NeqSim provides three complementary approaches for mineral scale prediction:

  1. Rigorous EOS-based (CheckScalePotential) — Uses Electrolyte CPA or Pitzer activity coefficients from a full thermodynamic flash. Best accuracy, requires complete fluid definition.
  2. High-salinity standalone (PitzerScaleActivityModel + MultiMineralScaleEquilibrium) — Uses published binary Pitzer parameters and coupled shared-ion precipitation for NaCl-dominated brines.
  3. Empirical standalone (ScalePredictionCalculator) — Uses Davies equation with ion pairing corrections. Fast dilute-brine screening, needs only water analysis data.

Both approaches share improved Ksp correlations, pressure corrections, and T-dependent parameters.

Architecture Overview

┌─────────────────────────────────────────────────────┐
│               Scale Prediction in NeqSim            │
├──────────────────────┬──────────────────────────────┤
│   Rigorous (EOS)     │     Empirical (Standalone)   │
│                      │                              │
│  SystemElectrolyte-  │  ScalePredictionCalculator   │
│  CPAstatoil          │  ├── Davies equation         │
│  ├── TPflash()       │  ├── Ion pairing (6 pairs)   │
│  ├── ion activities  │  ├── T-dep Ksp (Monnin etc.) │
│  └── CheckScale-     │  ├── Pressure correction ΔV° │
│      Potential       │  └── Debye-Hückel A(T)       │
│                      │                              │
│  SystemPitzer        │  Helper Classes              │
│  ├── PhasePitzer     │  ├── ScaleMassCalculator     │
│  ├── Pitzer params   │  ├── BariteCelestiteSolid-   │
│  └── T-dep β₀,β₁,Cᶲ │  │   Solution                │
│                      │  ├── WaterCompatibility-      │
│                      │  │   Screener                 │
│                      │  └── FlowlineScaleProfile    │
└──────────────────────┴──────────────────────────────┘

1. ScalePredictionCalculator (Empirical)

Package: neqsim.pvtsimulation.flowassurance

Fast standalone calculator for 5 mineral scales using water analysis input.

Setup and Usage

ScalePredictionCalculator calc = new ScalePredictionCalculator();
calc.setCalciumConcentration(400.0);     // mg/L
calc.setBariumConcentration(10.0);       // mg/L
calc.setStrontiumConcentration(5.0);     // mg/L
calc.setIronConcentration(2.0);          // mg/L
calc.setMagnesiumConcentration(1300.0);  // mg/L (for ion pairing)
calc.setSodiumConcentration(11000.0);    // mg/L (for ion pairing)
calc.setBicarbonateConcentration(150.0); // mg/L
calc.setSulphateConcentration(10.0);     // mg/L
calc.setTotalDissolvedSolids(35000.0);   // mg/L
calc.setTemperatureCelsius(80.0);        // °C
calc.setPressureBara(100.0);             // bara
calc.setCO2PartialPressure(2.0);         // bar
calc.enableAutoPH();                     // estimate pH from CO2
calc.calculate();

double siCaCO3 = calc.getCaCO3SaturationIndex();
double siBaSO4 = calc.getBaSO4SaturationIndex();
boolean risk = calc.hasScalingRisk();
String json = calc.toJson();

Python Usage

from neqsim import jneqsim
ScalePredictionCalculator = jneqsim.pvtsimulation.flowassurance.ScalePredictionCalculator

calc = ScalePredictionCalculator()
calc.setCalciumConcentration(400.0)
calc.setBariumConcentration(10.0)
calc.setStrontiumConcentration(5.0)
calc.setBicarbonateConcentration(150.0)
calc.setSulphateConcentration(10.0)
calc.setTotalDissolvedSolids(35000.0)
calc.setTemperatureCelsius(80.0)
calc.setPressureBara(100.0)
calc.setCO2PartialPressure(2.0)
calc.enableAutoPH()
calc.calculate()

print(f"CaCO3 SI = {calc.getCaCO3SaturationIndex():.3f}")
print(f"BaSO4 SI = {calc.getBaSO4SaturationIndex():.3f}")

