HEC-RAS pipe network computations support hydraulic control structures that influence flow within and into pipe systems. Two categories of hydraulic structures are currently supported: gates and weirs. These structures may operate under both open-channel and pressurized flow conditions and are integrated directly into the numerical solution of the pipe network equations.

Gates in Pipe Networks

Two gate types are supported in pipe network modeling: sluice gates and flap gates. 

Gates may be assigned to the downstream end of any conduit using the DS Gate Type column in the conduit attribute table in RAS Mapper. A gate may be located:

  • Internally within a pipe network,

  • At the boundary between a pipe network and a surface water region, or

  • On an external boundary of the pipe network.

Sluice Gates

Flow through sluice gates is modeled using the formulations described in the Sluice Gate section. 

Sluice gate operations are defined through the Boundary Condition Editor, where the user specifies gate opening heights. Gate operations may be controlled using:

  • Time series of gate openings,

  • Elevation controlled logic, or

  • Scripted rules.

Flap Gates

Flap gates permit flow only in the downstream direction of the conduit and prevent backflow. Flow in the downstream direction is computed as if no structure is present. Flow in the upstream direction is prevented using the frictional adjustment explained below and a high face friction factor.

Flap gate state (open or closed) is determined using the hydraulic head differential across the flap gate. A minimum head differential of 0.05 feet must be present across the gate in order to lift the gate and initiate flow. This "cracking head" represents a physical constraint on the system, and also limits computational instabilities resulting from surface waves reflecting in the pipe network opening and closing flap gates. No corresponding minimum head differential ("reseat head") is required to close a flap gate.

Computational Treatment of Gates

For sluice gates, the computational procedure is as follows:

  1. The effective gate opening height is determined.

  2. Flow through the gate is computed based on the gate opening and upstream and downstream hydraulic heads.

  3. The frictional resistance at the downstream computational face of the conduit is modified to achieve the computed gate flow.

This approach avoids numerical instabilities that can occur when internal flow boundary conditions prescribe more inflow than volume available within a closed or pressurized portion of the pipe network. Both open-channel and pressurized flow conditions are supported.

For flap gates, the same procedure is applied, but the friction factor is set to a very large value to enforce near-zero flow in the reverse direction.

The expression used to set the friction factor for a given gate flow is derived by solving the Manning equation for the roughness coefficient

n = \frac{A R_h^{2/3} S_f^{1/2}}{Q_{\textrm{gate}}}

where

n = Manning’s roughness coefficient [T/L1/3],
A = the gate face flow area [L2],
R_h = the gate face hydraulic radius [L],
S_f = the slope of the hydraulic grade line [-],
Q_{\textrm{gate}} = the computed gate flow [L3/T].

At high flows, the computed flow through the gate may be limited by the capacity of the conduit. At low flows, computed gate flows are limited to not exceed critical velocity.

Weirs

Unlike gates, weirs are not explicitly specified by the user as hydraulic structures at conduit ends. Instead, a weir is implicitly generated by the computational engine when a conduit meets the following criteria:

  1. The conduit is short and connects two closely spaced nodes.

  2. The conduit is assigned a maximum cell length such that it contains only one internal computational face.

  3. Elevation offsets are specified at both the upstream and downstream ends of the conduit.

  4. Each elevation offset must be at least 0.5 ft.

These requirements result in a raised computational face with a vertical offset relative to the adjacent computational cells, which functions as a weir crest.

Full Shallow Water Equation Treatment

When the full shallow water equations are used, the weir face is treated using the same numerical routines described in Plunging Flow in Pipe Systems.

The weir equation is:

1) Q_w = C_w A \sqrt{H}

where

Q_w is the flow over the weir [L3/T],
C_w is the dimensional weir coefficient [L0.5/T],
H is the height of the water surface above the crest of the conduit face [L], and
A is conduit face wet cross-sectional area [L2].

In pipe network computations, the user does not specify a weir coefficient directly. Instead, the sum of the conduit entrance and exit loss coefficients is used as the minor loss coefficient, K_{ML}​, for the raised computational face.

For conduits with rectangular cross sections, a relationship between the minor loss coefficient K_{ML}​ and the equivalent weir coefficient C_w​ can be derived by assuming critical flow at the weir crest, and conservation of energy along a streamline from a point upstream of the weir, to the weir crest.

2) C_w = \frac{2}{3} \sqrt{\frac{2g}{3}} \left( \frac{1}{1 + \frac{K_{ML}}{3}} \right)^{3/2}

K_{ML}=0 corresponds to a weir coefficient of C_w = 3.09, which is the theoretical maximum. K_{ML}=0.5 corresponds to C_w = 2.45, and K_{ML}=1.0 corresponds to C_w = 2.01.

For non-rectangular conduit cross sections, no closed-form relationship exists. In these cases, the relationship between K_{ML}​ and C_w​ depends on the upstream depth above the weir crest.

Additional losses may be present due to the user-specified pipe roughness, n. These losses are included in the solution at the weir face and are:

h_L = \left( \frac{3}{2} \right)^{1/3} n^2 L g H^{-1/3}

where L is the conduit length [L]. 

When pipes become pressurized, the raised conduit face no longer functions as a free-flowing weir, but transitions to orifice flow. The orifice flow equation is:

3) Q_o = C_o \sqrt{2g} A_o \sqrt{\Delta H}

where

Q_o is the flow through the orifice [L3/T],
C_o is the non-dimensional orifice coefficient [-],
\Delta H is the hydraulic head differential across the conduit face [L], and
A_o is conduit face cross-sectional area at the orifice [L2].

Application of the Bernoulli equation from a location upstream of the face yields the following expression for an orifice coefficient corresponding to the user-defined value for the minor loss coefficient.

4) C_o = \sqrt{ \frac{1}{1 + K_{ML} - \left( \frac{A_o}{A_{us}} \right)^2}}

where A_{us} is conduit face cross-sectional area at a location just upstream of the orifice [L2].

Additional losses due to the user-specified pipe roughness, n, are:

h_L = 2 L \frac{n^2 g}{R_h^{4/3}}

where R_h is the hydraulic radius of the full-flowing conduit [L]. 

Diffusion Wave Treatment

The diffusion wave solution only considers a local balance between frictional losses and the hydraulic grade line. As a result, energy head conservation through flow contractions is not explicitly considered, and different computational routines are required to obtain results consistent with the weir and orifice equations. 

When the diffusion wave equation set is used, the user-entered minor loss coefficients are internally converted to equivalent weir and orifice coefficients using equations 2) and 4). Flow through the raised face is then computed using the weir equation 1) or orifice equation 3). Once the flow is determined, the Manning’s roughness value at the weir face is dynamically adjusted during the computation to generate the required discharge, using the same approach applied to gate faces within pipe networks. A transition region similar to that described in Hydraulic Computations Through Gated Spillways is used to handle flows between fully submerged orifice flow and weir flow.