High-Salinity Mineral Scale and Production-Chemical Validation

This workflow combines five distinct engineering questions. Keeping the layers separate is essential: a scale inhibitor changes nucleation and growth, while pH adjustment changes carbonate thermodynamics, and an H2S scavenger consumes a reactive contaminant and may create a non-mineral spent product.

Layer NeqSim class Result
Aqueous screening chemistry ScalePredictionCalculator Free-ion concentrations and T/P-corrected Ksp
High-salinity activity PitzerScaleActivityModel Ion-specific activity coefficients for NaCl-dominated brines
Competing minerals MultiMineralScaleEquilibrium Coupled BaSO4, SrSO4, CaSO4, CaCO3 and FeCO3 precipitation
Chemical treatment ProductionChemicalScaleScenario Before/after pH, dissolved H2S and thermodynamic SI
Explainable diagnosis RootCauseAnalyser Ranked treatment-induced and untreated root-cause hypotheses

Choosing the activity model

MultiMineralScaleEquilibrium.ActivityModel provides three screening choices:

Model Recommended use Important limitation
DAVIES Dilute brines, approximately I <= 0.5 mol/kg Diverges at oilfield-brine salinity
BDOT Intermediate screening and comparison Empirical common divalent coefficient; not ion-specific
PITZER_BINARY NaCl-dominated high-salinity brines Binary/equivalent-I trace-ion mapping, not the full theta/psi mixture model

For compositionally complex brines, benchmark the screening result against PHREEQC with its Pitzer database or against a fully speciated NeqSim electrolyte-EOS calculation. Do not label PITZER_BINARY as full multicomponent Pitzer.

ScalePredictionCalculator water = new ScalePredictionCalculator();
water.setTemperatureCelsius(80.0);
water.setPressureBara(150.0);
water.setSodiumConcentration(70000.0);   // mg/L
water.setCalciumConcentration(3000.0);
water.setBariumConcentration(200.0);
water.setSulphateConcentration(1000.0);
water.setBicarbonateConcentration(500.0);
water.setTotalDissolvedSolids(150000.0);
water.setPH(7.0);

MultiMineralScaleEquilibrium equilibrium = new MultiMineralScaleEquilibrium(water)
    .setActivityModel(MultiMineralScaleEquilibrium.ActivityModel.PITZER_BINARY)
    .setPitzerIonicStrengthMolal(3.0) // density-corrected laboratory value, when available
    .solve();

The solver reports the coefficient used for each ion, each mineral’s cation-anion activity-coefficient product, the initial SI, equilibrium SI, precipitated mass, and residual shared-ion inventory.

Published-data activity benchmark

The regression dataset contains all 29 NaCl points from Hamer and Wu (1972), Table 16, from 0.001 to 6.0 mol/kg at 25 degrees C. The test evaluates the temperature-dependent PHREEQC binary parameters without fitting them to the benchmark.

Model RMSE in mean activity coefficient, m >= 0.5 mol/kg
Pitzer binary 0.000962
B-dot defaults 0.05049
Davies 1.00446

Across all 29 points, Pitzer mean absolute error is 0.000707 and maximum absolute error is 0.00336. The data are stored in src/test/resources/neqsim/pvtsimulation/flowassurance/nacl_mean_activity_hamer_wu_1972.csv and exercised by PitzerScaleActivityModelTest. Existing ScaleKspLiteratureBenchmarkTest separately checks the mineral solubility correlations, including the calcite value log10(Ksp) = -8.48 at 25 degrees C.

Primary sources:

Production-chemical scenarios

ProductionChemicalScaleScenario scenario = new ProductionChemicalScaleScenario()
    .setTemperatureCelsius(60.0)
    .setPressureBara(50.0)
    .setPH(6.5)
    .setCalciumMgL(2500.0)
    .setSodiumMgL(50000.0)
    .setSulphateMgL(500.0)
    .setBicarbonateMgL(1000.0)
    .setTotalDissolvedSolidsMgL(100000.0)
    .setDissolvedH2SMgL(100.0)
    .setActivityModel(MultiMineralScaleEquilibrium.ActivityModel.PITZER_BINARY)
    .addChemical(ProductionChemical.phStabilizer("MDEA", 1000.0))
    .addChemical(ProductionChemical.h2sScavenger("MEA-triazine", 1000.0))
    .addChemical(ProductionChemical.scaleInhibitor("phosphonate", 20.0));
scenario.evaluate();

The pH treatment solves a closed aqueous carbonate-alkalinity balance using active-product equivalents. NaOH, sodium carbonate, MDEA, MEA and HCl have explicit equivalent models. The H2S calculation uses active mass and practical stoichiometry for MEA/MMA triazine, iron and glyoxal. MEA-triazine also raises a spent-product/dithiazine evidence flag, consistent with the experimentally established stepwise H2S reaction chemistry.

The calculation does not infer a pH shift from an H2S scavenger’s neat-product pH. Commercial formulations and reaction products are vendor-specific. An unrecognised active remains in the audit report and raises an unsupported-model warning rather than receiving an assumed effect.

For a qualified proprietary product, override the family default with setH2SCapacityKgPerKgActive(...) or setAlkalinityCapacityMolEqPerKgActive(...). These values are included in the JSON audit trail.

Primary treatment sources:

Inhibitor semantics

ScaleControlAssessor now calls its empirical treatment output a kineticRiskIndex. The thermodynamic SI remains available through getThermodynamicSaturationIndex and is unchanged by inhibitor dose. Use MIC, dynamic tube-blocking, static bottle and vendor qualification data to decide whether a particular inhibitor controls a supersaturated brine.

Root-cause evidence

Attach the evaluated or configured scenario to the RCA:

RootCauseAnalyser rca = new RootCauseAnalyser();
rca.setChemicalTreatmentScenario(scenario);
rca.addSymptom(new Symptom(Symptom.Category.DEPOSIT,
    "Deposit downstream of pH stabiliser and scavenger injection"));
rca.analyse();

The analyser can add three explainable candidates:

These are competing hypotheses, not automatic conclusions. Close the RCA with deposit XRD/SEM/organic analysis, treated-water pH and alkalinity, actual chemical active concentration, injection history, temperature profile and water mixing history.