Thermodynamic property packages
The property method sets your VLE, flash and enthalpies — pick the right one and the rest of the simulation follows. MaximaLabs ships the packages engineers actually use, from cubic equations of state to electrolyte and reference models. Choose one, then run flash, a distillation column or a full flowsheet — free in your browser, or from the Python SDK.
NRTL
Activity-coefficient model for polar / alcohol-water VLE (azeotropes).
Explore packageNRTL + vapor association
NRTL liquid activity with chemical-theory vapor dimerization — carboxylic acids (acetic/formic), where the vapor phase associates and a plain NRTL/cubic gets the VLE wrong. Reduces to NRTL for non-associating components.
Explore packagePeng-Robinson (EoS)
Cubic equation of state for hydrocarbons / light gases and high pressure.
Explore packageSoave-Redlich-Kwong (EoS)
Cubic EoS for hydrocarbons / light gases; an alternative to Peng-Robinson.
Explore packageChao-Seader (K-value)
Semi-empirical K-value method for hydrocarbon / hydrogen-rich systems (hydrotreaters, reformers, hydrocrackers) — Curl-Pitzer liquid fugacity, regular-solution activity, SRK vapor fugacity (Chao-Seader 1961). The Grayson-Streed H2/CH4 special coefficients are a follow-up; light-gas K is approximate. SRK enthalpy/density.
Explore packagePC-SAFT (EoS)
Molecular SAFT EoS (chain + dispersion + 2B association) — long-chain alkanes, CO₂/light gases, high-pressure gas solubility, and hydrogen-bonding fluids (water, alcohols) via association.
Explore packageCPA (Cubic-Plus-Association)
SRK cubic + Wertheim association — natural-gas water content / dehydration, methanol & glycol hydrate-inhibitor injection, and aqueous-alcohol systems. Reduces to SRK for non-associating components; association parameters for water/methanol/ethanol.
Explore packageSAFT-VR Mie (EoS)
Full Mie-segment perturbation SAFT (a1/a2/a3 monomer + TPT1 chain; Lafitte 2013). Non-associating fluids only; cited parameters for methane (validated to NIST saturated-liquid density / vapour pressure). A specialist/validation package.
Explore packageUNIFAC (predictive)
Group-contribution activity model: predicts liquid non-ideality from a molecule's functional groups, so a binary with NO regressed parameters still gets real physics. Use when your pair is not in the fitted NRTL/UNIQUAC tables — and prefer a fitted binary where one exists.
Explore packageModified UNIFAC (Dortmund) — predictive
The refined UNIFAC: a ¾-power combinatorial term plus temperature-dependent group interactions (Weidlich & Gmehling 1987 + Gmehling-group revisions). Markedly better than original UNIFAC in the DILUTE limit — mean infinite-dilution activity-coefficient error 16% vs 42% against experiment — which is the regime absorbers, trace removal and solvent selection turn on. About the same on bulk VLE, so it is offered, not defaulted.
Explore packageIonic liquid (non-volatile entrainer)
Modified-UNIFAC (Dortmund) activity model extended with ionic-liquid group decompositions, where the IL is treated as INVOLATILE (K = 0) in the flash. For extractive-distillation entrainer screening — an IL such as [EMIM][BF4] binds water and raises ethanol's relative volatility ~2.8×, breaking the azeotrope. Components use the IL ids emim_bf4 / bmim_bf4 / hmim_bf4 / bmpy_bf4 (bmim_pf6 for non-aqueous). Bounded: the phase split is validated; the IL's pure-component enthalpy is omitted (screening), so a rigorous IL-column energy balance is a follow-up.
Explore packagePeng-Robinson + MHV1 (Gᴱ mixing)
Cubic EoS whose a_mix comes from NRTL's excess Gibbs energy instead of a kij — the package for polar / hydrogen-bonding mixtures AT PRESSURE (alcohol-water, water-amine), where a kij cannot describe the non-ideality and an activity model cannot handle a supercritical component. MHV1 matches Gᴱ at zero pressure, so the solver's existing low-pressure NRTL binaries are the correct parameters with no refit.
Explore packageSoave-Redlich-Kwong + MHV1 (Gᴱ mixing)
The same zero-pressure Gᴱ mixing rule on SRK rather than Peng-Robinson, with Michelsen's SRK-specific q₁ = −0.593.
Explore packagePSRK (predictive SRK)
SRK + MHV1 + UNIFAC (Holderbaum & Gmehling 1991, q₁ = −0.64663) — a fully PREDICTIVE high-pressure package: UNIFAC supplies the activity coefficients from functional groups, so a binary with no regressed parameters still gets real non-ideality. Use when the pair you care about is not in the NRTL tables.
Explore packagePeng-Robinson + Huron-Vidal (Gᴱ mixing)
The original infinite-pressure Gᴱ mixing rule (Huron & Vidal 1979) on Peng-Robinson. Rigorously it wants activity parameters regressed AT infinite pressure; driven by the bundled low-pressure NRTL binaries it is markedly better than a plain kij but measurably worse than MHV1 — prefer pr-mhv1 unless you are supplying your own infinite-pressure-fitted parameters.
