STOCKHOLM SVERIGE 2020,
Energy Storage System for Local Generation in a Grid-connected Microgrid
Sizing and analyzing an energy storage system for the Tezpur University campus
STINA CARNEHEIM
KTH
SKOLAN FÖR ELEKTROTEKNIK OCH DATAVETENSKAP
Energy Storage System for Local Generation in a Grid-connected Microgrid
Stina Carneheim
Abstract—Reducing the emissions is an important step in order to reduce the global warming. At Tezpur University in the Assam region in Northeast India a project is being performed which will try to reduce the use of diesel for back-up generation and partly replace this with other sources of energy. In 2018 a 1 MW PV- plant was installed as a part of this goal. However, since the consumption and the PV-production are not synchronized some of the energy goes to waste.
This thesis will consider how an energy storage system (ESS) can help to increase the usage of the power produced by the PV-plant. It will also assess the best type and size of the system.
In addition to this, a simple economical analysis was performed to determine the profitability of the project.
First, data regarding the system at Tezpur University was gathered. The data of interest was the production of the PV panels as well as the consumption. This data was then processed so that the consumption and production could be observed for each hour a typical day. By comparing these two the overproduction which goes to waste could be estimated. By evaluating how much energy produced from diesel which could be replaced by the ESS an assessment of the savings per year could be made. From this the payback period was calculated for the different size ESS.
The results show that a battery energy storage system (BESS) using lithium-ion batteries is the preferred solution in this case.
Assuming that 50 % of the life expectancy of a battery is a reasonable payback period the maximum size of the battery is 127 kWh. The optimal placement of the BESS is at substation 4 as the overproduction is the greatest in this area and as there is also a large load the stored energy would be used fully each day.
A battery size of 90 kWh was suggested by the E4T MicroGrid project and considering the payback period this is a reasonable size for the BESS.
Sammanfattning—Att minska utsläppen är ett viktigt steg i att minska den globala uppvärmningen. Vid Tezpur Universitet i Assam i nordöstra Indien genomförs nu ett projekt som skall minska användningen av diesel vid avbrott genom att delvis ersätta dessa generatorer med andra energikällor. 2018 installerades ett solkraftverk om 1 MW som en del i detta mål.
Eftersom konsumtionen och produktionen från solkraftverket inte är helt synkroniserade är det delar av den producerade elektriciteten som skickas tillbaka ut i nätet och därmed går förlorad.
Det här projektet har undersökt hur ett energilagringssystem kan användas för att öka användningen av energin producerad av solkraftverket. En annan del som undersökts är vilken typ av system och vilken storlek det bör ha. Efter detta görs en enkel ekonomisk analys för att utreda hur ekonomiskt gynnsamt projektet är.
Det första som gjordes var att samla data om microsystemet på Tezpur Universitet. Den data som samlades var om produk- tionen från solkraftverket och elkonsumtionen i de olika delarna av universitetet. Genom olika metoder kunde man undersöka hur konsumtionen och produktionen var per timme en typisk dag. Då man jämförde dessa kunde överproduktionen per dag estimeras. Besparingarna som görs beräknades genom att byta ut en del av dieselanvändningen med kostnaden av att ladda
energilagringssystemet. Detta gav tillräckligt med information för att uppskatta återbetalningsperioden.
Den bästa lösningen i det här fallet är att installera ett batterilagringssystem bestående av litiumjonbatterier. Under an- tagandet att återbetalningsperioden maximalt får var 50 % av livslängden av batteriet kommer den största tillåtna storleken att vara 127 kWh. Den optimala placeringen av systemet är vid transformatorstation 4 eftersom det är där som större delen av överproduktionen uppstår. Det är även till den som större delen av lasten är kopplad vilket garanterar att hela batteriets laddning kan användas varje dag. Batteristorleken om 90 kWh som föreslås i E4T MicroGrid-projektet är en bra storlek med tanke på återbetalningsperioden.
