 So, welcome to the second class on the course on topics in power electronics and distributed generation. In the last class, we briefly covered some of the issues related to the course. One of the issues was about comparison of central power plant and the distributed generation for providing energy to a load. We looked at issues of complexity of protection, cost related issues, efficiency related issues, then reliability related issues. As you are aware, there are large variety of DG technologies out there. However, there is quite a bit of commonality in the way in which the power is eventually processed and fed to the grid or to the load. And we are especially when it comes to the interconnection with the low voltage system. So, in this course, instead of focusing on individual technologies, we will be looking at this broader issues that are more common for independent of the particular technology. But, in today's class, we will briefly look at the range of DG technologies that are out there. And in particular, we will look at what are the technology trends related to the power conditioning or the power electronic side aspect of it. Because, the DG can be considered as a bigger system. It might have subsystems related to generation. It might have subsystems related to the power conditioning, interconnection, protection, etc. So, we will take a brief overview of DG technologies today. .. If you look at the DG technologies out there, quite prominent among them is solar electric generation, both photovoltaic and solar thermal. And it is one of the fastest growing electric generation methods today, although from a smaller base. Compatibly, wind energy is relatively mature. But, if you look at wind energy at a distribution scale, you are talking about individual turbines or small turbines. The common way in which large amounts of wind energy is interconnected today is in the form of wind farms. Also, micro hydro is getting a closer look for a variety of reasons. Then, you have tidal generation, wave based generation. You have geothermal energy systems in locations where geothermal energy is available. The most common form of distributed generation out there is actually generator sets. It is commonly diesel generator sets. And these generators need not just be filled with diesel. There can be a variety of fuels and that makes it especially attractive. Also, the gen sets can be modified to actually not just provide electric output. You can have combined heat and power units, where you co-generate electricity and also the thermal requirement, whether it is cooling or heating. You also have micro turbines, which is another way of generating electricity from a fuel source. And you have fuel cells. You have a variety of energy storage systems. Systems out there, batteries, ultra capacitors, flywheels, etcetera. And you can see that there is increasing use of power electronics in the DG systems, especially for the power conditioning aspect. If you look at typical role of power electronics, you can think of a block which might consider the source or the storage element. The output would go to a power conditioning system. The output would be fed to the grid. So, the output would be fed to the grid. Often, the source might be generating power at DC level or sometimes at AC. The power conditioning system, it need not necessarily just be a power converter or inverter. It can be something like a machine. For example, you can have DC machines, AC machines, etcetera. Often, you will find induction machines being used with wind turbines. So, you can think of the machine as a power conditioning unit or a synchronous generator when you are having a gen set. So, the power conditioning unit need not be power electronics. In fact, often machines are cheaper than inverters for a given power rating. So, often people look at machines much more closely rather than looking at inverter as the default option. Here, it is commonly AC. Mostly, the AC at you are considering low voltage. So, you are talking about 415 volts, 50 hertz or 230 volts single phase grids. You can also have DC. The off late people are looking at DC distribution, especially in applications like data centers, etcetera. So, you can have a combination of AC or DC depending on the configuration or the end application that you have in mind. If you look at the dominant technology that is emerging today, it is actually the solar electric generation. The photovoltaic is an important part of the solar electric generation. The cost of the photovoltaic system has been dominated by that of the panels. If you look at the photovoltaic system, you have the panels and the balance of plant. One third of the cost typically goes into the panels and the remaining into the balance of plant. In the balance of plant, you have the mechanical structures for mounting the panels. You have the power electronic inverters, the interconnection which could be switch gear, metering required for measuring how much power is actually generated. There will be wiring requirement from the panel to the power conditioning system and from there to the outside. Depending on the type of application, you might also need energy storage in the system. What I mentioned is that the panel cost is a dominant cost today. If you look at the overall cost of photovoltaic system, today people are talking about cost in the range of 1 and a half lakhs per kilowatt for the overall system