Blade Maintenance Database for Improved Asset Management Strategies

Blade Maintenance Database for Improved Asset Management Strategies Houston, 11 November, 2015 Confidential and proprietary information of Wind. Com Page 1

About Wind. Com Group 2 nd Largest Blade Manufacturer Worldwide Blade Services Confidential and proprietary information of Wind. Com Page 2

About Wind. Com: Facts Over 15, 400 blades repaired around world Blade data management system with over Projects in 18, 000 detailed cases 23 countries addressing challenging customer needs Built a team of over 230 professionals Confidential and proprietary information of Wind. Com Page 3

About Wind. Com: Global Presence Serviced 320 sites in NA Head Offices Dispatch Centers Presence in Projects Confidential and proprietary information of Wind. Com Page 4

Data Flow – From field to office to customer Input Data Analysis Reporting Quoting How long should this take? Confidential and proprietary information of Wind. Com Page 5

• Common input sources and formats • HD ground inspections • Rope Access • Suspended Scaffold • Aerial Platform Input • New input sources • Drones inspections • 3 D Scanner • Under development • Methods which do not require stopping the turbine • Image recognition Offline vs Online reporting systems Confidential and proprietary information of Wind. Com Page 6

• Algorithm based analysis: • Bresenham's line algorithm • Colour Algorithms • Texture Algorithms Data Analysis • Can determine damage type and severity • Engineering time is reduced. • Processing of images may be completed in hours instead of days. • Damage type and severity standardization. • Damage progression monitoring Confidential and proprietary information of Wind. Com Page 7

Damage Classification Method 5: Major Structural 4: Non-Critical Structural 3: Major Cosmetic / Light Structural Cosmetic Structural 2: Minor Cosmetic 1: Light Cosmetic 10 9 8 7 6 5 4 3 2 1 0 Hours Identify Damage Repair Time-Damage Type LE Erosion Severity Level 0: No Issues One of 10 pre-established damage types One of 17 pre-established damage types 1 2 3 Severity 4 5 Recommendation A: No action required B: Can operate, monitor progress in future inspections C: Can operate, repair soon to avoid futher degradation (12 months) D: Can operate, repair at the earliest opportunity (3 months) E: Stop machine, repair as soon as possible Confidential and proprietary information of Wind. Com Page 8

Damage Classification Method • Patterns in damage concentrations are extracted from the database. This information is crucial for optimized maintenance plans • Damage incidences may be correlated with the turbines’ Power curve. Confidential and proprietary information of Wind. Com Page 9

Sample Report Summary Reports • Are generated in seconds • Can be organized according to customer preference: • Damage type Reporting • Damage severity • Location on blade • Budget • Others Our reports allow for fast review of the work performed, yet include all pertinent technical data which may be accessed with one click. Confidential and proprietary information of Wind. Com Page 10

As damages have been analysed and classified into pre programed categories, the bulk of the quotation process may also be done automatically. ADVANTAGES Quoting An efficient and transparent quotation method saves time and money Allows customer to clear view of the repair campaign they are facing System considers variables beyond the damages found; such as average stand by time per location and time of year, park accessibility, curing time of materials used, etc. Confidential and proprietary information of Wind. Com Page 11

Benefits Fast response time; essential for critical damages and keeping the turbine running Creation of secure and customer owned data base which may be exported or transferred upon request Predictive maintenance plan Disruptive blade failure pattern identification Damage progression motorization Reduce downtime, critical damage incidences and overall cost with a long term maintenance plan Confidential and proprietary information of Wind. Com Page 12

Total time for repair proposal after site inspection. Input Data Analysis Reporting Quoting Do you believe 72 hours is enough? Confidential and proprietary information of Wind. Com Page 13

Taking Data a Step Forward Use data for predictive maintenance plan Use data to predict occurrences of various damages by evaluating data of similar turbines/blades Geographic location Consider geographic area to bring in erosion and lightning damage Disruptive blade failure pattern identification Help the user to invest in preventative maintenance wisely getting the highest ROI Confidential and proprietary information of Wind. Com Page 14

Thank you. Questions? “Information's pretty thin stuff unless mixed with experience. ” – Clarence Day Confidential and proprietary information of Wind. Com Page 15

For all your wind turbine composites servicing needs North America Houston - USA T. +1 281 227 5130 F. +1 281 227 5192 Sales 281 906 9572 Europe Madrid -Spain T. +34 91 241 10 02 F. +34 91 293 51 96 South America Sao Paulo - Brasil T. +55 15 2105 4800 F. +55 15 2102 4875 info@windcomservices. com | windcomservices. com Confidential and proprietary information of Wind. Com Page 16