Ksp Correlations

Scale Source Formula
CaCO3 (calcite) Plummer & Busenberg (1982) $\log_{10} K_{sp} = -171.9065 - 0.077993T + 2839.319/T + 71.595\log_{10} T$
BaSO4 (barite) Monnin (1999) $\log_{10} K_{sp} = 136.035 - 7680.41/T - 48.595\log_{10} T$
SrSO4 (celestite) Monnin (1999) $\log_{10} K_{sp} = 155.889 - 7862.38/T - 56.625\log_{10} T$
CaSO4 (anhydrite) Blount & Dickson (1973) $\log_{10} K_{sp} = 85.685 - 4279.82/T - 30.219\log_{10} T$
FeCO3 (siderite) Greenberg & Tomson (1992) $\log_{10} K_{sp} = -59.3498 - 0.041377T + 2.1963/T + 24.5724\log_{10} T + 2.518 \times 10^{-5} T^2$

Pressure Correction

All Ksp values are corrected for pressure using molar volume change:

\[\ln\frac{K_{sp}(P)}{K_{sp}(P_0)} = -\frac{\Delta V^\circ (P - P_0)}{RT}\]

Where $\Delta V^\circ$ (cm³/mol) values are: CaCO3 = -58.4, BaSO4 = -46.4, SrSO4 = -47.0, CaSO4 = -52.4, FeCO3 = -52.9.

Ion Pairing

Six aqueous ion pairs are modelled with association constants (25°C):

Ion Pair $\log_{10} K_{assoc}$ Effect
CaSO4⁰ 2.31 Reduces free Ca²⁺ and SO4²⁻
MgSO4⁰ 2.37 Reduces free Mg²⁺ and SO4²⁻
NaSO4⁻ 0.70 Reduces free Na⁺ and SO4²⁻
CaHCO3⁺ 1.11 Reduces free Ca²⁺ and HCO3⁻
MgHCO3⁺ 1.16 Reduces free Mg²⁺ and HCO3⁻
CaCO3⁰ 3.22 Reduces free Ca²⁺ and CO3²⁻

Temperature dependence uses van’t Hoff correction with $\Delta H / R$ coefficients.

2. CheckScalePotential (EOS-Based)

Package: neqsim.thermodynamicoperations.flashops.saturationops

Uses full thermodynamic model (Electrolyte CPA or Pitzer) to compute activity-based ion activity products.

Usage via ThermodynamicOperations

// Create electrolyte system
SystemInterface brine = new SystemElectrolyteCPAstatoil(273.15 + 80.0, 100.0);
brine.addComponent("CO2", 0.01);
brine.addComponent("water", 0.90);
brine.addComponent("Na+", 0.03);
brine.addComponent("Cl-", 0.035);
brine.addComponent("Ca++", 0.005);
brine.addComponent("Ba++", 0.001);
brine.addComponent("SO4--", 0.002);
brine.addComponent("HCO3-", 0.005);
brine.chemicalReactionInit();
brine.createDatabase(true);
brine.setMixingRule(10);
brine.setMultiPhaseCheck(true);

ThermodynamicOperations ops = new ThermodynamicOperations(brine);
ops.TPflash();
brine.initProperties();

// Check scale potential
int aqPhase = brine.getPhaseNumberOfPhase("aqueous");
ops.checkScalePotential(aqPhase);
String[][] table = ops.getResultTable();

// table rows: [saltName, scalePotentialFactor, ""]
// scalePotentialFactor > 1.0 means supersaturated
for (int i = 1; i < table.length; i++) {
    System.out.println(table[i][0] + " : " + table[i][1]);
}

