Well log analysis

Skill: Well Log Analysis

Read, analyze, and manipulate well log data in LAS format using lasio library.

Purpose

Handle well log operations across all oil & gas pipeline skills. Provides code patterns for reading LAS files, extracting metadata, performing petrophysical calculations, and exporting data.

Reference Materials

Primary source: lasio documentation

Library: lasio

Related packages:

• welly - Extended functionality for curves, wells, projects
• lascheck - LAS specification validation

Dependencies

pip install lasio pandas matplotlib numpy

Capabilities

1. Read LAS Files

import lasio

# Read LAS file
log = lasio.read('well_log.las')

# Access well metadata
print(f"Well: {log.well['WELL'].value}")
print(f"Field: {log.well['FLD'].value if 'FLD' in log.well else 'N/A'}")
print(f"Company: {log.well['COMP'].value if 'COMP' in log.well else 'N/A'}")

# Access curve data
print(f"Curves: {[curve.mnemonic for curve in log.curves]}")
print(f"Depth range: {log.index_min} to {log.index_max} {log.index_unit}")

2. Export to DataFrame

import lasio
import pandas as pd

log = lasio.read('well_log.las')

# Convert to pandas DataFrame
df = log.df()

# Access specific curves
depth = df.index
gamma_ray = df['GR']
density = df['RHOB']
neutron = df['NPHI']

print(df.head())

3. Access Header Metadata

import lasio

log = lasio.read('well_log.las')

# Version section
print(f"Version: {log.version['VERS'].value}")
print(f"Wrap: {log.version['WRAP'].value}")

# Well section
print(log.well)

# Parameters section
print(log.params)

# Other section
for item in log.other:
    print(item)

# Curves section
for curve in log.curves:
    print(f"{curve.mnemonic}: {curve.unit} - {curve.descr}")

4. Petrophysical Calculations

import lasio
import numpy as np

log = lasio.read('well_log.las')
df = log.df()

# Calculate porosity from density
# Phi = (matrix - bulk) / (matrix - fluid)
matrix_density = 2.65  # g/cc (sandstone)
fluid_density = 1.0    # g/cc (water)

bulk_density = df['RHOB'].values
porosity = (matrix_density - bulk_density) / (matrix_density - fluid_density)
porosity = np.clip(porosity, 0, 0.5)  # Clamp to reasonable range

print(f"Average porosity: {porosity.mean():.2%}")

# Calculate water saturation (Archie equation)
# Sw = ((a * Rw) / (Phi^m * Rt))^(1/n)
a = 1.0
m = 2.0
n = 2.0
Rw = 0.1  # ohm-m (formation water resistivity)

phi = porosity
Rt = df['RT'].values  # true resistivity

Sw = ((a * Rw) / (phi**m * Rt))**(1/n)
Sw = np.clip(Sw, 0, 1)

print(f"Average water saturation: {Sw.mean():.2%}")

# Calculate shale volume from gamma ray
GR = df['GR'].values
GR_min = GR.min()
GR_max = GR.max()

Vsh = (GR - GR_min) / (GR_max - GR_min)
Vsh = np.clip(Vsh, 0, 1)

print(f"Average shale volume: {Vsh.mean():.2%}")

5. Visualize Well Logs

import lasio
import matplotlib.pyplot as plt

log = lasio.read('well_log.las')
df = log.df()

fig, axes = plt.subplots(1, 4, figsize=(15, 10), sharey=True)

# Gamma Ray
axes[0].plot(df['GR'], df.index)
axes[0].set_xlabel('GR (API)')
axes[0].set_ylabel('Depth (m)')
axes[0].invert_yaxis()
axes[0].grid(True, alpha=0.3)

# Resistivity
axes[1].semilogx(df['RT'], df.index)
axes[1].set_xlabel('RT (ohm-m)')
axes[1].grid(True, alpha=0.3)

# Density-Neutron
axes[2].plot(df['RHOB'], df.index, label='RHOB')
axes[2].set_xlabel('RHOB (g/cc)')
axes[2].grid(True, alpha=0.3)

