Basin-Scale Density-Dependent Groundwater Flow Near a Salt Repository

Anke SchneiderGesellschaft für Anlagen- und Reaktorsicherheit

Kristopher L. KuhlmanSandia National Laboratories

Middelburg, The Netherlands

September 5-7, 2017

Sandia National Laboratories is a multi-mission laboratory managed and

operated by National Technology and Engineering Solutions of Sandia LLC, a

wholly owned subsidiary of Honeywell International Inc. for the U.S. Department

of Energy’s National Nuclear Security Administration under contract DE-

NA0003525. SAND2017-9392 C.

WIPP Hydrogeology

Repository in Salado bedded salt formation >500-m thick salt unit

Hydrogeology of formations above salt Rustler Formation

Culebra dolomite Magenta dolomite Anhydrite Mudstone/Halite

Dewey Lake Red Beds Silt/sand stones + clay

Dockum Group Silt/sand stones + clay

2

Rustler Conceptual Model

3

West

(Nash Draw)

East West of WIPP Shallow units High permeability Relatively fresh water

East of WIPP Deeper units Low permeability Saturated brine

Regional groundwater Flow used in WIPP PA Long-term geological

stability of salt

4

Corbet (2000) Model Domain

WIPP LWB

NM/TX Border

1996 WIPP PA Model

2004 PA Model

2014 WIPP PA Model

Scale [km]

0 10 20 4030

Corbet (2000) WIPP Model

5

Most of Delaware Basin Transient Simulation

Climate variation (dry vs. wet) 14,000 y → present → 10,000 y

Model Implementation “water table” moving boundary

model ~8700 km2 region (78 km × 112 km) Coarse mesh (2 km square cells) 12 model layers (10 geo layers) 1,500 cells/layer ~18,000 elements total

Halite areas (low k)

Dissolution area

(high k)

Motivation

6

Benchmark against existing solution (Corbet, 2000) Comparison with original model

Old mesh, model parameters & boundary conditions

Include new processes, features & data Include density-driven flow (e.g., Davies, 1989) Include chemistry & mineral dissolution Investigate flow & chemistry boundary conditions Test and update hydrogeological conceptual model Incoporate current data: 81Kr GW age data, water level data

Comparison and Development of Models PFLOTRAN (SNL)

Add density dependent flow

d3f++ (GRS)

Update of Corbet model

7

Corbet (2000): Hydraulic conductivity [m/s]

d3f++/PFLOTRAN: Permeability [m2]

density-driven groundwater flow

salt and heat transport

fluid density and viscosity depending on salt concentration and temperature

porous and fractured media

free groundwater surface – levelset function

sources and sinks

transport of radionuclides

decay and ingrowth

equilibrium and kinetically controlled sorption

precipitation/dissolution

diffusion into immobile pore water

colloid-borne transport

numerics based on UG, G-CSC, Frankfurt University

finite volume methods

geometric and algebraic multigrid solvers

completely parallelized (UG: scaling invest. some 100,000 proc.)

8

d³f++: distributed density-driven flow

Applications of d3f++

Applications 9

Porous media, overburden of host formations

• Gorleben Site: 2D density-driven flow and RN

transport in high saline environment

• Cape Cod: 2D contaminant transport with

pH-dependent sorption

Low permeable media

• Generic German Site in clay: 3D diffusive transport

in a low permeable anisotropic clay formation

Fractured media

• Yeniseysky site: Flow and transport in fractured rock

• Äspö (URL): Flow in the repository near field

• Grimsel (URL): Colloid-facilitated transport in clay

WIPP Site: „Basin-Scale“ model

SNL: Data of „Basin-scale“ groundwater model after Corbet & Knupp 1996

raster data of 10 hydrogelogic units

source: SNL, SECOFL3D

10

d³f++

ProMeshwww.promesh3d.com

Dewey Lake/TriassicAnhydrite 5Mudstone/Halite 4Anhydrite 4Magenta DolomiteAnhydrite 3Mudstone/Halite 3Anhydrite 2Culebra DolomiteLos Medanos Member

Forty-Niner Member

Tamarisk Member

unit permeability [m²]

