Reliability Assessment of Wind-Diesel Hybrid System

Transcription

1 Reliability Assessment of Wind-Diesel Hybrid System S. Kishore Kumar, Abha Rajoria, E.Fernandez Abstract Wind-Diesel Hybrid Power Systems are of interest to remote areas for power needs. They are preferred in remote area power systems as they have a reduced fuel consumption which leads to lower cost and lesser green house emissions. Besides, the hybrid combination needs lesser fuel storage requirements than a system consisting of a only conventional diesel generator sets. However the intermittent characteristics of wind energy seriously affect the reliability of wind diesel hybrid systems. Hence it is a matter of research interest to determine the adequacy levels for such schemes. In this paper an attempt has been made to evaluate Energy Index of Reliability (EIR) which is express as a function of Expected Energy Not Supplied () using Wind-Diesel hybrid System consisting of one or more wind and diesel genset units. This will help in arriving at the most suitable schemes having expectable level of adequacy. Index Terms Reliability evaluation, Wind diesel system, Hybrid generation Energy Index of Reliability (EIR), Expected Energy not supplied () designed wind speed prediction model in the Monte Carlo simulation can lead to very reliable results coming from large amount calculation time on sequential simulation. In [6] the authors demonstrate a model of constrained wind generation output below a certain ratio of system load considering conventional generation and load at the same time. In the time frame approach [7], discrete wind speeds are applied to wind turbine power output function with the probabilities of discrete speeds based on the Weibull distribution to model wind turbine generator (WTG). A short term system reliability index can then be obtained by associating this model with load duration curves within each discrete frame. In this paper, an attempt is made to evaluate a reliability index called the Energy Index of Reliability (EIR). This index is obtained in terms of another reliability index Expected Energy Not Supplied () and the load demand for the area, being supplied by the hybrid system. The scheme proposed for the study consists of variable mix of wind turbines and diesel generators. The basic scheme is shown in Figure. The number of wind and diesel units can be varied. I. INTRODUCTION He most difficult problem in the operation of a hybrid Tenergy system involving wind energy is due to, to the intermittent characteristics of wind energy. The most difficult issue is to assess the capacity adequacy of the hybrid system in addressing the electricity demand of the consumers. However, the well developed techniques applied to conventional generation system reliability evaluation, allocate fixed capacity outputs to generating units and cannot readily be extended to include wind energy sources that have highly fluctuating capacity levels[]. A number of authors have reported a variety of models to deal with the reliability issue. All these models can be roughly classified by their techniques into two categories; the Monte Carlo simulation and the analytical method. The application of the Monte Carlo simulation to wind energy generation system utilizes prediction techniques to obtain time series wind speed data and integrate this fluctuant attribute with generation unit indices[-4]. In this method, the most critical step is the estimation of wind speed data, which requires historical wind speed data and a wind speed prediction model [5]. A well S. Kishore Kumar is an M Tech student, Abha Rajoria is a Research Scholar and E Fernandez is a Faculty with the Electrical Engineering Department, IIT Roorkee, Roorkee (Uttarakhand) The objective of the study is to evaluate the EIR for different generator unit mix combinations supplying a given load demand to a remote area. This information will provide the energy planner sufficient support in making decisions in regard to desired adequacy levels under economic constraints. The approach proposed in this paper studies the performance of wind-diesel hybrid power system in discrete wind speed frames on the Weibull wind speed distribution to obtain long term reliability indices. The wind turbine generator modeling procedures deal with the single or multiple turbines. For the wind turbine generator, we assume that the outage events are 67

