Simulis® Thermodynamics

Simulis® Thermodynamics
Mixture properties and fluid phase equilibria calculation server ProSim / NTP Truboprovod Seminar (Moscow) – October, 28th, 2011 Stéphane Déchelotte (ProSim)

Outline The importance of thermodynamics What is included in Simulis® Thermodynamics How can you use Simulis® Thermodynamics How can you enrich Simulis® Thermodynamics Concluding remarks

(Stead-state, without chemical reaction)
The importance of thermodynamics Example of a flash unit (Stead-state, without chemical reaction) Feed Vapor Liquid Q Flowrate: F Composition: z Enthalpy: HF Flowrate: V Composition: y Enthalpy: H Flowrate: L Composition: x Enthalpy: h P T Data: F, z, HF Parameters: P, T, Q Variables: V, L, y, x, H, h Equations Global mass balance: L + V - F = 0 Partial mass balances: L . xi + V . yi - F . zi = 0 Enthalpy balance: L . h(T,P,x) + V . H(T,P,y) - F . HF(T,P,z) - Q = 0 Thermodynamic equilibrium relations: fiv(T,P,y) = fiL(T,P,x) or yi = Ki (T,P,x,y) . xi Constraints: S xi = S yi = S zi = 1 Models are used for equilibrium constants and enthalpies Quality of results rely on quality of models

The importance of thermodynamics
% ERROR IN NUMBER OF STAGES 30 25 20 15 10 5 % ERROR IN RELATIVE VOLATILITY ( a ) a = 10 a = 4 a = 2 a = 1.2 a = 1.05 On va considérer un cas concret, à savoir la distillation. Pour ce faire, on va introduire la notion de volatilité relative, alpha, qui est le rapport des constantes d’équilibre entre deux constituants. Elle caractérise donc la difficulté à séparer deux constituants : plus alpha est grand, plus les constituants sont faciles à séparer, plus il est proche de l’unité, plus la séparation de ces constituants devient difficile. On peut quantifier, en appliquant les équations relatives à la distillation, l’impact d’une erreur d’estimation de alpha sur l’erreur faite sur l’estimation du nombre d’étages théoriques lors du dimensionnement d’une colonne. C’est ce qui est présenté sur ce graphique : en abscisse est présentée l’erreur sur la volatilité relative, en ordonnée est représentée l’erreur sur l’estimation du nombre d’étages théoriques. On constate alors que si deux constituants sont faciles à séparer, à savoir si alpha est grand, une grosse erreur au niveau de la thermo entraine une erreur relativement faible sur le dimensionnement de la colonne. En revanche, si l’on a des constituants très proches, avec un alpha proche de 1, on voit qu’une faible erreur sur alpha implique une grosse erreur sur le dimensionnement de la colonne. Cet exemple montre bien entendu l’impact de la thermodynamique sur la simulation de procédés, mais il met aussi l’accent sur le fait que l’élaboration d’un modèle thermodynamique doit être faite en considérant le domaine d’utilisation et la finalité de ce modèle.

The importance of thermodynamics
Fluid phase equilibria: fugacity, activity coefficients,… Distillation, extraction, absorption, crystallization,… Thermodynamic properties: enthalpy, entropy, specific heat,… Heat balance, compressors calculation,… Transport properties: viscosity, thermal conductivity,… Heat exchangers, pressure drop, distillation towers sizing,… La simulation de procédés nécessite la connaissance de plusieurs grandeurs thermodynamiques : Tout d’abord, la connaissance des équilibres entre phases est nécessaire dans les opérations de séparation, telles que la distillation (diphasique ou éventuellement triphasique), l’extraction liquide-liquide, l’absorption de gaz ou encore la cristallisation qui fait intervenir des équilibres liquide-solide. Pour ce faire, les calculs de coefficients de fugacité ou de coefficients d’activité, en fonction de l’approche choisie, seront nécessaires. Outre les bilans matière, les bilans énergie sur un procédé sont requis. On demandera donc à la thermodynamique de calculer des propriétés thermodynamiques telles que l’enthalpie, la chaleur spécifique ou encore l’entropie, nécessaire par exemple pour le calcul de compresseur. Les dimensionnements d’équipements (on peut citer les échangeurs de chaleur, les internes de colonne ou encore des calculs de pertes de charge dans des canalisations) vont nécessiter la connaissance de propriétés de transfert, telles que les viscosités, les conductivités thermiques ou encore la tension superficielle Enfin, d’autres propriétés d’utilisation plus anecdotiques pourront être requises, telles que la vitesse du son dans un mélange

Quality of results depends on model used
En revanche, si l’on utilise une équation d’état telle que SRK pour calculer l’équilibre thermodynamique d’un système contenant des constituants polaires, tels que l’eau et l’acétone, on voit que le modèle est particulièrement mal adapté. Dans ce cas là, il convient d’utiliser un modèle d’enthalpie libre d’excès, tel que NRTL dans le cas présent. Les paramètres d’interaction binaire du modèle sont estimés en régressant des valeurs expérimentales : sans paramètre d’interaction binaire, cela revient à avoir une modélisation avec un modèle idéal, très éloigné lui aussi du comportement réel du mélange.

