Use Java 8: All codes have been rebuilt with java 8, so update your JVM if you aren't there yet
Converted-waves: The download section now has a link to the Chen/Glinsky converted-wave extension of Delivery
First public release of time-lapse Delivery code(Delivery4D). See the Delivery4D Paper, plus code tarball available from the Download link. This subsumes all the functionality of the standard Delivery, so you can do all your standard single-vintage inversions using this. Time--lapse saturation effects modelled via Gassmann, and stress effects from linearised elasticity. Now with arbitrary numbers of stacks, and possibly PS data imaged in PP time.
First public release of 1D CSEM code (DeliveryCSEM). See the DeliveryCSEM Paper, plus code tarball available from the Download link. Snapshot below. It's a longish paper, so charge your coffee.
Delivery is an open-source, trace-based Bayesian seismic inversion code for use in oil reservoir characterisation at the early development and appraisal stage. See the Computers and Geosciences paper for most of the details, and the Delivery4D preliminary writeup for the time-lapse modelling.
WaveletExtractor is a well-tie and wavelet extraction program bundled with Delivery. It can extract optimal wavelets for use in delivery from seismic, log data and checkshot (time-to-depth) information. It works on 3D,2D or single trace seismic, deviated wells, and multistack seismic. It can produce uncertainty estimates for wavelets, time-to-depth information, and noise level estimates for seismic inversion. See the Computers and Geosciences paper for details.
The DeliveryMassager code is a tool for moving the stochastic inversion results coming out of Delivery over to the cornerpoint grids (Eclipse-style) typically used in geomodelling packages like Petrel and Gocad. The code can integrate inter-layer and inter-property correlation structures induced by the seismic data with transverse correlation requirements coming from geological knowledge and hard well data from logs. Outputs include most-likely cormerpoint models, stochastic cornerpoint models, and on-the-fly volumetric statistics. See the Computers and Geosciences paper for details.
- Will operate best in an environment with Seismic Unix installed, and BHP extensions.
- Delivery and WaveletExtractor require java 1.5 or later.
- DeliveryMassager code is a native C code and will require compilation: i.e. a compiler & a few standard packages. Builds fine on Linux, MacOSX, and almost certainly cygwin (POSIX environment for Windows). See the relevant READMEs.
DeliveryCSEM is an open-source code for performing 1D CSEM inversion with various Bayesian flavours. The code is mainly java, with native kernels for the forward modelling and derivatives. See the paper for much of the gory detail. The code can be used for forward modelling, inversions of various flavours, and parameter uncertainty estimates via MCMC algorithms or a Bayesianised bootstrap approach.
- Delivery CSEM is a multilanguage code, chiefly java, but using also layers of C and fortran. It has been tested on Linux, MacOSX, and Windows with mingw+cygwin compilers. Read the README* files carefully, especially if rebuilding.
Major new code functionality for handling second impermeable rock types in Delivery, plus new ways to manage fluid probabilities on input. See appendices 14 and 15 in the Delivery Paper
Major fluid-contact functionality added to Delivery. See Appendix 10 in the Delivery Paper. This models reflections from depth-specified fluid contact within layers. Also: update your schema files: noise levels now use the tag
Aug 2006: DeliveryMassager code and paper added!
July 2006: Modified to use Java 1.5's faster formatted i/o, and probably generics. Users should install Java 1.5 or later.
Mar 2006: many bugfixes etc to all codes. Updated pdfs for delivery and the wavelet extractor. Delivery-to-ECLIPSE DeliveryMassager code coming soon!
Mar 2004: Full wavelet extractor code and documentation available.
Oct 2003: Latest downloads contain most up-to-date XML editor, with new schema and help.