MANAGEMENT OF CONVEYANCE
bar


 

 

 

 

The term conveyance refers to the amount of flow a channel can carry with a given energy slope; it represents the frictional controls imposed on discharge rate by channel cross-sectional shape and roughness.

Bed and bank stabilization treatments can modify cross-sectional shape, influence roughness, and otherwise influence channel conveyance. The incorporation of vegetation and LWD into channel and bank stabilization measures frequently results in rougher channel boundaries than more traditional measures. For example, plantings of woody vegetation may provide more flow resistance than riprap revetment. Changes in conveyance properties of a channel can have both engineering and ecological implications. If flooding or upstream drainage is an issue at the site in question, the engineer may have to estimate the impact of proposed measures on flood stages. During low flow periods conveyance properties affect both depth and velocity which have important implications for fish and other stream dwelling organisms (see Special Topic: Physical Aquatic Habitat). In either case, an environmentally-sensitive approach to channel protection implies a thoughtful review of the consequences of any large alteration in existing channel conveyance properties.

UNIFORM FLOW EQUATIONS

The oldest and perhaps simplest approach for computing channel conveyance for steady flow conditions involves the use of a uniform flow equation like the Manning or Chezy equation that contains a coefficient that represents the combined effect of all of the channel characteristics that contribute to flow resistance.  For example, the coefficient n in the Manning equation:

reflects channel bed material, bank conditions, planform, and cross-section shape for an entire reach, while Q is the discharge in ft3/s, A is the cross sectional area of the flow in ft2, R is the hydraulic radius in ft, and S is the energy slope.  R is equal to A/P, where P is the wetted perimeter in ft, and S is equal to the bed slope when flow is uniform.  If SI units are used, the conversion factor in the numerator is simply 1.0 instead of 1.486.

image005
Figure 1.  Definition of terms in uniform flow equations

Two other widely used uniform flow formulas are the Chezy formula, which states that

where C is a flow resistance coefficient, and the Darcy-Weisbach equation

Where f is a flow resistance coefficient and g is the acceleration of gravity.  Both of these formulas are applied using metric (SI) units.  Discharge may be computed by multiplying velocity by the cross sectional area: Q = AV.

Regardless of the uniform flow equation used, conveyance is given by the ratio of discharge, 1 to the square root of the energy slope, S:K = QS-1/2.

Formulas are available for computing flow resistance coefficients (n, C and f ) based on the size of the bed material and the flow depth (Chow 1959, Brownlie 1983).  Tables of values are also available in reference books (e.g., Chow 1959, Henderson, 1966), and photographs of river channels for which measured n-values are available are also published in similar works (Barnes, 1967).  In the absence of experience or data, the engineer may examine these photographs and attempt to match the site in question to one or more of the photographed sites.  Alternatively, the engineer may assign incremental n-values to each of about 8 channel characteristics and sum the incremental values to obtain a composite n for the reach (Cowan, 1956 in Chow, 1959).

n = (n0 + n1 + n2 + n 3 + n4) m

where:
        n0 = base value of n for a straight, uniform, smooth channel in natural materials   
        n1 = correction for the effect of surface irregularities
        n2 = correction for variations in cross section size and shape
        n3 = correction for obstructions
        n4 = correction for vegetation and flow conditions
        m = correction for degree of channel meandering

Appropriate values for these coefficients are provided in Table 1.

TABLE 1:  Suggested values for terms and factors in Cowan equation for Manning's n
image007

Experienced engineers often select n-values based on experience or calibrate their n-values based on stage-discharge curves from nearby gaging stations.  Typically the Manning equation as expressed above is applied to an entire reach only for preliminary estimates or as a first step in design.  More detailed analyses are typically based on computer models that represent the channel as a series of cross sections, for example HEC-2 or HEC-RAS
(http://www.hec.usace.army.mil/software/hec-ras/hecras-hecras.html; Brunner 2001).  Preliminary design computations for individual cross sections may also be performed using the SAM package (http://hlnet.wes.army.mil/software/sam/).

