Performance analysis of a microgrid with renewable energy sources

1 JOSIP JURAJ STROSSMAYER UNIVERSITY OF OSIJEK FACULTY OF ELECTRICAL ENGINEERING, COMPUTER SCIENCE AND INFORMATION TECHNOLOGY OSIJEK Undergraduate study PERFORMANCE ANALYSIS OF A MICROGRID WITH RENEWABLE ENERGY SOURCES Bachelor’s degree final paper Francesc Garcia Ferrando Osijek, 2018 2 Table of contents 1. Introduction _______________________________________________________________ 3 1.1. Main goal of the paper _________________________________________________________ 3 1.2. Economic concepts ____________________________________________________________ 3 2. Microgrids _________________________________________________________________ 5 3. PV panels theory ____________________________________________________________ 8 3.1. The solar resource _____________________________________________________________ 8 3.2. Functioning of the PV panels ____________________________________________________ 9 4. Batteries theory ____________________________________________________________ 13 5. Inverters theory ____________________________________________________________ 15 6. Study Case Microgrid _______________________________________________________ 17 6.1. PV Panels ___________________________________________________________________ 20 6.2. Batteries ____________________________________________________________________ 26 6.3. Inverter ____________________________________________________________________ 29 7. Simulations _______________________________________________________________ 31 7.1. On-grid microgrid with cheapest equipment ______________________________________ 31 7.2. On-grid microgrid with middle priced equipment __________________________________ 36 7.3. On-grid microgrid with most expensive equipment _________________________________ 42 7.4. Islanded microgrid with cheapest equipment ______________________________________ 48 7.5. Islanded microgrid with middle priced equipment _________________________________ 53 7.6. Islanded microgrid with most expensive equipment ________________________________ 57 7. Conclusions ______________________________________________________________ 63 References __________________________________________________________________ 64 Abstract: ___________________________________________________________________ 66 Curriculum Vitae ____________________________________________________________ 67 3 1. Introduction In a world that tends to collapse, finding new and more efficient ways to take advantage of renewable energy sources may be the key to changing that trend and redirecting it towards a more sustainable future. To do so, a paradigm change must be done; the microgrids powered by renewable energy have to take over. Following the European Union energy policies in these last years, and with the increasing electricity demand and dependence from it, distributed generation based on renewable energy sources can be the alternative to the traditional centralized and non-renewable energy production model. They may also be vital to accomplish the EU objectives known as the three twenties 20-20-20; which tries to achieve a 20% of power generation coming from renewable energy sources, greenhouse gases emissions 20% lower than the 1990 levels, and 20% increase in energy efficiency in all the EU countries before 2020. 1.1. Main goal of the paper The main goal of this paper is to compare different types of equipment, depending on the price, in two case studies: on-grid microgrid and isolated microgrid. To do so, an electrical performance and cost-benefit analysis using HomerPro will be conducted. The objective is to see if these types of microgrids are viable both economically and technically. 1.2. Economic concepts In a society where the importance of a project is determined by its economic viability, is important to define first some economic concepts that will be important in this project, the Levelized Cost of Energy, the Net Present Cost and the discounted payback period. The Levelized Cost of Energy (LCOE), according to [1, pp. 34], is a method to compare different types of energy sources and technologies over the lifetime of the projects, considering also how the passing of time can affect the prices of maintenance and fuel, as well as the possibility of a major breakdown of the plant. LCOE is calculated as in (1-1). 𝐿𝐶𝑂𝐸 =[𝑅·𝑐𝑝 𝐻·𝑓]+[𝑙·(𝑐𝑜 𝐻·𝑓)]+[𝑙·(𝑐𝑓 𝐻·𝑓)] (1-1) As we see in the formula, the LCOE it is based on its initial investment (cost of the plant: cp), the operation and maintenance costs (co) and the fuel costs (cf). It also takes into account the capacity 4 factor (f), the hours per year that the plant is functioning (H), the capital recovery factor (R) and the levelization factor (l). The Net Present Cost (NPC) of a component is the present value of the installation, operation and maintenance of that component, minus the present value of the profit generated by it, over its lifetime. The discounted payback period is a method used to determine how profitable is a project, reflecting the amount of time necessary to recover the initial investment taking into account the different cash flows for every period. 