In a fully deregulated market, a case study is performed, with the Microgrid connected to the utility grid. Microgrid under consideration has two sources (Solar and Diesel) of generation and battery storage. The trading model uses historic data and various forecasting techniques to get Day-ahead forecasts of solar irradiation, load and the market prices. Based on the forecasts available the model issues directives to all the components of the Microgrid to maximize the Trade balance and in turn increases the return on investment (ROI). It also greatly reduces the energy produced from the Diesel Generator and hence helps in cutting the CO2 emissions making it more eco-friendly.
In economic terms, electricity is a commodity capable of being bought, sold and traded. An electricity market is a system for effecting purchases, through bids to buy; sales, through offers to sell; and short-term trades, generally in the form of financial or obligation swaps. Bids and offers use supply and demand principles to set the price. Long-term trades are contracts similar to power purchase agreements and generally considered private bi-lateral transactions between counter parties. Wholesale transactions (bids and offers) in electricity are typically cleared and settled by the market operator or a special-purpose independent entity charged exclusively with that function. Market operators do not clear trades but often require knowledge of the trade in order to maintain generation and load balance. The commodities within an electric market generally consist of two types: Power and Energy. Power is the metered net electrical transfer rate at any given moment and is measured in Megawatts (MW). Energy is electricity that flows through a metered point for a given period and is measured in Megawatt Hours (MWh). To move to a more open and democratic energy market, certain changes have to be made to the current market. There is very little knowledge of how to properly design a retail electricity market and how to effectively incorporate other services. The energy distribution service requires quality improvements for the new market to function correctly, because of the higher granularity of the energy contracts. Because of this increased granularity they would need to handle a huge amount of operations in the system. Also, searching for the best sellers in the new market with millions of suppliers should be done autonomous. The concept of intelligent control for regulating the power network variables is to be realized. The intelligent multi-agent based control can be a solution in today’s power network to maintain the dynamics such as adequate power balance along with quality voltage under changing system conditions such as load and power injection. The technology with multi-agent intelligent control may be main module of Smart Grid architecture. The idea behind any multi-agent system is to break down a complex problem handled by a single entity – a centralized system – into smaller simpler problems handled by several entities – a distributed system.
Micro grids are modern, small-scale versions of the centralized electricity system. They achieve specific local goals, such as reliability, carbon emission reduction, diversification of energy sources, and cost reduction, established by the community being served. Like the bulk power grid, smart Micro grids generate, distribute, and regulate the flow of electricity to consumers, but do so locally. Smart Micro grids are an ideal way to integrate renewable resources on the community level and allow for customer participation in the electricity enterprise. They form the building blocks of the Perfect Power System. Micro grids allow power generation and consumption to be managed so that the load is balanced with the supply. Smart power meters allow power to be stored in batteries and reused at times of peak demand. This smart approach to managing energy use can result in lower energy costs to the consumer because it encourages making smarter choices about power use.
- Micro grids can be broadly classified into four categories as
- Remote grids, which are necessary due to geographical features, such as islands. Consider a country like Indonesia that has more than twenty-three-thousand islands. It is simply not practical to connect all these to a single national grid. The power sources in these grids are on the same conventional lines are using fossil fuel.
- Military and security are grids necessary to maintain data and security during a national catastrophe. The power sources in these grids are also on the same conventional lines using fossil fuel.
- Commercial or industrial grids catering to a specific industrial community. The energy sources could be fossil fuel based or energy recovered for the process like waste heat, bio fuels, or waste products. These are mainly captive energy systems.
Community grids that optimize and utilize the specific regional renewable resources to give cost effective power supply. Fossil fuel usage is only used as an emergency backup. This is the really effective Microgrid.
We also need to know why Microgrids play an important role in the future. The centralized transmission grid system is definitely the backbone of the electricity distribution system, but has its drawbacks.
- The energy loss is almost 8 -10%.
- There are high investment costs in transmission lines, step-up and step-down transformers, right of way and other legal issues.
- Grid management is a constantly juggling act where it balances the generation and the demand over a wide geographic area.
- The generating capacity has to match the peak load, which means a lot of excess capacity is built into the system, which increases the investment cost.
- All the users feel the grid disturbances, outages, frequency changes and voltage fluctuations, blackouts and brownouts. This can affect the performance and life of electrical equipment.
The Microgrid, even though not a replacement of the national grid, improves certain aspects especially for communities and regions that have adequate renewable resources.
- They have much smaller financial commitments.
- They use renewable resources hence are more environmentally friendly with lower carbon footprints.
- They require fewer technical skills to operate and rely more on automation.
- They are isolated from any grid disturbance or outage.
- They place the consumer out of the grip of large corporations that run the generation networks.