2 independent of each other and have no correlation with wind speed. II WIND DIESEL HYBRID SYSTEM CHARACTERISTICS A Power Curve of Wind Turbine Generator The Power curve gives a quantitative relationship between wind speed and wind turbine power output. It describes the operational characteristics of a wind turbine generator. Thus, to any power output, a corresponding wind speed can be found on this curve. A typical curve is shown in Figure.2 C Wind speed frames and wind turbine power output Some of the earlier developed models use discrete wind speed points to describe the fluctuating characteristic of wind speed in time sequence, which means that turbine power output are studied one speed point after another. This requires a large amount of computation time to conduct calculation for numerous time intervals. However, by observing the Weibull wind speed distribution curve, it is not difficult to find that wind speed profile can be easily depicted by a series of speed frames and their probabilities. Figure 3. Shows the wind speed frame on the Weibull distribution curve. Fig 2: A typical power curve of wind turbine generator V c is the cut-in speed V R is the rated speed V F is the cut-out speed C WTG is the rated Power output. This curve comes available from the wind turbine manufacturer or plotted by polynomial curve fitting technique using recorded wind speed and corresponding power output data. B The Weibull Wind Speed Distribution It is widely accepted that the Weibull distribution fits actual wind speed distribution quite well. Like many other studies, the Weibull distribution is utilized to describe the principle wind speed variation in this study. Its probability density function is given by: Fig3. Wind speed frame on the Weibull distribution The Weibull distribution can be split along the horizontal axis into N equal frames. The first frame and last frame are from 0 to V and from V N- to V N respectively, with V N V around m/s is sufficiently small. T i is the duration of i-th speed frame T is the whole study period. P wi is the availability of the i-th speed frame. (2C-) p wi = (2C-2) P WTGji is the average power output of j-th wind turbine generator operating with i-th wind speed frame, (2C-3) f v v exp v (2B-) D Wind farm modeling V is the wind speed (m/s) is the shape parameter and is the scale parameter (m/s) A Hybrid system may contains m wind turbines generators. If the Forced Outage Rate (FOR) of the wind turbine is FOR WTG, the availability rate can be represented by A WTG FOR WTG (D-) 68

3 If h wind turbines out of m wind turbines were operating in ith speed frame, the wind farm remaining capacity can be given by C = RWTGi P h (2D-2) WTGi P WTGi = average power output of any identical turbine generator operating within ith speed frame. The Probability of the h identical turbines available in calculated according to the Binomial distribution m h m h P WTG h AWTG AWTG (2D-3) A = availability of any identical wind turbine WTG Generator The above procedure for calculating capacity and probability applies when all the turbines are of same rating. When there are different ratings the corresponding probability can be obtained by utilizing the convolution algorithm. A Expected Energy Not Supplied () The loss of energy when g available diesel generators and h available wind turbines operating within ith speed frame can be denoted by the area E i the expected energy not supplied throughout N speed frames, gh can be given by. gh N i p E i i (3 A-) The above equation can be generalized operating throughout the whole range of N wind speed frames, the corresponding expected energy not supplied can be represented by xy so the of the hybrid system can be obtained by n m x 0 y 0 xy (3A-2) Then the Energy Index of Reliability (EIR) can be obtained by If g out of n Diesels are available. If all these diesel sets belong to same type, the remaining capacity of the system can be given by C g RD P (2D-2) D Corresponding probability can be given by n n g D g D D g P A A (2D-3) E Hybrid System Modeling Hybrid system model can be easily obtained by integrating the turbine model and the diesel model. The hybrid system capacity and corresponding probability can be given by the expressions E- and E-2 as: And CRDWi CRD C RWTGi (2E-) p p p DWi D WTG (2E-2) III RELIABILITY EVALUATION OF HYBRID SYSTEM After modeling the wind turbine, and diesel generator, the next task is the evaluation of the reliability indices of the whole system. The main reliability indices are LOLE and and EIR. EIR Energy. demand IV RESULTS AND DISCUSSION (3A-3) The methodology presented in this paper was applied to the data of a wind-diesel hybrid power system reported in [ ] The minimum and Maximum and average speeds of the test site are 0.2m/s,2.08 m/s respectively. By studying the wind speed records, the shape Parameter k and scale parameter c of the Weibull distribution are computed as and m/s respectively. The Weibull distribution is shown in Figure 4. PROBABILITY DENSITY FUNCTION WIEBULL DISTRIBUTION CURVE WIND SPEED m/s Fig.4 Weibull speed distribution curve 69

