Plant Optimization & Performance

Software

Samir Pandya
Power Industry Specialists
z Emerson Power and Water Solutions
has over 50 years of experience in
Power
z Unmatched Application expertise
z Boiler and Turbine controls for
different equipment manufacturer’s.
z With a rich experience in Power we
are uniquely qualified to offer
optimization solutions like advanced
process control.
Changing Face of the Industry
z Deregulation has spawned fierce market competition
– Mergers, acquisitions and internal restructurings for increased
cost competitiveness are common place
– Independent power producers build, own and operate plants at
highly competitive rates
z Government and environmental regulations are increasing
– Power Producers have to reduce Nox to meet compliance

To stay ahead of the competition, it is imperative that

power producers improve plant efficiency and increase
output, decrease costs and maintain environmental &
regulatory compliance.
Improvement goal
z Heat rate improvements 1/2% to 1 1/2%
z NOx Reductions 15% to 35%
z Opacity Reductions 15% to 30%
z Increased Capacity 1% to 2%
z 1% of MCR per minute improvement in ramp rate
z Reduced tube leaks and associated forced
outages
z Improve fleet management capabilities
Environmental Management Unit Performance

z NOx / SO2 cap compliance z Heat Rate

z NOx / CO minimization z Boiler Efficiency
z Opacity Reductions z LOI

Generation Management Operational Flexibility

z Performance Advisor z Ramp Response

 Optimization Software
Fleet Emissions Optimizer
Economic Optimizer

Combustion Optimizer (NOx, Heat Rate, Opacity)

Cyclone Boiler Optimizer

Steam Temperature Optimizer

Sootblower Optimizer

Global Performance Advisor

Enterprise Data Server

SmartProcess Features Two Types of Plant-
Specific Modules That Use Intelligent Software
Tools

Use knowledge-based software tools such as linear

models, neural networks and fuzzy logic to model
Optimizers and optimize generating units and send set-points &
biases directly to the existing plant DCS
(closed-loop integration)

Use state-of-the-art mathematical and modeling tools

to analyze process performance and provide operator
Advisors advisory messages identifying problem areas that are
generating unnecessary costs
r ld to
(open-loop integration) e wo
th
in izrs
e
e few pt im
f t h o p o
e o d l o
On cl o se
f fe r
O
What makes SmartProcess different?

Advanced control and optimization solutions incorporate

fuzzy logic, neural networks, model predictive
control, and optimization engines designed specifically
for the power industry needs

Browser-based user interfaces

Closed loop integration – Integrates directly with any

DCS or can be deployed via other protocols like OPC or
OSISoft PI

Automated testing tools for quick, efficient

implementation
Model Predictive Control

measurements

Process model Prediction

(dynamic) over horizon
sensors

optimization
BFGS, Powell stop?
setpoints

constraints
control module
 wc wg 1 1
1 ∑ ƒ ∑ Π

Neuro-Fuzzy ∑ ƒ 1
∑
Π

Π
∑ ƒ
Architecture 1
∑ ƒ
-1

1
∑

wc wg 1
1 ∑ ƒ ∑

∑ ƒ 1
∑ Π
∑ ƒ -1
x1 Π
1 1
z The way to combine NN’s (A)
∑
(B)
ƒ
(C) (D)
∑ Π
(E)
learning properties with P R E M I S E S
wa
qualitative knowledge of 1 1
∑
y

∑
fuzzy systems 1
∑
x2
wa
1
Π
ws

wa
1

ws
∑

(F) (G) (H) (I) (J) (K) (L)

C O N C L U S I O N S
What makes SmartProcess different?

Versatility to optimize for multiple objectives under

varying conditions

Dynamic routines that steady-state approaches cannot

match – Optimizer runs every 10 to 20 seconds,
outperforming all other comparable products

Adapts and learns changing plant conditions

DCS platform independent

What makes SmartProcess different?

