 1000 0.1 1 10 100 1000
Drainage Area (sq.mi.) Drainage Area (sq.mi.)

Figure 4: Regional curves showing “poor” (Figure A) and “good” (Figure B) relationships between
the dimensions of the effective discharge channel and drainage area.

7
 Drainage Ditch in Hillsdale County

714

Top of Ditch
712

710
Elevation (ft)

708 Bench or "Bankfull"

Grade Break

706

Channel Bed
704

702

700
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100
Station or Channel Distance (ft)

Figure 5: Profiles of the Bed, Bench, and Top of Ditch in Hillsdale County, Michigan.

712
711
710

Elevation (ft)
709
708
707
706
705
704
703
702
-30 -20 -10 0 10 20 30
Width from River Left to Right (ft)

Figure 6: A (left): Pre (dashed line) and post (solid line) construction geometries.
B (right): Construction of the benches for a ditch in Hillsdale County, Michigan.

View publication stats