Dewey Lake/Triassic 10-14-10-12

Forty-Niner Member 10-20-10-12

Magenta Dolomite 10-18-10-12

Tamarisk Member 10-20-10-12

Culebra Dolomite 10-17-10-11

Los Medanos Member 10-17

WIPP-Site: Prism grid, 6 layers

N

example:Culebra Dolomitesource: Corbet 2000

182,784 prisms (2x refined) 18,000 hexahedrons SECOFL3D 50x vertical exaggeration

11

WIPP-Site: Initial and boundary conditions

N

closed boundariesc=1(saturated brine)

12

assumed recharge ratessource: Corbet 2000

initial condition:water table 14,000 years ago

source: Corbet &Knupp 1996

WIPP-Site: Initial and boundary conditions

N

closed boundariesc=1(saturated brine)

recharge 2.0 – 0.0/0.1 mm/year, c=0 / seepage

initial condition:water table 14,000 years ago

source: Corbet &Knupp 1996

13

WIPP-Site: d³f++ simulations

14

density-driven flow, fixed water table (top boundary), permeability const. (layers)

(280,000 prisms)

model time 10,000 years, computing time 15 minutes, timestep 100 years

concentration

Darcy’s velocity

WIPP-Site: d³f++ simulations

15

density-driven flow, free water table

level 2 (182,784 prisms)

free water table

salt concentration

WIPP-Site: d³f++ simulations

16

density-driven flow, free water table

level 2 (182,784 prisms)

model time 100 years,

timestep 0.005 year (levelset method)

Darcy’s velocity, water table

Summary and outlook

Difficulties:

non steady-state density-driven flow model

strongly anisotropic (thin layers, jumping coefficients)

free groundwater surface 8,700 km²

Current work:

BMWi-funded joint project GRUSS (GRS, G-CSC Frankfurt University)

improve grid generating/refinement

improve robustness of solvers (convergence, timesteps)

implement volume of fluid (VOF) method to speed-up free surface handling

Next steps:

increase timestep levelset method

simulation 14.000 years past

reproduction of SECOFL3D results (Corbet & Knupp, 1996)

17

Reactive multiphase flow and transport code for porous media

Open source license (GNU LGPL 2.0)

Object-oriented Fortran 2003/2008 Pointers to procedures

Classes (extendable derived types with

member procedures)

Founded upon well-known (supported) open source libraries MPI, PETSc, HDF5, METIS/ParMETIS/CMAKE

Demonstrated performance Maximum # processes: 262,144 (Jaguar supercomputer)

Maximum problem size: 3.34 billion degrees of freedom

Scales well to over 10K cores18

SNL PFLOTRAN version

19~25x vertical exaggeration

SNL PFLOTRAN version

20Original Mesh: 13-layer hexahedral (cuboid) elements (18,000 elements)

100x vertical exaggeration

Issues Encountered

21

Old Mesh is very coarse PFLOTRAN and d3f++ have difficulty with mesh Mesh violates conventions regarding

Regularity (Δz varies too much in space) Connectivity (must build mesh “by hand”) Aspect ratio (2 km × 2 km × 1s-100s m)

Anke (GRS): re-mesh using modern tools (LARGE) Kris (SNL): struggle with old mesh (COARSE)

Moving water table ≠ Richards equation Unsaturated flow parameters are guessed Recharge applied at water table vs. applied at land surface

CFL condition requires very small time steps Too few elements to capitalize on parallel Smaller elements → smaller time steps!

Schedule

22

SECOFL3D data provided by SNL GRS begins building d3f++ model SNL begins building PFLOTRAN model SNL consults

GRS builds d3f++ model equivalent to Corbet (2000) SNL builds PFLOTRAN equivalent to Corbet (2000) GRS ‘includes’ density-driven flow SNL includes density-driven flow to PFLOTRAN

WIPP basin-scale model is: Numerically difficult Uses non-ideal mesh (pancake elements) Has complex boundary conditions

Try benchmarking simpler (2D) problems: Compare processes Use PFLOTRAN QA suite problem?

Ye

ar

1Y

ea

r 2

+Y

ea

r

23

Thank you for your attention!