5 2. Microgrids Nowadays, production of energy is based on big, centralized power plants fueled by nonrenewable primary energy, such as coal, petroleum or natural gas. These electricity-producing power plants normally use synchronous generators and turbines; moreover, they are usually far away from the loads and consumers of the energy they are producing, making the transmission and distribution lines to be long. These facts of the traditional large energy producing systems generate a series of problems, which can be divided in three levels:  Generation: Inefficient generating systems, very high renewal costs when obsolete or broken, emission of gases that contribute to greenhouse effect and other pollution due to combustion of fossil fuels, nuclear waste and possibility of reactor failure, etcetera.  Transport: Transient and dynamic instability due to consumption, voltage instability due to reactive loads, thermal limit to transmitted power, possibility of failure by natural elements or other accidents in long transmission lines, etcetera.  Distribution: Necessity of power transformation stations, non-linear loads, possibility of failure due to natural elements or other accidents, etcetera. The scheme of the typical electric network is shown in Figure 2.1 Figure 2.1. Simplified one-line diagram of the actual electrical network, source [2] 6 These problems can be minimized at the same time as we maintain or even improve the capabilities of the network by changing the paradigm of centralized energy production and introducing microgrids to the system. A microgrid, as defined by the U.S. Department of Energy [3], is a “group of interconnected loads and distributed energy resources within clearly defined electrical boundaries that acts as a single controllable entity with respect to the grid. A microgrid can connect and disconnect from the grid to enable it to operate in both grid-connected or island-mode”. Professor Robert H. Lasseter first used “microgrid” in 1998 referred to it as a combination of three key elements: DG (Distributed Generation) + PFC (Power Flow Control) + ESS (Energy Storage Systems). In Figure 2.2 is depicted the typical elements of a microgrid. Figure 2.2. Typical elements of the microgrid, source [4] As we see in Figure 2.2., and taking it as an example of a typical microgrid connected to the main grid, we distinguish between the distributed generation (photovoltaic panels and auxiliary genset group), the energy storage system (the batteries) and the power flow control (the interconnected converters). This microgrid distribution brings a series of advantages for the system:  Bust the usage of renewable energy sources, contributing to the fight against climate change and thus helping the environment.  Improves the resiliency and power quality of the network, making it more reliable. 7  Has the capacity to operate (some of them, mostly microgrids connected to main grid, only for a certain time if they don’t have an appropriate) without the electricity supply of the main grid, making them suitable for emergency operations while blackouts.  Enables the participation in new markets for demand response and auxiliary services.  Optimization of the usage of energy. However, not everything is good about microgrids, as they are facing some problems also, namely:  The price of some components of the microgrid can be too expensive for most of the population.  Large differences between load and production for intermittent renewable sources such as solar panels and wind generators.  Durability of the components  Protection  Legal and regulatory problems in some countries 8 3. PV panels theory 3.1. The solar resource Most of the energy we are using today comes from the Sun; even fossil fuels obtained their energy from it in the past. The Sun is the star on our planetary system and releases a huge amount of energy because of its nuclear fusion where hydrogen is converted into helium. From all this energy, around 5.6·1024𝐽 arrives to the atmosphere of our planet [5], where the 31% of it is reflected. The rest enters the atmosphere, and apart from the atmosphere absorbing a small amount of it, reaches the surface where an average of 4.2% is reflected back into the atmosphere. In a particular spot, the sum of the direct radiation arriving from the Sun and the diffuse radiation is the global radiation (Gg), in Figure 3.1 is shown a map of the amount of global irradiation arriving to Croatia. Figure 3.1. Global horizontal irradiation in Croatia, source [6] This solar resource can be exploited using three methods:  Passive solar energy: Using architecture to take advantage of the Sun and its benefits in the buildings (for example with an efficient placement of the windows). It is the simplest way to use the solar resource. 9  Thermal solar energy: Converts the solar irradiation in heat, being able to use this heat afterwards for buildings’ hot water or heating system.  Photovoltaic solar energy: Converts the solar radiation in electricity through solar cells. We will be deepening this one in this project. 3.2. Functioning of the PV panels The photovoltaic solar panels are made up of a transparent top sheet with an anti-reflective layer and a lower enclosure. In between, we can find the electric connexions and the converter substrate. Figure 3.2 shows the typical structure of a solar cell and its equivalent circuit diagram. Figure 3.2. Structure of a typical solar cell [5] The photovoltaic effect converts the electromagnetic radiation of the light into electricity. When the photons make contact with a semiconductor (such as silicon), they transfer their energy to electrons from the valance band, causing excitation on them, enabling these free electrons to cross the junction, and thus creating a current, because one side of the junction will have a positive charge and the other will have a negative charge. This current created by the photovoltaic effect can be used to power an electric circuit. The current created by the photovoltaic panels is direct current. It can be directly used if we want to power a device that needs DC to function or if we want to store it in batteries. However, it can be transformed into AC using an inverter and, this way, use it to give current to a load that needs AC or to be sold into the main grid, with the subsequent subventions. 