No daily maintenance – SmartProcess self adapts to the

subtle long term changes in the plant dynamics

Data validation tools and comprehensive support

capabilities available

Easily upgraded with base plant control system

upgrades/migrations
Typical SmartProcess Architecture
Plant LAN

GPA
Operator
EWS
Station
2nd Ethernet ports

DCS Network

DCS
Controllers

I/O
Economic Optimizer
The Economic Optimizer
enhances energy
allocation and plant
operation, based on a
number of factors,
including operating
costs, equipment
efficiencies, and
operating schedules.
Unify islands of optimization
with an overall plant model
z Applications
– Fleet wide economic
analysis
– Reduces operating costs on
multiple equipment type
plant configurations
– CHP, Combined cycle
plants, Co-generation
facilities
– Pumping networks
– Fuel blending strategies
– Cooling tower optimization
Fleet Optimizer
z Results
– Replicates and/or
The Fleet Optimizer diversifies calculations used
manages for business decisions
environmental – Provides reporting for O&M
compliance on a costs and real time heat
rate
fleet-wide scale with – Predicts emission cap
portal technology. compliance based on load
forecasts
– Actively optimizes plant
settings to achieve desired
compliance target margins
– Provides data redundancy
Operate cost-effectively while of key variables
achieving SO2 or NOx
compliance.
 WWW
Xcel Energy
Cherokee
1&2
PC Clients WWW
User input data
GUI

Performance calculations
Linux WWW User input data
Redundant Application Web server
Distributed Link
Server Interface generator
DCS Link
Relational Database Distributed Link
Customer Data Storage
Distributed Link
Client Visualization
Solution

intranet Fuel Policy Solution

Lookout Center
Cherokee
3&4
Performance calculations
WWW User input data
Performance calculations Distributed Link
User input data WWW DCS Link
Distributed Link Performance calculations
DCS Link User input data WWW
Distributed Link
DCS Link
Arapaho
Units 3 & 4 Valmont Unit 5
Combustion Optimizer
z Results
The Combustion
– Reduces NOx and CO
Optimizer reduces emission levels
NOx emissions boiler – Improves heat rate up to
efficiency while 1.5%
improving boiler – Reduces plant maintenance
efficiency, and costs
maintaining loss on – Maximizes staged
ignition. combustion efficiency
– Controls and reduces
measured opacity levels

Increase the efficiency of your

boiler combustion process.
SmartProcess Boiler Optimizer
dynamically adjusts and coordinates
the combustion process
z Stoichiometry
– Primary air flow
– Secondary air flow
– Tertiary air flow
– Fuel flow bias
z Elevation Windbox DP
z Excess oxygen
z Over fire air flow or
damper position bias
z Forced draft fan bias
z Induced draft fan bias
 Typical
SmartProcess
Architecture
Combustion Optimizer tools:SmartEngine
Web based interface
Sensitivity Analysis User interface
Case Study
Plant Optimization – Ameren Newton Unit #2
z 20.1% Average NOx Improvements
Company: Ameren Energy z 1% HR Improvements at high loads
Generating

Site: Newton Station z Issues driving the need for change

– 2003 emissions mandate to maintain NOx
below 0.13 #/mmBTU
– Avoid installing SCR
– Sell/Trade NOx credits
Unit: Unit #2
z 4 month project cycle
Location: Newton, IL, USA
MW:
Boiler:
615
Alstom (CE)
z No outage required
Turbine: GE
Primary Fuel: Coal z Payback of 9 months on NOx improvements
Application: NOx
8 months on (>$400K) from Heat rate
Optimization
Steam Temp
improvement
Optimization
Ameren Newton Unit #2 NOx Optimization
Results Overview – NOx Mode (100%)
NOx reduction

80

75

70

65

NOx_on
NOx [ppm]

60
NOx_off
55

50

45 Change in Heat Rate

40
10100.00
200 300 400 500 600 700

Load [MW] 10000.00

9900.00

NOx Reduction

Heat Rate [Btu/Kwh]

9800.00

Heat rate_on
9700.00
Heat rate_of f

9600.00

C O e m is s io n _IV Y_ON
9500.00

500
9400.00

450
9300.00
200 300 400 500 600 700
400
Load [MW]

350

300

Maintain Heat rate

CO [ppm]

CO _ a vg
250
CO _ ch a r
200

150

100

50

0
200 300 400 500 600 700

Load [Mw]

Below Average CO Levels

Ameren Newton Unit #2 NOx Optimization
Results Overview – HR Mode (100%)
Net Unit Heat Rate

10400 10328
10213
Heat Rate [BTU/kWh]

10200 10139 10157

10000 IVY ON

9800 9723 IVY OFF

9619
9600

9400
250 450 600
unit load [MW]

