The goal of the timed-trial was to create a real-life scenario that E&P operators may face in the near-future as the industry’s seasoned geologists begin to retire, or simply decide not to return to the industry. To understand how an early- to mid-career geologist might go about creating a new reservoir model for an unfamiliar field, we invited a geologist with 12 years’ experience working in field development across several oil companies (IOC and independent) to utilize a dataset from the Crux Field consisting of 5 wells (4 with reservoir penetrations), full log suites for each well, drilling reports, checkshot data and 270 km2 3D seismic data to generate a new facies model from scratch. The geologist had no prior experience or knowledge of the Crux Field.
From this dataset, the participant was asked to create an analogue-based reservoir model in Petrel* E&P software platform under two different scenarios, both of which would be timed. The comparison sought to understand the differences between the two workflows in terms of time and model robustness.
In the first scenario, the subject was provided only the dataset and a computer with access to the Internet and a license of Petrel installed. In the second scenario, the subject was provided the exact same dataset and computer, only this time the participant had access to Ava Clastics and was asked to utilise the cloud-based application to perform their work. In both scenarios, the geologist was asked to document the process.
THE RESULTS
Methodology
Phase 1
Phase 2
Phase 3
Total
Number of Models Produced
Ava Clastics
4 hours
15 hours
1.5 hours
20.5 hours
2
Standard Practice
4 hours
36 hours
16 hours
52 hours
1
Utilising Ava Clastics, the geologist was able to quickly gain relevant analogue data, automatically quantifying the dimensions of the facies objects in the Depositional Concept for the study area. This saves the time required to source the relevant papers and extract the required information for building an object based facies model. Also the time taken to perform variogram analysis is drastically reduced using Ava Clastics. Ava Clastics uses analogue data from the FAKTS database, together with published geostatistical methods (Ritzi 2001) to inform the variogram ranges, meaning that the process of selecting an appropriate analogue not only populates the parameters for object modelling, but the variograms as well.
In addition, information regarding wavelength and amplitude was also available when using the analogue database. This helped the geologist create a more robust model, because without this information from Ava Clastics, the Petrel defaults would have be applied due to the difficulty sourcing the required data. The downside of using these standard defaults is that their origin is unknown and their appropriateness to the model can be questioned. When a geologist is asked to defend their decisions, using defaults is a less reliable answer than citing the research, technical papers and resources backing their data input. The audit trail captured in Ava Clastics helps the team understand the origin of the model and the changes over time.
*Mark of Schlumberger
| Methodology | Phase 1 | Phase 2 | Phase 3 | Total | Number of Models Produced |
|---|---|---|---|---|---|
Ava Clastics |
4 hours |
15 hours |
1.5 hours |
20.5 hours |
2 |
Standard Practice |
4 hours |
36 hours |
16 hours |
52 hours |
1 |
Case Study (Extended Version) |
|
Each to their Own - the bias of analogues |
In this whitepaper you will see outcome of what happened when several geologists were asked to interpret between gamma ray and sedimentary logs (Fig. 3b). The panel contains two logs over a 150m section. |
Ready for a Free Trial? |
|