16 - Pure sine wave: This type of inverters provides a waveform to its output that, for practical purposes, can be considered identical to that of the general electrical network, thus allowing the power supply of any consumer device or, where appropriate, the connection to the network. - Modified sine wave (or trapezoidal): Intermediate between the two previous ones, allows to expand the spectrum of consumption and power elements, limited in the modulated square wave. 17 6. Study Case Microgrid We will be conducting two case studies, one of them with our microgrid working in isolated mode, and the other one working connected to the main grid. We will be using the program HOMER Pro. In this program, specialized in microgrid analysis, we can find all the tools we need to perform the studies we want. Moreover, this program counts in its database with all the equipment necessary to create a microgrid. First, all the simulations and case studies will be conducted during a project lifetime of 25 years, and taking into account a discount rate of 8%, an inflation rate of 2% and a 0% annual capacity shortage. The location will be the one from the Faculty of Electrical Engineering in Osijek, as shown in Figure 6.1: Figure 6.1. Location of the project. All the environmental data (temperature, irradiation, clouds, hours of sun, etcetera) is taken automatically from the HOMER software for this location. 18 In every one of our two case studies, we will simulate our microgrid performance for three different types of equipment depending on the price. The inverter will be the same in all case studies, and so will the load. However, apart from having the main grid in the first case study and not in the second, we will perform the analysis with the following equipment:  Load: A prototypical house provided by HOMER Pro, will have these characteristics: An average energy consumption of 11.27 kWh/day and an average power consumption of 0.47 kW, with a peak of consumption of 2.81 kW and a load factor of 0.17. The load will use Alternate Current. The peak month will be July, and the average load (kW) per hour of every month will be: In Table 6.1 is shown the average load per hour on Weekdays: Table 6.1. Average load (kW) per hour on weekdays. Hour Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0.087 0.09 0.098 0.109 0.12 0.128 0.131 0.128 0.12 0.109 0.098 0.09 1 0.076 0.079 0.085 0.095 0.105 0.111 0.114 0.111 0.105 0.095 0.085 0.079 2 0.076 0.079 0.085 0.095 0.105 0.111 0.114 0.111 0.105 0.095 0.085 0.079 3 0.076 0.079 0.085 0.095 0.105 0.111 0.114 0.111 0.105 0.095 0.085 0.079 4 0.262 0.271 0.294 0.327 0.36 0.383 0.392 0.383 0.36 0.327 0.294 0.271 5 0.4 0.415 0.45 0.5 0.55 0.585 0.6 0.585 0.55 0.5 0.45 0.415 6 0.44 0.457 0.495 0.55 0.605 0.644 0.66 0.644 0.605 0.55 0.495 0.457 7 0.4 0.415 0.45 0.5 0.55 0.585 0.6 0.585 0.55 0.5 0.45 0.415 8 0.336 0.349 0.378 0.42 0.462 0.491 0.504 0.491 0.462 0.42 0.378 0.349 9 0.344 0.357 0.387 0.43 0.473 0.503 0.516 0.503 0.473 0.43 0.387 0.357 10 0.396 0.411 0.446 0.495 0.545 0.579 0.594 0.579 0.545 0.495 0.446 0.411 11 0.426 0.442 0.48 0.533 0.586 0.624 0.64 0.624 0.586 0.533 0.48 0.442 12 0.553 0.574 0.622 0.691 0.76 0.808 0.829 0.808 0.76 0.691 0.622 0.574 13 0.415 0.431 0.467 0.519 0.571 0.607 0.623 0.607 0.571 0.519 0.467 0.431 14 0.334 0.347 0.376 0.418 0.46 0.489 0.502 0.489 0.46 0.418 0.376 0.347 15 0.318 0.33 0.357 0.397 0.437 0.464 0.476 0.464 0.437 0.397 0.357 0.33 16 0.327 0.339 0.368 0.409 0.45 0.479 0.491 0.479 0.45 0.409 0.368 0.339 17 0.526 0.546 0.592 0.658 0.724 0.77 0.79 0.77 0.724 0.658 0.592 0.546 18 0.985 1.022 1.108 1.231 1.354 1.44 1.477 1.44 1.354 1.231 1.108 1.022 19 0.802 0.832 0.903 1.003 1.103 1.174 1.204 1.174 1.103 1.003 0.903 0.832 20 0.541 0.561 0.608 0.676 0.744 0.791 0.811 0.791 0.744 0.676 0.608 0.561 21 0.384 0.398 0.432 0.48 0.528 0.562 0.576 0.562 0.528 0.48 0.432 0.398 22 0.24 0.249 0.27 0.3 0.33 0.351 0.36 0.351 0.33 0.3 0.27 0.249 23 0.163 0.169 0.184 0.204 0.224 0.239 0.245 0.239 0.224 0.204 0.184 0.169 19 In Table 6.2 is shown the average load per hour on weekends: Table 6.2. Average load (kW) per hour on weekends. Hour Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0.087 0.09 0.098 0.109 0.12 0.128 0.131 0.128 0.12 0.109 0.098 0.09 1 0.076 0.079 0.085 0.095 0.105 0.111 0.114 0.111 0.105 0.095 0.085 0.079 2 0.076 0.079 0.085 0.095 0.105 0.111 0.114 0.111 0.105 0.095 0.085 0.079 3 0.076 0.079 0.085 0.095 0.105 0.111 0.114 0.111 0.105 0.095 0.085 0.079 4 0.262 0.271 0.294 0.327 0.36 0.383 0.392 0.383 0.36 0.327 0.294 0.271 5 0.4 0.415 0.45 0.5 0.55 0.585 0.6 0.585 0.55 0.5 0.45 0.415 6 0.44 0.457 0.495 0.55 0.605 0.644 0.66 0.644 0.605 0.55 0.495 0.457 7 0.4 0.415 0.45 0.5 0.55 0.585 0.6 0.585 0.55 0.5 0.45 0.415 8 0.37 0.383 0.416 0.462 0.508 0.541 0.554 0.541 0.508 0.462 0.416 0.383 9 0.378 0.393 0.426 0.473 0.52 0.553 0.568 