industrial gas turbine health and …...gt2016-57722 industrial gas turbine health and performance...

GT2016-57722 Industrial Gas Turbine Health and Performance Assessment with Field Data 1

INDUSTRIAL GAS TURBINE HEALTH and PERFROMANCE ASSESSMENT

with FIELD DATA


Contents Introduction

• Scope of the paper • Test Case description

Performance Models • The PROOSIS Platform • GT Model Adaptation Process • Integrated Models

GTs Health Assessment • Performance Shift • Performance Deterioration

Economic Assessment • Washing Routine Assessment • Washing Routine Optimization

Summary-Conclusions







Summary-Conclusions


Advanced Diagnostic Methods application in Combined Cycle Gas Turbine Power Plants (CCGT) is sparse

CCGT operating mode was predominantly base load. Now days it shifts to load following operation, thus making methods capable for part load health assessment necessary

Health and Performance Monitoring can help in controlling Recoverable Deterioration such as Fouling, benefiting plant profitability

This work presents elements of an on-line integrated system developed for health and performance assessment of a Gas Turbine CHP power plant

For demonstrating the information that can be obtained, results derived by a procedure where a GPA (adaptive modeling) method is coupled with a water/steam cycle model and an economic module are presented

The results are used for the health/performance of the plant GTs and for the power plant performance and economic assessment by the operator

Scope of the paper


Test Case Description

Usually operates in 1x1x1 configuration

Process Steam Prioritized

GTs: IHB for DLN combustor

GTs: Chillers operation


Health and Performance Assessment

A s/w system developed by LTT/NTUA has been installed on-site for the on-line power plant performance and GTs health assessment

Several methods developed and validated by LTT/NTUA are integrated to the system as modules

DATA

ACQUISITION

DECISION

MAKING

DATA

PROCESSING

Assessment

Data/Report

GT ADAPTIVE & PERFORMANCE MODEL

Data Analysis

Signal processing

Heat balance,

Corrected data

EGT reduced dev.

Vibration Monitoring

Data Cleaning

Noise filtering &

Sensor validation

(range, PNN)

Data Acquisition

Monitoring data from

Historian Data Base

CCGT

HRSG & ST MODELS

GT Adaptive Model

EGT Monitoring

HRSG/ST Monitoring

Techno Economic


DATA

ACQUISITION

DECISION

MAKING

DATA

PROCESSING

Assessment

Data/Report

GT ADAPTIVE & PERFORMANCE MODEL

Data Analysis

Signal processing

Heat balance,

Corrected data

EGT reduced dev.

Vibration Monitoring

Data Cleaning

Noise filtering &

Sensor validation

(range, PNN)

Data Acquisition

Monitoring data from

Historian Data Base

CCGT

HRSG & ST MODELS

GT Adaptive Model

EGT Monitoring

HRSG/ST Monitoring

Techno Economic

Health and Performance Assessment

A s/w system developed by LTT/NTUA has been installed on-site for the on-line power plant performance and GTs health assessment

Several methods developed and validated by LTT/NTUA are integrated to the system as modules


Contents

Introduction • Scope of the paper • Test case description

Performance Models • The PROOSIS platform • GT Model adaptation process • Integrated Models



Summary-Conclusions


Engine Adaptive Model

The PROOSIS platform

Object-Oriented Steady State Transient Mixed-Fidelity Multi-Disciplinary Distributed Multi-point Design Off-Design Test Analysis Diagnostics Sensitivity Optimisation Deck Generation Connection with Excel & Matlab Integration of FORTRAN, C, C++


The PROOSIS platform TURBO library of gas turbine components

Industry-accepted performance modelling techniques Respects international standards in nomenclature, interface & Object Oriented programming

