Tuesday, September 29, 2009

Modeling Formaldehyde

Ever since my days on the support team, I’ve been fascinated by the formaldehyde-methanol-water system. Formaldehyde is one of the most important chemicals made; in 2005, worldwide production was approximately 21 million tons. It’s also highly reactive, and thus is usually handled in aqueous solution, sometimes with methanol as well to inhibit oxidation and polymerization reactions. The solution is usually referred to as formalin. This reactivity is what makes formaldehyde so interesting, valuable, and challenging to model.

Formaldehyde reacts with itself in water to form chains of poly-oxymethylene glycols, and with methanol to form chains of hemiformal. In the vapor, formaldehyde can react with water to form methylene glycol and with methanol to form hemiformal.

In a typical formalin solution, the bulk of formaldehyde is bound into methylene glycol (MG) and hemiformal (HF) molecules.

For a given solution of formaldehyde, methanol, and water, the following reactions apply:


CH2O + H2O =HO(CH2O)H (MG)



CH2O + H2O =HO(CH2O)H (MG)

HO(CH2O)nH + HO(CH2O)H = HO(CH2O)n+1H+ H2O


HO(CH2O) n CH3 + HO(CH2O)CH3 = HO(CH2O) n+1 CH3 + CH3OH

Distribution of hemiformal concentrations as a function of formaldehyde concentration in methanol

MGn and HFn have very different vapor pressures than does pure formaldehyde, water, or methanol. It is not possible to determine the pure properties of these polymers directly, as they exist only in solution. The reactions also take longer to reach equilibrium than typical vapor-liquid equilibrium, which is one reason why much of the literature data is not thermodynamically consistent.

To model this behavior in CHEMCAD, we solve the reactions and thermodynamics together in the thermo model. By combining the equilibrium reactions with an activity coefficient model such as UNIFAC to predict the polymer activity coefficients, we can accurately predict the overall thermodynamic behavior of the system. We named our model the “Maurer” method based on the work of Professor Gerd Maurer and his group at Kaiserslautern University. Future CHEMCAD updates will include adaptations to the density and viscosity mixing models.


  1. I use last version of Chemcad. When When I use Maurer model the entalpy calculation of the column outlet streams is not correct (calculated enthalpy is -1.#QNAN).
    Beside that when I use ESD model during formaldehyde calculation I am not able to add HTS file.
    Oleg Pajalic (oleg.pajalic@perstorp.com)

  2. Oleg, you should probably talk to your support rep on this. I'm not sure what your column composition is, but the enthalpy model seems fine when I run the maurer model.

    For formaldehyde, I don't recomend ESD and HTS, but you should be able to use a HTS file there

  3. As a test of Oleg's comment, I created a simple absorber.
    Inlets are both at 50F and 14.7
    Inlet liquid:
    100 lbmol/h water
    10 lbmol/h formaldehyde

    Inlet vapor:
    10 lbmol/h formaldehyde
    10 lbmol/hr nitrogen

    the outlet liquid is at 158F with an enthalpy of -13.69559 MMBTU/Hr its a 25% wt solution of Formaldehyde

  4. Hi again

    I just checked Heat of Evaporation for a mixture of 37% Formaldehyde and rest of water (by weight) by using Maurer method at boiling point at 1 bar. Calculated value in Chemcad was 1615 kJ/kg. Correct value (litterature and internal measurement is 2108 kJ/kg.
    Please explain why you have different value.

    Oleg Pajalic (oleg.pajalic@perstorp.com)