0.553 0.52 0.473 0.426 0.393 10 0.436 0.452 0.49 0.545 0.599 0.637 0.653 0.637 0.599 0.545 0.49 0.452 11 0.469 0.487 0.528 0.586 0.645 0.686 0.704 0.686 0.645 0.586 0.528 0.487 12 0.608 0.631 0.684 0.76 0.836 0.889 0.912 0.889 0.836 0.76 0.684 0.631 13 0.457 0.474 0.514 0.571 0.628 0.668 0.685 0.668 0.628 0.571 0.514 0.474 14 0.368 0.382 0.414 0.46 0.506 0.538 0.552 0.538 0.506 0.46 0.414 0.382 15 0.349 0.362 0.393 0.437 0.48 0.511 0.524 0.511 0.48 0.437 0.393 0.362 16 0.36 0.373 0.405 0.45 0.495 0.526 0.54 0.526 0.495 0.45 0.405 0.373 17 0.526 0.546 0.592 0.658 0.724 0.77 0.79 0.77 0.724 0.658 0.592 0.546 18 0.985 1.022 1.108 1.231 1.354 1.44 1.477 1.44 1.354 1.231 1.108 1.022 19 0.802 0.832 0.903 1.003 1.103 1.174 1.204 1.174 1.103 1.003 0.903 0.832 20 0.541 0.561 0.608 0.676 0.744 0.791 0.811 0.791 0.744 0.676 0.608 0.561 21 0.384 0.398 0.432 0.48 0.528 0.562 0.576 0.562 0.528 0.48 0.432 0.398 22 0.24 0.249 0.27 0.3 0.33 0.351 0.36 0.351 0.33 0.3 0.27 0.249 23 0.163 0.169 0.184 0.204 0.224 0.239 0.245 0.239 0.224 0.204 0.184 0.169 20 The scaled data daily profile per month of the load will be the one shown in Figure 6.2.: Figure 6.2. Scaled graphs of daily load consumption per month.  Grid: The price of purchasing energy from the grid will be 0.140 €/kWh, while the price of selling the electricity to the main grid will be 0.126 €/kWh. Both of them at simple rates. The emissions coming from generating that electricity using the traditional power plants on which the actual grid is relying on will be: 632 g/kWh of Carbon Dioxide, 2.74 g/kWh of Sulphur Dioxide, and 1.34 g/kWh of Nitrogen Oxides.  PV panels: Jinko JKM 275-60, CanadianSolar MaxPower CS6U-330p, Sharp ND-250QCS  Batteries: Trojan SAGM 06 375, Hoppecke 24 OPzS 3000, Tesla Powerwall 2.0  Inverter: Leonics STP-219Cp 15 kW. 6.1. PV Panels In this project, three types of solar panels from three different manufacturing companies will be used: Jinko, Canadian Solar and Sharp. The prices of each one of the solar panels have two components, the installation price, which is calculated to be 450 € per solar panel, and the price of the panel itself. The operation and maintenance costs of the panels have been calculated assuming a 2 % of the initial cost per year We installed 5 kW of solar panels in each of the case studies because is enough to satisfy the consumption required by the load. The current generated by the panels is DC, they have no tracking system and the default panel slope is 45.46º. 21 Apart from the datasheets (that can be found in the annexes) of the equipment, we also took info from the HOMER database [9] The Jinko JKM275-60 [10] is a flat plate solar panel, composed of 60 polycrystalline cells. We can see its most remarkable characteristics in Table 6.3. Table 6.3. Jinko solar panel characteristics [10] Capital cost/Replacement 572.40 € O&M costs 11.45 €/year Derating factor 88 % Temperature coefficient -0.41 %/ºC Operating temperature 45 ºC Efficiency 16.8 % Ground reflectance 20 % Maximum power (Pmax) 275 W Maximum voltage (Vmp) 32 V Maximum current (Imp) 8.61 A Open-circuit voltage (Voc) 39.1 V Short-circuit current (Isc) 7.44 A The electrical performance and temperature dependence graphs are detailed in Figure 6.3 and 6.4 Figure 6.3. Current-Voltage and Power-Voltage curves. [10] 22 Figure 6.4. Temperature dependence of Isc, Voc, Pmax.[10] In Figure 6.5 we can see the engineering drawing of the Jinko solar panel. Figure 6.5. Dimensions of the Jinko solar panel.[10] 23 The CanadianSolar MaxPower CS6U-330P [11] is a flat plate solar panel, composed of 72 polycrystalline cells. We can see its most remarkable characteristics in the Table 6.4 Table 6.4. CanadianSolar solar panel characteristics. [11] Capital cost/Replacement 632.12 € O&M costs 12.64 €/year Derating factor 88 % Temperature coefficient -0.41 %/ºC Operating temperature 45 ºC Efficiency 16.97 % Ground reflectance 20 % Maximum power (Pmax) 330 W Maximum voltage (Vmp) 37.5 V Maximum current (Imp) 8.80 A Open-circuit voltage (Voc) 45.9 V Short-circuit current (Isc) 9.31 A The electrical performance and temperature dependence graphs are detailed in Figure 6.6. 24 Figure 6.6. I-V curves of the CanadianSolar solar panel.[11] In Figure 6.7 we can see the engineering drawing of the CanadianSolar solar panel. Figure 6.7. Dimensions of the CanadianSolar solar panel. [11] 25 The Sharp ND-250QCS [12] is a flat plate solar panel, composed of 60 polycrystalline silicon cells. We can see its most remarkable characteristics in the Table 6.5. Table 6.5. Sharp solar panel characteristics. [12] Capital cost/Replacement 688,10 € O&M costs 13,76 €/year Derating factor 88 % Temperature coefficient -0,485 %/ºC Operating temperature 47,5 ºC Efficiency 15,3 % Ground reflectance 20 % Maximum power (Pmax) 250 W Maximum voltage (Vmp) 29,8 V Maximum current (Imp) 8,40 A Open-circuit voltage (Voc) 38,3 V Short-circuit current (Isc) 8,90 A In Figure 6.8 we can see the engineering drawing of the Sharp solar panel. Figure 6.8. Dimensions of the Sharp solar panel. [12] 32 While this is a summary of the microgrid using the solar panels, HOMER also offers you the results of the simulation in the case where only the main grid is being used. In that case, the LCOE would have been of 0.140 €, the NPC 7,445 € and the operating cost 575.90 €. Of course, we would not have an initial capital investment. To compare the economics of the two possible solutions to the simulation, HOMER offers the Table 7.2: Present worth 1,977 € Annual worth 185 €/year Return on investment 11 % Internal rate of return 10.3 % Simple payback 7.52 years Discounted payback 13.55 years Table 7.2. Compared economics between optimal solution and only using the grid. In Figure 7.2 we can see the graph of the cost summary and in Table 7.3 the breakdown of the NPC. Table 7.3. Breakdown of the costs. Component Capital Replacement O&M Salvage Total Grid 0 € 0 € -1,773.66 € 0 € -1,773.66 € PV system 2,862 € 0 € 739.97 € 0 € 3,601.97 € Inverter 1,800 € 1590.19 € 465.39 € -215.60 € 3,639.98 € System 4,662 € 1,590.19 € -568.30 € -215.60 € 5,468.29 € As we’ve seen, the operation and maintenance costs of the grid are negative, that’s because of the energy we sell to the utility. Figure 7.2. Summary of costs. 