4 Figure 5 shows the load probability distribution curve at the site. Fig5. Load distribution curve at test site The Matlab program based on the method presented in this paper wasdeveloped to calculate the Expected Energy Not supplied () and Energy Index of Reliability (EIR). The proposed model can be adopted to analyze different hybrid system schemes to obtain an optimal system configuration. A series of case studies were conducted to calculate the reliability indices of various schemes based on current load data. The results are shown in the following figures. Fig 7: Energy Index of Reliability (EIR) for different System Configurations As shown in above Fig7, which indicates wind turbines can provide significant contribution to system reliability. It should be noticed that the effect of wind turbine on system reliability is limited. The increase of diesel generator number leads to dramatic improvement in system reliability.when diesel generator number reaches 3, the Energy Index of Reliability (EIR) still increases substantially. By comparing these figures, it can be seen that although the increase in system capacity can make the system more reliable, the improvement is very marginal after some level,and the investment also increases greatly. Therefore an appropriate system scheme not only keeps system stable, but also makes system cost effective. In view of this, it would appear that a hybrid system consisting of two wind turbines and a single diesel genset unit will give an EIR close to unity. The same result can also be obtained by considering a hybrid system made up of two diesel units and one wind turbine (seefigure 7) Thus as these different combinations yield the same levels of reliability, the system planner will be able to decideon the lesser initial cost option. Further, a trade-off may alsobe possible if the system planner is prepared to accept a littlelower value of EIR at a saving in initial cost. In that case, the option of either a system of one wind turbine and one dieselgenset or only two gensets will yield an EIR above 0.9, and either can be chosen. Fig 6 Expected Energy Not Supplied () for different System Configurations V CONCLUSIONS A new technique of reliability evaluation of wind-diesel hybrid power system by utilizing discrete wind speed frame is presented in this paper. The main aspect of this technique is 70

5 studying wind turbine generator performance within each discrete speed frame on the Weibull distribution curve. This model is applied successfully to calculate and EIR of an existing wind-diesel hybrid power system. Battery Bank Modeling was not considered in this paper, by using Battery Bank Modeling the system reliability can improved much. Maintenance and repair characteristics of both conventional and unconventional units are not considered in this paper, but they can be easily taken into account by this model Additionally, this model can help system planner and utility administrators to evaluate, upgrade, size and optimize their hybrid power systems. REFERENCES [] R. Karki and R. Billington Cost Effective Wind Energy Utilization for Reliable Power Supply. IEEE Trans Energy Convers, Vol 9 pp , June 2004 [2] R. Billington, H. Chen and R. Ghajar.A Sequential Simulation for Adequacy Evaluation of Generating Systems including Wind Energy.IEEE Trans Energy Convers, Vol, pp , Dec 996. [3]R. Billington and H. Chen Assessment of Risk-based Capacity Benefit Fraction Associated with Wind Energy. IEEE Trans on Power Systems, Vol 3, pp 9-95, Aug 998. [6]E.S.Gavanidou, AG Bakirtzis and P.S. Dokopoulos. A Probabilistic Method for Evaluation of the Performance and the Reliability of Wind-Diesel Energy Systems. IEEE Trans on Energy Convers. Vol. 8(2) pp , Jun 993. [7] S.H. Karaki, R.B. Chedid and R. Ramadan Probabilistic Production Costing of Diesel Wind Energy Conversion Systems. IEEE Trans on Energy Convers. Vol. 5(3) pp , Sept [8] X. Liu and S Islam. Reliability Evaluation of a Wind- Diesel Hybrid Power System with Battery Backup using Discrete Wind Speed Frame Analysis., 9 th International Conference on Probabilistic Methods Applied to Power Systems, KTH, Stockholm, Sweden, June -5, Appendix: Details of Wind-Diesel System used for study Details of Sub-System Ratings Diesel Genset 320 kw /unit No. of Units= 03 Forced outage Rate(FOR)=0.06 Wind Turbine Generator (WTG) 300 kw/unit [4] R. Billington and G. Bai Generation Capacity Adequacy Associated with Wind Energy. IEEE Trans on Energy Convers,,Vol 9, pp 64-46, Sept [5] R. Billington, H. Chen and R. Ghajar Time Series Models for reliability Evaluation of Power Systems including Wind Energy. Microelectron. Reliab, Vol 36(9) pp253-6, 996. No. of Units = 0 Forced outage Rate(FOR)=0.0 Cut in Speed= 3.0 m/s Rated Speed = 4 m/s Cut out Speed = 25m/s. 7