in MS-Excel®, MATLAB® or other applications
Simulis® Thermodynamics Software component for computing thermophysical properties and phase equilibria on pure substances or mixtures in MS-Excel®, MATLAB® or other applications Simulis Thermodynamics

Phase envelope and hydrate line of a natural gas mixture with PR EOS
Example of calculation performed with Simulis® Thermodynamics Phase envelope and hydrate line of a natural gas mixture with PR EOS

in Simulis® Thermodynamics?
What is included in Simulis® Thermodynamics? Databases (pure, BIP) Thermodynamic Functions To compute thermo-physical properties

Thermo-physical properties supported
Transport properties Isobaric specific heat (Cp) Dynamic viscosity Cinematic viscosity Thermal conductivity Density Molar Volume Molar density Surface tension Molecular weight Thermodynamic properties Enthalpy (H) Entropy (S) Internal energy (U) Isochoric specific heat (Cv) Enthalpy of vaporization Non-ideal properties Activity coefficients Fugacity coefficients and Fugacity Ln of fugacity coefficients Henry's constant Compressibility properties Compressibility factor Gamma (Cp/Cv ratio) Sound speed Joule-Thomson coefficient Derivatives of the properties with respect to temperature, pressure and number of moles are also given

What is included in Simulis® Thermodynamics? Databases (pure, BIP) Thermodynamic Functions Flashes (LV, LLV, LL,...) To compute phase equilibria

Flashes supported T P w V H S U Liquid-Vapor equilibria
Bubble and dew temperatures and pressures Flash at given temperature (T) and pressure (P) Flash at given vaporization ratio (w) and P (or T) Flash at given enthalpy (H) and P (or T, or V, or U) Flash at given entropy (S) and P (or T, or V, or H, or U) Flash at given internal energy (U) and P (or T, or V) Flash at given volume (V) and P (or T) Phase envelope RVP (Reid Vapor Pressure) TVP (True Vapor Pressure) Liquid-Liquid-Vapor equilibria Bubble temperature Flash at given enthalpy and pressure Flash at given temperature and pressure Flash at given vaporization ratio and pressure Liquid-Liquid equilibria Flash at given temperature and pressure

What is included in Simulis® Thermodynamics? Databases (pure, BIP) Thermodynamic Functions Flashes (LV, LLV, LL,...)

Pure component properties and binary interaction parameters
Supplied with a database of over components based on AIChE's DIPPR® database (public release 2006) 34 constant properties (molar weight, critical temperature,…) 15 temperature dependant properties (Cp, Pi°, , Hvap …) Optionally the last public version of the DIPPR® database can be supplied All the properties of pure components can be accessed, modified, plotted,… Binary Interaction Parameters: When required (NRTL, UNIQUAC,…) BIP supplied for a number of systems

What is included in Simulis® Thermodynamics? Databases (pure, BIP) ThermodynamicModelss Thermodynamic Functions Flashes (LV, LLV, LL,...)

A wide set of thermodynamic models
Equations of State Soave-Redlich-Kwong (SRK) Peng-Robinson (PR) Predictive Peng Robinson 78 (PPR78) Lee-Kesler-Plöcker (LKP) Benedict-Webb-Rubin modifié Starling (BWRS) Nakamura PPC-SAFT (IFP) NRTL-PR etc… Activity coefficients models NRTL UNIQUAC UNIFACs (Larsen, Dortmund,…) Wilson Combined approach models MHV2 MHV1 PSRK Specific systems Pure Water (NBS/NRC steam tables - IAPS,1984) Chao-Seader, Grayson-Streed Sour-Water Carboxylic acids Formaldehyde Electrolytes Edwards UNIQUAC electrolyte ULPDHS Amines New methods are continuously implemented to cover a wide field of applications: oil & gas, fine chemicals, etc…

EoS/GE models: Ethanol - Water
SRK, MHV2 and UNIFAC Larsen are used: fully predictive model Prediction is pretty good even for isotherms over Tc of ethanol (~240 °C)

EoS/GE models: Ethane - Toluene
PSRK is used: fully predictive model Prediction is correct even for isotherms over Tc of toluene (~319 °C)

EoS/GE models: CO2 - Ethane
Predictions with both PPR78 (Jaubert et al.) and PSRK (Gmehling et al.) are satisfactory

Graphical User Interface
What is included in Simulis® Thermodynamics? Graphical User Interface Databases (pure, BIP) Thermodynamic Functions ThermodynamicModelss Flashes (LV, LLV, LL,...)

A graphical user interface
To select components

A graphical user interface
or to configure the property model The various available methods can be combined to model a specific system

What is included in Simulis® Thermodynamics? Graphical User Interface Set of Services Databases (pure, BIP) Thermodynamic Functions ThermodynamicModelss Flashes (LV, LLV, LL,...)

A full set of services available
An interactive calculation service With Simulis Thermodynamics, an interactive calculation service is also provided and user can compute any property or flash. The results are available as arrays of numbers, but user can trace the results as he wants.

