FEFLOW - Modelling Unsaturated-Flow and Infiltration Processes (27 Oct, Webinar)

Modelling unsaturated-flow and infiltration
processes using FEFLOW
Dr. Carlos Rivera ([email protected])

Global Product Specialist – FEFLOW
DHI
Agenda
• Introduction
• Unsaturated-flow modelling
• Live demonstration
• Q/A session
© DHI #2
MIKE Modelling & Simulation
© DHI
With FEFLOW, you can simulate a multitude of
groundwater processes involving flow,
contaminants, groundwater age and heat
transport under fully or variably saturated
conditions.
© DHI
Development History
1979 Birth and first manifestation based on finite-element predecessor program FINEL developed since 1973.
1986 FEFLOW 1: FEFLOW was already able to compute 2D transient groundwater flow and transport.
1992 FEFLOW 3: Commercial development, Motif GUI implementation, FEFLOW became a registered trademark.
2001 FEFLOW 4: 3D Visualization, GIS interfacing, Adaptive meshing, Unsaturated flow, IFM Interface (API)
2012 FEFLOW 6: New Qt-based GUI, data management and architecture based on C++.
2013 FEFLOW 6.2: New Problem Class “Groundwater Age”, FePEST became true.
2015 FEFLOW 7: Support of fully 3D Unstructured meshes, significant improvements in meshing workflows.
2017: FEFLOW 7.1: Dual-node support, coupling 2D surface water (M21FM.)
2018: FEFLOW 7.2: Significant extension to Python interfacing.
2019 FEFLOW 7.3: Cloud-computing support in FePEST.
2020 FEFLOW 7.4: Hydromechanical support, new OpenGL GUI, extended Python API support, meta information.
2021 FEFLOW 7.5: Hydrodynamical support, PEST++, model conceptualization

FEFLOW’s core values
All-in-one solution Save time & boost efficiency Extensive visualisation

• Full development engine & GUI • Import/export industry standard file formats • 2D/3D views, animations and virtual
• Minimises dependences on • Interoperability with 3D geological software reality
external packages • Smart and fast meshing algorithms • Verification of model settings
• One software for all modelling • Friendly programming interface to speed up • Professional result presentation
processes workflows
© DHI
With FEFLOW, you can have it all
GET THE RESULTS YOU NEED WITH THE USER EXPERIENCE YOU DEMAND
• Unstructured meshes (2D/3D) • Premium technical support

• Stable and efficient numerical • Ongoing professional training through
schemes and methods The ACADEMY
• Professional and flexible user • Free Viewer mode for post-processing
interface • Free convenient MODFLOW to
• PEST user interface for automatic FEFLOW conversion tool
calibration, uncertainty analysis,
predictive analysis and more
© DHI
Get all your questions answered using a single sofware
THERE’S NO NEED FOR DIFFERENT PACKAGES
• From the simple to the complex - • One project = one file

FEFLOW makes it possible! • DHI develops both the engine and
• Multi-scale support: from local to GUI - eliminating dependency on
regional third-party software
• A single user-interface
© DHI
All modelling workflow in one software
SAVE TIME AND BOOST EFFICIENCY
• GIS, CAD and geological model interfacing to define region/areas of interest

• Smart algorithms for meshing in 2D/3D. Discover really 3D unstructured meshes!
• Multiple methods for data regionalization
• Run every problem (flow, mass and heat) within the same interface in parallel
• Full integration with PEST including parallel computing (local, server and cloud)
• Use Monte Carlo methods to quantify uncertainty behind the predictions
© DHI
Unsaturated-Flow
© DHI #10
Application Examples
Dam seepage
© DHI
Construction of 0 d
underground
storage and
facilities 1080 d
© DHI
Efficient dewatering at mine sites
© DHI
Physical Aspects of Unsaturated Flow
• Two fluid phases present: water + gas (air)
• Capillary pressure:
pressure difference
between phases gas
water
• Gas phase reduces
conductivity:
blocking of flow paths