Python Usage

from neqsim import jneqsim
SystemElectrolyteCPAstatoil = jneqsim.thermo.system.SystemElectrolyteCPAstatoil
ThermodynamicOperations = jneqsim.thermodynamicoperations.ThermodynamicOperations

brine = SystemElectrolyteCPAstatoil(273.15 + 80.0, 100.0)
brine.addComponent("CO2", 0.01)
brine.addComponent("water", 0.90)
brine.addComponent("Na+", 0.03)
brine.addComponent("Cl-", 0.035)
brine.addComponent("Ca++", 0.005)
brine.addComponent("SO4--", 0.002)
brine.addComponent("HCO3-", 0.005)
brine.chemicalReactionInit()
brine.createDatabase(True)
brine.setMixingRule(10)
brine.setMultiPhaseCheck(True)

ops = ThermodynamicOperations(brine)
ops.TPflash()
brine.initProperties()

aq_phase = brine.getPhaseNumberOfPhase("aqueous")
ops.checkScalePotential(aq_phase)
table = ops.getResultTable()

for i in range(1, len(table)):
    print(f"{table[i][0]:20s} SR = {table[i][1]}")

Key Features

Dissolved-salt saturation diagnostics

calcSaltSaturation(String) adds dissociated salt formula units until the aqueous ion activity product reaches the COMPSALT solubility product. Use calcSaltSaturationWithDiagnostics(String) for the same system mutation plus immutable convergence and work diagnostics:

SystemPitzer fluid = new SystemPitzer(404.15, 995.0);
fluid.addComponent("water", 55.508);
fluid.setMixingRule("classic");

ThermodynamicOperations operations = new ThermodynamicOperations(fluid);
SaltSaturationResult result =
    operations.calcSaltSaturationWithDiagnostics("CaSO4_A");

double initialSaturationRatio = result.getInitialSaturationRatio();
double finalSaturationRatio = result.getFinalSaturationRatio();
double addedFormulaUnitsMol = result.getAddedSaltMoles();
int fullThermodynamicInitializations = result.getThermodynamicInitializationCount();
boolean converged = result.isConverged();

The result also reports bracket and bisection evaluation counts, whether the input was already saturated, whether the solve reached its iteration limit, and the absolute residual from unit saturation. These diagnostics expose the existing calculation’s work without adding flashes or changing its 1e-6 saturation-ratio tolerance. They do not create a solid phase, describe precipitated mass, or qualify the COMPSALT, Ksp, reaction-volume, or activity-model parameters.

Activity-consistent pure-mineral precipitation

ThermodynamicOperations.precipitateScale(String) removes one named COMPSALT mineral stoichiometrically until its aqueous activity saturation ratio is one. It returns an immutable SaltPrecipitationResult; the thermodynamic system is the residual gas/oil/aqueous fluid, while the result is the corresponding pure-solid material ledger. The solid is deliberately not inserted as a NeqSim phase.

SystemPitzer fluid = new SystemPitzer(298.15, 50.0);
fluid.addComponent("water", 55.508);
fluid.addComponent("Na+", 1.0);
fluid.addComponent("Ca++", 0.2);
fluid.addComponent("Mg++", 0.15);
fluid.addComponent("Cl-", 1.3);
fluid.addComponent("SO4--", 0.2);
fluid.setMixingRule("classic");
fluid.init(0); // Automatically selects the complete PHREEQC Pitzer catalog topology.
fluid.setMultiPhaseCheck(true);

ThermodynamicOperations operations = new ThermodynamicOperations(fluid);
SaltPrecipitationResult solid = operations.precipitateScale("CaSO4_A");

double residualSaturationRatio = solid.getFinalSaturationRatio();
double solidAmountMol = solid.getPrecipitatedMoles();
double solidMassGram = solid.getPrecipitatedMassGrams();
double materialResidualMol = solid.getMaximumIonBalanceResidualMoles();

The default SystemPitzer policy automatically selects the bundled PHREEQC catalog when every interaction required by the active aqueous topology is present; this complete Ca/Mg/Cl/SO4 example needs no manual dataset-selection call. Missing mixed interactions still fail closed rather than becoming zero. useLegacyPitzerParameters() is an explicit compatibility opt-out that must be called before the first property evaluation.

The operation uses the selected aqueous activity model without transferring Pitzer parameters into the mineral-reaction database. Every trial extent is evaluated on a fresh clone; the accepted composition is reflashed and physical properties are reinitialised. Consequently the residual fluid can continue through Stream, Heater, and ProcessSystem calculations. Ions remain in the aqueous phase, while gas and oil phases retain their EOS roles.

