home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Hall of Fame
/
HallofFameCDROM.cdr
/
oilfield
/
spe-27.lzh
/
AAREADME.027
next >
Wrap
Text File
|
1988-09-27
|
6KB
|
117 lines
************************************************************************
SOCIETY OF PETROLEUM ENGINEERS MICROCOMPUTER USERS GROUP DISK # 27
PC/MS DOS 2.0 & UP FORMAT: DSDD 9 SECTOR
------------------------------------------------------------------------
NOTICE TO USERS: Neither the contributor nor the Society of Petroleum
Engineers makes any express or implied warranty of any kind with regard
to these program materials, including, but not limited to warranties of
of merchantability and fitness for a particular purpose. Neither the
contributor nor the SPE shall be liable for incidental or consequential
damages in connection with or arising out of the furnishing, use or
performance of these program materials.
------------------------------------------------------------------------
READ.ME 5760 The file you are reading
MODEL.FOR 116608 Source code of Waterflood Model (FORTRAN)
MODEL.EXE 105038 Compiled & compressed object code
INPUT.DAT 1536 Example input data (steady state)
INPUT2.DAT 1536 Example input data (compressible)
START UP INSTRUCTIONS
Additional support and documentation can be obtained through
the SPE BOOK ORDER DEPT. (request paper # 16491).
This program requires a Fortran compiler (MICROSOFT FORTRAN,VERSION 4 OR
RYAN-MCFARLAN, PROFESSIONAL FORTRAN) to be run. SPE has included a compiled
version in object code with example input data.
For a print-out of the program's source code, you must type "MODEL.FOR".
Description: A single phase, two-dimensional, compressible
front tracking model for microcomputers, which can be used
to simulate waterfloods. It can be used for any type of
heterogeneous reservoir with variable thickness, permeability,
porosity and flow boundries. The program can be used to
analyze the anticipated pattern of an injected fluid migration
within the reservoir as a function of time.
It is recommended the program be run on a hard disk because of its large
size. If run from the floppy disk first make a back up copy of the
original disk. This will be your working copy. There is only 113K of free
space left for the output file. This should be sufficient for most runs
unless you specify frequent print outs or have a long run. If in doubt
replace your working disk with a blank formatted disk when the model asks
for "UNIT 2" or your output file name.
The model expects the input data to be on an existing file called "INPUT.DAT.
Also a file named "RESTART.DAT" must exist on the default drive. If they
are not present, the program will terminate.
To run the program example, type "MODEL". A message will appear on
the screen that reads "File name missing or blank - please enter name."
UNIT 2?... enter "OUTPUT.DAT" (or any name desired)
The program is long and will take some time to run. When the program
has finished, a message appears that reads:
"STOP - Program terminated."
Run time for an IBM XT (4.7 mh) with a 20x20 grid and 30 wells is about
3 minutes per year.
HELPFUL HINTS:
--------------
Obtain a copy of SPE 16491 paper and read thoroughly.
Maximum number of grids is 20 X 20 with origin (x=1,y=1) upper left corner.
Maximum number of wells is 30 with maximum injection wells of 20.
Do not put wells on the boundry columns or rows.
Just for information, the model considers the input file as UNIT 1 and the
restart file as UNIT 4. UNIT 2 the output file is the only one that has
to be named when the model first starts. After termination the file can
be copied to printer but carriage controls will be lost. If page controls
are desired for neat output specify "PRN" as the file for Unit 2 when the
model requests it, and output will go direct to the printer. BE SURE THE
PRINTER IS THE COMPRESED MODE OR YOU HAVE A 136 COLUMN PRINTER.
Two sets of example input data are on the disk. INPUT.DAT is for a steady
state run and INPUT2.DAT is for a compressible run. The files will have
to be renamed before running INPUT2.DAT. For example:
rename INPUT.DAT to INPUT1.DAT
rename INPUT2.DAT to INPUT.DAT
The model is a simple but good single phase model and it has no material
balance checks. Do not assume your data will automatically run the first
time, especially in the compressible mode. Generally the problem that
occurs is non-convergence. Usually this can be corrected by raising the
convergence tolerance variable "TOLR". Never use a value less than 0.1.
A value of 0.1 seems to work very well in the steady state mode (ie. non
compressible). In the compressible mode the value of "TOLR" may have to
be raised as high as 0.5. This will tend to induce some small error in
your pressures. However, we are talking about possibly one to two per-
cent error which is probably better than the data anyway.
Run time is longer in the compressible mode as the pressures have to be
updated after each time step. However, this mode is a more realistic
approach to the actual field performance since existing production and
injection rates can be input. If the rates entered are severe and are
not compatible with the transmisibility, instability may exist in the
calculated pressures and non-convergence will occur.
To make a restart run change input variable "IREST" from zero to one.
Change "TIME" to the number of days you want in this run. The model will
automatically cumulate the time.
This version has been compiled using the standard math library. If you
have an 8087 co-processor recompile the model using the 8087 library to
increase the speed.
standard math library. If you
have an 8087 co-processor recompil