Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
1
Energy Systems Initiative Center for Advanced Process Decision-Making
L. T. Biegler Department of Chemical Engineering
Carnegie Mellon University Pittsburgh, PA 15213
http://capd.cheme.cmu.edu
March, 2012
2
Provide intellectual leadership on complex design and operational problems faced by process industries Science base: optimization, control, computer science, systems engineering, business
CAPD Goals
3
CAPD Collaborations w/ NETL and More
Objectives • Accelerate R&D on advanced models,
methods, and tools for process systems engineering
• Apply to existing and emerging fossil energy systems, especially for carbon capture, utilization and storage (CCUS)
• Address technical barriers across power plant lifecycle-process innovation, design, operations, and management
Additional ESI Efforts • Process development, modeling and control
for PV solar cells
• Related interests with CAPD members
• New Initiatives with GS E&C as well as MERL
Transport Gasifier
Gas TurbineCombustorHRSG
EntrainedFlow
Gasifier
Transport Gasifier
Gas TurbineCombustorHRSG
EntrainedFlow
Gasifier
APECS Co-Simulation of IGCC-CCS Plants
CCSI for Power Plants
Energy Plant Lifecycle
4
ESI Agenda – March 11, 2012
1:30 Introduction Larry Biegler 1:35 AVESTAR Center for the Operation and Control of Steve Zitney
Clean Energy Plants
2:05 Planning, Synthesis of Energy Processes Ignacio Grossmann
2:10 Supply Optimization of Biofuels in Argentina Federico Andersen
2:30 Heat and Water Integration Linlin Yang
2:50 Optimization Modeling of Energy Processes Larry Biegler
2:55 Optimization of PSA units for Carbon Capture Alex Dowling
3:15 Reduced Order Modeling for CFD Units Yi-dong Lang
3:35 Refreshment Break
5
ESI Agenda – March 11, 2012
4:00 Learning, Sequestration and Green Computing Nick Sahinidis
4:05 Derivative Free Optimization for CO2 Capture Alison Cozad
4:25 Risk assessment for CO2 sequestration Yan Zhang
4:45 Modeling and Control of Silicon Solar Cells Erik Ydstie
5:05 Crystallization Modeling for Solar Cells German Oliveiros
5:10 Chemical Looping Control Tim McFarland 5:30 An Overview of the US Department of Energy’s David Miller (NETL)
Carbon Capture Simulation Initiative
6:00 Discussion and Wrap-up
7:00 CAPD Welcome Reception, CAPD Conference Room (DH 4200)
6
Optimization of PSA units for Carbon Capture (Alex Dowling)
Compressor
Feed ( L + H )
Pbed
L
Pbed
H
Feed Pressurization
Feed (Adsorption)
Counter-current Depressurization
Light reflux (Desorption)
7
PSA Bed Model
( ) izvC
tq
tC ii
sbi
b ∀=∂
∂+
∂
∂−+
∂
∂ 0)1( ρεε
)( *iii
i qqktq
−=∂
∂
2,1exp
11
4321
12
22
11
11*
=∀⎟⎟⎠
⎞⎜⎜⎝
⎛=+=
++
+=
∑∑==
miTkkbTkkq
Pyb
Pybq
Pyb
Pybqq
mimimimimi
smi
nc
jjj
iisi
nc
jjj
iisi
i
( ) ( ) vvCM
dv
dzP i
iw
bp
b
bp
b⎟⎟⎠
⎞⎜⎜⎝
⎛−+
−=
∂
∂− ∑
1000175.11150332
2
εε
εεµ
Component mass balance
LDF equation
Dual-site Langmuir Isotherm
Energy Balance
Ergun equation
( ) ( ) ( )
∑
∑∑
=
==
=+++=
=−+∂
∂+
∂
∂Δ−
∂
∂⎟⎟⎠
⎞⎜⎜⎝
⎛⎟⎠
⎞⎜⎝
⎛++−
nc
i
ipgi
ic
ic
ic
ic
ipg
wc
nc
i
iis
wpwwpss
nc
i
ipgit
TCChTdTcTbaC
TTDh
zvh
tqH
tT
DhCCRCC
1
32
11044
ρρρε
Ideal gas ∑=i
iRTCP
8
Two-bed PSA Superstructure Systematic formulation to develop, evaluate and optimize PSA cycles
Co-currentBed
(CoB)
Counter-current Bed
(CnB)
Light Product (LP)
Pressure-reducingValve
Top reflux (TR)
b(t)
Bottom reflux (BR)a (t)
Heavy Product (HP)
Heavy-product compressor
Feed compressor
Vacuum Generator
Patm
Pd(t)
Pa(t) Cd,i(t), Td(t),vd(t), Pdes(t)
f (t) Feed
Ca,i(t), Ta(t),va(t), Pads(t)
Input flux (F)
Inlet gasPfeed
Inletcompressor(optional)
Pinlet
Allows all steps (P-FD-DP-EV-EQ-HP-LP) Includes most steps with 2-bed interactions Extend to product tanks
9
Motivation for CO2 Capture
Challenges
- Which Sorbent works best for CO2 capture? - Which PSA cycle for high purity CO2 capture? - Computationally efficient flowsheet simulation/optimization with PDAE-based PSA model.
IGCC Existing pulverized coal plants
Post-combustion capture Pre-combustion capture
Can we use PSA for carbon capture?
10 3/30/12 10:20 10
Reduced Order Modeling for CFD Units(Yi-dong Lang)
CFD Model
11
ROM Development for CFD Models PCA-based Input-Output Mapping
Exp’l Design (e.g., LHS)
Modeling Strategy
Set of inputs
Set of solutions
PDE CFD
Input
Output
Snapshot Developing ROM
Yred
Mapping Scores
PCA
F(u)
(ROM)
Input-Output Mapping (Kriging)
?
Lang, Y-D et al., Energy and Fuels, 23, 1695 (2009)
12
ROM for Turbine Combustor
FLUENT
PCA
ROM
12
Average Fluent Case ~2000 CPU sec
Each case in ROM < 1 CPU sec Lang, Y-D et al., Energy and Fuels, 23, 1695 (2009)
13
ROM for Entrained-Flow Gasifier
Avg. Fluent Case 72 000 CPUs Avg. ROM Case < 1 CPUs
14
TNO-IGCC Process Flowsheet
TNO-Report R98/135 (1998)
Coal Gasification
Air Separation
Unit
Water gas shift
Reactor I
CO2 Capture
TSA
Gas Turbine
Water gas shift
Reactor II
Steam Cycle and
Steam Turbines
CO2 Compressor CO2, 110 bar
CO2, 19 bar
Heat Exchangers
Heat flux Steam
Coal
Air
Combustor
• Detailed Optimized Economic Study for IGCC with CCS • Various CCS technologies considered (e.g., Selexol, TSA) • Can Optimization with CFD models improve on this design?
15
TNO-IGCC and integration of replaced CFD models
Water shift
TSA and CO2 compress
Gas-Combustor-Turbine
Steam cycle
Gasif
USER3
Cmbst USER3
• Allows fast EO-based optimization with Aspen • Leads to 7% increase in energy efficiency in IGCC process