Abstract
The wind industry is trying to find tools to accurately predict and know the reliability and availability of newly installed wind turbines. Failure modes, effects and criticality analysis (FMECA) is a technique used to determine critical subsystems, causes and consequences of wind turbines. FMECA has been widely used by manufacturers of wind turbine assemblies to analyze, evaluate and prioritize potential/known failure modes. However, its actual implementation in wind farms has some limitations. This paper aims to determine the most critical subsystems, causes and consequences of the wind turbines of the Moroccan wind farm of Amougdoul during the years 2010–2019 by applying the maintenance model (FMECA), which is an analysis of failure modes, effects and criticality based on a history of failure modes occurred by the SCADA system and proposing solutions and recommendations.
General introduction
Morocco, among other countries in the world, faces increasing energy needs and limited conventional energy resources. The country is heavily dependent on energy imports. In addition, it is strongly affected by the effects of climate change (CC) and presents a high vulnerability to global warming at different levels. Forests, agriculture, fisheries, water supply and tourism are among the most vulnerable ecosystems. Morocco has adopted a new green energy policy that will allow the country to benefit from its abundant wind potential estimated at 25,000 MW (Ezzaidi et al., 2017). The assessment of achievements and projects underway at the end of 2019 shows that the installed capacity amounts to 3685 MW, including 700 MW for solar, 1215 MW for wind and 1770 MW for hydropower (MÉDIAS24, 2020). The integration of renewable energy resources is becoming a reality today. Wind power, offering lower costs than other energy resources, has grown rapidly around the world and has recently gained much attention due to its great benefits in reducing GHG emissions (Ezzaidi et al., 2017). A wind turbine is a device that transforms part of the kinetic energy of the wind into mechanical energy and then into electrical energy via a generator. Knowing that the energy efficiency of wind turbines depends on the reliability of the whole system, therefore it is necessary to optimize the design of the wind turbines and minimize the costs of electricity production, which is a cost forecasting tool for investors. A study focuses on the evaluation of the performance of the Amougdoul wind farm by Ezzaidi et al. (2017). They estimated the energy production of a wind turbine which is respectively 134.5 kW and 194.19 KW corresponding to 27% and 39% of the available wind energy, which confirms that the conversion efficiency does not exceed 40%. In various regions of the world, the deployment of wind turbines for power generation has become increasingly competitive Mostafaeipour (2010). Onshore wind energy can be conceptualized as a “global innovation system” in which resources can drive technological development and diffusion: knowledge, markets, investment and legitimacy Rohe (2020). The growing contribution of wind power to the overall energy structure underscores the need to comply with production standards in a safe and environmentally friendly manner and to ensure energy security at acceptable electricity prices (Burrows and Communications, 2018). The main reasons for the shortage of wind farms were found in the first batch of studies are:
Low reliability of wind turbines (Slot et al., 2020).
Underestimating access restrictions for asset maintenance (van Bussel and Zaaijer, 2001).
The application of corrective maintenance; the initiation of maintenance activities reactively after the failure of a component or part of the wind system (Gonzalez et al., 2017).
The application of this strategy and the corresponding impact on the availability of the wind turbine have been studied in some publications (Shafiee and Sørensen, 2019). A study conducted a maintenance strategy of around 12 million euros per year for a 500 MW wind farm in order to assess the impact of production losses due to downtime. The direct cost of corrective maintenance is approximately four times the cost of preventive activities (Scheu et al., 2012). In order to keep wind turbines running to generate electricity reliably, great efforts have been made. The availability depends on the conditions of the site of the wind farm and also on the technical characteristics. Efforts to achieve usability goals are preventive maintenance activities, such as regular replacement of wearing parts, oil, or grease and responding to unpredictable conditions such as failure of the machine. Wind turbine by correctly passing the helicopters and with spare parts, tools and technicians always available. As the industry has evolved, maintenance strategies have been developed and concepts such as condition based monitoring (CBM) have been developed. CBM is a form of preventive maintenance, including changes in physical conditions, analysis and all subsequent maintenance actions. CBM strategies can avoid failures by understanding the physical principles of failures and the corresponding sequence of subsequent target maintenance activities (El-Thalji and Jantunen, 2012). The implementation of CBM requires the installation of sensors and the application of analysis tools for various WT operating signals, which increases the complexity and cost of operating the wind farm (Scheu et al., 2019).
The main objective of this article will be to apply a maintenance model for the main onshore wind systems for the Amougdoul wind farm in ESSAOUIRA during the period 2010–2019, taking into account all the main assemblies of the kinematic chain and the effects of their failure on the overall performance of the turbine. A primary objective will be to examine how the method handles a significant design and optimization change proposed for the WT. A logarithmic scale for the FMECA procedure is also proposed in this article. The document is organized as follows: section 2 presents a theoretical background on risk prioritization, FMECA and PARETO analysis, section 3 concerns the baseline and survey platform. Section 4 presents the FMEA methodology and procedure used for the wind farm using the maintenance management tools and logarithmic scales, it also shows the quantitative results of FMEA. Section 4 describes the wind turbine’s preventive maintenance plan, proposed solutions, improvements and recommendations. And by the end a conclusion.
