Energy flow management of a hybrid renewable energy ... - IEEE Xplore

12 downloads 745 Views 2MB Size Report
a Laboratory for Electrical Engineering and Renewable Energy Systems, Institute for Energy ... In the model two energy storage devices, a battery and a.
Energy flow management of a hybrid renewable energy system with hydrogen Lars Baumanna, b,*, Ekkehard Boggascha, Mark Rylattb, Andrew Wrightb a Laboratory for Electrical Engineering and Renewable Energy Systems, Institute for Energy Optimized Systems – EOS, Ostfalia University of Applied Sciences Wolfenbüttel, Germany b Institute of Energy and Sustainable Development, De Montfort University Leicester, UK

In the model two energy storage devices, a battery and a hydrogen reservoir are included. To control the energy flow a simple algorithm based on state-charts was developed. The aim of the controller is to optimize the self-consumption of renewable electrical energy in the system. Primarily, the controller uses renewable sources to meet the electrical load. If the demand exceeds the supply, the storage devices are discharged.

Abstract: This paper presents a preliminary study on a hybrid renewable energy system at the Ostfalia University of Applied Sciences Wolfenbüttel, Germany. The test-bed is made up of solar photovoltaics (PV), a micro wind turbine (MWT), a micro-CHP, a fuel cell system (FC), and two storage devices (a battery system and an electrolyzer). All the installations are in the range of 1 to 6 kW electrical power output/input; the focus of this research is renewable energy systems for residential applications. In addition to the presentation of experimental test results, a model of the hybrid system will be introduced.

*

Corresponding author. Tel.: +49 5331 939 39740. E-mail address: [email protected]

could be useful to utilize all available power from sun and wind [4]. Such installations with different power sources in combination with storage devices are generally referred to as hybrid energy systems.

1. Introduction The European Union has set targets for reducing the carbon dioxide emission to 30% below 1990’s levels and to increase the proportion of renewable energy sources in its energy mix to 20% by 2020 [1]. In Germany the Renewable Energy Sources Act (EEG) has promoted application of renewable energy sources. Therefore, the number of small scale power sources in low-voltage networks has significantly increased [2]. It can be noticed that the traditional centralised power generation is changing to a distributed power system. The so-called microgeneration technologies (e.g. small scale photovoltaic solar system, wind turbines, combined heat and power, or fuel cells) put the generation of electrical power closer to the end-user to supply their buildings.

For short- and mid-term electricity storage the most commonly used device is nowadays the lead acid battery. An option to store surplus renewable electric energy for a longer period would be to convert it into hydrogen by using an electrolyzer. If the hydrogen is produced from renewable sources and used in fuel cells it would be a clean and reliable energy source. The utilization of hydrogen as a storage opportunity is triggered by the energy policy of many countries and it would help to get more independent from fossil fuels [5]. Recent studies have shown that hydrogen could play an important role in storing large amounts of surplus electricity from wind farms and consequently could help to balancing the variations in renewable energy production [3, 6]. Besides the large scale utilization of hydrogen storage systems an additional application would be as part of small hybrid renewable energy systems for stand-alone or grid connected residential applications. In [7-10] the performance of hydrogen systems for buildings was investigated and a feasible installation of using hydrogen in homes was demonstrated. A review of reported renewable driven hydrogen systems was undertaken in [11] and its conclusion suggests that there is a potential of using hydrogen systems in the housing stock. The results of an analysis reported in [12] underline that hydrogen technologies would be a promising option for heating and power generation for residential application and could help to reduce the carbon dioxide emissions.

Renewable energy sources are on one hand free of greenhouse gas emissions and they are unlimited, but on the other hand they are site-dependent. Furthermore, renewable energy sources – such as wind and solar – have an intermittent nature. The power outputs of PVand wind-systems are very dependent on weather conditions (irradiation and wind speed respectively); therefore their power production varies with month, days, hours, and seconds. It is generally known that electricity should be produced just in time to guarantee grid stability. In future with a wide penetration of renewable sources these requirements may not be fulfilled and power variations could influence the power quality such as voltage and frequency. A recent published study [3] shows the importance of implementing storage devices to tackle the intermittency problem. In addition, the introduction of power systems using more than one energy source –at least one of them must be controllable-

978-1-4244-6078-6/10/$26.00 ©2010 IEEE

78

To emulate different load scenarios three programmable electronic loads with a total rated power of 10.8 kW are currently being integrated into the test-bed. These give the opportunity to stress the system with realistic load profiles with a high temporal precision. The typical load profile of a residential building is intermittent around a low base load of a few hundred Watts with short peak demands usually in the range of 7…10 kW. Therefore the temporal resolution of load profiles using for experimental or simulation studies is essential for the results. A study untaken in [13] has revealed that using hourly data rather than minutely data can dramatically affect the power balance of demand and supply throughout a year.

