Unlocking a century of subsurface intelligence

Drill cuttings are an invaluable source of organic and inorganic geological data. Yet, they are often underutilised, with hundreds of thousands of cuttings stored in warehouses and repositories across the globe, spanning decades of drilling activity. In Malaysia alone, PETRONAS, through Malaysia Petroleum Management (MPM), is storing more than 6,000,000 cuttings and core samples from across Malaysia, covering some 100 years of exploration and development.

With the onset of new analytical technologies and AI-driven software applications, all these samples can generate high-quality, consistently generated digital datasets that will assist in our detailed understanding of subsurface plays at a fraction of the cost of drilling a new exploration well. Standard techniques are now available to correctly and accurately depth-match fresh and historic samples to wireline log suites.

Figure 1 outlines a cuttings sample digitisation workflow that is currently being utilised for the digitisation of the MPM cuttings data repository in Kuala Lumpur, Malaysia. The key to unlocking this digital data revolution from cuttings is to understand how to properly and consistently prepare each sample for analysis, and which types of analyses will yield valuable and correlatable data. Consistent sample preparation is key when trying to interpret and correlate data from samples of different vintages acquired by different vendors. Understanding mineralogy, potential organic richness and mud type are some of the important factors to consider when making sure a sample is properly prepared. In addition, correct washing and drying procedures are fundamental for preserving mineral species and evidence of hydrocarbons.

Once correctly washed and dried, a base analytical workflow should be undertaken to ensure that every sample is consistently measured, thus allowing for confident correlations sample to sample, well to well, and both vertically and laterally across the basin / play / asset. Typically, this base workflow should consist of white light (WL) and ultraviolet light (UV) quantitative digital imaging, color extraction (Red-Green-Blue (RGB) and Brightness), particle size and shape segmentation and elemental data analysis.

Beyond this base workflow, on a caseby- case basis, additional valuable data can be obtained, such as mineralogical content, organic richness and maturity and biostratigraphic data. All these collected digital data can then be used to help constrain rock physics models or as input to regional or field subsurface 3D models by creating accurate image analysis and elementally constrained lithofacies models.

The Sotong gas field case study

The Sotong field lies to the NE of the Tenggol Arch, offshore east of Peninsular Malaysia in the South China Sea. It is located down-dip from other hydrocarbon- producing fields such as Tembakau and contains reservoir sections characterised by stacked fluvial channelised facies, tidal flats and shoreface complexes. Analysed cuttings data as part of the MPM-led digital transformation initiative were used to build a new lithofacies model for the field, thus offering a new and greater insight into the subsurface complexities.

Figure 2 cross plot shows an image analysis based on brightness derived from Red-Green-Blue colour parameters versus key elemental data (log Si/Al ratio) generated from Sotong field drill cuttings samples. This cross plot forms the basis of a new lithotype model with cut-offs defined as Brightness > or <80 and log Si/Al > or < 3.8, leading to the creation of four newly defined and quantitative lithotype families:

1. Clean Sandstone: Brightness >80; log Si/Al >3.8
2. Heterolithic Sandstone: Brightness >80; log Si/Al <3.8
3. Clay-dominated Silt/Sandstone: Brightness <80: log Si/Al <3.8
4. Heterolithic Muddy Siltstone: Brightness <80; log Si/Al > 3.8

As Brightness is a known facies indicator in clastic sediments, each of the four Lithotype Families was subdivided into Lithotype Associations based upon the individual cuttings sample Brightness readings, e.g. for the Clean Sandstone Family, Ss1, Ss2-Ss5 etc. and then further subdivided into individual lithotypes.

These interpretations were further cross-checked against log data (Gamma Ray (GR), Neutron / Density and Deep Resistivity) and other elemental and elemental ratio data such as Si%, Al%, Zr/Nb and Si/Ca for sedimentological context. Once complete, the newly developed Lithotype Scheme was used to identify small to medium-scale cyclicity in the form of darkening upwards and brightening upwards cycles, which stack to form a series of newly developed field-wide correlatable reservoir layers.

