CWELL lives in the $CWPROOT/src/Fortran/Cwell and comes with SeisUn*x from the Center for Wave Phenomena. The program comes with demos that help you learn how to run the program. I have combined the demos in this HTML file for easier reading. Your best bet is to run the demos as well.
This demo produces a plot of the model and the quits. Only 4 records are required in file PARAM1 for this run....
syncline :model file 4 :#interfaces in model plotcolors :model colors file mq :first plot descriptor (mq)Record 1 of PARAM1 contains the name of the file decribing the model; the file is called syncline here. (See below for a description of the format of the model file.) Record 2 of PARAM1 gives the number of interfaces described in the model file (not including the upper surface). Next comes the name of a file describing the colors of various elements of the plot. It is possible, on a color screen, to plot the interfaces, rays, source/receiver locations etc., in different colors. The colors are defined in file plotcolors in this demo. A key is given at the end of the file. The last record in PARAM1 for this run is a plot descriptor. To get a plot of the model enter m here. To quit immediately after the model plot enter mq.
Format of the model file
An interface is described by a list of (x,z) coordinates, through which the program fits a cubic spline. The program expects to find descriptions for the upper surface and n interfaces in the model file, where n is the integer in Record 2 of PARAM1. The program considers the upper surface to be Interface 0 and splines it just like the deeper interfaces. The upper surface should come first in the model file. An interface description is terminated by a large negative z-coordinate (z<-9999.).
0. 0. :start of upper surface 10000. 0. 1. -99999. :end of upper surface 0. 1100. :start of interface 1 1000. 1100. 5000. 1200. 8000. 900. 10000. 700. 1. -99999. :end of interface 1 0. 2100. :start of interface 2 2000. 2100. 5000. 2200. 8000. 1700. 10000. 1400. 1. -99999. :end of interface 2 0. 3100. :start of interface 3 2000. 3100. 5000. 3200. 8000. 2500. 10000. 2200. 1. -99999. :end of interface 3 0. 4000. :start of interface 4 2000. 4000. 5000. 5500. 8000. 3400. 10000. 2900. 1. -99999. :end of interface 4
5 receivers 5 sources 0 well color 5 rays that don't intersect well (7/30/91 this option turned off) 2 rays that hit well 4 interfaces key: (CWP's xgraph colors) 0 black 1 white 2 red 3 green 4 dark blue 5 light blue 6 violet 7 yellow
Run the program with the following shell script Xcwell
#! /bin/sh
# Run cwell with CWP X Graphics
version=1
x1beg=0 x1end=20 x2beg=-1 x2end=10
cwell |
cshotplot > bhrs${version}plot outpar=bhrs${version}par
xgraph < bhrs${version}plot par=bhrs${version}par style=seismic
-geometry 600x400+0-250
title="Model for BHRS"
label1="Depth (m)" label2="Range (m)"
x1beg=$x1beg x1end=$x1end x2beg=$x2beg x2end=$x2end &
In this demo plots are made of the model, the receiver well, and the receiver locations are plotted. Two more records have been added to PARAM1...
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wgq :receiver plot descriptor (wgq)
Record 5 of PARAM1 is the name of the file containing the receiver well description. This file, called rwell here, also specifies the receiver locations (see below). Record 6 of PARAM1 is another plot descriptor. To plot the receiver well enter w in Record 6. To plot the receivers enter g (for geophones). To plot both well and receivers enter wg. You can also quit at this point by tagging on a q in Record 6. (Notice how Record 4 has been changed to m - plot the model, do not quit after plotting the model.)
Format of the well file
First in the well file are a list of (x,z) coordinates describing the well. The well may be curved (the coordinates will be splined by the program) but should not be too curvy. The first record in the file contains the x-coordinate of the top of the well. The program calculates the z-coordinate of the top of the well for you (you may not know it if the upper surface is curved). The well definition is terminated by a large negative z-coordinate.
Next in the well file is the depth (arc length) down the well to the first receiver location. This is followed by the number of receivers and the receiver spacing.
IMPORTANT CONSTRAINTS
To aid in bookkeeping the program assumes the following:
(1) The well intersects ALL interfaces in the model. Make sure you define the well so that it cuts through the deepest interface in the model.
(2) No receivers are located below the deepest interface.
Example receiver well file
6000. :x-coord. of top of receiver well 6200. 3000. 7000. 6000. 1. -99999. :end of well definition 500. :depth to first receiver 20 200. :#receivers, receiver spacing
Here we have plots of the model, receiver well+geophone locations, and source well+source locations. Four records have been added to PARAM1...
