 |
| Titel |
Coupling R and PHREEQC: an interactive and extensible environment for efficient programming of geochemical models |
| VerfasserIn |
Marco De Lucia, Mich. Kühn |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250080795
|
|
|
|
|
|
| Zusammenfassung |
| PHREEQC [1] is a widely used non-interactive open source software for speciation,
batch-reactions, one-dimensional transport, and inverse geochemical caclulations. It
represents the tool of choice for many researchers and practicioners for a broad set of
geochemical problems, underground CO2 storage among others. Its open source nature, the
flexibility to program arbitrary kinetic laws for the chemical reactions, as well as a thorough
implementation of the Pitzer formalism explain its success and longevity. However, its
non-interactive architecture make it cumbersome to couple PHREEQC to transport programs
to achieve reactive transport simulations [2], but also to overcome the limitations of
PHREEQC itself regarding the setup of large numbers of simulations - for example exploring
wide ranges of conditions - and the graphical evaluation of the results. This has
been the main motivation leading to the development of an interface with the high
level language and environment for statistical computing and graphics GNU R
[3].
The interface consists of minor modifications in PHREEQC’s C source code, only affecting
data I/O, plus on the R side a bunch of helper functions used to setup the simulations -
basically automated manipulation of PHREEQC’s input files, which are text files - and to
collect and visualize the results. The most relevant subset of PHREEQC’s capabilities and
features are fully usable through the interface.
Illustratory examples for the utility of this programmable interface were given in the
framework of the research project this developement originated from: CLEAN [4], a project
investigating the feasibility of enhanced gas recovery combined with CO2 storage. This
interface allowed us to successfully and easily manipulate, compare and refit chemical
databases, perform sensitivity analysis by combinatory variations of parameters, and all
that in an environment which is both scriptable and interactive, with all results
directly available for further manipulations and visualization in a powerful high level
language, and benefiting from an enormous amount of third-party open source R
extensions.
The possibility to rapidly prototype complex algorithms involving geochemical
modelling is in our opinion a huge advantage. A demonstration is given by the
successful evaluation of a strategy to reduce the CPU-time needed to perform reactive
transport simulations in a sequential coupling scheme. The idea is the “reduction”
of the number of actual chemical simulations to perform at every time step, by
searching for “duplicates” of each chemical simulations in the grid: such comparison
involves typically a huge number of elements (one chemical simulation for grid
element for time step) and a quite large number of variables (concentrations and
mineral abundances). However, through the straightforward implementation of the
prototype algorithm through the R/PHREEQC interface, we found out that the scan is
extremely cost-effective in terms of CPU-time and typically allows a relevant speedup
for simulations starting from a homogeneous or zone-homogeneous state. This
speedup can even greatily exceed that of parallelization in some favorable but not
unfrequent case. This feature should therefore be implemented in reactive transport
simulators.
References
[1] Parkhurst D, Appelo C (1999) Users guide to PHREEQC (version 2). Tech. rep, U.S.
Geological Survey.
[2] Beyer C, Li D, De Lucia M, Kühn M, Bauer S (2012): Modelling CO2-induced
fluid-rock interactions in the Altensalzwedel gas reservoir. Part II: coupled reactive transport
simulation. Environ. Earth Sci., 67, 2, 573-588.
[3] R Core Team (2012) R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL
http://www.R-project.org/.
[4] Kühn M, Münch U (2012) CLEAN: CO2 Large-Scale Enhanced Gas Recovery.
GEOTECHNOLOGIEN Science Report No. 19. Series: Advanced. Technologies in Earth
Sciences, 199 p, ISBN 978-3-642-31676-0. |
|
|
|
|
|
|