Callers should require a complementarity residual such as solid.getComplementarityViolation() <= 1e-5 and independently check total and elemental balances.

Simultaneous competing pure minerals

ThermodynamicOperations.precipitateScales(String...) enforces non-negative pure-solid amounts and aqueous saturation complementarity for several named COMPSALT minerals. The active set precipitates supersaturated minerals and redissolves an undersaturated present mineral. Names are sorted before iteration, so caller ordering cannot select a different solid topology.

MultiSaltPrecipitationResult scales =
    operations.precipitateScales("CaSO4_A", "CaSO4_G");

SaltPrecipitationResult anhydrite = scales.getMineralResult("CaSO4_A");
SaltPrecipitationResult gypsumCorrelation = scales.getMineralResult("CaSO4_G");
double complementarity = scales.getMaximumComplementarityViolation();
double componentBalanceMol = scales.getMaximumComponentBalanceResidualMoles();

The returned ledger is absolute for that call and is not inserted into the NeqSim phase list. Carry it with the residual fluid. After a heater, pressure change, dilution, or composition change, pass the same ledger back so available solids can dissolve as well as precipitate:

MultiSaltPrecipitationResult updatedScales =
    new ThermodynamicOperations(changedFluid).equilibrateScales(scales);

The continuation API conserves dissolved plus ledgered formula units and fails closed if the bounded active set cannot reach 1e-6 log10-SR complementarity. Non-reactive component and element ledgers use a 1e-10 mol absolute balance tolerance. Reactive element ledgers use the same absolute floor plus a 1e-8 relative tolerance on each element inventory, matching the numerical closure of the chemical-equilibrium solver without weakening trace-element checks. The result reports both the maximum absolute residual and the maximum residual normalized by its quantity-specific tolerance; the normalized value must not exceed one. It also reports the update count, maximum complementarity violation, per-mineral saturation ratios and non-negative amounts. It is serializable and its mineral map is defensively immutable for Java and Python process workflows. A thermodynamic system remains mutable and must not be shared between threads. Sequential clones are deterministic, and independently constructed systems are covered by a parallel determinism regression. Concurrent use of clones from one Pitzer instance is not qualified because current shared Pitzer internals can race; construct each parallel system independently until that separate prerequisite is resolved.

The calcium-sulfate COMPSALT rows distinguish anhydrite CaSO4_A from gypsum CaSO4_G. Gypsum explicitly carries two waters per formula unit: its saturation ratio includes the active aqueous model’s solvent activity squared, precipitation removes two moles of water into each mole of solid, dissolution returns them, and the solid ledger reports the hydrated molar mass. Anhydrite remains the water-free CaSO4 reaction. The implementation follows the PHREEQC 3.9.0 reactions CaSO4 = Ca+2 + SO4-2 and CaSO4:2H2O = Ca+2 + SO4-2 + 2 H2O from the USGS public-domain pitzer.dat catalog at commit b0b3be767158ccc3322d2c816625cf470045e67e.

A separate dilution continuation proves that an existing solid can redissolve without stale-state carryover. These regressions establish standard-state and material-ledger semantics; they do not independently validate the COMPSALT temperature/pressure correlations. In particular, the current COMPSALT pressure-volume coefficients differ from PHREEQC mineral molar volumes, so quantitative high-pressure gypsum/anhydrite phase-boundary use remains outside the qualified scope.

Calcium-sulfate phase-boundary evidence

Use ThermodynamicOperations.qualifyCalciumSulfatePhaseBoundary() to inspect the mineral-standard-state evidence separately from Pitzer or electrolyte-CPA aqueous parameter qualification:

CalciumSulfatePhaseBoundaryQualification evidence =
    operations.qualifyCalciumSulfatePhaseBoundary();

double transitionC = evidence.getPredictedPureWaterTransitionCelsius();
double requiredWaterActivity25C = evidence.getRequiredWaterActivityAt25Celsius();
boolean publicationReady = evidence.isPublicationReady();