Compressor map Turbine map

SW2 SE2

SW4 SE4

Component Health

Parameter Symbol

Compressor Flow SW2 Efficiency SE2

Turbine Flow SW4 Efficiency SE4


Engine Adaptive Model

Create required engine configuration

(schematic) using TURBO library components

Design point sensitivity analysis &

optimisation relative to design point

performance

Base Load

performance data

after overhaul or

off-line washing Design point sensitivity analysis &

optimisation relative to off-design

performance

Map scalars, secondary air system flows,

duct pressure losses coeff. etc

Global Map scalars, Isolines re-labeling

Dedicated IGVs related factors

Off-Design optimization for accounting IGVs

operation

Part Load

performance data

after off-line washing

or/and corrected for

fouling

Manual local adaptation for

compressor & turbine

Dependency of

SW/SE on OP

and amb. cond


Engine Adaptive Model Predictions lie within ±0.5% for both Base and Part Load

No dependency on operating point and ambient conditions

0

0.2

0.4

0.6

0.8

1

SW2 SE2 SW4 SE4

RM

S (%

)

Full Open (ΔIGVs=0)

Part Load (ΔIGVs up to 30deg)

0

0.2

0.4

0.6

0.8

1

3000 3010 3020 3030

RM

S (%

)

Corrected Rotational Speed (rpm)

SW2 SE2 SW4 SE4

Interesting Notes

NG Fuel properties (LHV, RD) should be measured especially for the case of multiple NG entries to the national system

Threshold selected at historian (minimum deviation for changing the stored value) should be evaluated and refined if needed


Water/Steam Cycle Model The HRSG modeling is based on NTU-effectiveness method

The steam turbine is modeled utilizing Stodola law and efficiency correction for part load

HRSG and steam turbine models were calibrated versus operation data

Predictions lie within ±1% for both base and part load and for all available pressure & temperature measurements

-1

-0.6

-0.2

0.2

0.6

1

700

710

720

730

740

750

760

770

780

790

70 80 90 100 110 120 130 140

%D

iffe

ren

ce

HP

Hea

der

Tem

per

atu

re [

K]

GT Power [MW]

PredictedMeasured%diff


Integrated Procedure

Model at

Diagnostic Mode

Measurements

(WF, CDP, CDT, EGT)

Health factors

EGT

Model at

Simulation Mode

for healthy

condition

Operating Point

Conditions (CIT, CIP,

WAR, IGV, IBH, TF,

HHV, RD, EGP, Speed)

Powerref

HRref

Power (meas)

SW2,SE2,SW4,SE4

GT Performance

Deterioration

Water / Steam

Cycle Model

Plant Performance

Deterioration

Cost Analysis Fixed

Parameters

&

Number of Washings

Total Profit

Maximum or

Specific Energy

Cost

Minimum?

MODIFY

Washing Time

Intervals

Total Profit &

Specific Energy

Cost

NO

YESOptimization

Code

Washing Time

Intervals

ECONOMIC

MODULE

Performance

Deterioration

Rates

RESULTS

Optimum Washing

Time Intervals

MODIFY

Washes Number

Cost Analysis Fixed

Parameters

&

Number of Washings

Total Profit &

Specific Energy

Cost

Washing Time

Intervals

ECONOMIC

MODULE

Performance

Deterioration

Rates



• Scope of the paper • Test case description

Performance Models • The PROOSIS platform • GT Model adaptation process • Integrated Models



Summary-Conclusions


GTs Health Assessment

This data set was used by the operator as blind test for the system

The operator applied sporadically off-line and on-line washes

Base load operation is dominant but part load operation periods exist

A distinct non-recoverable performance shift occurred at the end of the period observed from full open IGVs data (ΔIGVs=0)

Relative Corrected Performance (~ 14 months)

-40

0

40

80

120

160

200

240

280

0.5

0.6

0.7

0.8

0.9

1

1.1

1/12/12 30/1/13 31/3/13 30/5/13 29/7/13 27/9/13 26/11/13 25/1/14

Re

lati

ve C

orr

ect

ed

Po

we

r O

utp

ut

(-)