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell1 :source well coordinates don't care :file containing coords. of sources wsq :source plot descriptor (wsq)
First, in Record 7, is the shooting mode. This is either d, for downhole, or g for general. The receivers are always located inside the receiver well, but there is a choice for locating the sources. In downhole mode, the sources are distributed throughout a source well; thus, we must provide a source well description (see below). The name of the file describing the source well and source locations is given in Record 8 of PARAM1. In the general shooting mode (Record 7 = g) the sources are not confined to a well, but can be anywhere in the model. In this case, a file containing a list of the source (x,z) coordinates is provided. The name of the source file is given in Record 9 of PARAM1. (See Demo06 for an example of the general shooting mode.)
Record 10 of PARAM1 is a plot descriptor for the sources. Enter w here to get a plot of the source well, s for the source locations, and/or q to quit at this point.
Format of the Source Well File (Downhole Mode)
The format of this file is the same as that for the receiver well:
Format of Source Description (General Mode)Example source well file
3000. :x-coord. of top of source well 5000. 6000. 1. -99999. :end of source well definition 500. :depth to first source 1 500. :#sources, source spacing
This file simply contains a list of source locations:
x1, z1 :location of source 1 x2, z2 :location of source 2 ... ... xn, zn :location of source n
3000. :x-coord. of top of source well 5000. 6000. 1. -99999. :end of source well definition 500. :depth to first source 1 500. :#sources, source spacing
This demo goes quite a bit further and makes a plot of rays corresponding to the direct wave. There are a lot more entries in PARAM1 for this run....
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell1 :source well coordinates don't care :file containing coords. of sources ws :source plot descriptor (wsq) r :job descriptor (rlt) don't care :output file name -90. 90. :range of takeoff angles 1. .1 3. :coarse, fine angle increment; ray density 4000. 6000. 7000. 7500. 12000. :velocities y :direct wave? n :primaries? n :head waves?
First, Record 10 has been changed to ws (plot the source well, plot the source locations, do not quit). Record 11 is job descriptor. To get a ray plot enter r in Record 11. To get a listing of some ray information enter l. To get time sections (ie, wiggle traces - built later by program CSHOT2) enter t. To get all three at once, enter rlt, and so on. See Demo07 for an example of generating listings. The first wiggle traces are produced in Demo08.
When generating a listing or time data, the program creates output files. The first part of the name given to these files comes from Record 12. Appended to this name are .listing for the listing file, and .shot for the time section file. No output files are generated when producing only ray plots, so Record 12 is not used here.
Record 13 is the range of takeoff angles over which rays will be shot out from the source. The program rotates the range of angles you specify in Record 13 depending on the direction of the target. For example, for upgoing rays (aimed at a target reflector above the source location) the angles are measured from the upward pointing vertical, as follows
0
^
|
|
-90 <------ -----> +90
Downgoing rays are measured from the downward pointing vertical
-90 <------ -----> +90
|
|
0
The coordinate system for direct waves depends on whether the source is to the left or right of the receiver well. For example, for a source to the left of the receiver well the angles are measured as
r
-90 | e
| c
source |---------- 0 e
| i
+90 | v
e
r
s
The program rotates the coordinate system automatically. A range of -180 degrees --> +180 degrees covers all possible takeoff angles; however, a more restricted range results in a faster run
Record 14 contains three entries: coarse change in takeoff angle, fine change in takeoff angle, and ray density. Beginning with the starting takeoff angle (given in Record 13) the program shoots rays out from the source until a ray intersects the well. The program uses the coarse change in takeoff angle for this search. Once a ray that intersects the well has been found, the program backs up to the previous takeoff angle and approaches the well more carefully using the fine change in takeoff angle. The aim is to find a ray that intersects the well near to the well top (or the bottom, depending on the type of ray). The reason for this is that the program uses interpolation to obtain traveltimes and amplitudes at receiver locations. We want rays to intersect the well all the way along its length (and not miss large sections at the ends). When the well is located for the second time (using the fine change in takeoff angle), the program starts to monitor the change in end point of the rays that intersect the well. This is where the ray density takes over: the program now tries to adjust the change in takeoff angle to satisfy the ray density parameter. Ray density is defined here as the number of rays that hit the well between each pair of receivers. For example, given a ray density of 3 the program will try to find 3 rays between each pair of receivers (it is not possible to guarantee this). Larger values for the ray density may result in more accurate traveltime and amplitude calculations in the interpolation procedure.