In this paper an experimental set-up and a first simple MATLAB/Simulink® model of a hybrid renewable energy system with hydrogen will be introduced. The main objectives of the ongoing IGES1 project are to analyze the energy flows in complex hybrid energy systems and to develop a suitable management system for residential applications, which considered both electrical and thermal energy. 2.

Experimental setup

A schematic of the hybrid renewable energy system configuration considered in the current work is shown in Fig. 1. The test-bed at the Laboratory for Electrical Engineering and Renewable Energy Systems consists of a 5.1 kWp PV-array, a 1 kWp PV-array with adjustable tilt angle, a 4 kW micro wind turbine, a 1.2 kW fuel cell system, and a 6 kW electrical power micro-CHP unit. To store surplus electricity two storage devices are installed; a three-phase battery system (approx. 20 kWh) and a 6 kW alkali electrolyzer (1Nm³/h H2 @ 30 bar) combined with a hydrogen tank (18 Nm³). The production rate of the electrolyzer is controllable between 20…100 % by using a 4…20 mA signal.

All of the devices are connected via a building automation system using the LON2 networking platform to a distributed information system. LON is characterized by a high data performance, reliability, and a user friendly system configuration. All of the process values are stored via OPC3 interface into a MySQL database with a sample rate of one second.

Fig. 1: Hybrid renewable energy system at the Laboratory for Electrical Engineering and Renewable Energy Systems 1

Official project title is: IGES Intelligente-Gebäude-Energie-Systeme Local Operating Network 3 OLE for Process Control 2

79

This allows a very detailed view of the dynamic aspects of such a system. For the data mining the MATLAB® software is used.

2.1 Dynamic Data of the test bed The electrical power supply of the PV-array and the MWT for the 4th September 2007 at a 5 sec resolution and corresponding 5 min averages are shown in Fig. 2. It can be seen that the 5 min mean values are much lower and smoother than 5 sec averages. Large but brief power spikes (see the lower diagram) hardly affect the 5 min average. These relatively high power gradients must be considered in the energy flow management for small offgrid hybrid systems, and for utility grid districts with a high penetration of such devices.

The installed hybrid renewable energy system could be used to observe both autonomous and grid connected configurations. In comparison to other studies [14-17], the work described here deals with an existing testfacility and the installed communication infrastructure allows to monitor and to control the system with a high temporal resolution (1 sec). Furthermore, the developed system model and management algorithm could be validated against the real setup.

Fig. 2: Supply profile of the renewable energy system at 5 sec and 5 min time resolution for a single day in September 2007. The upper diagram shows the entire day; the lower diagram shows a more detailed two hour period of the day.

80

3. Simulation Model In order to analyze the performance of the hybrid renewable energy system and to develop an energy management algorithm for residential applications, a first simple model is introduced based on MATLAB/Simulink®. The model contains all components of the test-bed, a simple battery model and an electrolyzer with hydrogen storage. The electrolyzer model is based on a look-up table which shows the measured hydrogen production rate depending on the input electrical power of the installed system. The minimum hydrogen flow rate is around 0.113 Nm³/h at 1.6 kW and the maximum exceeds 0.9 Nm³/h at 6 kW electrical power. Figure 3 shows the implemented relationship between the production rate and the power input.

Fig. 3: Measured hydrogen production rate [Nm³/h] against electrical power input [W] taken from the installed electrolyzer (AccaGen AGE 1.0, H2 quality 99.99%) with an accuracy of ± (0,8 % Rd. + 0.004Nm³/h)

To simulate the electrical load two power demand tables for Simulink® were created either based on the standardized German VDEW4-H0 load profile or on an individual load profile. A schematic of the model considered in the current work is presented in Fig. 4.