Daily on-line washing

ΔIGVs angle

1 2 3 4 5 6

-40

0

40

80

120

160

200

240

280

0.8

0.9

1

1.1

1.2

1.3

1/12/12 30/1/13 31/3/13 30/5/13 29/7/13 27/9/13 26/11/13 25/1/14

Re

lati

ve C

orr

ect

ed

He

at R

ate

(-)

ΔIGVs angle


GTs Health Assessment-Shift

GT Health Indices

-5

-4

-3

-2

-1

0

1

2

3

4

5

SW2

(%)


1 2 3 4 5 6

-5

-4

-3

-2

-1

0

1

2

3

4

5

1/12/12 30/1/13 31/3/13 30/5/13 29/7/13 27/9/13 26/11/13 25/1/14

SE2

(%)

-5

-4

-3

-2

-1

0

1

2

3

4

5

SW4

(%)


1 2 3 4 5 6

-5

-4

-3

-2

-1

0

1

2

3

4

5

1/12/12 30/1/13 31/3/13 30/5/13 29/7/13 27/9/13 26/11/13 25/1/14

SE4

(%)

Compressor: Recoverable performance deterioration & marginal change (SW=-1%)

Turbine: Significant reduce of effective throat area and increase of efficiency

Verdict: Permanent positive change on Hot Section: overhaul or engine up-rate


GTs Health Assessment-Shift EG

T (o

C)

PR(-)

12/2012 to 10/2013

1/2014

5deg

-40

-35

-30

-25

-20

-15

-10

-5

0

5

ΔIG

Vs

(de

g)

Qexh (MWth)

12/2012 to 10/2013

1/2014

2 deg

Basic Control Parameters

Cross-checked via heat balance calculations and measurements analysis

No significant change on basic control logic (e.g. PR vs EGT & Qexh vs ΔIGVs)

Operating line change consistent with throat area decrease (PR increase)

The Diagnostic Verdict was confirmed by the operator (engine up-rating)


PR

(-)

W2,cor (kg/s)

12/2012 to 10/2013

1/2014

ΔIGVs=+2οΔIGVs=0ο

Cross-checked via Heat Balance Results and Measurements Analysis

No significant change on basic control logic (PR vs EGT & Qexh vs ΔIGVs)

Operating line change consistent with throat area decrease (PR increase)

The Diagnostic Verdict was confirmed by the operator (engine up-rating)

GTs Health Assessment-Shift

W2,cor CIT=W41,cor×

CPR Tt41

T(o

C)

Pgross, cor (MW)

12/2012 to 10/2013

1/2014

COT, cor

TITREF

20 deg

Operating Parameters


GTs Health Assessment-Fouling

Fouling Monitoring

The traditional fouling indicators: power output and compressor efficiency are consistent for base load but the provided information quality significantly degrades for part load operation

The diagnostic model calculated performance deterioration is consistent throughout engine operation (base & part load)

0.8

0.85

0.9

0.95

1

1.05

1.1

1.15

8/7/13 15/8/13 23/9/13

Re

lati

ve T

ITER

F

Re

lati

ve C

orr

ect

ed

Po

we

r O

utp

ut

(-)

Correction Curves

Diagnostic Model

1TITREF

4 5



0.9

0.95

1

1.05

1.1

1.15

8/7/13 15/8/13 23/9/13

Re

lati

ve T

ITER

F

Re

lati

ve C

om

pre

sso

r Ef

fici

en

cy (

-)

Heat Balance

Diagnostic Model

1TITREF

4 5Daily on-line washing


GTs Health Assessment-Fouling

Fouling Monitoring

The application of adaptive modeling gives clear indication throughout operation for:

Turbomachinery components performance deterioration

Washing routines effectiveness

Overall engine performance deterioration

0.9

0.92

0.94

0.96

0.98

1

1.02

1.04

3/3/13 2/4/13 2/5/13

Re

lati

ve TΙT

RΕF

Re

lati

veC

orr

ect

ed

Po

we

r O

utp

ut

(-)

2 3

1TΙTRΕF

-3

-2

-1

0

1

2

3

SW2

(%

)