Risk prioritization, FMECA and PARETO analysis
Several qualitative and quantitative methods can be applied to determine the priority of system or failure scenarios for further investigation in our case of SCADA-type wind turbines. FMECA is classified as a semi-quantitative reliability method. It is defined as a system process that can identify potential design and process failures before they occur, either to eliminate them or to minimize the associated risks. The concept of risk is used to prioritize the system and assess what can happen, for example, component failure, the probability of occurrence and the consequences. The results can be divided into different categories depending on the objectives and the risk assessment. In general, consequences include the impact on the cost and value of assets as key performance indicators, employee perception and safety considerations and the impact on the environment. In addition, Scheu et al. (2019) have applied the FMECA maintenance model to offshore wind turbine subsystems. They determined an overview of the overall distribution of the identified failure modes for each of the main systems with a categorical classification.
In order to apply the FMECA maintenance model, it is necessary to distribute the system according to the functional location of the equipment and to collect information on the operation of the system: First, fully understand the operations that the system must perform during normal operation; second, divide the system into several parts and use schematics and flow charts to identify the relationships between the subassemblies; third, prepare a complete parts list for each subassembly and then identify the main operational and environmental constraints that cause failure modes of each subassembly. Tables 1 and 2 below present the root causes and effects of failure modes in wind energy systems.
Root causes of the failure modes in wind turbine systems (Ozturk et al., 2018).
Locations, causes and effects of the failure (Qian et al., 2019).
The Pareto analysis method identifies the main causes (20%) that lead to 80% downtime. Once the main causes have been identified, diagnostic techniques such as the Ishikawa diagram can be used to identify the deeper causes of the problems (Mulder, 2012; Powell and Sammut-Bonnici, 2015).
To apply Pareto analysis in practice, some basic steps have been defined that can be followed to obtain a complete analysis: a complete analysis are presented in Table 3.
Pareto analysis in practice (Mulder, 2012).
Context and survey platform; ESSAOUIRA Amougdoul wind farm
The wind farm is located in Morocco at the city of ESSAOUIRA, the site (Coordinates: 31.413263, −9.802698) is located on one of the westernmost points of the central Atlantic coast of Morocco (Cap Sim). This park includes 71 asynchronous wind turbines, of the GAMESA G52-850 type, with a nominal power of 850 kW (Ozturk et al., 2018). All wind turbines are connected via an internal underground electricity network to the national MV 60 KV network. The park also includes an MV (Medium Voltage)/HV (High Voltage) transformer station for the evacuation of energy using digital technology. The park was built as part of the national strategy to promote renewable energies. Main onshore wind systems are shown in Figure 2. Table 4 shows the data for Gamesa G52/850 type wind turbines. Table 5 presents the characteristics of the Amougdoul wind farm.
SCADA data history for Amougdoul wind farm during 2010–2019.
Main systems of onshore wind turbines (Scheu et al., 2019).
Gamesa G52/850 wind turbines data (The Wind Power, 2018).
Characteristics of the ESSAOUIRA Amougdoul wind farm (RCREEE, 2012).
The database recorded in Excel according to the SCADA system is organized as follows (Figure 1): Type of maintenance, date of failure, nature of failure, end of failure, nature of intervention, downtime, Year.
Material and method
The park has gathered several problems in terms of maintenance management. It has recently suffered serious consequences as a result of this mismanagement. Consequently, a loss of production due to the organization, unreliability of the equipment and the method of maintenance monitoring.
Our objective is to study the duration and the number of breakdowns of the wind turbine, the possible and main causes which present an obstacle to the good approach of the maintenance the problems which lead to the degradation of the production and to make proposals improvement of the maintenance function for the ESSAOUIRA wind farm in 2010–2019, the aim of which is to optimize the production and cost of maintenance and to have a good operation of the maintenance department according to the following approach:
Analyze the current state of wind power and detect availability issues.
Justify the choice of critical elements to be studied, based on the calculation of maintenance indicators.
Apply the FMECA method.
Carrying out maintenance follow-up with proposed solutions.
Application of the PARETO analysis
The PARETO analysis will be applied, focusing for the moment on finding the critical elements that cause the system to fail and that reduce the availability of the machine. To properly select the critical components, we will react to the history of wind turbine failures during 2010–2019.
From the Table 6 “PARETO analysis” we draw the PARETO diagram to determine the most critical elements of the wind turbine as shown in Figure 3.