33 In Figure 7.3 we have the discounted cash flow by cost type of our microgrid in the 25 years of the duration of the project. Figure 7.3.: Cash flow. As to the electrical performance of our microgrid, the total production of our system is 9,084 kWh/year: 6,472 kWh/year produced by the Jinko solar panels (71.2%) and 2,612 kWh/year from grid purchases (28.8%). The consumption of our load is 4,114 kWh/year, and adding the 3,991 kWh/year of grid sales, we have our total consumption that ascends to 8,105 kWh/year. Is remarkable than almost half of our consumption (49.2%) is destined to grid sales, making the microgrid economically viable. We have an excess electricity of 750 kWh/year (the 8.26% of our production) but we don’t have unmet electric load or capacity shortages, so the microgrid is reliable in terms of electrical performance. In Figure 7.4 we can see the average electricity production per month which, as stated before, will match perfectly with our electricity needing from the load and also, especially during the daylight, we will be able to sell electricity to the grid. Figure 7.4. Monthly average electric production. The renewable fraction of our microgrid is 67.8%, but the maximum renewable penetration is 177%. This high value of more than the 100% is because we take into account the sales to the grid; 34 if we divide the total renewable production by the load, the value is 79.8%, and if we divide the same renewable production by the generation, the fraction is 71.2%. The rated capacity of our Jinko solar panels, as we said before, is 5 kW. In Table 7.4 we can see some remarkable values from their functioning in our microgrid. Table 7.4. Jinko solar panels Mean output 0.739 kW Mean output 17.7 kWh/day Capacity factor 14.8 % Hours of operation 4,379 hours/year Total production 6,472 kWh/year Levelized cost 0.0431 €/kWh In Figure 7.5 there is the graph of the PV power output from the Jinko solar panels. We can appreciate how, logically, they produce more during the central hours of the day and during the summer. Figure 7.5. Jinko solar panels power output per day and hour. The main details of the interactions with the utility are depicted in Figure 7.6, while the graph of the energy we purchase from the grid and the energy we sell to it are shown in Figures 7.7 and 7.8. As we will see, the majority of the grid sales are made also during the central hours of the day, especially on summer months, when the PV panels are functioning at the maximum. On the other 35 hand, we make the majority of the grid purchases during the evening, when there is not enough sun for the PV panels to work so we need the back up from the utility. Figure 7.6. Grid data per month. Figure 7.7. Energy purchased from the grid per day and hour. Figure 7.8. Energy sold to the grid per day and hour. 36 In Table 7.5 we can see some remarkable values from the functioning of the Leonics inverter in our microgrid. Table 7.5. Leonics inverter data. Mean output 0.627 kW Maximum output 3 kW Capacity factor 20.9 % Hours of operation 2,980 hours/year Energy out 5,493 kWh/year Energy in 5,722 kWh/year Losses 2,29 kWh/year In Figure 7.9 is shown the graph of the inverter output, which, as expected, is functioning at the same time as the PV panels. Figure 7.9. Inverter output per day and hour. The gases emissions we are saving to the planet by not purchasing energy to the grid, and in addition selling to it are: 872 kg/year of Carbon dioxide, 3.78 kg/year of Sulphur dioxide and 1.85 kg/year of Nitrogen oxides. 7.2. On-grid microgrid with middle priced equipment The scheme of our microgrid with the CanadianSolar panels is shown in Figure 7.10. 37 Figure 7.10.: Scheme of the microgrid. After 8 simulations the HOMER software shows us a summary of the results as in Table7.6: Table 7.6. Summary of the simulation. LCOE 0.0558 € NPC 5,843 € Operating cost 68.29 € Initial capital 4,961 € Renewable fraction 67.8 % Capital cost of the PV panels 3161 € Production of the PV panels 6,473 kWh Inverter mean output 0.627 kW Inverter capital cost 1,800 € Energy purchased from the grid 2,612 kWh Energy sold to the grid 3,992 kWh The comparison of the economics between the optimal solution using PV panels and the solution using only the grid (with the same costs as the simulation before), is offered by HOMER in the Table 7.7. Table 7.7. Compared economics between optimal solution and only using the grid. Present worth 1,601 € Annual worth 150 €/year Return on investment 10.2 % Internal rate of return 9.2 % Simple payback 8.08 years Discounted payback 14.88 years 38 In Figure 7.11 we can see the graph of the cost summary and in Table 7.8 the breakdown of the NPC. Table 7.8. Breakdown of the costs. Component Capital Replacement O&M Salvage Total Grid 0 € 0 € -1,774.31 € 0 € -1,774.31 € PV system 3,160.60 € 0 € 817.17 € 0 € 3,977.77 € Inverter 1,800 € 1,590.19 € 465.39 € -215.60 € 3,639.98 € System 4,960.60 € 1,590.19 € -491.74 € -215.60 € 5,843.44 € Also in this simulation, we have a negative maintenance of the grid because of the sales from the PV panels. Figure 7.11. Cost summary. In Figure 7.12 we can find the discounted cash flow by cost type of our microgrid in the 25 years of the duration of the project. 