MANNINGS-N VALUES FOR COMPLEX BOUNDARIES

Channel Bends

One-dimensional flow models do not reproduce the complex flow phenomena found in channel meander bends, and the engineer must compensate for this by increasing the Manning n values for reaches that are not straight by some factor, as shown in the Cowan equation above.  There are a variety of semi-empirical techniques for determining an appropriate value for m, the factor to allow for meandering in the literature.  A review is provided by James (1994), who tested eight common techniques using three data sets based on meandering trapezoidal laboratory channels.  One of the best-performing methods was also one of the simplest, the linearized SCS method, which states that

m = 0.43s  + 0.57 for s < 1.7

m = 1.30 for s > 1.7

where m is the meandering correction factor in the Cowan equation above and s is the channel sinuosity, or the ratio of channel thalweg length to straight line distance between the channel endpoints.  This method produced an average absolute error in computed discharge for a given stage of 8%.  This method produces values in close agreement with the suggested values for the Cowan equation (e.g., Chow 1959).  Other methods often show the dependence of m on the ratio of channel width to bend radius of curvature, B/rc with higher values for m for higher values of B/rc.

Composite n Values

Environmentally sensitive channel- and bank-protection measures often feature treatments that create highly non-uniform channel boundaries.  For example, large stone may be used along the bank toe, with woody vegetation on the middle and upper bank.  In addition, complicated cross-sectional shapes are common, particularly for high flows (Figure 2).  Wide berms or benches that support various types of vegetation may occur next to more uniform main channels, and meander bends in the main channel, floodway or both are common.  In order to use a one-dimensional computational model, the engineer must either compute a single resistance coefficient for the entire cross-section, or he must subdivide the cross section into regions of more or less homogenous flow conditions. 

The second approach is discussed below under the heading "conveyance method."  Many of the aforementioned computer models have the capability of simulating the effects of irregularly-shaped cross sections with varying types of roughness in different parts of the section using either approach.  The models typically contain formulas for combining n-values for different parts of a cross section into a single value, and then the Manning equation shown above is applied to the entire cross section using the "composite" n-value.  In order to use such a model to simulate a reach with vegetated boundaries, the engineer must select appropriate n-values for each of the segments ("panels") of the cross section, and when using some models, select an appropriate approach for combining these n-values.  Other models provide only one n-composition approach.  Information about several composition approaches is summarized in Table 2 below. 

image009
Figure 2. Representing a cross section with varying hydraulic roughness in a one-dimensional flow model.

Regardless of the method used to compute composite n-values, the magnitude of the hydraulic influence of vegetation or woody debris is closely related to the fraction of wetted perimeter covered with the vegetation or debris. Clearly, bank vegetation has more influence on the conveyance of narrow channels than wide ones. For example, Masterman and Thorne (1992) presented a case study for a gravel-bed channel with bed material size of 118 mm (4.6 in) and flow depth of 1 m (3.3 ft). For a width/depth ratio of 5, addition of dense vegetation to banks reduced discharge capacity by 38%, but for width/depth ratios of 20 and 30, discharge capacity was reduced only 8% and 6%, respectively.

TABLE 2:  Formulas for computing composite Manning n values, nc.   A= cross-sectional area, P = wetted perimeter, R = hydraulic radius and R = A/P. The subscript i refers to the ith panel in the cross section.   Panels are line segments between coordinate points.
Method
Formula for nc
Works well when

Total force
Chow (1959)


Floodplain flow depth is more than 30% of main channel flow depth
Equal velocity (used by HEC-RAS for n variation within the main channel)
Chow (1959)

There are rough vertical walls or steep side slopes
Lotter Motayed and Krishnamurthy (1980)


Conveyance Method

When the channel consists of a central main channel and a wide overbank, berm or floodplain on one or both sides, the conveyance method is recommended for flow simulation.  In the conveyance method, the channel is treated as several parallel channels for flow computation purposes, and results are simply summed (Figure 3).  For example, for the channel shown in Figure 3, the total conveyance, Ktotal = Klob + Kch + Krob. SAM uses the conveyance method as another technique for producing composite hydraulic properties, but HEC-RAS uses conveyance computations in the more orthodox fashion--to analyze flow in overbank and main channel areas more or less separately. 

image020
Figure 3.  Representation of complex cross section using conveyance approach.  Variables are defined above.