Fig. 4: Schematic of the hybrid system model The energy management and control algorithm for a first analysis of the dynamical behaviour was developed using Stateflow®. Figure 5 illustrates the flowchart of the implemented algorithm. Transition between states is based on “if-then” rules. First the power balance PSources minus PLoads is calculated before the algorithm takes the storage capacities into account.

The data from the MySQL database can be used directly by MATLAB/Simulink®. All existing components can either be totally simulated or measured data can be used. Moreover, the OPC interface allows for a connection between the Simulink® environment and the existing installations of the hybrid system. The simulation step size is in this case 1 second. 4

VDEW: Verband der Elektrizitätswirtschaft in Deutschland

81

Fig. 5: Flowchart of the implemented energy management algorithm used in Simulink®/Stateflow®

Fig. 6: State-chart of the hydrogen storage control. All other components of the systems were represented by a model. The storage size was assumed to be 10 kWh battery capacity (minimum depth of discharge of 50 %) and a 18 Nm³ hydrogen tank which is equivalent to approximately 18 kWh by using the Nexa fuel cell. The used load data is derived from a load model developed in [18]. Figure 7 shows the generated load profile of typical 4 occupants’ household and the measured electrical power output of the renewable sources for a week in July 2009 both with a temporal resolution of 1 minute.

The energy management system consists of three statecharts for the battery, hydrogen, and CHP control. For instance figure 6 illustrates the state-chart for the hydrogen storage control. The optimization function of the energy management prioritises renewable energy to meet the current demand. If there is a surplus of energy it is stored in the battery. If the battery is fully charged the energy is stored as hydrogen. Not until both stores are completely full is the energy exported to the grid. The residential power demand is only provided by the grid when the limits of storages are exceeded. 3.1 Simulation Example The simulation uses measured data for the PV, MWT and weather conditions of one week in July 2009 with a simulation step-size of 1 sec.

82

Fig. 7: Domestic load profile (red line) for 4 person home generated with the model in [18] with a temporal resolution of 1 minute and electrical power (green line) drawn from renewable sources. The resulting electrical energy produced by the renewable sources was 164.03 kWh (153.1 kWh PV energy + 10.93 kWh wind energy) and the total consumption of the residential home was 72.81 kWh. Therefore, there is surplus of electrical energy of 91.22 kWh. Fig. 7 illustrates that the power from the PV is dominant during the summer in comparison to the output of the micro-wind turbine. In addition, it reveals that the greatest demand of the occupants occurs in the morning and evening, in times with a lower solar irradiation. During these times the demand must be met by the storage devices or the grid. Fig. 8: Power difference between load and renewable sources

The system was simulated for a complete week to observe the energy flow. Initial values for the hydrogen storage pressure and battery capacity were assumed to be 20 bar and 7.5 kWh, respectively. Figs. 8-10 show the power difference between the load and the renewable sources, the battery state of charge and the pressure in the hydrogen, respectively, during the simulation of one week. It can be seen from Fig. 9 that the battery is alternating between its limits of 50 % and 100 % state of charge. The battery power is used first because of its higher efficiency in comparison to the fuel cell and electrolyzer. The pressure in the hydrogen tank is increasing during the week (Fig. 10). The total energy consumption of the electrolyzer amounts to 32.37 kWh and the energy output of the fuel cell amounts to 3.49 kWh during the week.

Fig. 9: Battery state of charge

For a better understanding the energy flow and dynamics in a hybrid renewable energy system a snapshot of day 4 is presented in Figs. 11-14. The figures show the moving average (2 minutes) of the power difference between load and sources, the power flow of the battery, the power of the hydrogen system, and the net power flow. Fig. 10: Pressure in the 0.6 m³ hydrogen tank

83

Fig. 11: 24h snapshot of the power difference between load and renewable sources (2 minutes moving average)

Fig. 13: 24h snapshot of the hydrogen system power (2 minutes moving average)

Fig. 12: 24h snapshot of the battery power (2 minutes moving average)

Fig. 14: 24h snapshot of the power import and export (2 minutes moving average)

It can be seen from the Figs. 11-14 that there is a high dynamic behaviour in the power variations. For the first analysis it is assumed that the battery and the hydrogen system are capable of meeting these transient changes of the power flow. However, a more detailed simulation as well as more experimental studies are necessary to get a better understanding of the dynamic performance. Based on these investigations a suitable management algorithm could be defined.