39 Figure 7.12. Discounted cash flow. Referring to the electrical performance of this microgrid using the CanadianSolar PV panels, the total production of our system is 9,085 kWh/year: 6,473 kWh/year produced by the solar panels (71.2%) and 2,612 kWh/year from grid purchases (28.8%), the same amount as the previous equipment. The consumption of our load is also the same as before (and it won’t change neither in the next simulation, as the load doesn’t change) 4,114 kWh/year, and adding the 3,992 kWh/year of grid sales, we have our total consumption that ascends to 8,105 kWh/year. Is remarkable than almost half of our consumption (49.2%) is destined to grid sales, making the microgrid economically viable. We have an excess electricity of 751 kWh/year (the 8.26% of our production) but we don’t have unmet electric load or capacity shortages again, so the microgrid is also reliable in terms of electrical performance. In Figure 7.13, we can see the average electricity production per month. Figure 7.13. Monthly average electric production. The renewable fraction of our microgrid is 67.8%, and the maximum renewable penetration is 177% again. If we divide the total renewable production by the load, the value is 79.9%, and if we divide the same renewable production by the generation, the fraction is 71.2%. 40 In Table 7.9 we can see some remarkable values from the functioning of the CanadianSolar panels in our microgrid. Table 7.9. CanadianSolar PV panels Mean output 0.739 kW Mean output 17.7 kWh/day Capacity factor 14.8 % Hours of operation 4,379 hours/year Total production 6,473 kWh/year Levelized cost 0.0475 €/kWh In Figure 7.14 we can find the graph of the PV power output from the CanadianSolar PV panels. Figure 7.14. PV power output per day and per hour. The main details of the interactions with the utility are depicted in Figure 7.15, while the graph of the energy we purchase from the grid and the energy we sell to it are shown in Figures 7.16 and 7.17. The interactions remain nearly exactly the same from the ones in the previous simulation; in fact, we’ve seen that the difference of price between the Jinko solar panels and the ones from CanadianSolar doesn’t reflect in the electrical performance, and also the LCOE is higher than the one before, so we can say that is a better option to choose the cheaper equipment between these two. 41 Figure 7.15. Grid data per month. Figure 7.16. Energy purchased from the grid per day and month. Figure 7.17. Energy sold to the grid per day and month. In Table 7.10 we can see some remarkable values from the functioning of the Leonics inverter with the CanadianSolar panels. 48 Figure 7.27. Inverter output per day and hour The emissions we are saving are: 853 kg/year of Carbon dioxide, 3.70 kg/year of Sulphur dioxide and 1.81 kg/year of Nitrogen oxides. 7.4. Islanded microgrid with cheapest equipment Now we start with the simulations of the off-grid microgrid. We can see the scheme of it in Figure 7.28. Figure 7.28. Scheme of the off-grid microgrid. The summary offered by HOMER is shown in Table 7.16. Notice that now we left HOMER to optimize the number of batteries. This will increase a lot the cost of the overall microgrid, and also means that the renewable fraction is 100%, as we are relying only in the PV panels and the batteries. 49 Table 7.16. Summary of the simulation. LCOE 0.549 € NPC 30,782 € Operating cost 498.16 € Number of batteries 41 Initial capital 24,342 € Capital cost of the PV panels 2,862 € Production of the PV panels 6,472 kWh Inverter mean output 0.469 kW Inverter capital cost 1,800 € Autonomy of the batteries 172 hours Annual throughput of the batteries 2,591 kWh In the cost summary, depicted in Figure 7.29, we can clearly see how the batteries are the equipment that rises the cost compared to the on-grid microgrids; this is because of the necessity to have a back-up source of power when the PV panels are not working and the load still needs to receive electricity. In that case, we need a large amount of batteries to face the supply of the house on their own. Figure 7.29. Cost summary. In the following Table 7.17 we can find the detailed cost summary. 50 Table 7.17. Detailed cost summary of the equipment. Component Capital Replacement O&M Salvage Total PV system 2,862 € 0 € 739.97 € 0 € 3,601.97 € Inverter 1,800 € 1,590.19 € 465.39 € -215.60 € 3,639.98 € Batteries 19,680 € 0 € 5088.27 € -1,228.20 € 23,540.07 € System 24,342 € 1,590.19 € 6,293.63 € -1,443.80 € 30,782.02 € As we saw in the previous table, the batteries take the 76.47% of the total cost of the microgrid. The discounted cash flow is presented in Figure 7.30. Figure 7.30. Discounted cash flow. The electrical performance of this microgrid is based purely in the PV panels with the support of the batteries. In Figure 7.31 we can appreciate how almost all of the electric load is met by the PV panels generation. However, unlike the on-grid microgrid, where we counted with the backup of the utility that works 24/7, in this case study we don’t have that lifesaver. In this case, we have an unmet electric load of 2.50 kWh/year, which represents the 0.0607 % of the total, we also have a capacity shortage of 2.86 kWh/year, the 0.0695 % of the total. 