Integrated characterisation of mud-rich overburden sediment sequences using limited log and seismic data: Application to seal risk

Abstract

Muds and mudstones are the most abundant sediments in sedimentary basins and can
control fluid migration and pressure. In petroleum systems, they can also act as source,
reservoir or seal rocks. More recently, the sealing properties of mudstones have been
used for nuclear waste storage and geological CO2 sequestration. Despite the growing
importance of mudstones, their geological modelling is poorly understood and clear
quantitative studies are needed to address 3D lithology and flow properties distribution
within these sediments. The key issues in this respect are the high degree of
heterogeneity in mudstones and the alteration of lithology and flow properties with time
and depth. In addition, there are often very limited field data (log and seismic), with
lower quality within these sediments, which makes the common geostatistical modelling
practices ineffective.
In this study we assess/capture quantitatively the flow-important characteristics of
heterogeneous mud-rich sequences based on limited conventional log and post-stack
seismic data in a deep offshore West African case study. Additionally, we develop a
practical technique of log-seismic integration at the cross-well scale to translate 3D
seismic attributes into lithology probabilities. The final products are probabilistic
multiattribute transforms at different resolutions which allow prediction of lithologies
away from wells while keeping the important sub-seismic stratigraphic and structural
flow features. As a key result, we introduced a seismically-driven risk attribute (so-called
Seal Risk Factor "SRF") which showed robust correspondence to the lithologies
within the seismic volume. High seismic SRFs were often a good approximation for
volumes containing a higher percentage of coarser-grained and distorted sediments, and
vice versa.
We believe that this is the first attempt at quantitative, integrated characterisation of
mud-rich overburden sediment sequences using log and seismic data. Its application on
modern seismic surveys can save days of processing/mapping time and can reduce
exploration risk by basing decisions on seal texture and lithology probabilities.