Offshore Wind Turbine Blade Coating Deterioration Maintenance Model

Offshore Wind Turbine Blade Coating Deterioration Maintenance Model
Jesse A Andrawus and Laurie Mackay

School of Engineering, Robert Gordon University, Schoolhill, Aberdeen, AB10 1FR, UK

Abstract
Maintenance of offshore wind turbine blades has significant impact on the overall cost of managing offshore wind farms. Effective maintenance of protective coatings of wind turbine blades is one of the key challenges of offshore wind farms given that the current condition monitoring systems for wind turbines are limited in detecting coating deterioration. The durability and adhesion of coatings are affected by varying complex offshore weather conditions. This impact is further aggravated as access to offshore wind farms are costly and weather dependent. In this paper an effective predictive maintenance strategy for protective coating of wind turbine blades is developed. The strategy is based on risk analysis approach and coating codes for offshore installations. Practical implementation of the developed strategy will significantly reduces the overall maintenance cost of offshore wind farms.
Keywords: Offshore Wind farms, Wind Turbine, Blades, Coating deterioration, Maintenance.
1

Introduction

The European Wind Energy Association target of 40GW by 2020 and 150GW by 2030 [1] will require a significant number of grid connected offshore wind turbines. Currently there are about 43 offshore wind farms operating in Europe with a total capacity of about 2,400 mega watts (MW) [1]. The estimated total investment in European offshore wind energy generation is about €100-€150 billion ($130-$200 billion) [1]. Future forecast shows a very steep rise in offshore wind farms installation and most of it will be located in the United Kingdom
Continental Shelf (UKCS) North Sea [2]. However, degradation of protective coating of wind turbine blades is posing a serious challenge to the offshore wind industry. Wind turbine blades in offshore locations are fully exposed to UV, wind and water [3]. The combined effect of these variables reduces the durability of coating design life and hence affects the coating adhesion and blade surface performance. Thus developing and implementing a maintenance programme for coatings [4] is an important aspect of effective maintenance regime for offshore wind farms. This paper uses risk analysis technique to determine and optimise maintenance of offshore wind turbine blade coatings.
2
Design materials of wind turbine blades
Wind turbine blades are designed to harness wind movement and then transmit the rotational energy to the gearbox via the hub and main shaft. The blades of a wind turbine are usually made from composite materials. Composite materials are often preferred because of the possibility of achieving high strength and stiffness-to-weight ratio [5]. Composite materials are also corrosion resistant and good electrical insulators. These properties are advantageous in an offshore environment where corrosion is a critical factor to be considered. Figure 1 shows a typical wind turbine blade and the cross sectional area revealing the internal composition of a blade.

Leading Edge

Leading
Edge
Cross Section
Figure 1: A typical wind turbine blade and the cross sectional area
Table 1 shows some composite materials and binders commonly used in the manufacture of wind turbine blades. Composite blades can be described as carbon fibre reinforcing or Woodepoxy laminates or fibreglass reinforced plastic (GRP). GRP is the most commonly used blade because it is cheaper than other composite materials [6]. Furthermore, fibre glass has good tensile strength.
Table 1 Composites and binders used in manufacturing wind turbine blades
Composites

Binders or Resins

a

Fibreglass

a

Polyester (unsaturated)

