Modeling and Simulation of Nuclear

Fuel Recycling Systems

David DePaoli
Oak Ridge National Laboratory

Managed by UT-Battelle
for the Department of Energy

Presented at:
Introduction to Nuclear Chemistry and Fuel Cycle
Separations
North Las Vegas, Nevada
July 20, 2011

Outline
Introduction
Some applications to date
Solvent extraction
Plant-level modeling
Agent design

Key research needs

Advancing to the future
NEAMS vision
Separations M&S development

Used Nuclear Fuel Recycling Entails

Many Interconnected Steps
Used fuel

Products

varied composition

meeting stringent
specifications

(burn-up, cooling time)

Environmental, safety, accountability constraints

shearing
solidification

dissolution

Fuel
voloxidation
feed & accountability

solvent extraction
cycles

off-gas treatment

waste treatment

recovery systems

Waste forms

Need new processes to meet future goals

Emerging modeling and simulation capabilities can improve
development and implementation (better, cheaper, faster)

Benefits of modeling and simulation of

nuclear reprocessing systems
Reduced cost of process development by guiding
and minimizing the amount of experimental and
piloting work required
Compare different separation and fuel cycle strategies
Develop new chemical processes with lower cost and waste
generation

Optimized system designs, with reduced design

margins
Scale up with confidence
Reduce hot-cell footprint (surge capacity, throughput)
Process control

Increased safety and acceptance of regulatory

bodies

Reduced risk of material diversion by providing

accurate predictions of materials streams

Modeling and simulation

Modeling is the development of an approximate
mathematical description of physical and chemical
processes at a given level of sophistication and
understanding.
Simulation utilizes computational methods to obtain
predictions of process performance.
Important Note: Advanced
modeling and simulation do
Together modeling and simulation
not replace the need for
experiments!!!
enhance understanding of known
systems,

provide qualitative/quantitative insights and

guidance for experimental work, and
produce quantitative results that replace difficult,
dangerous, or expensive experiments.
(Basic Research Needs for Advanced Nuclear Energy Systems,
http://www.sc.doe.gov/bes/reports/abstracts.html#ANES)

Modeling and simulation of nuclear separations

has primarily focused on solvent extraction

TBP

TBP

M ( NO3 ) ai bi TBP

A. D. Ryon,
ORNL-3045,
1961

NO

ai

NO3 M

a
M ( aqi ) ai NO3 ( aq) bi TBP( org)

Ki

ai

M ( NO )

bi TBP(org)

3 ai

M ( NO ) b TBP
M NO TBP
3 ai

ai

aq

org

ai
3 aq

bi
org

Original predictions:
Graphical stage
calculations using
experimental equilibrium
data and operating lines

Existing models are based on empirical fits to

experimental data

KU'

UO2 ( NO3 ) 2 2 TBPorg

UO NO TBP
2

'
KTh

aq

org

Th( NO3 ) 4 3 TBPorg

Th NO TBP
4

aq

K H'

2
3 aq

4
3 aq

3
org

HNO3 3 TBPorg

H NO TBP

aq

3 aq

org

KU' C1 C2 C3 2 C2 3
'
KTh
C5 C6 C7 2 C8 3

K H' C9 C10 C11 2 C12 3

H 3 UO22 10 Th 4
Rainey and Watson, 1975

AMUSE Models Solvent Extraction

Speciation
Aqueous
Phase
Speciation

Feed
Compositions

D value

100

SASPE
Activities

1000

Distribution
Ratio
Calculation

10
1

D( M )

Organic
Phase
Speciation

0.1
TBP , NO
, H
, H

OK 0 , K1 ,...K n ,

1 ,... n

0.01
0.001
0.1

10

100

[HNO3 ], M

Input

Mass
Balance
SASSE

Process
Parameters

Flowsheet
Module

y1i
Mass
Balance

y1o

y2i y2o

y3i

y3o

yi

Org.

Org.

