OVERVIEW

OF

Performance Analysis
Diagnostics & Optimization
( PADO )
How to Improve Efficiency ?

Performance Analysis, Diagnostic and Optimisation (PADO)

First Step: low hanging fruits

 Improve efficiency by improving operation

and maintenance

Next Steps:
 Audit or „map“ the plant using off line
modelling tool e.g. Ebsilon

 Online optimization tool PADO

 Online fault detection system using

SPC and Fault trees
 Online life time monitoring SR1 for better
planning of inspections and maintenance
Why to use the PADO tools ?

DCS also has the information needed to do the calculation

of efficiencies & heat rates

But what are the limitations ?

1. DCS does not give the system wise efficiencies so you do not
know where the losses occur

2. Data from I/O points e.g. Temperature, pressure mass flow could
be wrong because of sensor errors, bad connectors etc. That
makes calculation erroneous.

3. DCS does not give advice on what to do

PADO: Online Optimization Sequence

1. All relevant data from DCS which goes into calculation

need to be validated i.e. all implausible values have to
be replaced by plausible values

2. All calculations must be done every 5 minutes so as to

continuously monitor component and heating surface
efficiencies.

3. Results should be presented in user friendly manner:

state of comonents indicated by green, yellow, red and
losses expressed in monetary value per hour
PADO Functions

OFFLINE SRv
Data Validation
WhatIf

SRx (Datamanagement)

SR4
Diagnosis +
SRx
Optimization
Visualisation

DCS
PADO in India

80 units order for PADO have been placed on SESI

27 units successfully commissioned till date

National Thermal Power Corporation (NTPC)

The largest power generating major of India generating power from Coal
and Gas with an installed capacity of 34,194 MW, has standardised on
Steag PADO for all future units including Super-Critical Units.

Bharat Heavy Electricals Limited (BHEL)

The largest supplier of power equipment with 70% of current installed
market of Thermal Power Plants has a Framework Agreement with
SESI for installation of PADO for all future units
including Super-Critical Units.
PADO in India (2)

55 units where PADO is commissioned or under commissioning

 NTPC Simhadri 2x500 MW  RVUNL Stage 1 and 2 Chhabra 3 x 250 MW

 NTPC Ramagundam 1x500 MW  Shree cement Ltd. RAS
 NTPC Rihand 2x500 MW  DVC Maithan 2x500 MW
 NTPC Talcher 4x500 MW  GEB Ukai 2x500 MW
 NTPC Kahalgaon 3x500 MW  NTPC Korba Extn 1x500 MW
 NTPC Sipat 2x500 MW  NTPC Bongaigaon 3x250 MW
 NTPC Vindhyachal 2x500 MW  TNEB North Chennai 2x600 MW
 NTPC Korba 1x500 MW  CSEB Marwa 2x500 MW
 NTPC Dadri 2x500 MW  CSEB Korba 1x500 MW
 NTPC Farakka 1x500 MW  L&T Rajpura 2x700 MW
 Mahagenco Khaparkheda 1x500 MW  L&T Koradi 3X660 MW
 Mahagenco Bhusawal 2x500 MW  Sterlite Jharsuguda 4X600 MW
 NTPC Simhadri (stage II) 2x500 MW
 NTPC Jhajjar 3x500 MW
 KPCL Bellary (KPCL) 1x500 MW
Signing of Framework Agreement
Advantages of PADO

● Improving the quality of measurements by

data validation
● Evaluation of boiler, turbines, condenser and other
components

● Optimization of unit operation (sootblowing, setpoints)

● Calculation of what-if scenarios
● Generation of daily and monthly reports

● Enhance the efficiency of the power plant !

Modules of PADO System

Boiler Fault
Setpoint Tree
Soot Statistical
Optimizer
Blowing Process
Optimizer Control

Data Ebsilon
Data Data
Management
Visualizer Validation Model
System

Lifetime Performance
Base Modules Monitoring Monitoring
Fault Detection Metal
Optimization Temperature What-If
Physical Condition Analysis
Performance Analysis
SR::x Data Management System

SR::x is the central data management in the

SR product family
– Competitive server featuring a “state-of-the-art“
visualization
– Long-term storage of measured and computed
values in time-oriented archives;
base time class is ‘minute values’
– Automatic aggregation to higher time classes
such
as 5‘-, quarterly-, hourly-, daily-, monthly- or yearly-
values
– Integrated mathematical formula editor
– Excel-Add-In and HTML-List Generator allow
the generation of extensive reporting systems
SR::x Data Management System

SR::x is the central data management with “state-of-the-art”

visualization
Data Validation System

Data Validation to replace data errors due

to defective sensor or cable problems:
Incorrect Data – Wrong results
No data – No results