with panel cost today being of the order of as cheap as 50,000 rupees per kilowatt. Four years back, it used to be a panel cost by itself would be the 1 and a half lakhs. So, you can see that the panel cost has reduced quite drastically by a factor of at least 3 over the last four years. If you look at the remaining aspects, the cost have not fallen that drastically. The cost of power electronics in such a system is actually on the smaller side. It is typically in the 10 to 15 percent range. So, you might say for a kilowatt, you might pay up to may be 20,000 rupees. So, you might say then the entire focus should be on the panels and the electronics is secondary. But the main issue is if you look at it from a reliability perspective, the cost of the life of the panels is expected to be about 20 to 30 years. Whereas, if you look at the reliability of electronics today, it does not last that long. If you look at one end of the spectrum, today's laptops, cell phones, etcetera, they get obsolete in a couple of years. So, and things stop functioning. Even high end electronics, you might be able to work for a longer duration, but you might have to replace some semiconductors or capacitors, etcetera. So, if you look at typical systems, you are talking about say, may be 5 years, may be 10 years. But you can see that if you have cost of a system which cost 20,000 rupees, but it has to be replaced every 5 years, the overall lifetime cost can actually accumulate quite a bit. If you look at the details of the inverter topology in this photovoltaic systems today, the earlier topology was a centralized inverter, where you take all the panels, you connect the panels together in arrays and you bring them to a single inverter and use that inverter to actually send power out to the grid. So, you have issues of if the inverter for some reason is down, then you do not have any power that is sent out at all. The second thing is if you look at the maximum power point at which the overall array can work, it would depend on how the cells are connected together. You might have issues of things like partial shading, which might not allow the array to go at its best performance. So, people have looked at connecting strings on one particular inverter and having parallel such inverters. So, that approach is called the string inverter approach. People are also looking at having one inverter per module, which means that you could actually operate at the best power level on a per module basis. So, there has been a evolution in the topology of the inverter, that people look at advantages and disadvantages of going across this particular spectrum. An important part of the inverter is also packaging consideration. For example, if you have a solar inverter, which is sitting indoors, the thermal issues are relatively mild compared to a solar inverter, which you want to mount along with your module outdoors. Or if the entire central inverter has to sit outdoors, you have to deal with it and package it in a manner, which will handle the heat, the dust, humidity, etcetera outside, which makes thermal management more challenging, which in turn would affect the reliability. So, packaging consideration of the inverter is also an important issue. So, the other aspect is, if you look at the largest solar photovoltaic installations today, commonly it across the world, people are looking at grid connected, but in situations, where you are operating in a remote situation, where you do not have the grid. Or if you have a grid, which where the power quality of the grid is not good, you need energy storage and people are looking at dual mode systems, which mean that you can operate either in a grid connected mode or in a standalone mode disconnected from the grid. If you look at the solar thermal electric technology, it consists of large reflectors, which concentrate the solar thermal energy and make use of some fluid to actually drive a turbine and a generator and send the power to the grid. So, in many ways, it is similar to the technology of the traditional generator, depending on how much power is collected. Typically, solar tracking is done in this technology, which means that depending on the time of the day, you would actually track the sun. You also have to track the peak power that can be extracted from such a system, so as to fully utilize it. You have a variety of technologies for the dish collector, you would have parabolic collectors, you have dish reflectors, you have a frenel reflectors. Frenel reflectors are essentially small planes, which together form part of a dish, so it is more easily handled. So, the operation of such a system is similar to what you do with a thermal plant, where you have a hot fluid, you generate use that to run a turbine and then a generator and the power level depends on what is available from the sun. People are also looking at the possibility of storage in such a system, where you can make use of the collected energy during the day and you store it as stored thermal energy in things like molten salts. Then, you use that energy to drive the turbine at a time at night or a time when the solar energy is not available. So, people are looking at possibility of storage even with solar thermal technology. The recent people are looking at solar thermal closely is that the cost of solar thermal is actually lower on a cost of energy basis compared to the photovoltaic and the efficiency