Org.

yo

xo

Aq

Aq

Aq

xi

x1i

x1o

x2i x2o

x3i

x3o

Calculated
Flowsheet

Strip
b

Scru
n

4.50E-02-5.00E-02

o
acti
E x tr

1.00E+00
1.00E-02
1.00E-04

3.50E-02-4.00E-02

0.050

2.50E-02-3.00E-02

0.045

2.00E-02-2.50E-02
1.50E-02-2.00E-02

0.040

1.00E-02-1.50E-02

0.035

Conc., M

4.00E-02-4.50E-02
3.00E-02-3.50E-02

Org. Conc. of Metal

Output

1.00E-06
1.00E-08
1.00E-10

Org

1.00E-12

Aq
1.00E-14

5.00E-03-1.00E-02
0.00E+00-5.00E-03

0.030

1.00E-16
1.00E-18

0.025

Stage

0.020
0.015
0.010
0.005

Stage

0.000

Dec. Scrub
Concentration

Product

Raffinate

M. Regalbuto and C. Pereira, ANL

AMUSE has been used for process upset and product

diversion analysis
Metal, aq. conc.

-1-0
-2--1
-3--2
-4--3
-5--4
-6--5
-7--6
-8--7
-9--8
-10--9
-11--10
-12--11

Stage

Scrub conc.

-1-0
-2--1
-3--2
-4--3
-5--4
-6--5
-7--6
-8--7
-9--8
-10--9
-11--10
-12--11
-13--12
-14--13

Metal, org. conc.

AMUSE was used to bracket the

operational window for a plant
conceptual design
Four fuel compositions were used
as the initial process feed
High and low burn-up; long- and
short-cooled fuels
Results showed little difference with
cooling time but stronger effect due
to burnup differences
More recently AMUSE has been used to
examine the effect of changing specific
process parameters on the behavior of
different elements
Design of instrumentation to track
material
Process control
Product purity determination
Product diversion detection

Stage\
HNO3, feed

Changes in feed composition lead to changes in

the concentration profiles in the aqueous and
organic phases

Recent work Linkage of AMUSE to plant-scale flowsheet in ASPEN

M. Regalbuto and C. Pereira, ANL

1980s a full plant model

Consolidated Fuel Reprocessing Program - a complete plant simulation run in

the Advanced System for Process ENgineering (ASPEN) simulator

52 components tracked throughout a preconceptual design of a plant

containing 32 systems and approximately 700 streams, including:

fuel cleaning and storage

disassembly and shearing
dissolution and feed preparation
hulls drying
feed clarification
feed preparation and accountability
solvent extraction
solvent extraction ancillary systems (concentration, backcycle, storage, high-activity
waste concentration, solvent recovery)
process support (acid and water recovery and recycle, process steam, and sump)
product conversion
cell atmosphere cooling and purification
process off-gas treatment
vitrification
vitrification off-gas treatment

Large simulation for computers of the time

broken down into three segments that were executed separately to achieve a steadystate material balance for the complete plant.

R. T. Jubin, ORNL

Sandia Safeguards Performance Model

ModelMBA1: Front End

Each unit operation

is a subsystem that
models the process

Blue blocks represent the

numerous accountancy or
process monitoring
measurements

Each signal is an array that contains the

mass flow rate of each element as well as
bulk liquid and solid flow ratearrays are
expandable
Red blocks are potential
diversion points that can
be turned on by the user
B. B. Cipiti, SNL

Advanced Monitoring System Results:

Direct Diversions of 8 kg of Pu

Abrupt, Direct Diversion

Protracted, Direct Diversion

B. B. Cipiti, SNL

Current state of separations process modeling

Solvent extraction most developed
Useful aid in process development and
analysis
Semi-empirical fits
Many species not modeled well, or not
at all
Leading codes use equilibrium stages
Current codes do not predict well:

Example of transient output from SEPHIS process

model for U/Pu solvent extraction step.

Mass transfer and reaction kinetics

effects
Effects of micellization, third-phase
formation, radiolysis, etc.

Few transient codes

Other important processes not well

modeled
Legacy modules for some important
unit operations are available
e.g., dissolver, acid recovery

Full plant models are crude

Simple descriptions of important unit
operations
Not fully transient

Height of bars indicates concentration of uranium

in organic (top) and aqueous (bottom) phases in a
48-stage mixer settler bank. Time indicates the
time in minutes from the start of operation with zero
concentration throughout the system.

