ProcessChem

Accuracy and Limits

About

This page says what ProcessChem is good at, what needs care, and what it does not calculate. Every statement rests on a comparison: the app's number beside a published value (steam tables, NIST reference equations, measured phase-equilibrium data, textbook worked examples) or a hand calculation. The comparison covers about 250 cases and is run again before each release.

In the app, the ⋯ menu on every screen and the ⓘ sheet beside a value both link here.

ProcessChem is a calculator for engineering estimates, teaching and checking. It is not a substitute for a validated process simulator or for measured data in safety-critical design: check relief sizing, design pressures and guarantees against an independent source.

How the app tells you

  • Amber warnings under a result name anything that weakens it: an estimated value, a model outside its range, a mixture model that does not suit the mixture.
  • Solved with: names the equation of state or correlation that produced the result.
  • ⓘ beside a value opens its source: the method, its source, the valid range and whether your temperature is inside it.
  • Methods used lets you switch method where there is a choice, to see how much the model moves the answer.
  • Method changes are said in a note: Peng-Robinson taking over from a reference equation, a corrected phase split, an estimated latent heat or solid heat capacity, a boiling point taken at 1 atm.
  • Balance checks run on every calculation: phase amounts must add back to the feed, and mixing must conserve energy. A failure shows as a warning instead of a silent wrong number. Stream tables state the mass balance under them ("Mass balance: in = out = …").
  • Errors are given in words and in your units, with the reason and what to try.

Good at

What Measured against Typical deviation
Pure fluids with a reference equation (water, CO₂, nitrogen, methane, propane, ammonia and more than 100 others) Steam tables, NIST Under 0.1 % on density, boiling point and heat of vaporisation
Heater and Heat Exchanger duty, 1 to 200 bar Real-fluid enthalpy difference Under 0.1 % (water, nitrogen, methane, CO₂, propane, ammonia, ethanol)
Light hydrocarbon and gas mixtures: bubble point, dew point, vapour composition Reference mixture equations 0.3 to 1.3 °C; vapour mole fraction within 0.01
Polar mixtures with fitted pair data (methanol/water, ethanol/water, benzene/toluene) Measured vapour–liquid data Bubble point within 1 °C; vapour mole fraction within 0.01; ethanol/water azeotrope at 88.7 mol % (measured 89.4)
Liquid density and other properties on the Property tab Handbook values 0.1 to 0.2 % for common liquids
Salt, acid and base solutions: density and heat capacity Handbook tables Under 0.5 % (NaCl and HCl up to 20 mass %)
Reaction heat, ΔG and equilibrium constant Textbook values K within 0.2 to 12 % (limestone calcination, steam reforming, ammonia synthesis)
Gas–solid equilibrium compositions at low to moderate pressure Measured compositions Ammonia at 450 °C and 10 atm: 2.08 mol % (measured 2.04)
Solids that sublime (dry ice, iodine, ice) Measured sublimation points and heats Sublimation point within 2 °C; heat within 2.5 %
Fluidized bed: minimum fluidization and terminal velocity Kunii & Levenspiel worked examples Under 0.5 %
Mass and energy balances Must close exactly Closed to 1 part in a million or better
Throttling steam, 40 to 1 bar Steam tables Within 0.2 °C

Use with care

What Why How far off
Liquid density from Peng-Robinson: any substance or mixture without a reference equation, on Phase Equilibrium and the sealed vessel A cubic equation is fitted to vapour pressure, not liquid volume 4 to 7 % for hydrocarbons; 10 to 17 % for water, ammonia, methanol, nitrogen and methane if you pick Peng-Robinson yourself. The Property tab uses measured liquid densities instead and is not affected.
Boiling point and heat of vaporisation without a reference equation Same cubic Usually within 1 to 2 °C and 5 % (aniline, styrene, tetrahydrofuran)
Natural-gas throttling Mixture is on Peng-Robinson Outlet 1.5 to 3 °C colder than the reference
Liquid-full vessel, mixture Liquid compressibility from Peng-Robinson Pressure rise about 20 % low (propane/butane). A single substance with a reference equation is exact.
Polar mixtures without fitted pair data UNIFAC estimates the interactions from chemical groups; marked Estimated An estimate: good for a first look, not for column design. With fitted pair data (NRTL) the tested azeotropes are within 0.02 mole fraction and 1 °C.
Chemical equilibrium above about 50 bar The gas is treated as ideal Ammonia synthesis: 5 % low at 100 atm, 13 % low at 300 atm
Reaction and equilibrium far outside a heat-capacity correlation's range The correlation is extrapolated; the app flags each species Indicative only, typically above 1700 °C
Supercritical mixtures and estimated critical constants Phase labels near the critical region are a convention The numbers hold; the label "gas" or "liquid" may not mean much

