Day 1 - Wednesday 18 October 2017

Registration And Refreshments


Chair’s Opening Remarks

Thomas Pump, Head Of Asset Information Systems, E.ON

Using Data To Reach The Full Potential Of Your Wind Farm


Keynote Address: Using Data To Improve O&M Efficiency And Drive Down The LCOE

  • What has moved on with data capabilities?

  • Utilising automation and smart systems to reduce the number of times you send employees offshore

  • Planning operational activities based on weather forecasting

  • Driving down costs without eroding the certainty in level of results

  • How to increase the life of a turbine and decrease maintenance required for less OPEX expenditure

  • How can you demonstrate return on investment for data initiatives?


Maintenance And Reliability Data And Getting The Best Out Of Drone Inspections

  • How can you use data from maintenance activities to improve operations?
  • What can be done on a data front to support technicians going up towers?
  • How to get maintenance data in a structured way so that an algorithm can use it
  • What is done with the data from drone inspections? and what impact does it have in monitoring the health of a turbine?

Latest Analysis Techniques: How Are Failing Turbines Detected?

  • Can you create relative comparisons with the rest of your fleet?
  • Predictive analytics using weather forecasts to plan maintenance
  • What data is required for recognising underperforming turbines?
    • What tools do you need to have in place?
    • What analysis platforms do you need?
    • How to get to the root cause of the problem and find an actionable recommendation for technicians on what to fix

Morning Refreshments And Networking


Applying Operational And Event Data To Understand The Turbine’s Performance And Reliability Behaviour

  • Why data is needed to describe continuous condition of wind turbines
  • Identifying abnormal behaviour to assess and improve performance
  • Enabling predictive maintenance by association analyses

Stefan Faulstich, Reliability Analyst, Fraunhofer IWES

Forecasting Failure And Performance Degredation With Better Measurement Tools And Analysis


Understanding Power Performance Degradation And Possible Power To Assess Performance Over Lifetime

  • Exploiting SCADA to understand how the performance of a turbine develops over its lifetime 
  • Taking ageing and erosion data into consideration when calculating power output
  • Best techniques for reducing maintenance required whilst maintaining optimal power output 
  • The possible power: what a wind farm could produce without environmental restrictions 

Getting Better Wind Measurements For Improving The Output Of Your Assets 

  • Employing more accurate anemometers for enhanced wind predictability and using wind flow and mast flow models

  • Comparing the output vs. the predicted power and using recursive loops to fit your wind farm output

  • Comprehending the difficulties of calculating loads and wind speeds in complex/mountainous terrains to better predict performance? 

  • Desktop linear flow models vs. cfd models with climatological physics

  • Maintaining uncertainty at an acceptable level and lifting environmental restrictions to increase performance

Networking Lunch


Measurement And Optimisation Of Blade Angle Deviations And Quantification Of Subsequent Performance Improvements

  • How to measure blade angle deviations pros/cons of different systems on the market
  • Development of an calculation method to quantify the performance improvements after correction of the blade angles
    • Impact of blade angle misalignments on the measured wind speed on top of the nacelle
    • Is the performance correlation with neighbouring turbines a reasonable approach to quantify the improvement?
  • Verification of the developed calculation method using a LiDAR for the complete characterization of the incoming wind field

Dr Thomas Burchhart, Fleet Performance Analyst, VERBUND Hydro Power GmbH

Understanding Preventative Maintenance, Prognostics And Artificial Intelligence To Improve O&M Activities


Turning Predictive Analysis Into Preventative Maintenance: Fleet-Based Operational Optimisation As An A.I. Task

  • Getting enough good data for the task
  • How combining data sources can result in better predictions
  • Turning data alarms into workable task orders
  • Using signals from vibrations, scada data, temp data and pressure data to get a more effective overall picture of maintenance required

Elena Tsiporkova, Team Leader Data Innovation, Sirris


Condition Monitoring Utilising Diagnostics, Prognostics And Sensor Flow

  • Understanding fixed analysis and images to recognise specific error, deviation and pattern matching
  • Using virtual simulation models whilst combining scada and vibration data to assess for maintenance needs and possible malfunctions
  • Learning to interpret data earlier and faster to spot potential failures and reduce risk
  • How to predict what will happen based on measurements in the turbines
  • Monitoring temperature from gearboxes and generators from scada data to optimise production
  • Electrical component monitoring

Networking Break And Afternoon Refreshments


Applying Machine Learning And Complex Algorithms For Easy Data Computing And Better Results

  • Addressing underlying issues and difficulties in machine learning
  • Best techniques and real life case study examples
  • Using machine learning to crunch the predictive maintenance data for greater efficiency
  • Could machine learning be employed in security of data?
  • Why an algorithm is only as good as the data used and the problems of applying machine learning to real world applications?

Is The Wind Industry Ready For Artificial Intelligence (A.I.)?

  • What have we learned in A.I that can be implemented to improve wind industry operations and maintenance?
  • Tools that can be used in the A.I. suite alongside machine learning network to network
  • Allowing enough time for machines to learn from mistakes
  • Why the need for cloud computing in order for A.I. to work
  • Why its application to the real world is so difficult and how data quality slows progress and ability to learn
  • What work is currently being carried out using neural networks?

Speed Networking

Realising the importance of connecting with your peers, we organised a moderated networking session where delegates are prompted to meet others in brief 3/4 minute rounds. The moderator will be keeping track of time and announcing participants when to switch partners. Make sure you bring lots of business cards along with you!

Chair's Closing Remarks 

Thomas Pump,  Head Of Asset Information Systems,   E.ON


End of Day 1