Wind Power System for Isolated Applications - Research Paper Example

Wind power system for Isolated Application Introduction Because of increasing public awareness of global warming, fast depletion of existing non-renewable energy sources, and the introduction of Kyoto Protocol, researchers are seeking for new alternative energy sources that are sustainable and renewable. Among the possible energy sources, wind energy is one of the most promising and feasible sources in some places. (Mutschler, 2002, 11). Wind energy, as a renewable energy source, has the potential of being cost effective and has a high capacity factor. Currently, most commercially operating wind turbines have horizontal axis designs with variable speed electric power generation. Variable speed wind power generation is superior to fixed speed generation in that, when compared to the latter, it has a high energy yield (Goodfellow, 2004, 250) allows good and robust control of power generation, and has lower power and torque pulsations. A typical configuration of a wind energy generation system is shown below. Configuration of Wind Energy Generation System Naturally, the blades interact with wind and convert wind energy into mechanical energy at the shaft. The wind turbine, which includes couplings, dampers, and gear boxes, conveys the mechanical energy harvested at the blade shaft to the shaft of the generator system. (Heier, 2006, 2) The generator system, with its generator and power electronics, performs electro-mechanical energy conversion to generate electrical power to the power grid. Typical wind turbine characteristics are shown below. Wind Turbine Characteristics The turbine speed has to vary, according to the wind speed, to maximize the energy being harvested. If the wind speed is higher than the rated wind speed, the turbine speed has to be decreased to limit the power output to its rated value. As controlling the blade pitch angle can only limit but not to maximize the power output, a good design to the control of the generator system becomes essential for variable speed systems so as to maximize as well as to limit the power generation There are several types of generators for variable speed wind power generation, which are, namely, cage induction generators, synchronous generators, wound rotor induction generators with variable rotor resistances, doubly fed induction generators (DFIGs), brushless doubly fed induction generators, and switched reluctance generators. (Heier, 2006, 01) Among them, cage induction generators, synchronous generators and switched reluctance generators require power electronic converters having the same ratings as that of generators, because all generated electric power has to be processed by the converters. The large ratings of these converters, apart from being very costly, will lead to large size and hence overall heavy generation units. In contrast, the ratings of the power electronic converters of DFIGs and brushless doubly fed induction generators are only fractions of the total generator output power, hence these controllers are cheaper, simpler and more reliable. For wound rotor induction generators with variable rotor resistances, slip rings are unnecessary, however, the feasible operating speed range is limited to 100% to 110% of the synchronous speed of the generators. They also suffer from having poor control of real and reactive power. The rotor power in wound rotor induction generators with variable rotor resistances is dissipated as heat, instead of being recovered as electrical energy in DFIGs. In the wind power industry, brushless doubly fed induction generators are however not popular because they are conceptually new to the wind turbine manufacturers. On the other hand, DFIGs are considered to be most suitable for isolated locations and have the largest market share and highest annual growth rate in the wind power industry because of their low converter power ratings, high energy output, good utilization of the generators and features that allow independent control of real and reactive output power over a wide speed range. Optimum Design Procedure In order to design a wind power system for isolated application following steps will be taken. First, a feasibility study will be developed. Once the feasibility of such a device is verified, an optimum design must be found. Due to the range of conditions, an optimization model must be built to predict how a rotor will behave under the specified circumstances. A shroud must also be designed for the application. Once an optimum design or designs has been found, the results should be compared to any similar studies and experiments that have been performed. Each of the steps will be covered in the following sections. Feasibility Since there is such a wide range of different conditions, no one apparent windspeed can selected. A wind velocity range from 50 – 500 ft/min will be used for the design range, which fits the MSHA requirements. This range will cover the conditions found in most underground mines. As a standard, unless otherwise specified, the generalized mine entry will have dimensions of 8 ft tall by 20 ft wide. The power required to charge the radio batteries is approximately 2 watts. Based on generator and electrical losses and a safety factor, the power out of the wind turbine rotor must be about 4 watts. Power out of an unshrouded wind turbine rotor is related to the swept area of the rotor, windspeed, density and the coefficient of performance of the rotor. Eq. 1 displays the power out of a turbine rotor, Equation 1 Density will be taken at standard sea level conditions, 1.225 kg/m3. For an ideal rotor which has the maximum possible coefficient of power of 0.593, at a windspeed of 350 ft/min (1.778 m/s), a rotor diameter of about 5.15 ft is required to produce 4 Watts of power. This windspeed, chosen for this calculation, is in the middle of the velocity range. Realistically, rotors with Cp’s as high as this do not currently exist. The figures below display a first order look at the relationship between the windspeed, power and rotor size. Figure 