Petroleum Engineer Manager · KNOC · Ulsan, Korea
Petroleum AI
Reservoir and production engineering tools that run in your browser. No install, no server, and your data never leaves the tab.
Energy powers our world, yet its hidden complexity is not easily understood.
Therefore, we unite petroleum with AI, so that knowledge may become clearer and decisions wiser.
May this intelligence serve all, making energy safer, smarter, and better for the world.
Everything below is the working half of that — the tools, the record, and the notes on where both break.
The tools
Single HTML files. Open one from disk on a plane.
Standard physics, every assumption written out where you can read it. Free, and free of a login.
Next
Four today. The list is not finished.
Every project leaves behind a diagnostic I wish had existed. As the AI and machine-learning side of this work matures, more of them will end up here — still free, still offline, still open about their assumptions. The aim is simple: keep giving the industry tools that make an engineer's judgement easier, not tools that pretend to replace it.
If there is a calculation you keep redoing by hand, tell me what it is — that is usually where the next tool comes from.
Suggest a toolWhere AI stops
Six pressure points. Thirty Bscf of doubt.
Drag the OGIP and watch the straight line pivot. Anything inside the shaded band fits the same six points to within gauge noise — the solver cannot tell you which one is the field.
That is what the AI work is for: surface every match the data allows, then let the engineer bring the drive mechanism, the geology and the operating history to the decision.
Fit residual 27 psi · the shaded band is every OGIP inside gauge noise.
Two ledgers, one practice
What I do on the subsurface side, and what I do in code.
- Reservoir simulation, history matching and prediction
- Reserves estimation and field development planning
- EOR / IOR screening for mature assets
- ESP and artificial lift optimisation, production surveillance
- Technical due diligence and data room review
- Browser-native engineering tools, offline by design
- Automated history matching — Nelder–Mead, multi-tank solvers
- Ensemble search that surfaces every match the data allows
- Open public datasets rebuilt into usable interfaces
- ML on production and pressure data — screening, not deciding
The record
Nine countries, seven reservoir types.
Asset evaluation and reservoir engineering at KNOC since 2009 — from new-ventures screening in Europe & Africa to a three-year operations secondment at Cuu Long JOC in Vietnam.
| Region | Reservoir | Scope | Period |
|---|---|---|---|
| United States | Shale gas · offshore | Reserves evaluation, gas lift optimisation | 2025 — present |
| Canada | Bitumen · tight gas | Reservoir simulation (SAGD), rate transient analysis and reserves evaluation | 2015 — present |
| Vietnam | Fractured basement · gas-condensate | Operations secondment — ESP and gas lift optimisation, water-injection simulation, acidizing, field surveillance and production allocation; compositional simulation, FDP assessment and approval | 2021 — 2024 · 2015 |
| Iraq | Carbonate | Reserves evaluation | 2020 |
| United Kingdom | North Sea offshore | Reservoir simulation, FDP assessment and approval, polymer EOR evaluation | 2018 — 2020 |
| Kazakhstan | Heavy oil | Reservoir simulation (cyclic steam stimulation) | 2014 |
| Peru | Offshore | Reservoir simulation (gas injection), reserves and recovery evaluation | 2013 |
| Poland | Shale gas | New ventures screening, technical due diligence and data room review | 2011 — 2012 |
| Africa | Onshore | New ventures screening, technical due diligence and data room review | 2009 — 2011 |
DEC 2009 — PRESENT · ULSAN
Petroleum Engineer Manager
Korea National Oil Corporation · American & Europe Technology Team
MAR 2021 — AUG 2024 · HO CHI MINH CITY
Senior Petroleum Engineer
Cuu Long JOC — seconded from KNOC
1998 — 2006 · BUSAN
B.Eng Natural Resources · BBA
Dong-A University
Publications
Published with the Society of Petroleum Engineers.
SPE-176489-MS · OCT 2015 · BALI
Rate Transient Analysis in Hydraulic Fractured Tight Gas Reservoir
Hyeonseog Joo, Seil Ki · Korea National Oil Corporation
SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition
Four rate transient analysis methods — flowing material balance, type-curve matching, square-root-of-time and an analytical model — applied to 2.3 years of rate and flowing wellhead pressure from a hydraulically fractured tight gas well, alongside micro-seismic and log data. The methods agree that boundary-dominated flow begins about eight months in, and cross-validate the OGIP, EUR and SRV estimates. Tracking how those estimates move as the analysis point advances gives a guide for history matching the reservoir simulation.
Read the abstract on OnePetrodoi:10.2118/176489-MSTechnical notes
What the tools are doing, and where they break.
Well performance monitoring in a single HTML file
No install, no server, no data leaving the PC — and why in-house LLMs matter once AI touches real production data.
Read →Material balance non-uniqueness
Why three OGIP values fit the same six points, and how to choose between them.
Read →Pressure data first, solver second
Gauge drift, datum corrections and the errors that survive every optimiser.
Read →What AI cannot decide for a reservoir engineer
Drive mechanism is an interpretation. Models rank matches; they do not know the field.
Read →Reading 25 years of NSTA data
Rebuilding an open production dataset into something you can actually query.
Read →Found a bug? Want a feature?
If a tool gives you a result that looks wrong, tell me and I will fix it as fast as I can.
- Emailnasel777@gmail.com
- LinkedInlinkedin.com/in/hsjoo1307
- Based inBusan, Republic of Korea · open to international projects