The need of the hour today is the transcendence from the fast-depleting conventional energy resources to the cleaner and greener alternative sources of energy. The universal need for a smart, self sufficient and healing, and efficient energy management system has led to rise in smart microgrid system technology. Initially microgrids composed of DG were used only for the self healing purposes, i.e, these DGs were used to supply the load region with fault and break down. The DGs supply the partial load at times of faults and the time consumed for healing purposes. However, today the microgrids are also used to supply the main course of power share. This is due to the advancements in power electronic industry as the modern power electronics can be used to overcome all the power quality issues.
Introduction: The load pattern in India is very non linear and is hugely dependent on the various factors. As shown in the fig, The Blue line shows the load profile which seems highly nonlinear and similarly the distributed generation profile as shown in fig2 (in red) shows the variation in generation which is also hugely non linear. Let us analyze the power system supply and generation for various phases of the graph. At point P1, the generation of the power is equivalent to the requirement of load and thus could be solely be supplied by the distributed hybrid generation and storage units and thus for the system, there is no need for connectivity to grid.However at point P2, The Generated power is greater than the load power required, thus there is a extra power generated, this extra power at point P2, Pdg-Pgen should either be dumped or dissipated through the dump loads or if there is a possibility of the two way communication in the particular grid code, then the extra generated power could be sold back to the utilities. This way at point 2, there is an equalization of load and the excess energy is either dissipated or is
sold out to the utility sectors provided there is a allowable two way networking in the grid code of the nation. The grid code varies from nation to nation and the code mainly depends on the requirement of national security level, built infrastructure, national electricity policy and allowable voltage and frequency regulation levels.
In the Point P3, The load is more than the capability of the distributed generation units now, the system should automatically switch to the DC grid and it would be two sources sharing the load according to the capability of the load. At this point through modern power electronics, it must be ensured that the voltages at the bus from each source is equal as, in the other case one of the sources might act as a load and a current produced due to this difference in voltages might flow in the opposite direction.
Fig1: LOAD PROFILE
Fig2 GENERATION AND LOAD PROFILE
There is a need for smart systems which should be capable of the faster and accurate data acquisition of the real time power information like demand, capability of distributed energy systems at the moment, fuel availability etc, and intelligently predict according to developed algorithm and switch between the distributed utilities and if the demand is beyond the capability of the distribution utilities, the load is supplied through the grid supplied by the electricity board. The system needs sustainability, security and reliability. The demand and the generation of the electric power should be coherent as the loss of the energy produced by the carbon based emissions harming the environment should be reduced. The real fact is that both the load and generation profiles in the system are very unpredictable and in the system and this problem remains very much peculiar in the small scale distribution systems and the systems involving external factors like wind, illumination and temperature. Thus both the generation and load profiles are variable and the main challenge for the utilities is to meet demand without losses greater than permissible level being fed to dump loads.
In this project, a hybrid distributed generation system of photovoltaic array, battery backup and diesel generator has been employed to supply to the load of 100V DC, 5A. As shown in the block diagram, (ref fig1), the system is an hybrid combination of various distributed generation units involving power electronics to equalize the voltages at the bus so as to avoid circulating currents in the bus. The photovoltaic array is a typical 80Watt, 21V output array which is boosted to 100V DC using a DC-DC converter employing PWM technique. This 100V system is thus connected to the main bus bar supplying the non critical loads. The loads have been categorized according to the priority and the quality required as per the demands of the user into critical and non critical loads. The critical loads are the ones which require good quality and huge quantity of powers and are prone to damage the equipments if it meets the transients. These loads are to be supplied by the grid supply which has lesser issues with the quality constraints.
The other loads which are not susceptible to the quality issues like the domestic lights and fans can be supplied by the distributed generation units. This project is a design of dynamic system which instantaneously acquires data within the specified amount of delay and intelligently changes parameters so as to optimize the cost of the electricity and reduce the emissions. Initially, the systems acquires the data like the capability of photovoltaic generator array, the battery bank system discharge position, the average predicted illumination of the day, the fuel levels in the diesel generator, the load requirement etc.