of the conversion is also typically higher. In the photovoltaic system, we are talking about efficiencies of 7 to 14, 15 percent for common commercial cells depending on whether it is amorphous or crystalline, etcetera. Whereas, you could get efficiencies of the order of 25 percent in the solar thermal. So, you end up having better cost of energy. So, it can be quite cost effective. The challenge in the solar thermal system is to make it truly distributed. Often, the solar thermal system is actually large installations, where you have large collection points from which you generate power almost on a centralized basis. It is difficult to scale many of these collectors tracking systems, so that you could put it on say individual rooftops or on commercial establishment. So, to make it truly distributed is actually a challenge. If you look at the wind technology today, it is compared to the photovoltaic is much more mature. In India, I think today the amount of energy generated from wind technology is actually comparable to for example, nuclear power. So, it is I would say fairly mature in terms of what capacity is out there. If you look at the evolution of these wind turbines, it started off with some things in the order of kilowatts in the 80s to megawatts today on a individual turbine basis. If you look at the wind turbine, the reason why it has scaled up to higher power level is if you look at the energy captured by a individual turbine, the energy captured is proportional to the area covered by the blades of the turbine, which is proportional to say the dimension square like radius square. Whereas, if you look at the material that goes into the construction, it is proportional to the dimension. So, you have cos that scale with dimension, but energy which scales with the square of the dimension. So, if you make it larger, you could actually bring down the cost per unit of power produced. So, the limits on how big it gets is actually related to constraints like how strong a material you could actually use for constructing the turbine to make it even bigger or you could have limits on what could be the speed beyond which your acoustic noise becomes too severe. So, even though it is rotating at slow r p m's the turbines, the tip speeds can almost reach to close to the speed of sound. So, you could have really loud noise coming in from things like the tip surfaces etcetera. So, the limits on how large you can go is actually limited by the technology of the materials that and that go into the turbine. And today commercial turbines are available in the 1 to 3 megawatts quite commonly and there are research turbines, prototypes that are installed in the tens of megawatts on a per turbine basis. People have looked at individual turbines that go up to 25 megawatts etcetera. If you look at the small wind turbines, you are considering turbines which can be connected to the distribution. Typically, you are talking about small turbines which are rated at less than 100 kilowatts. In terms of the mechanical structure of wind turbine, you can have a variety of mechanical configurations. You can have the common variety is the horizontal axis what we commonly see. You can also have vertical axis turbines. The commonly seen turbines are the 3 bladed turbines. You can have multi bladed turbines. If you look at the really old turbines that were used for pumping water etcetera, you would see a large number of blades and they would rotate at lower speed, but have a sufficient torque to pump water. You can actually also reduce the blades. You can have 2 bladed turbines. People have actually built prototypes of single bladed turbines. If you have just 2 blades, it is easier to carry. It is just in one dimension. Single blade again reducing components. So, people have looked at a variety of blade configurations. In terms of operation, the earlier turbines used to be fixed speed, stall regulated. What it means is that the speed would be dependent on the grid frequency, the motor that is connected and the gearbox. The limitation of power would come in by the design of the blade. For example, if the wind power becomes more, the blades were designed to stall rather than capture the increased speed of the wind. They are called stall regulated turbine. The other option with the blade is rather than keep the blade at a fixed position and design the blade to stall. You could physically change the pitch angle of the blade. If you have excessive power, you could actually increase the pitch angle and let the power not be captured by the wind turbine. They are called pitch regulated blades. You could also off late. People have looked at variable speed operation. This is what is commonly the state of the art today, where you not operate at a fixed speed, but you operate at speed which gives you the best energy capture. So, even at low speeds, you can actually operate at a lower speed and capture the best possible energy under low speed conditions. When you go to high speed, you pitch out the blades and reduce the amount of power that is captured. If you look at the power train, the power train of a wind turbine is essentially the main path of power flow starting from the blades where you have the energy from the wind being converted to the mechanical energy typically through a gear box, a step up gear box. So, if you are having blades that are rotating at of the order of 20 R p m, you and say your