Microgrids are cost effective only if you can tap into locally available renewable energy resources. Solar energy is available everywhere but with limitations. Wind, mini hydro, geothermal and biomass are regionally available and can augment Solar energy. This combined with a storage device, battery or super capacitors and backup diesel generator makes Microgrids highly reliable and cheap. Storage devices in very large grid systems are not economically or technically proven. The advent of latest technologies in nano batteries and nano super capacitors makes electricity storage a reality in the smaller capacity range. This is an advantage for Microgrids. Advances in computerized control technology make it possible to have simple and efficient controls with less human interference and is the key ingredient that makes Microgrids feasible.
High technology products like nano solar cells, nano super capacitors, nano batteries and fuel cells will make the Microgrid with storage capacities a reality. Advances in automation, power electronic control systems will also help in the popularity of Microgrids. Even hybrid cars plugged into the home wiring can act as a generating sources or a storage device. In larger communities mini nuclear plants could be the ideal source of energy for the Microgrids.
Government policies and Effects of deregulation
The Central Electricity Regulatory Commission (CERC) on February 6, 2007 issued guidelines for grant of permission to set up power exchanges in India. Financial Technologies (India) Ltd responded by proposing then tentatively named ‘Indian Power Exchange Ltd’ and applied for permission to set it up and operate it within the parameters defined by CERC and other relevant authorities. Based on the oral hearing on July 10, the CERC accorded its approval vide its order dated 31st August, 2007. IEX thus moved from the conceptual level to firmer grounds. On 9th June 2008 CERC accorded approval to IEX to commence its operations and 27th June 2008 marked its presence in the history of Indian Power Sector as Indian Energy Exchange Ltd (IEX), India’s first-ever power exchange goes LIVE. Within 5 years it gets and Average Daily Volume for Q1 FY 2011 – 20,921 MWh with 86% market share in India. It also took the market price of a KWh to a record low of 13 paisa in November 2009, demonstrating the true potential of supply demand economics. Also, Benefits of having a deregulated market are worth talking about.
Access a diversified portfolio: Exchange offers a broader choice to generators and distribution licensees so that they can trade in smaller quantities and smaller number of hours without additional overheads.
- Payment security: Exchanges stand in as the counter-party for all trades; so participants need not be concerned about the risk-profile of the other party.
- Minimal transaction overheads/charges: All charges are public information and due to the economies of scale the charges are minimal.
- Efficient portfolio management: Exchanges enables participants to precisely adjust their portfolio as a function of consumption or generation. Participants, especially distribution licensees, are enabled to precisely manage their consumption and generation pattern.
Energy trading is a basic economic concept that involves multiple parties participating in the voluntary negotiation and then the exchange of one’s goods and services for desired goods and services that someone else possesses. Energy trading is much more than this and there are a few factors which make is special. The development of electricity markets is based on the premise that electrical energy can be treated as a commodity. There are, however, important differences between electrical energy and other commodities such as barrels of oil or even cubic meters of gas. These differences have a profound effect on the organization and the rules of electricity markets. The most fundamental difference is that electrical energy is inextricably linked with a physical system that functions much faster than any market. In this physical power system, supply and demand – generation and load – must be balanced on a second-by-second basis. If this balance is not maintained, the system collapses with catastrophic consequences. Such a breakdown is intolerable because it is not only the trading system that stops working but also an entire region or country that may be without power for many hours. Restoring a power system to normal operation following a complete collapse is a very complex process that may take 24 h or more in large, industrialized countries. The social and economic consequences of such a system wide blackout are so severe that no sensible government would agree to the implementation of a market mechanism that significantly increases the likelihood of such an event. Balancing the supply and the demand for electrical energy in the short run is thus a process that simply cannot be left to a relatively slow-moving and unaccountable entity such as a market. In the short run, this balance must be maintained, at practically any cost, through a mechanism that does not rely on a market to select and dispatch resources.
Another significant, but less fundamental difference between electrical energy and other commodities is that the energy produced by one generator cannot be directed to a specific consumer. Conversely, a consumer cannot take energy from only one generator. Instead, the power produced by all generators is pooled on its way to the loads. This pooling is possible because units of electrical energy produced by different generators are indistinguishable. Pooling is desirable because it results in valuable economies of scale: the maximum generation capacity must be commensurate with the maximum aggregated demand rather than with the sum of the maximum individual demands. On the other hand, a breakdown in a system in which the commodity is pooled affects everybody, not just the parties to a particular transaction. Finally, the demand for electrical energy exhibits predictable daily and weekly cyclical variations. However, it is by no means the only commodity for which the demand is cyclical.