Sequestering agents are the basis for separations

Influence of ligand architecture

Large effects on binding affinity:
O

O
O

O
O
O
O O

O O
O

O
O
O O

106

1010

O
O O

105

103

and significant impacts on selectivity:

N
N

N
N

Hg2+/Cu2+ = 5

N
N

N
N

Cu2+/Hg2+ = 104

HO2C
HO2C

CO2H

HO2C

CO2H

HO2C

Eu3+/Al3+ = 13

CO2H
CO2H

Al3+/Eu 3+ = 500

B.Hay, ORNL

Experimental development is slow and expensive

Identify
candidate
structures

Optimize:
binding affinity
selectivity
solubility
stability
cost
Design
criteria

Computational
screening

Molecular
modeling

Ligand synthesis

Structural characterization
including X-ray analysis

Thermodynamic studies,
Binding constant measurement

Testing under process conditions

Computer-aided ligand development

can acclerate progress

B.Hay, ORNL

140 g
360 g
500 g

Experimental validation

3.90
3.30
N

Extraction of Sr2+ from 1M HNO3 using 0.1 M ligand in n-octanol, 25 C

N
N

O
Extraction into t-butylbenzene from Oaqueous
solution
containing 1 M NaNO3, 1.5 mM HNO13, 0.1 mM
Eu(N03)3, and 1-L of 15 5Eu tracer solution.

Horwit z, E.P.; Diet z, M.L.; Fisher, D.E. Solvent Ext r. Ion Exch.
O 1991, 9, 1.

1
N

1,6
7
2

Extraction into t-butylbenzene from a

containing 1 M NaNO3, 1.5 mM HNO
Eu(N03)3, and 1-L of 15 5Eu tracer s

E
u

Ext
con
Eu(

E
u

3ICJ
3IDJ
3IEJ

N
O

10

102

log

DEu

Sr

E
u

104

100
10-2

N
O

[ligand], mollit -1

[ligand], mollit -1

10-4

-1

10-6
0.0001

0.001

0.01

0.1

[ligand], M
10 million times more effective

-2
10

12
14
U,
kcal/mol
U,
kcal/mol

16

Lumetta, G. J.; Rapko, B. M.; Hay, B. P.; Gilbertson, R. D.; Weakly,

T. J. R.; Hutchison, J. E. J. Am. Chem. Soc. 2002, 124, 5644.

B.Hay, ORNL

Recent workshops have identified key needs

for contributions by modeling and simulation
Separations Challenges
Plant-scale simulation
integrated toolset to enable full-scale simulation of a plant chemistry, mass
transport, energy input, and physical layout
dynamic plant models

Computational fluid dynamics

Multiple fluid phases, fully developed turbulence, non-Newtonian flows, interfacial
phenomena, radical chemical processes due to the presence of ionizing radiation

Predictive methods for thermodynamics and kinetics data as input

to process simulators
extend currently limited thermodynamics data reliably into broader ranges of
parameter spaces
incorporate limited experimental data and use computational chemistry
approaches

http://www-fp.mcs.anl.gov/anes/SMANES
/gnep06-final.pdf

Rational design of the separations system from first-principles

physics and chemistry
predict what molecules will have the desired properties and can be synthesized
reliably predict the properties of liquids, solvation, and kinetics in solution

Connecting/crossing time and length scales, with uncertainty

quantification
access longer times without dramatic changes in theoretical and algorithmic
approaches
span spatial regimes; critical regime is the mesoscale (1 nm-1 m)
Below 1 nm, computational chemistry; above 1 m, continuum approaches

Data management and visualization

Data must be captured, managed, integrated, and mined from a wide range of
sources to enable the optimal design and operation of separation processes
Computer resources and access
Export control issues

http://www.sc.doe.gov/bes/reports/
abstracts.html#ANES

Advanced Modeling and Simulation has

become an Essential Part of DOE-NE
R&D

R&D Objective 1 Develop technologies and

other solutions that can improve the reliability,
sustain the safety, and extend the life of
current reactors.

R&D Objective 2 Develop improvements in

the affordability of new reactors to enable
nuclear energy to help meet the
Administration's energy security and climate
change goals.

R&D Objective 3 Develop sustainable

nuclear fuel cycles.

R&D Objective 4 Understand and minimize

the risks of nuclear proliferation and terrorism.