CONTENTS
I Nomenclature 2
II Introduction 3
III Project Background 3
III-A Tezpur University . . . . 3 III-B E4T Microgrid Pilot for Tezpur Univer-
sity Campus . . . . 3 III-C Performance Optimization during Grid
Outages using Demand Side Management 4
IV Problemization 4
V Aims and Objectives 4
VI Methodology 4
VI-A Initiation and background . . . . 4 VI-B Analysis . . . . 4 VI-C Findings . . . . 4
VII Microgrids 4
VIII Solar Energy 5
VIII-A Model for PV-production . . . . 5
IX Energy Storage Systems 5
IX-A Mechanical energy storage . . . . 5 IX-A1 Flywheel energy storage . . 6 IX-A2 Pumped hydro storage . . . . 6 IX-A3 Gravity Energy Storage . . . 6 IX-A4 Compressed air energy storage 6 IX-B Chemical energy storage . . . . 6
IX-B1 Fuel cell - hydrogen energy storage . . . . 6 IX-C Superconducting magnetic energy storage 7 IX-D Electrochemical energy storage . . . . . 7
IX-D1 Energy storage in capacitors and super-capacitors . . . . . 7 IX-D2 Lead-acid battery . . . . 7 IX-D3 Lithium-ion battery . . . . . 7 IX-D4 Sodium sulfur battery . . . . 7 IX-D5 Nickel-cadmium battery . . . 7 IX-D6 Sodium nickel chloride battery 8 IX-D7 Flow battery energy storage . 8 X The Grid-Connected Microgrid at Tezpur Uni-
versity 8
X-A Grid Connection . . . . 8 X-B PV-system at Tezpur University . . . . 8 X-B1 Equipment used . . . . 8 X-B2 Distribution of the system . . 8 X-C Diesel generators . . . . 8 X-D Loads . . . . 9 X-E Electricity prices at Tezpur University . 9
XI PV-production 9
XII Outages 10
XIII Loads 10
XIII-A Student hostels . . . . 10 XIII-A1 Load profile for the hostels . 10 XIII-A2 Load per month . . . . 10 XIII-B Departments . . . . 11 XIII-B1 Load profile for departments 11 XIII-B2 Load per month . . . . 11 XIII-C Water treatment plant . . . . 12 XIII-D Total load . . . . 12
XIV Sizing Energy Storage System 12
XIV-A Store overproduction . . . . 12 XIV-B Store for specific functions . . . . 14 XIV-C Store constant charge during off peak
hours . . . . 14 XIV-D Suggested solution from E4T MicroGrid 14 XV Comparison of Energy Storage System 14
XV-A Comparison of lithium-ion and lead- acid batteries . . . . 14
XVI Economical Analysis 15
XVII Results 15
XVII-A Store the total overproduction . . . . 15 XVII-B Store the overproduction in substations
separately . . . . 16 XVII-C Size of storage for water treatment center 16 XVII-D BESS size needed to cover the Depart-
ment of Energy . . . . 16 XVII-E Constant charge during off peak hours . 16 XVII-F 90 kWh BESS . . . . 17
XVIII Discussion 17
XVIII-A Type of battery . . . . 17 XVIII-B Size and placement of ESS . . . . 17 XVIII-C Limitations of the analysis . . . . 18
XIX Conclusions 18
XIX-A Choice of energy storage system . . . . 18 XIX-B Placement of energy storage system . . 18 XIX-C Size of energy storage system . . . . . 18
XX Future Work 18
XXI Ethics 19
References 19
Appendix A: Technical Data of Equipment 21 A-A Technical data for PV-panels . . . . 21 A-B Technical data for inverters . . . . 21
Appendix B: PV-system 22
I. NOMENCLATURE
Abbreviation Explanation
Acca-1 Academic Building 1 Acca-2 Academic Building 2
APPCPL Arunachal Pradesh Power Corperation Private Limited BWH Bordoichilla Women’s Hostel
CHP Combined Heat and Power CMH Charaideo Men’s Hostel CT Current Transformer
DCS Department of Chemical Sciences DE Department of Energy
DES Department of Environmental Science
DG Diesel Generators
DoD Depth of Discharge
DMBBT Department of Molecular Biology and Biotechnology DSM Demand Side Management
DWH Dhansiri Women’s Hostel EES Energy Storage System EMS Energy Management System
INR Indian Rupees
KBRA Kalaguru Bishnuprashad Rabha Auditorium KMH Kachanjungha Men’s Hostel
KTH Royal Institute of Technology KWH Kopili Women’s Hostel LMU Load Management Unit
MG Microgrid
NMH Nilichal Men’s Hostel NWH New Women’s Hostel PF Power Factor
PMCWH Prabitora Madame Curie Women’s Hostel PMH Patkai Men’s Hostel
PWH Pragiyotika Woman’s Hostel PV Photovoltaic
RES Renewable Energy Sources RESCo Renewable Energy Service Company SAC Student’s Activity Center
SAIC Sophisticated Analytical Instrumentation Center SEA Swedish Energy Agency
SMH Saraighat Men’s Hostel SWH Subansiri Women’s Hostel USD United States Dollar WTP Water Treatment Plant
II. INTRODUCTION
ELECTRICITY has become a basic need for many people around the world. It is no longer a symbol of wealth and prosperity, but a matter of survival. In 2015 India signed the Paris Agreement promising to transform their trajectory toward a more sustainable development [1]. The aim is to limit the warming to between 1.5 and 2 ◦C above the pre- industrial levels. India’s goals are by 2030 to reduce the emissions intensity by 33-35 % and also to to increase the non-fossil share of the cumulative power generation capacity to 40 % [2]. In recent years India has invested more money in renewables than into fossil fuels making the country a world leader in renewable energy. As a consequence of this, the country will most probably meet the 2◦C goal. One of the largest contributions to the high investments in renewable energy is that the price of solar photovoltaic (PV) energy has become very competitive in comparison to coal.
Despite India’s role as a leader in renewable energy, the CO2 emissions in India increased by 4.8 % in 2018 [2]. This increase is mostly due to the use of coal power plants and there are plans to increase the scale of the coal production further. In order for India to reduce these emissions one must consider which energy sources are used.
Even if these goals must be tackled on a large scale, one must also consider what can be done on the smaller scale. In rural parts of India the grid is not so stable. This causes a lot of outages, which must be covered in some other way. In many cases there are diesel generators (DG) that can supply power during these outages. If one could reduce the usage of the DGs one could on a small scale also reduce the CO2
emissions. A very useful source for small scale production is PV-power. However, PV-panels have one major drawback;
power is only produced when there is sun. For PV-power to
become the main source of energy in India one must develop a way of storing the energy produced during the day, so that it can be used during the night time as well.