51 Figure 7.31. Monthly average electricity production of the Jinko panels. The details on the energy production of the Jinko PV panels is shown in the following Table 7.18 Table 7.18. Jinko PV panels Mean output 0.739 kW Mean output 17.7 kWh/day Capacity factor 14.8 % Hours of operation 4,379 hours/year Total production 6,472 kWh/year Levelized cost 0.0431 €/kWh And in the following Figure 7.32 we can see the Jinko power output Figure 7.32. PV power output per day and hour. In the Table 7.19 we can find the general information about the functioning of the 41 Trojan SAGM batteries of this microgrid. 52 Table 7.19. Functioning data of the batteries. String size 1 battery Strings in parallel 41 strings Bus voltage 6 V Energy in 2735 kWh/year Energy out 2388 kWh/year Storage depletion 69,4 kWh/year Losses 416 kWh/year Annual throughput 2591 kWh/year Autonomy 172 hours Storage wear cost 0,244 €/kWh Nominal capacity 101 kWh Usable nominal capacity 80,5 kWh Lifetime throughput 87580 kWh Expected life 33,8 years We can see the state of charge of the batteries in Figure 7.33. Figure 7.33. State of charge of the batteries per day and hour. As is noticeable in the graph, the batteries are almost fully charged in the central hours of the day, except for those more yellow lines that correspond to cloudy days, and the winter months, where the batteries are almost discharged, and when is more suitable to have the unmet electricity load that we commented before. The inverter in this microgrid also does the function of a rectifier and load-flow controller, as it has to distribute the energy sometimes from the solar panels to the batteries, other times from the batteries to the load and others from the solar panels directly to the load. In the following Table 7.20 we can see the important features of the Leonics inverter functioning in this microgrid. 53 Table 7.20.: Functioning of the inverter. Rectifier capacity 2.40 kW Inverter capacity 3 kW Mean output 0.469 kW Maximum output 2.39 kW Capacity factor 15.6 % Hours of operation 8750 hours Energy out 4,111 kWh/year Energy in 4,282 kWh/year Losses 171 kWh/year In Figure 7.34 we can see how the inverter works more in the evening hours, when it has to deliver the power from the batteries to the load because the solar panels don't have eneough sun to work. Figure 7.34. Inverter output per day and hour. In these off-grid microgrids, as we use only renewable energy sources, the emissions are zero. 7.5. Islanded microgrid with middle priced equipment Now we are going to use the CanadianSolar PV panels and the Hoppecke batteries. Here in Figure 7.35 is the scheme of the microgrid. Figure 7.35. Scheme of the microgrid. 54 The summary of the results of the optimization with HOMER are shown below in Table 7.21 As the Hoppecke batteries are better than the Trojan ones, we will need less of them, but they are also more expensive. Table 7.21. Summary of the simulation. LCOE 0,774 € NPC 41162 € Operating cost 943,85 € Number of batteries 16 Initial capital 28961 € Capital cost of the PV panels 3161 € Production of the PV panels 6473 kWh Inverter mean output 0,469 kW Inverter capital cost 1800 € Autonomy of the batteries 171 hours Annual throughput of the batteries 2577 kWh The cost summary is shown in Figure 7.36 and it’s also detailed in Table 7.22. Figure 7.36. Cost summary. Table 7.22. Detailed cost summary of the equipment. Component Capital Replacement O&M Salvage Total PV system 3,160.60 € 0 € 817.17 € 0 € 3,977.77 € Inverter 1,800 € 1,590.19 € 465.39 € -215.60 € 3,639.98 € Batteries 24,000 € 7,651.38 € 6,205.21 € -4,312.04 € 33,544.54 € System 28,960.60 € 9,241.56 € 7,487.77 € -4,527.64 € 41,162.29 € 55 As we see clearly in the cash flow in Figure 7.37 the main difference between this microgrid and the previous one is that the Hoppecke batteries need its replacement before. Figure 7.37 Cash flow. About the electric performance, in Figure 7.38 we see the PV panels generation. In this case, we have an unmet electric load of 0.855 kWh/year, which represents the 0.0208 % of the total, less than the previous microgrid with the cheapest equipment. We also have less capacity shortage, 1.05 kWh/year, the 0.0255 % of the total. So we can say that with this equipment the electrical performance of the microgrid has improved. In Table 7.23 we can see the basic information about the solar panels in this microgrid. Table 7.23. CanadianSolar PV panels. Mean output 0.739 kW Mean output 17.7 kWh/day Capacity factor 14.8 % Hours of operation 4,379 hours/year Total production 6,473 kWh/year Levelized cost 0.0475 €/kWh Figure 7.38. Monthly average electric production of the PV panels. 56 In the next graph, in Figure 7.39 we can see the power output from the CanadianSolar PV panels. Figure 7.39. Power output of the solar panels per day and hour. The functioning data of the Hoppecke batteries is shown in Table 7.24. Table 7.24. Functioning data of the batteries. String size 1 battery Strings in parallel 16 strings Bus voltage 2 V Energy in 2,707 kWh/year Energy out 2,390 kWh/year Storage depletion 66.6 kWh/year Losses 384 kWh/year Annual throughput 2,577 kWh/year Autonomy 171 hours Storage wear cost 0.160 €/kWh Nominal capacity 114 kWh Usable nominal capacity 80.1 kWh Lifetime throughput 51,546 kWh Expected life 20 years The state of charge of the batteries is presented in Figure 7.40. 