Explore packagePeng-Robinson + Wong-Sandler (Gᴱ mixing)
The one Gᴱ rule that also reproduces the theoretically correct quadratic composition dependence of the second virial coefficient, so it stays sound at low density as well as high. Carries a kij alongside the activity model; like Huron-Vidal it performs best with parameters refit for it rather than the bundled low-pressure binaries.
Explore packageUNIQUAC
Activity-coefficient model for polar / alcohol-water VLE (azeotropes).
Explore packageWilson
Local-composition activity model for miscible polar / alcohol-water VLE (azeotropes). Cannot represent liquid-liquid splitting — use NRTL/UNIQUAC for LLE.
Explore packageVan Laar
Two-constant, temperature-independent regular-solution activity model for miscible polar / alcohol-water VLE — the oldest of the family and cheap to fit, at the cost of accuracy over a wide temperature range. Cannot represent liquid-liquid splitting — use NRTL/UNIQUAC for LLE.
Explore packageElectrolyte NRTL (CO₂–MEA)
Bounded electrolyte model for CO₂ capture VLE in aqueous MEA (carbamate/bicarbonate speciation + Davies activity).
Explore packageElectrolyte NRTL (CO₂–AMP)
Same speciation model with 2-amino-2-methyl-1-propanol (AMP) constants — a hindered amine with a much weaker carbamate, so it can be pushed past MEA's ~0.5 mol CO₂/mol amine ceiling. Screening-grade: no independently published VLE dataset validates this system (unlike MEA's Jou-Otto-Mather anchor).
Explore packageElectrolyte NRTL (CO₂–DEA)
Same speciation model with diethanolamine (DEA) constants — a canonical secondary gas-treating amine. Its protonation anchor is genuinely cited (pKa 8.883, Bower-Robinson-Bates 1962/NBS); the carbamate stability is a disclosed engineering estimate calibrated to DEA's cited operating range, so treat capacity as screening-grade and basicity as cited-grade.
Explore packageElectrolyte NRTL (CO₂–MDEA)
Same speciation model with N-methyldiethanolamine (MDEA) constants — a tertiary amine that forms no carbamate (proton-acceptor mechanism only). Screening-grade single-amine MDEA; the mixed MDEA/PZ blend is the richer acid-gas-treating package.
Explore packageElectrolyte NRTL (CO₂+H₂S — MDEA/PZ)
Bounded electrolyte model for CO₂ AND H₂S over aqueous MDEA promoted with piperazine (mixed-amine acid-gas treating): protonation + PZ carbamate/dicarbamate + bicarbonate/bisulfide speciation with Davies activity. Equilibrium capacity/selectivity only — NOT rate-based (PZ's kinetic CO₂ promotion and MDEA's kinetic H₂S selectivity are not modeled); the PZ carbamate constants and H₂S path are screening-grade.
Explore packageSour water (NH₃ + H₂S)
Bounded weak-electrolyte model for refinery sour water — ammonia and hydrogen sulfide over water, with the mutual suppression that defines the system: ammonia raises pH and holds H₂S down as bisulfide, H₂S lowers pH and holds ammonia down as ammonium, and heat reverses both so a stripper drives them off together. Equilibrium speciation with Davies activity; NH₃/H₂S only — a CO₂-bearing sour water is out of scope.
Explore packageGERG-2008 (natural gas / LNG)
Reference multiparameter Helmholtz EoS (GERG-2008-family, via CoolProp HEOS) for natural-gas / LNG mixtures — cryogenic density and phase behaviour several percent more accurate than the cubics. Natural-gas / light-hydrocarbon components only.
Explore packageIAPWS Steam Tables
IAPWS-95 water/steam properties for boilers, turbines and utility loops.
Explore packageCoolProp (ideal)
Pure-component / ideal-solution properties from CoolProp.
Explore packageFlory-Huggins (Polymer)
Polymer-solution activity model (single polymer + single volatile solvent) for devolatilization / residual-monomer stripping. Requires flowsheet.polymer_params.
Explore packageSanchez-Lacombe (Polymer EoS)
Multi-solvent lattice-fluid equation of state (one polymer + any number of volatile solvents) for polymer devolatilization / solvent recovery — the multi-solvent step up from Flory-Huggins. Requires flowsheet.polymer_params (caller-supplied r/P*/T*/ρ* characteristic parameters).
Explore packageSugar solution (Norrish)
Water + dissolved sugars (glucose, fructose, sucrose, maltose) with a real Norrish-derived boiling-point elevation — syrup evaporation in corn wet milling/HFCS, cane and beet refining. The non-electrolyte sibling of the brine package: without the BPE feedback a syrup evaporator is bimodal (never boils, or boils to dry solids).
Explore packageBrine (Pitzer electrolyte)
Water + dissolved salts (nacl, kcl, cacl2, mgcl2, na2so4, mgso4, na2co3) with a real Pitzer-derived boiling-point elevation — multi-effect evaporators, desalination, and hydrometallurgical/kraft-liquor brines.
Explore package