Micro Grid Energy Trading
The Microgrid under consideration, as shown in the Fig 1 has multiple sources of generation i.e. Solar panels and Diesel generator set. The Microgrid is connected to the utility grid through an Energy trading model which takes profitable decisions satisfying all the constraints. What makes the trading decisions complicated is the presence of the Battery, which is the only element with memory. In simple words, using it in the ith hour will have an effect on using it in the (i+1)th hour unlike any other element. The electricity markets intrinsically have very cyclical patterns, each cycle being 24 hours. So to take care of the battery (memory element) we forecast the data 24 hours in advance and obtain the most profitable trades for the next 24 hours and implement only the next hour’s trade. Thus every hour we project the next 24 trades and implement the next hour trade only. This way, we sell the energy stored in the battery at the maximum possible price, thus generating maximum returns. The working of the model is explained in detail in the further chapters. The roadmap of the project is also briefly discussed to give the readers a better understanding of the approach to the problem. The project is split into two parts, the first being forecasting and the second being trading.
Fig. 1: Single line diagram of the Microgrid with arrows indicating the direction of the flow of power
- Part A: Forecasting
The most essential part of any forecast is acquiring good quality data. Once the data is acquired various techniques can be used for day ahead forecasting. For working of the model we need to forecast Solar energy output (SG), non schedulable demand (D) in the Microgrid and the market price of power (MP) a day in advance. Day ahead forecasting of solar irradiation is imperative in finding out the solar energy output and hence we need to find out on what measurable data it depends on. Forecasting is discussed in detail in the next chapter.
- Part B: Trading
This is the core part of the project, where we build the model which makes key decisions on trading to maximize the trade balance. The detailed working of the model is discussed in the chapter on trading.
Microgrid systems targeted in this study are autonomous areas serving their power demand of several kilowatts with diesel engines (DE), photovoltaic panels (PV) & batteries (B). Fig. 3 shows the structure of the proposed Microgrid. Microgrids can also be connected to the external power system by tie lines for reducing frequency/voltage fluctuation in the normal and emergency conditions. Such Microgrid systems are operated independently with zero tie-line flow under normal conditions. In this project a dynamic simulation is conducted assuming the Microgrid is operated connected to external electric power systems actively trading energy. To maintain the frequency near constant under this operation, the demand-and-supply balance is controlled by the diesel governor and battery output control.
Diesel engine output is controlled by the governor which is installed in the generator set. This achieves generally a good load following operation.
The operating condition of the battery storage is decided depending on the frequency responses of the Microgrid system. The steady-state output of the battery storage is the amount of load with the total output of diesel engine and of the PV panels subtracted.
Photovoltaic (PV) Panels
The output of photovoltaic cells change by the weather, so in this project, we have used the pattern of output that has been measured by the field tests of photovoltaic panels available on National Renewable Energy Laboratory website. For each daylight hour average output is assumed under the different weather conditions such as clear, cloudy, or rain. In addition, the max capacity of the photovoltaic panels installed in the Microgrid is assumed to be around 500 kW.
The Microgrid is connected to the utility grid through an Energy trading model which takes profitable decisions satisfying all the constraints. What makes the trading decisions complicated is the presence of the Battery, which is the only element with memory. In simple words, using it in the ith hour will have an effect on using it in the (i+1)th hour unlike any other element. The electricity markets intrinsically have very cyclical patterns, each cycle being 24 hours. So to take care of the battery (memory element) we forecast the data 24 hours in advance and obtain the most profitable trades for the next 24 hours and implement only the next hour’s trade. Thus every hour we project the next 24 trades and implement the next hour trade only. This way, we sell the energy stored in the battery at the maximum possible price, thus generating maximum returns. The most essential part of any forecast is acquiring good quality data. Once the data is acquired various techniques can be used for day ahead forecasting. For working of the model we need to forecast solar energy output (SG), non schedulable demand (D) in the Microgrid and the market price of electrical energy (MP) a day in advance. Day ahead forecasting of solar irradiation is imperative in finding out the solar energy output and hence we need to find out on what measurable data it depends on.
Fig. 2: Information flow diagram of the Microgrid with arrows indicating the direction of the flow of information
Energy Trading Formulation
Trading is done to maximize the total profit by minimizing the total cost incurred in the next 24 hours (Fundamental Cycle) and the optimal values for the current hour (hour 0) are implemented at that time. As shown in Fig 2 the trading model receives data from all the sources and loads and it will use the forecasted data to make profitable trading decisions without compromising on the constraints. It will also make the decision of taking power from the main grid, Battery and the emergency Diesel Generator set. It will make sure that the schedulable load is supplied power when the price of the power is the least. There are two sets of independent variables i.e. 48 variables which can be varied to find the minimal value of the total cost.