machine over here is running at of 50 hertz. Say at 1000 R p m, you would need a gear box which is having a ratio of the order of 1 is to 50. So, there is a large speed step up ratio that is required for the gear box. When you talk about 20 R p m, it means that in a minute you have 20 rotations. So, for one rotation, it takes 3 seconds for the blade to go around which appears slow, but because the dimension of the blade is really large of the order of 50, 70 meters etcetera, the tip speed is actually quite large even with a slow rotational speed of the blade. So, you typically have a gear box and the earliest technology where you would have a fixed induction machine used that is used to connect the output of the gear box to the electric grid and the electric grid being 50 hertz. It means that the synchronous speed of the machine is fixed once you fix the pole number of the machine. You could have may be a couple of fixed speed points depending on the pole number. You could have a 6 pole slash 8 pole machine. You would also typically add a back to back connected thyristors to limit inrush during startup of the machine etcetera and this was the typical early configurations of wind turbine. If you look at wind turbines today, the state of the art you replace the induction machine, the squirrel cage induction machine with a doubly fed machine. So, you have windings on both the stator and the rotor. The stator is connected to the grid and the rotor is connected to the grid through back to back power converters. So, this type of configuration is what is called slip energy recovery scheme in doubly fed machines. The power handled by the rotor is proportional to the slip speed. So, if your slip speed is small range say 30 percent of your nominal speed, then the power rating required by this power converters would be a smaller percentage one third of the power required for the rated power of the overall turbine, which means that you would need smaller rated converters over here. This is actually the state of the art. If you look at commonly available commercial turbines today, there would be doubly fed machines with back to back connected converters with the rotor energy recovery scheme. If you look at what the people are looking at today, as you go to higher and higher penetration levels of the grid and when you are dealing with weaker grids, people are looking at instead of handling the just the rotor power with a power converter, you replace it with now a fully rated power converter, which can handle the entire power. Now, if you have a fully rated power converter, which handles the entire power, then the generator can be replaced with may be a synchronous machine or a permanent magnet machine, which has higher, which can have higher efficiency than an induction machine. Also, if you design now a special machine, which can run at lower electrical frequency, then it means that potentially you can eliminate the gearbox, because this particular machine can be made to run at a rated speed of may be 20 rpm itself, rather than at 1000 rpm or a higher rpm. So, you can potentially eliminate a large element such as a gearbox in such configuration and this is being looked at more closely by many manufacturers as potential as turbines that can be installed on, especially in the condition when you are having high penetration of wind. Turbines already in the system and you want to install more turbines or when you have weak grids, where the power that is put out by the turbine can actually cause problems to the grid, then having a fully rated power converter will allow rapid control of power flow, which will help stabilize the grid. So, if you look at the technology trend that was there, it is going in the direction of one increased energy capture from the wind. The second is to improve reliability. For example, if you can eliminate the gearbox, then the issue of gearbox reliability goes away. You also have high grid penetration issues. For example, not all parts of the world have the best wind resource. If you look at India, where you have the best possible wind resources available, you are talking about places like maybe Gujarat, Tamil Nadu, etcetera, you already have a large number of turbines. So, if you want to put more turbines in such a location, then you need to have better ways of dealing with the grid. So, that is another trend, which is demand requirement, which is being imposed from the grid side. So, granting some of the requirements like dispatching reactive power to support the grid, responding to grid frequency changes. So, if the grid frequency rises suddenly, you could cut down on the power that is sent out to the grid and help stabilize the grid. Obviously, the turbine cannot increase the power beyond what is already available from the wind, but you can actually reduce the power and still operate. Also, another issue that has become important is fault right through. So, for example, if you have a fault under grid and all the power converters shut off, then you lose a large amount of generation at one go. So, you need to have turbines with power converters that can actually stay in there when the fault protection logic of the power system operates. Power system protection typically operates in the many fundamental cycles time frame. So, you are talking about hundreds of millisecond whereas, if you look at the protection of IGBTs in a power converter, you are talking about