Automated facies classification from well logs
Facies of a formation are classified from data charactering properties of a portion of the formation as a function of depth, wherein the number of facies is determined automatically in an unsupervised manner without human input. In one embodiment, a layer-based methodology is provided that performs facies classification based on layer-based properties which are determined from well log data obtained from a plurality of different well logging tools. In another embodiment, a depth-based methodology is provided that performs facies classification based on well log data obtained depth-by-depth from a plurality of different well logging tools. The number of facies can be determined automatically without human input, for example using the Bayesian Information Criterion or a method which determines the optimal number of clusters based on the repeatability of the clustering results. In embodiments, the facies classification can be performed using the Gaussian mixture model (GMM) method.
1. A method of classifying facies of a geological formation, the method comprising:
a) generating data characterizing properties of a portion of the geological formation as a function of depth based on well log data obtained from a plurality of different well logging tools, the well log data comprising a plurality of images having varying resolution, the well log data being part of a layer model that includes a number of layers each associated with a vector of properties of the portion of the geological formation and layer boundaries, the generating the data characterizing the properties of the portion of the geological formation as the function of depth based on the well log data obtained from the plurality of different well logging tools comprising:
identifying the layer boundaries from the well log data;
generating the layer model based on the identified layer boundaries and the well log data;
determining a respective resolution of facies classification of each of the well logging tools that obtained the plurality of images of the well log data at the varying resolution;
normalizing the well log data based on the determined resolution of the facies classification of the well logging tools that obtained the well log data; and
determining the properties of the vectors of the layer model by assigning weights to appropriately scale the normalized well log data obtained from the plurality of different well logging tools based on the resolution of the facies classification of the well logging tools that obtained the plurality of images of the well log data of the varying resolution such that corresponding well log data for each of the well logging tools is scaled to a highest resolution among the respective resolutions of the well logging tools;
b) processing the data of a) to automatically classify or identify a set of facies for the portion of the geological formation including automatically determining, without human input, a number of facies in the set;
c) automatically generating a two-dimensional image of the portion of the geological formation as a function of depth, the two-dimensional image comprising a plurality of layer lines each having a thickness corresponding to the highest resolution such that each of the plurality of layer lines has a color corresponding to a facies classification; and
d) outputting and displaying the two-dimensional image of the portion of the geological formation identifying the set of facies based on the facies classification corresponding to the color of each respective layer line.
2. The method according to claim 1 , wherein:
the data of a) is generated by preprocessing the well log data obtained from the plurality of well logging tools.
3. The method according to claim 2 , wherein:
the preprocessing involves performing a dimension reduction on the well log data obtained from the plurality of well logging tools.
4. The method according to claim 1 , wherein:
the layer boundaries are identified by processing well log data obtained from at least one predetermined well logging tool.
5. The method according to claim 4 , wherein:
the layer boundaries are identified by application of log squaring or a changing point algorithm to the well log data obtained from the at least one predetermined well logging tool.
6. The method according to claim 1 , wherein:
the data of a) is based on well log data obtained depth-by-depth over the portion of the geological formation.
7. The method according to claim 1 , wherein:
the data of a) is determined by at least one of:
i) upscaling well log data obtained from the plurality of well logging tools;
ii) normalizing well log data obtained from the plurality of well logging tools; and
iii) assigning weights to well log data obtained from the plurality of well logging tools based on quality or resolution of facies discrimination of the well logging tool that obtained the well log data.
8. The method according to claim 1 , further comprising:
processing the data of a) to automatically determine the number of facies in the set of facies determined in b).
9. The method according to claim 8 , wherein:
the processing prior to b) uses a Bayesian Information Criterion (BIC) to automatically determine the number of facies in the set of facies without human input.
10. The method according to claim 1 , wherein:
the processing of b) uses a method in which an optimal number of facies clusters is based on a repeatability or consistency of clustering results and there is no human input regarding the optimal number of clusters.
11. The method according to claim 1 , wherein:
the processing of b) uses a Gaussian mixture model (GMM) method to perform facies classification for the portion of the geological formation.
12. The method according to claim 11 , wherein:
a clustering analysis uses the Gaussian mixture model (GMM) method.
13. The method according to claim 1 , wherein the well log data is not restricted to a specific type of well log.
14. The method according to claim 1 , wherein:
operations a) and b) are performed by a processor.
15. The method according to claim 1 , further comprising:
storing in electronic form data representing the set of facies determined in b) or using said data to output and display visual representation of the set of facies determined in b).
16. A system for classifying facies of a geological formation, comprising:
one or more processors configured to:
a) generate data charactering properties of a portion of the geological formation as a function of depth based on well log data obtained from a plurality of different well logging tools, the well log data comprising a plurality of images having varying resolution, wherein the well log data is part of a layer model that includes a number of layers each associated with a vector of properties of the portion of the geological formation and layer boundaries, the generating the data characterizing the properties of the portion of the geological formation as the function of depth based on the well log data obtained from the plurality of different well logging tools comprising:
identifying the layer boundaries from the well log data;
generating the layer model based on the identified layer boundaries and the well log data;
determining a respective resolution of facies classification of each of the well logging tools that obtained the plurality of images of the well log data at the varying resolution;
normalizing the well log data based on the determined resolution of the facies classification of the well logging tools that obtained the well log data; and
determining the properties of the vectors of the layer model by assigning weights to appropriately scale the normalized well log data obtained from the plurality of different well logging tools based on the resolution of the facies classification of the well logging tools that obtained the plurality of images of the well log data of the varying resolution such that corresponding well log data for each of the well logging tools is scaled to a highest resolution among the respective resolutions of the well logging tools;
b) process the data of a) to automatically classify or identify a set of facies for the portion of the geological formation including automatically determining, without human input, a number of facies in the set;
c) automatically generate a two-dimensional image of the portion of the geological formation as a function of depth, the two-dimensional image comprising a plurality of layer lines each having a thickness corresponding to the highest resolution such that each of the plurality of layer lines has a color corresponding to a facies classification; and
d) output and display the two-dimensional image of the portion of the geological formation identifying the set of facies based on the facies classification corresponding to the color of each respective layer line.