NEAMS Program Elements

Fundamental Methods
and Models
Cross-cutting
Methods and
Tools

Integrated Performance and Safety Codes (IPSC)

Waste Forms

Safeguards and
Separations

Reactors

Integrated
Performance
and Safety
Codes

Nuclear Fuels

Verification, Validation &

Uncertainty Quantification
Capability Transfer
Enabling Computational
Technologies

Continuum level codes that will predict the performance

and safety of nuclear energy systems technologies
Attributes include 3D, science based physics, high
resolution, integrated systems
Using interoperability frameworks and modern software
development techniques and tools

Program Support Elements

Develop crosscutting (i.e. more than one IPSC) required

capabilities

Fundamental Methods and Models

Verification, Validation and Uncertainty Quantification
Capability Transfer
Enabling Computational Technologies

NEAMS Safeguards and Separations

Scope
Instrument
Models

Safeguards

Used
Fuel

Disassembly

Head-end operations

Voloxidation

Dynamic Inventory
Analysis

Dissolution

Separations
Aqueous Separation
Steps
Electrochemical
Separation

Alternative Separation

Conversion

To
Recycled
Fuel
Production

Solid Waste
to
Repository

Solidification and
packaging

Offgas
Treatment
Gaseous Emissions

GOAL: Predictive capability for performance and safeguards aspects of

reprocessing plants, to guide further research and development.

NEAMS Reprocessing Plant Simulator

Toolkit
Vision
Dynamic plant-level simulation capability encompassing entire plant and integrated safeguards
Modules should be extensible, interoperable, and hierarchical to ensure continued usability
Structured as a three-tiered collection of modules to aid in development and implementation
Toolkit is open-source, but will enable use of controlled modules

Scientific Needs and Features

Plant-level modeling and simulation of
major sub-systems
Dynamically coupled plant-level
physicochemical modules.
Integrated instrument models and
inventory analysis modules.
Loosely coupled unit operation modules as
higher fidelity models
Support stand-alone first-principles
simulation capability
Support analysis and design of alternative
reprocessing technology
Support safeguards by design

Software Needs and Features

Complex plant simulations
through dynamic integration of
user plug-in software modules
Detailed visualization of realtime, dynamic simulation data
and analysis
Job launch capability on multiple
computational platforms

Distributed computing
capabilities
Support verification, validation,
and uncertainty quantification
Documentation and materials
that facilitate usage of the toolkit

Plant-level models are under development for main unit

operations. Example: Dissolver

Modern M&S for Solvent Extraction

Centrifugal Contactor Simulations Using OpenSource CFD

Goal Provide a pathway for:
Predicting stage efficiency for given conditions
Developing lower fidelity models for plant-level unit operations models

FY10 Results
Finite Element-based (unstructured) Lattice Boltzmann Method (FELBM)
Testing of modifications to codes solver and BC implementation
Verified excellent parallel scaling performance up to 2048 cores on Argonnes BlueGene/P

OpenFOAM
Demonstration of limits to approach for three phase liquid-liquid-air simulation
Coupled mixing/separation zone contactor models exploring BC and mesh dependence

K. E. Wardle, ANL

Animation of Liquid-Liquid Mixing

oil is red
water is
transparent
blue
Only large droplets are resolved (~1mm)

Actual droplet size, ~25 m

~5 m mesh (x, ~50x smaller)
N ~ 11011 cells
t ~ 110-7 s
Cr limit, as x , t

K. Wardle Multi-Fluid Flow in Liquid-Liquid Contactors ANS, June 30, 2011

25

Comparison of effect of vane

geometry on mixing
4-Vane

8-Vane

Curved

K. E. Wardle, ANL

Interface with Experimental Work

Contactor CFD Validation Using Electrical Resistance
Tomography (ERT)
current in
current out

Contactor CFD Validation Facility

Engineering-scale contactor (CINC V-5)
Multiphase measurements using ERT
Detailed operational diagnostics

reconstructed
phase fraction
maps

electrodes
K. E. Wardle, ANL

Circular 32-Electrode Array Tomography Data: 4-Vane

Acquired data is a temporal and spatial average

Temporal: ~0.75 s (overall rate of ~1.4 Hz)
Spatial:

Out of plane: electrode size, 7.6mm x 7.6mm

In plane: ITS quotes 5% of vessel diameter (5% of
6 = 0.3 = 7.6mm)

Approximate vane location is shown

Measurements are relative to a reference
measurement (avg of 100 frames)
Effect of internals is masked

Tomographic reconstruction based on

Sensitivity Conjugate Gradient (SCG) method
[Wang 2002]
Result is generally asymmetric

Wang, M. Meas. Sci. Technol. 13:101 (2002).