An interactive calculation service Graphical display of properties on temperature, pressure or composition ranges This calculation service displays also graphical of properties on temperature, pressure or composition ranges. One can see for example phase envelope and temperature - entropy diagram

An interactive calculation service Graphical display of properties on temperature, pressure or composition ranges Calculation of petroleum fractions properties Simulis Thermodynamics offers also a management tool for petroleum fractions to the user.

An interactive calculation service Graphical display of properties on temperature, pressure or composition ranges Calculation of petroleum fractions properties Management of group contribution methods versions Several versions are supported: UNIFAC original UNIFAC (Dortmund) modified UNIFAC (Dortmund) LL UNIFAC (Lyngby) modified Larsen UNIFAC formaldehyde PPR78 A group contribution models editor is supplied Several other services are implemented inside Simulis Thermodynamics, whose goal is to provide user with quite all tools required for thermodynamics analysis. One can note for example that some estimation methods of pure component properties are also available in Simulis Thermodynamics, Unit conversions management tool is also available insight Simulis Thermodynamics.

To rapidly integrate a new matrix To modify parameters (ri, qi) of selected groups To add new groups or sub-groups Several other services are implemented inside Simulis Thermodynamics, whose goal is to provide user with quite all tools required for thermodynamics analysis. One can note for example that some estimation methods of pure component properties are also available in Simulis Thermodynamics, Unit conversions management tool is also available insight Simulis Thermodynamics.

An interactive calculation service Graphical display of properties on temperature, pressure or composition ranges Calculation of petroleum fractions properties Management of UNIFAC versions Estimation of pure component properties Data regression of pure components experimental properties Unit conversions management tool etc… Several other services are implemented inside Simulis Thermodynamics, whose goal is to provide user with quite all tools required for thermodynamics analysis. One can note for example that some estimation methods of pure component properties are also available in Simulis Thermodynamics, Unit conversions management tool is also available insight Simulis Thermodynamics. All these services become automatically available in any software embedding Simulis® Thermodynamics

What is included in Simulis® Thermodynamics? Graphical User Interface Databases (pure, BIP) Set of Services Simulis Thermodynamics Thermodynamic Functions ThermodynamicModelss Flashes (LV, LLV, LL,...) Simulis® Thermodynamics have a very "rich" content

Simulis® Thermodynamics can be used in ProSim software suite
Since Simulis® Thermodynamics is a software component it must be embedded in another application ProSim Software Suite Simulis Thermodynamics Simulis® Thermodynamics is the thermodynamic "heart" of all ProSim software suite

Simulis® Thermodynamics can be used within MS-Excel®
Simulis® Thermodynamics is provided as an add-in to MS-Excel MS-Excel Add-In ProSim Software Suite Simulis Thermodynamics Rigorous thermodynamics become available in MS-Excel®

Example of use in MS-Excel®
Step 1: embed a “calculator” in a spreadsheet 2. Click on "Insert" menu 1. Select cells range

All the properties of the pure components can be accessed, modified, plotted,… Step 2: select compounds Supplied with a database of over 2000 components including AIChE's DIPPR® database and access to your "private" databases Access to several databases Mixtures with up to 200 compounds can be investigated Several research criteria to easily find a component

Refine the predefined model Access to several predefined models
Example of use in MS-Excel® Step 3: select the thermodynamic model Refine the predefined model Access to several predefined models

Select the desired thermodynamic function
Example of use in MS-Excel® Step 4: insert function in the spreadsheet Select the desired thermodynamic function

Step 5: provide the parameters of the function… ...as for any Excel function... … and get the results in your spreadsheet.

Simulis® Thermodynamics can be used within MS-Excel®
Thermodynamic functions are added to Microsoft® Excel …and used in spreadsheets as native functions… MS-Excel Add-In ProSim Software Suite Simulis Thermodynamics …to perform more or less complex engineering calculations (with rigorous thermodynamics)

Simulis® Thermodynamics can be used within MATLAB®
Toolbox Simulis® Thermodynamics is provided as a toolbox in MATLAB® MS-Excel Add-In ProSim Software Suite Simulis Thermodynamics Rigorous thermodynamics become available in MATLAB® without further programming effort

Simulis® Thermodynamics can export result files to other packages
MS-Excel Add-In MATLAB Toolbox ProSim Software Suite Simulis Thermodynamics Tabulated results to : MS-Excel Export files Aspen TASC (PSF file) OLGA (PVT file)

Simulis® Thermodynamics can be used within your software
A complete Application Programming Interface (API) is also provided Your software (C++, VB, FORTRAN,…) A.P.I. Simulis® Thermodynamics can be easily embedded in any application supporting the COM/DCOM technology Visual Basic C++ Delphi FORTRAN C# etc… MS-Excel Add-In MATLAB Toolbox ProSim Software Suite Simulis Thermodynamics Export files However, the interface between the embedding application and Simulis® Thermodynamics must be coded

Example of interface developed
With DPP: tool for regression of model parameters (from Dechema) MS-Excel Add-In MATLAB Toolbox ProSim Software Suite Simulis Thermodynamics Export files A.P.I. Simulis® Thermodynamics can be used within DPP

Simulis® Thermodynamics can be used within CO compliant packages
MS-Excel Add-In MATLAB Toolbox CAPE-OPEN "plug" Modeling tool implementing CO Thermo Socket Aspen Plus, ProII, Aspen HYSYS, HTRI, gPROMS … ProSim Software Suite Simulis Thermodynamics Export files A.P.I.