1 Power curves for varying rotor diameters and performance coefficients. From Figure 1 it can be seen that for a realistic rotor (Cp = 0.46) at the high end of the velocity range a rotor diameter of about 3.5 ft is needed to produce 4 watts. Also, it can be seen that a rotor diameter of 5.25 ft will produce 4 watts at about 375 ft/min. For an 8ft by 20 ft mine entry way, a 5 ft diameter rotor is feasible. The rotor will fit in the contained area. It can also be seen from the plot that the low end of the velocity range is not productive at all. Even for rotors of much larger size, the low end of the windspeed range will still not produce power sufficient enough to charge the radio batteries. Figure 2 emphasizes the fact that an unshrouded rotor cannot supply the required power for the application in the lower one third of the windspeed range. Figure 2 Power curve for a rotor diameter of 20 ft Figure 2 displays the power results for a rotor much larger than the rotors previously discussed. The data is only reported to show that at the low end of the air velocity range (less than 150 ft/min), it is impossible (with a traditional enshrouded turbine as large as the mine entry) to produce the power required. This means that if air speeds in a particular mine were at the low end of the range, a wind turbine system is not a feasible solution to charge the radio batteries (unless future electrical technology enables the batteries to be charged with less power). However, since the power out is directly related to the cube of the windspeed, as the air speed in the mine increases, so does the feasibility of such a solution. And as stated, the upper end of the wind range should provide ample speed to build a sufficiently small but efficient wind turbine. With the added power augmentation of a diffuser shroud (not included in Figure’s 1 or 2 ), a wind turbine should be able to meet the needs presented by the issue. Optimization Model The first order study shows that this idea can work on a portion of the wind conditions in the study. A numerical model was developed to quickly and easily output the geometry and design parameters of an optimum wind turbine rotor based upon inputs related to the mine (mainly size and airspeed). Matlab was used as a calculation/design tool. The computer codes for the model can be seen in Appendix C. Blade element momentum theory was utilized in the model. It was chosen due to its proven performance qualities in previous relevant research . The method and procedure used in developing the numerical model are discussed in detail in the following section. Due to the complex nature of rotor blade aerodynamics, it was decided that the first model would simply deal with an unshrouded (thus unaugmented turbine). The diffuser shroud will be added in as an augmenter at a later stage in the research/design process. (Nicolas, 2005, 3317) Blade element momentum theory is a combination of general momentum theory and simple strip theory. Strip theory utilizes the concept of a blade element or “strip”. Each rotor blade is split into span-wise segments. Each element covers a small radial distance on the blade, dr. Figure 3 displays the forces and angles associated with a 2- D cross-section of an element of a common rotor blade. Figure 3 Flow on the cross section of a rotor blade element. The forces on each strip can vary and each contributes to the total force on the rotor blades. As air flows through the turbine, the rotor interrupts the normal flow stream. Blade element momentum methods calculate axial and tangential interference factors that are used in defining this interaction. In Figure 3 these factors are seen as a (axial) and ’ (tangential). These interference factors are very influential in the calculations of the aerodynamics of the rotor. (Williamson, 2005, 115) The angle between the plane of blade travel (rotation) and the resultant wind velocity is φ. The angle of attack, α, is the difference between the relative in-flow angle, φ ,of the wind and the blade twist (pitch) angle, β. If the airfoil that is to be used for the blades is known, then the lift coefficient Cl and the drag coefficient Cd can be found from given airfoil data. This airfoil data comes from testing at a specified Reynolds number. Cl is found to stay fairly constant for different Reynolds numbers. However, Cd changes if the Reynolds number changes. Eq. 2 shows how to update Cd with changing Reynolds number. Cd will have to be updated for each case where chord length changes since Reynolds number is dependent on chord length. Equation. 2 The resultant thrust, dT, and the resultant torque, dQ,(not shown in Figure 3 but acting equal and opposite to the tangential resultant velocity term) are defined by Eq.3 and Eq. 4 respectively below: Equation 3 Equation 4 where q is the dynamic pressure, B is the number of blades, c is the chord length of the element and dr is the radial distance covered by the element. The coefficients for these resultant forces CdT and CdQ can be calculated using Eq.’s 5 and 6: Equation 5 Equation. 6 Momentum theory then looks at a columnar annulus of air that effects the same element discussed above. New relationships can then be developed for the elemental thrust and torque that involve the axial and tangential interference factors. These equations are displayed below as: Equation. 7 Equation 8 The F term in the above equations is called the Prandtl tip/hub loss factor. This factor accounts for losses from airflow effects at the hub and tips of the rotor blades. The tip and the hub both have separate but similar functions. However, hub losses are often neglected in describing rotors with small hubs (Hansen, 2003). The Prandtl tip loss factor is defined in Eq. 9 as, Equation 9 It should be noted that when the diffuser shroud is added, the tip losses will become very small and can be neglected or (F = 1). Since the design constraints require the turbine being designed to be protected by a shroud, the tip losses will not be included in the design calculations. By equating Eq’s 7 and 9 and Eq’s 3 and 4, the interference factors can now be solved as; Equation 10 and Equation 11 where r σ is the local solidity of the rotor blade for a particular element. Eq. 12 defines the local solidity. The r variable in Eq. 9 and 12 is the defining radius of the element or the radius from the center of the rotor to the center of the element and Equation. 