• •

Design guidelines for control systems that will be capable of optimising highly hybrid systems with short and long term storage capacity. Quantification of the benefits arising from the proposed control systems extrapolated over residential stock under different scenarios, including estimates of carbon reduction.

Acknowledgement: The ongoing IGES project is funded by the European Union and the Government of Lower Saxony under grant EFRE W2-80026918.

4. Future prospects In this paper a test-bed of a portfolio of microgenerators has been presented. In addition to the physical installation a first simple model of a hybrid renewable energy system for residential applications was introduced using MATLAB/Simulink®. The presented work are the first results of the IGES project, which started in November 2008. The ongoing IGES project will focus on the following work: • A study of the issues and opportunities for bidirectional energy management arising from the complex load and supply balancing requirement in highly hybrid systems.

References: [1] An energy policy for Europe; Commission of the European Communities, Brussels 10.01.2007; www.eur-lex.europa.eu [2] Development of renewable energy sources in Germany in 2009 - Graphics and tables; Federal Ministry for the Environment, Nature Conversation and Nuclear Safety (BMU) Germany, March 2010

84

[3] VDE-Study: Energy storage in power supply systems with a high share of renewable energy sources, German Association for Electrical, Electronic and Information Technologies VDE, 2009

[15] Bilodeau A, Agbossou K. Control analysis of renewable energy system with hydrogen storage for residential applications. Journal of Power Sources, 162 (2006), 757-764

[4] Paska J, Biczel P, Klos M. Hybrid Power System – An effective way of utilising primary energy sources. Renewable Energy, 34 (2007), 2414-2421

[16] Dufo-López R, Bernal-Agustín J L and Contreras J. Optimization of control strategies for stand-alone renewable energy systems with hydrogen storage. Renewable Energy, 32 (2007), 1102-1126

[5] Yilanci A, Dincer I, Ozturk H K. A review on solarhydrogen/fuel cell hybrid energy system for stationary applications. Progress in Energy and Combustion Science, 35 (2009), 231-244

[17] Ipsakis D, et al. Power management strategies for stand-alone power system using renewable energy sources and hydrogen storage. International Journal of Hydrogen Energy, 34 (2009), 7081-7095

[6] Stolzenburg K, et al. Hydrogen as a Means of Controlling and Integrating Wind Power into Electricity Grids – The HyWindBalance Project., http://www.hywindbalance.com/gb/HyWindBalance_Res ults.pdf

[18] Stokes M. Removing barriers to embedded generation: a fine-grained load model to support lowvoltage network performance analysis. IESD, Leicester, De Montfort University, PhD thesis 2005

[7] Ulleberg Ø, Mørner S O. TRNSYS Simulation models for solar-hydrogen systems. Solar Energy, 59 (1997),, 271-279 [8] Yvon K, et al. Evalutation of a 5 kWp photovoltaic hydrogen production and storage installation for residential home in Switzerland. International Jornual of Hydrogen Energy, 25 (2000), 97-109 [9] Maclay J D, Brouwer J and Samuelsen G S. Dynamic analysis of regenerative fuel cell power for potential use in renewable residential applications. International Journal of Hydrogen Energy, 31 (2006), 994-1009 [10] Stewart E M, et al. Modeling, analysis and control system development for the Italian hydrogen house. International Journal of Hydrogen Energy, 34 (2009) 1638-1646 [11] Deshmukh S S, Boehm R F. Review of modeling details related to renewably powered hydrogen systems. Renewable and Sustainable Energy Reviews, 12 (2008), 2301-2330 [12] Lubis L I, Dincer I, Naterer G F, Rosen M A. Utilizing hydrogen energy to reduce greenhouse gas emissions in Canada’s residential sector. International Journal of Hydrogen Energy, 24 (2009), 1631-1637 [13] Wrigth A, Firth S. The nature of domestic electricity-loads and effects of time averaging on statistics and on-site generation calculations. Applied Energy, 84 (2007), 389-403 [14] Maclay J.D., Brouwer J. and Samuelsen G.S. Dynamic modelling of hybrid energy storage systems coupled to photovoltaic generation in residential applications. Journal of Power Sources, 163 (2007), 916925

85