An important part of a power system is that the production of power is in sync with the usage of power, which is often not the case, therefore it is important to have a buffer such as an energy storage system [3]. This can be especially important in rural areas where the power grid is not as stable. Here an energy storage system can be the difference between having or not having any clean water, a very important resource when running a school or university.
III. PROJECTBACKGROUND
This project is a collaboration between the KTH Royal Institute of Technology (KTH) and Tezpur University (TU).
It is also a continuation of the project E4T which was a joint project lead by the company Pamoja Cleantech with Swedish Energy Agency (SEA), KTH and TU. E4T will be discussed further in section III-B.
A. Tezpur University
Tezpur University is located outside of the city Tezpur in the Assam province in Northeast India. TU is a state university and was founded in 1994. The campus is spread out over approximately one square kilometer and facilitate the 3000 students and 500 staff members. Both staff and students live and work on campus. The students are housed in student hostels divided into seven women’s hostels and five men’s hostels. The teachers reside in different sized quarters.
Except for housing and academic buildings there are stores, restaurants, a sports center and other amenities.
The power system in this region is, as in many rural areas, unstable. This creates a need for alternative sources of energy.
The site of the university has an average of 4557 hours of sun each year making it quite suitable for PV-power [4]. There is also an abundance of biological bi-products, such as paddy husk, in the area which could be used for electrification [5].
B. E4T Microgrid Pilot for Tezpur University Campus E4T MicroGrid is a project in cooperation between Pamoja Cleantech, SEA, KTH, and TU. The goal is to create a smart polygeneration microgrid at the Tezpur University campus [5]. As the Energy Department at TU specializes in rural electrification using both biomass and PV technologies it is the perfect location for this pilot project. In the original scope of the project a sterling engine was to be tested at the site in cooperation with the company INRESOL. However, since the initiation of the project, INRESOL has been liquidated as a result of technical difficulties with the sterling engine.
In replacement of INRESOL the company Modio has been involved. Modio will supply an internal combustion combined heat and power (CHP) engine and generator to replace the sterling engine.
The scope of the project is to integrate a biomass power plant with a cogeneration gas engine, an electricity storage system, an energy management system and a demand side
management system. Since before the initiation of the project there is a PV plant installed at the site. The integration of the PV plant is also included in the scope of the project. Another part of the project is to utilize the excess heat from the biomass power plant to satisfy the hot water supply of the campus.
C. Performance Optimization during Grid Outages using De- mand Side Management
As part of E4T MicroGrid, a previous masters thesis was performed examining the possibility of implementing demand side management (DSM) strategies at TU [6]. This project included a two week study trip to TU performing extensive studies including interviews and surveys. This was then used as a base of how the facilities are used a typical day at TU. Using the findings three DSM strategies were suggested and examined: air conditioning outage energy control, hostel daytime energy control and water pumping energy control. By applying these strategies the consumption could be reduced significantly. The results from this report are now being used in the development of the E4T MicroGrid project.
IV. PROBLEMIZATION
The power grid in the Assam region is very unstable. With outages approximately 3 % of the time, TU is dependent on diesel generator (DG) for backup when there is an outage [6].
This is a considerable cost as the price of diesel is much higher than the electricity from the grid. It is also a great source of emissions which could be reduced if a different source of energy could be used.
When the PV system was installed there was no meter installed to measure which way the power was flowing. When there is overproduction from the PV plant the university is charged both for the PV-production as well as the extra power produced which goes out into the grid. By installing an energy storage system (ESS) one hopes to store the overproduction and thereby reduce the costs further.
V. AIMS ANDOBJECTIVES
The aim of the project is to create and analyse a potential ESS for Tezpur University in Assam, India. The ESS should reduce the diesel consumption of the university as well as en- sure a more stable supply of energy to the university. The ESS should also reduce the overproduction of the PV-production at the university. The question that will be answered is: which type of ESS should be chosen, how should it be sized and where in the system should it be placed in order to reduce the cost of the diesel consumption and the overproduction of the system?
VI. METHODOLOGY
The project will be divided into three main parts. First a period of understanding the problem and researching different methods to solve it. Then a period of executing the chosen method and lastly a period of analysing the results and finishing the project.
A. Initiation and background
The aim of the initiation phase is to gain a better under- standing of the problem as well as finding a method of solving that problem. During the initiation phase a thorough literature research is necessary in order to get a good understanding of the problem and find the optimal solution for the project.
The research will focus on similar projects, different methods used for energy storage, how to perform an estimate of the PV power produced and grid-connected microgrids. In order to get a full picture of the existing power system and the problem at hand a field trip to Tezpur University was performed. The aim of the field trip was to map the current system with respect to equipment and set up. Another aim was to gather additional data regarding the PV production of the system, loads at the university as well as current consumption of diesel. In the initial contact with TU it was mentioned that the university occasionally pay more than double for the production of the PV-panels. Therefore, another objective of the field trip was to gain a better understanding in regards to the contract and set up with the power company to understand how this is possible.
B. Analysis
The next phase of the project was to assemble the infor- mation from the initiation phase and analyse the data. Using the information gathered at TU as well as information from the previous projects performed at TU the microgrid can be modelled. This model contains loads, diesel generator, PV- panels and other parts of the microgrid. The data gathered at TU was then analysed to create a load profile as well as an estimation of the PV-production. Next, different methods of using an energy storage system were created and an economical analysis was done.
