Application of Advanced Diagnostic Techniques for Plant Performance - PowerPoint PPT Presentation

Application of Advanced Diagnostic Techniques for Plant Performance and Availability Improvement Michael Fox Douglas Eakle Supervisor, Data Utilization Performance/CBM Engineer FirstEnergy Corp. FirstEnergy Corp. IDEA Center Harrison Power Station Ron Griebenow, P.E. Director, Energy Services GP Strategies Corporation USAID Clean Coal International Conference 21 & 22 November 2013

• Six million customers • 18,000+ MW capacity • Operations in six states • 65,000 square mile service territory • 20,000 miles of high- voltage transmission • $50 billion in assets • $15 billion annual revenue • 16,500 employees

• Performance Improvement Leader since 1966 • Servicing 1/3 of Fortune 500 • Energy, Manufacturing & Government Sectors • Headquartered in Columbia, Maryland • 3,000+ Employees in 13 countries • $400M US Revenues (2012) • NYSE: GPX “Integrating People, Processes & Technologies”

GLOBAL PRESENCE Our Office Locations: Where We Operate:

Energy Services Improving Plant & Workforce Performance WORKFORCE PERFORMANCE ASSET PERFORMANCE HUMAN PERFORMANCE MANAGEMENT “More than 150 Power Generation SME’s.”

Background • FirstEnergy has utilized on-line performance monitoring since the late 1980s • Allegheny Energy (now FirstEnergy) implemented APR for validation of on-line data in 1998 – Expanded to on-line equipment health monitoring in 2004 • Performance And Condition Monitoring included multiple, independent systems – Each required maintenance, training and application expertise – diagnosis of a specific problem often required the subject matter expert (SME) to switch back and forth between the systems • Standardized on EtaPRO Suite in 2011 • Established Information Diagnostic Evaluation and Analysis (IDEA) Center 7

The EtaPRO System Synergistic Technologies Core Technologies 2010 2012 2004 ANOMALY MACHINERY THERMODYNAMIC DETECTION DYNAMICS MODELING

On-Line Performance Monitoring

Thermodynamic Modeling Design Data Engineering Principles VirtualPlant™ Model  Boiler  Steam Turbine  Feedwater Heaters  Condenser  Air Heaters  Steam Seal System  Boiler Feed Pumps 10

11

EtaPRO Predictor Machinery Health Monitoring Asset Sensors Fault Symptom Monitoring Signature Processing Unit Dynamic Sensors measured  Processes dynamic signals  Vibration into machine signatures  Air gap using advanced signal  Flux analysis techniques.  Current  Compares new data with  Phase reference pre-classified references. Semi Static Data imported  As supplement to the alarm  Send changes to the  Power, Active, Reactive system, the Plot Manager EtaPRO Predictor Server  Inlet pressure provides advanced diagnostic for warning, diagnosis,  Flow presentation and analysis tools prediction and storage.  Bearing and Winding Temps. etc. for fault symptom forecasting

EtaPRO Predictor AutoDiagnosis

Advanced Pattern Recognition Asset Sensors History APR Model  Power Output  Reactive Load  Exhaust Temp  IGV Position  Inlet DP  Bearing Temp  Bearing Vibration  Wheel Space Temps  Etc.

Alarm Range Measured Value Expected Value Difference between expected and measured Alert

Concern Management • Concern History • Multi-concern View • Asset/Hierarchy Views • Expected Values – APR – VirtualPlant – EtaPRO – Predictor

FirstEnergy Case Studies • All Examples from Harrison Power Station 3x650 MW Original Rated Capacity – Foster Wheeler Opposed-Wall, Coal-Fired, Supercritical Boilers – Westinghouse Single Reheat Turbines – Turbine Upgrades (~670 MW) – • Closely track issues identified, resolution and value, based on EPRI guidance Fleet Wide Monitoring for Equipment Condition Assessment (TR-1010266, – March 2006) On-Line Monitoring Cost-Benefit Guide (TR-1006777, November 2003) – • 2012 probability-weighted value of validated EtaPRO concerns almost $2.5 million total potential savings exceeded $24 million – first half of 2013 (January – June) = almost $500,000 –

FirstEnergy Case Studies 19

Feedwater Heater Isolation/Bypass • May 23, 2013, EtaPRO alerted plant staff to a high feedwater EtaPRO 10.1 heater level – Harrison Unit 2 – heater 24B • Heater level pots "bumped" during teardown of scaffold – caused the heater to isolate – bypass valves to open – no annunciator alarm was received in control room • Unit derate of 5 MW – would likely have remained isolated for one week – value of the lost generation = $29,400 20