CO thermodynamic "plug" Simulis® Simulis®
Ability to generate CAPE-OPEN "Property Packages" to be used within compliant modeling tools Successfully tested in Aspen Plus Aspen Hysys PRO/II gPROMS® Xist (HTRI) COCO COFE Belsim VALI INDISS COMSOL etc… Thermo 1.0 and Thermo 1.1 are implemented Simulis® Thermodynamics Client Software Microsoft® Excel ProSimPlus, MATLAB, etc CO Property Package generated by Simulis® Thermodynamics Modeling tool implementing CO Thermo Socket Aspen Plus, ProII, Aspen HYSYS, HTRI, gPROMS … The thermo of ProSim can be used in software without satisfactory thermo routines A thermo expert can prepare with Simulis® Thermodynamics the model for a complex system, and provide it to his/her colleagues for further "safe" use in their traditional simulation tool (Aspen Plus, ProSimPlus,…)

Simulis® Thermodynamics can be used in a number of ways
MS-Excel Add-In MATLAB Toolbox Modeling tool implementing CO Thermo Socket Aspen Plus, ProII, Aspen HYSYS, HTRI, gPROMS … ProSim Software Suite Simulis Thermodynamics CAPE-OPEN "plug" Export files A.P.I. Tabulated results to : MS-Excel Your software (C++, VB, FORTRAN,…) Aspen TASC (PSF file) OLGA (PVT file)

You can enrich Simulis® Thermodynamics with your components
New components can be added and properties modified estimation methods included regression of experimental data tool provided Existing in-house property databases can also be included (interface format supplied) Simulis Thermodynamics Databases import r m l s DH Cp kij In-house DB

Example of regression of temperature dependant pure component properties

You can add a specific library to Simulis® Thermodynamics
Your thermo (C++, VB, FORTRAN,…) Here the interface must be specifically developed by ProSim team A good option if you have a big code and you don't want to take care of its integration in Simulis® Thermodynamics Specific libraries Simulis Thermodynamics Databases import

Standard Reference Database 23
Example of specific library added: REFPROP NIST reference database for refrigerants 84 pure fluids and 5 pseudo-pure fluids (air…) Helmoltz energy equation of state, MBWR, Bender… REFPROP Standard Reference Database 23 Specific libraries Simulis Thermodynamics Databases import

Example of specific library added: REFPROP
A specific DLL to match the Simulis Thermodynamics entry point predefined syntax has been developed No more development required when a new version of REFPROP is released All Simulis Thermodynamics services become available with REFPROP models: calculation service, PSF file export, diagram service, etc… The NIST Reference Fluid Thermodynamic and Transport Properties Database (REFPROP) last version (8.0) includes 84 pure fluids, 5 pseudo-pure fluids (such as air) and allows calculation for mixture up to 20 components. This package uses the most accurate equations of state and models currently available (Helmoltz energy equation of state, MBWR, Bender…). NIST provides REFPROP customers with a DLL for which entry points are fully documented. In order to use REFPROP with the expert mode of Simulis® Thermodynamics, a specific DLL has been developed to wrap the REFPROP DLL in order to match the predefined syntax of each entry point available in Simulis® Thermodynamics. Then, with new refprop release, no more development will be required on the Simulis Thermodynamics side. These developments were done with the financial support of Air Liquide

You can develop your own thermo package within Simulis® Thermodynamics
Specific libraries Simulis Thermodynamics "Expert mode" VBScript or DLL Databases import Your thermo (C++, VB, FORTRAN,…)

"Expert mode" Provide thermodynamic experts with a simple and standardized development framework: to develop their own new thermodynamic models or to integrate in Simulis® Thermodynamics existing thermodynamic models taking advantage of Simulis® Thermodynamics environment (pure components properties, unit conversions and management,…) in view to use these developments in other applications (commercial software, Microsoft® Excel, MATLAB®, legacy codes …) Two possibilities are offered: VBScript models External DLL models The main objective of the expert mode is to provide thermodynamic experts with a simple and standardized development framework in order to integrate their know-how in Simulis® Thermodynamics. This know-how can be a new one and then user will benefit from the standardized integration structure (entry point definition, user parameters management…)., or it can be an existing one and then user will benefit from a validated development framework for integration and exploitation of existing codes. Furthermore, this framework supplies developers with tests and debugging facilities. A strong point of this feature is that pure component properties databases are fully available from the expert mode and release user from pure properties management. Like this they can make their models available in all client applications: Microsoft® Excel, Matlab®, all Process Modeling Environment CAPE-OPEN compliant (Aspen Plus, PRO/II, gPROMS, Aspen Hysys, HTRI Suite, Unisim Design, COCO COFE, Belsim VALI, INDISS...) or in any application supporting the COM/DCOM technology (C#, VB.NET, Visual Basic, C++, Delphi, FORTRAN,…). This feature is available from two modes: VBScript or external DLL.