This was unexpected, but feed is also asymmetric

Addition of windows to vane plate could visually
verify this result
K. Wardle Multi-Fluid Flow in Liquid-Liquid Contactors ANS, June 30, 2011

28

Flow Regime Visualization

State-of-the-art high-speed digital video imaging,
solid-state light, and optics provide needed insight
Sub-millimeter flow regime never seen before
Reveals significant time and length scales under realistic
system and operating conditions
1:5 O/A flow ratio

1 mm

Elapsed time: 37 ms
Spatial resolution: ~7 m

1 mm

5:1 O/A flow ratio

1 mm

Elapsed time: 18.5 ms

Organic-rich flow regimes possess
greater air entrainment

seeing is believing . . .

These videos provide more than insight for modeling . . .

V. de Almeida, ORNL

Computer-Aided Image Analysis

3.5 mm

Large data sets are obtained (8 GB for 1-s elapsed time)

Can utilize powerful tools of computer image analysis (machine vision)
Computing intensive need to inspect every pixel and its vicinity

Generate data
useful for model
development
bubbles
noise

bubbles

300 m

Area (pixels)

drops
drops

Drop and bubble

size distribution
Velocity tracking
Calibration of phase
mixture turbulent
momentum balance
V. de Almeida, ORNL

Solvent Extraction Agent Design

HostDesigner code parallelized for minor actinides solvent extraction
agent design
Previous simulation capability: limited MM analysis on 10 to 20 ligands (~days wall-clock time)
Current capability: MM conformer search on > 1000 ligands in ~1.5 hours wall-clock time

Current focus:

Improving functionality by adding quantum mechanics based

scoring of candidates
Improving computing throughput by embedding a quantum
mechanics library in HostDesigner

Chemical Transport
Important chemical reactions occur at the interface
Challenges
Lack of basic understanding of how species move
across interfacial region
Strongly coupled physical and chemical processes

H2O
H+

HNO3TBP

UO22+

An approach to understand
solvent extraction

Molecular dynamics simulation

NO3

UO2(NO3)2 2 TBP

HNO32TBP
C12H26

?
10 nm 1

Calibration from experimental data

Insight from molecular quantum chemistry calculations when
experimental data are not available

T BP

Interfacial Transport Visualization by

Molecular Dynamics Simulation
t = 10 ns

t = 0 ns

All uranyl adsorbed on interface

Some nitrate ion also adsorbed

Uranyl

Nitrate Ion

Water

Onset of extraction of UO22+ , NO3 , H2O

Species have crossed the TBP surfactant layer
H2O and organic hidden; TBP butyl tails hidden

TBP

Dodecane

Phys. Chem. Chem. Phys., 2010, 12, 1540615409

Solv. Extr. Ion Exch., 2010, 28, 118
J. Phys. Chem. B 2009, 113, 98529862

V. de Almeida, ORNL

Quantifying Water Extraction by

TBP/Dodecane via MD

WaterTBP-NDD

Molecular dynamics of aqueous/organic interfacial transport of water

Modeling TBP (QM parameterized)
Testing various FF: AMBER, CHARMM, GROMOS, OPLS
Large scale simulation 1 M atoms (ORNL/INL machines)
Prediction of thermodynamic (density/enthalpy) and transport properties
Extraction simulations in progress
NEUP: UTK (B. Khomami) / ORNL (V.F.de Almeida)

Sharp Interface Tracking in Rotating

Microflows of Solvent Extraction

WaterTBP-NDD
mixture

Front

tracking two-phase flow

3 mm

Development of incompressible flow, high-order accuracy interface tracking

Simulation on New York Blue Gene (1024 cores)
Porting code to INL machines
Further development on correcting interface impermeability
NEUP: SUNY-SB (Glimm) / ORNL(V. F. de Almeida)

E-chem modeling
Goal: Predictive capability of electrorefiner
performance

Mark-IV

Phenomena in an electro-refiner include:

Solid electro-dissolution, Electrohydrodynamics, Electrodeposition composition and electrode morphology,
Thermodynamics, Electro-chemical reaction
Recent Progress: Initial model developed. Results for codeposition of metals at cathodes are in agreement with
experimental studies
Uses current available data for activity and diffusion
coefficients, apparent standard electro-potential; assumes
the fuel is an ideal solid solution

0.010

Considers surface area changes due to dissolution and

deposition

i(U)
i(Np)
i(Gd)
i(total)
Experimental Data
i(Np-ss)
i(total-ss)

0.006

Current (A)

Calculates behaviors of co-dissolution at the solid anode

and co-deposition at the solid cathode of Zr-U-Pu alloys.

