TM

Toward rigorous heat integration tools for coal-firedToward rigorous heat integration tools for coal fired power plants with CO2 capture and compression

N.V.S.N. Murthy Konda and David C. MillerNational Energy Technology LaboratoryNational Energy Technology Laboratory

11th Annual Conference on Carbon Capture, Utilization & Sequestration April 30 - May 3, 2012, Pittsburgh

OutlineI t d ti• Introduction– Carbon capture: Key challenges and opportunities– Key objectives and scope of this worky j p

• Overview of heat integration approach

• Modeling of supercritical power plant

• 2 stage sequential optimization approach– Results and discussion– Results and discussion

• Conclusions and future work

2

TM

Carbon Capture Simulation Initiative

Risk Analysis & Decision Making Framework

Uncertainty Quantification Framework

n rk Particle &

New Capabilities

New Capabilities

New Capabilities

Uncertainty Quantification Framework

Inte

grat

ion

Fram

ewor

Basic Data

ROMsParticle & Device Scale

Simulation Tools

Plant Operations & Control

Tools

Process Synthesis &

Design Tools

CarbonCapture

CarbonCapture

CarbonCapture

Tools

pDevice Models

System Models

pDynamic Models

3

TM

Surrogate Models

Process Synthesis

Superstructure for Optimal

Process

Simultaneous Superstructure

ApproachP H t

Flexible Modular Models

PC Plant ModelsThermoflowAspen Plus

(Alison Cozad)Process

Configurations(Zhihong Yuan)

Power, Heat, Mass Targeting

(Linlin Yang)

New SorbentSolid Sorbent Carbon

Capture Reactor Models

Aspen Plus(John Eslick)

Heterogeneous Simulation-Based Optimization Framework

el

New Sorbent Models

(Task 1, EFRC)

Capture Reactor ModelsACM, gPROMS

(Andrew Lee, Hosoo Kim)

Compression System

Pla

nt M

odel

sion

Sys

tem

Mod

e

Black Box Optimization(Hosoo Kim)

Compression System Models

Aspen Plus, ACM, gPROMS

(John Eslick)

PC

Com

pres

s

Other carbon capture models

Aspen Plus, ACM, gPROMS, GAMS

(J. Eslick, Juan Morinelly)

Heat/Power Integration

Automated GAMS Formulation/Solution(Murthy Konda)

External Collaboration

(Prof. Phil Smith, ICSE)

Industry Specific

Collaboration(ADA)

Oxy-combustionGAMS, Aspen Plus,

ACM, gPROMS, (Alex Dowling)

4

TM

M d l f th i iti l d t ti f CCSI P S th i d D i T l t

Heat integration approachModels from the initial demonstration of CCSI Process Synthesis and Design Toolset

CO2 Compression System (Simulation Model)

CO2 Capture System(Simulation Model)

Steam cycle(Simulation Model)

Excel interface for exporting ‘stream data’ (SINTER)

Steam Cycle Heat Exchanger Network yOptimization Model

(Mathematical model)(HEN) Optimization Model

(Mathematical model)

Integrated AnalysisMinimize energy penalty on power plant

5

TM

Minimize Cost Of Electricity (COE)

Supercritical power plant steam cycle

6

TM

Modeling of power plant steam cycle

• Basis: Simulation developed in Thermoflex

• Optimization model is developed in GAMS

• GAMS model: algebraic representation of steam cycle– Mass and energy balances – Correlations for steam/water enthalpy prediction– Corrections for turbine efficiency

7

TM

Modeling of power plant: some aspectsA t th l l ti f ti f t t d• Accurate enthalpy correlations as a function of temperature and pressure

0.20%

0.30%

or

% error in liquid enthalpy predicted by correlation

0 20%

-0.10%

0.00%

0.10%

% e

rro

• Exhaust losses are taken into account and developed efficiency correlations for the last LP turbine stage

-0.20%0 50 100 150 200 250 300 350

Temp (C)

for the last LP turbine stage

100

y (%

)

Isentropic Efficiency Vs Steam Inlet Flow (last turbine stage)

0

50

0 20 40 60 80 100 120Effic

ienc

y

8

TM

Steam inlet flow (10^4 kg/hr)

• Solid sorbent carbon capture process is from Chang et al., 2011Steam extraction for sorbent regeneration

p p g ,– 2 bubbling fluidized bed adsorbers and 1 moving bed regenerator– Steam required: 138 GJ/hr/train (10 parallel regenerator trains)

• Steam extraction from IP/LP crossover (@100 psi)C d d i d h d i h BFH i– Condensed steam is returned to the deaerator in the BFH section

CO2 Capture Process IP/LP Steam extraction

9

TM

Chang et al. (2011). Synthesis of optimal adsorptive carbon capture processes. AIChE annual meeting, Minneapolis, MN

2 stage sequential optimization methodology

1. Optimize steam cycle for required steam extraction ratei Determine feasible temperature profile in BFHi. Determine feasible temperature profile in BFH

section subjected to the amounts/quality of available ‘heat sources’ from capture & compression systems

ii. Min. parasitic loss assuming available heat can be used.