Net Unit Heat Rate improvement

1.2 1.07 1.11

Ave. of 1% improvement improvement [%] 1.0
in Heat Rate over the 0.8
typical Load range of 450 0.6
to 600 MW equals 0.4
0.17
0.2
$407,000 0.0
250 450 600
unit load [MW]
Case Study #44
Plant Optimization – Dynegy Hennepin #2
z 13% Average NOx Improvements
Company: Dynegy z Issues driving the need for change
Site: Hennepin – Drive plant average below .13 #/mmBTU
– Prior solution ineffective
z Real-time optimization of NOx emissions
and heat rate optimization
z 4 month project cycle
Unit: Unit #2 z No outage required
Location: Hennepin, IL
MW: 250
Boiler: Alstom (CE) z Head to head comparison against “other “
Turbine: GE
Primary Fuel: Coal competitive solution
Application: NOx
Optimization
Summary
Normal plant operation NOx reduction

18.00
16.00

NOx reduction in %
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
Data Set 1 Data Set 2 Average

SmartEngine ON Baseline - SmartEngine OFF

Load > 190MW Load > 190MW
January 25th, 2004 to March 3rd, 2004 June 1st, 2003 to August 16th, 2003

NOX NOX
0.112 0.141
Change in %
NOX
-21.17%
Steam Temperature Optimizer
z Results
The Steam Temperature
Optimizer provides
precise responses to
disturbances for accurate
temperature control.
Improve steam temperature for
faster ramp rates.
– Improves ramp rates up to
1% of MCR per minute
– Minimizes temperature
variations by up to 75%
– Controls spray valves, tilts,
pass dampers, for accurate
temperature
– Multivariable control
strategy to maintain
optimum steam
temperature
Case Study: Neal North
Steam Temperature Optimization
Application
z 515 MW, coal-fired, drum boiler, GE turbines,
WDPF control system
Challenge
z Excessive steam temperature oscillations limited
unit response to about 1% per minute at low loads
and 0.3% per minute at high loads.
z Stress on boiler pressure parts caused operational
falters
z Temperature variations and high steam pressure
frequently proved harmful to boiler and turbine
components.
Results
z Maintained steam temperature oscillations to within
5ºF while lowering overall main steam temperature
deviations by almost 40% (compared to the data
from testing with the new tuning parameters).
 Radiant SH Outlet Response to Sootblowing

„ Steam temperature variation

decrease from +/40 deg F to
+/-5 deg F across normal load range
and +/-10 deg F across full load
Model
range
Signal

Advanced Control
Prevents Overshoot!

Spray
Valve
 Auto STO Auto
Results!! 25 MW ramp 90 MW ramp
-35 To +25 -6 To +9
Sootblower Optimizer
z Results
The Sootblower Optimizer
uses an intelligent
modeling tool to develop
heat rate absorption
models that accurately
reflect the numerous
interrelationships of
various heat transfer
sections.
Ensure efficient sootblowing.
– Delivers optimal cleanliness,
resulting in a 0.5% heat rate
improvement
– Decreases soot accumulation
– Improves overall boiler
efficiency
– Balances blowing sequences
– Minimizes unnecessary steam
usage
– Reduces opacity spikes
– Reduces NOx formations
– Enhances steam temperature
variability
Sootblower Control Illustration

Scheduler

Master Sequencer

Blower
SmartEngine Sootblower
Case Study: Sootblower Optimization
Southern California Edison - Mohave Unit #1
z Heat transfer rate increases
Company: Southern – 8-10 % water wall and div superheaters
California
Edison
– 6-7% final superheater and front reheat
Site: Mohave Station
– 2-4% rear reheat and economizer
z Opacity reduction
z Issues driving the need for change:
– Reduced opacity spikes during sootblowing
and load ramps
Unit: Unit #1
Location:
MW:
Laughlin, NV
800
z Real-time sootblower optimization
Boiler: Alstom (CE)
Turbine: GE z 5 month project cycle
Primary Fuel: Coal