The program will continue to search for rays that intersect the well until it reaches the final takeoff specified in Record 14.
Record 15 of PARAM1 contains the layer velocities (shallowest layer first). The program expects to find one more velocity than the number of interfaces specified in Record 2.
Next come the event specifications. To generate a direct wave enter y in Record 16; else enter n. To generate primary reflections from all interfaces in the model enter y in Record 17; else enter n. To generate head wave events from all interfaces in the model enter y in Record 18; else enter n.
This time the program has been instructed to seek a primary refection from each interface (Record 17 = y). The direct wave has been turned off.
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell1 :source well coordinates don't care :file containing coords. of sources ws :source plot descriptor (wsq) r :job descriptor (rlt) don't care :output file name -90. 90. :range of takeoff angles 1. .1 3. :coarse, fine angle increment; ray density 4000. 6000. 7000. 7500. 12000. :velocities n :direct wave? y :primaries? n :head waves?
Here we have switched to general shooting mode (Record 7 = g), specified 2 source locations in file sources (Record 9), turned off the source well plot (Record 10), and asked for a multiple reflection event (Record 19). Any number of events can be specified at the end of PARAM1. Multiple bounce reflection and head wave events are allowed. In this example, the letter r in Column 1 of Record 19 specifies a multiple reflection. (The other choice is h for multiple bounce head wave. See Demo11 for an example of this.) The integers in Record 19 are the interface numbers of the reflections; thus, Record 19 specifies bounces from Interface 0 (the upper surface), Interface 2 and then Interface 1.
As with Record 19, each event is described by a list of interfaces at which the ray reflects.
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wg :receiver plot descriptor (wgq) g :shooting mode (dg) don't care :source well coordinates sources :file containing coords. of sources s :source plot descriptor (wsq) r :job descriptor (rlt) don't care :output file name -90. 90. :range of takeoff angles 1. 3. :angle increment, ray density 4000. 6000. 7000. 7500. 12000. :velocities n :direct wave? n :primaries? r 0 2 1 :multiple reflection event
Here we are back in downhole mode (Record 7) with 3 source locations (see file swell2), we have asked for a reflection from interface 2 (Record 19), for a ray plot and a listing (Record 11). The first part of the file name given to the listing comes from Record 12. Appended to this is .listing, so the file generated here will be called demo7.listing. The file begins with a list of the source and receiver locations. This is followed by the intersection coordinates with the interfaces and the traveltime for each ray. Note that the ray density has been reduced (Record 14) so as not to make the listing too long.
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell2 :source well coordinates don't care :file containing coords. of sources ws :source plot descriptor (wsq) rl :job descriptor (rlt) demo7 :output file name -90. 90. :range of takeoff angles 1. .1 1. :coarse, fine angle increment; ray density 4000. 6000. 7000. 7500. 12000. :velocities n :direct wave? n :primaries? n :head waves? 2 :reflection from Interface 2
This is the first demo in which wiggle traces are generated. This time it necessary to run program CSHOT2 in addition to CWELL. CSHOT2 reads the time, amplitude and phase file output by CWELL and actually constructs the traces. Program CSHOT2 and its input parameter file PARAM2 have been described in the CSHOT documentation. (For example, see Demo08 in the CSHOT directory.)
In order to make time data enter t in Record 11 (to make rays as well enter rt). The time data file output by CWELL takes its name from Record 12 of PARAM1. Appended to this is .shot, so the file will be called demo8.shot here. This file name must be entered in PARAM2 for CSHOT2. Here we are generating data for one shot, direct wave and all primaries.
syncline :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell1 :source well coordinates don't care :file containing coords. of sources ws :source plot descriptor (wsq) rt :job descriptor (rlt) demo8 :output file name -90. 90. :range of takeoff angles 1. .1 3. :coarse, fine angle increment; ray density 4000. 6000. 7000. 7500. 12000. :velocities y :direct wave? y :primaries? n :head waves?Example of PARAM2 file:
s :job option (s,r) 1 1 :first, last shot for sort 1 100 :first, last trace OR first last receiver 10. 25. 35. 50. :frequency spectrum of wavelet .150 :wavelet length (secs) .004 :sample rate (secs) 2. :record length (secs) demo8shot :input filename demo8traces :output filename
Explanation of PARAM2 file:
The following notes refer to the records in param2.
