New Techniques of Reliability and their Application to Offshore Wind Farms

European Offshore Wind 2009, StockholmTechnology and InnovationOffshore Wind Turbine Reliability16th September 2009

Introduction - Michael Starling

Background

• Chartered Mechanical Engineer, started work in 1979, worked for BMT since 1990

• worked in Renewable Energy since 2004

• specialise in engineering and risk

• applied to transport, energy and the built environment

Current/recent projects

• construction, transportation and installation system for concrete offshore wind turbine foundations (with Gifford)

• reliability, maintainability and survivability guide for the European Marine Energy Centre in Orkney

• navigation impact assessment of a tidal fence across the Severn

Introduction - BMT

British Maritime Technology Ltd

established 1985 (now BMT Group Ltd)

NMI LtdBritish Ship

Research Association

Reliability – Some major projects

Air to Air Refuelling Tanker Aircraft

Reliability – Some major projects

Channel Tunnel Trains

Reliability – Some major projects

Ro Ro Ferries

Reliability – Some major projects

Offshore and Subsea Oil & Gas

Reliability – Personal contact with my work

Aircraft Fuel Pumps – Airport Baggage Handling – Airport Trains – Metros - Escalators

Reliability – A current project

Pulse Tidal Generator

Who achieves high reliability?

Aircraft achieve high reliability

An A330 will typically achieve greater than 98.5% operational availability

and they guaranteed it from day 1

How do they achieve high reliability?

Fundamental economic driver

A complete common purpose between safety, reliability, performance and profitability

International standards driven

Everything is specification, certification and approvals led

Technical drivers

Complete hierarchy of specification and certification from the smallest component to the whole aircraft and from an individual maintainer to the operator

Approvals are technical, organisational and individual

There is international commonality and transferability

Functional drivers

Aircraft design based on equipment functionality and integrity

and on appropriate redundancy

Appropriate redundancy

Redundancy “enhances high integrity”

It does not “compensate for low integrity”

What process do they follow?

Formal processes of assurance

Defining what the equipment, operation or service has to do

Designing, operating and maintaining it to do it

Finding some assurance that it will “work and keep on working”

Summary

Aircraft Achieve High Reliability By

Reliable Design

demonstrated by

Reliability Assurance

based on

Integrity, Functionality, Appropriate Redundancy and Comprehensive Testing

mandated by

Specification, Certification and Approval

and controlled through life by

Monitoring and Modification

Design for reliability

Maintain for reliability

Success-based reliability

Three topics for rest of this paper

Design for Reliability

How reliable does a device have to be?Common measure of reliability is Mean Time Between Failure (MTBF)

• Common belief that a 10 year MTBF means that the equipment will last about 10 years

• That is a 10 year life not a 10 year MTBF

• After 10 years running approximately 63% of “10 year” MTBF equipment will have failed

• For 1% failed the MTBF needs to be approximately 1,000 years

• Some MTBFs– Offshore Wind Turbine, 1 month– Domestic Boiler, 5 years – Double Glazing, 10 years

Reliability of the Device

(MTBF in years)

Reliability over Time Period

Unreliability over Time Period

Number of Failed Devices at end of 10

year time periodNumber of Devices on Site 100 1 0% 100% 100Time Period (Years) 10 10 37% 63% 63

50 82% 18% 18100 90% 10% 10

1,000 99% 1% 1

Does redundancy help?Typical solution to poor reliability is redundancy

• Works well for repairable systems

• Works badly for non repairable systems

• It works better for non-repairable systems when the equipment is reliable

• It is often better to spend money on increasing integrity rather than fitting redundancy

Dual RedundancyReliability of the

Device(MTBF in years)

Number of Failed Devices at end of 10

year time period

Number of Failed Devices at end of year time period

Number of Devices on Site 100 1 100 100Time Period (Years) 10 10 63 40

50 18 3100 10 1

1,000 1 0

Design for reliability - conclusion

Design for high integrity

Backup with redundancy only if easy to repair

Maintain for Reliability

What type of maintenance can I do?Preventive Maintenance

• The routine activities to prevent failure, i.e. the servicing

• Typically done to a planned schedule based on time or usage

• Ideal is to do when no wind resource available

Corrective Maintenance

• The activities required to respond to failure, i.e. the repairs

• Typically done to a reactive schedule

• Ideal is to avoid

Predictive Maintenance

• The activities required to respond to an indicator of future failure, i.e. maintenance triggered by some measurement of condition

• Ideal is to be able to defer predictive maintenance to times when preventive maintenance takes place

What type of maintenance should I do?