The immutable Java result is available through JPype and uses the exact COMPSALT solubility-product calculation used by precipitation. For simultaneous anhydrite and gypsum equilibrium, the ion activities cancel and the required water activity is $a_{\mathrm{w}}=\sqrt{K_{sp,\mathrm{G}}/K_{sp,\mathrm{A}}}$. The registered independent evidence is Voigt and Freyer (2023), DOI 10.3389/fnuen.2023.1208582, CC BY 4.0:

Observable Independent envelope Current COMPSALT result
Pure-water transition 42 ± 1 °C 60.445190 °C
NaCl crossing at 25 °C $a_{\mathrm{w}}=0.8551$–$0.8634$ $a_{\mathrm{w}}=0.7736299$
NaCl crossing at 40 °C $a_{\mathrm{w}}=0.9370$–$0.9587$ $a_{\mathrm{w}}=0.8437837$

The NaCl intervals retain Bock (1961), DOI 10.1139/v61-228, as their primary experimental lineage without redistributing the copyrighted primary table. The current correlations fail all three registered envelopes, so the result deliberately reports REJECTED and publicationReady=false; it does not tune a coefficient toward the evidence. Absolute-solubility residuals, primary-row uncertainty, mixed-brine qualification and the high-pressure reaction-volume convention remain explicit follow-on boundaries.

The process-system test carries the solid ledger beside the residual fluid through a Stream -> Heater -> ProcessSystem calculation, then re-equilibrates it at the outlet. Its charge-balanced feed contains nonzero Ca++, Mg++, Cl-, and SO4–; Ca/SO4 close against the solid ledger while Mg/Cl remain unchanged spectators. Fluid-phase density, enthalpy and heat capacity remain finite and ions remain aqueous. This exercises the complete four-ion PHREEQC topology but does not by itself qualify quaternary mixed-brine observables. Solid density, enthalpy, heat capacity and heat of precipitation are not yet represented, so rigorous process energy balances with a material solid stream remain a separate model/property boundary.

Neither pure-mineral API solves solid solutions, nucleation, kinetics, deposition, or inhibitor performance. A Pitzer calculation must first select a parameter dataset complete for its active aqueous topology. Missing binary, same-sign, ternary, or neutral interactions remain an error; they are not silently set to zero.

SaltPrecipitationPerformanceBenchmark records explicit-operation cost separately from the neutral control. On OpenJDK 17 in the development container, its median fresh-system calculations were 78.9 ms for aqueous anhydrite precipitation and 1.018 s for the complete gas-oil-aqueous case. The unchanged neutral SRK control measured 0.215 ms before and 0.055 ms after the Pitzer batches (ratio 0.254, reflecting JIT warmup rather than a regression). This operation is invoked only by precipitateScale or precipitateScales; neutral PR/SRK/CPA calculations execute no new branch or allocation.

With simultaneous CaSO4_A/CaSO4_G enabled in the same benchmark, median fresh-system times were 80.5 ms for the aqueous calculation and 1.206 s for gas-oil-aqueous. The active set required one solid update, reached 9.995e-9 maximum log10-SR complementarity violation, and closed the component ledger to reported machine zero. The unchanged neutral SRK control measured 0.205 ms before and 0.049 ms after the electrolyte batches (ratio 0.238, JIT dominated). Timings are diagnostic and not portable hardware guarantees.

Reaction-level saturation diagnostics

ChemicalReaction.getSaturationRatio(system, phaseNumber) evaluates the same thermodynamic definition for reaction-backed minerals:

$\mathrm{SR}=\frac{\mathrm{IAP}}{K_{sp}},\quad \mathrm{IAP}=\prod_i a_i^{-\nu_i}$

Here $\nu_i<0$ identifies each dissolved reactant and $a_i$ is its dimensionless activity. Electrolyte-CPA uses its established mole-fraction/activity-coefficient convention; Pitzer uses solute molality and molality-scale activity coefficients while retaining the solvent convention. calcLogSaturationRatio(...) returns $\ln(\mathrm{SR})$ directly for trace systems where the linear ratio may underflow.

This activity-based definition is consistent with USGS PHREEQC saturation-index reporting and the calcite equilibrium treatment of Plummer and Busenberg (1982), DOI 10.1016/0016-7037(82)90056-4. The regression uses synthetic compositions and an analytical identity; it copies or fits no external numerical data.