The processor has some predefined values prefixed in it which has to be changed when there is a change in the system generation components. The prefixed entities which are primarily required are the capacity of the battery, the maximum load demand at any time and the system bus voltage. In the step1, the comparison is made so as to check if the photovoltaic power output (Ppv) is able to meet the demands of the load at that instant (Pp2). If its able to meet the demand then the battery position is checked and if the battery discharge level is lesser than what it is calibrated for, then the excess amount of power is used up to charge the battery. Thus battery will act as an additional variable load which could equalize the demand and supply curves not allowing the power dissipation to earth through dump loads.
Now, if the photovoltaic power output is unable to meet the requirement of the load at the instant, the processor puts the bidirectional DC-DC converter connected as an output of the battery bank to the boost mode to discharge from the battery to supply to the load. The processor then checks if the load is supplied with adequate amount of power and if not then it maintains the mode of operation of the bidirectional converter to boost and then triggers on the diesel generator. The diesel generator is put into the system so as to improve the reliability of the system. The diesel generator is preferred over the fuel cells because of cost and overall efficiency of the former is much higher than the later as for the fuel cells, the hydrogen fuel is to be consistently supplied and the chamber is to be maintained at the requisite temperature. Thus diesel generator has been used to boost the reliability factor involved in the power plant.
The base load which is usually 25-30% of the peak load (0.25P2max), for a more reliable system, the base load is to be supplied by the electrical grid and the
other load is to be shared between the DG units, storage systems and the grid.
The whole project has been verified in Matlab Simulink and the code of the algorithm has been applied in C, Mfile, Labveiw and verified. The Data acquisition was tried and performed through different systems like ARM microcontroller, PIC microcontroller
System Block Diagram
and LabVeiw DAQ card. The most precise, accurate, user friendly and with least time delay was found with the LabVeiw DAQ card and thus we have employed it for the model verification.
As seen in the fig2.0, the storage element, battery was connected to the bidirectional DC-DC converter, which could either function as buck converter or boost converter according to the reference signals given by the controller. The battery is capable of giving an output of 50V under fully charged condition up to nearly 78% of the battery discharge. Thus the bidirectional converter shall act in the boost mode up to this point and then automatically switches on to the buck mode where it sucks current and the battery goes into the charging mode. Refer the nominal battery discharge profile. The voltage at the bus is always maintained as 110-100V, as the simulation results suggests. The charging or discharging of the battery is decided by the external equipment and when it reaches 78% of the discharge level, the external equipment triggers the DAQ to send signals to reverse the modes of the battery form either Boost (Discharging) to Buck (Charging) or the vice versa.
The second phase represents the PV source, which is assumed as the constant output source for the simulation purpose the PV gives an output of 50V DC max and Imax of 2.5Amps. The PV panel is connected to the DC-DC converter triggered by the PWM pulses so as to avoid the harmonic distortion and power quality disturbances.
The Alkaline Solid Oxide Fuel cell is chosen for the siumation purpose with a open circuit voltage of 65V DC and nominal operating point of 4A, 50V is set for the purpose. The nominal stack efficiency is optimally chosen as 55% with operating temperature of 65oC.
There is a DC-DC converter in the system to level up the voltage from 40-55V range to 105-115V range and a capacitor of 4000 Micro-Farad is put in parallel to the output of the bus so as to stabilize the voltage profile.
The forth element is the Diesel Generator. The diesel generator might not be advisable as it intakes fossil fuel and also gives out emissions adding up to environmental problem but still the diesel generator is preferred in the system because of the fact that it improves the reliability and is very inexpensive when compared to power supplied by EB and PV. Thus
Fig 2.5 The battery bidirectional DC-DC converter circuit.
Fig2.6- Nominal battery discharge characteristics.
Diesel Generator is also used to supply the peak loads. One specific problem with Diesel Generator is that the frequent switch on and switch off of the plant unit might cause a lowering to the battery life. The delay time between the switch on and switch off of the generator should at least be half hours. Thus if the diesel generator is switched on, the unit should at least run for half hour to support the extended life of the generating equipment. Thus if the load profile changes in the mean delay time of half hours, the power is to be supplied only through this generator and other options are to be considered later.
