NeqSim provides general KineticReaction models, an experimental CO2ImpurityKineticReactor, and transport-timescale diagnostics. These APIs support screening; they do not supply a qualified kinetic dataset or apply reaction source terms to a pipeline.

Finite-rate results depend on temperature, absolute pressure, phase state, water availability, concentration and rate bases, and sometimes wall or catalyst material. A numerical rate is therefore not evidence that its parameters are suitable for engineering use.

What the two contracts do

KineticReactionQualification stores a source citation, stable source identifier, validation status, inclusive temperature and pressure ranges, and limitations. Its requireValidatedAt method rejects an unvalidated parameterization or a state outside those declared ranges.

KineticReactionDiagnostics.evaluate screens one initialized active phase and a caller-supplied residence time. For reactants with negative stoichiometric coefficients, it estimates the shortest consumption time from the absolute reaction rate and local concentration. It then reports

\[\mathrm{Da} = \frac{t_{\mathrm{transport}}}{t_{\mathrm{reaction}}}.\]

The diagnostics object does not call the qualification object. The caller must enforce qualification first, as the complete Java 8 helper below does.

Safe Java helper

Provide a separately constructed reaction and qualification whose names, source, rate basis, phase, composition range, and material limitations have been reviewed together. The thermodynamic system must already be flashed and initialized.

import org.apache.logging.log4j.LogManager;
import org.apache.logging.log4j.Logger;
import neqsim.process.equipment.reactor.KineticReaction;
import neqsim.process.equipment.reactor.KineticReactionDiagnostics;
import neqsim.process.equipment.reactor.KineticReactionQualification;
import neqsim.thermo.system.SystemInterface;

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

  private QualifiedKineticsScreen() {}

  public static KineticReactionDiagnostics evaluate(
      KineticReaction reaction,
      KineticReactionQualification qualification,
      SystemInterface fluid,
      int phaseIndex,
      double residenceTimeSeconds) {
    qualification.requireValidatedAt(fluid.getTemperature(), fluid.getPressure());

    KineticReactionDiagnostics diagnostic =
        KineticReactionDiagnostics.evaluate(
            reaction, fluid, phaseIndex, residenceTimeSeconds);

    logger.info(
        "reaction={} Da={} regime={} limitingReactant={}",
        diagnostic.getReactionName(),
        diagnostic.getDamkohlerNumber(),
        diagnostic.getRegime(),
        diagnostic.getLimitingReactant());
    return diagnostic;
  }
}

getTemperature() returns K and getPressure() returns bara, matching the qualification constructor. residenceTimeSeconds is in s and must be finite and non-negative. phaseIndex is an active phase index, not a phase-type identifier; resolve and validate the intended phase before calling the helper.

The qualification and reaction objects are not automatically bound to one another. Matching names do not prove that the rate parameters, stoichiometry, phase, or evidence belong together. Treat that association as project-owned provenance and verify it before calling evaluate.

Diagnostic semantics

The current diagnostic accepts only KineticReaction.RateBasis.VOLUME. Catalyst-mass and catalyst-area bases are rejected because a local catalyst loading or area is required to convert them to a volumetric consumption timescale.

Damköhler number Regime Screening interpretation
no finite consumption time INACTIVE no finite local conversion timescale is available
Da < 0.1 TRANSPORT_DOMINATED reaction is slow relative to residence time
0.1 <= Da <= 10 COUPLED reaction and transport occur on comparable timescales
Da > 10 REACTION_DOMINATED reaction is fast relative to residence time

The thresholds at exactly 0.1 and 10 are COUPLED. The reported rate keeps its sign, while the timescale uses its absolute value. A missing or zero-concentration reactant, zero rate, or otherwise non-positive limiting time produces INACTIVE, an infinite reaction time, and Da = 0. Non-finite rates, invalid phase indexes, and unsupported rate bases fail explicitly.

