[Users] fluid-effect-in-channel

[Friso Brouwer]

Dear Abdhesh,

Assessing fluid effects is complex matters and there is no one answer to 
what is the right procedure, it all depends on factors as AVO class of 
target, which frequency effects are present and do they have a causal 
relation to HC saturation, what kind of data quality do you have. The 
best procedure to go ahead is to start with the normal HC indicators and 
try to develop attributes for each that work in the dataset under 
consideration. For example:

*Bright spot/dim spot *(depending on AVO class):
use "energy attribute".

*AVO*:
"energy attribute" applied on the far stack and/or ratio of "energy far 
stack" divide by "energy near stack" using "mathematics attribute".

*Polarity reversal (depending on AVO class): *
Assuming the horizon you work with is the "top sand" and consistently 
picked on a peak. If there is a polarity reversal the gas (or light oil) 
section is characterized by a strong through above this peak and weak 
negative side lobe below the peak, while the none gas section is 
characterized by a weak negative sidelobe above the peak and stronger 
through below the peak. Using the "volume statistics attribute" with 
appropriate time windows one can extract the minimum values in a window 
above and below the horizon. Divide the MinimumAbove and MinimumBelow 
using the "mathematics attribute" will give you a ratio between upper 
and lower minimum and reveal where phase reversal takes place. Ofcourse 
this assumes all else (thickness, embedding lithology, ...) remains 
constant. Variants of this attribute should be used when data is 
European polarity  (through is AI increase), the horizon is interpreted 
on zero-crossing etc.  Another use of the same kind of attribute (in AVO 
class 2b or 3) is that it can be used to separate bright spots from hard 
streaks.

*Frequency effects:*
The use of frequency effects to explore for HC is a widely debated 
subject, with no consensus on how to handle this in a unified way, but 
the general idea is that HC are associated with a bias to lower 
frequency. Among the many ways to extract frequency attributes in 
OpendTect I will explain two:
/Simple but robust:/ Look at the amplitudes of the low frequencies using 
a "CWT" or "Frequency filter" -> "energy attribute". This will work only 
in a bright spot regime and emphasize area where there is and a 
frequency bias to low frequency and overall increase in reflection 
coefficient, which both can be related to HC presence.
/More complex:/ Use bandpass filters to isolate parts of the spectrum 
well below and well above the dominant frequency using the "frequency 
filter attribute" (for example Fdom=30 take LP  <15hz and HP >45hz). 
Take the energy of each frequency band. Divide the EnergyOfLP by 
EnergyOf HP. This emphasizes areas where the spectrum is biased to the 
low pass and is due to the normalization in the division independent of 
amplitudes (measures only frequency effect). This attribute has some 
pittfalls, among other the choice of extraction windows in relation to 
the horizon.

*Flatness: *
The OpendTect volume viewer is a good tool to verify suspected flatspots.

*QC:*
Major QC's  1) structural conformance: attributes should highlight the 
(complete) updip parts of the channels, unless you have a goods reason 
to assume exceptions such as compartments in the channel, or updip 
decrease in reservoir quality or thickness. 2) (if pressure 
communication) each HC indicator should have a single downdip limit 3) 
all valid HC attribute should have mutually consistent lower limit 
cut-offs,

*Using well information and neural networks:*
if you have multiple well penetrations in the channel (minimal one wet 
and one HC) you can use a neural network trained using well calibration 
to predict HC phase. Input attribute can be the one described above 
and/or seismic waveform. You can use OpendTect default attribute set 
"Unsupervised Segmentation 2D" as a shortcut to create waveform attributes.

*Before applying attributes analyze and acknowledge factors as these:*
seismic data quality: does the data allow meaningful analysis
seismic data processing: e.g. agc, adaptive spectral compensation and 
other cosmetics good for horizon tracking can be fatal for attributes, 
when using partial offset/angle stack, be sure you know how they are 
extracted and compensated for fold variations per depth (especially if 
your horizon has significant structural relief).
geology: are there carbonates or coals which have significant amplitude 
responses, can your channel be clayfilled (and how to detect that, 
attributes will work for that to), are there gas chimneys - gas chimney 
are a useful exploration tool themselves, but will skew frequency 
response below the gas chimney.
rock properties: which AVO class is my channel.
fluid properties: a heavy oil would give you another (or none) AVO 
response compared to gas or high GOR oil.
geophysics: is the target very shallow, or very deep. What are the 
implications for my attributes (intrinsic attenuation, angle of 
incidence, mutes and folds)
attributes: using the above factors decide what attributes to apply and 
assign confidence level to each of them.

I hope this helps you out.

-- 
Friso GC Brouwer
Chief Geophysicist

dGB-USA
One Sugar Creek Center Boulevard
Suite 935
Sugar Land, TX, 77478

Tel 281 240-3939 (main)
Tel 281 240-6957 (direct)
Fax 281 240-3944
e-mail friso.brouwer at dgb-group.com
http://www.dgb-group.com






Abdhesh Kr. Upadhyay wrote:
>
> Hi All,
>
>  
>
> I'm using CWT ( continuous wavelet transform) attribute in OpendTect. 
> CWT -- spectral decomposition technique delineating channel very 
> clearly but I need some suggestions on it, "How can I see the fluid 
> effect in the delineated channel??"
>
> You advise will be very helpful.......
>
>  
>
> Thanks & regards,
>
> Abdhesh
>
>  
>
>  
>
>  
>
> Abdhesh Kr. Upadhyay
>
> Lead Geophysicist
>
> Cairn Energy India Pty. Ltd.
>
> Phone: +91 124 4141360 Extn: 3353 Fax: +91 124 2889320
> Direct: +91 124 2703353 Mobile: +91 9910486059
>
>  
>
>
> This email contains privileged and/or confidential information and is 
> meant for the named recipient(s) only. Please be aware that if you 
> enter into email correspondence with Cairn India Limited, its 
> subsidiary Cairn Energy India Pty Limited or any of its other 
> subsidiaries (collectively referred to as the Cairn India Group) this 
> communication may be subject to monitoring and all emails will be 
> stored in archives. If you are not a named recipient you should 
> neither copy nor distribute this email nor disclose its contents to 
> any person. If you have received this email in error, please notify 
> the sender immediately by way of a return email or telephone +91 124 
> 414 1360 and then delete it from your system. The views, opinions, and 
> judgments expressed in this email are solely those of the author. The 
> contents of this email have not been reviewed or approved by any 
> company of the Cairn India Group. This email and any attachments are 
> not guaranteed to be free from computer viruses and it is recommended 
> that you check for such viruses before downloading it to your computer 
> equipment.
> ------------------------------------------------------------------------
>
> _______________________________________________
> Users mailing list
> Users at opendtect.org
> http://lists.opendtect.org/mailman/listinfo/users
>   


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opendtect.org/pipermail/users/attachments/20080304/5b15a2eb/attachment.html>