1. No category

Datenanalyse, Methoden und
Prozesse für Ertragsmaximierung
Spreewind 2014
Author
Date
Reference
GRAY
28.03.2013
V1.0
Overview

History, motivation

Infrastructure, software, processes

Diagnostics, prognostics and CBM

Summary
© Uptime Engineering GmbH
14.11.2014
What‘s So Great About SCADA?
Advantages and opportunities provided by available data
© Uptime Engineering GmbH
14.11.2014
How Did I Get Here?
Optimisation of
Engine Reliability
Optimisation of
Wind O&M
© Uptime Engineering GmbH
14.11.2014
Multiple Failure Modes
Proportion Of All Failures [%]
35
Germany (WS) [6]
30
Germany (LWK) [6]
Ref: Reliawind 2010
Denmark (WS) [6]
25
Sweden [7]
Finland [7]
20
Germany [7]
15
10
Generator
Hydraulics
Gearbox
Blades / Pitch
Mechanism
Control
System
Electrical
System
0
Yaw System
5
Ref: Gray & Watson, Wind Energy 2009
Wind turbine reliability affected by a high volume and frequency of failure modes
© Uptime Engineering GmbH
14.11.2014
Wealth of Information
Large volumes of information available to support analysis tasks
© Uptime Engineering GmbH
14.11.2014
Multiple Applications of SCADA Analytics
WS
RP
BA
YM
LH
HS
RS
PD
0,0
Production forecasting
0,2
0,4
0,6
1,2
1,0
0,8
Normalised Parameter [-]
1,4
1,6
1,8
2,0
Failure prognostics
Yaw misalignment detection & correction
Outlier detection
SCADA can be applied for a large range of tasks relating to wind turbine operations
© Uptime Engineering GmbH
14.11.2014
Infrastructure, Software, Processes
Requirements for effective application of SCADA analytics
© Uptime Engineering GmbH
14.11.2014
Next Level SCADA
EFFICIENT PROBLEM IDENTIFICATION  OPTIMISATION
Condition Monitoring, Business Intelligence, Reporting, Analytics
Centralised SCADA Database
Multiple Information
Sources
(bill of materials, service
reports, CMS etc)
SCADA
SCADA
SCADA
OEM #1
OEM #2
OEM #3
© Uptime Engineering GmbH
14.11.2014
Software & Analytics
Software should allow the operator to quickly evaluate past, present and future operations
© Uptime Engineering GmbH
14.11.2014
Process Integration
FOCUS
ANALYSE
INSPECT
CORRECT
CONFIRM
MONITOR
Data analytics must be effectively integrated into operational processes for value generation
© Uptime Engineering GmbH
14.11.2014
Practical Examples
Diagnostics, Prognostics, CBM
© Uptime Engineering GmbH
14.11.2014
Diagnostics & Prognostics
Failure mechanisms can be understood, modelled, quantified and monitored
© Uptime Engineering GmbH
14.11.2014
Knowledge Database
Uptime HARVEST™ knowledge DB (example: Blade)
Structured information storage within an integrated system support continuous improvement
© Uptime Engineering GmbH
14.11.2014
Bearing Temperature [°C]
Outlier Detection
= Measured
= Model
Residual [°C]
Measurement Point [-]
Measurement Point [-]
Fully automated and accurate outlier detection using physics-based models
© Uptime Engineering GmbH
14.11.2014
Physics of Failure
0
5
10
15
Failure Rate vs Time at HIGH LOAD
20
Proportion of failed turbines [%]
Proportion of failed turbines [%]
Failure Rate vs Time
0
0.5
1.0
1.5
2.0
2.5
Time at Condition [years]
Survival
Probability [%]
Operational Time [years]
Turbine
Physics of Failure modelling combined with SCADA improves prognosis accuracy
© Uptime Engineering GmbH
14.11.2014
Cost Minimisation
Planned vs Unplanned Repair Costs
Optimium
replacement
interval
Introduction of well-timed preventative maintenance reduces overall costs
© Uptime Engineering GmbH
14.11.2014
Improvement: Recommendations and Potential
Recommendations
•
Intermediate replacement, instead of run-to-failure
•
Prioritisation according to accumulated damage
•
Online monitoring (SCADA, alarms)
•
Failure rate tracking and adaptation of plan
Planned replacement
Failure Probability [%]
Failure Probability
End of Life
online monitoring & regular inspection
2
4
6
8
12
10
Time [years]
14
16
18
20
Combination of maintenance scheduling, monitoring and prognostics for minimum total cost
© Uptime Engineering GmbH
14.11.2014
Summary

Data analytics can support a wide range of optimisation activities

Requirements: data infrastructure, software, analytics, process integration

Physics based methods provide transparency and are highly effective

A combination of diagnostics, prognostics and CBM provides financial gains

Power of SCADA: available, plentiful, continuous, detailed, growing
© Uptime Engineering GmbH
14.11.2014
Thank you
© Uptime Engineering GmbH
14.11.2014