Wind Turbines allow us to extract the energy which is present in the wind and depending on how efficient the Wind Turbine, this allows us to make the best use of the cleanest source of energy. This form of generation produces no environmentally damaging C02 emissions which harms the Earth's outer atmosphere. Wind Energy is converted into Wind Power through the help of Wind Turbines that converts the Kinetic Energy in the Wind into Mechanical Energy which can then be turned into Power.
Our main concern is the actual Wind Turbine because if the Wind Turbine is performing at almost 100% efficiency, this will then allow us to capitalize on the Wind Energy and have the best Power co-efficient available. The efficiency of a Wind Turbine is related to its Performance, which in a way can be also related to the Turbine Aerodynamics.
Included in the report are the different factors that can affect the performance of a given Wind Turbine and how its power curve can be distorted due to this. The performances of wind turbines are measured using SCADA data which is Real Time data which can be past or present. Also outlined in the report is how we take the Power Curve that is formed by SCADA data and distort it in a way in which it has been influenced by Turbulence. The objective of this is to allow us to monitor how this will be plotted in the Performance Metric space to give us an idea how they way in which Turbulence is plotted on the Metric space to help us for future referencing.
Aims & Objectives
The overall aim of this project is to observe the different ways in which a Power Curve of a Wind Turbine can be distorted. Then once this is done for each of the different factors we have to devise a simple way in which we can look at any distorted power curve and then determine what factor actually caused the distortion (underperformance).
The fist key aim of this project is to review the necessary literature related to the Performance of Wind Turbines to allow us to gain a basic insight into the main factors at present that are responsible for the distortion of Wind Turbine Power Curves
The second key aim is to take a perfectly healthy Power Curve that has been plotted using SCADA data and distort it in the same way it would be distorted if for instance Turbulence was present, this should be done for the majority of the factors that have been reviewed in the literature. After doing this and producing a distorted Power Curve, this can be viewed in the SgurrTrend Software to show where it lies on the Performance Metric plot.
The overall objective of this project is to observe the points that are out-of-sequence on a metric space plot and then without having to look in detail at the different Power Curves for each point, be able to identify what factor caused the underperformance. If we are able to do this, then it would make it much easier to identify what factor caused the distortion therefore causing the point to be out of sequence. This would make it more efficient when analyzing SCADA data in the future.
The last objective to look at is drawing up conclusions regarding the use of the SgurrTrend software and even perhaps further development of the software to make it more efficient in observing the Performance Metrics of Wind Turbine Power Curves.
Introduction to Literature Review
This section basically reviews the literature that is relevant and enables us to have a better understanding of the topics that are related to this project. By reading through this literature, it has allowed me get a better insight and understanding of relevant issues that surround this topic.
As well as looking at the common factors that normally cause the distortion of a Power Curve, text regarding aerodynamics of a wind turbine has been viewed as there can be problems that occur in the physical parts of the Wind Turbine such as the Yaw or Rotor.
Aerodynamics of a Wind Turbine
The Aerodynamics of a Wind Turbine need to be taken into consideration as there are cases in which the Power Performance has been affected as there have faults in the Turbine components.
Most modern Wind Turbines are measured on their performance by their Power Coefficient which is Cp and no Wind Turbine can extract more than 59% Energy which is carried by the wind. The maximum theoretical Power Co-efficient exists and is calculated as follows:-
The maximum theoretical Power Co-efficient is also denoted by the Betz Limit. Most efficient modern wind turbines operate substantially below this limit close to 0.5 at which they have optimized performance.
The two different types of Modern Wind Turbines are:-
- Vertical Axis Wind Turbine (VAWT)
- Horizontal Axis Wind Turbine (HAWT)
For the Vertical Axis Wind Turbine (VAWT) the blades of the Turbine are connected to a vertical shaft, whereas for the Horizontal Axis Wind Turbine they are connected to a horizontal shaft. The difference between them is that a HAWT has higher Power efficiencies than VAWT's. However, the wind direction does not affect the a VAWT and this saves time as there is no need for it to adjust to the Wind direction at any given time, thus saving time and also power. In conditions where there is Turbulence present and other factors that can have an effect on the Power Curve, a Vertical Axis Wind Turbine is the better even though it has a lower efficiency.
Loading on Wind Turbines
There are three important sources that are responsible for the loading on Wind Turbines. These can affect the components of the Wind Turbines such as the Blades, Rotor and the Rotor Shaft.