3. Pitzer Activity Coefficient Model

Package: neqsim.thermo.phase.PhasePitzer, neqsim.thermo.component.ComponentGePitzer

Enhanced Pitzer model with temperature-dependent parameters loaded from database.

Temperature-Dependent Debye-Hückel

The Debye-Hückel osmotic coefficient $A_\phi$ is computed from:

\[A_\phi = \frac{1.4006 \times 10^6 \sqrt{\rho_w}}{(\varepsilon \cdot T)^{3/2}}\]

Where $\rho_w$ uses Kell (1975) water density and $\varepsilon$ uses Archer & Wang (1990) dielectric constant.

Temperature-Dependent Binary Parameters

Parameters follow the form:

\[\beta(T) = \beta_{25} + T_1 \left(\frac{1}{T} - \frac{1}{298.15}\right) + T_2 \ln\left(\frac{T}{298.15}\right)\]

The Pitzer parameter database (PitzerParameters.csv) currently contains 30 cation-anion rows. Of these, 23 non-estimated rows have populated binary parameters and are covered by regression tests for database loading plus finite mean ionic activity and osmotic coefficients. The covered ions include Na, K, Ca, Mg, Ba, Sr, Fe and H with Cl, SO4, HCO3, CO3 and OH.

At 298.15 K and 1.01325 bara, NaCl has a separate public reference validation against the traceable recommended values in Tables 6 and 10 of Partanen and Partanen (2020), DOI 10.1021/acs.jced.0c00402, licensed CC BY 4.0. The validation covers 0.2, 0.5 and 1.0 mol/kg water; 2.0 and 3.0 mol/kg water are retained as a concentrated hold-out. Pointwise acceptance limits are 2% relative for the mean molal activity coefficient and 0.75% relative for the osmotic coefficient. The tabulated values are rounded to 0.001 and were derived by the authors from traceable electrochemical, isopiestic, vapor-pressure and solubility evidence. NeqSim parameters are not fitted or changed by this validation.

This evidence applies only to binary NaCl(aq) at 298.15 K on the molality standard state. It does not validate temperature dependence, mixed salts, carbonate speciation, mineral parameters, precipitation complementarity or transfer of Pitzer parameters to electrolyte EOS models.

Using the Pitzer Model

SystemInterface pitzer = new SystemPitzer(313.15, 50.0);
pitzer.addComponent("methane", 5.0);
pitzer.addComponent("CO2", 0.05);
pitzer.addComponent("n-heptane", 2.0);
pitzer.addComponent("water", 55.5);
pitzer.addComponent("Ca++", 6.0e-4);
pitzer.addComponent("Cl-", 2.0e-4);
pitzer.addComponent("HCO3-", 1.0e-3);
pitzer.chemicalReactionInit();
pitzer.createDatabase(true);
pitzer.setMixingRule("classic");
pitzer.setMultiPhaseCheck(true);

ThermodynamicOperations ops = new ThermodynamicOperations(pitzer);
ops.TPflash();
pitzer.initProperties();

double calciteScalePotential = ops.getRelativeScalePotential("CaCO3");

The example couples SRK gas and oil phases to Pitzer aqueous chemistry. Remove n-heptane for a gas-aqueous case. Its ionic feed is electroneutral: 2 m(Ca++) = m(Cl-) + m(HCO3-). The primary-salt coverage topology is therefore the qualified binary Ca/Cl pair; bicarbonate remains part of the reactive carbonate subsystem.

Do not extend this fixture to a mixed primary salt by adding Na, Ba, Sr, Mg, sulfate, or another ion without defining the complete binary, same-sign theta, and ternary psi family from one convention-mapped dataset. PhasePitzer.getPitzerParameterCoverage() reports the exact missing tuples, and standard initialization fails closed when a mixed primary-salt topology is incomplete. An explicit zero is a scientific parameter definition, not a placeholder.

The returned value is the calcite saturation ratio, where values above one indicate thermodynamic supersaturation. It does not calculate precipitated mass or deposition kinetics. The fixed-role hybrid flash currently rejects explicit solid- and wax-phase checks.

