Dealing with thermodynamics: the difference between petroleum industry and chemical industry

As hard as it may seem to accept, using a simulation software more adapted to the chemical industry (Aspen Plus) for a more chemicals like mixture doesn't mean that all the results out of the box, even for common mixtures, are more reliable than what comes from a simulator more typically used in the petroleum industry (HYSYS). By petroleum industry I mean both the oil&gas industry and the refining industry.

Just a couple of days ago I had an Email exchange where I tried to explain this and unfortunately the person at the other end didn't accept. "The problem" occurs when someone from the petroleum world needs to solve a problem that is more chemicals oriented and turns to a simulator like Aspen Plus believing that this action alone is going to guarantee accurate results.

In the petroleum world it is self evident that the simulator provides you with pretty accurate answers for common mixtures like any hydrocarbon mixture within a broad range of operating conditions. But chemicals are not labeled to form non-idela mixtures for nothing. Aspen Plus has loads of pure component data, it has loads of binary interaction coefficients and loads of methods. If you select the proper methods you can expect pretty accurate results for the vapour liquid equilibrium. Even for enthalpy calculations some more caution is already required. The problem here is that you want your calculations to provide consistent results which means that the enthalpy of the vapour and the enthalpy of the liquid need to be different by exactly the heat of vapourisation. This may be not too hard for pure components, but once you start mixing components things are not that easy anymore. When you calculate the two enthalpies and the heat of vapourisation from the pure component data you are almost inevitably going to find that the result is not consistent and so you need to make a choice. You can only calculate two out of three from the pure component data and the third one is a consequence. That of course mean that the third one may be quite different from what you expect.
For physical properties (mass density) and transport properties (viscosity, thermal conductivity, surface tension), the story is similar. The example I was faced with was the viscosity of a water - methanol mixture. The Aspen Plus result is not good. BUT, if you go looking at the predicted viscosities for pure water and pure methanol they are spot on. The problem is with the mixing, this mixture doesn't behave like most and to get a better result you have to change the mixing rule. This is where Aspen Plus is much stronger. In HYSYS you would not find a large selection of mixing rules.

So face it: calculating properties for non-ideal mixtures, even common ones is not solved by simply switching simulators, some more work is needed to verify results and select the proper methods and mixing rules to get reliable results.