# US 2020/0401951 A1 · Estimating permeability values from well logs using a depth blended model

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0401951 A1
- Document ID: US20200401951A1
- Status: Published application
- Kind: A1
- Application number: 16/787,245
- Filed: 2020-02-11
- Published: 2020-12-24
- Earliest claimed: 2019-06-21
- Assignees: Cgg Services SAS; Kuwait Gulf Oil Company
- Applicants: Cgg Services SAS; Kuwait Gulf Oil Company
- Inventors: Chiranjith RANGANATHAN; Francisco BRITO; Ahmad S D S M ALBUSSAIRI
- Web page: https://patentgrep.com/patent/US20200401951A1

## CPC

- G06N 20/20: Ensemble learning
- G01V 1/306: {for determining physical properties of the subsurface, e.g. impedance, porosity or attenuation profiles}
- G01V 1/50: Analysing data
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G01V 2210/6169: using well-logging
- G01V 2210/6246: Permeability
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G01V: GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS (means for indicating the location of accidentally buried, e.g. snow-buried, persons A63B29/02)

## Abstract

Permeability values are estimated based on well logs using regression algorithms, such as gradient boosting and random forest. The training data is selected from well logs for which core-analysis-based permeability values are available. The estimated permeability values are used to plan hydrocarbon production. The well logs used to build the depth blended model may include total porosity, gamma ray, volume of calcite, density, resistivity, and neutron logs. Selecting the training data may include grouping the well logs according to regions expected to have similar characteristics, choosing a subset of the well logs corresponding to wells expected to provide stable models according to pre-determined criteria, and/or identifying training zones on the chosen well logs according to one or more rules. Validation and consistency checks may also be performed.

## Claims

### Claim 1

A method for estimating permeability values based on well logs, the method comprising:
selecting training data from well logs for which core-analysis-based permeability values are available;
training regression algorithms using the training data to generate a depth blended model for estimating permeability values, wherein the depth blended model selectively outputs values obtained via at least two of the regression algorithms; and
applying the depth blended model to the well logs other than the training data to estimate model-based permeability values,
wherein the model-based permeability values are used to plan hydrocarbon production.

### Claim 2 (depends on 1)

The method of claim 1, wherein the well logs used to build the depth blended model include total porosity, gamma ray, resistivity, volume of calcite, density, and neutron logs.

### Claim 3 (depends on 1)

The method of claim 1, wherein the selecting includes grouping the well logs according to regions expected to have similar characteristics.

### Claim 4 (depends on 1)

The method of claim 1, wherein the selecting includes choosing a subset of the well logs corresponding to wells expected to provide stable models according to pre-determined criteria.

### Claim 5 (depends on 4)

The method of claim 4, wherein the selecting further includes identifying training zones on the chosen well logs according to one or more rules.

### Claim 6 (depends on 1)

The method of claim 1, wherein the selecting includes pre-processing the training data by scaling and/or interpolating.

### Claim 7 (depends on 1)

The method of claim 1, wherein the depth blended model yields a first permeability value estimated by a primary regression algorithm among the regression algorithms if the first permeability value meets a predetermined requirement and yields a second permeability value estimated by a secondary regression algorithm otherwise.

### Claim 8 (depends on 1)

The method of claim 1, wherein the primary regression algorithm is a gradient boosting, the secondary regression algorithm is a random forest algorithm and the predetermined requirement is the first permeability value be positive.

### Claim 9 (depends on 1)

The method of claim 1, further including validating the depth blended model by applying the depth blended model to well logs included in the training data but not used to generate the depth blended model, wherein if the depth blended model-estimated permeability values are not consistent with the core-analysis-based permeability values, the training is repeated.

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
performing well-by-well quality checks for the model-estimated permeability values.

### Claim 11

A permeability estimating apparatus, comprising:
a processor configured
to select training data from well logs for which core-analysis-based permeability values are available,
to train regression algorithms to generate a depth blended model for estimating permeability values based on well logs included the training data with their corresponding core-analysis-based permeability values, and
to apply the depth blended model to the well logs other than the training data for obtaining model-estimated permeability values; and
a communication interface connected to the processor and configured to exchange data with other devices and/or enable interaction with a user,
wherein the estimated permeability values are used to plan hydrocarbon production, and wherein the depth blended model selectively outputs values obtained via at least two of the regression algorithms.

### Claim 12 (depends on 11)

The permeability estimating apparatus of claim 11, wherein the well logs used to build the depth blended model include total porosity, gamma ray, volume of calcite, density, and neutron logs.

### Claim 13 (depends on 11)

The permeability estimating apparatus of claim 11, wherein when selecting the training data, the processor groups the well logs according to regions expected to have similar characteristics.

### Claim 14 (depends on 11)

The permeability estimating apparatus of claim 11, wherein, when selecting the training data, the processor chooses a subset of the well logs corresponding to wells expected to provide stable models according to pre-determined criteria.

### Claim 15 (depends on 14)

The permeability estimating apparatus of claim 14, wherein, when selecting the training data, the processor identifies training zones on the chosen well logs according to one or more rules.

### Claim 16 (depends on 11)

The permeability estimating apparatus of claim 11, wherein the processor is further configured to perform pre-processing of the training data by scaling and/or interpolating.

### Claim 17 (depends on 11)

The permeability estimating apparatus of claim 11, wherein the depth blended model yields a first permeability value estimated by a primary regression algorithm among the regression algorithms if the first permeability value meets a predetermined requirement, and a second permeability value estimated by a secondary algorithm otherwise.

### Claim 18 (depends on 11)

The permeability estimating apparatus of claim 11, wherein the primary regression algorithm is a gradient boosting, the second regression algorithm is a random forest algorithm and the predetermined requirement is the first permeability value be positive.

### Claim 19 (depends on 11)

The permeability estimating apparatus of claim 11, wherein the processor is further configured to validate the depth blended model by applying the depth blended model to well logs included in the training data but not used to generate the depth blended model, wherein if model-estimated permeability values are not consistent with the core-analysis-based permeability values, the training is repeated.

### Claim 20 (depends on 11)

The permeability estimating apparatus of claim 11, wherein the processor is further configured to perform well-by-well consistency checks for the model-estimated permeability values.