C. Findings
After analysing the different methods the final results were compared and an optimal solution was suggested. This final section of the project also includes writing the report and presenting the findings.
VII. MICROGRIDS
A microgrid is a clearly confined system of generators and loads [7], [8]. The definition of a microgrid is that the system can be controlled locally as an independent system and operate in a so called island mode but it can also operate as part of the centralized larger grid. In many cases the microgrid is primarily connected to the grid but has the capacity to operate on its own in island mode. When in island mode the system can supply the energy needed for the local load using local generation. This means that the system can continue operation even if there is a large outage in the main system. This is one of the great advantages of having a microgrid; the local grid is less dependent on outside factors and becomes more dependable.
Typical generators in a microgrid are PV-panels, fuel cell as DGs and micro turbines connected on the distribution level [9]. However, generation in microgrids can be any type of
generation. A typical set up of a microgrid can be seen in Figure 1.
Since many microgrids are powered by PV and wind power, which are not constant, one must have methods of coping with differentiation in the production. One way of ensuring that the demand is met is to integrate demand regulation to the system [6]. This enables the load to be regulated when the power supply is limited. Another method is to have an ESS. By storing energy during the hours when there is power connected one can be more prepared when there is an outage.
By implementing microgrids into the main grid one can make many improvements to the network operations. Some of these benefits are that losses can be reduced, voltages controlled and having a more reliable system [8]. One can also improve the energy efficiency of the system by controlling the local system. The energy consumption can be minimized using i.e. demand regulation. Better control also leads to a reduced environmental impact.
Utility Grid
Backup generator User
Renewable energy sources
Battery Microgrid
Figure 1: A typical microgrid
VIII. SOLARENERGY
A technology that is used increasingly more frequent today is to generate power using the sun. This technology uses electromagnetic waves and turns these into electricity using semiconductors [10]. The technology was first created in 1954 and was then used for space crafts and small calculators. Today PV-technology has been developed and is used for large scale production of energy.
The most common type of PV-cell is made out of silicon [11]. These cells consist of two layers of silicon, a metal grid on top and a metal sheet on the bottom. This structure can be seen in Figure 2. By adding a small amount of atoms the silicon becomes doped. The upper silicon layer is doped with phosphorous creating a negatively charges carrier. This doped silicon is called n-silicon. On the bottom layer the silicon is doped with boron creating a positively charged carrier called p-silicon. The metal grid on the top will collect the charges
Grid
Top
Rear
p-silicon n-silicon
Figure 2: Set up of a PVpanel
formed from the solar radiation on the cell. When a photon hits the PV-cell, the energy can excite the electrons in the silicon. These electron are attracted to the n-silicon. When the electrons are excited they leave a hole. These holes are attracted to the p-silicon. Because of this a voltage will be created. Since the top and the bottom collectors are connected there will be a current.
Light can be measured as power density, also called solar irradiance (W/m2). This data can then be used to approximate the energy produced by the PV panels. In the end, the amount of energy produced is a matter of meteorology and climatology as well as the energy conversion system [12].
A. Model for PV-production
In order to estimate the average power produced by a PV- panel one must find the irradiance at a specific location. This can be done by collecting data at that point over a long time period. This can however be a slow and expensive project. In- stead one can use data that already exists. A popular method of compiling data is to reanalyse meteorological data [13]. This is due to it being on a global scale, it has gathered data over several decades and is often freely available. Another method of compiling data is to use satellite images. The accuracy of this method is higher than when using the meteorological method. However, freely accessible data from satellites are more scarce and does therefore not cover the whole globe.
IX. ENERGYSTORAGESYSTEMS
An ESS is a systems which stores energy when there is an abundance and allows you to use that energy at a later time when it is needed. Many different approaches can be used some of which will be discussed in this section.
A. Mechanical energy storage
Mechanical Energy Storage (MES) uses gravity, kinetic energy or elasticity in order to store energy [14]. It uses potential energy in the form of the working principles of pres- surized gas, forced springs and other ways as well as kinetic energy [15]. The life expectancy of a MES is usually long as it is mainly determined by the life time of its mechanical components [16]. This section will discuss the most common types of MES.
1) Flywheel energy storage: A Flywheel Energy Storage (FES) stores the kinetic energy of a mass spinning at a high velocity [15]. It consists of a massive rotating cylinder that is mounted on a shaft and levitated between magnetic bearings.
The FES, as all ESSs, has three operating modes: charging, discharging and idling [16]. In charging mode a motor rotates the shaft and kinetic energy is transferred to a rotor which stores the energy. The faster the rotor in spinning the more energy is stored [15]. When the energy is needed the shaft decelerates and the motor acts as a generator to for the main system. The energy stored in the FES is given by:
Wkin=1
2Jmω2m (1)
where Jmis the moment of inertia calculated using the radius, r, and the distribution of mass, m, of the flywheel:
Jm= Z
m
r2 dm (2)
ωmis the angular velocity of the flywheel given by:
ωm= 2π · nm (3)
where nm is the rotational speed.
The instantaneous efficiency of a FES is 85 % [17]. How- ever, due to high friction losses the efficiency would drop to 78 % after five hours and to 45 % after one day. Considering this FES is more suitable for short term storage.