Energy Trading Architecture
The Energy Trading Model is shown in Fig. 3. The trading model receives data from all the sources and loads and it will use the forecasted data to make profitable trading decisions without compromising on the constraints. It will also make the decision of taking power from the main grid, Battery and the emergency Diesel Generator set. It will make sure that the schedulable load is supplied power when the price of the power is the least.
Fig. 3: Information flow diagram of the Microgrid with arrows indicating the direction of the flow of information
The Objective is to make profitable trading decisions, adhering to constraints.
So, Introducing a function,
where, is a vector consisting of all the independent variables in
The output variables of the model are the independent variables associated with and theyare assigned the values they take at the global minimum of within the boundary conditions by the model.
Using the energy trading model drastically reduces the over costs and increases the bottom-line considerably. We have taken a case study with near – real time values and simulated it with both, the model operating on it and without the model operating on it.
The model has two sources of generation, they are:
- Solar panels with a maximum generation of 500 kW
- Diesel generator with a maximum generation of 300 kW, with the cost generation function:
DC (Rs/h) = a + b (DG) + c (DG)2 (2)
a = 200 Rs/h
b = 10 Rs/ kW-h
c = 0.005 Rs/ (kW)2-h
The model also has a super storage facility (Battery) with maximum usable energy range of (BEMAX) of 500 kWh and maximum power (BPMAX) of 200 kW. There are two sources of revenue, one from trading and other from consumers in the Microgrid who are charges and fixed rate of Rs.8 / kW-h. (all the data used in this project is from S1-S2 regions of India and hence consumer charge is also fixed at Rs.8 / kW-h).
Advantages of having a trading model in place are clearly visible from the Fig. 4. The area below the graph gives the total cost incurred in a day after trading. The trading model makes optimal trading decisions and reduces the cost by over 60%.
Fig. 4: Plot showing hourly cost incurred after trading during a day (24 hours)
The results are surprising but we have to remind ourselves that we did not take into consideration the losses in charging and discharging the battery and also the usage cost of the battery. The life of a battery depends on its usage; excessive usage will reduces the life of the battery at a very high rate. Also, the cost of capital on such a high investment will considerably reduce the return on investment (ROI). But, such a high increase in the bottom-line is worth discussing as there is a lot of research going into improving the storage devices which have a very long lifetime and a very high efficiency in converting it to electricity again. Let us go a step further and analyze how the model makes such a high profit. We can make some inferences from Fig 5. Where we can see the generation curves and also the trading decisions the model has taken to increase the bottom line. It is clear that Microgrid Energy Trading Model (MTEM) has bought power from the utility grid when the net-generation is negative, and has sold power to the utility grid when there is excess of power in the Microgrid. But, this is very intuitive and simple and the working of the model is not very clear from the above plot, i.e. even without the model nothing different would have happened but what is counter intuitive from the plot is that it sold power it generated using diesel generator and at the same time the net generation was negative. That means the power was sourced from the storage device, but what compelled the trading model to take that decision can be understood from the plot showing the usage of battery and the hourly market price.
Fig. 5: Plot giving the details of hourly generations from Diesel Generator, Solar Panels and also hourly trade volumes
Fig. 6 gives a better idea of working of the METM. The model charged the battery when the market price was low and sold it when the Market price was relatively high. Looking at both the plots together gives a much better idea of what really happened. From hour 20-22 (peak time) Market price went up to around 16 `/kW, making the use of Diesel generator with the incremental cost (b) around `10/kW economically viable. Hence, the Trading Model sold power from the battery and also sold the power generated using the Diesel generator to reach a local maximum in trade volume and to make maximum profits. This was possible only because it could see the opportunity of making such profits before hand and charge the battery hours in advance when the cost of power was considerably less, i.e., cheaper than the diesel generator also and seized the opportunity of making such high profits.
Fig. 6: Plot showing the battery usage and the Market price for every hour
The Microgrid Energy Trading Model (METM) built using multiple coding languages and applications is has achieved very good results. In the case study, where we have assumed multiple sources of generation and a storage facility, it has reduced the net costs incurred by over 60% by trading efficiently and making maximum use of the storage facility. It has also increased the bottom-line by over 175%, this is a little surprising but once we take losses in charging and discharging into consideration it will reduce to a little over 100% increase. In addition to increase of profits and increase in the Return on Investment (ROI), it also reduces the carbon dioxide emissions thus producing clean energy. A typical 300 kW Diesel Generator produces with a load factor of 20-50% produces approximately 0.8 kg of CO2 for every kW-h generated. With the METM in place the Diesel Generator in the Microgrid produces approximately 900 kW-h less than what it produces without the METM. This implies that the model reduces CO2 emissions by 720 kg/day.