# Caliper
axes[3].plot(df['CALI'], df.index)
axes[3].set_xlabel('CALI (in)')
axes[3].grid(True, alpha=0.3)

plt.tight_layout()
plt.savefig('well_logs.png', dpi=150)

6. Export Data

import lasio

log = lasio.read('well_log.las')

# Export to CSV
log.to_csv('output.csv')

# Export to Excel
log.to_excel('output.xlsx')

# Write new LAS file
log.write('modified.las')

7. Handle Non-Standard Files

import lasio

# Read malformed LAS files
log = lasio.read('problematic.las', ignore_header_errors=True)

# Handle encoding issues
log = lasio.read('file.las', encoding='latin-1')

# Read only headers (fast for large files)
log = lasio.read('large.las', read_policy=[])

# Custom null policy
log = lasio.read('file.las', null_policy='common')

8. Create LAS File from Scratch

import lasio
import numpy as np
from lasio import LASFile, CurveItem, HeaderItem

# Create new LAS file
las = LASFile()

# Add well metadata
las.well.append(HeaderItem('WELL', value='Well-001'))
las.well.append(HeaderItem('FLD', value='Test Field'))
las.well.append(HeaderItem('LOC', value='North Sea'))
las.well.append(HeaderItem('SRVC', value='Acme Logging'))

# Add curves
depths = np.arange(1000, 2000, 0.5)
gr = np.random.uniform(20, 150, len(depths))
rhob = np.random.uniform(2.2, 2.7, len(depths))

las.add_curve('DEPT', depths, unit='m')
las.add_curve('GR', gr, unit='API', descr='Gamma Ray')
las.add_curve('RHOB', rhob, unit='g/cc', descr='Bulk Density')

# Write file
las.write('new_well.las')

Workflow Integration

Used by:

• oil-gas-pipelines/exploration - Formation evaluation, log correlation
• oil-gas-pipelines/drilling - Real-time LWD interpretation
• oil-gas-pipelines/reservoir-production - Reservoir characterization

Common Issues

Missing curves

# Check if curve exists
if 'GR' in log.keys():
    gr = log['GR']
else:
    print("GR curve not found")
    print(f"Available: {list(log.keys())}")

Depth units mismatch

# Check depth unit
print(f"Index unit: {log.index_unit}")  # 'm' or 'ft'

# Convert if needed
if log.index_unit == 'ft':
    df = log.df()
    df.index = df.index * 0.3048  # Convert to meters

Null values

# Check for null values
df = log.df()
null_value = log.null
print(f"Null value: {null_value}")

# Replace nulls
import numpy as np
df = df.replace(null_value, np.nan)

Software Tasks

If user requests:

• Well log database → invoke petropowers:oil-gas-delegation
• Automated log interpretation tool → invoke petropowers:oil-gas-delegation
• Log visualization web app → invoke petropowers:oil-gas-delegation

Quick Reference

Operation	Code
Read LAS	lasio.read('file.las')
Get DataFrame	log.df()
Access curve	log['GR'] or log.df()['GR']
Get curve mnemonics	[c.mnemonic for c in log.curves]
Export CSV	log.to_csv('output.csv')
Write LAS	log.write('output.las')

Example Session

User: "Calculate porosity from this LAS file"

Agent:

import lasio
import numpy as np

log = lasio.read('well.las')
df = log.df()

# Density porosity
matrix = 2.65
fluid = 1.0
phi = (matrix - df['RHOB']) / (matrix - fluid)

print(f"Average porosity: {phi.mean():.2%}")
print(f"Porosity range: {phi.min():.2%} to {phi.max():.2%}")

Output:

Average porosity: 23.45%
Porosity range: 5.2% to 41.3%

OSDU Integration

LAS to OSDU WellLog Schema

OSDU uses standard schemas for well log data in the cloud:

• WellLog - Log curves (GR, RHOB, NPHI, etc.)
• Wellbore - Wellbore trajectory and geometry
• Well - Well header and metadata

Schema Mapping: LAS → OSDU WellLog

LAS Field	OSDU WellLog Field
WELL	FacilityName (via Well)
FLD	FieldName
DEPT	LogCurveDepth
GR, RHOB, NPHI	LogCurveData.Curves
Company	Operator

