Total
39
Done
0
Progress
0%
Left
39
Est. time
~26 h
Industry & value chain
industry 0/8 ▾Upstream, midstream and downstream
Who explores and produces, who moves and stores, who refines and sells. Where data work sits in each.
EIA — Oil and petroleum explained
The petroleum system
Source rock, reservoir, seal, trap and migration: the conditions an oil or gas accumulation needs.
Hyne
Conventional vs unconventional
Permeable reservoirs that flow on their own, against shale and tight rock that need horizontal wells and fracking.
Hyne · EIA
The E&P lifecycle
Exploration, appraisal, development, production and abandonment, and the data each phase produces.
Hyne
Reserves and resources (PRMS)
1P, 2P and 3P reserves, contingent and prospective resources. A reserve is an economic estimate, not a volume in the ground.
SPE — PRMS 2018
Benchmarks and prices
Brent, WTI and Henry Hub, and how local prices relate to them.
EIA spot prices
Natural gas and LNG
Gas as its own business: processing, pipelines, storage and liquefaction.
EIA — Natural gas explained
Vaca Muerta
The shale play in Neuquén: its geology, the operators, the move from pilots to development, and the public data on its wells.
Secretaría de Energía — open data
Subsurface & reservoir
subsurface 0/6 ▾Porosity, permeability and saturation
The three numbers that say how much fluid a rock holds, how easily it flows, and how much of it is oil.
Hyne · Ahmed
Well logs
Gamma ray, resistivity, density, neutron and sonic: what each one measures and what it says about the rock.
Hyne
LAS and DLIS files in Python
Reading logs with lasio, welly and dlisio. Depth units, null values and curve mnemonics that change between vendors.
lasio · welly · dlisio docs
Reflection seismic and SEG-Y
How seismic images the subsurface, and reading SEG-Y volumes with segyio.
segyio docs
Reservoir fluids and PVT
How oil, gas and water behave with pressure and temperature: bubble point, formation volume factor, gas-oil ratio.
Ahmed
Reservoir simulation
Grids, flow equations and history matching. Run an open simulator on a public field model.
OPM Flow
Drilling & completion
drilling 0/5 ▾Well construction
Rig, drill string, bit, drilling fluid, casing and cement: a well is built in sections, each one cased before the next.
Hyne
Directional and horizontal wells
Surveys, inclination and azimuth, and why long horizontal laterals made shale viable.
Hyne
Drilling data and WITSML
Real-time parameters (ROP, WOB, RPM, torque), mud logs, and WITSML, the standard that moves them between rig and office.
Energistics — WITSML
Hydraulic fracturing
Stages, fluid and proppant, and the completion design choices that drive what a shale well produces.
Secretaría de Energía — Adjunto IV
Drilling KPIs and NPT
Days per well, meters per day and non-productive time: how operators compare rigs and crews.
Secretaría de Energía — drilling data
Production & operations
production 0/6 ▾Production data
Monthly oil, gas and water per well, and cumulative production. Reading the Capítulo IV files.
Secretaría de Energía — Capítulo IV
Decline curve analysis (Arps)
Exponential, hyperbolic and harmonic declines. Fitting the initial rate, Di and b, and estimating EUR.
Ahmed
Artificial lift
Rod pumps, ESPs and gas lift: when a well needs help to produce, and what each method's data looks like.
Hyne
Well testing
Pressure buildup and drawdown tests, and what they reveal about the reservoir around a well.
Ahmed
Water, gas and flaring
Water cut, produced-water handling, and why gas gets flared where there is no pipeline.
World Bank — GGFR
Production allocation
Meters sit at facilities, not at every well, so per-well volumes are often allocated rather than measured. What that does to your models.
Energistics — PRODML
Field data & platforms
data 0/4 ▾SCADA and historians
Sensor data from wells and plants, stored in historians like AVEVA PI: tags, compression and interpolation.
AVEVA PI System docs
OPC UA and edge gateways
Reading PLC data in the field and sending it over poor links without losing it.
OPC Foundation
OSDU and industry data standards
The OSDU data platform, and the Energistics family: WITSML for drilling, PRODML for production, RESQML for models.
The Open Group — OSDU
Open subsurface Python
The Software Underground community and its tools (lasio, welly, segyio, striplog), with the Transform tutorials.
Software Underground — Transform
Machine learning applied
ml 0/7 ▾Facies classification from logs
Predict lithology from log curves. The SEG 2016 and FORCE 2020 contests are the public benchmarks.
Pandey · Hall (2016)
Predicting missing logs
Synthesize a sonic or density curve from the others. Validate across wells, never within one.
Pandey
Production forecasting: Arps vs ML
When ML beats decline curves and when it just overfits. Split by well, not by date.
Pandey
Artificial lift failure prediction
Classifying dynamometer cards from rod pumps, and predicting ESP failures from current, pressure and temperature.
Pandey
Anomaly detection on sensor streams
Leaks, stuck pipe and failing sensors in historian data. Labels are rare, so start with rules and unsupervised baselines.
Pandey
Deep learning on seismic
Fault and salt detection as image segmentation, trained on synthetic data and tested on the F3 block.
Wu, Liang & Shi (2019)
Physics and ML together
Why purely data-driven models break outside their training range, and how hybrid and physics-informed models handle it.
Pandey
Emissions & regulation
emissions 0/3 ▾Methane emissions
Where methane leaks along the chain, how it is measured, and how satellites spot large plumes.
IEA — Global Methane Tracker
Flaring from space
Nightly VIIRS fire detections turned into flared-gas volumes by country and field.
Colorado School of Mines — VIIRS Nightfire
Argentine regulation and reporting
What operators report to the Secretaría de Energía (Capítulo IV, Adjunto IV) and where that public data lives.
Secretaría de Energía
// bibliography
The books, courses and practice sites behind this roadmap.
// books
Nontechnical Guide to Petroleum Geology, Exploration, Drilling & Production
The whole upstream for non-engineers. Start here.
Reservoir Engineering Handbook
Rock and fluid properties, well tests and decline curves, with the equations.
Machine Learning in the Oil and Gas Industry ↗
ML applied to subsurface, drilling and production, with Python.
// courses
Software Underground — Transform tutorials ↗
Hands-on Python for geoscience: logs, seismic and ML.
// practice
Capítulo IV — production by well ↗
Monthly production of every well in the country, CC-BY-4.0.
Equinor Volve field data ↗
A full North Sea field: logs, seismic, drilling and production.
FORCE 2020 lithology contest ↗
Labeled logs from 100+ Norwegian wells.
SEG 2016 facies contest ↗
The classic facies-classification benchmark, with notebooks.