(1) The first record in param2 specifies the sort option. Besides constructing shot records it is also possible to produce a receiver sort. The choices for Record 1 are s or r. To build shot records choose s. For a receiver sort choose r.
(2) The second record of param2 defines the shot records to include in the sort.
(3) The third record of param2 also depends on your choice of s or r.
(i) Record 1 = s For the shot option, Record 3 contains the first and last trace to build from each shot, eg, to build traces 30-50 from each shot just specify 30 50 here.
(ii) Record 1 = r For the receiver sort option, Record 3 contains the first and last receiver sort to build. The number of traces you get in each sort depends on the number of shots for which each receiver was live, or active, within the range of shots given in Record 2.
(4)-(7) These records used to be in file ftdata in the old version of the code. Note that the sample rate is specified in seconds now.
(8) Record 8 of param2 is the name of the file generated by cshot1. It contains traveltimes, amplitudes, etc. for all the shots in the cards file.
(9) The last record in param2 is the name to be given to the file containing the actual traces built by cshot2.
It is also possible to generate a common offset sort. This is done with the shot option by specifying only one trace to build in Record 3, eg, to build trace 10 only from all shots in Record 2, specify 10 10 in Record 3. (I am assuming, of course, that the trace in question is recorded at a receiver which is always the same distance from the shot. You need to be careful enough in the shot cards file if you want to make this happen. What it means is that the shot and receivers must move up the same amount from one shot to the next.)
Must also use more lines in Xcwell
#! /bin/sh
# Run cwell with CWP X Graphics
demo=8
x1beg=0 x1end=6500 x2beg=-100 x2end=10000
cwell |
cshotplot > demo${demo}plot outpar=demo${demo}par
xgraph < demo${demo}plot par=demo${demo}par style=seismic \
-geometry 600x400+650-400 \
title="Direct+Primaries" \
label1="Depth (ft)" label2="Range (ft)" \
x1beg=$x1beg x1end=$x1end x2beg=$x2beg x2end=$x2end &
cshot2
suaddhead < demo${demo}traces ftn=1 ns=501 |
sushw key=dt a=4000 |
sugain gagc=1 wagc=0.5 |
suxwigb title="Crosswell Data (agc)" \
ybox=70 \
label1="Time in Seconds" label2="Trace" &
The plotting (model, rays etc) has been turned off for this demo. Data is generated for 10 shots in a well (file swell3). Direct wave and primaries are calculated.
syncline :model file
4 :#interfaces in model
plotcolors :model colors file
:first plot descriptor (mq)
rwell :receiver well coordinates
:receiver plot descriptor (wgq)
d :shooting mode (dg)
swell3 :source well coordinates
don't care :file containing coords. of sources
:source plot descriptor (wsq)
t :job descriptor (rlt)
demo9 :output file name
-90. 90. :range of takeoff angles
1. .1 3. :coarse, fine angle increment; ray density
4000. 6000. 7000. 7500. 12000. :velocities
y :direct wave?
y :primaries?
n :head waves?
PARAM2 file:
s :sort option (s,r) 1 10 :first, last shot for sort 1 20 :first, last trace; or first, last receiver 10. 20. 30. 40. :frequency spectrum of wavelet .200 :wavelet length (secs) .004 :sample rate (secs) 2. :record length (secs) demo9shot :input file demo9traces :output tracesXcwell file for Demo09
#! /bin/sh
# Run cwell with CWP X Graphics
demo=9
cwell
cshot2
suaddhead < demo${demo}traces ftn=1 ns=501 |
sushw key=dt a=4000 |
sugain gagc=1 wagc=0.5 |
suxwigb title="Crosswell Data (agc)" \
ybox=70 xbox=650\
label1="Time in Seconds" label2="Trace" &
Here we have asked for head waves from all interfaces (Record 18). Also, an easier model has been specified (Record 1), and the velocities have been adjusted.