2. Determine optimal matches (location of HX’s) to integrate heat from capture & compression back into g p psteam cycle to achieve 1.

10

TM

Stage 1: Steam cycle and HEN model integrationAll the ‘heat so rces’ in the CO capt re and compression processes generall• All the ‘heat sources’ in the CO2 capture and compression processes generally provide ‘low-grade heat’ (i.e., < 2500C). – This heat can be fed into the steam cycle through the BFH section and

cannot directly be used in the boiler to produce HP steam– In practice, part of this heat may also be used to produce LP steam to drive,

for instance, auxiliary equipment• Heaters (red circles) are assumed in the LP BFH section and these heaters can

use ‘hot energy’ from CO2 capture & compression processes

11

TM

H t ( d i l ) h tili d h t f t d

Stage 1: Results• Heaters (red circles) have utilized heat sources from capture and

compression process (and hence temperature rose across heaters)• Steam extraction at turbine stages is lowered to improve efficiency.

• Legend:• Basecase (before heat integration) conditions are shown in black• Case 2 (after heat integration) conditions are shown in red

15.1 kg/s17.9 kg/s17.0 kg/s0.0 kg/s 0.0 kg/s 3.4 kg/s

LP1 LP2 LP3 LP4

Temp (C) profile in the BFH section:128 1 128 1 96 6 96 6 62 4 62 4 34 7 34 7

Condenser

128.1 128.1 96.6 96.6 62.4 62.4 34.7 34.7120.2 114.0 87.4 69.5 66.2 66.1 51.1 37.8

HXLP3HXLP2HXLP1

12

TM

Stage 1: Results (continued)

• Basecase (without heat integration): – Net efficiency: 44 3% 33 8%Net efficiency: 44.3% 33.8%– Gross power: 709 MWe 553 MWe

• Case 2: i.e., with heat integration (by using heat sources available in CO2 capture and compression processes)– Net efficiency: 33 8% 36 4%Net efficiency: 33.8% 36.4%– Gross power: 553 MWe 591 MWe

13

TM

Stage 2: Model for optimal HEN design• Given:

– Process and utility stream data • Tin, Tout, heat capacity & heat transfer coefficients

– Cost dataCost data

• Objective is to:– Minimize total cost of the network

• While optimizing (i.e., decision variables):– Hot and cold stream matches (binary variables)– Load approach temp and area of each heat exchanger (continuousLoad, approach temp and area of each heat exchanger (continuous

variables)

• Formulation (Yee and Grossmann, 1990)– Multi-stage network with isothermal mixing

Yee and Grossmann (1990) Simultaneous optimization models for heat integration 2 Heat exchanger network synthesis

14

TM

Yee and Grossmann (1990). Simultaneous optimization models for heat integration – 2. Heat exchanger network synthesis.

Stage 2: Stream data for hot and cold sources

HOT STREAM DATA COLD STREAM DATA

StreamName

TIN (C)

TOUT (C)

MCp(MJ/hr-C)

StreamName

TIN (C)

TOUT (C)

MCP(MJ/hr-C)Name (C) (C) (MJ/hr-C) Name (C) (C) (MJ/hr-C)

FlueGas_Cooler 81. 43. 13646. BFW heater3 38. 51. 2,200ProdCO2_Cooler 71. 40. 22933. BFW heater2 66. 66. 2,201

IC_01 86. 43. 612. BFW heater1 69. 87. 2,209IC 02 88 43 643 BFW h 114 120 2234IC_02 88. 43. 643. BFW heater0 114. 120. 2234IC_03 86. 43. 563.IC_04 88. 43. 513.IC_05 87. 43. 490._IC_06 89. 43. 488.IC_07 254. 60. 802.

15

TM

Stage 2 Results: HEN configurationL dLegend:Dashed lines represent HEN connections and circles on the dashed lines represent l ti f h t h ith

10 stage CO2 compressor trainlocations of heat exchange with BFH section

Note:HEN fi ti iHEN configuration is preliminary and further improvements/investigation are necessary

254.5 C151.6 C

120.2 114.0 87.4 69.5 66.2 66.1 51.1 37.8

Intercooler 7 (from compression system) is sufficient to satisfy heating l d th BFW h tloads on the new BFW heaters (red circles). Heat available in flue gas cooler and product CO2 cooler (from capture

) t d d

BFWheater1

BFWheater2

BFWheater3

BFWheater0

LP boiler feed heating (BFH) section

16

TM

process) are not needed.LP boiler feed heating (BFH) section

Conclusions and future workRi ti i ti d l f iti l l t i• Rigorous optimization model for a supercritical power plant is developed– Optimization of BFH temperature profile can help reducing energy p p p p g gy

penalty due to steam extraction (for sorbent regeneration).

• Net efficiency can be improved from 33.8% to 36.4%.

• Work in progress:– Integrate into single stage, simultaneous algorithmg g g , g

17

TM

Thi t ti d t f k

DisclaimerThis presentation was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, 

f h i l knor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, 

d di l d h iapparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade 

d k f h dname, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

18

TM

18

TM