VBScript models: a very easy way
Code is directly entered in Simulis® Thermodynamics A skeleton is provided Many available functions This mode is an integrated solution as VBScript language is an interpreted one and then doesn't require any external compiler. The code is directly entered in Simulis Thermodynamics. Many functions are available (more than 100) and for each one a skeleton is provided. Function parameters must satisfy a predefined syntax, which is fully described for all available functions. Function parameters must satisfy a predefined syntax (name, type, units)

VBScript models: a very easy way
VBScript is a well known (or easy to learn) language Access to pure components properties, units management,… However, VBScript is an interpreted language  not very efficient Useful for prototyping or simple functions (Cp,…) Can be a first step If we do a quick comparison between these two expert modes, the VBScript mode use an interpreted language close to Visual Basic which is a language well known from most engineers but which is not efficient in term of calculation time. It is very useful for the prototyping or for implementation of simple functions, like heat capacity for example and then it can be a first step implementation. The DLL expert mode allows DLL written using any language and is very useful when reusing existing codes, which just requires a wrapping to match the predefined syntax of each calculation routine, being in fact an entry point in the DLL. A big advantage with the other mode is the calculation performance. This last mode will be highlighted by showing what has been done with REFFPROP. In both cases, developers have access to pure components properties, units management and all Simulis thermodynamics facilities. For more complex developments an other possibility is offered: DLL models

"Expert mode": DLL models
An external DLL is plugged (C++, Fortran,…) User parameters are supported Test facility Many available functions If the engineers in electronic material were to start from a sand heap each time that they design a new device, if their first stage always had to consist in extracting the silicon to make integrated circuits, they would not progress very fast. Function parameters must satisfy a predefined syntax (name, type)

"Expert mode": DLL models
Can be built using any language (FORTRAN, C++,…) Allows re-use of existing codes (wrapping required to match the predefined syntax) More efficient Can be used to make thermo legacy codes CAPE-OPEN compliant Easy configuration (access to standard pure components databases,…) Predefined access to pure components properties User-friendly existing GUIs to access parameters Full rewriting is not required (possibility to mix with native existing models) VBScript, external DLL and native models can be mixed (each one computing a different property) If we do a quick comparison between these two expert modes, the VBScript mode use an interpreted language close to Visual Basic which is a language well known from most engineers but which is not efficient in term of calculation time. It is very useful for the prototyping or for implementation of simple functions, like heat capacity for example and then it can be a first step implementation. The DLL expert mode allows DLL written using any language and is very useful when reusing existing codes, which just requires a wrapping to match the predefined syntax of each calculation routine, being in fact an entry point in the DLL. A big advantage with the other mode is the calculation performance. This last mode will be highlighted by showing what has been done with REFFPROP. In both cases, developers have access to pure components properties, units management and all Simulis thermodynamics facilities.

Priorities must be given
"Expert mode" VBScript, external DLL and native models can be mixed (each one computing a different property) Native models will be used except when a VBScript model or a DLL model is available If the engineers in electronic material were to start from a sand heap each time that they design a new device, if their first stage always had to consist in extracting the silicon to make integrated circuits, they would not progress very fast. Priorities must be given

You can use other commercial thermo package within Simulis® Thermodynamics
External software able to generate CO packages (Aspen Properties, Multiflash, PPDS,…) CAPE-OPEN "socket" Specific libraries Simulis Thermodynamics "Expert mode" VBScript or DLL Databases import

CO thermodynamic "socket"
Ability to use an external thermodynamic model (CAPE-OPEN "Property Package") Successfully tested with Multiflash (Infochem) PPDS (TUV-NEL) Aspen Properties (AspenTech) COM Thermo (AspenTech) IVCSEPThermoSystem (IVC-SEP) COCO TEA (AmsterCHEM), etc… Thermo 1.0 and Thermo 1.1 are implemented External CAPE-OPEN Property Package External software able to generate CO packages (Aspen Properties, Multiflash, PPDS,…) Simulis® Thermodynamics Client Software MS-Excel ProSimPlus, MATLAB, etc If required, a third party commercial thermo package can be used within any application embedding Simulis® Thermodynamics

Simulis® Thermodynamics can welcome your know-how
External software able to generate CO packages (Aspen Properties, Multiflash, PPDS,…) Your thermo (C++, VB, FORTRAN,…) REFPROP Standard Reference Database 23 CAPE-OPEN "socket" Specific libraries Simulis Thermodynamics "Expert mode" VBScript or DLL Databases import r m l s DH Cp kij In-house DB Your thermo (C++, VB, FORTRAN,…)

Example of interesting combination
Your thermo legacy code become CAPE-OPEN "compliant": "Usable" in other applications No knowledge of CO technology required Reduced development time CAPE-OPEN "plug" Modeling tool implementing CO Thermo Socket Aspen Plus, ProII, Aspen HYSYS, HTRI, gPROMS … Simulis Thermodynamics "Expert mode" VBScript or DLL Your thermo (C++, VB, FORTRAN,…) You don't need anymore to rely on your sw provider to maintain the specific interface with your code

CAPE-OPEN "socket" External software able to generate CO packages (Aspen Properties, Multiflash, PPDS,…) MATLAB Toolbox Simulis Thermodynamics You can use your favorite thermo package in MATLAB® without any programming effort