These models can then be used to define cuttings-based resolution downhole interpretations on a well-by-well basis. Figure 3 is an example of such a downhole interpretation for one well from the Sotong field. From the lefthand side, tracks 1 to 6 show depth (ft, MD) stratigraphy and log data, including the computed gamma ray calculated from the K (%) Th (ppm) and U (ppm) elemental data. Tracks 7 to 11, outlined in red, show the downhole detail associated with the lithotype model, including interpretation. Tracks 12 to 18 show the colour extraction metadata and interpreted brightness calculation from the image analysis data. Finally, tracks 19 to 36 show selected elemental and elemental ratio data extracted from the XRF analysis, including the Si/Al ratio (track 31) and other elemental data used to support the interpretation (e.g. U/Th ratio, Ti/Si ratio, Rb (ppm)).

The detail associated with the Lithotype interpretation (outlined in red in Figure 3) clearly shows cyclicity at different scales from small scale (10 – 20 ft) darkening upwards cycles shown in Track 7 (Lithotype Code) to medium scale (50 – 100 ft) generally darkening upwards cycles shown in Track 10 (Gross Cyclicity) and finally to large scale cycles depicted as Reservoir Layers and shown in Track 11. Reservoir Layers C.1 to C.3 are dominated by the Clean Sandstone Lithotype Family (high brightness coupled with high log Si/Al ratio) with minor Heterolithic Muddy Siltstone and Clay-dominated Silt / Sandstone. These layers are associated with a dominantly fluvial depositional setting. Above this, Reservoir Layer D.1 probably represents a transition zone, grading upwards into a more Clay-dominated Silt / Sandstone facies with lower brightness and lower log Si/Al ratio. Associated with this layer is a decrease in U/Th ratio, an increase in Ti/Si ratio and an increase in Rb (ppm). The upper section of the well (upper Lower Miocene to lower Upper Miocene) is dominated by the Clay-dominated Silt / Sandstone facies with low brightness, low log Si/Al ratio, higher but stable Rb (ppm) and upwards increasing P (%) and probably represents a more quiescent marginal marine depositional environment.

Once applied to other Sotong field wells, this newly developed rock-based Lithotype Model will greatly assist with the updated, spatial subsurface understanding of this play and will create rock-constrained input into the updated 3D reservoir model for this field, which is due to be completed during 2027.

An inexpensive but powerful tool

Understanding subsurface plays typically requires a substantial initial investment (e.g. seismic and log data) and long analysis times at multiple labs. However, the datasets generated by the Digital Cuttings project, coupled with the added-value and consistently measured rock-based workflow presented here, offers increased efficiencies by providing a cost-effective and rapid method to collect, analyse, define, model and predict key reservoir drivers. The ability to identify elemental and image analyses specific lithotypes from routinely collected fresh and legacy drill cuttings will allow operators to better predict across-field spatial reservoir definition and understanding, thus improving subsurface model-build and overall field development. The workflow example for the Sotong field presented here can easily be utilised for all other fields and plays within PETRONAS’ current and future portfolio.

Drill cuttings can be a relatively inexpensive but incredibly valuable dataset that could help to unlock the future potential of our energy plays for years to come. As an industry, we need to make the most of these datasets by digitally unlocking their data potential and having the confidence to use these data in our subsurface evaluations. Reliable, accurate and consistent data is the key to unlocking or confirming regional and field-specific reservoir / seal trends. With large data centers containing enormous compute power now readily available to all, having access to these vast ‘integraiton / AI-ready’ consistently measured datasets will revolutionise the way we explore for, or appraise our energy assets.

The authors would like to thank and acknowledge Malaysia Petroleum Management (MPM), PETRONAS, and Geolog for granting permission to publish this work. In addition, the authors would like to thank all our colleagues who have worked tirelessly to generate the data for this publication.