[Opm] flow: Usage of restart utility

Joakim Hove joakim.hove at gmail.com
Thu Jun 9 16:15:43 UTC 2016


>
>
> My open question is, what is meant by time step. I assume that this is
> the simulation step number in the ouput of flow.



OK;

this is my mental picture of the timeline in an ECLIPSE simulation:


X  : Report step - given with the DATES or TSTEP kewyord in the Schedule
section of the input file.
 |  :  A timestep chosen by the adaptive timestep algorithm

X-------|------|----|---|--X------|---|----|-----X--------|--------X
[](                        ](                     ](                 ]
0                          1                     2                 3

This is a simulation with three report steps 1,2 & 3 and in addition there
is an initial report step 0 which only contains the initial equilibrium
solution. Observe the half open intervals (...] indicated under the
timeline, there we see that the results which 'happen'  in the timeinterval
(R-1,R] are attached to report step R - but the report step 0 is a
singularity - spanning the time interval [0,0]. Understanding these
intervals is most important when it comes to summary files, where we have
results for all simulator timesteps - but they are "attached" to report
steps as indicated above.

All IO configuration is in terms of report steps, so the comment
<https://github.com/OPM/opm-parser/blob/7ef3a7b4c1db3ff99dc5bf5fbb6ee6ce8504c8b7/opm/parser/eclipse/EclipseState/IOConfig/IOConfig.hpp#L38>s
you
linked to are unfortunately misleading - every single[1] occurence of
'timestep' in the RPTRST configuration should be replaced with REPORT STEP.

HTH - Joakim

 [1]: Of course - there is always an exception, there  is a RPTRST setting
( Ithink it is BASIC=6) which will actually get utput for every simulator
timestep - that is not supported in flow.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <//opm-project.org/pipermail/opm/attachments/20160609/aa1a2ef5/attachment.html>


More information about the Opm mailing list