CAPE-OPEN "socket" External software able to generate CO packages (Aspen Properties, Multiflash, PPDS,…) You can use your favorite thermo package in your own software (heat exchanger design,…) Simulis Thermodynamics A.P.I. Your software (C++, VB, FORTRAN,…)

Standard Reference Database 23
Example of interesting combination MS-Excel Add-In Specific libraries REFPROP Standard Reference Database 23 Simulis Thermodynamics REFPROP components and models can be used in a spreadsheet You take advantage of both the accuracy of REFPROP for refrigerants and the user-friendliness of Simulis® Thermodynamics

Concluding remarks Thermodynamics is fundamental to get good simulation results Simulis® Thermodynamics is a comprehensive tool A full set of property models including advanced ones Extended databases and robust flash algorithms A full set of services Its seamless integration in many codes is allowed All the services available in Simulis® Thermodynamics become available in the embedding application Any application embedding Simulis® Thermodynamics automatically inherits from its CAPE-OPEN standard compliance

Concluding remarks Simulis® Thermodynamics is really an open structure
Simulis® Thermodynamics allows several levels of use Within MS-Excel or MATLAB® to perform more or less complex engineering calculations Embedded in other applications taking advantage of the interfaces proposed (CAPE-OPEN, etc…) As a framework to welcome existing thermodynamic models or to develop new ones in view to use them in other applications

Concluding remarks These levels of use can coexist in the same organization allowing: Minimization of learning time Consistency of data & results between several applications Easy deployment of "thermo expert" analysis Repository of internal knowledge It's not because you already have a thermo package that you don't need Simulis® Thermodynamics Due to our "low-cost" business model, the license price must not be a limitation

Thank-you for your attention!
Simulis® Thermodynamics Thank-you for your attention! Pour conclure sur la situation de ProSim aujourd'hui, nous avons vu que :

Software Solutions for Process Modeling, Simulation and Optimization
Contact Us ProSim Software Solutions for Process Modeling, Simulation and Optimization Web: ProSim SA Stratège Bâtiment A BP 27210 F Labège Cedex France Phone: +33 (0) Fax: +33 (0) ProSim, Inc. Science Center 3711 Market street, 8th floor Philadelphia, PA 19104 U.S.A. Phone: Fax: Enfin la troisième tendance est pour nous de conserver une activité forte en prestations. Lire Il s'agit en fait de maintenir notre souplesse et notre réactivité qui sont je crois un de nos points forts.

The two generally used approaches
Two phases system at thermodynamic equilibrium (Liquid-Vapor for example) For wide T and P ranges but non polar compounds For polar compounds but at low pressure only Equation of state (Tc, Pc, w, MR, BIP) GE model (BIP) Pi°, EoS, vi° Si l’on considère un système diphasique en équilibre thermodynamique liquide vapeur par exemple, cet équilibre est caractérisé par l’égalité des fugacités de chacun des constituants dans chaque phase. Si l‘on considère un système contenant des corps peu ou non polaires, l’utilisation d’une équation d’état est particulièrement recommandée. Avec cette approche, les fugacités en phase liquide et vapeur seront calculées par la même expression qui fait intervenir un coefficient de fugacité, calculé par la résolution de l’équation d’état. Compte tenu de cette remarque, l’approche est appelée homogène. Elle permet d’effectuer des calculs quels que soient les niveaux de température et de pression. En revanche, si l’on a un système contenant des corps polaires, l’utilisation des équations d’état avec leurs règles de mélange standard est à proscrire. On va alors utiliser pour le calcul de la fugacité en phase liquide un modèle de coefficient d’activité qui va venir représenter l’écart à l’idéalité dans la phase liquide. Le calcul de la fugacité en phase liquide étant toujours fait avec une équation d’état, on parle alors de modèle hétérogène. Bien entendu, compte tenu du fait que l’on calcule les fugacités en phases liquide et vapeur par deux méthodes différentes, cette approche ne doit être utilisée qu’à faible pression.

EoS/GE models Principle : an equation of state is used to cover the whole fluid domain. This EoS includes sophisticated mixing rules based on activity coefficient models: Equation of State (CEoS) SRK PR Activity coefficient model Wilson NRTL UNIQUAC UNIFACs Mixing rules MHV1 MHV2 PSRK Combine advantages of EoS approach (the whole fluid domain is covered) with advantages of GE models (highly non-ideal liquid phase can be tackled)

EoS/GE models: Acetone - Water
SRK, MHV2 and NRTL are used BIP identified with experimental values at 1 atm can be used to investigate high pressure domains

Other example: amines sweetening
Species: H2S, HS-, S2-, CO2, HCO3-, CO32-, RR’R”NCOO-, RR’R”N, RR’R”NH+, H2O, H+, OH- Chemical reactions : Dissociation of alkanolamine: RR’R”NH+  RR’R”N + H+ Dissociation of water: H2O  H+ + OH- Dissociation of H2S: H2S  H+ + HS- HS-  H+ + S2- Hydrolysis of CO2: CO2 + H2O  H+ + HCO3- HCO3-  H+ +CO32- Formation of a carbamate (except when ternary amine): CO2 + 2 RR’R”N  RR’R”NH+ + RR’R”NCOO- Outres les hydrocarbures présents dans le gaz à traiter, les espèces globales sont représentées par le H2S, le CO2, l’alkanolamine et bien entendu l’eau qui représente le solvant dans ce système. La liste des espèces vraies correspond à la liste des espèces globales, augmentée des ions qui vont être automatiquement rajoutés par le modèle en fonction des réactions précodées. On peut citer (donner la liste des réactions présentées sur le transparent)

