Well log interpretation that stays on your machine
A browser-based petrophysics workspace for loading LAS and DLIS well logs, running Vsh through net pay in structured views, and staging zone averages for Volumetrics, without uploading your log files or saved cases to our servers.
Included on the Starter plan and above. Formation Evaluation projects save to a folder you choose on your computer; they do not count toward Starter’s cloud saved-case limit. See Starter pricing
Formation Evaluation with RRS-1 EAGLE loaded, 7,800–8,800 ft, 12 curves, 0.5 ft sample rate. A synthetic well, so nothing here is client data.
Your well data stays local
Formation Evaluation saves projects to a folder you choose on your computer using the File System Access API: case JSON, LAS files, trend lines, and display templates live in that project folder. We do not store well log files or interpretation cases in Supabase. DLIS parsing runs on our server only to return LAS-shaped JSON to your browser; the uploaded file is not retained. When you bridge zone results to Volumetrics, the handoff stays in your browser (local module bridge). CSV import elsewhere on the platform follows the same browser-only pattern where enabled.
Confidentiality with the operator
Petrophysicists working under operator confidentiality terms often cannot send raw logs to third-party cloud storage. This module is designed so interpretation work remains on the device you control; sharing is manual (export project package or case files) when you choose to involve a teammate.
Workspace views
Each case moves through dedicated views (same order as the in-app View menu):
Log viewer: load LAS or DLIS, alias curves, QC, and apply display templates
Vsh workspace: shale volume from GR, SP, neutron–density, and related indicators
Porosity workspace: density, neutron, sonic, and combined models with shale endpoints
Sw workspace: 6 saturation model families (Archie, Simandoux, Indonesia, dual-water, Waxman-Smits, and Juhasz)
Net pay workspace: cutoffs, coal handling, and net/gross flags
Crossplots: neutron–density, Pickett, Hingle, M–N, Buckles (BVW), and GR histogram
Results and sensitivity: zone tables, exports, and cutoff sensitivity
The Compute suite
Beyond the seven core views, the Compute menu adds specialist interpretation you would normally leave the workspace for. Every mode reads the same curves, depth reference, zones, alias table, and unit system as the rest of your case, and writes results back as depth curves you can plot, cross-plot, and export.
Porosity and lithology
Neutron–density crossplot porosity
Apparent porosity from neutron–density crossplot tie lines, using the NPHI calibration already set in your interpretation parameters. No display template needed; the calibration follows the case.
Methods: Crossplot tie-line solution with matrix and fluid end-points
Neutron–sonic crossplot porosity
Apparent porosity from neutron–sonic tie lines for intervals where the density log is washed out, missing, or affected by barite mud.
Methods: Crossplot tie-line solution with NPHI calibration and fluid transit time
Neutron matrix conversion
Convert a raw neutron curve recorded on one matrix to the matrix you are interpreting on, so limestone-scaled and sandstone-scaled logs across a field can finally be compared.
Methods: Standard neutron matrix correction table, calibration lithology from interpretation parameters
Fuzzy lithology classification
Classify sandstone, limestone, dolomite, and shale from a curve suite without hand-drawing cut-offs on every crossplot. Output is a class-code curve you can carry into net pay and zone work.
Methods: Fuzzy c-means clustering of gamma ray, neutron, density, and resistivity
Multimineral solver
Deterministic multimineral inversion that solves component volumes from your log end-points under physical constraints (every volume at or above zero, all volumes summing to one) rather than letting an unconstrained fit return mineral fractions that cannot exist.
Methods: Constrained least-squares inversion from user-defined log end-points
Permeability
Permeability from logs
Estimate permeability from effective porosity and water saturation using the published transform that matches your rock, with the choice visible in the case rather than buried in a spreadsheet formula.
Methods: Timur, Tixier, Wyllie–Rose, Kozeny–Carman, and Coates free-fluid transforms
Neural network permeability
Where a published transform does not fit the rock, train a small network on your own core calibration points and predict permeability from porosity plus up to three further log curves.
Methods: Feed-forward neural network trained on your core plug data
Rock typing
Rock typing and hydraulic flow units
Split the core porosity–permeability cloud into populations that each obey their own transform, then predict which one every log sample belongs to. A single global k–φ transform is usually the largest error in a computed permeability curve; two plugs at the same porosity can differ by two orders of magnitude in permeability, and one regression through the whole cloud is wrong for both. Each rock type carries its own fitted transform, its own average k and φ for a saturation-height model, and a published class name where the method has one.
Methods: Amaefule RQI/FZI flow units, Winland R35 port size, Lucia rock-fabric number, silhouette-scored clustering, and a Gaussian classifier that reports its own confidence
Saturation and pressure
Saturation-height modelling
Model water saturation from height above the free water level instead of from resistivity, the independent check that matters in thin beds, low-resistivity pay, and wells where the resistivity answer is disputed.
Methods: Power-law, Swirr-anchored, and Leverett J-function forms built from capillary pressure and k/φ
Formation pressure analysis
Import RFT and MDT pressure points, fit gas, oil, and water gradients, and read fluid contacts straight off the gradient intersections rather than inferring them from logs alone.
Methods: Gradient fitting with GOC, OWC, and FWL from gradient intersections
Geomechanics
Geomechanics (1D mechanical earth model)
A full wellbore stability model along the well: elastic properties with shear synthesis, overburden, pore pressure, rock strength, poroelastic horizontal stresses, the safe mud weight window, a stress polygon, and sanding onset pressure, calibrated against imported RFT/MDT and LOT/FIT points.
Methods: Eaton and Bowers pore pressure, McNally / Horsrud / Golubev UCS, Kirsch hoop stresses with Mohr–Coulomb failure
Tie the well to seismic from sonic and density: acoustic impedance, integrated two-way time, normal-incidence reflectivity, and a convolved synthetic trace, without exporting curves to a separate geophysics package.
Methods: Acoustic impedance, time–depth integration, and Ricker wavelet convolution
Crossplots
Custom crossplot
Cross-plot any two curves the well carries, logged or calculated, when the standard neutron–density, Pickett, and Hingle views do not frame the question you actually have. An axis can be committed back to the well as a depth curve, so a relationship you find on the plot becomes a curve the rest of the interpretation can use.
Methods: Arbitrary curve pairs with optional axis commit back to the well
Bridge to Volumetrics
When zone averages are ready, stage NTG, porosity, Sw, and net pay for the Volumetrics module through an explicit in-browser handoff. You review the payload before confirming; nothing is applied silently downstream.