The power electronics interface of the diesel generator is slightly different as a single phase synchronous generator is considered here and it supplies an output of 50V AC, single phase. The output power is then connected to the DC-DC converter through a freewheeling capacitor of 200 Micro Farads. This capacitor also acts like an AC filter and regulator to reduce the ripples of pulsating DC. The DC-DC boost chopper boosts the voltage levels to 110V level to equalize the DC bus voltage level.
Main Circuit of SEMS:
The Simulation results:
Bus Voltage and Load Current profiles:
The above curves show the voltage and current profiles of the system at the 110V DC bus and resistive load of 25Ohms. The voltage fluctuates between 108.8V to 106.7V that implies a voltage regulation lesser than 2% which is fairly acceptable in Indian Grid Code. The current varies between 10.85 and 10.7 Amps. A regulator of higher rating will give a better regulation.
fig 3.3 Output of PV and FC: (fig 3.3)
Output of Diesel Generator (fig 3.4)
The outputs of the FC and DG where we can see that each of them partially sharing the load current requirement.
The Data Acquisition and Control through LABVIEW- DAQ card:
LabVIEW (short for Laboratory Virtual Instrumentation Engineering Workbench) is a system design platform and development environment for a visual programming language from National Instruments. It is commonly used for data acquisition, instrument control, and industrial automation on a variety of platforms including Microsoft Windows, various versions of UNIX, Linux, and Mac OS.
The National Instruments PCI-6221 is a low-cost multifunction M Series data acquisition (DAQ) board optimized for cost-sensitive applications.
ï‚· Product Name: PCI 6221
ï‚· Analog I/P Channel: 16
ï‚· Resolution: 16
ï‚· Sample Rate: 250kSamples/Sec
ï‚· Max Voltage: 10V
ï‚· Max Voltage Range: -10V to +10V
ï‚· Max Voltage Range Accuracy: 3100μV
ï‚· Max Voltage Range Accuracy: 97.6μV
ï‚· Output Channels: 2
The following VI was made in LabView 2010 to emulate the above explained algorithm. The VI
simulates 4 Sources of Energy namely Photovoltaic Cell, Fuel Cell, Battery Backup and Diesel Generator. The VI verifies the algorithm via LED's which represent the switches in the circuit. Therefore when an LED is turned ON it represents the switch is in ON state.
S2, S3 and S6 represents 3 MOSFET switches used to control power flow out of the 3 sources namely Diesel Generator, Fuel Cell and Grid Supply. The Photo Voltaic Cell is always connected to the load as the Primary Source of Power. The Grid connection which is controlled by switch S6 is used as a standby when all the available sources cannot supply the demand of the load connected. The toggle switch 'opt' is used to represent a choice which the user has between using the diesel generator and the fuel cell since both have their own associated advantages and disadvantages. Fuel Cell in our current scenario is expensive in terms of cost but it represents a cleaner, more environment friendly option as an energy source. Diesel Generator is comparatively cheaper now but as our fossil fuel reserves run out it will become economically unviable. So the switch represents a choice which the user has to choose between the available options.
G41 and G42 represent switches that control flow of power to and from the Battery Source. A bidirectional converter is attached to the battery to enable the battery to charge by utilising any available excess power from any of the sources. G41 represents the Gate pulse required to operate the bidirectional converter in Boost mode i.e to allow the battery to act as a source. This mode is only operated after checking the current status of the battery i.e whether it contains more than 40% charge or not. In case the battery charge is less than 40% and there is excess energy available from the PV cells then it is redirected to charge the battery by operating the converter in Buck mode. This is represented by the switch G42.
The VI shown in fig 4.1 was simulated using LabVIEW 2010. Various test case were given containing different values for the current available from the various sources. The corresponding switch outputs were observed from the LED outputs to check if the algorithm is working correctly for various scenarios.
Fig 4.1: Software Algorithm for the logic.