Deshmuk and Mather model, parameters fitted by Wieland et al. Thermodynamic equilibrium relation (convention asymmetry) : Solvent (water) : Solutes : Equation of State: Peng Robinson Le modèle implémenté dans Simulis thermodynamics pour ces systèmes est le modèle de Deshmuk et Mather, les paramètres du modèle étant ceux identifiés par Wieland et ses collaborateurs. L’approche thermodynamique retenue est une approche de type hétérogène, avec une convention assymétrique : le coefficient d’activité de l’eau tend vers 1 quand la fraction molaire de l’eau tend vers 1 et les coefficients d’activité des solutés tendent vers 1 quand leur fraction tend vers 0 (c’est à dire à dilution infinie). L’équation d’état retenue pour le calcul de la fugacité en phase gaz est l’équation d’état cubique de Peng et Robinson. Le modèle de coefficients d’activité est le modèle d’Edwards, qui est une modification du modèle de Pitzer. Activity coefficient calculation model: Edwards

Sur ce transparent, nous avons présenté les courbes de parité pour le CO2 et l’H2S dans une solution de MDEA à 35% massique, pour des températures comprises entre 40 et 100°C. L’abscisse correspond à la pression partielle calculée, l’ordonnée correspond quant à elle aux pressions partielles expérimentales. Ainsi, quand le point se situe sur la bissectrice des axes, le modèle représente parfaitement les données expérimentales, ce qui est le cas.

Conventional simulator
Integration in ProSimPlus® Thermodynamics Chemical Reactions Unit Operations Numerical Methods GUI Conventional simulator Monolithic program (generally divided into several source files and DLLs) ProSimPlus Thermodynamic calculations are performed within a component (Simulis® Thermodynamics) Unit Operations Chemical Reactions GUI Numerical Methods Thermodynamics Simulis® This component implements CAPE-OPEN interfaces (Plug & Socket) ProSimPlus also implements CO Unit Socket

Example of formaldehyde process
Formaldehyde is highly reactive and commonly handled in solutions with water and/or methanol Mixtures of formaldehyde with water and/or methanol are complex to model since formaldehyde reacts with both of them [1] Michael Albert, Baudilio Coto Garcia, Cornelius Kreiter and Gerd Maurer Vapor-Liquid and Chemical Equilibria of Formaldehyde-Water Mixtures AIChE Journal, September 1999, Vol. 45, N° 9, pp [2] Michael Albert, Baudilio Coto Garcia, Christian Kuhnert, Roger Peschla and Gerd Maurer Vapor-Liquid Equilibrium of Aqueous Solutions of Formaldehyde and Methanol AIChE Journal, August 2000, Vol. 46, N° 8, pp [3] Y.Q. Liu, H. Hasse and Gerd Maurer Enthalpy Change on Vaporization of Aqueous and Methanolic Formaldehyde Solutions AIChE Journal, November 1992, Vol. 38, N° 11, pp Model proposed by Pr. Gerd Maurer (Kaiserslautern Unversity) has been implemented in Simulis® Thermodynamics

Chemical reactions involved
Formation of methylene glycol (MG) Formation of polyoxymethylene glycols (MGn) n  2 Formation of hemiformal (HF) Formation of polyoxymethylene hemiformals (HFn) n  2

Silver catalyst process simulated with ProSimPlus®
Polyoxymethylene glycols are taken into account up to MG5 - HO(CH20)5H Polyoxymethylene hemiformals are taken into account up to HF5 - CH3(OCH2)5OH

System modeling Vapor phase: ideal but chemical equilibria taken into account (Formation of methylene glycol and formation of hemiformal) Liquid phase: real chemical reactive mixture (water, methanol, formaldehyde, methylene glycol,polyoxymethylene glycols, hemiformal and polyoxymethylene hemiformals) non-ideality taken into account applying a specific UNIFAC Vapor-liquid equilibrium described by extended Raoult's law Specific enthalpy calculation method (taking into account the oligomer distribution in the liquid phase)

Example of results obtained with Simulis® Thermodynamics
Partition coefficient of FA in the FA-water-methanol system Concentration of HF, HF2 and HF3 in chemical equilibria at 276 K

Example of interface developed
With DPP: tool for regression of model parameters (from Dechema) MS-Excel Add-In MATLAB Toolbox ProSim Software Suite Simulis Thermodynamics Export files A.P.I. Simulis® Thermodynamics can be used within DPP

Why data regression is needed?
Binary interaction parameters (BIP) stored in databases provided with process simulators have some limitations: They are available for a limited number of systems They have been regressed with models which are not always exactly the same as those used in the process simulator Vapor pressure laws of pure components stored in the simulator does not correspond exactly to the equilibrium data available They have been regressed using only vapor-liquid equilibrium data. Consequently, they give good results for vapor-liquid equlibrium prediction, but often very bad results for other properties: excess enthalpies,…

DPP and Simulis® Thermodynamics
DPP: Data Preparation Package from Closes the gap between raw thermophysical data (measured discrete values) and model based process simulation packages Besides selection and graphical display of data sets DPP permits the regression of model parameters as well as the comparison of models with each other (eg comparison of the gamma models NRTL vs. Wilson vs. UNIQUAC or comparison of vapor pressure equations like Antoine vs. Wagner).