Not for

What What the app does instead
Heat of mixing of unlike liquids (ethanol into water, acid into water) The mixed temperature ignores it and a warning says so. Ethanol into water at 25 °C: measured about 34 °C, app 19 °C. Concentrated sulfuric acid into water: really about 110 °C, app 35 °C.
Boiling-point rise, vapour pressure and heat of dilution of salt solutions A solution is always liquid, with fitted density, heat capacity and viscosity. No boiling point is shown.
How much water dissolves in oil, and oil in water The split into two liquids is found, but the trace amounts dissolved in each are wrong by 10 to 1000 times with Peng-Robinson and by up to about 40 times with the activity model. Do not use them.
Polar mixtures on Peng-Robinson (a fixed-volume calculation on water with an alcohol, for example) Carries a Not reliable warning. Azeotropes are misplaced by 5 to 10 °C.
Acetic acid and other acids that pair up in the vapour Heat of vaporisation is 60 % high and the boiling point 4 °C low on Phase Equilibrium. The Property tab's measured values are fine.
Azeotropes that boil higher than both components (acetone/chloroform) Not found by either model.
Solid solutions, melts and slags, ions in water in Chemical Equilibrium Each solid is a pure phase; ionic species are left out.
Reaction rates beyond the ideal reactors of the Reactor case Equilibrium and Reaction give where a reaction ends, not how fast.
Heat exchanger sizing Duty, outlet temperatures and approach only: no area, no heat-transfer coefficient.
Relief sizing and design pressure The sealed vessel gives the unrelieved pressure as an indication.

Composition tables

Results that involve more than one component or phase carry a table of what is where:

  • Phase Equilibrium: the whole stream, then one table per phase (gas, each liquid, solid), in and out, each with a Total. With several streams, the mixed feed and each incoming stream to open.
  • Property, for a mixture: the composition for 1 mol, and what each phase holds when it is part gas and part liquid.
  • Heater, Heat Exchanger, sealed vessel: the amount in each phase in and out, and every component for a mixture.
  • Reactor: feed and outlet for every component.

Shares can be read in mol % or mass %, and amounts in any amount or flow unit. The Total rows are the check: what goes in comes out.

Where the numbers come from

The comparison is a set of reference cases; each gives one row with the reference, its source, the app's value and the deviation. A check before each release fails if any case becomes worse than its recorded verdict. The methods and data sources are described in Methods and data sources.

References

The comparison values in the tables above come from the sources listed here. Each case in the comparison names its own source.

  • Steam and water: the IAPWS-IF97 steam tables (Wagner et al., Journal of Engineering for Gas Turbines and Power 122, 2000). The app itself answers with IAPWS-95 (Wagner and Pruss, Journal of Physical and Chemical Reference Data 31, 2002). Ice Ih: IAPWS-06 (Feistel and Wagner, Journal of Physical and Chemical Reference Data 35, 2006). The sublimation pressure of ice: the IAPWS equation (Wagner et al., 2011).
  • Pure fluids and light mixtures with a reference equation: the Helmholtz equations in CoolProp (Bell et al., Industrial & Engineering Chemistry Research 53, 2014), including Span and Wagner for carbon dioxide and GERG-2008 type mixing for light gas mixtures.
  • Boiling points, liquid densities, heats of vaporisation, sublimation points and heats: the CRC Handbook of Chemistry and Physics and the NIST Chemistry WebBook.
  • Salt solutions: the CRC Handbook, "Concentrative properties of aqueous solutions", for densities and the boiling point of brine. Perry's Chemical Engineers' Handbook for the heat capacity of brine and for the enthalpy-concentration chart of sulfuric acid in water.
  • Vapour-liquid equilibrium: Perry's Chemical Engineers' Handbook, table 13-1, for methanol and water. McCabe, Smith and Harriott for benzene and toluene. The azeotrope table in the CRC Handbook (Horsley). The IUPAC-NIST Solubility Data Series for water with a hydrocarbon in two liquid phases.
  • Reaction equilibrium constants and heats: NIST-JANAF Gibbs energies of formation (Chase, NIST-JANAF Thermochemical Tables, 4th ed., 1998) for steam reforming, water-gas shift and ammonia synthesis. The NIST lower heating value for methane combustion. The measured decomposition temperature of limestone, 895 °C.
  • Ammonia at equilibrium at 450 °C, 10, 100 and 300 atm: Larson and Dodge (1923, 1924), measured.
  • Fluidized bed: Kunii and Levenspiel, Fluidization Engineering, 2nd ed., Example 3.1, for the minimum fluidization velocity. The Wen and Yu and Haider and Levenspiel correlations and Stokes' law are checked by hand calculation.
  • Dry ice: Giauque and Egan, Journal of Chemical Physics 5, 45 (1937), for the heat capacity.

See also the combined list on Methods and sources.