|
|
|
 |
| |
|
| |
Mini-Workshops Accepted
|
| |
 |
|
|
| |
|
|
Computational
Mathematics in the Oil and Gas Industry
I
|
|
Organizers:
German Larrazabal & Saul Buitrago
Speakers:
Carlos Torres M., Pablo Guillen,
& Saul Buitrago
|
|
|
| Abstract
|
| The oil and gas industry
are a perfect target for applications
of mathematics techniques. From subsurface
exploration to production, refining,
transport and finally distribution
of oil, gas and products, real life
mathematical and decision problems
arise continuously. This mini-workshop
aims at bringing together scientists
and engineers from academia, research
laboratories, and industry, to promote
the application mathematics techniques
to the oil and gas industry problems.
Papers on all aspects of mathematical
theory, and computational and applied
mathematics focus to solve oil and
gas industry problem are invited to
be presented at the mini-workshop. |
| |
Carlos
Torres
Comparative Study of Non-Isothermal
Transient Gas Flow Models in Pipelines |
| |
Pablo
Guillen
Parallel Approach for Gas Pipeline
Network Simulation |
 |
In this paper we develop and implement
a parallel approach for a simulation
model of steady and transient regimes
in a pipeline network. The gas pipeline
network is composed of a set of nodes
and segments (pipeline sections).
The solution model is characterized
by the estimation and correction of
pressures in the network nodes until
a convergence criterion in mass flow
balance is satisfied. This model allows
the concurrent execution of the calculation
on all the segments, which have a
large computer processing requirement
compared with the flow balance process.
Our parallel approach was coded in
MPI based on master/slave paradigm,
where the master process reads the
required information (network topology,
border and initial conditions, and
the mechanical and thermal parameters);
assigns the load to each process and
calculates the global balance. Simultaneously,
the slaves calculate the mass flow
for each pipeline section assigned.
The parallel approach was implemented
and tested on a synthetic network
with 12 nodes and 14 sections; we
obtained a maximum acceleration of
7.25 for 16 processors.
Keywords: Simulation, Gas pipeline
network, Parallel Computing.
|
| |
|
|
|
|
|
|
|
SDSU: Computational Science
and Engineering Gateway to Latin America
|
Last
updated:
June 3, 2010 10:06 AM
|
|
|