Which simulation software

[simple70]

Hi,

Which simulation software is considered the best and most accurate for crude refining ? Hysys, Chemcad, Pro II, Design II .. ?

Did anyone try to cross check the simulation with manual calculation ?

 

[MortenA]

I think that they will come out identical in any calculation simple enough to be checked by hand calculations.

The main difference today would be in features, user friendlyness, cost and number of users.

Best regards

Morten

 

[EmmanuelTop]

For refining applications, I consider HYSYS and its relatives (Ref-SIM, Petro-SIM) as the best modeling tool. HYSYS does not support refinery reactor systems, which is very well covered in Ref-SIM and Petro-SIM (built-in moduls of these unit operations).

Aspen Plus is one of the most user-unfriendly softwares (for my taste), while ChemCAD and Design II are more like academic purpose software. I do not have experience with Pro/II.

Also include ProMax and VMGSim in your evaluation list. Bot are rated high in process industry. VMGSim comes from the original Hyprotech team, developers of HYSYS.

Best regards,

 

[dcasto]

how can the accuracy be defined, they all use an EOS state that will give the same answer everywhere on the planet.

I useually do a hand calculation first before using a simulation. Not a tray by tray ect, but generalized tables/charts. I may use a simple energy or mass balance to get an idea of what is going on. This allows me to give the simulator a good set of initial guesses or where recycles are and how big.

I consider WinSim the best bang for the buck. It isn't as pretty as the others, but you have to love the cost and ease of use, yeah, once you get to know its tricks. The same is true with them all. So much for the commercial.

Morten is also with me on this, "it isn't the software that makes a simulation, its the user."

 

[MortenA]

dcasto

I also like winsim - unfortunately as a consultant i have to go with that package that my clients prefer - and here it seems like HYSYS has the larger userbase. One of our major cliens hwoever uses winsim - and therefore we (still) have licenses for winsim.

Best regards

Morten

 

[dcasto]

MortenA and OP,

I'm not a power user, but I keep an eye open for people with questions in that forum along with the other simulators.

 

[simple70]

Thank you all for your comments.

Since the basis of distillation is vaporization, the simulation software should be able to provide the phase change of crude correctly.

I don't know if it is possible for people who have access to different simulation software to try to find how much of a specific crude is vaporized at specific conditions, for example the heater outlet conditions or the flash zone conditions, for a crude they already know about. It is possible to do the hand calculations cross checking with known crude such as the crude provided in the example shown in Watkins book "Petroleum Refinery Distillation".

The manual calculation of the crude Equilibrium Flash Vaporization curve provides a certain % vaporization at heater outlet conditions, if the simulator is able to provide similar result, then it is a good starting point for its accuracy.

Would anyone try that ?

 

[Longinthetooth]

Emmanualtop is right. Apsen is very user unfriendly. I had to convert from Pro2 to Apsen and I rue the day it happened. Pro2 (or SimSci) with it's colour coded windows, containing all the directions you need, is by far the easier. I think they give equivalent results, although they both have different options with pluses and minuses.

Cost difference....well that depends what deal you can prize out of them.

 

[MortenA]

Geordie87

You hint that you have been able to get a discount on Aspen products. I usually dont stop untill i get some discout - in some cases up to 50% - but with Aspen i couldnt get anything.

Best regards

Morten

 

[Longinthetooth]

Morton,

It was not my decision to convert to Aspen. I think it was the company's need for access to a polymers package...namely Polymers Plus from Aspen.

My (admittedly cynical and simplistic) view is that Aspen was written by scientists for scientists and SimSci was written by engineers for engineers.

Best wishes

Geordie87

 

[dcasto]

Aspen will not discount and if your company is bought, then they won't assign the contract to the new owner. The original simulator built by engineers is ChemShare. Its the best in its class. IMHO.