a micro second range. So, you need to operate your power converter without tripping for many hundreds of milliseconds range, which is a challenge. Also, another challenge is to look at whether you could make use of advanced metrology. So, you have better weather prediction, better prediction of what the wind conditions are and use that to actually plan on how much wind power will be there, which means how much other resources would be needed. So, combined dispatch ability is something that people are looking at especially with power electronics, you could actually adjust things at a faster rate. Also, people are looking at offshore wind technology, large number of installations out there in the world. If you look at the load centers of the world, the bigger cities of the world in India, Bombay, say, Cochin, Chennai, Calcutta, they are all coastal cities and those are the load centers. So, if you look at London, New York, etcetera, you can have the power generation source cited close to the loads, then you have a better possibility of feeding the load from a source that is close by. So, the offshore wind technology is actually quite appealing. However, with offshore technology, you need higher reliability. For example, if you have a turbine, which is down on land, you can send a service person to go and repair it. Whereas, if you have a turbine, which is sitting in the Arabian sea in monsoon season, the rough sea might prevent you from repairing it during the monsoon season. So, you would need higher reliability for actually doing things like proper servicing and power collection is also an issue. At sea, you need to collect the power and transmit it to the land typically through subsea cables or some form of DC conversion and transmission, etcetera. People are also mechanical challenges of how to install the turbines in the water. People are looking at the technology from offshore oil and gas platforms and making use of that to actually use that in offshore wind technology. Another technology, which is coming up and is quite appealing is micro hydro. If you look at the large dams, that is already the best sites are taken. In the Indian context, you have a large dam, you submerge a lot of land and then there are rehabilitation and relocation issues. So, micro hydro is actually quite appealing in that context, where you are taking the traditional hydro system and trying to simplify it. If you look at the simplification, the big cost in hydro is actually the civil, the construction of the dam and the lands submerged, etcetera. So, if you can eliminate the construction of the dam, that eliminates a fairly big cost. So, if you eliminate the dam construction, what you are doing is you are taking water at a increased height and taking a pipe and draining it out at a lower height, where such possibility exists and making use of a fixed head. You do not have storage, but you have a fixed head depending on the height of the intake and the outlet. The second reduction of cost can be in terms of say, to control the power in a traditional dam, you have governors which regulate the amount of water that is going in, which would regulate the amount of power that is generated. So, if you have a fixed height and if you are not having a governor, it means that your power output is constant. If you have loads that are connected which are variable, if you can have dump loads such that the sum of the load, actual load plus the dump load is constant, then you do not need a governor. You are always operating at constant power output from the machine, because the total load is constant. So, you could potentially eliminate another costly system such as the governor. Then you could also for example, replace the machine, induction machine is cheaper than a synchronous machine. So, replace a synchronous machine with an induction machine, which means that then you have cost reduction in terms of the exciter not being there. The excitation in this case is then provided by a bank of capacitors which would, so it would be a self excited induction machine. So, for a variety of these reasons people are starting to look at possibility of small hydro where it is possible. In India you are talking about the western, eastern guards. So, in the monsoon season when water is available, you could generate power. Similarly, in the Himalayan mountain range, you could actually generate power with small hydro and people are looking at these as possible solutions. So, if you look at what would be the role of poly electronics potentially in such a system, you could say for example, if you take a load typically it is not just real power you might also need reactive power, which means that the total output of the induction machine, the voltage would not stay constant even though the total real power stays constant. If you add more inductive watts from the load, your voltage might drop, which means that you cannot have just a fixed capacitor, you would have need a bank of capacitors. Also by rapidly responding to the load requirements, you could improve the power quality from say a power electronic connected converter that is used in parallel. Also if you have limited some amount of storage available in the power converter, if the load on a transient basis for example, on a startup of the load, you need a higher level of power than what is available from the machine. Then potentially the storage element could provide