열역학물성연구회 2003 년 10 월 13 일발표자료 열역학물성연구동향 및 KDB 응용프로그램의개발 고려대학교화공생명공학과강정원
발표내용 열역학물성관련연구동향 CAPEC, Technical University of Denmark 센터소개 열역학물성분야의표준화방향 KDB 응용프로그램의개발 KDB 응용프로그램개발현황및문제점 향후개발방향
CAPEC, DTU 소개 Computer Aided Process Engineering Center Professor Rafiqul Gani Professor Sten Bay Jorgensen Professor Jens Abildskov Professor Niels Jens Research Area Program A : Physical Properties Program B : Process Modeling, Simulation and Idetification Program C : Process Synthesis, Product / Process Design and Analysis Program D : Process Control and Operation Program E : Numerical and Computational Aspects Program F : Safety and Risk Analysis
Research Activities List of Softwares ICAS : Integrated computer aided system TML : Thermodynamic Model Library ProPred : Property Prediction TMS : Thermodynamic Model Selection PDS : Process Design Studio PROCAMD : Computer Aided Molecular Design MoT : Model Testbed
Collaboration Consortium FMC, USA Bayer, Germany Mitsubishi, Japan ICI, UK Syngenta, UK Novo Nordisk, Denmark PDC, Germany Danisco Ingredient, Denmark and many more (About 20 ) Academic Prof. J. P. O Connel (Univ. of Virginia)
Tools for Thermodynamic Properties and Phase Equilibrium Thermodynamic Properties Database TML (Thermodynamic Model Library) TMS (Thermodynamic Model Selection) ProPred 3.5 (Pure component property prediction Too) ProCAMD (Computer-Aided Molecular Design Software)
Activities in CAPE Problems Choosing among Process Alternatives Feasibility Study Thermodynamic Model Selection Parameter fitting for the Thermodynamic Model Simulation Design Calculation Optimization Properties of Equilibrium unknown molecules Analysis Propred 3.5 Model Selection Guide TMS Thermodynamic Model Selection Regression Input Regression Result TML Thermodynamics Model Library Process Conditions Properties Values Pure component properties predeiciton Pure properties Export / Import CAPEC Thermophysical Properties Database - Pure component properties and parameters - Mixture properties and parameters - Experimetal data
Research Development of 2 nd order UNIFAC method KT-UNIFAC 1 st order and 2 nd order New parameters Matrix Development of KT-UNIFAC Utility Program Phase equilibrium for surfactant solution
Development of UNIFAC Method Uncertainties for prediction using current GC method are quite high Current GC methods are generally unable to distinguish isomers and reflect effects of adjacent groups Second order UNIFAC can overcome deficiencies of current GC methods C lnγ ln ln R2 i = γ + γ + w i i R lnγ R2 i First oder contribution Second order contribution
Development of new UNIFAC method Example of group decomposition of molecules 2-Propanol M-Xylele CH 3 CH CH 3 OH CH 3 CH 3 First-order groups 2 CH3 1 CH 1 OH Second-order groups 1 CHOH First-order groups 4 ACH 2 AC-CH3 Second-order groups 1 ARS1S3
KT-UNIFAC, Model Description Combinatorial Term Residual Term Second-Order Term C C i C i SG i FH, ln, ln ln γ γ γ + = i J i J C i,fh γ ln 1 ln + = + = i i i i i C SG i L J L J q Z ln 1 2 ln, γ = NMG k k ki s ki G k ki s i L i q i R η η γ ln ) ln (1 ln ( ) + = NSOG j NSG j m NSG m n ni ni m i n ni m n m i j m n R i s mn G G mn V G RT u τ η τ ϑ ϑ δ γ ' ' ', ' 2 ln ( ) mi nm m i ni m ni s G G V n G RT u nm m i m j nm τ η τ ϑ ϑ δ ' ' ' ',