OSDU Dataset Service API

import requests

headers = {
    "Authorization": "Bearer <token>",
    "data-partition-id": "mypartition"
}

# Get storage instructions
storage_resp = requests.post(
    "https://api.osdu.com/api/dataset/v1/storageInstructions",
    headers=headers,
    json={
        "kind": "osdu:wks:dataset--File.LAS:1.0.0",
        "dataContentType": "text/plain"
    }
)

# Upload file to provided URL
upload_url = storage_resp.json()["storageLocation"]["signedUrl"]
with open("well_log.las", "rb") as f:
    requests.put(upload_url, data=f.read())

# Register dataset
register_resp = requests.post(
    "https://api.osdu.com/api/dataset/v1/register",
    headers=headers,
    json={
        "id": "dataset--File.LAS:1654321",
        "kind": "osdu:wks:dataset--File.LAS:1.0.0"
    }
)

Create OSDU WellLog Manifest

import json
from datetime import datetime

def create_welllog_manifest(well_id, wellbore_id, curves, legal_tags):
    """Create OSDU manifest for well logs"""
    
    # Map LAS curves to OSDU curves
    curve_data = []
    for curve in curves:
        curve_data.append({
            "CurveID": f"log-curve-{curve['mnemonic']}",
            "CurveUnit": curve['unit'],
            "LogCurveValues": curve['values']
        })
    
    manifest = {
        "kind": "osdu:wks:Manifest:1.0.0",
        "Data": {
            "WorkProduct": {
                "kind": "osdu:wks:work-product--WellLog:1.0.0",
                "id": f"work-product--WellLog:{well_id}-{datetime.now().strftime('%Y%m%d')}",
                "legal": legal_tags,
                "data": {
                    "Name": f"Well Log {well_id}",
                    "WellboreID": f"master-data--Wellbore:{wellbore_id}",
                    "LogCurveDepth": {
                        "Start": curves[0]['start_depth'],
                        "End": curves[0]['end_depth'],
                        "Unit": "m"
                    },
                    "LogCurveData": {
                        "Curves": curve_data
                    }
                }
            }
        }
    }
    return json.dumps(manifest, indent=2)

LAS to OSDU WellLog Workflow

import lasio
import json

def las_to_osdu_welllog(las_file, wellbore_id, legal_tags):
    """Convert LAS file to OSDU WellLog schema"""
    
    # Read LAS file
    log = lasio.read(las_file)
    
    # Extract curves
    curves = []
    for curve in log.curves:
        curves.append({
            "mnemonic": curve.mnemonic,
            "unit": curve.unit,
            "values": curve.data.tolist(),
            "start_depth": log.index_min,
            "end_depth": log.index_max
        })
    
    # Create manifest
    manifest = create_welllog_manifest(
        well_id=log.well['WELL'].value,
        wellbore_id=wellbore_id,
        curves=curves,
        legal_tags=legal_tags
    )
    
    return manifest

# Example usage
manifest = las_to_osdu_welllog(
    "well_001.las",
    wellbore_id="master-data--Wellbore:wellbore-001",
    legal_tags={"legaltags": ["my-legal-tag"], "otherRelevantDataCountries": ["US"]}
)

From LAS to OSDU Workflow

1. Read LAS file using lasio, perform quality checks (NULL values, depth consistency)
2. Create Wellbore master data if not exists
3. Create WellLog work product with curve data
4. Create Dataset referencing original LAS file
5. Submit manifest via OSDU ingestion API

OSDU Reference Documentation

• OSDU WellLog schema: https://community.opengroup.org/osdu/data/data-definitions/-/tree/master/Schema/WellLog
• OSDU Data Loading Guide: https://gitlab.opengroup.org/osdu/r3-program-activities/docs/-/raw/master/R3%20Document%20Snapshot/13-osdu-data-loading-guide.pdf
• Dataset Service API: https://community.opengroup.org/osdu/platform/system/dataset

Software Tasks

If user requests:

• OSDU ingestion pipeline → invoke petropowers:oil-gas-delegation
• Well log visualization app → invoke petropowers:oil-gas-delegation
• LAS to OSDU conversion script → stay in domain skill (conversion logic)