Wear-out

Degradation

Initial Success

Steady

Early Life Failure

Classical “Bathtub”

Depends on the nature of the failure

Effect of time scheduled maintenance

Wear-out

Degradation

Initial Success

Steady

Early Life Failure

Classical “Bathtub”

No difference

Made worse

Made better

Bit better

Bit better

Bit worse

Example using of aircraft data

Wear-out

Degradation

Initial Success

Steady

Early Life Failure

Classical “Bathtub”

No difference

Made worse

Made better

Bit better

Bit better

Bit worse 4%

2%

5%

7%

14%

68%

Maintain for reliability - conclusion

Define maintenance based on understanding the types of failure

Success-based reliability

A bit of history

These success-based techniques grew out of failure

• failure of reliability techniques to lead to change

• failure of techniques to improve reliability

• failure of techniques to be value for money

Led to questioning the fundamental reliability techniques

• techniques are focussed on failure

• should they be focussed on success?

Focus on success

For many years those of us in reliability have concentrated on understanding and eliminating failure.

• why things fails, when they fail, where they fail and how to stop them failing are questions that are examined in great detail.

However in doing so we may have overlooked the equal importance of understanding and creating success.

• why things work, when they work, where they work and how to make them work are equally, or perhaps more important, questions.

How to achieve success? – via assurance

• to define what the equipment, operation or service has to do

• to design it and operate it to do it

• find some evidence that it will work and keep on working.

• identify and eliminate threats to success.

Define

Evidence

Threats Design

Assurance via developing a reliability case

Part Technical Process

• that aims to provide the “Evidence of Success” and

• identify and eliminate the “Risk of Failure”

• and produce a “Reasoned Argument” supporting expected performance

Part Management Process

• that aims to provide “Scrutiny” that the evidence and argument is valid

Assessing the quality of the evidence

Proof?E8 Current Situation

E7Trials

(or Industry Standard Tests)

E6 Validated Simulations

E5Quantitative Calculations

(or Simplistic Quantitative Tests)

E4Qualitative Analysis(or Qualitative Tests)

E3 Expert Opinion - Written

E2 Expert Opinion - Verbal

E1 No Evidence

Best Evidence

Worst EvidenceExample Evidence Quality Matrix

Evidence?

Faith?

Producing a reasoned argument

The reasoned argument leads to

• a claim of expected reliability performance

• an assessment of the level of risk associated with the claim

There is an obligation to use all evidence, the supporting and the opposing

All the Evidence All the Conditions

Supporting

Evidence

Evidence

Evidence Science

Evidence Environment

Evidence Operations

Evidence

EvidenceChallenging

Concept of a Reasoned Argument

Requirements and

Assumptions

Reliability Claims

Reasoned Arguments

Typical Management Process – providing scrutiny

Scrutiny Processes

Authority to Proceedor

Need for Further Work

Issued Reliability

Case Report

Production Processes

Management Process - Production and Scrutiny of a Reliability Case

Plus Project

Plus Customer

Reliability Case Owner

Basis of Design

Solution Build Up

Reliability Case

Draft Reliability

Case Report

Updated Reliability

Case Report

HearingRed Team Review

Peer Review

This philosophy is not new

Rene Descartes

1596 - 1650

Knowledge should be based on

“Proof and evidence rather than just faith”

and

“Nothing should be accepted unless subject to scrutiny”

And finally

Some advice on how to achieve high reliability

• Specify the reliability you want– and specify it in terms meaningful to your business

• Design to achieve it– but beware the false promise of redundancy

• Build up a Reliability Case– and expose it to scrutiny (and don’t always believe your

experts!)

• Maintain for reliability– base your maintenance on understanding failure

Discussion

Michael Starling

BMT Fleet Technology

www.fleetech.com

mstarling@fleetech.com

+44(0)780 3925110

Typical Technical Process – building up evidence

Strategy to Build Evidence

Identify Need for Evidence

Reliability Cycle

Action to Build

Evidence

Evidence of Success

and

Risk of Failure

Identify Risk of Failure

Risk CycleAction to

Reduce Risk

Strategy to Reduce Risk

Technical Process - Building Evidence and Reducing Risk

Operational and Financial

Requirements

Reliability Requirements

Proposed Solution

Level of Risk associated with the

Claim

Claim ofExpected Reliability

Performance