Application: Steam
z No outage required
Temperature
Optimization z Estimated payback of 8 months
Global Performance Advisor
z Results
The Global Performance
Advisor allows operators
to identify controllable
losses, track equipment
performance against
design specifications, and
quickly identify
problematic process areas
to reduce operating costs.
Monitor and benchmark plant
performance.
– Reduces operating costs by
tracking unit heat rate
penalty costs over time and
indicating dollars lost due to
equipment performance
deviations
– Calculates net unit heat rate
and tracks heat rate
deviations
– Displays deviations and
cost of deviations to help
operators determine
corrective action or flag
equipment repair and
maintenance needs.
 Global Performance Advisor
z Unit Heat Rate Module
z Turbine Heat Rate Module
z Turbogenerator Heat Balance
z Condenser Performance Module
z Boiler Performance Module
z Economizer Performance Module
z Boiler Feedwater Feedheater
Train
z Boiler Feedpump Turbine Module
z Fan Efficiency Module
z Large Pump Performance Module
z Cooling Tower Module
Condenser Design Data Screen
SmartProcess Design Advantages
z SmartProcess modules dynamically optimize the process throughout
the full plant operating range

SmartProcess Optimizes Over Full Load Range

MW

Competitive Steady State

Optimization Packages Optimize
Only During Fixed Load

Time
Typical SmartProcess Project Steps

Site
Site Assessment
Assessment Model
Model Design
Design &
& Validation
Validation

Project
Project Start
Start Installation
Installation of
of Model
Model

Development
Development ofof Plant
Plant Test
Test Plan
Plan Advisory
Advisory Mode
Mode Operation
Operation
and
and Design
Design of
of control
control modifications
modifications && Operator
Operator Training
Training

Execution
Execution of
of Plant
Plant Test
Test Plan
Plan Closed-Loop
Closed-Loop Mode
Mode Operation
Operation

Final
Final Training
Training &
&
Data
Data Collection
Collection &
& Conditioning
Conditioning Performance
Performance Assessment
Assessment
 Systems Interconnect Capabilities
OPC Serial Link
(RS232/SCSI) Existing
SmartProcess
ABB/Bailey
PC OPC CIU Infi90/Net90
Network, Consoles
and PCU’s

CIU Exception
z RS232 Throughput @ 9600b, @19.2kb Device Points Reports Control System
NSPM01 500 No No
Scheme 1 CIU* 1 Port - ~250pps, 500pps IMSPM01 500 No No
Scheme 2 CIU 2 Ports - ~500pps, 750pps IMCPM02 500 No No
IMCPM03 500 Yes Yes
Scheme 3 shown above similar to Scheme NCIC01 500 Yes Yes
1 or 2 capabilities NCIU01 500 Yes Yes
NCIU02 2,500 Yes Yes N90
Note * Support only a single connection
NCIU03 10,000 Yes Yes N90
NCIU04 10,000 Yes Yes I90
INPCI01 500 Yes Yes
INPCI02 10,000 Yes Yes I90/S
INICI01 10,000 Yes Yes I90/S
INICI12* 10,000 Yes Yes I90/S
INICI03* 30,000 Yes Yes I90/S
Bailey CIU Components
Bailey MFP to I/O
Bailey CIU
Termination Unit
Termination Unit
(no I/O attached)

Printer port

Network
Serial port to Interface Slave
OPC Server (NIS)
Computer
Transfer Module
(ICT)
Power Plant Smart Process Solution

OPC CIU EWS

Serial
OPC Client /
Smart Process OPC Server Interface
Bailey DCS
Software
Bailey Data Highway
Station
Advanced Control

Set point

disturbance

error Output
Controller boiler

Model (prediction) –
what will be the future

Optimization –
choosing best control
strategy

feedback
 Advanced control Model (Prediction))
Optmization

Modification of set points

Base controllers

Set points

1 2 3 4
SmartProcess Installations - Number
80

king for
70
ns wor
st a lla tio
c e s s In ob a lly
60
r t Pro MW’s g l
Sm a 0 0
2 5,0
50
over
40 80

30 53

20 33
24
10
15
5
1
0
1997 1998 1999 2000 2001 2002 2003
Major Utilities Choose
Emerson SmartProcess
z Considered a competitive advantage
z SmartProcess is latest technology
z Won competitive bids
Florida Power & Light
z Proven sustainable benefits

Wisconsin
Northern States Electric
Power
Summary
z Comprehensive package for complete and precise optmization of
Power Plant Process.
z Total solution from initial site assessment to final commissioning
z Each SmartProcess module is fully engineered, validated and
configured to meet specific generating unit requirements
z Platform independence allows SmartProcess products to be
implemented on WDPF, Ovation or any other distributed control
system
z Plant model incorporates automatic, adaptive error detection and
correction
z Proven Technology with large global installations