b

Carbon fibre

b

Vinyl ester

c

Wood

c

Epoxy

Epoxy is a thermosetting polymer formed from reaction of an epoxide "resin" with an amine or amide curing agent. Epoxy has a wide range of applications, including fibre-reinforced plastic materials, coatings and adhesives. The two epoxies used in manufacturing process of the
Wind Turbine blades are: Epoxy Bisphenol A and Epoxy Bisphenol F. The epoxy resin is injected into the core materials (wood and foam) and reinforced with glass-fibre strands or woven glass fabric. The process is called Vacuum Assisted Resin Transfer Moulding
(VARTM). Due to its superior properties and low surface tension, epoxy is the most favoured resin for composites.
Polyurethane is any polymer consisting of a chain of organic units joined by urethane
Carbamate links. Polyurethane polymers are formed through step-growth polymerisation by reaction of a monomer containing at least two Isocyanate functional groups with another monomer containing at least two hydroxyl (alcohol) groups in the presence of a catalyst. The polyurethane is normally sprayed onto reinforced plastics in the moulding process as the adhesive due to the extremely strong bonding that takes place when the polyurethane crosslinking takes place. It is also applied for its superior erosion resistant and elastomeric properties. 3
Failure Characteristic of wind turbine blades
Interaction between wind turbine blades’ centrifugal and gravitational force as well as varying wind thrust and turbulence induce the blades to a cyclic and flap-wise loading. As a result, the
IEC TS 61400-23 requires full-scale blade test for strength, static and dynamic fatigue, stability and critical deflection to validate design certification. Wind turbine blades in normal operating conditions are known to fail as a result of cracks arising from fatigue [7-9], defects in materials accumulating to critical cracks [10-11] and lightening strikes [12]. Ice build-up is also known to cause failure of blades. Blade pitting, coating cracking and coating erosion have been reported to cause significant disruption to operational effectiveness and efficiency of wind turbines. The main thrust of this paper is effective management of protective coating failures. 4
Current maintenance of wind turbine blades
The current maintenance practises of wind turbines rely on periodic examination of turbines at a pre-determined calendar schedule or running hours, and the utilisation of turbines until it breaks-down before repair is carried out. However, inspection carried out during the regular services often results in extra visit to the turbines for repair purposes [13]. This requires additional access and personnel time which increases the overall maintenance cost. Access and logistics are sometimes restricted by weather conditions such as wave height and wind speed; this increases downtime and consequences of failures. Thorpe [14] explained that
“unexpected equipment failures occur between time-based maintenance intervals. Manpower, time, and money are wasted because periodic maintenance is performed with little awareness of the current equipment condition”. This is particularly of significant effect while considering the remote location of offshore wind farms.
Condition Based Maintenance (CBM) constitutes maintenance tasks carried out in response to deterioration in the condition or performance of an asset or component as indicated by condition monitoring processes [15]. Saranga & Knezevic [16], Arthur & Dunn [17] stated that
CBM is the “…most cost-effective means of maintaining critical equipment”. The monitoring of the structural integrity of turbine blades using thermal imaging and acoustic emission [18], the use of performance monitoring [19], on-line analysis systems [7], etc exist to determine incipient failures of a wind turbine. The following enumerate some condition monitoring techniques specifically applied to wind turbine blades.
4.1
Strain measurement involves attaching strain-gauges to the surfaces of a wind turbine’s blades are used to measure strain in the blades. This is done by measuring changes in electrical resistance in the strain gauge. The technique is used for laboratory life-time prediction and safeguarding of the stress level of blades [13].
4.2
Acoustic Analysis Technique involves attaching acoustic sensors to wind turbine blades, and then listening to the sounds generated by the blades. The sensors are attached to the blades by using flexible glue with low attenuation. Abnormal sounds which are not related to the dynamic loading of the turbine are indicators of possible blade failure. The variability of wind speed, wave, turbulence as well as the dynamic operating nature of wind turbines can limit the use of acoustic analyses technique [20].
4.3
Fibre Optics Measurement involves embedding optical fibre sensors in the blade structure to enable the measurement of five parameters which are critical to blade failure. The five parameters include; strain measurement which monitors the blade loading and vibration level, temperature measurement for likely over-heating, acceleration measurement to monitor pitch angle and rotor position, crack detection measurements, and lightning detection which measures front steepness, maximum current and specific energy. The loading data from blades sensors can be used for real-time pitch control. This reduces significantly the out-of balance loading on the tower and foundation. Drewry and Georgiou [21] argued that the fibre optics measurement is limited in detecting design faults which may lead to catastrophic failure during the operating life.
The above techniques focus on structural integrity and crack detection with no consideration to blade coating degradation which eventually leads to blade failure.

