Geometric Data

The basic geometric data consist of cross-section data, reach lengths, energy loss coefficients (friction losses and contraction and expansion losses), stream junction information and hydraulic structure data. Similarly, data collection is required downstream of the modelled system to prevent any user-defined boundary condition from affecting the results.

Cross-sections

Coupled with the hydrological data, the geometric data determine the conveyance of water downstream, both within the channel and across the floodplain. The boundary geometry for the analysis of flow is specified in terms of ground surface profiles (cross-sections) and the distance between them (reach lengths). The cross-sections are located at intervals along the river in order to characterise and accurately represent the flow

37 carrying capacity and geometry of the channel and its floodplain. Cross-sections are required where changes to roughness, discharge, shape and slope occur, as well as before and after weir and bridge structures. In HEC-RAS the cross-sections extend across the entire floodplain, perpendicular to anticipated flow.

Each section is labelled with a river, a reach and a river station label. The cross-section is described by entering the station and elevation (x-y) data from left to right. All cross-section data are defined looking in the downstream direction. The specific points at which ‘overbank flow’ is defined along the channel cross-section are called ‘left and right overbank stations’. The reach lengths between cross-sections must be specified for the left bank, right bank and channel. Channel lengths are typically measured along the thalweg and overbank distances along the expected centre of mass of the overbank flow.

These values differ more greatly as river bends.

Levées can be designated within the geometry data by identifying a left or right bank levée station and elevation on a cross-section. Once established, water can only pass the levée station if the water level exceeds the elevation of the levée. The levées are usually established at an existing point on the cross-section.

The spacing of cross-sections depends on slope, stream size and uniformity, but also the purpose of the study in question. For example, studies analysing the effect of local geomorphology on flow depths will require more closely spaced cross-sections than a study investigating the deposition of sediment in reservoirs. Interpolation is often required where the velocity head is too large to accurately determine the change in energy gradient. Inadequate cross-sectional spacing can result in significant computational errors as a result of inaccurate integration within the profile computations. This error can be effectively removed by adding interpolated cross-sections. Increasing the density of the cross-sections will improve the accuracy of the profile computations (the solution of the equations). It is necessary to have frequent cross sections in order to describe the hydraulic behaviour of the channel with acceptable precision, but this must be balanced with time pressures of this procedure (Samuels, 1990). Interpolating over a very small distance between cross sections creates too much data and unnecessary expense, whilst very large spacing can result in calculation instabilities (Samuels, 1990).

38 Roughness

The Manning’s n selected for each cross-section is a major influence on the friction slope and hence energy losses. Thus it influences the conveyance of water and therefore the speed at which water can flow through a channel. The amount of friction combined with the channel geometry determines the depth of flow within a channel at a certain location.

The application of Manning’s n values contains an element of subjectivity but there are standard values recommended for different channel and floodplain types. The criteria for selecting n values are well documented in U.S. Army Corps of Engineers (2008) and Chow (1959), the main source for obtaining n values (Table 3.1). Values for Manning’s n for left overbank, right overbank and channel are required for each cross-section.

Expansion and contraction coefficients

Energy losses can occur between two cross-sections as a result of contraction or expansion of flow. The loss is computed using the contraction and expansion coefficients specified in the cross-section data editor. The energy loss is calculated by multiplying the coefficients by the absolute difference in velocity heads between one cross-section and the next one downstream. If the change in cross-section is small or gradual, the contraction and expansion coefficients are typically 0.1 and 0.3 respectively. These values are recommended by the HEC-RAS manual to account for gradual changes in river sectional area (U. S. Army Corps of Engineers, 2008). Where the change in effective cross-section area is abrupt (e.g. at bridges) contraction and expansion values of 0.3 and 0.5 are typically used.

39 Type of Channel and Description Minimum Normal Maximum

Natural Streams

h. Very weedy reaches, deep pools, or floodways with heavy stands of timber and brush.

i). Scattered brush, heavy weed 0.035 0.050 0.070

ii). Light brush and trees in winter 0.035 0.050 0.060 down trees, little undergrowth, flow below branches (Source: Chow, 1959; U. S. Army Corps of Engineers, 2008).

40 Structures: bridges

In order to model bridge structures, HEC-RAS requires four user-defined cross-sections in order to compute energy losses around the structure. Cross-section one is located far enough downstream so that the flow is not affected by the structure. Field investigations during high flow should determine this distance. A second cross section should be located just downstream of the bridge to represent the topography of the channel and floodplain.

A third cross-section should be placed just upstream of the bridge to represent the channel and floodplain upstream. A fourth cross-section should be located upstream of the bridge where flow lines are approximately parallel. This method allows for the changes in water surface profile and energy losses apparent in the vicinity of bridges to be calculated. For a 1-D approach, as for HEC-RAS, expansion and contraction coefficients are used to quantify energy losses as a result of contraction and expansion through a bridge based on the abruptness of the change. Typical bridge sections are given contraction and expansion coefficients of 0.3 and 0.5 respectively which must be specified by the user. The coefficients must be adjusted back to an appropriate value downstream of the structure.

Structures: weirs

HEC-RAS has the ability to model broad-crested, ogee shape and sharp-crested weirs as inline structures across the main river. Like bridges, weirs have an effect on flow as water backs up behind them causing a localised increase in width and depth. The presence of a weir also has an effect on river-floodplain interaction and both the bank and bed can be subject to erosion. In HEC-RAS the flow over a weir is computed using the standard weir equation:

Equation 3.4

where Q is discharge, C is the weir flow coefficient (typical values range from 2.6 to 4.0 depending upon the shape of the spillway crest), L is the length of spillway crest and H is the upstream energy head above the spillway crest.

During very high discharges, a weir can become ‘submerged’ and, if a flow gauging station is present (such as Westwick and Hunsingore), it will be unable to provide accurate measurements (Rickard et al., 2003). HEC-RAS automatically accounts for submergence if the tailwater is high enough to slow the flow. As submergence increases, the weir flow coefficient is automatically reduced. The shape of the weir selected determines how

HEC-41 RAS calculates submerged weir flow. In order to model inline structures it requires the same cross-section framework as for bridges. The geometry of the weir is entered using elevation and station data across the river. The same contraction and expansion coefficients apply for weirs as for bridges.

Storage areas

Floodplains and storage areas in a river system are represented in one of two ways in HEC-RAS: either by using a series of extended cross-sections which includes the properties of the floodplain (topography and roughness) within the specification of each cross-section, or by representing the floodplain as a series of discrete flood cells, hydraulically connected to neighbour cells and/or the main channel, through which water is routed (Maidment, 2002; Tayefi et al., 2007). In order to add a storage area to the model, two different methods can be used. Firstly, the area times depth method requires the area of the storage and minimum elevation to be entered. Secondly, the elevation versus volume method requires volume measurements for each elevation of the storage area. This is the recommended method where possible as it is more detailed. HEC-RAS uses lateral structures to link two storage areas together and to link the storage areas to cross-sections. As in Tayefi et al. (2007) and Hornberger et al. (1998), flow exchanges between the storage areas are calculated using the standard weir equation for a broad-crested weir (Equation 3.4). For a lateral structure connecting a cross-section to a storage area there must be a cross-section upstream and downstream of the lateral structure. For each lateral structure station data, elevation data, weir width and a weir coefficient are required.