Other electrolyte GE models

The reactive hybrid solver is selected through the HybridEosGeFlashModel contract; it does not contain a Pitzer type check. SystemDesmukhMather and SystemKentEisenberg provide the same SRK-gas / SRK-oil / GE-aqueous roles. For example, a Desmukh-Mather amine system can retain an oil phase, solve aqueous speciation and evaluate the same activity-based scale-potential API:

SystemDesmukhMather fluid = new SystemDesmukhMather(313.15, 5.0);
fluid.addComponent("methane", 5.0);
fluid.addComponent("CO2", 0.2);
fluid.addComponent("n-heptane", 2.0);
fluid.addComponent("MDEA", 1.0);
fluid.addComponent("water", 9.0);
fluid.addComponent("Ca++", 1.0e-4);
fluid.addComponent("Na+", 1.0e-3);
fluid.addComponent("Cl-", 2.0e-4);
fluid.addComponent("HCO3-", 1.0e-3);
fluid.chemicalReactionInit();
fluid.createDatabase(true);
fluid.setMixingRule("classic");
fluid.setMultiPhaseCheck(true);

ThermodynamicOperations operations = new ThermodynamicOperations(fluid);
operations.TPflash();
double calciteSaturationRatio = operations.getRelativeScalePotential("CaCO3");

Every SystemEosGE subclass can explicitly configure the same topology through enableHybridEosGeFlash(). This includes Wilson, NRTL, UNIFAC and specialised activity models, but it only supplies the phase-equilibrium topology: a model must still support water, the requested ions, aqueous reactions and appropriate interaction parameters before its scale result is meaningful. Pitzer remains the broadly parameterised choice for concentrated mineral-scale brines; Desmukh-Mather and Kent-Eisenberg are primarily amine/electrolyte screening models. SystemDuanSun is not included because its current system API accepts CO2 only and therefore cannot represent a gas-oil-aqueous electrolyte feed.

4. WaterCompatibilityScreener

Package: neqsim.pvtsimulation.flowassurance

Screens formation water and injection water for compatibility by evaluating scale risk at every mixing ratio.

Usage

WaterCompatibilityScreener screener = new WaterCompatibilityScreener();

// Formation water (high Ba, low SO4)
screener.setFormationWater(
    400,  // Ca mg/L
    200,  // Ba mg/L
    50,   // Sr mg/L
    2,    // Fe mg/L
    150,  // HCO3 mg/L
    10,   // SO4 mg/L
    50000,// TDS mg/L
    90,   // T °C
    200,  // P bara
    3.0,  // CO2 pp bar
    6.2   // pH
);

// Injection water (seawater: low Ba, high SO4)
screener.setInjectionWater(
    400, 0, 5, 0, 140, 2700, 35000, 15, 200, 0.3, 8.1
);

screener.setMixingRatios(new double[]{0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100});
screener.calculate();

System.out.println("Worst case: " + screener.getWorstCaseScale()
    + " at " + screener.getWorstCaseRatio() + "% IW, SI = "
    + screener.getWorstCaseSI());

5. FlowlineScaleProfile

Package: neqsim.pvtsimulation.flowassurance

Computes scale saturation indices along a pipeline with linearly varying T and P.

Usage

FlowlineScaleProfile profile = new FlowlineScaleProfile();
profile.setWaterChemistry(400, 10, 5, 0, 150, 10, 40000, 2.0, 6.5);
profile.setInletConditions(90.0, 250.0);   // wellhead: 90°C, 250 bara
profile.setOutletConditions(15.0, 50.0);   // host: 15°C, 50 bara
profile.setNumberOfSegments(50);
profile.calculate();

// Find worst location for each scale
System.out.println("Max CaCO3 SI: " + profile.getMaxSI("CaCO3"));
System.out.println("Max BaSO4 SI: " + profile.getMaxSI("BaSO4"));

String json = profile.toJson();

6. ScaleMassCalculator

Package: neqsim.pvtsimulation.flowassurance

Estimates the mass of scale precipitated per litre of produced water when SI > 0.

