Kring KNGMG Noord: Lezing Februari – Correlation and interpolation in subsurface geology

De lezing heeft als ondwerp:
Correlation and interpolation in subsurface geology by Jan Lutgert.

Subject:        Application of K-Means Cluster Analysis and Linear Discriminant Analysis for electrofacies analysis from well logs:
                      discussion and demonstration of two common statistical techniques used to discover structure in large multivariate
petrophysical datasets.
                      With practical examples using public well data from The Netherlands.
Presenter:    Jan Lutgert (petrophysicist, ex-TNO, ex-EBN)
Time:            Tuesday 7th February 2023, 17:00-18:00.
The meeting room will be open from 16:30 onwards. Soft drinks and some snacks will be available.
Place:           NAM office, Schepersmaat 2, 9405 TA Assen: Gele Kamer (“Yellow Room”). 
 
Abstract:    
When it comes to characterizing and modeling reservoirs we never have enough data. And usually  most of our data is scattered over a large subsurface volume, with lots of detailed, high resolution information at the well locations and much lower resolution seismic information to guide us over the inter-well areas and beyond. So, for geoscientists, linking the information from these two scale domains has always been a common goal. Unfortunately there are as many ways to persevere these goals as there are geoscientists. Also, many of the existing software implementations require an intimate knowledge of statistics while many of the users only have a limited knowledge on the working assumptions. So where do we start?
In this presentation we investigate two of the most commonly applied statistical techniques used in multivariate classification: K-Means Clustering (K-Means) and Linear Discriminant Analysis (LDA). The K-Means technique mostly is used to group data that is ‘similar’ in some sense. In this situation it is used to define reservoir intervals sharing some typical petrophysical characteristics. Once the K-Means clustering algorithm has derived a classification scheme, LDA can be used to test the validity and robustness of the classification and can also be used to propagate the classification to other wells. Practical examples will show the strengths and limitations of this workflow and suggest some  improvements. During the discussion it also will be argued that robust, fully automated electrofacies classification sometimes may be beyond what the data and software allow…

About our Speaker:
Jan Lutgert worked for TNO as a geoscientist from 1987 until 2010. Assignments involved geological and petrophysical R&D studies, advisory work for the Ministry of Economic Affairs, and international consultancy. From 2010 to 2018 he worked for Energie Beheer Nederland (EBN), dominantly focusing on and hydrocarbon exploration, the de-risking of tight reservoirs and reservoir unitizations. He retired in 2018 and currently enjoys parttime lecturing petrophysics at the Vrije Universiteit of Amsterdam and the Anton de Kom University in Surinam.
—————

Als KNGMG-lid bent u van harte welkom om onze lezingen bij te wonen! Als u geïnteresseerd bent, graag van tevoren aanmelden d.m.v. een e-mail naar kngmgnoord@gmail.com

 

Datum

feb 07 2023
Expired!

Tijd

16:30 - 18:00

Locatie

NAM, Assen
Schepersmaat 2
mei 2024
Geen evenementen gevonden!