These sources are shown below:-
This is caused by the Earth's gravitational field which causes a sinusoidal loading effect on each blade. When this sinusoidal loading is applied to each blade, there is also a frequency which is equal to the rotation of the rotor. This form of loading causes different stresses on the blades whilst they are rotating.
This type of loading occurs when the turbine changes speed, either accelerating or decelerating.
Inertia loading is applied when for instance the rotor is braking and therefore a torque () is applied at the rotor shaft. Whilst this occurs a small section of the blade will happen to feel a force (dF) in the direction of rotation. The force dF can be calculated as follows:-
This loading is caused by the flow that passes through the actual structure of the Wind Turbine which are the blades and the tower. For this type of loading, turbulence can play a part and this causes the real wind field to not be steady. The wind field is characterized by a Mean Wind Speed which varies with the Turbine height (z) above the ground. The equation below allows us to calculate the average wind speed over a ten-minute period in regards to the Turbine Height (z) and the Surface Roughness (Zo)
Depending on the different factors or conditions that are present, this can cause the surface roughness to be changed and therefore the average wind speed over a 10 minute period. The different factors or conditions can cause variations in the wind speed.
Different monitoring techniques
There are two different monitoring techniques that can be used to assess Power Performance of a Wind Turbine and these are Condition Monitoring and Performance Monitoring.
Condition Monitoring is when you assess the Wind Turbines of a Wind Farm by bringing in additional instrumentation, this allows you to monitor the Turbines mechanical components and predict any failures that can occur. By carrying out condition monitoring it is basically allows early detection of any breakdown or mechanical failure. Condition Monitoring can be costly as you require to bring in extra instruments that are need to monitor the Turbines but if there is failure is detected in the system this can reduce or avoid downtime and lost power generation.
This is the other way in which the Power Performance of a Wind Turbine can be measured and this requires no extra instrumentation. This means that there is no extra cost associated with the actual monitoring of the performance. The way in which Performance Monitoring is carried out is that it takes past real time data that has been gathered from Wind Turbines which is known as SCADA data and then analyses it.
Performance Monitoring basically uses a specifically designed software tools and previous data to analyze many different factors of a Wind Turbine. This is important for our project as by through Performance Monitoring we can identify the points plotted on the metric space that are not in sequence and therefore are underperforming and require further analysis.
The way in which Performance Monitoring is done is that it provides a snapshot of a Wind Turbine for a given time and then allows you to compare it another snapshot of the same Wind Turbine but in a different timeframe. By doing this you can monitor the behavior changes of the Wind Turbine and can also the power curve to give a better insight to what caused the changes and how to optimize performance. A Performance Monitoring software can allow you to do the following compare the warranted Power Curve with it actual measured Power Curve, to view the Power Curve of the other Wind Turbines on the same Wind Farms and lastly to view the Power Curve of the same Wind Turbine but in different times.
The Performance of the Wind Turbine is actually determined not by the actual Power Produced but by the "Yield Deficit, this is the gap between the benchmark and the actual operational performance in real time. Our task would to figure out what caused the Yield Deficit and fix this problem to Optimize Performance of the Wind Turbine.
The overall effect of Performance Monitoring is that it can automatically find relationships between different data types and since everything is done by analysis of data, it can save time since everything is automated. By any chance, if this leads to the slightest of Performance increases then it would save money in the long run.
The Power Curve of a Wind Turbine is a graphical plot that shows the exponential incline of how the Electrical Power Output varies with increasing Wind Speed. A warranted Power Curve is one which shows the Power Output for varying Wind Speeds when the Wind Turbine has Optimized Performance and this can also act as a Reference plot to compare to if the Wind Turbine is not performing as it should.
A warranted Power Curve does not show the effects that can distort a Power Curve such as Turbulence, Wind Shear, Wind Veer or Blade problems. The graph in Figure 1 shows a perfectly healthy Power Curve. Also, shown below are the different factors that can cause distort a Power Curve and therefore reducing its performance.
Factors affecting the Power Curve
Now, the factors that can affect the Power Curve of a Wind Turbine and also its performance will be discussed. Basically, there are several common factors that usually occur due to weather conditions. There is no way at present to overcome them unless you construct a Wind Turbine in an environment where you are sure not to be affected by the conditions. Instead, we have to devise a way in which they are easy to identify and can save less time in figuring out what effect caused what distortion on the Power Curve. The most common factors have been highlighted below and this is important to us as it allows us to distort the Power Curve ourselves to figure how it is plotted in the Metric Space of Performance.
