Electricity production losses in wind farms: expectations vs. reality
20 October 2022Article by José Carlos Matos, director of the Wind Energy area, and Filipa Magalhães, senior consultant at INEGI
The energy potential of a wind farm is not always the same on paper and in reality. During the design phase, studies and measurement campaigns of meteorological characteristics are carried out to assess the wind resource, define the best location for the wind turbines and select the most appropriate technology for the wind regime. After all, the estimates of annual electricity production are a critical aspect for the decision making on the execution of the investment.
However, during the operation of wind turbines, the effective production of electricity does not always correspond to the operator's initial expectations. Why?
In general, potential deviations arise from the following aspects:
- On the estimation of electricity production during the project phase;
- Temporal variability of the wind resource;
- Insufficient technology performance.
As regards the 1st and 2nd reasons, it is appropriate, after a few years of operation, to reassess the estimates accompanied by a comparison with the actual production data of the wind farm, thus adjusting the value of the estimate to a more realist. This is a useful exercise not only insofar as it provides a more effective picture of the park's capacity, both in past and future terms, but also insofar as the new estimate is accompanied by a reduction in the uncertainty associated with it. .
Figure 1 - Productivity of a wind farm (fictitious example)
Regarding the 3rd aspect, the importance of continuous monitoring and evaluation of the performance of the wind turbines during their operation, in view of the observed wind conditions, with the purpose of preventing unnecessary stops, minimizing losses, and maximizing their production according to the available resource.
This type of monitoring allows, for example, to detect cases of misalignment of the wind turbines in relation to the wind direction, warning signs of the wear of a component, among other situations. With a more detailed analysis, it is possible to categorize the wind turbine stops according to their cause and assess the associated production loss. This information will support the operator's decision-making in terms of the operation and maintenance of its asset.
The origin of deviations
The causes that may be at the origin of production losses in a wind turbine can be divided into three main groups:
a) external origin to the wind farm;
b) origin internal to the wind farm, but external to the wind turbine;
c) causes intrinsic to the wind turbine.
The first group includes exceptional situations, including the unavailability of the electrical network, due, for example, to incidents in medium or high voltage overhead lines, breakdowns in the wind farm substation or even damage to the connection cables. from the park to the substation. Stops requested by the operator or weather conditions (parameters outside the normal operating range of the wind turbine) can also be included in this group of external factors. These events can result in the total shutdown of the wind turbines for a certain period of time or, in the conditioned production, through the imposition of a power limitation of delivery to the electricity grid by the wind farm as a whole, or the imposition of a power limitation to one or more wind turbines in particular.
In other cases, the impositions of the point of connection to the electricity grid and the legal framework may limit the maximum power injection value, so that some wind farms will have to operate with a limitation in relation to the installed power. As a condition applied to the wind farm, it will not impact the normal operation of the wind turbine, being in this case considered as an external factor.
The last group mentioned will have direct implications for the operation of the wind turbine and encompasses all events that affect its production, such as failures and problems arising from the various control and safety systems of the wind turbine and its components.
Normative approaches
In the past, the performance of a wind turbine was seen by the sector from a perspective of availability based on operating time, contractually guaranteed by the technology providers. However, this assessment becomes reductive in the sense that it is based solely on whether the wind turbine is able to generate energy or not, ignoring whether, when in operation, it is effectively generating what is expected, taking into account the observed wind regime.
With the publication of the IEC 61400-26 series of standards, updated in 2019 to its latest version through the IEC 61400-26-1 standard, models for calculating availability indicators based on time and production were defined. Applying the principles of this standard, INEGI implemented internal methodologies for the evaluation of wind assets, using performance indicators of a wind turbine taking into account its electricity production potential.
This upgrade in the availability assessment, now in terms of production, allows for a more realistic notion of the asset's operation. After categorizing the events, an estimated energy loss is allocated to periods of anomalous or conditioned operation.
The choice of methodology for calculating the estimates may depend, among other factors, on the source of information available:
- only data recorded at a meteorological station or stations installed on the site of the wind farm,
- only operational data of the wind turbines, namely those that are registered in the Supervisory Control and Data Acquisition systems, commonly known as SCADA systems,
- both sources mentioned above or,
- atmospheric reanalysis data.
