In order to acquire realistic data and to have distributed sources connected in the laboratory network three systems were designed and built. First is the system for solar characterization with purpose of better measurement of solar radiation and different photovoltaic technology in operation. Then complete photovoltaic system was built in order to measure grid connected operation in real conditions. Finally, hybrid system was built in order to measure conditions when wind power, solar power and energy storage are working together.

Description for these three systems follows with some initial results. 

1. Solar power characterization (RE3)

   Photovoltaic panels produce electricity, depending on their characteristics and solar radiation. For more realistic assessment of electricity generation from photovoltaic panels, it is necessary to make better assessment of the behavior of photovoltaic panels in different conditions and predict the intensity of solar radiation better.

   For empirical monitoring of the behavior of different photovoltaic panels in all common types of installations it is necessary to make the same number of test installations. Taking into account that there are at least five photovoltaic panels technology, now widely available, the number of necessary installation becomes unfeasible for an empirical test. Therefore, this project is designed so that only with six photovoltaic panels of different technologies (poly, mono and amorphous silicon, and two types of thin film without silicon) placed in the two positions with the measurement system critical set of possible positions is covered (with necessary analytical processing).

   Already designed and installed pilot unit (Figure 20) and project ready for installation are aimed to empirically explain  the behavior of different photovoltaic panels installed at FER and at four different locations in Croatia (i.e., Osijek, Karlovac, Rijeka and Split).

   For both positions, current-voltage characteristics of the panel, temperature and solar irradiation on horizontal surfaces are measured. In this way very detailed results of the behavior of photovoltaic panels in two most important positions at different actual solar irradiation and different weather conditions (temperature) is obtained. The production forecast of photovoltaic panels installed at each position is done analytically, by applying correlation terms where horizontal values are converted to values of solar radiation reaching the surface of the panel position. Correlation models assume an average share of direct and diffuse components in a certain area and can more or less correspond to a specific location.

   Expected outcomes of annual measurements from this system could be used for different purposes. First, the results obtained would help improving the correlation and verification procedures. Then, the results obtained would demonstrate the actual performance of certain photovoltaic technology depending on the actual conditions (solar irradiation, temperature, etc.). Finally, this would empirically show the advantage, or lack of it, of the application of advanced solutions that position photovoltaic panels toward the sun direction. It would also improve prediction of best fixed position of photovoltaic panels for maximizing the annual or seasonal electricity production.

2. Integration of solar power in power system (RE1)

   RE1 or PV1kW_D stands for the solar system located on the top of the D building at Faculty of Electrical Engineering and Computing.

First the software predictions and calculations were made using the HOMER software. After the simulation of the PV system in HOMER all necessary hardware was bought (PV modules, inverter, communication and monitoring system, protection). Momentarily PV system is installed and ready to use.

System components

• PV module SV-60

   Using the meteorological data for city of Zagreb as an input for the HOMER software helped reaching the decision of choosing the output power of the PV system. For the installation 5 SV-60 PV modules with 200 W peak power per unit were chosen. SV-60 PV modules were constructed in Solvis d.o.o. in Croatia. Electrical data for SV-60 module are displayed in Table 2, mechanical in  Table 3and thermal in Table 4.

Table 2 Electrical data for SV-60 module

• The SMA Sunny Boy 1200 inverter

   Input data (DC side) for SMA Sunny Boy 1200 are displayed in the Table 5and output data (AC side) is in Table 6.

Table 5 Input data (DC side) for SMA Sunny Boy 1200

Table 6 Output data (AC side) for SMA Sunny Boy 1200

   Inverter is connected in the PV system between PV modules and distribution network as it is shown on the figure below. The inverter was installed following the producers recommendations.

• Sunny WebBOX

   As a high-performance communication hub, Sunny WebBOX represent central interface for communication of the PV system. Sunny WebBOX collects all the data from PV modules and also allows controlling system monitoring, remote diagnosis and visualization of the data. Connection between Sunny WebBOX and personal computer or portable device can be provided through global internet or GSM or locally by 10/100 Mbit Ethernet. RS485 communication bus is used for connection between Sunny Boy inverter and Sunny WebBOX. A constraint for communication range of RS485 is 1200 m and for Ethernet is 100 m.

• Connection electricity boxes

   Two types of boxes are needed: protective and measuring. Protective equipment has function to protect PV modules and inverter from short circuit current, overvoltage or from overload etc. Measuring equipment is connected to the LV grid to measure power which is produced from PV system.

Integrated system overview

            All the mentioned components are installed and ready to use. PV system will be started when the server component of the project will be physically established. Test runs of the PV system could be done before implementation of the server component. This system can be used to study PV effects on the grid or to study PV power production but on the small scale. PV system which will be an important component in the smart grid testbed is shown on the figure below (Figure 23).

Figure 23 Photovoltaic system PV1kW_D connected to the grid (RE1)

3. Hybrid distributed generation system (RE2)

   First the software predictions and calculations were made using the HOMER software. After establishing certain scenarios which are most likely to happen the search for all the necessary hardware had begun. A couple of different offers were analyzed and a decision was made to buy certain hardware (PV, WT, inverter, controller, data communicator, battery bank) which will be listed later.

   The software simulation using HOMER was conducted with local wind and sun resources as depicted in figures below:

Figure 24 Solar resource

Figure 25 Wind resource

Based on the results the optimal solution which had all the components (a battery, PV, wind) was chosen:

Figure 26 Possible solutions and system configuration

   HOMER gives the predicted energy productions, battery states throughout the year etc. but that is not of great importance once the components were chosen and bought.

   The software ANSYS was used to further determine the optimal position of the wind turbine depending on the available wind. ANSYS can give distributions and simulate the wind flow based on the input data of the developed model. Since the predefined place for this hybrid system is the roof of the FER building C model was used to check if there is acceptable placing available on the rooftop. The geometry of building C is simple but nevertheless the wind varies significantly.

   The results are depicted in the following Figure 27.

Figure 27 Wind passage over the C building (h=55m) in north-south direction

   The position was chosen based on two most frequent wind directions, north-west and north.

   The system will be installed like shown in the scheme below (Figure 29). In the figure it was assumed that all the connections can be made directly, without the need to install some current sensor for example. Also for efficient control certain switches will be needed.

   Basic idea of the control of this simple hybrid system is to charge the batteries depending of their SOC (State of Charge) and the batteries would power the DC load. Production of the wind aggregate and PV array is measured and the data is stored. Also the possibility to transfer suffice of power produced to the grid is planned. This would be the case when the batteries are fully charged and the production is high.

   PV array is connected to the MPPT (Maximum Power Point Tracker) regulator. The battery is also connected to the regulator over the DC bus alongside with the MATE controller. In this case the controller decides based on the production rates and SOC level of battery whether to charge the batteries or to relay the power to the grid. The same case is with wind aggregate. Wind aggregate is connected to its own hybrid controller. This controller is also connected through the DC bus with the batteries. Finally to enable the transfer of power to the grid and from the grid in case the hybrid production is not enough to power the load inverter is connected to the DC bus and on the other side to the AC grid.

   In order to efficiently manage the control more equipment is needed such as different voltage and current measurements (Hall sensors for example and switches and simple control units). As it was already stated the goal of the design of this small hybrid system is to constantly power the load from the DC bus and in the same time have the ability to transfer power to the AC grid through the inverter. Parallely all the measurements are logged and all the production rates and desired variables are organized and stored on the central SCADA system run on the server.

Figure 29 Basic wiring scheme of the HybridC system (RE2)