modeling of engineered systems inthe vadosezone

SRNL-MS-2010-00070

Modeling of Engineered Systems in the Vadose Zone

Greg Flach

13 April 2010

Richland WA

Performance Assessment Community of Practice Technical Exchange

2

SRNL-MS-2010-00070

Outline

Engineered Systems at the Savannah River Site

Key failure / degradation modes

Modeling philosophy

Modeling practice

Opportunities for ASCEM and CBP

3

SRNL-MS-2010-00070

Engineered systems

Solid waste disposal, E-area

Intermediate-level and low-activity waste vaults

Components-in-grout trench disposal

Tank closures, F- and H-area

Saltstone disposal facility, Z-area

Decontamination and Decommissioning (D&D)

reactor buildings

canyons

other building slabs and below-grade infrastructure

4

SRNL-MS-2010-00070

Intermediate Level Vault (ILV)

5

SRNL-MS-2010-00070

Intermediate Level Vault (ILV)

6

SRNL-MS-2010-00070

Low Activity Waste (LAW) Vault

7

SRNL-MS-2010-00070

Low Activity Waste (LAW) Vault

8

SRNL-MS-2010-00070

Components-In-Grout (CIG) trench disposal

9

SRNL-MS-2010-00070

Components-In-Grout (CIG) trench disposal

10

SRNL-MS-2010-00070

Tank closures

11

SRNL-MS-2010-00070

Tank closures

12

SRNL-MS-2010-00070

Saltstone Disposal Facility

Groutplant

Vault 1

Vault 4

Future disposal

cells

Future disposal

cells

Groutplant

Vault 1

Vault 4

Future disposal

cells

Future disposal

cells

13

SRNL-MS-2010-00070

Saltstone Disposal Facility

14

SRNL-MS-2010-00070

Decontamination and Decommissioning (D&D)

Disassembly Basin

Dashed red line indicates the cross section location

Reactor Building

15

SRNL-MS-2010-00070

Decontamination and Decommissioning (D&D)

Alternative 2A:• reactor vessel remains (capped with concrete)• below-grade portions of building grouted• above-grade portion of disassembly basin removed

Assembly Area

Purification Area

Process Area

Disassembly Basin

- Existing Building Concrete (Walls, Floors, Ceiling and Roof)

- Grout

- Above Grade, Concrete Cover Over Disassembly Basin

- Ground

- Reactor Vessel

0’

-20’

-40’

0’

-20’

-40’

-51’ -51’

Water Table

NOT TO SCALE

- Concrete

Actuator Tower

ReactorVessel

- New Roofs

16

SRNL-MS-2010-00070

Key failure / degradation modes

Saltstone

sulfate attack and rebar corrosion for concrete

oxidation for HDPE liners

differential settlement / seismic events for structure

Waste tanks

acid leaching/decalcification for saturated concrete

carbonation / rebar corrosion for unsaturated concrete

general and pitting corrosion for steel tank

17

SRNL-MS-2010-00070

Key failure / degradation modes (continued)

E-area vaults

structural cracking due to seismic events and cover system overburden

rebar corrosion for concrete

Components-In-Grout trenches

grout credited for 300 years

D&D

varies

18

SRNL-MS-2010-00070

Modeling philosophy

Under, but close to,performance objective

simple is often sufficient

tailor waste stream to disposal options

no containment

components-in-grout

vaults

avoid unnecessary decontamination, cleaning, pre-treatment, and over-engineered closures → Graded approach

19

SRNL-MS-2010-00070

Modeling philosophy (continued)

Reasonable assurance

historically

deterministic analysis with conservative assumptions

sensitivity analysis

increasingly

probabilistic analysis

using abstracted models

typically a hybrid blend

deterministic results from higher fidelity model

probabilistic results from simpler, abstracted, model

Compliance?

SimplifiedTime

Mean

5%

95%

limit

Abstraction

Benchmark

Dose vs. Limit

Time (yr)

Detailed

HYBRIDHYBRID

Compliance?

SimplifiedTime

Mean

5%

95%

limit

Abstraction

Benchmark

Compliance?Compliance?