smoothmodel :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell1 :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell4 :source well coordinates don't care :file containing coords. of sources ws :source plot descriptor (wsq) rt :r to plot rays demo10 :output file name 0. 90. :range of takeoff angles 1. .1 2. :coarse, fine angle increment; ray density 4000. 6000. 8000. 10000. 15000. :velocities n :direct wave? n :primaries? y :head waves?Xcwell file:
#! /bin/sh
# Run cwell with CWP X Graphics
demo=10
x1beg=0 x1end=6500 x2beg=-100 x2end=10000
cwell |
cshotplot > demo${demo}plot outpar=demo${demo}par
xgraph < demo${demo}plot par=demo${demo}par style=seismic \
-geometry 600x400+0-70 \
title="Head Wave Rays" \
label1="Depth (ft)" label2="Range (ft)" \
x1beg=$x1beg x1end=$x1end x2beg=$x2beg x2end=$x2end &
cshot2
suaddhead < demo${demo}traces ftn=1 ns=501 |
sushw key=dt a=4000 |
sugain gagc=1 wagc=1.0 |
suxwigb title="Head Wave Data (agc)" \
ybox=20 xbox=650 \
label1="Time in Seconds" label2="Trace" &
PARAM2 file:
s :sort option (s,r) 1 1 :first, last shot for sort 1 20 :first, last trace; or first, last receiver 10. 20. 30. 40. :frequency spectrum of wavelet .200 :wavelet length (secs) .004 :sample rate (secs) 2. :record length (secs) demo10shot :input file demo10traces :output tracesand finally the swell4 file:
1000. :x-coord. of top of source well 2000. 6000. 1. -99999. :end of source well definition 100. :depth to first source 1 500. :#sources, source spacingnot so interesting.
Example smoothmodel file
0. 0. :start of upper surface 10000. 0. 1. -99999. :end of upper surface 0. 1100. :start of interface 1 1000. 1100. 5000. 1200. 8000. 900. 10000. 700. 1. -99999. :end of interface 1 0. 2100. :start of interface 2 2000. 2100. 5000. 2200. 8000. 1700. 10000. 1400. 1. -99999. :end of interface 2 0. 3100. :start of interface 3 2000. 3100.
Records 19 and 20 specify multiple bounce head waves. The letter h in Column 1 defines these events as head waves. The integers that follow are Interface numbers where reflections occur, and the refractor number. Record 19 describes rays that leave the source, reflect from Interface 2, then reflect from Interface 1, then intersect Interface 2 (the refractor) at the critical angle, thus becoming head waves. So, the last integer in the list specifies the refractor. Record 20 describes rays that reflect from Interface 4 before meeting a refractor (Interface 3 in this case). Notice how the velocities allow these refractions to occur in this model. (If you ask for a lot of bounces, you might find that no rays are found. This is often because the offset between source and receiver well is not large enough. Multiple bounces require larger offsets or the rays never go critical at the refractor.) Note that the refracted rays from Interface 2 go up, while the refracted rays from Interface 3 go down, in this example. This explains the move-out on the shot record.
smoothmodel :model file 4 :#interfaces in model plotcolors :model colors file m :first plot descriptor (mq) rwell1 :receiver well coordinates wg :receiver plot descriptor (wgq) d :shooting mode (dg) swell4 :source well coordinates don't care :file containing coords. of sources ws :source plot descriptor (wsq) rt :r to plot rays demo11 :output file name 0. 90. :range of takeoff angles 1. .1 2. :coarse, fine angle increment; ray density 4000. 6000. 10000. 6000. 20000. :velocities n :direct wave? n :primaries? n :head waves? h 2 1 2 :multiple bounce head wave h 4 3 :multiple bounce head waveXcwell file:
#! /bin/sh
# Run cwell with CWP X Graphics
demo=11
x1beg=0 x1end=6500 x2beg=-100 x2end=10000
cwell |
cshotplot > demo${demo}plot outpar=demo${demo}par
xgraph < demo${demo}plot par=demo${demo}par style=seismic \
-geometry 600x400+650-70 \
title="Head Wave Rays" \
label1="Depth (ft)" label2="Range (ft)" \
x1beg=$x1beg x1end=$x1end x2beg=$x2beg x2end=$x2end &
cshot2
suaddhead < demo${demo}traces ftn=1 ns=501 |
sushw key=dt a=4000 |
sugain gagc=1 wagc=1.0 |
suxwigb title="Head Wave Data (agc)" \
ybox=70 \
label1="Time in Seconds" label2="Trace" &
and the PARAM2 file:
s :sort option (s,r) 1 1 :first, last shot for sort 1 20 :first, last trace; or first, last receiver 10. 20. 30. 40. :frequency spectrum of wavelet .200 :wavelet length (secs) .004 :sample rate (secs) 2. :record length (secs) demo11shot :input file demo11traces :output traces
There is also documentation for the CSHOT programs cshot1 and cshot2. These
live in $CWPROOT/src/Fortran/Cshot.
© William P. Clement 2000