More than just reflections – Unleash the value of your seismic data

By Graeme Eastwood, Fugro-Jason

Geophysicists have struggled for years to get more information from their seismic data than that provided by a simple structural interpretation, which is still the main use of the data today. One of the goals has been direct identification of lithology and estimates of the associated reservoir properties, if possible with associated uncertainties. Typically this has been attempted at a basic level through the use of seismic attributes and seismic character recognition, but there is much more information contained in the seismic data than this – it just takes a different approach to extract it.

There are a number of factors which influence the validity and usefulness of traditional attribute analysis and character recognition. Some of these include:

1. Without knowledge of the underlying wavelet in the seismic data it is difficult to predict the seismic response from well based synthetics
2. The relationships established between reservoir properties and seismic attributes are empirical and are difficult to extend laterally.
3. Attributes are usually based on reflectivity data and relate to layer boundaries rather than to layer properties.

Fugro-Jason has, over the last 20 years,developed a set of tools which when combined with the output from our industry leading inversion processes can provide reliable 3D estimates of reservoir properties based on rigorous integration with well log information and well established geophysical elastic rock property relationships along with a robust way of interpreting them.

Well Tie and Wavelet Estimation

The first step in this workflow (assuming the well data has been edited and corrected for incorporation into a seismic driven workflow) is to generate an appropriate time to depth relationship for each well and use this to get a reliable estimate of the wavelet imbedded in the data. This is actually essential even for traditional character and attribute analysis to ensure that the correct seismic information is correlated to the well data, as shown in Figure 1.

Figure 1 - Well Tie showing seismic data, synthetic seismic, impedance and well data

Having the correct wavelet also allows the wells to be used more effectively to predict the seismic response. The 'wavelet' may actually be changing both vertically and spatially, so any estimation and use must be able to account for this.

Lithology Definition and Estimation of Elastic Properties

The next step in the process is to define a set of discrete lithologies which differentiate between good and bad reservoir based on both the physical rock characteristics such as porosity and shale/clay content as well as fluid saturations. Often these will be the same as geological lithotypes, but it is important not to overcomplicate the situation by trying to extract too much in terms of depositional setting - often lithotypes such as pay sand, wet sand and shale will accomplish the goals of the study.

Once the lithotypes have been defined one needs to understand the elastic properties associated with each lithotype to determine whether seismic data can be used to identify the different lithologies.

This is done by cross-plotting the elastic log data points associated with each lithology. This is first done at the well log scale to determine the most optimistic separation and then repeated with the well data filtered to seismic frequencies to determine the separation than can be expected when using seismic data as the source of information. In actual fact, it is necessary to create a petro-elastic model from the well data which enables the petrophysical data to be used to correctly model the elastic response. The seismic and well data are then effectively tied by this petro-elastic model, which is the cornerstone in allowing interpretation of seismic properties in terms of reservoir properties.

The crossplots can also be used to create probability density functions for each lithotype which can be used to determine the likelihood that a particular combination of elastic parameters represents a sample from that lithotype.

Inversion for Elastic Properties

To use the seismic for lithology estimation it is necessary to derive elastic property volumes (P-Impedance, S-Impedance, Density, etc) from the seismic.  This is typically done by simultaneous inversion of partial offset or angle stacks for estimates of Vp, Vs and density or some combination of these variables. In order to predict properties accurately away from well control, such an inversion process must be driven by the seismic data and not by a model derived from the well data, since it is the seismic data which is sampling the unknown. Obviously any such process must be able to honour the well data within the seismic bandwidth to a high degree.

Fluid and Facies Prediction

Once the elastic property volumes have been created, the probability density functions defined by the well data can be used to determine the probability that a sample from the seismic volume is associated with a particular lithology.

The Jason FFP module provides a simple way to build lithology probability volumes and most likely lithology volumes using probability density functions defined from well data and elastic property volumes generated by seismic inversion.  A priori information regarding the expected proportion of each lithology is used to ensure the probability density functions are weighted appropriately.

QC processes are available to allow for proper calibration of the seismic to the wells.

Body Capture

Jason's Body Check routine is used to indentify connected bodies which meet defined criteria - size, connectivity, etc - and traditionally is used on the output from inversion processes to identify "geobodies" exhibiting the same properties, usually P-Impedance, S-Impedance, Vp/Vs, etc. An example is shown in Figure 2.

Figure 2 - Captured Geobodies colored to show connectivity

Lithology probability volumes and most likely lithology volumes provide an alternate set of input volumes which can be used to identify bodies of a particular most likely lithology or to identify bodies which exceed a probability threshold. For example, capturing all voxels in which there is a greater than 50% chance of sand occurrence.

Geological Model

The final step in the workflow is to incorporate the lithology information derived from the seismic data into geologic models.

Fugro-Jason's EarthModel FT software can be used to generate new, or import existing, geologic models.  Once the models are defined the lithology and reservoir property volumes can be easily upscaled and mapped into the models with complete zonal integrity, ensuring there is no "leakage" of properties into the wrong lithological unit - a common problem with traditional depth resampling.

The lithology probability volumes can also be used indirectly as secondary trends to guide sequential indicator simulation. Alternatively the lithology volumes can be used directly to define lithology within the model. Whilst this has some limitations caused by seismic frequency limits for deterministic inversion results, lithology volumes resulting from geostatistical inversion are not limited by seismic frequencies and, so, exhibit detail compatible with traditional geological models.

Lithology based relationships between elastic parameters and reservoir properties can be used to further define the reservoir properties within the bodies.

ABOUT FUGRO-JASON

Fugro-Jason is dedicated to delivering innovative products and services to help our clients identify and produce hydrocarbon deposits by integrating information from the various geoscience disciplines. In 1993 Fugro-Jason introduced the Jason Geoscience Workbench, making it possible to integrate geological, geophysical, geostatistical, petrophysical and rock physics data information into a single consistent model of the earth. In 2001 Fugro-Jason added the PowerLog family of petrophysical applications, and in 2003 added the EarthModel FT family of geological modeling applications.

The application of Fugro-Jason's technology, either through use of the software or through our consultancy services, substantially improves returns on E&P investments by adding invaluable reservoir model information that reduces the risks, costs and cycle-times associated with field development.

Fugro-Jason is part of the Fugro group of companies.

Graeme Eastwood

Region Manager, Fugro-Jason

c/o Fugro Survey (Middle East) Ltd

PO Box 43088

Abu Dhabi

United Arab Emirates

Tel.: +971 2 554 1011

Fax: +971 2 554 7811

geastwood@fugro-jason.com