Subsurface Modeling
Subsurface modeling builds 3D geological representations (structure, stratigraphy, rock properties) for oil & gas, mining, and groundwater work. SLB's Petrel dominates enterprise E&P shops: proprietary subscription, latest public build Petrel 2024.8, AI assistant added October 2025, reported annual cost ~$30K-$50K per user. Emerson's SKUA-GOCAD uses patented UVT-transform gridding for complex structures; its Subsurface Science & Engineering unit joined AspenTech in 2022, and AspenTech became a wholly owned Emerson subsidiary in March 2025. Baker Hughes' JewelSuite, co-developed with Shell, uses patented gridding for well placement via Azure Marketplace. RockWare's RockWorks undercuts both: perpetual licenses $2,000-$6,000, annual $860-$2,580/year. The open-source option, GemPy (Python, EUPL-1.2), hit stable release 2026.0.3 in June 2026 with 1.3k GitHub stars.
Pick Petrel if you're inside an SLB-standardized enterprise workflow needing seismic-to-simulation in one desktop — nothing else matches its ecosystem lock-in, but $30K-$50K/user/year is dead money for a two-person consultancy or a university lab. For structurally violent plays (salt, thrust belts), SKUA-GOCAD's UVT-transform beats Petrel's grid-first approach, though its 2025 folding into Emerson-owned AspenTech makes the roadmap worth watching. If you're not enterprise, skip all three and start with GemPy — free, scriptable, EUPL-1.2 — until you outgrow implicit modeling; its own docs admit it struggles with overturned and reversed-dip geometries that explicit gridding handles natively. Known weakness: GemPy's implicit-interpolation approach — the trade-off for being free and scriptable — struggles with overturned beds, reverse faults, and other reversed-dip geometries that explicit gridding tools like Petrel and SKUA-GOCAD handle natively.