This group includes all inflows and outflows, seepage, potential heads, concentrations, well withdrawals, and the like (see figure).

 

Possible boundary conditions

 

Minimum required boundary conditions in a flow model:

For an unambiguous definition of a model, either a fixed potential head or a rate must be specified for at least one node. That node can have a value for the fixed potential head (data type POTE) or a lekage boundary condition (VORF and LERA or LEKN or LEEL defined in a at least one adjacent element).

 

The following options are available for the nodes of the model boundary in a groundwater flow model:

• 1^st type or Dirichlet boundary condition:

Dirichlet, also known as fixed potential- or constant-head, boundary conditions can be assigned with the attribute POTE. Once a 1^st type boundary conditions has been assigned to a node, then a 2^nd type boundary condition (KNOT (sink/sources) or RAND (on-/outflow rates) cannot be assigned to the same node. However, inflows from the data type FLAE or GW-N (surface infiltration, which is simulated as a 2^nd type boundary condition in 3D models) in the neighbouring elements of horizontal or 3D models is allowed. Inflow or discharge rates associated with constant-head boundary conditions are calculated automatically from the mass/water balance (as per the principle of conservation of mass).

 

• 2^nd type or Neumann boundary condition:

By specifying the data types KNOT or RAND, inflow or outflow rates are specified as Neumann, also known as flux, boundary conditions. In 3D models, the recharge- / surface infiltration rates (FLAE or GW-N) are also 2^nd type boundary conditions, i.e. they define an inflow rate over the upper model boundary.

 

• 3^rd type or Cauchy boundary condition:

The Cauchy, or head-dependant flux, boundary condition is an amalgamation of the 1^st and 2^nd type boundary conditions, and is referred to as the 3^rd type of boundary condition in groundwater flow modelling. It corresponds to a defined leakage condition (see also material characteristics) where the flux is dependent on a specified head.

 

If no boundary condition is explicitly defined for a node on the model boundary, then it is automatically treated as an impermeable (no-flow) boundary condition, where flow across the boundary is assumed to be zero (2^nd type boundary condition with the rate equal to zero).

 

Impermeable/no-flow boundary conditions can be used to represent, for example, sheet pile walls or watershed/catchment boundaries, and can also be useful perpendicular to known hydraulic head contours.

Regardless of whether a 1^st type boundary condition (POTE) was defined for a boundary node or not, the inflow from the top of the model can be defined for a horizontal model using the FLAE or GW-N data type.