a transient power requirement keeping the average power coming in from your induction machine, the micro hydro and the power electronic device providing you the dynamic power requirement as and when required. If you look at wave and tidal energy, these are similar to in a way if you can think of it as something that is similar to wind, wind is a fluid which is not that dense, water is a fluid which is much more dense. So, if you look at the turbine required in a tidal application, the diameter of the blades in the tidal turbine would be much smaller because of power density available in the flow of water is actually much higher. So, you can think of there is some similarity to what is being done in wind and what could be done in the tidal or the wave turbines. So, you have the typical effect that the water flow does not change as dynamically as the change of wind. So, if you look at the tidal regions and wave energy regions in India, the Gulf of Kutch is a good location in India for potentially harvesting wave or tidal energy. Similarly, the Park Strait, the Gulf of Manar area, they are good locations for potentially harvesting tidal and wave energy. The challenge in these systems is that you have now turbines which are submerged in water. So, it is exposed to sea water which is has salt, it is quite a corrosive environment. So, packaging is a important consideration. If you look at geothermal, again depending on where the resources are available, I think in India there are geothermal stations in the Himalayas, a smaller basis. So, depending on where the resources are available, people are looking at the possibility of geothermal. In all these systems, you have turbines that drive generators to produce power. If you look at the role of power electronics in these applications, it is still in the evolutionary stage and the starting stage. If you look at the most common distributed generation technology that is actually out there, it is the gensets that are commonly used. If you go to any commercial area of the Indian city, you would have people who have gensets outside their shops. If there is a power voltage, all the gensets will turn on. So, there is actually a very large installed capacity of gensets out there. They are not connected to the grid, they are acting as backup systems. The cost of the genset is quite low, because it leverages the IC engine technology from automotive applications. So, the cost is actually quite low. If you look at the system, you have the fuel. The fuel can be a variety of fuel, diesel, natural gas, propane. It can also be landfill gas, fuel that is flared out in some chemical processing, industrial byproducts, etcetera. It could also be agricultural byproducts, ethanol blends, biomass, gasified biomass or biogas. So, you could have a variety of fuels and in the genset, what you do is you try to provide a fuel so as to regulate the speed, the output of your machine, so that you maintain your 50 hertz frequency required by the AC grid or the load. So, you would have this particular machine operating at constant speed depending on the required frequency, say if this is 50 hertz, you would have the corresponding speed depending on the pole number of the machine. You would also have a exciter system. So, with standalone loads, you are ensuring that you are maintaining the required load voltage or if you are operating it connected to the grid, you are using the exciter system to control the reactive power flow from the machine to the outside. So, one constraint that you can immediately see is that irrespective of the load that is connected or the power that is sent out, the IC engine has to operate at a fixed speed at light load or full load depending on the frequency that is to be generated at the output, which is in a way, area where power electronics can come in and that is essentially by allowing the IC engine to operate at variable speeds. So, you take the output from the IC engine through a machine rectified and then you use a inverter to provide the fixed frequency, which means that potentially now at light loads, you can run your IC engine at a much lower rpm. So, reducing the fuel consumption in the IC engine. So, you could see that also in many applications, your gen set might need to run continuously if you want to provide backup in a rapid basis. So, if you start a gen set, even fast starting gen sets, you are talking about maybe 10 seconds for it to start up. If you want to transfer to a gen set in a time duration shorter than 10 seconds, then you would run the gen set at almost close to no load and whenever there is a disturbance, you would transfer to the gen set. Also, many times when you size a gen set, you will size it for your maximum load and most of the time you will be running it at a partial load or at a light load, which means that you are not making the best possible use of the gen set in terms of operating at the best possible efficiency point and by being able to run at variable speed, you can actually operate now at the best efficiency for a given power that is being generated. So, also when you run a gen set at a fixed speed, your acoustic emission is actually related to the speed of your machine. You can see when you drive a vehicle at high speeds, the sound from your engine goes up quite a bit. If you can actually run it at low speeds, the sound can come down. So, for a variety of reasons, this variable speed