Shortcomings of These Approaches

Recent research shows that the boundary between slow-moving flow on the floodplain or berm and faster-moving flow in the main channel is the location for considerable turbulence, leading to momentum transfer and energy losses that are not well represented in either the composite-n or in the conveyance approach.  However, these losses are likely much smaller than those due to solid boundary-induced shear except for very narrow channels.

SELECTING N-VALUES FOR VEGETATED BOUNDARIES

Clearly, in order to use the methods prescribed above, the engineer must select reliable n-values for vegetated portions of the channel boundary.  Different approaches are needed for flexible and rigid vegetation.

Flexible Vegetation

Knowledge regarding interactions between vegetation and flow in rivers is expanding rapidly.  Rigid vegetation, like large trees or woody debris, should be analyzed differently from flexible types like grasses, and vegetation that protrudes through the water surface (emergent) has a different effect on flow than fully submerged vegetation. In general, flow resistance due to flexible vegetation declines with increasing discharge as stems are flattened by the flow.  For example, Oplatka (1998) reported that field tests conducted on 3 to 6 year old willows grown from cuttings showed that the area of the plants perpendicular to flow decreased by a factor of 4 to 5 at a flow velocity of 1 m/s (3.3 ft/s) and by a factor of 20 to 40 at a flow velocity of 4 m/s (13 ft/s).  Since resistance due to flexible vegetation is a function of the shear stress applied to the vegetation, iterative solutions are always required.

Darcy f values for parts of the boundary covered with flexible vegetation may be selected using approaches developed by Kouwen (1988), but this approach requires an estimate of the stiffness of the vegetation and local boundary shear stress.

where

a and b are coefficients that are based on the ratio of total boundary shear , ( t0 = g ynS) to a critical shear for the vegetation,  yn is the depth of flow, h is the length of the vegetation, M is the vegetation stem density, E is the modulus of elasticity and I is the stem area’s second moment of inertia.  Together, the term MEI represents the overall resistance of the stems to deformation by the flow.  Kouwen presents a table of values for a and b that suggest that a and b are equal to 0.15 and 1.85, respectively for erect grasses.  The coefficient a varies from 0.20 to 0.29 and b from 2.70 to 3.50 for prone grasses.

Kouwen (1988) suggests that vegetation stiffness may be measured in the field using a simple test that involves dropping a board on the erect grass.  He also showed that the results of the board test were highly correlated with grass height for the grasses he tested.  For example,

Green grass MEI = 319 h3.3

Dormant grass MEI = 25.4 h2.26

A similar approach for finding n-values for grassed channels was developed by the Soil Conservation Service (1954) that allows selection of n-values based on the type of grass and the product of velocity and hydraulic radius.  This technique is included as an option within SAM, but the types and sizes of grass are limited.

Less information is available for flexible woody plants than for grasses.  Kouwen and Fathi-Moghadam (2000) present results of tests on four woody coniferous species designed to simulate conditions when nonsubmerged flexible vegetation occurs in vegetated zones of river cross sections.  An iterative procedure is again required.  The user may read Manning n values from the table below given a flow velocity.  Tabulated n-values must be corrected as follows

Corrected n = (tabulated n)[(a/at)(yn/h)]1/2

Where a = total top-view area of the canopy of a typical individual tree or shrub, at = total area of simulated floodplain divided by the number of trees or shrubs,  yn = depth of flow, and h = height of vegetation (without deformation due to flow).  For example, a floodplain conveying flow 1 m deep at 1.0 m/s supporting 75% cover of cedar trees 2 m high would have a Manning n value of

Corrected n = 0.112*[0.75*(1/2)]1/2 = 0.069

Additional species-specific empirical relationships for Manning values for regions covered by flexible woody vegetation are provided by Copeland (2000), who provides a synthesis of findings for a flume study involving 220 different experiments and 20 shrubby plant species tested over a range of depths and velocities (Freeman et al. 2000).  In general, it was found that flow resistance increased with flow depth for partially submerged shrubs and woody plants.  However, as flow depth increased, the plants bent and became submerged at flow depths less than 80% of the plant height.  Resistance decreased with flow velocity for submerged plants as they bent and presented a more streamlined profile to the flow.  The leaf mass or foliage canopy diverted flow beneath the canopy, resulting in significant velocities along the channel bed and general scour.  The aforementioned empirical relations are rather complex power-law regression functions.

