Trust your 3D model: Three ways to avoid an expensive exploration mistake
By Barry de Wet & Nick Williams
3D models play a critical role in modern mineral exploration programs, but not all 3D models are created equal: unreliable models lead to poor drill targets and missed opportunities for explorers.
Today, 3D models are easily generated and found everywhere from conference presentations to annual reports and on every exploration company webpage. But too many explorers rush into choosing drill targets without asking the important questions first.
Is this model reliable?
3D models need to be sufficiently accurate to justify spending exploration dollars on a drilling campaign.
Virtually all geological and geophysical datasets, observations or measurements can be converted into a 3D model. Each observation is extrapolated through a volume based on known trends, processes or physics to produce a 3D model of cells or points with specific inferred properties.
Each step of the process has uncertainties that propagate through the resulting 3D model. Every geological and geophysical observation has some uncertainty or measurement error.
The processes used to extrapolate parameters through a 3D model are based on specific sets of assumptions or numerical models that may or may not be correct. Nearly all direct observations are made at locations on the surface or along very skinny drill holes. Unless most of the subsurface is directly sampled, a complete 3D model is impossible.
There are many critical modelling parameters that effect the reliability of drill targets selected from a 3D model. For example, most 3D models are discretised into cells or blocks of specific sizes.. Larger cell sizes make computation easier, but provide a more approximate and inaccurate representation of the earth, increasing the error in the model and making it more difficult to accurately position a drill hole.
So what's required to ensure we can trust our 3D models to help make expensive business decisions?
Know your data
This is possibly the most important point: know your data.
What is being measured? How was it measured? What are the limitations of the measuring tools? Can we separate the contributions of the geological targets of interest from other masking effects? How do we minimize the contributions of unimportant features? Does the data add to our existing understanding of the problem?
In gravity surveys some of the biggest data contributions come from topography, elevation and latitude, not from the target of interest.
In magnetic surveys, the observed magnetic response is an intimate mixture of two completely independent magnetic properties and only skilled analysts can separate those contributions to understand the true source features.
Electrical and electromagnetic datasets have even more potential masking agents, with contributions from the survey design itself, younger cover and overburden, groundwater and atmospheric effects. The target signal of interest can be very difficult to isolate.
For example, our Typhoon technology provides an extremely clean and accurate signal, minimising measurement error. However, a survey design process takes care to minimise the contributions of unwanted factors, minimising the potential sources for error in our models.
Insider tip: Prepare you data set to isolate the important features and minimise the contributions of unwanted factors.
Generating realistic models from infinite possibilities
In medical imaging, the MRI and CT scanners have access to all sides of a patient, however in mineral exploration we only have access to one side, the surface, and even that is usually covered.
As a result, we must extrapolate a small number of surface measurements over a large volume to model a complex arrangement of rock types, structures, fluids and more at depth. With such limited data it may be possible to generate an infinite number of models that equally explain a dataset.
Realistic models - those that identify the features and structures you are seeking to - are possible by using the best 3D modelling algorithms available for each problem. By including as much extra supporting information as possible we can constrain the modelling to limit the results to a smaller subset of models that are realistic.
Insider tip: To ensure the model identifies the feature you are looking for, intentionally promote features of interest, include all supporting information for your target and minimizing unwanted contributions.
Critique your model with confidence
Ultimately our goal is to accurately drill targets we are confident in by extracting every last drop of value out of the available data.
With an intimate understanding and tight control over the input data, the 3D modelling processes and the assumptions used, you can ensure which areas of the recovered 3D models will have the highest reliability and which areas are more speculative.
Understanding the building process and limitations of 3D models helps us to prevent misdirected drilling and maximise opportunities for success.
Insider tip: Don’t jump in and start picking drill targets until you understand the limitations of your data and how they can manifest in the model. A 3D model is only the best prediction of what is below the surface based on the data used and assumptions made during the modelling.
Work with the best in the business
A lot of work goes into building a reliable 3D model.
HPX and strategic partners at Computational Geosciences Inc. are industry leaders at every step of 3D modelling, from survey design and data collection, through processing and modelling the data, to critical interpretation and understanding of resulting models.
We have applied our techniques successfully to generate all kinds of the 3D geophysical modelling process and integrate them with geological models. In some cases we have drilled prospective targets within 5 m of their predicted position.
Barry de Wet is the renowned and respected former Chief Geophysicist for BHP Billiton and Ivanhoe Australia. As Chief Geophysicist and a Director of HPX, he leads a team of talented Geophysicists deploying HPX proprietary technology and applications on projects around the world.
Nick Williams produces 3D models every day, even on the weekend. He loves physics and pushing the boundaries to develop advanced 3D geophysical strategies for exploration. He specialises in integrating geological and geophysical models and data in 3D to find buried mineral deposits.