by gas phase
© DHI
Mathematical Description
Model is described by the Richards equation
Nonlinearity
o Describes the water phase with its constitutive relationships
o Capillary pressure function S(Ψ)
o Relative conductivity Kr(S)
o For saturated conditions, Kr = 1 and S = 1, Richards equation becomes identical to the
groundwater flow equation based on Darcy’s law
© DHI
Simplification by Richards‘ Assumption
Assumption:
o Gas phase is infinitely mobile
o Influence of gaseous shear forces are neglected
o Reduction to single phase problem
Benefits:
o Computational efficiency
o Valid for vadose and saturated zone
Not applicable if:
o Gas flow is dominating
(e.g. remediation by soil ventilation)
o Phase transitions occur
(e.g. steam production in nuclear waste site)
© DHI
Challenges
...resulting from constitutive relations s() and Kr(s())
o Nonlinearity
o Parametric expressions
o Hysteresis
© DHI
Parametric Functions
For constitutive relationships:
Soil-water Retention Relative Conductivity
S Kr
p S
© DHI
Parametric Functions
 Empirical models:
• Van Genuchten
• Brooks&Corey
• Exponential
• Van Genuchten modified
• Haverkamp
• Linear
© DHI
Parametric Functions Measurements
Alternative to empirical models:
 Splines
• Very flexible
• Can strictly honor measured data
Spline model
Empirical model
© DHI
Typical Soil Retention Curves
© DHI
Typical van Genuchten Model Parameters
© DHI
Pedrotransfer Functions for Soil Unsaturated Parameters
Source: Rajka et al. 2004.

Estimating the water retention
curve from soil properties:
comparison of linear, nonlinear
and concomitant variable
methods.
© DHI
Upstream Weighting (Spatial Integration of Kr)
Standard approach: Relative conductivity is evaluated at element centroids.
Upstream Weighting: An upstream possition is selected.
Kr is function of nodal saturation (or

pressure). Advantage: Kre is evaluated
only along element edges
© DHI
Hysteresis Effects
Scott model
PDewatering
PImbibition
© DHI
Seepage Face Boundary Condition
• Atmospheric pressure on the outside
• Outflow direction: gradient-driven outflow
• Inflow direction: no inflow
• Implementation: hydraulic-head BC + fluid-flow constraint:
Boundary condition: h = z
Fluid-flow constraint: Qmax = 0 Q=0 Seepage-face BC
p<0
(unsat. p>0
model)
p=0
Q<0
© DHI
Release of FEFLOW 8.0
A major FEFLOW release since 2015.

Release focused on model
conceptualization and significant https://www.theacademybydhi.com/training/webinars
workflows improvements!
Conceptual modelling and new 3D Supermesh Repair New Well Manager editor MIKE Cloud for FEFLOW and FePEST
workflow for 3D model building
© DHI
Live Demonstration
© DHI
Thank you!
Dr. Carlos Rivera Villarreyes
Global Product Specialist – FEFLOW
E-Mail: [email protected]
Join the FEFLOW community:

https://ocean.dhigroup.com/sign-up-for-dhi-group-newsletters/
© DHI #29

FEFLOW - Modelling Unsaturated-Flow and Infiltration Processes (27 Oct, Webinar)

Uploaded by

Copyright:

Available Formats

FEFLOW - Modelling Unsaturated-Flow and Infiltration Processes (27 Oct, Webinar)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FEFLOW - Modelling Unsaturated-Flow and Infiltration Processes (27 Oct, Webinar)

Uploaded by

Copyright:

Available Formats

Modelling unsaturated-flow and infiltration

processes using FEFLOW

Dr. Carlos Rivera ([email protected])

2017: FEFLOW 7.1: Dual-node support, coupling 2D surface water (M21FM.)

2018: FEFLOW 7.2: Significant extension to Python interfacing.

2019 FEFLOW 7.3: Cloud-computing support in FePEST.

2021 FEFLOW 7.5: Hydrodynamical support, PEST++, model conceptualization

All-in-one solution Save time & boost efficiency Extensive visualisation

• Unstructured meshes (2D/3D) • Premium technical support

THERE’S NO NEED FOR DIFFERENT PACKAGES

• From the simple to the complex - • One project = one file

SAVE TIME AND BOOST EFFICIENCY

• GIS, CAD and geological model interfacing to define region/areas of interest

Efficient dewatering at mine sites

blocking of flow paths

...resulting from constitutive relations s() and Kr(s())

Soil-water Retention Relative Conductivity

For constitutive relationships:

Alternative to empirical models:

Source: Rajka et al. 2004.

Upstream Weighting: An upstream possition is selected.

Kr is function of nodal saturation (or

A major FEFLOW release since 2015.

Join the FEFLOW community:

You might also like