5
Coating Deterioration Model
The first step to determine an effective maintenance regime for protective coating of wind turbine blades is to identify the causes of protective coating degradation using Failure Modes
Effect and Causes Analysis (FMECA) technique. The second step is to determine the risk of blade coating failure taking into account the failure frequency and the potential consequences. 5.1
Failure Modes Effect and Causes Analysis (FMECA)
Table 2 shows the FMECA of a wind turbine blade coating failure, the FM denote failure mode while the FC denotes failure causes. It is worth noting that the FMECA analysis in the Table 2 did not include failure effects as the intention is to eliminate or prevent the causes of blade coating failure.
Table 2 Failure Mode Effects and Causes Analysis (FMECA)
Failure mode
FM1 Blade coating failure

Causes (level 1)
FC1 Blistering
FC2 Cracking

Causes (level 2)
_
FC2-1 Primary failure of the Matrix or resin all the way through to the inner wing
FC2-2 Widespread crazing of the resin or matrix from topcoat to first layer of the matrix
FC2-3 Longitudinal or transverse surface checking anomalies on the topcoat only

FC3 Fatigue

FC4 Flaking

FC3-1 Fatigue or stress cracks on the leading edge

FC4-1 Topcoat surface failure only
FC4-2 Topcoat and matrix delamination from the resin base coat

FC5 Erosion

FC5-1 Topcoat worn down to laminate
FC5-2 Topcoat and matrix worn down to base resin
FC5-3 Topcoat, matrix and resin worn down to primary frame FC6 Adhesion

FC6-1 Catastrophic: Failure of all coats from the inner skin or frame without warning
FC6-2 Premature failure of one or more coats to the substrate or inter-coat over a short period of time
FC5-3 Gradual failure of the topcoat to expose the matrix and resin during the life of the blade

FC7 Corrosion

FC7-1 Coating breakdown

5.2
Governing standards for classification of service environments
The International Standard Organisation (ISO) 12944 has classified six environments and durability ranges as listed in tables 3 & 4 based on experiments that have measured the rate of metal loss for uncoated steel. The classification of environments applies to structural steel exposed to ambient (less than 120°C/248°F) conditions. Wind turbine blades can be classified under the “C5M” in the table 4.

Table 3 ISO 12944 Classifications
ISO 12944
Classification Typical Environments
C1
Rural areas, low pollution. Heated
C2
buildings/neutral atmosphere.
C3
Urban and industrial atmospheres.
Moderate sulphur dioxide levels.
Production areas with high humidity.
C4
Industrial and coastal.
Chemical processing plants.
C5I
Industrial areas with high humidity and aggressive atmospheres.
C5M
Marine, offshore*, estuaries, coastal areas with high salinity.
* Corrosion protection in ISO 12944 C5M - Offshore environments is being addressed via a new standard (ISO 20340) dedicated to this environment
Table 4 ISO 12944 Durability (Time to First Major Maintenance)
High Durability
Medium Durability

>15 years design life
5-15 years design life

Low Durability

£400,000

3-8

1

5

£100,000£400,000

8-20

2

4

£25,000£100,000

20-40

3

3

£5,000£25,000

>40

4

2

£0

0-3

Minor

Major

Critical

Table 6 expresses coating and laminate deterioration as a percentage of the surface area.
When a condition survey is carried out, the surveyor can non-destructively examine the presented surfaces for evidence of failure and make a considered judgement on what intervention is required. During the sentencing of the coating condition the Table 6 suggests a period of time in which repairs should be implemented to minimise risk impact on the business efficiency and regulatory compliance obligations. Based on the out put of the risk matrix (Table 5) and the description of tasks required for each category of consequences shown in Table 6, optimal maintenance regime for wind turbine blade coatings can be develop as shown in Table 7.

Once the coating condition is known the engineering maintenance regime is selected to assist in the development of job step plans, risk assessments and life-cycle costing of each option.
The content of the Table 7 is dynamic to coincide with operator intervention strategies and geographical demands placed upon the wind turbine operator.