2) Pumped hydro storage: The basic principles of Pumped Hydro Storage (PHS) is simple; water is pumped from one reservoir to another reservoir at a higher level using excess power produced when consumption is low [15], [16]. This water can then be used when there is a need. The energy is stored in the form of gravitational potential energy. Today, PHS is mostly used in high power applications however it can be used in small scale projects [18]. The main shortcoming of this method is that there is a need for a substantial height difference between the reservoirs [17].
3) Gravity Energy Storage: As in PHS, Gravity Energy Storage (GES) uses gravity to store energy [16]. It is an umbrella name for many technologies that all use gravity to store energy. The energy can for example be stored by raising a piston higher when there is an abundance of energy which then is released when the energy is needed [19]. The energy stored by the GES and be described as:
E = mrgzµ (4)
where mr is the relative mass of the piston, g is gravitational acceleration, z is the height of the piston and µ is the efficiency of the GES. One example of GES has been developed by the company Energy Cache. The system uses the excess energy during off peak periods to transport buckets of gravel up a hill to then release it during peak hours. This type of system has the same requirement for height difference as PHS.
4) Compressed air energy storage: Compressed Air Energy Storage (CAES) uses off peak electricity to compress air and store until peak hours [16]. The air can be stored in underground caverns, in pipes or vessels above ground. Energy is produced by releasing the compressed air through an air
turbine at a high pressure [20]. One can increase the output by boosting the air pressure through burning natural gas with the compressed air. By using an expansion machine, three times the power can be produced with the same fuel consumption [17]. The expansion machine consists of a volumetric expander driving an electric generator [21]. A general setup of a CAES can be seen in Figure 3. CAES generally has a low impact on the environment [15]. It is favourable to use for large scale storage.
El. in
Compression Air in
Reservoir
Expansion Air out
El. out
Figure 3: A general setup of a CAES [21]
B. Chemical energy storage
Chemical Energy Storage (CES) uses the chemical bonds of atoms and molecule for storage of energy [14]. The energy is then released in a chemical reaction [15]. CES is defined as all technologies that used electrical energy to produce chemical compounds to be stored until needed [16]. The energy density of a chemical compound usually has a higher energy density than CAES and pumped hydro which makes it a very effective storage system. Examples of the most common produced chemical fuels are coal, gasoline, diesel fuel, natural gas and hydrogen [15].
1) Fuel cell - hydrogen energy storage: Hydrogen energy storage is a very popular type of chemical storage [16].
When produced using renewable electricity hydrogen is a very clean renewable fuel as well as being storable and transportable [15]. A Fuel cell - Hydrogen Energy Storage (FC-HES) transforms the chemical energy in the hydrogen and oxygen into electricity [22]. The only bi-product is water vapor. To create the hydrogen an electrolyzer can be used.
An electrolyzer is an electrochemical converter which converts water into oxygen and hydrogen [15]. One drawback for FC- HES is that hydrogen is extremely flammable in gas form [14]. Another problem is the efficiency. Combining a fuel cell with an electrolyzer will give an efficiency of approximately only 35 %, this is however partly compensated due to the high storage density [16], [17].
+
A
Load
− B
Anode Electrolyte Cathode Cation
Anion
Figure 4: General set up for secondary batteries
C. Superconducting magnetic energy storage
By using electrodynamic principles a Superconducting Magnetic Energy Storage (SMES) stores energy in the electro- magnetic field [15]. The field created by inducing a DC current into a coil made of superconducting cables with a resistance close to zero [15], [16], [17], [22]. The resistance is reduced by cooling the coil to under its critical temperature. The biggest drawback of a SMES is the need for a cooling system which is very expensive [17].
D. Electrochemical energy storage
Electrochemical energy storage stores chemical energy which is then converted into electrical energy when needed [14]. To release the electric power at least two chemical components undergo a reaction [15]. Two types of electro- chemical energy storage will be discussed in this section:
electrochemical batteries and electrochemial capacitors.
The most commonly used method for small scale energy storage is a battery. A battery energy storage system (BESS) is also often used for backup energy. A battery consists of two electrodes and an electrolyte in an isolated container [14]. The electricity is created when an external source or load is con- nected to the electrodes. The electrodes are called anode and cathode. When the external component is attached the anode releases positive ions into the electrolyte and release electrons into the circuit. At the cathode the ions are absorbed using electrons from the circuit. This creates a current. Batteries are categorized into three categories: primary, secondary and flow batteries. Primary batteries are not rechargable and will therefore not be discussed further [23]. A secondary battery is a chemical composition which can be reversed and therefore be recharged. The general set up of a secondary battery can be seen in Figure 4. A flow battery is a type of secondary battery but with two liquids separated by a membrane [24].
1) Energy storage in capacitors and super-capacitors: The simplest and most direct way of storing electrical energy is a capacitor. It consists of two plates made of metal separated by a non-conductive dielectric [15]. The energy is stored in the field between the two plates. A capacitor is able to accept and deliver large powers.
A super-capacitor has the same principle as the regular capacitor. The difference is that the dielectric is replaced by an electrolyte ionic conductor [17]. In the conductor ions move along a very large conducting electrode.