It should be noted that the methodology that uses sources 2) or 3) suggests that the wind farm was not totally inoperable, since there are records available in the SCADA system. In the case of the situation referred to in 1), it is assumed that the park will have been completely inoperable, since the SCADA system records are not available. In this case, data collected from meteorological stations, preferably located in the vicinity of the park, is used. The source of information referred to in 4) will be of last resort when none of the other sources are available.
All the methodologies mentioned above assume the existence of a history of operation of the wind farm of at least 1 year, which will allow establishing a relationship between the production of wind turbines and the observed wind conditions, as exemplified in Figure 2.
Figure 2 – Relationship between wind farm production and wind speed (fictitious example)
This relationship will result in an estimate of the production that was not generated by the wind turbine during the period considered and that will serve as a decisive input to different situations, such as the activation of insurance by the park promoter, the optimization of the operation of the assets by the operators or simply monitoring and monitoring the performance of wind turbines throughout their operation.
INEGI's experience
INEGI has been involved, since 2009, in evaluating the performance of more than 800 MW of installed power (national and international), using operational data recorded in SCADA systems. In 2014, the Institute presented to the market a platform for monitoring the performance of wind turbines – Wind-to Power (W2P), developed with internal resources.
This tool allows, in an expeditious way, to scrutinize the operational data of each wind turbine, associating different categories at each moment, depending on the wind turbine's operating mode. This categorization includes not only the events that refer to the contractual unavailability of the wind turbine (recorded through errors in the SCADA system), but also to other causes that somehow affect the operation of the wind turbine, such as the detection of ice or power limitations that are not imposed by the network. This analysis results in a direct comparison between the number of occurrences of an event, its duration and the estimated production loss associated with that event.
It should be noted that, depending on the time when the event occurred (period with greater or lesser wind resources available), its duration may have a greater or lesser impact on the estimated loss of production, as can be seen in the example in Figure 3. More Once again, the importance of evaluating the availability of the wind turbine in terms of production and not just availability of time is highlighted here.
Figure 3 - Occurrences, duration and estimated production lost (fictitious example)
For wind farms with a significant number of wind turbines, this type of analysis becomes useful from an operation management point of view, as it allows one to identify the wind turbine that most contributes to the loss of revenue. Through a more detailed analysis, it is even possible to identify the possible cause of the anomalous operation of the wind turbine.
As exemplified in Figure 4, the categorization of events during the wind turbine operation allows the estimation of the energy that the wind turbine would be expected to produce, for a certain period, based on an average power curve considering only the normal operation data (curve of blue power) and the observed frequency histogram. Likewise, it is possible to calculate the energy produced by the same wind turbine, based on the average power curve, considering, in this case, all the operating data (green power curve), for the same period and the frequency histogram. The ratio between expected and produced energy results in a performance indicator of the wind turbine, depending on its production, for the period under review.
As exemplified in Figure 4, the categorization of events during the wind turbine operation allows the estimation of the energy that the wind turbine would be expected to produce, for a certain period, based on an average power curve considering only the normal operation data (curve of blue power) and the observed frequency histogram. Likewise, it is possible to calculate the energy produced by the same wind turbine, based on the average power curve, considering, in this case, all the operating data (green power curve), for the same period and the frequency histogram. The ratio between expected and produced energy results in a performance indicator of the wind turbine, depending on its production, for the period under review.
Figure 4 - Power curves (fictitious example)
It is therefore considered that a complete analysis is based on a complementarity between availability based on the operating time of the wind turbine and indicators that are based on availability in terms of production and the capacity factor.
Related Pages
Consulting | Wind Energy
References
1 IEC 61400-26-1, First edition 2019-05, Wind turbines – Part 26-1 Availability for wind energy generating systems. This first edition cancels and replaces the IEC/TS 61400-26-1:2011, IEC/TS 61400-26-2:2014 and IEC/TS 61400-26-3:2016.
2 Atmospheric reanalysis series are a temporal and geographic description of the climate, produced by combining models with observations, containing estimates of quantities for various meteorological parameters.