This paper presents and discusses an experience based working approach and philosophy for geological modelling. The aim is to encourage a discussion on how to best balance and synergize powerful modelling technologies available today with bright thinking of scientists working for the petroleum industry. This method consists of (1) proper model framing in order to meet business objectives, (2) analysis of a suitable modelling technique to enable effective delivery of results, (3) in-depth uncertainty analysis through a dissection of the subsurface into its individual components, (4) selection of appropriate range of subsurface scenarios which cover uncertainties and risks (5) a variety of manipulative tools which can help to implement subsurface concepts which cannot easily be addressed with geostatistics and (6) a threefold tactic ("self-criticism" - "supervision" - "peer review") to facilitate quality control.
This paper provides some nontechnical and technical methods which may help to achieve static geological modelling results faster and with a stronger link to business objectives.
The challenge today is not to fall prey to the "menace of the black box" by over-using powerful computer programs. Now-adays, with statistical algorithms being quite advanced, different models can be run in a very short time and the temptation is large to mechanically advance pushing buttons beyond the objectives and the aim of the modelling exercise planned. Modelling should not be chasing a "holy grail", but use the computing power efficiently and in a fit for purpose way to improve subsurface understanding, to plan and execute development projects in a safe and successful manner and to keep geological information and interpretation understandable and auditable.
Before the start of every modelling exercise a generic framing exercise should be organized. Framing the planned modelling effort is aimed at (1) ensuring the modelling exercise is set on the right track from the start (2) avoiding misunderstanding between different parties involved and (3) to identify gaps in input data/concepts, manpower/funding, technical infrastructure available.