The range of hire charges for dynamic positioning marine vessels with 10 - 50 tonne crane capability ranges between £75,000 and £150,000.
Table 6 Description of tasks required for each category of consequence

Risk

Criticality

Blade Coating condition (% of
Coating failure)

High

5

>40

Replace

12 Hrs

Yes

Yes

Med

4

30 - 39

Sheath or localised coating repair

48 Hrs

Yes

Yes

Low

3

19 - 29

Localised spot repair 7 days

Yes

No

Unlikely

2

>10

No repair

N/A

No

No

Rare

1

10<

No repair

N/A

No

No

Recommended
Intervention

Visual condition

Time to effect Business repairs Critical

Safety
Critical

Table 7 Maintenance regime for wind turbine blade coatings

Criticality

Task
Required

Number of technicians Surface
Preparation
methodology

Remedial
Coating

1

No requirement to rectify a cat 1 defect

1

N/A

N/A

1

0

2

No requirement to rectify a
Cat 2 defect

1

N/A

N/A

1

£5,000 £25,000

3

Surface preparation and rectification of a Cat 3 defect 4

Surface preparation and rectification of a Cat 4 defect 5

Surface preparation and rectification of a Cat 5 defect 3

Portable blasting equipment with Fast cure 2soft pack environmentally polyurethane friendly abrasive

4

Portable blasting Fast cure 2equipment with pack soft polyurethane environmentally or a *NPD friendly abrasive blade sheath

10

Blade replacement

Difficulty of
Cost in £ 's access factor

2

2

N/A

Comments

Survey only and update database with condition and replan next survey
Continue to monitor deterioration for propogation of stress cracks and coating erosion, update database and replan next survey level for
Intervention

effective and efficient remedial works.
£25,000Update database and
£100,000
replan next survey

Option 1: Prepare and liquid coat
Option 2: Replace with
£100,000 vacuum applied
£400,000
sheath update database and replan next survey
Uneconomical field repair. Replace blade
Fixed cost at earliest opportunity
> £400,000

Time to effect remedials in the
Integrated Asset
Plan (IAP)

Optimal survey period N/A

36 months

N/A

24 months

90 days

12 months

30 days

1 month

Immediate

36 months

6 Conclusion
This paper critically assessed wind turbine blade coating failures using the Failure Modes
Effect and Causes Analysis. Risk analysis matrix was developed based on coating history factor Chf and failure consequences FC . The risk matrix was used to determine optimal maintenance regime for offshore wind turbine blade coatings.
References
1 European Wind Energy Association (EWEA) http://www.ewea.org/index.php?id=203
2 Offshore Wind – Key statistics (2011). Factsheet produced by the Energy Generation and Supply Knowledge Transfer Network. www.innovateuk.org/energyktn
3 ISO 12944.
4 Axelsen, S.B., Knudsen, O.O. & Johnsen, R. (2009). Protective Coatings Offshore:
Introducing a risk based management system. NACE International Corrossion
Conference 2009. Paper No 09016.
5 Manwell, J. F., McGowan, J. G. & Rogers, A. (2002), Wind Energy Explained: Theory,
Design and Application, John Wiley & Sons, Ltd.
6 Burton, T., Sharpe, D., Jenkins, N. & Bossanyi, E. (2001), Wind Energy Handbook,
JohnWiley and Sons Ltd.
7 Philippidis, T. P. and Vassilopoulos, A. P., (2004) Life prediction methodology for
GFRP laminates under spectrum loading, Renewable Energy, 35, 657–666.
8 Infield, D. (2004), Condition monitoring of wind turbines, Technical report, Centre for
Renewable
Energy
Systems
Technology
Loughborough
University.
URL:http://www.hie.co.uk/Renewables-seminar-04-presentations/crest-davidinfield.pdf [Accessed 15th February 2011]
9 Dutton, A. (1991), Infra red condition monitoring techniques for composite wind turbine blades, in ‘British Wind Association Conference’.
10 Jorgensen, E. R., k. Borum, K., McGugan, M., Thomsen, C. L., Jensen, F. M.,
Debelog, C.P. & Sorensen, B. F. (2004), Full scale testing of wind turbine blade to failure- flapwise loading, Technical Report Riso-R-1392(EN), Riso National
Laboratory, Roskilde.URL: http://www.risoe.dk/rispubl/VEA/veapdf/ris-r-1392.pdf
[Accessed 15th February 2011].
11 Anastassopoulos, A. A., Kourousis, D. A., Dutton, A. G., Blanch, M. J., Vionis, P. &
Leku,D. J. (2002), ‘Real-time evaluation of wind turbine blades with acoustic monitoring during certification tests’, Non Destructive Techniques NDT.net 7(09).
URL: http://www.ndt.net/article/v07n09/12/12.htm [accessed 24th July 2010]
12 Conover, K., VandenBosche, J., Rhoads, H. and Smith, B (2000) Review of operation and maintenance experience in the DOE-EPRI wind turbine verification program,
Proceedings of American Wind Energy Association’s WindPower 2000,
NREL/CP-500-28620.
13 Verbruggen, T. (2003) Wind turbine operation and maintenance based on condition monitoring WT-Ω final report, Technical Report, ECN-C–03-047, Energy Centre
Netherlands.
14 Thorpe, C., (2006) Condition-Based Maintenance for CVN-21and DD (x), Empfasis,
2005, URL: http://www.empf.org/empfasis/july05/cbm705.htm [accessed August
2010]
15 Moubray, J. (1991), Reliability-Centred Maintenance II, 2nd ed. ButterworthHeinemann, Oxford.
16 Saranga, H. and Knezevic, J., (2001) Reliability prediction for Condition-Based maintained systems, Reliability Engineering and System Safety, 71, 219-224.
17 Arthur, N. and Dunn, M., (2001) Effective Condition Based Maintenance of reciprocating compressors on an offshore oil and gas installation, IMechE
International Conference on Compressor and their system, 2001.
18 Clayton, B. R., Dutton, A. G., Aftab, N., Bond, L., Lipman, N. H. and Irving, A. D
(1990) Development of structural condition monitoring techniques for composite wind turbine blades, proceedings of European Community Wind Energy
Conference, 10-14.
19 Learney, V.C., Sharpe, D.J., and Infield, D., (1999) Condition monitoring technique for optimisation of wind farm performance, International Journal of COMADEM, 2(1),
5–13.