PARETO analyses.
PARETO graph.
Interpretation of the curve:
The PARETO curve is made up of three zones
Class A: 20% of the causes responsible for 80% of the effect.
Class B: 30% of the causes responsible for 15% of the effect.
Class C: 50% of the causes responsible for 5% of the effect.
Based on the analysis of the failure history, we have noticed that the most critical items are: Generator, Gearbox. Then the next step will be the application of the FMECA method for wind turbines by identifying its critical subassemblies.
Application of the FMECA method
Rating scales
From a few pivotal works (N. Daujeard, Novembre 2009), the rating used is for a four-level scale established from the literature and validated in a working group, the stages of a FMECA have been identified, including strategies aimed at applying this knowledge. In order to make the study homogeneous, the criticality of the failures of all wind turbines will be evaluated according to the same rating scale, based on three independent criteria: the frequency or probability of occurrence (F), severity (G) and probability of non-detection (D). This made it possible to calculate a criticality score (C) according to the formula C = G × F × D. The criticality threshold made it possible to identify failures which should lead to improvement actions. To each criterion we have associated a rating scale defined according to levels based on: the history of maintenance department shutdowns and staff experience. Indeed, the rating scale is based primarily on the time of downtime as well as the number of wind turbine failures.
Frequency
Frequency of failure due to a particular cause. Table 7 present the frequency scale.
Frequency scale.
Gravity
Table 8 shows the gravity scale. This is the severity of the effects of the failure:
Productivity losses (production stoppage).
Cost of maintenance.
Safety, environment.
Gravity scale.
Non-Detection
Table 9 presents the non-detection scale in four levels.
Non-detection scale.
Criticality
The scale of criticality presented in Table 10. It discriminates against actions to be taken and calculates them on the basis of severity, frequency and failure of non-detection.
Scale of criticality.
FMECA grid on the critical elements of wind turbines
The database registered in the excel, it is arranged and organized according to the systems and subsystems of the wind turbines, then presented in the form of a diagram.
Generator
2880 hours as higher downtime for stator failure, with a frequency of 170 times for 9 years as shown in Figure 4. The generator step change ranks second for downtime of 1920 hours with a frequency of 100 times for 9 years. Overload; environmental effects; misalignment; fatigue; mechanical failure; Loss of transmission control are the relatively related causes of these generator failures.
Diagram of downtime and frequency of generator failures in Amougdoul for the period 2010–2019.
According to the graph showing the failures associated with the rotor (Figure 5), the K52 rotor fault is the highest for a downtime of 182.13 hours and frequency of 37 during the years 2010–2019.
Diagram of downtime and frequency of rotor failures for the period 2010 to 2019.
The diagram in Figure 6 shows a decrease in downtime and also the frequency of failures associated with the stator element. K52 stator fault and magnetic stator error are the most critical subsystems in the stator element with a downtime of 542.30 and 459.05 and frequency of 77 and 29 during the years 2010–2019. Figure 7 shows the effects of stator failures for the subsystems.
Diagram of downtime and frequency of stator failures for the period 2010-2019.
Generator failure: (a) bearing, (b) loss of magnetic wedge and (c) contamination (Zhu and Li, 2018).
The following Tables 11 and 12 show the failure modes of the generator and gearbox, their effects and their criticality values.
FMECA generator grid.
Gearbox FMECA grid.
Gearbox
The diagram presented in Figure 8 illustrates the distribution of downtime per part in the wind farm over the 9 years in real time and the frequency of gearbox failures that occurred from 2010 to 2019; the results show that gearbox failures related to change of gearbox and main shaft cause the most downtime (3360 hours) with a frequency of 20 times. We can see that the fault of the multiplier pump has the greatest number of failures (100 times), followed closely by that of the rubber change of the multiplier shock absorbers (100 times).We presented a state of the art on the different failure modes of the gearbox for the horizontal axis wind turbine (Lamhour and Tizliouine, 2019). Figure 9 shows the effects on bearing failures for the subsystems.
Diagram of downtime and frequency of gearbox failures for the period 2010–2019.
Bearing failure: (a) bearing in the gearbox; (b) and (c) appearance of bearing failure (Zhu and Li, 2018).
PARETO analysis
We will do the PARETO analysis to determine the most critical subsystems Table 13.
Critical components.
Interpretation
According to the FMECA analysis “Analysis of Failure Modes, their Effects and their Criticality,” we noticed that the most critical components those which have a high criticality, we considered four levels of criticality according to the next board Figure 10:
PARETO diagram for criticality.
Interpretation
According to the FMECA analysis “Analysis of Failure Modes, Their Effects and Criticality,” we noticed that the most critical components from Table 14 are those with a high criticality “32” which is act from:
K52 stator fault: the stator is a key part in the generator for the production of current. A simple failure in the circuit of this component can generate a reduction or even no energy production.
