Elsevier

Applied Energy

Volume 94, June 2012, Pages 156-165
Applied Energy

The assessment of the contribution of short-term wind power predictions to the efficiency of stand-alone hybrid systems

https://doi.org/10.1016/j.apenergy.2012.01.017Get rights and content

Abstract

Distribution of electricity to the rural areas, particularly in regions which have rough topography causes high costs and losses. Hybrid systems can provide electricity at a relatively economic price at these regions. This paper designs and tests a stand-alone hybrid system combining variable speed wind turbine (WT), fuel cell (FC) and battery. The main objective is to optimize the hydrogen utilization while guaranteeing the load balance implicitly as well as to achieve proper FC operation. To this end, a wind power prediction based controller is proposed in order to take action according to foreseen amount of power deficit or excess in the system. For the purpose of investigating the effects of predictions on the system efficiency, a case study is carried out on a coastal area with a high wind potential in Izmir, Turkey. The results obtained provide insights about the advantages of utilizing wind power predictions in a hybrid system.

Highlights

► Analysis of a stand-alone wind turbine/fuel cell/battery hybrid system. ► The utilization of wind power predictions in order to optimize the hydrogen consumption. ► Application of fuzzy logic for hybrid system energy management. ► Evaluation of hybrid system performance with and without prediction based control.

Introduction

Wind power has become increasingly competitive with other conventional power generation systems over the last decade and the share of wind energy in worldwide electricity production has increased dramatically. The cost-effectiveness, reliability and efficiency of wind energy compared to those of other alternative sources can be regarded as the main reasons for this development. However, it is evident that a wind energy system cannot provide a continuous supply of energy due to the intermittent nature of wind. Therefore in order to meet sustained load demands, particularly in remote rural areas where economics are not favorable to extend the grid supply, different energy sources need to be combined. Thus, it may be likely to utilize alternative energy resources more efficiently and alleviate the power quality problems caused by wind power.

To this end, a great number of concepts and system designs have been studied in the literature over the past decade [1], [2], [3]. In order to maintain a constant power under all conditions, Kalantar and Mousavi present a stand-alone system of wind turbine (WT), microturbine, solar array and battery storage [4]. Another stand-alone wind/fuel cell (FC) hybrid system is proposed by Khan and Iqbal [5]. They propose to store superfluous energy from wind by producing hydrogen with the aim of later use. Taking this issue into account, many authors utilize the storage systems at different scales such as ultracapacitors, electrolyzer/FC systems, flywheels, pneumatic systems and batteries [6], [7], [8], [9], [10]. Hence a share of power can be supplied from these systems while decreasing the size and cost of the WTs.

Lately, exploiting the energy from WTs during periods of over-production and using this energy with a high efficiency when wind power is not sufficient become a promising research area where many researchers have contributed. With the objective of increasing the hydrogen production according to the wind conditions, Valenciaga and Evangelista [11] propose a stand-alone system consists of a wind energy generation module, an electrolyzer and a battery bank. Khalid and Savkin [12] utilize wind power prediction so as to provide an efficient control to smooth the wind power output and prolong the battery life in a hybrid energy system. Hence, it is inferred that accurate wind power predictions would be utilized to help in achieving energy efficiency to some extent apart from its main applications such as unit commitment, economic dispatch and participation in the electricity market.

This study gives a further step in terms of improving efficiency by making use of wind power prediction in a system created by the combination of a WT, an FC and a battery bank. Thus it is aimed to minimize the hydrogen consumption without putting at risk the quality and continuity of the energy provision and reducing the operative life of the battery bank. For this purpose, operation of a residential application primarily powered by a WT and supported by a hybrid system made up of a PEM (Proton Exchange Membrane) FC stack and a battery bank is modeled and simulated using MATLAB, Simulink and SimPowerSystems software packages.

In the proposed system, when the WT produces enough power, the load demand will be supplied entirely from wind energy. In the case of probable low wind conditions, the control system examines whether the predicted wind and battery power will be sufficient for net load according to the outputs from the wind prediction system. Otherwise, the deficit power can be supplied from the FC and/or the battery with respect to control system. If the output power from the WT exceeds demand, the excess power may be stored by the battery banks to be used for insufficient generation.

This paper is organized as follows: Section 2 presents the proposed system and gives a brief description of each component. The prediction model developed is explained in Section 3. The case study and simulation results are given in Section 4 and Section 5 concludes our study.

Section snippets

System description

A simulation model of the hybrid system is built in MATLAB/Simulink and SimPowerSystems in order to study the performance of the proposed control strategy. The main components in the system proposed are a 2 kW PEMFC model with a DC/DC converter, a battery bank with a reversible converter which is capable of either charging the battery or recovering the storage energy, and a 5 kW WT model as illustrated in Fig. 1.

Among the various types of FC systems, PEMFC has been found to be especially suitable

Wind power prediction model

As stated earlier, the wind power prediction model is integrated with the hybrid energy system so as to optimize the hydrogen consumption as well as to ensure the continuity of load supply. Therefore, the controller utilizes the inputs from the wind prediction model and power capacities of the hybrid system components. The proposed wind power prediction model consists of four processes, as given in Fig. 4.

The first process is to utilize wavelet transform to cut up the original wind speed series

Case study and results

The location selected for the simulations is on Aliaga (Izmir), a town in western Turkey (Fig. 6). This region is chosen because of its high wind potential during all seasons. For the simulation studies, a period of one day is chosen that is representative of the general situation which can probably be seen in this region for these wind speeds and load profile. In order to avoid data redundancy, simulations are only carried out with the values concerning the autumn of 2010, as depicted in Fig. 7

Conclusions

As WT output power varies with wind speed, it can be combined with FC stack and battery storage to satisfy the load demand under all conditions and to assure the required degree of reliability. In this study, the wind generation system is used as main energy source while the FC and battery are used as back-up energy sources.

The main contribution of this work is the minimization of hydrogen amount consumed in the FC system utilizing short-term wind power predictions. The simulation results show

Acknowledgments

This work was supported in part by “Yildiz Technical University Scientific Research Projects Coordination Department” under Grant 2011-04-02-DOP02”. Also the authors are gratefully acknowledged to the General Directorate of Electrical Power Resources Survey and Development Administration (EIE) for providing the real wind speed data required for this study.

References (24)

Cited by (35)

  • Ultra-short term wind prediction with wavelet transform, deep belief network and ensemble learning

    2020, Energy Conversion and Management
    Citation Excerpt :

    More relevant researches could be seen in Ref. [13–15]. Even though the subseries of WT are more predictable, there still exists difficulty in high-accuracy prediction [16]. For time-series signal forecasting, another effective way to facilitate the prediction accuracy is extracting deeper features [17].

  • Data mining-assisted short-term wind speed forecasting by wavelet packet decomposition and Elman neural network

    2018, Journal of Wind Engineering and Industrial Aerodynamics
    Citation Excerpt :

    Based on this knowledge, the present study employed a decomposition space number of 3 and Daubechies 6 as the wavelet function as it these conditions have been widely employed in many studies (Catalão et al., 2011). The details of the parameters choices are presented in the previous studies (Avdakovic et al., 2012; Tascikaraoglu et al., 2012). The structures of the ENN must first be determined prior to wind speed forecasting.

View all citing articles on Scopus
View full text