2 3

-3

-2

-1

0

1

2

3

3/3/13 2/4/13 2/5/13

SE2

(%

)


GTs Health Assessment-Fouling Fouling Monitoring

-1

0

1

2

3

4

5

Po

we

r D

ete

rio

rati

on

(%

)

2 3

-0.5

0

0.5

1

1.5

2

3/3/13 2/4/13 2/5/13

He

at R

ate

De

teri

ora

tio

n (

%)

The overall engine performance deterioration due to fouling is calculated and translated to information deemed meaningful by the user for supporting a decision making procedure and for increasing the usability of the diagnostic procedure







Summary-Conclusions


Economic Assessment

Steam price is assumed equal to the cost of the fuel needed for producing the steam in a 80% efficiency boiler

GTs operate in an interchangeable way during 2013 & 2014 (off-line wash after shut down)

The up-rated engine model is used for the economic assessment for both cases utilizing the corresponding compressor health factors (2013 & 2014)

Test Case Definition

Washing Action

2013 2014

Off-Line (GT1&GT2)

8 15

On-Line Sporadic

for 10 months Daily

Daily for Aug & Sept

Economic Data Used

Fuel Cost 8.7$/MMBTU Elc Market Price 87$/MWh Life Expectancy 30 years People /MW 0.2 Salary 1500$/month Maintenance Cost 6$/MWh Off-Line EOH 20 Off-Line Cost $3000


Economic Assessment

Application of on-line washes significantly decreases performance deterioration

The washing routine applied in 2014 (frequent off-line washes and daily on-line washes) enhanced power production (+1.6%) and heat rate (-0.8%)

The performance enhancement is translated to over 800k$ additional gain for the operator

Washing Routine Assessment

-3%

-2%

-1%

0%

1%

2%

3%

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

Dif

fere

nce

Power Production Fuel Consumption Heat Rate

-1

0

1

2

3

4

5

6

7

0 200 400 600 800

Po

we

r D

ete

rio

rati

on

[%

]

Hours

Aug 2013 (on-line)

Aug 2014 (on-line)GT change (2014)


Economic Assessment

Time intervals optimization gives a gain of more than 50k$ by increasing the off-line washes frequency during summer time

Higher deterioration rate during summer months for the specific plant

Washing Routine Optimization

Washing Action

Nr of Washings

Change (k$)

Applied 2014

15 Reference

Optimized 15 +57

Typical Equidistant

15 +38

0.0

0.5

1.0

1.5

2.0

2.5

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

De

teri

ora

tio

n R

ate

[%/m

on

th]

Power Heat Rate







Summary-Conclusions


Summary Data obtained over two years from an on-line health and performance

monitoring system installed to a CHP power plant has been analyzed utilizing an integral approach for diagnosis, performance and economic assessment built in close cooperation with the power plant engineers

Gas Turbine performance model has been developed using a semi-automated adaptation procedure and used for health and performance assessment. Coupled with a water/steam cycle performance model and an economic analysis module was used for assessing power plant performance and economics

The adaptive modeling method correctly identified turbine as the source of an engine performance shift

The adaptive modeling method was used for successfully monitoring fouling throughout engine operation (base and part load)

The calculated overall plant performance deterioration was used for assessing two washing strategies and for optimizing the currently applied


Conclusions

Engine performance deterioration can be identified and quantified by “back of the envelope calculations” such as heat balance and performance parameters correction utilizing strictly base load data

Adaptive modeling can effectively identify and quantify component and engine deterioration along with component faults for the whole operating envelope

This information can further be used to support a decision making procedure regarding maintenance planning when coupled with overall plant performance and economic assessment tools

The economic comparison of two washing routines applied by the power plant personnel indicated that on-line washing results to performance enhancement which is translated to a significant economic gain

It is significant to decide in cooperation with the operator’s personnel how to present the diagnostic results and what post-processing parameters are deemed informative for everyday operation


Questions

감사합니다 - THANK YOU

industrial gas turbine health and …...gt2016-57722 industrial gas turbine health and performance...

Documents