2) Lead-acid battery: The lead-acid battery is a mature technology which results in a low cost and high reliability [16], [22], [25]. In charged mode the positive plate consists of lead dioxide (PbO2) and the negative plate consists of lead (Pb) [21]. When discharging both plates react with hydrogen sulphate (HSO4-) and become lead sulfate (PbSO4) and water (H2O). The global reaction will be:
P b + P bO2+ 2HSO−4 + 2H+→ 2P bSO4+ 2H2O (5) Lead-acid batteries are being used as back-up power supply and in energy management systems [22]. It has also been developed for vehicle-applications. Other versions of the lead- acid battery are carbon lead-acid batteries and advanced lead- acid batteries [16]. The lead-acid battery has a low energy density, low power density, long charge time, low cycle life and high self discharge rate [22], [25]. It is likely to cause environmental pollution if it is damaged.
3) Lithium-ion battery: The lithium-ion battery has lithium- ions that moves between a positive and a negative electrode [25]. The positive electrode is for example comprised by a lithium based compound such as lithium iron phosphate (LiFePO4) and lithium manganese oxides. The negative elec- trode is usually made of graphite or titanium. The use of graphite creates a risk due to dendrite-induced short circuits during rapid changing. Because of this new material for the negative electrode is being examined such as silicone, metallic tin and their oxides. The battery also contains a protection circuit [16]. The lithium-ion battery is lighter, smaller and more powerful than other batteries. It also has a high efficiency and a low self discharge. It is preferred in situations where fast response and small dimensions are important [22]. The biggest drawback for the lithium-ion battery is its high capital cost [16].
4) Sodium sulfur battery: A sodium sulfur battery has molten sodium and sulfur as negative and positive electrodes [25]. A ceramic tube serves as a solid electrolyte and separated the positive and negative electrodes. The battery has a high en- ergy density, high rated capacity and almost no self discharge [22]. It is comprised by inexpensive non-toxic materials so it is easy to recycle. It shows great promise to be used in high power applications. However, in order for the electrodes to stay molten the battery operates at a high temperature, 300- 350◦C [22][25]. If there is a rapture induced short circuit this can lead to severe fires [25]. The high temperature can also lead to corrosion of the ceramic tube. Research regarding the battery is focused on enhancing performance and lower the temperatures needed [22].
5) Nickel-cadmium battery: The electrodes of the battery consists of nickel oxy-hydroxide and metallic cadmium [16], [22]. The electrolyte is an aqueous alkali solution [22]. The nickel-cadmium battery is the most developed of the nickel batteries [16]. It is robust and does not require much main- tenance [22]. It has a fast recharge time, a long cycle time and deep discharge rates with no damage or loss of capacity [16]. However, the maximum capacity is greatly decreased if the battery is recharged after only being partially discharged [22]. Another disadvantage of the nickel-cadmium battery is
that both nickel and cadmium are highly toxic and can cause environmental damage if leaked [16], [22].
6) Sodium nickel chloride battery: The sodium nickel chloride battery (ZEBRA) has NaCl salt and Ni which is transformed into NiCl2and molten Na [16]. It operates at high temperatures (270-350◦C). The electrolyte is a ceramic wall which conducts Na+ but not electrons. The ZEBRA battery is maintenance free and has a relatively high life cycle [22]. It is typically used in electric vehicle applications [16], [22].
7) Flow battery energy storage: The active materials in a flow battery are in liquid state and mixed with the electrolyte [17]. The energy storage is based on electro-chemical reactions between the two fluids. A flow battery has a high efficiency [25]. It also has a long life time since the electrodes do not undergo physical or chemical change [16]. An advantage the flow battery has over a classic battery is that one does not have to maximize the energy density in order to optimize power acceptance and delivery properties. However, critical materials are scares which leads to high construction cost [25].
The most commonly used flow battery are the vanadium- based flow batteries [3]. The chemical energy is stored in a sulfuric acid electrolyte. The electrodes are vanadium ion solutions [16], [22], [25]. There are four types vanadium ions used: V2+, V3+, V4+ and V5+. When the battery is charged the positive electrode consists of V5+and the negative electrode 2+. When discharged the composition is V4+ in the positive electrode and V3+ on the negative electrodes.
Conduction is carried out using H+. Advantages include low maintenance costs, quick response and tolerance for over charging [16], [22]. It can also be deeply charged without effecting the life cycle of the battery. However, in addition to the battery itself it needs pumps, sensors, power management and secondary containment [16]. This results in that the battery is not suitable for small scale use.
Other types of flow batteries include zinc bromine flow batteries which uses an aqueous solutions of zinc bromine as electrolyte [16]. It is in the early stages of commercialization but only has an efficiency of 80 % compared to the efficiency of the vanadium-based batteries of 85 % [22].
X. THEGRID-CONNECTEDMICROGRID ATTEZPUR
UNIVERSITY
The system at TU can be defined as a grid-connected micro- grid. The demand side of the microgrid consists primarily of academic and domestic buildings spread out over the campus.
On the generation side one has a grid connection, a PV-plant and diesel generators.
A. Grid Connection
The system is connected to the main grid through a grid connection point on the edge of the campus marked in Figure 5. At the connection point there are two 2.5 MVA transformers.
These have a voltage ratio of 33kV / 11kV three phase. The electricity is then distributed to the rest of the campus.
In total there are seven substations spread out over the campus in order to even out the load. Five of the substa- tions have the set up of 2x315 kVA and a voltage ratio of
11 kV/0.433kV. The additional two substations have the set up 1x100 kVA and the voltage ratio 11 kV / 0.433 kV.
