Fluid characterization converts measured light components, true-boiling-point (TBP) cuts, and an unresolved heavy-end or plus fraction into the pseudo-components used by a compositional equation of state. The generated fluid description is a model input: it still requires comparison with representative PVT laboratory data before engineering use.

Input contract and order of operations

Use relative mole amounts for every component. NeqSim normalizes composition during initialization; the values do not have to sum to one.

Input Method Required basis
Identified component addComponent(name, moles) Component name from the NeqSim database and a non-negative relative mole amount
Pre-binned petroleum cut addTBPfraction(name, moles, molarMass, density) Molar mass in kg/mol; density argument as specific gravity/relative density, numerically equal to g/cm³
Unresolved heavy end addPlusFraction(name, moles, molarMass, density) Molar mass in kg/mol; density argument as specific gravity/relative density, numerically equal to g/cm³
Thermodynamic state system constructor or setters Temperature in K and absolute pressure in bara

The four-argument density methods accept the usual petroleum-cut relative-density range. Values above 1.5 are interpreted by the API as kg/m³ and converted. Prefer the documented relative-density basis so an accidental unit change does not silently change the characterization.

Apply the setup in this order:

  1. create the equation-of-state system at the intended temperature and pressure;
  2. select the TBP correlation before adding TBP or plus fractions;
  3. add identified components, TBP cuts, and one unresolved plus fraction;
  4. select the plus-fraction and lumping models;
  5. call characterisePlusFraction();
  6. set the mixing rule, flash the characterized system, and inspect the result.

Characterization and lumping routes

setPlusFractionModel("Pedersen") distributes the unresolved heavy end into single-carbon-number components using the selected TBP correlation. characterisePlusFraction() then applies the configured lumping model.

Lumping route Fluent configuration Meaning
Preserve the light TBP cuts and lump only the heavy end .model("PVTlumpingModel").plusFractionGroups(n) Keeps the explicit C6-C9 TBP cuts and creates n groups from C10+
Lump all heavy fractions from C6 .model("standard").totalPseudoComponents(n) Creates n total heavy pseudo-components
Retain every generated SCN component .noLumping() Highest component count; useful for diagnosis rather than routine simulation
Match project-owned carbon-number bins .customBoundaries(...) Uses the supplied starting carbon number for each group

Do not use plusFractionGroups and totalPseudoComponents interchangeably. Their counts have different meanings. The final component slate also contains the identified components and any TBP cuts preserved by the selected route.

Executable characterized-fluid example

The program below uses the public fluent lumping API, performs a TP flash, and checks the resulting component count and phase split. It is one complete Java 8 program and uses Log4j2 instead of console output. Run documentation examples with assertions enabled (java -ea) so the engineering checks execute.

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import neqsim.thermo.system.SystemInterface;
import neqsim.thermo.system.SystemSrkEos;
import neqsim.thermodynamicoperations.ThermodynamicOperations;

public final class PvtFluidCharacterizationExample {
  private static final Logger logger =
      LogManager.getLogger(PvtFluidCharacterizationExample.class);

  private PvtFluidCharacterizationExample() {}

  public static void main(String[] args) {
    double temperatureK = 298.0;
    double pressureBara = 10.0;

    SystemInterface fluid = new SystemSrkEos(temperatureK, pressureBara);
    fluid.getCharacterization().setTBPModel("PedersenSRK");

    fluid.addComponent("CO2", 1.0);
    fluid.addComponent("methane", 51.0);
    fluid.addComponent("ethane", 1.0);
    fluid.addComponent("propane", 1.0);

    // name, relative moles, molar mass [kg/mol], relative density [-]
    fluid.addTBPfraction("C6", 1.0, 0.090, 0.70);
    fluid.addTBPfraction("C7", 1.0, 0.110, 0.73);
    fluid.addTBPfraction("C8", 1.0, 0.120, 0.76);
    fluid.addTBPfraction("C9", 1.0, 0.140, 0.79);
    fluid.addPlusFraction("C10", 11.0, 0.290, 0.82);

    fluid.getCharacterization().setPlusFractionModel("Pedersen");
    fluid.getCharacterization().configureLumping()
        .model("PVTlumpingModel")
        .plusFractionGroups(9)
        .build();
    fluid.getCharacterization().characterisePlusFraction();
    fluid.setMixingRule("classic");

    ThermodynamicOperations operations = new ThermodynamicOperations(fluid);
    operations.TPflash();

    double vaporMoleFraction = fluid.getBeta();
    assert fluid.getNumberOfComponents() == 17;
    assert Double.isFinite(vaporMoleFraction);
    assert vaporMoleFraction > 0.0 && vaporMoleFraction < 1.0;

    logger.info(
        "Characterized {} components at {} K and {} bara; vapor mole fraction={}",
        fluid.getNumberOfComponents(),
        temperatureK,
        pressureBara,
        vaporMoleFraction);
  }
}

The 17-component assertion closes the intended slate: four identified components, four preserved C6-C9 TBP cuts, and nine groups created from C10+. It is a regression check for this stated example, not a universal component count for other assays or lumping choices.

Tuning, experiments, and reported properties

Characterization does not tune an equation of state. Match uncertain plus-fraction properties, binary-interaction parameters, volume translation, and viscosity parameters against project-owned measurements using a documented objective function, bounds, weights, and holdout data. Preserve the raw assay, unit conversions, model names, fitted parameters, software version, and residuals with the fluid description.

Do not calculate oil formation-volume factor or gas-oil ratio from a phase volume divided by total feed moles. Those quantities require named stock-tank and separator conditions plus an explicit standard-volume basis. Use the dedicated CCE, CVD, differential-liberation, separator, and viscosity workflows, and record every pressure as absolute or gauge and every gas/oil standard condition.

Engineering boundaries

The Pedersen and Whitson-style characterization routes are correlations, not measurements. Results can be sensitive to plus-fraction molar mass and density, TBP correlation, number and boundaries of lumped groups, equation of state, mixing rule, and volume translation.

Before relying on a characterized fluid:

  1. validate the input assay, composition closure, units, and sample provenance;
  2. compare saturation pressure, density, relative volume, liquid dropout, separator yields, and viscosity with representative laboratory data;
  3. rerun sensitivity cases for plausible heavy-end and lumping choices;
  4. freeze the exact component slate and fitted parameters used downstream; and
  5. obtain independent PVT/model review for reserves, facilities, flow assurance, custody, or safety-critical decisions.

TP flash rejects a present TBP or plus fraction with a non-finite acentric factor or a factor at or below -1 before changing its inventories. By definition, the acentric factor is $\omega = -\log_{10}(P_{sat}(0.7T_c)/P_c)-1$; a factor at or below -1 implies a saturation pressure at or above the critical pressure below the critical temperature. This guard preserves explicit failure for the legacy Twu sour-fluid fixture with an acentric factor of approximately -38.45. It does not establish that other fitted petroleum-fraction properties are qualified; retain the assay and laboratory validation steps above.

Asphaltene, wax, hydrate, electrolyte, and solid-phase behavior require their own qualified models and data. A characterized hydrocarbon slate does not establish those risks.