These labels do not establish that equilibrium chemistry is valid. They also do not update composition, conserve elements over a timestep, reflash the fluid, or solve charge balance.

Published aqueous CO2 hydration bridge

AqueousCO2HydrationKinetics implements the reversible aqueous reaction

\[\mathrm{CO_2(aq) + H_2O \rightleftharpoons H_2CO_3}\]

using the primary measurements and correlations of Soli and Byrne (2002), doi:10.1016/S0304-4203(02)00010-5:

\[\ln(k_H / \mathrm{s^{-1}}) = 22.66 - \frac{7799}{T}\]

and

\[\ln(k_D / \mathrm{s^{-1}}) = 30.15 - \frac{8018}{T}.\]

The implementation preserves the source’s narrow evidence boundary: 0.65 molal NaCl and 288.15–305.65 K. At 298.15 K it gives kH = 0.0302586 s-1, kD = 25.9844 s-1, and [H2CO3]/[CO2(aq)] = kH/kD = 0.00116449, or [CO2(aq)]/[H2CO3] = 858.744. The paper also reports a separately observed 25 °C ratio of 848. The 1.27% difference from the two fitted rate equations is a within-source consistency check, not an independent validation dataset. The implementation does not tune either correlation to remove the difference. Calls outside the published temperature range fail closed instead of extrapolating.

The equations are direct implementations of the primary-source fits, not a new NeqSim calibration or an independent validation dataset. Public DOI metadata and the institutional abstract are available without redistributing the paper’s tabulated data. The public abstract does not report uncertainty statistics for these two fitted equations, so exact equation-reproduction tests must not be interpreted as experimental prediction uncertainty.

createReaction() maps the two correlations exactly onto the generic reversible KineticReaction contract. Water has stoichiometric coefficient -1 but kinetic order zero because it is the solvent in the published pseudo-first-order model. advance(...) provides the analytical closed-batch update for the CO2/H2CO3 pair. The exact solution is non-negative and conserves that pair’s carbon without numerical timestep error.

Explicit/lumped electrolyte-speciation handoff

The finite-rate reaction distinguishes CO2(aq) and H2CO3, but NeqSim’s existing electrolyte reaction data does not contain a thermodynamic H2CO3 component. Its model-specific CO2water reaction uses one molecular CO2 inventory together with water, bicarbonate, and hydronium. The Pitzer row follows the molality-standard-state carbonate expressions in the public-domain USGS PHREEQC 3 database; electrolyte CPA retains its distinct STANDARD mole-fraction source.

collapseExplicitPairToLumpedCO2(...) therefore performs the conservative handoff

\[c_{\mathrm{CO_2^*}} = c_{\mathrm{CO_2(aq)}} + c_{\mathrm{H_2CO_3}},\]

before a caller initializes or reruns the selected electrolyte equilibrium model. partitionLumpedMolecularCO2(...) applies the source rate ratio to split that neutral pool again for reporting or to initialize the explicit kinetic pair. Both methods use mol/m3, reject negative or non-finite input, and expose a carbon-closure residual. They do not mutate a thermodynamic system, run a flash, or alter charge balance.

Do not add an explicit H2CO3 species beside the current CO2water reaction and do not substitute the kinetic ratio for the model-specific equilibrium constant. Either action would mix standard states and can count hydration twice. A future fully coupled solver needs an explicitly qualified thermodynamic H2CO3 species, true-ionization convention, phase properties, and reaction table.

This bridge does not qualify pressure dependence, dense-phase CO2, other salinities, acid dissociation, pH, charge balance, or phase transfer. Soli and Byrne did not vary pressure. Do not apply the correlation directly at pipeline pressure without separate evidence. Electrolyte speciation remains owned by issue #3144; pipeline source-term coupling remains owned by issue #3318.