Change – Gomes dampers multiplier (GD): By a permanent vibration of the wind system, climate change and temperature variation in the installation, the damper gomes of the multiplier get tired and deteriorate consequently the installation of the multiplier loses the alignment with the generator and may cause a fault in the latter.
Multiplier oil pump fault: A malfunction of the pump may lead to poor or absence of lubrication of all the multiplier elements. This may cause heating due to friction at the bearings and gears, consequently the system seizing up eminent.
Generator bearing: this is a very important element in the generator and for balance. The bearings are designed to operate in a clean environment and under ideal conditions. Bearing failures can be caused by fouling and the intrusion of foreign objects. Premature wear is often the result of inadequate or insufficient lubrication. If the bearing is running loudly and there is a rapid rise in temperature, this is an indication of insufficient lubrication or too much oil. If the diameter of the shaft is too large or the bore is too narrow, deformation of the rings occurs, resulting in spalling on the raceways. When one of the tracks of a groove bearing is poorly adjusted, the latter is subjected to indeterminate loads which are added to the normal load, causing an overload which deteriorates the assembly. Poor alignment, electrical current flow, poor lubrication, overloading, welding, oversized shafts and mesh on the bearing cages can cause bearing damage.
Criticality levels.
Critical analysis of the current maintenance of the ESSAOUIRA wind farm
According to the current maintenance policy (GAMESA manufacturer’s maintenance manual) and the maintenance carried out by ONEE technicians, we see that the main causes of failures are as follows:
Difficulties in obtaining spare parts: These difficulties are linked to insufficient maintenance or maintenance budgets, sometimes with the lack of after-sales services. These situations lead to a shortage of stocks.
Lack of training for maintenance technicians and users of wind turbine equipment: According to our findings, there is no training plan for maintenance technicians and equipment users. For maintenance technicians, training is limited to their participation in seminars and workshops devoted to maintenance management. User training is very often limited to that carried out by the supplier of the equipment during its installation, which leads to frequent equipment breakdowns resulting from handling errors. For this we note that training is essential for a good functionality of the equipment.
Exceeding the duration and neglect of some preventive maintenance tasks: Some technicians do not respect the preventive maintenance schedule of an equipment, they are not serious about the duration of each action, this leads to an increase in the probability appearance of a failure.
Lack of maintenance monitoring. There is no monitoring to verify the maintenance carried out (no maintenance report carried out).
The cost – For a breakdown outside the contract, additional cost for ONEE – Time between response and intervention: increase in unavailability and costs.
All that remains is to set a threshold beyond which we must implement preventive actions to reduce the risk. For this we will use the method.
To reduce criticality, it is necessary to reduce the three factors which are frequency, severity and non-detection by the following methods:
a. Reduce the frequency through preventive maintenance.
b. Reduce the severity by good preparation before the intervention across ranges and preparation of resources.
c. Reduce non-detectability by planning checks with the appropriate measurement resources (oil analysis, vibration analysis, etc.)
We will then focus on the actions to be taken to reduce the reappearance of failures, in addition to a presentation of some technical solutions to remedy the failures.
Recommendations
We make recommendations from Table 15 that will help improve maintenance service. For the maintenance of the multiplier pump element, the following must be done:
Recommendations that will help improve maintenance service.
Proposed solutions and improvement
Based on the current critical maintenance analysis, we note that this is a progressive failure requiring a maintenance check. The failure study made it possible to develop a maintenance monitoring manual to match the reliability between the maintenance performed by the technicians and the manufacturer’s schedule, as well as some suggestions for technical solutions. The process for producing the maintenance monitoring manual was based on the previous analyzes, with the use of the failure history since 2010–2019. This part aims to implement the results of the FMECA studies. We will present technical solutions to these failures Table 16.
Actions to be taken to reduce the recurrence of failures.
Conclusion
Maintenance is the critical factor affecting the life of wind turbines. Proper repairs are also essential to the reliability and longevity of the turbine generator. Once the actions are in place, the criticality is recalculated. All these actions therefore make it possible to reduce the frequency of breakdowns while optimizing the frequency of preventive interventions. At the end of this study, we can come out with the following recommendations:
The systematic maintenance instructions must be observed, such as the replacement of defective parts at the intervals recommended by the manufacturer.
Train maintenance technicians on the equipment to facilitate the detection of anomalies.
Train maintenance service personnel at AMDEC.
Maintain a safety stock of essential spare parts.
Footnotes
Appendix
Acknowledgements
We thank Mr. LAARAJ Abderrahim of ONEE (National Office of Electricity and potable water) and responsible of the Amougdoul wind farm, for kindly providingus with the data necessary to carry out this work.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