20 Jungert, A. (2008). Damage Detection in Wind Turbine Blades using two Different
Acoustic
Techniques.
The
e-Journal of Non-destructive
Testing.
www.ndt.net/search/docs.php3?MainSources=25.
21 Drewry, M.A & Georgiou, G.A., (2006). A review of NDT techniques for wind turbines.
The e-Journal of Non-destructive Testing. Wind turbine NDT.

References: 1 European Wind Energy Association (EWEA) http://www.ewea.org/index.php?id=203 2 Offshore Wind – Key statistics (2011) 4 Axelsen, S.B., Knudsen, O.O. & Johnsen, R. (2009). Protective Coatings Offshore: Introducing a risk based management system 5 Manwell, J. F., McGowan, J. G. & Rogers, A. (2002), Wind Energy Explained: Theory, Design and Application, John Wiley & Sons, Ltd. 6 Burton, T., Sharpe, D., Jenkins, N. & Bossanyi, E. (2001), Wind Energy Handbook, JohnWiley and Sons Ltd. 7 Philippidis, T. P. and Vassilopoulos, A. P., (2004) Life prediction methodology for GFRP laminates under spectrum loading, Renewable Energy, 35, 657–666. 8 Infield, D. (2004), Condition monitoring of wind turbines, Technical report, Centre for Renewable 13 Verbruggen, T. (2003) Wind turbine operation and maintenance based on condition monitoring WT-Ω final report, Technical Report, ECN-C–03-047, Energy Centre 14 Thorpe, C., (2006) Condition-Based Maintenance for CVN-21and DD (x), Empfasis, 2005, URL: http://www.empf.org/empfasis/july05/cbm705.htm [accessed August 16 Saranga, H. and Knezevic, J., (2001) Reliability prediction for Condition-Based maintained systems, Reliability Engineering and System Safety, 71, 219-224. 17 Arthur, N. and Dunn, M., (2001) Effective Condition Based Maintenance of reciprocating compressors on an offshore oil and gas installation, IMechE International Conference on Compressor and their system, 2001. 18 Clayton, B. R., Dutton, A. G., Aftab, N., Bond, L., Lipman, N. H. and Irving, A. D (1990) Development of structural condition monitoring techniques for composite 19 Learney, V.C., Sharpe, D.J., and Infield, D., (1999) Condition monitoring technique for optimisation of wind farm performance, International Journal of COMADEM, 2(1), 20 Jungert, A. (2008). Damage Detection in Wind Turbine Blades using two Different Acoustic 21 Drewry, M.A & Georgiou, G.A., (2006). A review of NDT techniques for wind turbines.