SimplifiedTime

Mean

5%

95%

limit

Abstraction

Benchmark

SimplifiedTime

Mean

5%

95%

limit

Abstraction

BenchmarkTime

Mean

5%

95%

limit

Abstraction

Benchmark

Abstraction

Benchmark

Dose vs. Limit

Time (yr)

Detailed

HYBRIDHYBRIDDose vs. Limit

Time (yr)

Detailed

Dose vs. Limit

Time (yr)

Detailed

HYBRIDHYBRID

20

SRNL-MS-2010-00070

Modeling philosophy (continued)

Core team approach

"core team" of stakeholders and subject matter experts

consensus on conceptual approach and anticipated outcomes

vetting of key results and interpretations

formalized for CERCLA / RCRA work

21

SRNL-MS-2010-00070

Modeling philosophy (continued)

Regulator constraints and considerations

CERCLA/RCRA

formal "core team" approach and decisions

agreement to use Groundwater Modeling System (GMS) graphical user interface and GMS supported codes by default

ready access to familiar software preferred

free, commonly used, well documented, open source

strong software pedigree preferred

QA, large user base and associated track record

22

SRNL-MS-2010-00070

Modeling practice

Typical practice is described

Vadose zone modeling only

no aquifer

Emphasis on

Performance Assessments

PORFLOW modeling

23

SRNL-MS-2010-00070

Software

HELP

cover system

PORFLOW

2D deterministic

GoldSim

1D probabilistic

The Geochemist's Workbench

pH and redox transitions, solubilities

STADIUM

24

SRNL-MS-2010-00070

Computational demands

Thousands of runs required

geometries

cases (best-estimate, sensitivity, . . .)

flow periods

transport species

re-runs (learning, do-overs, updated information, . . .)

Need turnaround in one to a few days

2D vadose zone with (only) several thousand nodes

temporal variability defined by a sequence of steady-stateflow fields

25

SRNL-MS-2010-00070

Computational resources

Windows (~50%)

typically multi-core personal desktops

interactive

Linux (~50%)

typically SRNL clusters

500 cores (subset available)

batch

Licenses

13 network for PORFLOW, 1+ node-locked for GoldSim

26

SRNL-MS-2010-00070

Workflow

Auxiliary supporting calculations

pH/Eh transitions

material degradation

Pseudo-transient vadose zone flow

sequence of steady-state flow solutions

single-phase, single-component

single-domain, porous-medium

Transient vadose zone transport

abbreviated radionuclide chains

linear sorption, sometimes with solubility constraints

27

SRNL-MS-2010-00070

Scripting for PORFLOW simulations

Manual approaches are often not practical

too many cases, materials, properties, species, etc.

do-overs in response to errors, new inputs, evolving needs

difficult to document and reproduce

Extensive use of pre- and post-processing scripts

more efficient, and provide a valuable electronic record

ad hoc development of several years by multiple analysts

often extensive linking (piping) of existing software

many languages and utilities

often unwieldy

28

SRNL-MS-2010-00070

Visualization

PORFLOW

Tecplot

Excel

Python

GoldSim

GoldSim GUI

Excel

29

SRNL-MS-2010-00070

Data management

Primary

Excel

Text files

GoldSim

Infrequent

"Capital D" Databases (e.g. Access)

30

SRNL-MS-2010-00070

Dimensionality

1D

probabilistic and deterministic analyses

2D

deterministic

axisymmetric or half-width

3D

rarely

D&D example

31

SRNL-MS-2010-00070

Grid resolution

D&D example

mtyp: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Tank example

5-10k nodes

32

SRNL-MS-2010-00070

Fast flow paths and fractures

Explicit fast-flow paths

usually a postulated condition

modeled as thicker seam of sand or gravel

one to a few in number

Smaller scale fractures

cracking is assumed form of concrete physical degradation

compute effective properties based on combined matrix andfracture flow

matrix and fracture assumed to be in equilibrium

33

SRNL-MS-2010-00070

Postulated fast flow path through Type IV tank

34

SRNL-MS-2010-00070

Hydraulic property variations (cementitious materials)

Frequently,

Instantaneous failure to a soil surrogate, or

Gradual increase in saturated conductivity, retainingcharacteristic curves (water retention, relative perm.)