B. PV-system at Tezpur University
In 2018 a PV-plant was installed at the TU campus. It was installed by the power company Arunachal Pradesh Power Corperation Private Limited (APPCPL) that owns and main- tains the system. The size of the plant is 1 MW and it is the first MW capacity grid connected PV-plant at an educational institute in the entire Northeast India [4]. The system is distributed over fifteen buildings on the university campus.
1) Equipment used: The PV-panels used at TU are of type Eldora VSP.72.325.03 04 in the Eldora Grand Ultima Silver Series by Vikram Solar [4]. The rated peak power is 325 W per panel. There are in total 3257 panels spread out over 15 buildings. The total capacity is 1055.5 kWp. The panels cover an area of 7350 m2. Each panel is connected to one of the 20 inverters spread out over campus. They are of type SMA Solid-Q 50. The technical data for the panels can be found in Appendix A-A and for the inverters in Appendix A-B.
2) Distribution of the system: The panels are arranged in nine clusters and connected via inverters to the substations on campus. Each group is divided into strings; each with approximately 20 panels in series. The arrangement of the panels depend on the available space on the roof [4]. The orientation of the roof as well as its angle also played a role in the design of the system. Further information about roof space and angles can be found in Appendix B. The number of panels and the capacity of each cluster can be seen in Table I.
One can also see to which substation the cluster is connected.
Most of the capacity is connected to substation 4. No PV- panels are connected to substation 2.
Table I: Placement and capacity of PV-panels
Placement Number of
Panels Capacity [kWp] Number of Inverters
Substation 1:
Kanchenjunga Men’s Hostel and
Nilachal Men’s Hostel 112 345 2
Subansiri Women’s Hostel and
Pragjyotika Women’s Hostel 113 348 2
Substation 3:
Patkai Men’s Hostel 165 512 3
Department of Energy and Community Hall 112 346 2
Substation 4:
Academic Building-II and Central Library 162 498 3 Department of Environmental Science and
Molecular Biology and Biotechnology 150 462 3
KBR Auditorium 55 171 1
Student Activity Center and
Department of Chemical Sciences 100 309 2
SAIC Building 87 266 2
C. Diesel generators
On the TU campus there are five back-up diesel generators (DGs). The generators are identical and of make Cummins, Model KTAA19-G10 [6]. They each have the nominal power
1 10
3 4
6
2
7
9 11 15
18 16 17 23
21 24 14
20 19 8 13
22 5 12 25
25
26
27
1 BWH 2 CMH 3 DWH 4 KMH 5 KWH 6 NMH 7 NWH 8 PMCWH 9 PMH 10 PWH 11 SMH 12 SWH 13 Acca-1 14 Acca-2 15 Community hall 16 Council hall 17 DCS 18 DE 19 DES 20 DMBBT 21 Library 22 KBRA 23 SAC 24 SAIC 25 WTP 26 Generator center 27 Grid connection
Figure 5: University campus
of 500kVA and operates at 50 Hz and 415 V [26]. The average cost of running the DGs are 19 INR/kWh [6]. Three of the generators are connected to all of the substations so that energy can be supplied to the whole campus in case of an outage. The two remaining generators are connected solely to substation 5 in order to supply for the School of Engineering. At peak all three of the generators supplying the whole campus must be operated as well as one or two of those connected to substation 5. During off peak it is usually enough to run two of the main generators and one of those connected to the School of Engineering.
The DGs are controlled by manual switches at each of the substations. Previously there were automatic switches, however, these were all damaged by lightning. Instead the generator stations are manned every hour of the day in case that there is an outage and the DGs need to be turned on.
D. Loads
The loads are in the form of educational buildings, housing for students and employees and other functions such as stores etc. The data given for this project does not cover all the loads of the campus. The loads which will be considered and discussed can be found in Table II. In order to make assumptions about the load profiles of the loads they have been divided into types based on the usage. Buildings which are used during office hours has been defined as Departments.
This includes department buildings, administrative buildings as well as library and other buildings used primarily during the day. The second category is Hostels. These are the buildings where the students of the university live. The hostels are used primarily during evening and night. Lastly data is given for the water treatment plant (WTP). This is where the water at the
campus is processed and pumped out to the rest of the campus.
The types of buildings and their function will be discussed further in Section XIII.
A substantial part of the load of the university is the School of Engineering [6]. However, as no data is given for this part of the load it will not be included in this project.
E. Electricity prices at Tezpur University
The PV system was purchased in a Renewable Energy Service Company (RESCo) model. This model entails that the power company payed for the initial installation costs for the panels. The energy produced is then sold to the university at a lower cost than that purchased from the grid [27], [28].
The installation company APPDCL is responsible for all maintenance and operation for 25 years after the installation [29].
APPDCL is also the main grid operator in the Assam region. The tariff system works in such a way that there are different tariff structures for different categories [30]. For domestic supply the the rate changes depending on the amount of electricity consumed each month. In industries the tariff changes depending on the time of day. However, in educational institutions such as TU the rates are fixed. The tariff for TU is 6.75 iR/kWh. A penalty or rebate can be used depending on the power factor (PF) of the consumer. If the PF is below 0.85 the consumer must pay a penalty. If the PF is over 0.85 the power company will give a rebate depending on how high the power factor is.