New parameters for 1 st order groups 1 CH2 1 2 C=C 2 X 3 ACH 3 X X 4 ACCH2 4 X X X 5 OH 5 X X X X 6 MeOH 6 X X X X X 7 H2O 7 X X X X X 8 AcOH 8 X X X X 9 X Parameters Available CH2CO 9 X X X X X X X X 10 CHO 10 X X X X X X X X 11 Parameters NOT Available CCOO 11 X X X X X X X X X 12 HCOO 12 13 CH2O 13 X X X X X X X X X X 14 CNH2 14 X X X X X X X 15 CNH 15 X X X X X X X X X 16 (C)3N 16 X X X X X X X X X 17 ACNH2 17 X 18 PYRIDINE 18 X X X X X X X X 19 CCN 19 X X X X X X X X X X X 20 COOH 20 X X X X X X X X 21 CCL 21 X X X X X X X X X X X X 22 CCL2 22 X X X X X X X X X X X X X X X 23 CCL3 23 X X X X X X X X X X 24 CCL4 24 X X X X X X X X X X X X X X X X 25 ACCL 25 X X 26 CNO2 26 X X X X X X X X X X X 27 ACNO2 27 28 CS2 28 X X X X X X X X X X X 29 CH2SH 29 X X X X X X X X 30 FURFURA 30 X X X X X X X X X X X 31 DOH 31 32 I 32 X X X X X X X X X X 33 BR 33 X X X X X X X X X X 34 C#C 34 X X X X X X X X 35 DMSO 35 X X X X X X X X X X X 36 ACRY 36 X X X X 37 ClC=C 37 X X 38 ACF 38 x X X X X X X X 39 DMF 39 X X X X X X X X X X X X X X 40 CF2 40 X X 41 COO 41 X X 42 SiH2 42 X X 43 SiO 43 44 NMP 44 X X X X X X 45 CCLF 45 X X X X X X 46 CONCH2 46 X X X X X X 47 OCCOH 47 X X X X X X 48 CH2S 48 X X X X X 49 Morpholine 49 X X X X 50 Thiophene 50 X X X X X X 54 CH2(cyc) 54 X X X X x X X X X X X X X X X X X X X X X X X X X X 60 ACO 60 X X 63 ACCN 63 X X X 69 ACBr 69 X X X 74 SO2 74 93 PO4 93 102 O=COC=O 102 X X X X X X X X 103 OCOO 103 X X 104 SO2(cyc) 104 105 EPOXY 105 X X X X X X 106 NCO 106 X X X T dependency is similar to Linear UNIFAC (Hansen et al.) τ Δu = exp( T ) mk Δu mk mk = a mk, 1 + a,2( T T0 mk First order parameters were obtained from 4413 Data Sets (VLE, H E, γ inf ) New groups were included (Nitriles, epoxy,...) )
New parameters for 2 nd order Second Order Group Parameters were fitted from 842 experimetal data sets (VLE, H E, γ inf ) Property UNIFAC Modified UNIFAC First Order New Method Second Order AAE ARE (%) AAE ARE (%) AAE ARE (%) AAE ARE (%) P 3.50 3.47 4.09 3.89 1.95 2.69 1.78 1.99 y 1 0.0153 3.64 0.0147 3.60 0.0129 3.04 0.0112 2.79 H E 388.82 192.03 476.62 1167.60 137.94 20.33 115.79 20.20 γ inf 89.53 39.41 68.07 58.81 134.96 25.78 124.97 25.11
Comparison of calculated result Improved prediction due to new groups 100 80 Experimental Data Original UNIFAC Modified UNIFAC KT-UNIFAC, First Order 50 45 40 Experimental Data Original UNIFAC Modified UNIFAC KT-UNIFAC, First Order Pressure (kpa) 60 40 20 Pressure (kpa) 35 30 25 20 0 15 0.0 0.2 0.4 0.6 0.8 1.0 Mole Fraction 10 0.0 0.2 0.4 0.6 0.8 1.0 Mole Fraction Figure 3. Comparison of experimental and predicted result for VLE (hexane + methoxy benzene (anisole) at 333.25 K) Figure 4.Experimental and predicted result for VLE (1,2-epoxybutane + n-heptane at 313.15 K)
Comparison of calculated result Improved prediction due to second order group 45 30 28 40 26 Pressure (kpa) 35 30 25 Experimental Data Original UNIFAC Modified UNIFAC KT-UNIFAC, First Order KT-UNIFAC, Second Order Pressure (kpa) 24 22 20 18 16 14 Experimental Data Original UNIFAC Modified UNIFAC KT-UNIFAC, First Order KT-UNIFAC, Second Order 20 0.0 0.2 0.4 0.6 0.8 1.0 Mole Fraction 12 0.0 0.2 0.4 0.6 0.8 1.0 Mole Fraction Figure 7. Experimental and predicted result for VLE (Hepatne + 2-Methyl-Isobutyl-Ketone at 348.15 K) Figure 8. Experimental and predicted result for VLE (hepatne + 2-methyl-2-butanol at 328.15 K)