gen set technology is actually quite appealing and there are manufacturers out there who offer variable speed gen sets. You can also say for example, in this particular slide, you look at machine, permanent magnet machine which is rectified along with the inverter to actually feed the loads. You can also have full and partial rated generators, similar to what we discussed in the wind application by having things like doubly fed machines to actually feed loads and even that has been manufacturer in Bangalore actually has started building doubly fed machine based variable speed gen sets. Another appealing aspect of gen set is when you are not just looking at the electrical output from the generator, but if you can potentially make use of the waste heat that would be invariably generated from the engine. So, if you look at the efficiency of conversion from fuel to electrical, you are talking about numbers in the 30 to 40 percent range. So, if you are only generating electricity, you are just making use of this part of the energy that is there in the fuel and there is a large amount of waste heat. So, if you could actually tap the waste heat for say a heating requirement or also for cooling requirement, then you might have traditionally the heating being done or air conditioning being provided by a compressor which might actually consume electrical energy. Now, because now the heating or cooling requirement is from the waste heat, the electric consumption can actually come down. So, that becomes an added advantage which means that now your fuel to your end use of both electrical and thermal energy can go up even above 70 percent compared to the 30 to 40 percent if you are just looking at making use of the electricity. Obviously, now for this you now need to have additional systems like heat exchangers, chillers, adsorption chillers etcetera and they are used for a variety of applications. For example, when you are a common application of such combined heat and power is in say a production of sugar where you have to actually boil the juice, the extracted sugar cane juice to actually concentrate it at higher temperatures lower pressures to actually get the sugar. So, you could actually make use of the waste baguette to actually drive your genset and then make use of the electricity to produce electricity to meet the requirements of the plant and make use of the waste heat to actually raise the temperature of your the syrup that is being concentrated. So, there are applications in sugar, steel, food processing where you also have data centers which need cooling, your hospitals where you need both hot water cooling, critical installations. If you look at the cooling requirement, typical cooling that will be done in say a refrigerator is with a compressor. So, that is one way of achieving cooling. The way for example, our body cools itself down is by adsorption cooling where essentially it depends on the phase change where you sweat and your sweat becomes vapor rises from the liquid state and the phase change actually cools down your skin. So, you have similar adsorption cycles with special fluids like lithium bromide, ammonia etcetera, which can actually make use of the waste heat to actually provide cooling to your end requirement. So, you could actually make use of the waste heat from the genset along with things like adsorption chillers for combined heating, cooling and power. If you look at the other technologies related to which consume fuel, another technology is the micro turbine technology. If you look at its difference with the IC engine, IC engine is producing pulses of power each time you ignite the fuel in the cylinder. Whereas, in a turbine you are continuously burning the fuel in the turbine combustor and again the micro turbines can be used with a variety of fuels. Because it is a continuous burn, you can optimize the burn to reduce emissions and pollution and because of the continuous burn, you can also have uniform power generation rather than pulsed power that is being generated or pulsed torque that is being generated, you can improve the reliability. However, the capital cost of a micro turbine is more than the IC engine because it is not able to leverage the technology that is commonly there in automotive applications. In fact, you can think of a micro turbine as a scaled down jet engine. So, its power scale can be from kilowatts to megawatt range and once you have the output of the micro turbine, the turbines are spinning at fairly high rpm, tens of thousands of rpm. So, you typically have a machine, you can have a high speed permanent magnet machine, which generates AC, you rectify it and then you use a inverter to actually feed power into the grid or feed the load at 50 hertz. So, in the next class we will look at some of the other possibilities such as fuel cells, storage and today we have looked at a range of technologies of distributed generation. If you look at the role of power electronics in these systems, it is actually increasing over time and the role of power electronics is to improve the energy conversion efficiency, the energy capture, to improve the type of power quality that is output from the DG systems, to improve reliability, to meet a variety of such requirements, the role of power electronics is actually increasing in these DG systems. So, in the next class we will discuss a few more DG technologies and take it from there. Thank you.