12 In most research applications an iterative solution procedure is used. Initial values for a and are guessed, ' a φ is calculated based on these values, α is also calculated, then the interference factors and are updated until there is convergence. The referenced method will produce a number of data curves that will display the performance of different rotor setups over the varying range of constraints. This will allow the designer to select the optimal setup for use in specific conditions. This method uses the same equations but, rather than initially guessing the interference factors, the flow angle φ is assumed and used as the starting point for calculations. From this point, the torque and thrust coefficients are calculated from Eq.’s 4 and 5. The local solidity at each element is determined by Eq. 11. Eq.’s 9 and 10 are then used to calculate the interference factors. These interference factors are then used to calculate an elemental speed ratio, r λ, defined as, Equation 13 Once the local speed is known, the elemental power coefficient can be calculated as, Eq. 14 This term can then be integrated over the entire blade to determine the coefficient of power of the entire rotor. A very close approximation can be obtained by taking the mean of the local Cp’s found from Eq. 14. However, for this to be accurate each elemental annulus must have the same area. Eq. 15 shows how to calculate element divisions so that the annular areas will be equal to one another. n is the number of the element for which the division is being calculated and N is the total number of elements in the blade or, Eq. 15 From this, the defining radius of each element is taken halfway between the element division for each particular element. Once the power coefficient is found by taking the mean of the local power coefficients, the power output of the rotor is easily obtained by using Eq. 1. This calculation method was implemented using MATLAB. Each inflow angle will produce a different power output (separate solution) for a specified rotor size, chord and blade number. The calculation process is iterated to solve each inflow angle for a varying number of blades and varying chord lengths. The size of the rotor is also varied to determine the smallest possible design configuration that will produce the required output. Results from running these calculations (for a single inflow angle) will give design configurations for constant pitch, constant chord rotor blades. The results then simply have to be searched for the designs that produce the largest power outputs. In this case, a certain inflow angle (combined with blade number and chord length) will provide an optimum design. It is not certain which blade number or chord length will produce the best design but this study makes it possible to analyze the underlying implications of these variables combined with the varying inflow angles. Each element has a different optimum inflow angle based on the flow properties through the rotor. This means that in order to achieve a true optimum rotor design the rotor blades will have to have a twist (pitch) distribution based on these optimum inflow angles. In order to determine the optimum twist distribution, the iteration results can be searched numerically to find the inflow angles that provide the highest local (elemental) power coefficient. (Goodfellow, 2004, 220)These values are saved within the model and can then be plotted against the blade station number to show the optimum twist (pitch) distribution. The same process can be used to determine the optimum chord (taper) distribution. These configurations, since they have a higher overall power coefficient, will produce more power than the constant pitch configurations found by just a single inflow angle. There will be an optimum configuration for each blade number simulated. The inflow angles will be specified based upon which angle provides the best coefficient of performance. This means that the only other variable that must be selected is the chord length. The model will return the number of blades and the chord length for each optimum case. These values determine the solidity of the rotor. References Ackermann, T. (2008) 'Wind power in power systems' John Wiley and Sons, pg 1 Chen, H., Mang, C, and Zhao, X.: (2005) 'Research on the switched reluctance wind generator system'. 2001 IEEE International Conference on Systems, Man, and Cybernetics, Arizona, USA ,pp. 1936-1941 Goodfellow, D., Smith, G.A., and Gardner, G.(2004) 'Control strategies for variable-speed wind energy recovery'. Proc. of 8th British Wind Energy Association Wind Energy Conference, Cambridge, England, pp. 219-228 Hansen, A.D., Jauch, C, S0rensen, P., Iov F., and Blaabjerg F.(2003): 'Dynamic wind turbine models in power system simulation tool DIgSILENT' Ris0 National Laboratory. Heier, S. (2006)'Grid integration of wind energy conversion systems' (John Wiley and Sons, pg 1-3 Hulle, F.V.(2005) 'Large scale integration of wind energy in the European supply analysis, issues and recommendations' European Wind Energy Association. Pg 1-5 Mutschler, P., and Hoffmann, R.(2002) 'Comparison of wind turbines regarding their energy generation'. IEEE 33rd Annual Power Electronics Specialists Conference, Queensland, Australia, pp. 6-11 Nicolas, C.V., Blazquez, F., Ramirez, D., Lafoz, M., and Iglesias, J.(2005) 'Guidelines for the design and control of electrical generator systems for new grid connected wind turbine generators'. Proc. 28th Annual Conference of the IEEE Industrial Electronics Society, Sevilla, Spain, pp. 3317-3325 Williamson, S., Ferreira, A.C., and Wallace, A.K. (2005) 'Generalised theory of the brushless doubly-fed machine. Part I. Analysis', IEE Proc. Electric Power Applications, pp. 111-122 Read More