The Turbulence effect on a Power Curve is the most common type that there and if there is Turbulence present at low wind speeds this causes a higher Power Output. However, if there is Turbulence present at high wind speeds, this causes the Power Output to be decreased.
This shows us that Turbulence cannot be bad factor if does increase the Power Output in some way but at higher wind speeds there is more Power Generated meaning that more power will also be lost due to Turbulence and this is why it is considered not to be a good factor.
The equation shown in (5) is one that can applied a zero turbulence Power Curve like the one shown in Figure 1 and distort it for different Turbulence Intensities. This can be done to notice the effect Turbulence has on a Power Curve and therefore on its plot in the Performance Metric space.
The equation above allows us to alter a perfectly healthy Power Curve for different Turbulence Intensities to our preference. The way in which this is done is that the equation shown in (5) is re-arranged to calculating average P(v) as this is the average Power Output that needs to be calculated for the different Wind Speeds so that we can plot a Power Curve showing the effect of Turbulence
The equations that have been mentioned regarding Turbulence Intensity have been used to show the effect of Turbulence on the Power Curve, the results for these are shown in Section 3.
Through the analysis of data it has been shown that high wind shear regardless to whether it is positive or negative, reduces the overall annual energy wind production. When looking at Wind shear the following parameters need to be taken into account the hub height and Wind Speed, the hub height over Rotor Radius (H/R) and the wind shear exponent (a). If these parameters are known then the Equivalent Wind Speed can be calculated as follows:-
Where, the limits are -1 and 1 for the integration.
By using the above equation (7), this allows us to calculate the Power output for each Wind speed as long as we know the wind shear exponential co-efficient. The Wind shear is associated with the hub height and the rotor radius.
Wind Veer is the vertical wind direction change and has no influence on the Power and Wind Speed measurement directly. This is able to change depending on the weather circumstances and is based upon the horizontal wind speed. Wind Veer can reduce Power Flow generally by the rotor having Skew Airflow and is usually caused by an error in the Yaw of the Turbine.
The veer that is present above the hub is positive and the veer that is present below the hub is negative. This influence can be expressed including the parameters hub height (H), rotor radius (R) and rotor angle (d), by simplification the equation below shows the relationship between the Power Output and Power if there is Wind Veer present:-
Distorting Power Curves
This section of the report basically shows how we take a perfectly healthy Power Curve and distort it depending on the influences as mentioned in Section 2. Till now we have been able to distort the Power Curve if it has been affected by Turbulence and how this has been done is shown below.
The way in which we distorted Power Curve due to the influence of Turbulence, is that values were selected for the different Turbulence Intensities as instead of plotting the one graph, many were plotted to show how if the Turbulence was increased or decreased, the position it would lie on the Performance Metric Space.
The Turbulence Intensity would increase in steps of 0.01starting at 0.05 all the way up to 0.18. This basically allows us to have a plot of 14 turbulence effected Power Curves on the one graph and therefore can be compared. Each Turbulence Intensity is represented by a different Wind Turbine in the Metric Space plot and this is how the data is arranged. Since, the turbulence intensity that we want to vary the Power Curve is known, the data can be entered in to the Turbulence Normalization equation (6) shown in Section 2 to allow us to calculate the average Wind Power Output for each different Wind Speed up to 25 m/s. One point that should be noted is that we should be actually using a reference Power Curve that has zero turbulence but in this case, the reference Power Curve itself has a Turbulence Intensity of 0.1 which is our reference. This does not in any way affect our final results but in future this scenario can be looked into but for zero turbulence Power Curve. The reference Power Curve has been shown in Figure 2 which is in Appendix B.
All the calculations have been done using Microsoft Excel by entering the Turbulence Normalization formula for each of the Turbulence Intensities and by using the values from the actual Reference Power Curve like Power Output for each Wind speed. By calculating the second derivative of the of the Reference Power Curve and each of the different variables, this allows us to draw up a Graph that includes all the separate varying Turbulence Intensity Plots on the one graph as shown in Figure 3. There are fourteen different plots for the fourteen different Turbulence Intensities and each is represented by a Wind Turbine.
The Performance of these Wind Turbines with different Turbulence Intensities are now monitored using a Performance Monitoring which is the SgurrTrend Software provided by SgurrEnergy. This software allows us to plot each Turbine depending on its Performance Metrics on the Metric space. Before this could be done, the software requires certain variables so that it can identify the parameters of each Turbine and plot the Performance Metric.