Postulated or assumed degradation / failure

Increasingly,

Modified saturated conductivity and characteristic curves to reflect combined matrix and fracture flow

Degradation based on an auxiliary, mechanistic, analysis(e.g. STADIUM for sulfate attack)

35

SRNL-MS-2010-00070

Combined matrix and fracture flow example

B b

matrix flow fracture flow

36

SRNL-MS-2010-00070

Effective unsaturated K based on Orr and Tuller (2000)

1.E-12

1.E-11

1.E-10

1.E-09

1.E-08

1.E-07

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

1.E-01

1.E+00

0.1 1 10 100 1000 10000 100000 1000000

Suction Head (cm)

Un

satu

rate

d C

on

du

cti

vit

y (

cm

/s)

Intact concrete

Ksat correction only

Cracked concrete

Saturated flow

Thick film flow

Thin film flow

Active fracture

Inactive fracture

Transition zone

Matrix flow

37

SRNL-MS-2010-00070

Sulfate attack example

degradedconcrete

intactconcrete

ettringitefront

saltstone(high

sulfateconc.)

soil (low

sulfateconc.)

degradedconcrete

intactconcrete

ettringitefront

saltstone(high

sulfateconc.)

soil (low

sulfateconc.)

38

SRNL-MS-2010-00070

Vault 2 Ettringite Front

0

5

10

15

20

25

30

35

0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000

Elapsed Time (yr)

Po

sit

ion

(cm

)

Vault 2 wall

Vault 2 floor

Vault 2 roof

Ettringite formation based on STADIUM simulations

39

SRNL-MS-2010-00070

Vault 2 Effective Properties

1.0E-11

1.0E-10

1.0E-09

1.0E-08

1.0E-07

1.0E-06

1.0E-05

1.0E-04

0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000

Elapsed Time (yr)

Eff

ecti

ve C

on

du

cti

vit

y (

cm

/s)

Vault 2 wall

Vault 2 floor

Vault 2 roofcompliance period

Vault 2 Effective Properties

1.0E-11

1.0E-10

1.0E-09

1.0E-08

1.0E-07

1.0E-06

1.0E-05

1.0E-04

0 2000 4000 6000 8000 10000 12000 14000 16000 18000 20000

Elapsed Time (yr)

Eff

ecti

ve C

on

du

cti

vit

y (

cm

/s)

Vault 2 wall

Vault 2 floor

Vault 2 roofcompliance period

Effective saturated conductivity

40

SRNL-MS-2010-00070

Chemical property variations

Chemical equilibrium analysis

The Geochemist's Workbench

pH, Eh transitions defined in terms of pore volumes

Timing defined by flow simulation

Kd and solubility defined by pH / Eh regime, e.g.

Region II versus III in parlance of Bradbury and Sarott (1995)

Oxidizing versus reducing conditions

literature and/or SRNL experiments

41

SRNL-MS-2010-00070

Saltstone or Disposal Unit

Material

Infiltrate

Eluate = Infiltrate equilibrated with

cementitious material

Volume of Infiltrate = Volume of Eluate for Each Step

Saltstone or Disposal Unit

Material

Infiltrate

Eluate = Infiltrate equilibrated with

cementitious material

Volume of Infiltrate = Volume of Eluate for Each Step

Saltstone example

8

8.5

9

9.5

10

10.5

11

11.5

12

0 500 1000 1500 2000 2500 3000 3500 4000

Fluid Reacted (Pore Volumes)

pH

Case 1Case 2

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

0 500 1000 1500 2000 2500 3000 3500 4000

Fluid Reacted (pore volumes)

Eh

(m

V)

Case 1

Case 2

Case 3

pH Eh

42

SRNL-MS-2010-00070

SRS modeling interests

Multiphase, multicomponent, reactive transport

predict evolution of bulk chemistry

pH, Eh (x, t) + empirical Kd (pH, Eh) → Kd (x, t)

oxidation / carbonation of unsaturated concretes and grouts

especially for Tc-99 in Saltstone slag concretes and grouts

gas-liquid partitioning and transport in air pathway analysis

Probabilistic "full-physics" analyses

take advantage of evolving hardware

43

SRNL-MS-2010-00070

SRS modeling interests (continued)

Fractured-media approaches for cracked materials

dual-porosity? explicit fracture networks?

Open-source flow and transport code

PORFLOW replacement?

eliminate licensing limitations and "black-box" mysteries

expanded accessibility and user base

Mechanistic material degradation predictions

damage mechanics and associated properties

corrosion