 

[simple70]

For those who have access to simulation software and are willing to check my inquiry, following are the analysis of the crude in the example of Watkins book:
Bulk Gravity : 36.3 oAPI
Light end analysis per 100 barrels
bbl
C2H6 0.1
C3H8 0.9
iC4H10 1.4
nC4H10 2.3
iC5H12 2.1
nC5H12 1.8

TBP distillation of crude , Temperature in oF

Vol.% TBP
5 62
10 144
20 255
30 344
40 437
50 531
60 623
70 717
80 819
85 897

I asked a friend who has access to Chemcad in a university to check the % vaporized provided by the simulator, comparing with the manual calculation it was very far and can't be accepted.

Would anyone try other software and share the results ?

Thanks ...

 

[simple70]

Conditions

Sorry for not providing the conditions to check for % vaporized

Watkins made his illustration for the crude distillation at the following conditions

Heater outlet
T = 700 oF, P = 35 psia

Flash Zone conditions
T = 695 oF , P = 25 psia

The heater outlet conditions are the main checking case, since the % vaporized at flash zone is affected by presence of stripping steam.

 

[mbt22]

I would suggest that the simulator is not important, but rather getting the correct thermodynamic method. As Morten stated, accuracy is not really a factor in the choice of simulation package. For a more exotic system, say Geordie87's polymers, the availability of specific methods may be an issue, but this is very unlikely to be the case for crude oil.

I would check the assumptions in the manual method - is it possible that the simulator with an appropriate equation of state is going to be more accurate than a (presumably short cut) hand calculation? I would also ask your friend what property method they used for the calculation.

 

[dcasto]

Stream Number 1 2
Stream Name Heater in Heater out

Thermo Method Option Esso Esso
Vapor Fraction 0.0227614 0.9143195
Temperature F 215 700
Pressure psia 35 25
Enthalpy btu/hr -85367587.1 384706965
Entropy btu/R/hr 1740.185 524631
Vapor Density lb/ft3 0.3099071 0.3084775
Density lb/ft3 48.31416 48.48441
Specific Gravity60F@STP 0.8561347 0.9722873
Vapor Cp btu/lbmol/R 29.22838 98.73733
Vapor Cv btu/lbmol/R 26.97994 96.5664
Liquid 1 Cp btu/lbmol/R 96.96243 387.37642
Vapor Viscosity cP 0.0096158 0.0124019
Liquid 1 Viscosity cP 1.1053 1.60212
Vapor Thermal Conductivity
btu/ft/hr/F 0.0159779 0.028927
Liquid 1 Thermal Conductivity
btu/ft/hr/F 0.0676442 0.0391321
Vapor Flowrate mmscf/day@STP 1.4213 57.09321
Liquid Flowrate gal/min@STP 2884.44235 625.47441

Molecular Weight 181.5572 181.5572
Molar Flowrate lbmol/hr 6857.1946 6857.1946
Mass Flowrate lb/hr 1244973.051 1244973.051
Note: All Liquid 1 Phase calculations exclude Free Water

 

[simple70]

To mbt22
My knowledge so far concerning crude distillation is that using the TBP curve to construct the EFV curve of the whole crude at heater outlet conditions and flash zone is the first step for successful distillation calculations, because it will give the actual % vaporized. Creating EFV is a manual method, but it is the only known method (as far as I know). I presume that the simulators do not have other ways to create the EFV.

Since my only available option is Chemcad, I was informed it doesn't provide EFV curve, but the % vaporized calculated by it should be close to what Watkins calculated since he used the only known method for creating EFV, otherwise the built in method in Chemcad for crude phase change won't be accurate enough.

To dcasto
Thanks, it seems you took a practial step, which software did you use ?
I think you should try the heater outlet conditions at pressure = 30 psia, it will give you slightly lower % vaporized, as the 25 psia is at flash zone and in this case you can't compare your results to the manual results obtained by Watkins since he used stripping steam which will give more vaporization than dry conditions.
In all cases, I think you arrived at about the same results which were obtained by Chemcad, the % vapor you got is much higher (much much higher) than the actual EFV curve results.