Validation – A 3 tier Process

1) Plausibility Check using Neural
Networks
2) Plausibility Check based on range
of Values
3) Data validation / reconciliation
based on “First Principle
Thermodynamic” model
Data Validation System (1)

based on Neural Network…

Data Validation System (2)

based on “First Principle Thermodynamics”…

Data Validation System (3)

check the quality of measurements …

Implausible Value - detection

main steam measurements

left and right
(left not plausible)
Implausible Value - replacement

Neural network generated the value 477 °C for not

available value of tag 70HAH30CT129_XQ03
Data Validation Report
Visualization Aid 1 :
Traffic light coding system

Red - Critical
Yellow - Suboptimal
Light Green – Optimal
Dark Green – Time gap to next
action
Visualization Aid 2 : Data values as tags &
in tables with coded Background colors
Interpretation of Background Colors

White – Measured Value

Grey - Calculated Value

Violet – Replaced value (Originally non

plausible or not available)
 Performance Monitoring

 Compares the actual values of critical parameters with the best

achievable under current operating conditions.
 Shows monetary loss against each sub optimal operating
parameter, defining the scope of improvement.
Turbine Performance Monitoring
Boiler Performance Monitoring
Boiler Mapping
Soot Blowing Optimization

 Fouling and Slagging of boiler heating surfaces cause…

- Loss of efficiency due to increased flue gas temperature.
- Loss of efficiency due to increased Reheater spray flow.
- Shutdowns for removing persistent slag.

 Use of soot blowers…

- Causes costs (blowing medium, wear-and-tear...).
- Impact on control actions.
- Erosion of the heating surfaces.
- May be prohibited or not reasonable in some operating states.
Soot Blowing Optimization

SR::EPOS::BCM is the SR product for Platzhalter Bild

optimizing
the soot blowing
● Controlled by costs or other criteria the
optimum points in time for activating the individual
blower levels are calculated
● Closed-Loop application possible if desired
● Application of fuzzy technology
 Criteria for Soot Blow Optimization

 Fouling/Slagging: Results from thermodynamic model - SR::EPOS

 Time intervals
- Minimum frequency of soot blowing
- Minimum-pauses

 Configurable priorities for each criterion

 Soot-blowing costs

 Process-engineering criteria, plant-specific

- Reheat spray flow
- Flue gas temperature before air preheater
- Furnace exit gas temperature
- RH Metal Temperature
- Mills Combination
- Coal mass flow
- NOx-emissions
Soot Blowing Optimization
Boiler Setpoint Optimization

Shows the optimal against current setpoints and improvement in heat rate.
Optimized Heat Rate

main steam flow = 450.33 kg/s

furnace exit temperature = 1242.7 °C
2340
mill influence 1 = 33.38
2330

2320

2310

2300

2290
20
15 0.05
10 0.045
0.04
5 0.035
0.03
0 0.025
In this case the optimum means to move
 the burner tilt down from the current 9 degree to 4.5 degree and
 the excess O2 from the current 4% to 3.55%.
Metal temperature Module

Shows the temperature profile of individual tubes of various

heating surfaces of boiler and identifies the Hot spots.
 Lifetime monitoring module

module aims to calculate the

remaining life of thick walled
components in boiler

the consumed life of an

equipment could be different
from the actual age of the
equipment
Lifetime monitoring module

Temperature

Free Capacity for Creep

Operating Point Scheduled

Plant Life

Pressure
Lifetime monitoring module
depends upon how stressful the life of equipment has been so far in terms
of temperature and pressure which effect fatigue and creep.

Monitoring Creep & Fatigue for Drum 1/top

Statistical Process Control

Statistical methods to evaluate partly automatic, early &

reliable detection of changes where deterioration is slow.

Air ingress

false alarms
07.09.2006
plausible alarm air 07.09.2006
ingress plausible alarm air ingress
 Key Performance Indicators
Key Performance Indicators (KPIs) can be used to detect component failures and to
measured
enable condition basedcalculated
maintainance. KPI should depend only on component condition
but not on load, ambient conditions or operation mode
vibration heatrate
bearing temperature Component quality
factor
oil temperature
efficiency
power consumption
…
…

Key measurements in power

plants usually depend on protection limit
- load,
- operation mode
- fuel quality
- ambient conditions
- etc.
And are superposed by noise
key performance indicators by Act /
Ref-comparison (KPI)

KPIs measure the qualitiy of the

process / component.

They do not depend on external

disturbance variables

KPI = act-value / ref-value

protection
limit
“KPI“-limit
normalized
parameter
=
KPI
SPC – Alarm Control Centre
 Fault Trees

• Models and analyzes faults in the

process.
• Composed of logic diagrams that
display the state of the system
and the states of the components
• Constructed using Drag & Drop
technique
• Does not need programming
expertise for building such
trees.
 Thank
You

... Ideas & Solutions for Tomorrow