Feedwater Heater Isolation/Bypass • Heater was returned to service12 hours after the initial alert EtaPRO 10.1 – actual value returned to normal 21

Forced Draft (FD) Fan IB Bearing • April 11, 2013, Unit 3 B-side FD fan inboard bearing temperature EtaPRO 10.1 increased measurably – Concern being monitored – Temperature ~25 o F higher than normal – DCS alarm point 180 o F • Boiler tube leak outage Week of July 2 – oil change – babbit material in oil – bearing rolled out – minor damage – Scraped, blued, and returned to service

Forced Draft (FD) Fan IB Bearing • Mechanics noticed sight glass indicating normal level after draining EtaPRO 10.1 – sight glass connection to housing plugged with dirt and sludge • Following repair, temperature dropped ~ 15 o F – Still above “normal” – no change with continued operation – model scheduled for tuning

Forced Draft (FD) Fan IB Bearing • Loss of the FD fan = unit derate of 300 MW EtaPRO 10.1 • Repair would have requires at least 48 hours • Would result in lost revenue of approximately $504,000 • Conservative 10% probability, results in probability-weighted savings of $50,400 • Excludes likely increased costs for the actual repair, had the bearing been run to DCS alarm level or possibly to failure

Gland Steam Temperature • Gland steam temperature not under automatic control EtaPRO 10.1 • Requires manual valve adjustment to control gland steam temperature – supply ~440-460 o F – temp feeding LP turbine seals Gland Steam to LP ~210-220 o F Gland Steam Temperature Instrument Desuperheating Station – often overlooked after a plant start-up Gland Steam Supply Temperature Instrument – scheduled for DCS integration – monitor in the meantime 25

Gland Steam Temperature • Tags added to existing LP steam turbine model on April 2, 2013 EtaPRO 10.1 26

Gland Steam Temperature • Alerted to high gland-steam temperature on June 3, 2013 EtaPRO 10.1 • Predicted value “noise” noted as modeling concern by GP Strategies Staff 27

Gland Steam Temperature • High temperatures lead to LP turbine vibration issues EtaPRO 10.1 28

Gland Steam Temperature • Vibrations did not hit alarm levels EtaPRO 10.1

Gland Steam Temperature • Model data review reveals “problem” data from 2005/2006

Gland Steam Temperature • Removal of “problem” data eliminates “noisy” expected value

Cycle Isolation • July 7, 2012, alerted to a high drain line temperature EtaPRO 10.1 • work order for replacement during next unit outage 32

Cycle Isolation • Valve was replaced during short outage starting late on July 13 EtaPRO 10.1 • Drain temperature excursion immediately following unit start- up July 18 – air solenoid leakage – insufficient pressure to close the valve – solenoid replacement resolved problem 33

Cycle Isolation • FirstEnergy estimated 2% of the reheat steam flow was being EtaPRO 10.1 dumped to the condenser – Based on the temperature, valve size and drain piping size and length • Thermodynamic model analysis (e.g., VirtualPlant) used to quantify the leakage effect on generation and heat rate • Resulted in increased generating costs of $38,985 per week 34

Air Heater Support Bearing • July 7, 2012, alerted to a high air-heater support bearing temp EtaPRO 10.1 • site glass indicated normal oil level • Added oil after further temp increase • Failure = unit trip – Repair – 72+ hours – loss in revenue > $4M million. • 10% probability of not being identified through other systems • Probability-weighted benefit of $436,440

Conclusion • FirstEnergy estimates system licensing and implementation costs EtaPRO 10.1 were recovered in less than 4 months • Single, integrated platform has resulted in increased visibility of issues through the availability of both performance data and equipment condition anomalies on a single platform and unlimited distribution rights of client software • Pilot installation in FirstEnergy nuclear operations (FENOC) indicates similar return on investment in nuclear stations. • FENOC expanding equipment models for Perry Station using in- house resources and will implement throughout the nuclear fleet (four units at three sites)

Questions? Ron Griebenow, P .E. Director, Energy Services 724 Whalers Way, Suite H100 Fort Collins, CO 80525 (970) 226-0812 rgriebenow@gpstrategies.com Knowledge. Performance. Impact.