DAQ in Hardware Implementation:
The hardware implementation of the model was done in a 12V system using the vi shown in the figure below. The inputs were acquired via ports ai0, ai1, ai2 and ai3 of the DAQ card. The inputs can be acquired in the form of voltages across a fixed 2Ω resistor although a current sensor can also be used for the same purpose. The outputs were given from the analog output ports of 2 DAQ cards named dev1 and dev2 via ports ao0 and ao1 of both instruments.
Since no comparison or Boolean operations can be performed on the inputs acquired from DAQ port hence all the comparison operations are reduced to a series of subtractions or other arithmetic operations from a reference quantity. The results are compared to zero to determine if the difference is positive or negative which in turn determines which input is greater in magnitude and which is smaller than the other. This is then used to trigger appropriate outputs which in turn trigger the gate pulses to the switches via optocouplers which provide electrical isolation between the gate circuit of the MOSFETs and the DAQ circuit. For triggering the switches a positive DC signal of +5Volts is given to the
appropriate output port. To turn off a MOSFET there are 2 options available. Either a signal if -5Volts can be given from the output port of the DAQ to act as forced commutation for the MOSFETs. Else a signal of 0Volts can be given to allow the switch to naturally commutate.
Therefore the VI shown in the picture below was used to practically implement the algorithm in the 12V system. Various load configurations were tried and the sources were appropriately found to be supplying the load requirements. The full switching strategy was also verified and the load sharing was found to be appropriate.
Alternative Control Techniques:
Apart from the DAQ to control the switches alternative options were also explored. This included use of PIC Development board with relays as well as use of ARM processors. Code was developed in assembly language for PIC 18 as well as ARM. However there were constaints regarding the resolution available from the input terminals of the board since they cannot distinguish between float inputs. DAQ cards allow for more resolution in the input data. Secondly use of the boards would involve interfacing various hardware components such as sensors and ADC using assembly language. Hence these approaches were abandoned in favour of DAQ card due to various resource and time constraints.
At first, a 12V model was tested and verified using the same logic mentioned above. The model showed quick responses whenever the load was manually changed. The system also responded and Fig4.0 LabView DAQ circuit and control model for hardware.
automatically switches between sources based upon load demand and capability of the system to provide the load at the particular instant. (As shown in the video)Then the model was also verified on 100V System and appropriate results were noted accordingly.
For switching between sources, MOSFETs (IRFZ044) have been used and have been triggered according to conditions through the gate pulses provided by the LabView DAQ card which feeds an input voltage signal from a very small resistor which drops a voltage of (0-5)V and thus DAQ reaches the signal and according to the magnitude of the signal takes appropriate action by means of the programmed logic in the lab view.
Project SEMS was successfully finished and verified and many important aspects have been inferred during the project. The importance of load sharing and voltage equalization at the bus has been practically seen and corrected using apt power electronics. The scheme used in SEMS can be implemented in the practical system so as to increase the system efficiency and decrease the overall cost per unit of electricity consumed. If two way communications is allowed, this logic also allows the users to sell back the excess energy to the grid and thereby earn while not using the loads. The importance of this project is that the future of electricity would majorly depend upon renewable energy, which is very expensive in terms of cost of installation and also the transmission losses. Whereas in this project, as at domestic level the consumer can
afford hybrid DG set (*Hybrid DG set is combination of two or more DG units so as to improve the reliability) and the huge transmission losses are avoided. The SEMS Controller has to be made cost effective and a specific microcontroller to function only to switch between sources and vary the firing angle is to be designed. DG units also are capable of increasing the power quality of the main power by providing ancillary services like reducing power quality disturbances, reactive power control, maintenance of voltage profile etc. The Circuit was connected accordingly and equal sharing was found initially when the sources were not disturbed and loads were changed. Then, the second phase was to change the supply capability and that was accomplished by connecting rheostats in parallel with the autotransformer and varied the same so as to distribute the current level among two of its branches. The share was uneven when the source capability was differed and various readings in the ammeters were seen and when summed up, that equalled the total load required by the loading rheostat.
Thus both variable load and variable source conditions were tested and verified successfully the same with the logic.
Demo Model: (12V DC bus system)
Main Model: (100V DC Bus System)