0.008

0.004

0.002

FY15 Target:
Model that incorporates electrochemical kinetics,
phase morphology, and property data.

0.000
-1.1

-1.3

-1.5

-1.7

-1.9

-2.1

Cathode Voltage vs Ag/Ag+ (V)

J. Zhang (LANL) & J. Willet (ANL)

Example of Safeguards Modeling: Neutron

Balance Approach for Head-end Safeguards

Concept: Correlate number of neutrons

escaping fuel bundle to number escaping
Input Accountability Tank (IAT).
Used nuclear fuel emits spontaneous and
(,n) neutrons primarily from 244Cm.
The ratio of Cm/Pu can be determined from
burnup to yield Pu mass in both places.
The number of neutrons emitted/gram from
the used fuel in an assembly should be the
same as in an Input Accountability Tank (IAT);
only these two locations in the plant are
currently analyzed.

Methodology:
Isotopic composition estimates of fuel assemblies calculated using Monte Carlo burnup
Determine neutron source terms and use MCNP for transport and multiplication calculations

H. R. Trellue, LANL

Example of Instrumentation Modeling:

Hybrid K-Edge Modeling
Hybrid K-Edge method uses two simultaneous xray measurement techniques to determine the
actinide content of a liquid sample
GOAL of the Modeling Activity: Simulate the act
of measuring a liquid specimen with a hybrid
densitometer
Develop a safeguards module for Hybrid K-Edge/X-ray
Fluorescence Densitometer (HKED) simulation
The module will read user-defined input parameters
(including simulated concentrations of thorium,
uranium, neptunium, plutonium, americium, and
curium) into the plant-level code and produce the
desired spectra.

M. Collins, LANL

Photorealistic and Physics-Realistic Interactive Models

for Test, Evaluation and Analysis

Mixer/Settler Equipment Computer Graphics 3-D Model

Model built and textured from scratch in 1.5 work days by the Los Alamos National Laboratory VISIBLE
development team, using only photos of the original equipment.
Each part of the modeled equipment can be manipulated and custom programmed for behavior.
K. Michel, LANL

Real-world vs. Virtual World

A virtual
model can
have as much,
or as little
detail as
needed.

K. Michel, LANL

Future Safeguards Data Review Interface: Safeguards Data Shown

in Context for Evaluation and Analysis of Events

12:05:40
12:05:36
12:05:34

Summary
Modeling and simulation have provided useful input
to the development of fuel cycle separations over the
past several decades.
With significant scientific advancements and vast
increases in computational power, modeling and
simulation can play an increasing role in solving the
complex challenges to be overcome in developing
advanced nuclear energy systems.
Acknowledgments
Ben Cipiti
Michael Collins
Valmor de Almeida
Ben Hay
Bob Jubin
Alex Larzelere
Kelly Michel
Candido Pereira
Monica Regalbuto
Pratap Sadasivan
Holly Trellue
Kent Wardle

Prepared by Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, Tennessee 378316285, managed by UT-Battelle, LLC, for the U.S. Department of Energy under contract DEAC05-00OR22725.
The submitted manuscript has been authored by a contractor of the U.S. Government under
contract DE-AC05-00OR22725. Accordingly, the U.S. Government retains a nonexclusive,
royalty-free license to publish or reproduce the published form of this contribution, or allow
others to do so, for U.S. Government purposes.
This report was prepared as an account of work sponsored by an agency of the United States
government. Neither the United States government nor any agency thereof, nor any of their
employees, makes any warranty, express or implied, or assumes any legal liability or
responsibility for the accuracy, completeness, or usefulness of any information, apparatus,
product, or process disclosed, or represents that its use would not infringe privately owned
rights. Reference herein to any specific commercial product, process, or service by trade name,
trademark, manufacturer, or otherwise, does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States government or any agency
thereof. The views and opinions of authors expressed herein do not necessarily state or reflect
those of the United States government or any agency thereof.