Usage

// First compute SIs
ScalePredictionCalculator calc = new ScalePredictionCalculator();
// ... set concentrations, T, P ...
calc.calculate();

// Then compute mass
ScaleMassCalculator massCal = new ScaleMassCalculator(calc);
massCal.setWaterVolume(1000.0);  // litres

// Individual mass calculations (mol/L inputs)
double caMolL = 400.0 / 40078.0;    // 400 mg/L Ca / MW
double co3MolL = 150.0 / 61017.0;   // approximate
double mass = massCal.calcCaCO3Mass(caMolL, co3MolL, calc.getCaCO3SaturationIndex());

7. BariteCelestiteSolidSolution

Package: neqsim.pvtsimulation.flowassurance

Regular solution model for (Ba,Sr)SO4 co-precipitation using Margules 1-parameter model.

At equilibrium:

\[x_{Ba} \gamma_{Ba}^{(s)} K_{sp,BaSO_4} = a_{Ba^{2+}} \cdot a_{SO_4^{2-}}\] \[x_{Sr} \gamma_{Sr}^{(s)} K_{sp,SrSO_4} = a_{Sr^{2+}} \cdot a_{SO_4^{2-}}\]

Where solid activity coefficients: $\ln \gamma_i^{(s)} = W (1 - x_i)^2$

Usage

BariteCelestiteSolidSolution ss = new BariteCelestiteSolidSolution();
ss.setAqueousActivities(0.001, 0.005, 0.01);  // aBa, aSr, aSO4
ss.setEndMemberKsp(1.08e-10, 3.44e-7);        // Ksp_BaSO4, Ksp_SrSO4
ss.setMargules(2.3);                           // W/(RT) parameter
ss.calculate();

System.out.println("BaSO4 in solid: " + (ss.getBaSO4MoleFraction() * 100) + "%");
System.out.println("Total SI: " + ss.getTotalSaturationIndex());

Coupled mineral-equilibrium diagnostics

MultiMineralScaleEquilibrium exposes the numerical evidence needed to decide whether a coupled precipitation result is acceptable:

The same values appear in the JSON diagnostics object. Acceptance remains case-specific: an iteration-limit flag is not itself a thermodynamic residual, and a small step does not prove that the mineral inequalities are satisfied. The reference regression requires maximum complementarity violation <= 1e-3 SI and maximum tracked-ion balance residual <= 1e-12 mol/L. A deliberately one-iteration solve remains material-balanced but fails complementarity, preventing silent classification as an equilibrated mineral state.

This contract follows the equilibrium-phase convention in Parkhurst (1995), U.S. Geological Survey Water-Resources Investigations Report 95-4227, doi:10.3133/wri954227: minerals in the stable phase assemblage satisfy the target SI equality, while absent minerals remain inequality constraints at or below the target. The current PHREEQC 3 EQUILIBRIUM_PHASES documentation states the same convention. USGS-authored information is public domain. No numerical data, parameter, correlation or PHREEQC code is copied or fitted by this diagnostic regression.

The diagnostic is limited to the standalone coupled solver’s tracked free ions and fixed 1:1 mineral stoichiometries. It does not prove elemental/charge closure for the predictor’s aqueous speciation, update activity coefficients during precipitation, select a globally stable phase topology, or validate Ksp and activity parameters against independent precipitation data. Full electrolyte EOS/GE calculations must continue to apply their model-specific activity, speciation and balance gates.

Comparison of thermodynamic routes

Feature Davies standalone Pitzer binary + coupled solids Electrolyte EOS / full Pitzer
Activity coefficients Charge-based Davies Ion-specific binary Pitzer Full fluid speciation/model
Ion pairing 6 explicit pairs 6 pairs plus trace-ion activity mapping Thermodynamic phase model
Input Water analysis Water analysis + salinity/molality Full fluid definition
Best for Dilute screening NaCl-dominated oilfield-brine screening Detailed design, complex brines
Ionic strength Approximately <=0.5 mol/kg Validated against NaCl data to 6 mol/kg Model/database dependent

See High-Salinity Mineral Scale and Production-Chemical Validation for published-data errors, treatment scenarios and limitations.