Performance Analysis of Twin-Spool Water Injected Gas Turbines using Adaptive Modeling

Transcription

1 Performance Analysis of Twin-Spool Water Injected Gas Turbine using Adaptive Modeling I. Roumeliotis K. Mathioudakis N. Aretakis Laboratory of Thermal Turbomachines National Technical University of Athens 1

2 Performance Analysis of Twin-Spool Water injected Gas Turbines using Adaptive Modeling Constitution of a Performance Model Elements of Adaptive Modeling Implementation to an Engine Study of Gas Turbine Behavior Components Behavior Water Injection Effect on Engine Diagnostics Conclusions 2

3 Constitution of a Performance Model W.I.1 W.I.4 W.I.5 W.I.2 Desire to represent the performance of an individual engine as accurately as possible in all situations 3

4 Performance Analysis of Twin-Spool Water injected Gas Turbines using Adaptive Modeling Constitution of a Performance Model Elements of Adaptive Modeling Implementation to an Engine Study of Gas Turbine Behavior Components Behavior Water Injection Effect on Engine Diagnostics Conclusions 4

5 Elements of Adaptive models Actual Engine Measured variables / Available Data Do data match? Engine Model Predicted variables Adjust model Parameters through MFs 5

6 The meaning of Modification factors f k x p, k = x x p, k p, ref, k : Actual value for parameter Pressure Ratio x p, ref, k : Reference value for parameter Corrected Mass Flow Transformation of component performance maps 6

7 Performance Analysis of Twin-Spool Water injected Gas Turbines using Adaptive Modeling Constitution of a Performance Model Elements of Adaptive Modeling Implementation to an Engine Study of Gas Turbine Behavior Components Behavior Water Injection Effect on Engine Diagnostics Conclusions 7

8 Implementation to a selected Engine 5 stage LPC 14 stage HPC with 6 V.G.S. 2 stage air cooled HPT 5 stage LPT 8

9 Implementation to a selected Engine Engine Data Data for the Nominal Operating Conditions Thermodynamic Analysis Create Components Initial Characteristics Adaptation Initial Adaptation to Nominal Operating Conditions Acquire Operating Data Points Determination of MFs according to Available Data Determination of new Components Characteristics 9

10 Implementation to a selected Engine 9000 Present Model S ha ft P ow e r (M W ) Load Schedule Ambient Temp. o C Shaft heat Rate (Btu/kWhr) Dry data Steam Inj. data Water Inj. Data Ambient Temp. o C 10

11 Performance Analysis of Twin-Spool Water injected Gas Turbines using Adaptive Modeling Constitution of a Performance Model Elements of Adaptive Modeling Implementation to an Engine Study of Gas Turbine Behavior Components Behavior Water Injection Effect on Engine Diagnostics Conclusions 11

12 Performance Prediction Control Parameter Low Pressure Turbine Inlet Temperature Compressor Discharge Temperature (T amb >15 o C) Test Cases Evaporative Cooling via Water Injection Intercooling via Water Injection 12

13 Evaporative Cooling L o a d /(L o a d )re f CDT constant LPTIT constant CDT constant & Ev. Cooling LPTIT constant & Ev. Cooling Ambient Temperature ( o C) Operation with constant CDT results in a Slight Higher Gain in Load 13

14 Evaporative Cooling H R /(H R )re f CDT constant LPTIT constant CDT constant & Ev. Cooling LPTIT constant & Ev. Cooling Ambient Temperature ( o C) Operation with constant CDT results in a Slight Higher Gain in Thermal Efficiency 14

15 Intercooling L o a d /(L o a d )re f water injection CDT constant LPTIT constant CDT constant & intercooling LPTIT constant & intercooling Ambient Temperature ( o C) For High Ambient Temperature operation with constant CDT results in an Important Load Gain 15

16 Intercooling 1.1 H R /(H R )re f water injection CDT constant LPTIT constant CDT constant & intercooling LPTIT constant & intercooling Ambient Temperature ( o C) For High Ambient Temperature operation with constant CDT results in Higher Gain 16

17 Performance Analysis of Twin-Spool Water injected Gas Turbines using Adaptive Modeling Constitution of a Performance Model Elements of Adaptive Modeling Implementation to an Engine Study of Gas Turbine Behavior Components Behavior Water Injection Effect on Engine Diagnostics Conclusions 17

18 Component Behavior 1.20 Low Pressure Compressor (πc-1)/(πc-1)ref Dry Chilling Intercooling Steam Injection Fraction of Design Air Flow Tamb = 15 o C, Load varies from 30 to 50 MW 18

19 Component Behavior (πc-1)/(πc-1)ref High Pressure Compressor Dry Chilling Intercooling Steam Injection Increasing Load Fraction of Design Air Flow Tamb = 15 o C, Load varies from 30 to 50 MW 19

20 Component Behavior (πc-1)/(πc-1)ref Low Pressure Compressor Dry Chilling Intercooling Steam Injection Decreasing Tamb Intercooling in Use (10 o C) Fraction of Design Air Flow Load = 45MW, T amb varies from -10 to 50 o C 20

21 Component Behavior (πc-1)/(πc-1)ref High Pressure Compressor Dry Chilling Increasing Tamb Intercooling Steam Injection Intercooling in Use Fraction of Design Air Flow Load = 45MW, T amb varies from -10 to 50 o C 21

22 Component Behavior Additional Information obtained for operation with Water Injection Original Adapted π q 22

23 Performance Analysis of Twin-Spool Water injected Gas Turbines using Adaptive Modeling Constitution of a Performance Model Elements of Adaptive Modeling Implementation to an Engine Study of Gas Turbine Behavior Components Behavior Water Injection Effect on Engine Diagnostics Conclusions 23

24 Engine Diagnostics f 1 ~W f 2 ~n p f 3 ~W f 4 ~n p f 5 ~Δp b f 6 ~n b f 7 ~W f 8 ~n p f 9 ~W f 10 ~n p Pictorial view of MF on engine components 24

25 Engine Diagnostics 0.5 Measurement Deviation Intercooling with Water Injection, Ww/Winletair = 0.02 D e viatio n (% ) Load = ct Tamb = 15 o C RHamb = 30% LPTIP CDP N25 CDT LPTIT EGT Monitoring Techniques based on Direct Observation of Measured Quantities may lead to erroneous conclusion if Water Injection is not taken into account 25

26 6 Engine Diagnostics Health Indices Deviation Intercooling with Water Injection (Ww/Winletair=0.02) f2 f3 f4 f7 f8 f9 f10 Monitoring Techniques based on Health Parameters may lead to erroneous conclusion if Water Injection is not taken into account 26

27 Conclusions Using appropriate procedure it is possible to estimate the engine performance even if limited information is available The structure of the model gives the ability of modeling water injection at various station and using different parameters as control variables. Data from water Injection gives additional information for determination of component characteristics Water Injection must be modeled for accurate diagnosis 27