Rigid Vegetation Including Large Woody Debris

Manning's n-values for regions covered with rigid vegetation or woody debris depend upon the size and spacing of the rigid objects (say, trees) and whether they are submerged or protrude through the free surface.  Semi-empirical equations are available that allow computation of Manning's n given vegetation density and the drag coefficient.  This approach is illustrated in some detail by Arcement and Schnieder (1989), but they use an "effective" drag coefficient based on field data that is about 10 times greater than laboratory measurements.  Shields and Gippel (1995) used a similar approach, but with drag coefficients based on laboratory flume data to compute the influence of large woody debris on Darcy f.  Manning n values for rigid vegetation are highest for emergent or just-submerged vegetation and decline as vegetation becomes more deeply submerged (Shields and Gippel 1995; Wu et al. 1999).

HEC-RAS allows n-variation with stage, but the user must decide how n depends on stage.   As noted above, flexible vegetation may be flattened at higher stages, reducing n, and submergence of low, rigid debris or vegetation also reduces n.  However, flow across a floodplain may encounter higher n values as stage increases and flow encounters the crowns of trees.

TABLE 3: Estimated Manning's n for vegetated zone of rivers and floodplains in metric (SI) units from Kouwen and Fathi-Moghadam (2000)

Velocity (m/s)

Velocity (ft/s)

Cedar

Spruce

White pine

Austrian pine

0.1

0.3

0.190

0.201

0.198

0.208

0.2

0.7

0.162

0.171

0.169

0.178

0.3

1.0

0.148

0.156

0.154

0.162

0.4

1.3

0.138

0.146

0.144

0.151

0.5

1.6

0.131

0.139

0.137

0.144

0.6

2.0

0.126

0.133

0.131

0.138

0.7

2.

0.122

0.129

0.127

0.133

0.8

2.6

0.118

0.125

0.123

0.129

0.9

3.0

0.115

0.121

0.120

0.126

1.0

3.3

0.112

0.118

0.117

0.123

1.1

3.6

0.110

0.116

0.114

0.120

1.2

3.9

0.107

0.114

0.112

0.118

1.3

4.3

0.105

0.111

0.110

0.115

1.4

4.6

0.104

0.110

0.108

0.113

1.5

4.9

0.102

0.108

0.106

0.120

1.6

5.2

0.101

0.106

0.105

0.110

1.7

5.6

0.099

0.105

0.103

0.109

1.8

5.9

0.098

0.103

0.102

0.107

1.9

6.2

0.097

0.102

0.101

0.106

2.0

6.6

0.096

0.101

0.100

0.105

Temporal Factors

When trying to simulate flow effects of vegetation, the user must decide if seasonal factors should be considered.  Case studies show that flow resistance due to deciduous vegetation in full leaf is much greater than for dormant (winter) conditions (Chow 1959, Wilson 1973).  HEC-RAS allows input of constant factors for adjusting n-values by month.  Finally, over the long term, vegetation may impact conveyance by inducing sediment deposition.  Inclusion of sedimentation in the hydraulic analysis for a project will increase the cost and complexity of the analysis several times.

REFERENCES
   
Aldridge, B.N., & Garrett, J.M. (1973). Roughness coefficients for stream channels in Arizona. (U.S. Geological Survey Open-File Report), 87 pp.