Comparison of calculated result Improved prediction due to second order group 400 1600 Excess Enthalpy (J/mol) 300 200 100 Experimental Data Original UNIFAC Modified UNIFAC KT-UNIFAC, First Order KT-UNIFAC, Second Order 0 0.0 0.2 0.4 0.6 0.8 1.0 Mole Fraction Excess Enthalpy (J/mol) 1400 1200 1000 800 600 400 Experimental Data Original UNIFAC Modified UNIFAC 200 KT-UNIFAC, First Order KT-UNIFAC, Second Order 0 0.0 0.2 0.4 0.6 0.8 1.0 Mole Fraction Figure 10. Experimental and predicted result for HE (benzene + 1,2,3,4-tetrahydronaphthalene at 298.15 K) Figure 11. Experimental and predicted result for Excess Enthalpy (benzene + isopropanol at 308.15 K)
UNIFAC Utility Database Management Tools UNIFAC Groups and Group Assignment Tables UNIFAC Interaction Parameters VLE Experimental Data (CAPEC DB and User Database) Parameter Regression and Analysis Tool
UNIFAC Group Data Management Tools
Regression Analysis Tool
Software Demo CAPEC Tools UNIFAC Utility
열역학물성관련연구동향 자료의표준화 (Standardization) 효율적인정보의공유와유통을위하여필수적임 자료의검증 (Validation) 적절한절차를통하여검증된정포를포함해야함 QM/MM 기반의물성 DB 상업화, 대형화, 집중화
Standardization Efforts PDXI (Properties Data exchange) by AIChE Move on to CAPE-OPEN CAPE-OPEN (European Committee) Too abstract COM Model Self-ML (CODATA Group, Kehiaian) Project ended (funding too!) CML (Chemical Mark-up language) Chemical Structure matml (Material Mark-up language, NIST) ThermoML (TRC, NIST) Collaboration with J. Chem. Eng. Data
Why XML? Minimum specification small set of rules (easily understood) Formats and actual data are integrated Don t have to worry about data structure Extensible Easily portable to RDBMS structure
XML 처리과정
ThermoML M. Frenkel et al., J. Chem. Eng. Data, 48, 2 (2003) Specification Experimentally measurable physical properties and transport properties data (120 properties) Mixture phase equilibrium and reaction data Cooperative data processing between J. Chem. Eng. Data and TRC
Example
Information flow architecture
국내연구동향 KISTI Accepted ThemoML as a new standard
KDB 연구연구개발동향 KDB 현황 웹버전으로개발되어서비스중 현재화학공학연구정보센터에서서비스중 매월 8000여사용자 지속적인업데이트가어려운실정임
KDB 연구연구개발동향 현재 KDB 의문제점 자료와계산프로그램이별도로운영되고있음 많은양의자료를효과적으로교육및연구에활용하기에불편함이많음 지속적인사업을하기위한예산확보의어려움 효과적인홍보전략이필요할것으로보임
열물성데이터베이스의교육용소프트웨어개발의사례 ICAS 5.0 CAPEC, DTU Phase W. Chapman, Rice University
문제점해결방안 -1 KDB 의활용 열역학및분리공정과목등의학생교육에활용 DDB 및 KDB 를통합한인터페이스를개발하여연구목적으로활용 모델의개발 (G-NLF-HB EOS) 기타물성그룹기여방법개발등 해결전략 VB.NET 를활용한통합인터페이스및응용프로그램의개발 KDB 및 DDB 의표준화 (ThermoML 의표준도입 )
KDB work in progress DB Format conversion (Incorporating ThermoML) Thermodynamic calculation methods EOS : SRK/PR/NLF-HB Activity : NRTL, UNIQUAC, UNIFAC Calculations VLE / SLE / LLE /SVE Regressions Pure component properties VLE regression Group Contribution regression Planning to finish before the end of November
KDB-Thermo LAB KDB Thermo LAB DB Utility Pure Properties Mixture Properties KDB Data Management Tools Properties Calculation Phase Equilibrium Calculation Properties Regression Phase Equilibrium Regression DDB Data Interface Group Contribution Methods Group Contribution Methods ThermoML