neqsim-reference-fluid-synthetic-generation

Reference-Fluid Synthetic Generation

Use this skill for the "common reference fluid → adjust split factor → match measured data → generate representative fluids" workflow. It provides three plant-agnostic, dependency-free helpers:

1. match_split_factor — a robust golden-section 1-D search that finds the split / characterization factor best reproducing a measured target.
2. generate_fluid_cases — build representative or synthetic fluid cases by applying a range of factors to a common reference fluid.
3. blend_compositions — combine several well or fluid compositions into a single field composition by molar-rate allocation.

The forward model / fluid builder is injected by the caller, so this skill has no dependency on a particular EOS. In practice it wraps the community pseudocomponent-split-characterization gamma split or a NeqSim characterization call, so the same factor drives both the split and the match.

When to Use

• When there is no PVT study at all — a discovery, a prospect, an early concept — and a reservoir model still needs a fluid. See No PVT data below.
• When you have a common reference EOS/fluid and want field-specific or per-case fluids by adjusting one heavy-end split factor (the "common factor" idea: reuse one characterization method with field-specific calibration).
• When a split factor must be calibrated so the model reproduces a measured saturation pressure, GOR, or stock-tank-oil density.
• When production is allocated across several wells or fluids and you need one representative field composition.
• When a complete PVT study is unavailable and you must generate a usable fluid from a reference plus available measurements.

No PVT data: build a best guess and say so

A reservoir model needs a fluid long before a laboratory PVT study exists, and often before a sample has been taken. The alternative to guessing silently is to guess explicitly.

from reference_fluid import build_analogue_fluid_basis

basis = build_analogue_fluid_basis(
    depth_tvdss_m=3590.0,
    province="northern_north_sea",
    formation="Brent",
    water_depth_m=381.0,
    measured_temperature_C=129.0,     # anything measured is used and labelled
    measured_pressure_bara=542.0,     # ... and anything absent is derived
)
basis.targets["gor_sm3_sm3"]      # value + provenance + GOR DEFINITION
basis.seed["c7_plus_cuts"]        # ready for addTBPfraction / addPlusFraction
basis.assumptions                 # each with the measurement that retires it
basis.acquisition_plan            # ranked by how much it reduces the answer