Published neutral-pair residence-time screen

screenResidenceTime(...) compares a residence time with the exact relaxation time of the reversible CO2(aq)/H2CO3 pair:

\[\tau = \frac{1}{k_H+k_D}, \qquad \mathrm{Da}_{\mathrm{pair}} = \frac{t_{\mathrm{res}}}{\tau}, \qquad \frac{\delta(t)}{\delta(0)} = \exp(-\mathrm{Da}_{\mathrm{pair}}).\]

Here delta is any initial concentration deviation from the pair equilibrium. The result reports the remaining deviation, the relaxed fraction, and the same transport-dominated/coupled/ reaction-dominated thresholds as KineticReactionDiagnostics. timeToRemainingDeviationFraction provides the inverse analytical calculation.

At 298.15 K and 0.65 molal NaCl, the published equations give tau = 0.0384399 s. A residence time of 0.177022 s leaves 1% of an initial pair disequilibrium, while 1 s gives Da_pair = 26.0147 and leaves about 5.03e-12.

This rapid relaxation applies only to the neutral molecular pair and its small H2CO3 equilibrium fraction. It is not a bicarbonate/carbonate, pH, total dissolved inorganic carbon, gas-to-water mass-transfer, water-dropout, or corrosion-equilibration time. It also carries the same Soli–Byrne temperature, salinity, and pressure limitations; no van Eldik–Palmer pressure multiplier is applied automatically.

Published pressure response

AqueousCO2PressureKinetics implements the pressure derivative measured by van Eldik and Palmer (1982), doi:10.1007/BF00649292, for the hydration of aqueous CO2 and dehydration of carbonic acid at 298.15 K and ionic strength 0.5:

\[\frac{k(P_2)}{k(P_1)} = \exp\!\left[-\frac{\Delta V^{\ddagger}(P_2-P_1)}{RT}\right].\]

The reported activation volumes are -9.9 +/- 1.9 cm3/mol for hydration and +6.4 +/- 0.4 cm3/mol for dehydration. From 1 to 100 bara at 298.15 K, the implementation gives a nominal hydration multiplier of 1.04032877 and a nominal dehydration multiplier of 0.974764734. The corresponding uncertainty intervals are 1.03246477–1.04825267 and 0.973208843–0.976323112. Calls outside the deliberately narrow 298.15 K and 1–1000 bara gate fail closed.

Conway et al. (2016), doi:10.1071/CH15271, independently report stopped-flow high-pressure hydration measurements from 400 to 1000 atm. That public abstract is retained as corroborating provenance but is not used as a numerical parameter source.

These are dimensionless pressure multipliers, not absolute rate constants. Do not automatically multiply the Soli–Byrne absolute rates by these factors: the studies used different solution compositions, so doing so is a separate cross-dataset assumption requiring its own validation. This class does not mutate a fluid, select a phase, calculate speciation, or apply a pipeline source term.

Intended transport integration

A future control-volume integration should:

  1. solve and initialize the local pressure, temperature, and phase state;
  2. identify the active phase required by each reaction;
  3. verify reaction-to-evidence provenance and call requireValidatedAt with the local K/bara state;
  4. evaluate the local rate and reaction/transport timescale;
  5. apply a bounded, element-conserving reaction extent over the timestep;
  6. reflash and update electrolyte speciation and charge balance when an aqueous phase exists;
  7. retain reaction extent, source identity, range status, and conservation diagnostics.

Operator splitting requires timestep-refinement evidence. A Damköhler regime alone does not qualify a model, choose a timestep, establish phase consistency, or approve a transport design.

CO2 impurity chemistry boundary

The current CO2ImpurityKineticReactor default constants are experimental and illustrative. Do not use them as design correlations without independent evidence and qualification. The guide does not embed facility-specific data, kinetic constants, or a validated CO2/water, SOx, H2S, NOx/O2, or cross-impurity reaction set.

For equilibrium speciation, mineral saturation, or precipitation, use the electrolyte framework and verify its reaction set, standard states, charge balance, and applicability separately.