XI. PV-PRODUCTION
In this project The National Solar Database (NSRDB) has been used to estimate the irradiance at TU. For estimation of
Table II: Connections between substations and loads
Substation 1 2 3 4
Departments: KBRA Acca-1 DCS DMBBT
SAC DE DES
Council hall Library SAIC Acca-2
Hostels: KWH PMCWH CMH
DWH PMH
PWH SMH
BWH NWH
KMH NMH SWH
Other: WTP
irradiance in India NSRDB uses the SUNY Semi-Empirical Model [31]. The data used is: hourly frames from a geo- stationary satellite covering the Indian subcontinent, ground elevation, climatological precipitable water and ozone from previous simulations, and monthly aerosol optical depth. This data is used to convert a cloud index into a clearness index.
The Solis clear sky model is then used to estimate the Global Horizontal Irradiance (GHI). This in turn is used to estimate the Direct Normal Irradiance (DNI) and the Diffuse Horizontal Irradiance (DHI). The in-plane irradiance is then calculated as:
G = DN I + DHI (6)
The next step is to sum up the in-plane irradiance for each day of the year. By dividing each hourly data with the total for that specific day the percentage per hour per day is calculated.
Data regarding the PV-production was gathered from Oc- tober 2018 to September 2019 [30]. Each month the daily production at each inverter was documented. By multiplying the daily production with the percentage of production per hour an estimation of the hourly production was given. By then using the Matlab-function mean the average production each day per month could be estimated and can be seen in Figure 6.
XII. OUTAGES
The data given for the outages is only over sixteen months and therefore the data is not detailed enough to determine the probability each month. Due to this the probability of an outage is assumed to be the same each day of the year.
The analysis is divided into two parts, the starting time and the outage length. First the distribution of when the outage occurs was estimated. By checking the starting time of each outage one could find the distribution. The outages occur in a normal distribution and can be seen in Fig. 7. By using the expected value of the distribution one can determine the most probable starting time for the outage.
Next the probable length of the outage is determined. This distribution was assumed to be an exponential distribution and can be seen in Fig. 8. The expected value of the starting time of the outage is tstart,exp =11:32 AM. By looking at the average consumption at the expected starting time the expected consumption was estimated. The energy needed was E(tstart,exp) = 86.8 kW. The length of the outage is expected to be tlength,exp= 1.87 minutes.
The probability of an outage occurring during any given day was calculated by dividing the number of outages with the number of days measured.
poutage= N umber of Outages N umber of Days = 483
485 = 99.59% (7) These distributions could then be used to estimate the power needed during outages per year.
Eexp= E(tstart,exp) · tlength,exp· poutage· 365 (8) The expected energy needed during the outages can be estimated to be 23.6 MWh per year.
XIII. LOADS
A. Student hostels
The hostels at the campus is where the students live. In this report the nine hostels will be analysed.
1) Load profile for the hostels: In December 2018 eight load management units (LMUs) were installed at substation 1 by Pamoja Cleantech [6]. These monitor the load with a high resolution. The LMUs measure the current and voltage of the loads. By using the formula:
P (t) = V (t) · I(t) (9)
where V is the voltage and I is the current, the power (P) per hour can be calculated. The data used for this project was recorded over ten days from 1st to 11th of December 2018.
By adding the consumption of all the measured loads and dividing by the total load of a day a hourly percentage of the daily consumption was calculated. These percentages form a load profile for the hostels which can be seen in Fig. 9.
2) Load per month: Consumption data for the hostels was given as a sheet with the meter readings recorded by hand from June 2018 to September 2019. The meters used at TU have a small transformer which converts the current to a lower more manageable level [30]. The ratio of this transformation is called the Current Transformer ratio (CT- ratio). The transformed current was then measured and the power was calculated. That is the data from the meter readings.
By calculating the difference between the meter readings per
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month
0 100 200 300 400 500
Production PV [kWh]
Figure 6: Average production per hour each month
-5 0 5 10 15 20 25 30
Hour 0
5 10 15 20 25 30 35 40
Number of outages
Figure 7: Distribution of time when outages start
0 2 4 6 8 10 12 14
Length [Hours]
0 50 100 150
Number of outages
Figure 8: Distribution of length of outages
month and multiplying with the CT-ratio, rCT of each meter the consumption could be calculated.
Econ,i= (Xi− Xi−1) · rCT (10) Where i is the month starting with i = 0 corresponding to June 2018. As readings were given for the months of July, August and September both 2018 and 2019 the average of the two was used in the analysis. This gives the consumption per hostel per month of one full year.
By dividing the consumption per month with the number of days for that specific month the consumption per day during that month could be estimated. By combining the load profile in Fig. 9 with the consumption per day, one gets the consumption per hour a typical day per month for each hostel.
This data can then be used either individually or be summed to get the total consumption of the hostels as shown in Fig.
12.
B. Departments
The departments are considered to be all the buildings which are primarily used during the work day. This includes department buildings, the auditorium, the student activity center, academic buildings, the library, the council hall and the SAIC building.
1) Load profile for departments: The occupancy of the departments was examined in the masters thesis regarding DSM [6]. A survey was conducted with TU employees which covered the occupancy among other things. 50 people an- swered the survey. The office is mostly occupied between 09:00-17:00 and some are there earlier in the morning or later at night. This data was used as a base to find the load profile for the departments. By dividing the occupancy each hour with the total amount of hours the deparments were occupied per day, the load profile in Fig. 10 was found.
2) Load per month: As with the consumption of the hostels, the consumption data for the departments was given as meter readings per month. Using the same method as in section XIII-A2 the consumption per month could be estimated. As