Neutral File Interface (IK-CAPE PPDX)
The architecture of DPP Databases: DETHERM (Dechema) DIPPR TRC In-house etc… Raw data GUI Graphics subsystem Optimizer Neutral File Interface (IK-CAPE PPDX) Thermo Interface DPP Thermodynamic modules: Aspen Properties IK-CAPE Thermodynamics Simulis® Thermodynamics Since parameters must be identified with the model where they will be used, DPP does not contain any thermodynamic model Simulators: AspenPlus ProII ProSimPlus etc… Model parameters

The optimizer: the heart of DPP
Allows simultaneous regression of several experimental data types for instance: VLE + LLE + HE + ∞ + Azeotropes Possibly giving different weights to some data sets Using different error functions Least-squares, 2, Absolute, Maximum-Likelihood + 2, ... Possibly bounding some variables or setting equality constraints to some of them Using optimization algorithms adapted to the kind of problem to be solved Simplex (Nelder-Mead), Powell, Fletcher-Reeves, Broyden-Flechter-Goldfarb-Shanno, Gauss-Newton modified

traditional thermo package
The workflow with a traditional thermo package Specify problem Selection of compounds Selection of models Save problem as *.yyy file Within X Thermo package Load raw data Databases : DETHERM (Dechema) DIPPR In-house etc… Raw data Load *.yyy file Within DPP Optimize DPP Optimizer X Thermo (DLLs) (models and VLE flashes) The work performed can be used only within X Inc. products The interface must be updated with new releases of X Inc. products (several interfaces must be maintained) X Inc. products Regressed model parameters Export parameters as *.xxx file

(AspenPlus, HTRI, gPROMS, …) (AspenPlus, HTRI, gPROMS, …)
The workflow with Simulis® Thermodynamics Specify problem Selection of compounds Selection of models Direct access to Simulis® Thermodynamics "calculators" editor Within DPP Databases DETHERM, DIPPR, TRC, In-house, etc… Raw data Load raw data CAPE-OPEN compatible modeling environment (AspenPlus, HTRI, gPROMS, …) CAPE-OPEN compatible modeling environment (AspenPlus, HTRI, gPROMS, …) Client software MS-Excel, MATLAB, ProSimPlus, etc Generate a "Simulis Thermo Package" CAPE-OPEN Compliant Property Package The thermo model which better fits the experimental data for the considered system Optimize DPP optimizer Simulis® Thermodynamics (models and VLE flashes) Simulis® Thermodynamics

The "Data Package" concept
In large organizations the thermodynamic model can be prepared by a "thermodynamics expert". He will: validate the pure components properties, adjust the thermodynamic approach, fit the binary interaction parameters, etc… in order to get a reliable modeling of a particular system With Simulis® Thermodynamics, the expertise of this specialist will be available to the whole organization

He will perpetuate and secure this expertise creating a "data package" The "data package" contain all the information requested to perform calculations on a particular system (data, models, parameters)

This "data package" will be easily dispatched thanks to the set-up script automatically generated… …which will be sent by or through the intranet of the company to end-users

End-users will use this reliable model: within their usual software (data packages are CAPE-OPEN compliant) with the same efficiency without thinking about thermodynamics anymore without risk of uncontrolled modification with full consistency with other calculations performed on the mixture

However, no "true re-entrance" between different modes
"Expert mode" Combination of VBScript, external DLL and native models is supported H (Cp) Cp User DLL for calculation of H and Cp H (Cp) However, no "true re-entrance" between different modes H (Cp) Cp If the engineers in electronic material were to start from a sand heap each time that they design a new device, if their first stage always had to consist in extracting the silicon to make integrated circuits, they would not progress very fast. Native DLL for calculation of H and Cp

You can use your favorite thermo package compliant only with Thermo 1.0… CAPE-OPEN "socket" External software able to generate CO packages (Aspen Properties, Multiflash, PPDS,…) 1.0 CAPE-OPEN "plug" Modeling tool implementing CO Thermo Socket Aspen Plus, ProII, Aspen HYSYS, HTRI, gPROMS … 1.1 …in an application only compliant with Thermo 1.1… …and vice-versa 1.1 1.0 Simulis Thermodynamics

Simulis® Thermodynamics

Similar presentations

Presentation on theme: "Simulis® Thermodynamics"— Presentation transcript:

Similar presentations

About project

Feedback

Log in

Auth with social network:

Simulis® Thermodynamics

Similar presentations

Presentation on theme: "Simulis® Thermodynamics"— Presentation transcript:

Similar presentations

About project

Feedback