Arcement, George J. Jr., & Schneider, Verne R. (1989). Guide for selecting Manning's roughness coefficients for natural channels and flood plains. (United States Geological Survey Water-Supply Paper 2339).  Denver, Colorado:  United States Government Printing Office. (pdf)

Barnes, H.H., Jr. (1967). Roughness characteristics of natural channels. (U.S. Geological Survey Water-Supply Paper 1849), 213pp. http://www.engr.utk.edu/hydraulics/openchannels/Index.html

Brownlie, W. R. (1983). Flow depth in sand bed channels.  Journal of Hydraulic Engineering, 109 (7), 959-990.

Brunner, G. W. (2001). HEC-RAS, River Analysis System User’s Manual.  US Army Corps of Engineers Hydrologic Engineering Center, Davis, CA (pdf)

Chow, V. T. (1959).  Open-channel hydraulics. McGraw-Hill Book Company, New York.

Copeland, R. R. (2000). Determination of flow resistance coefficients due to shrubs and woody vegetation.  Technical Note No. ERCD/CHL CHETN-VIII-3, U. S. Army Engineer Waterways Experiment Station, Vicksburg, MS. (pdf)

Cowan, W. L. (1956). Estimating hydraulic roughness coefficients. Agricultural Engineering. 37(7). 473-475.

Federal Interagency Stream Restoration Working Group (FISRWG) (1998).  Stream Corridor Restoration:  Principles, Processes, and Practices.   GPO Item No. 0120-A; SuDocs No. A 57.6/2:EN 3/PT.653.  ISBN-0-934213-59-3. (pdf)

Freeman, G. E., Rahmeyer, W. H., & Copeland, R. R. (2000). Determination of resistance due to shrubs and woody vegetation.  (Technical Report No. ERDC/CHL TR-00-25), U. S. Army Engineer Waterways Experiment station, Vicksburg, Mississippi, 62pp. (pdf)

Henderson, F. M. (1966). Open-channel flow.  New York, MacMillan Publishing Co., Inc., 522 pp.

James, C. S. (1994). Evaluation of methods for predicting bend loss in meandering channels. Journal of Hydraulic Engineering,. 120(2): 245-253.

Kouwen, Nicholas. (1988). Field estimation of the biomechanical properties of grass.  Journal of Hydraulic Research, 26(5):559-567.

Kouwen, N. & Fathi-Moghadam, M. (2000). Friction Factors For Coniferous Trees Along Rivers. Journal of Hydraulic Engineering, 126(10)732-740.

Masterman, R. & Thorne, C. R. (1992). Predicting Influence of Bank Vegetation on Channel Capacity. Journal of Hydraulic Engineering, 118(7):1052-1058.

Motayed A. K. & Krishnamurthy, M. (1980). Composite roughness of natural channels.  Journal of Hydraulic Engineering Division, 106(HY6):1111-1116.

Oplatka, M. (1998). Stablititat von weidenverbauungen an flussufern Versuchsanstalt fur Wasserbau, Hydrologie und Glazioloige der ETH Zurich. 156, 244.

Shields, F. D., Jr., & Gippel, C. J. (1995).  Prediction of effects of woody debris removal on flow resistance. Journal of Hydraulic Engineering. 121(4):341-354. (pdf)

Thomas, W. A., Copeland, R. R., Raphelt, N. K., & McComas, D. N. (1993). User's Manual for the Hydraulic Design Package for Channels (SAM) . U.S. Army Corps of Engineers Waterways Experiment Station, Vicksburg, Mississippi.

USDA (1954). Handbook of channel design for soil and water conservation. Prepared by Stillwater Outdoor Hydraulic Lab., Stillwater, OK., Soil Conservation Service, U.S. Department of Agriculture, Washington, D.C. (pdf)

Wilson, K. V. (1973). Changes in floodflow characteristics of a rectified channel caused by vegetation, Jackson, Mississippi.  Journal of Research of the U.S. Geological Survey. 1(5): 621-625.

Wu, Fu-Chun; Shen, Hsieh Weh, & Chou, Yi-Ju. (1999). Variation of Roughness Coefficients for Unsubmerged and Submerged Vegetation. Journal of Hydraulics 125(9) 934-942.

TOP