57 Figure 7.40. State of charge per day and hour. The inverter functioning in the microgrid is detailed in the Table 7.25 and we can see also the inverter output in Figure 7.41. Table 7.25. Functioning of the inverter. Rectifier capacity 2.40 kW Inverter capacity 3 kW Mean output 0.469 kW Maximum output 2.39 kW Capacity factor 15.6 % Hours of operation 8,759 hours Energy out 4,113 kWh/year Energy in 4,284 kWh/year Losses 171 kWh/year Figure 7.41. Inverter output per day and hour. 7.6. Islanded microgrid with most expensive equipment In this last simulation, we will use the Sharp solar panels and the Tesla batteries as we see in the scheme of the Figure 7.42. 64 References [1] P.A. Narbel, J.P. Hansen, J.R. Lien; Energy Technologies and Economics, Springer, Norway, 2014. [2] R. Strzelecki, G. Benysek; Power Electronics in Smart Electrical Energy Networks, Springer, Poland, 2008. [3] D.T. Ton, M.A. Smith; The U.S. Department of Energy’s Microgrid Initiative, The Electricity Journal, Elsevier Inc. 2012 [online], available on: https://www.energy.gov/sites/prod/files/2016/06/f32/The%20US%20Department%20of%20Ener gy%27s%20Microgrid%20Initiative.pdf [accessed in 14 April 2018] [4] Classes from the subject “Control de Sistemes Energètics”, Robert Piqué, EEBE, UPC, 2017 [5] M. Kaltschmitt, W. Streicher, A. Wiese; Renewable Energy Technology, Economics and Environment, Springer, Germany, 2007. [6] 2007SolarGIS, Global Horizontal Irradiation [online], available on: http://www.mappery.com/Solar-Radiation-Map-of-Croatia [accessed on 21 April 2018] [7] Classes from the subject “Energies Renovables”, Herminio Martínez, EEBE, UPC, [8] 2017NREL, Best research-cell efficiencies [online], available on: https://www.nrel.gov/pv/assets/images/efficiency-chart.png [accessed on 23 April 2018] [9] HOMER Pro, Components library [online], available on: https://www.homerenergy.com/products/pro/docs/3.11/components_library.html [accessed on 30 April 2018] [10] Jinko JKM 275-60 datasheet [online], available on https://jinkosolar.com/ [accessed on 2 May 2018] [11] CanadianSolar MaxPower CS6U-330p datasheet [online], https://www.canadiansolar.com/ [accessed on 2 May 2018] [12] Sharp ND-250QCS datasheet [online], available on http://www.sharpusa.com/ [accessed on 2 May 2018] [13] Trojan Solar SAGM 06 375 datasheet [online], available on https://www.trojanbattery.com/solar-agm-2/ [accessed on 2 June 2018] 65 [14] Hoppecke 24 OPzS 3000 datasheet, available on https://www.hoppecke.com/ [accessed in 2 in June 2018] [15] Tesla Powerwall 2.0 datasheet, available on https://www.tesla.com/powerwall [accessed in 2 June 2018] [16] Leonics STP-219Cp 15 Kw datasheet, available on http://www.leonics.com/product/renewable/inverter/dl/STP-210p-194.pdf [accessed in 2 June 2018] 66 Abstract: In this paper, we have carried out the electrical performance and the cost-benefit analysis of a domestic microgrid based on renewable energy sources. The software used in this work in order to perform the simulations is Homer Pro, which enabled modelling of microgrid and all its parameters in a period of one year, accounting the real weather conditions and usage hours of the electric loads of the building we are working with. The microgrid consists of polycrystalline photovoltaic panels, a set of batteries, the utility (depending on the case study), and the load, which represents a prototype house and an inverter and controller that allow us to manage the energy flows between all the elements of the microgrid. Two case studies were conducted; in first, our microgrid was isolated from the electrical network and in the other, we counted on its support. Keywords: microgrid, PV, batteries, cost-benefit analysis, electrical performance analysis. 67 Curriculum Vitae PERSONAL INFORMATION Francesc Garcia Ferrando C/ Puig den Galileu, 9, 07193 Bunyola (Spain) (+34) 616 39 50 48 [email protected] Sex Male | Date of birth 10/10/1994 | Nationality Spanish EDUCATION AND TRAINING 11/07/2012–Present Degree in Energy Engineering Universitat Politècnica de Catalunya, EUETIB, Barcelona (Spain) 01/09/2010–30/06/2012 Baccalaureate IES Ramon Llull, Palma de Mallorca (Spain) 01/09/2006–30/06/2010 Secondary Education Col·legi Sagrat Cor, Palma de Mallorca (Spain) 01/09/1997–30/06/2006 Primary Education Col·legi Sagrat Cor, Palma de Mallorca (Spain) WORK EXPERIENCE Agricultural worker PERSONAL SKILLS Mother tongue(s) Catalan/Valencian, Spanish Foreign language(s) UNDERSTANDING SPEAKING WRITING Listening Reading Spoken interaction Spoken production English C1 C1 B2 B2 C1 Cambridge English: First Certificate Levels: A1 and A2: Basic user - B1 and B2: Independent user - C1 and C2: Proficient user Common European Framework of Reference for Languages Communication skills Good communication skills earned through years of group projects in university and high school as well as the experiences of living in a student's residence in Barcelona (CMU Ramon Llull) and Erasmus+ in a foreign country (Osijek, Croatia) Digital skills SELF-ASSESSMENT 68 Information processing Communication Content creation Safety Problem solving Proficient user Independent user Independent user Independent user Independent user Digital skills - Self-assessment grid Driving licence B