BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Modified Sequential Fully Implicit Scheme for Compositional Flow S
 imulation - Arthur Moncorge\, TOTAL
DTSTART:20151112T113000Z
DTEND:20151112T123000Z
UID:TALK60915@talks.cam.ac.uk
CONTACT:Catherine Pearson
DESCRIPTION:The Fully Implicit Method (FIM) is widely employed for general
 -purpose reservoir simulation. However\, in the Multi-Scale Finite-Volume 
 (MSFV) approach\, sequential strategies are used to couple flow (pressures
  and velocities) and transport (saturations/compositions)\, that is\, eith
 er Implicit Pressure Explicit Saturation (IMPES) or Sequential Fully Impli
 cit (SFI) integration schemes. For problems with tightly coupled nonlinear
  interactions between flow and transport in multiphase\, multi-component s
 ystems\, the SFI approach may require many more Newton iterations\, and/or
  smaller time steps compared with FIM.\n\nWe analyzed the nonlinear coupli
 ng between flow and transport for multiphase\, multi-component systems tha
 t involve significant compressibility effects and interphase mass transfer
 . Therefore we isolated the nonlinear effects related to transmissibility 
 and compressibility from those due to interphase mass transfer. It could b
 e demonstrated that when the convergence rate of the FIM is far superior t
 o that of SFI integration\, there exist large differences between the spec
 tra of the SFI and FIM operators. Based on these findings we propose a mod
 ified SFI scheme\, that is\, the SFIM method\, whereby the SFI operator is
  complemented with a local approximation of the pressure-saturation coupli
 ng terms. We show across a wide parameter range that this SFIM algorithm h
 as convergence properties that are close to those of FIM and consistently 
 better than those of SFI. Black-oil and compositional systems in depletion
  and compression settings are presented and discussed.\n
LOCATION:Open Plan Area\, BP Institute\, Madingley Rise CB3 0EZ
END:VEVENT
END:VCALENDAR