Design and optimization of the energy supply for the Global Interactive Village Environment: Techno-economic feasibility of an off grid solution for electrification in India

(1)

Examensarbete 30 hp Juli 2017

Design and optimization of the

energy supply for the Global Interactive Village Environment

Techno-economic feasibility of an off grid solution for electrification in India

Marco Frigeni

Masterprogrammet i energiteknik

Master Programme in Energy Technology

(2)

Teknisk- naturvetenskaplig fakultet UTH-enheten

Besöksadress:

Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0 Postadress:

Box 536 751 21 Uppsala Telefon:

018 – 471 30 03 Telefax:

018 – 471 30 00 Hemsida:

http://www.teknat.uu.se/student

Design and optimization of the energy supply for the Global Interactive Village Environment

Marco Frigeni

In a energy scenario moving fast towards the deployment of renewable energy technologies and the need of reducing CO2 emissions, hybrid energy systems for rural electrification are a feasible alternative solution to the utilization of conventional Diesel generators. The project focuses on the design and optimization of an off-grid hybrid energy system for a village of around 250 inhabitants in Gujarat, India. The energy system is part of a bigger project, “G.I.V.E Center of Excellence”, which has an innovative concept on a more sustainable rural lifestyle. The system, which has to depend mainly on locally available resources, intends to serve three main services: electrical demand, water purification and thermal energy for cooking. Two system configurations were designed and optimized to supply the estimated demand. The main outcome is a techno-economic analysis of the different system performances, which leads to a conclusion: dealing with the services individually has lower costs of implementation, less than half if compared to the implementation of a conventional Diesel generator. Furthermore, CO2 emissions are drastically reduced.

A sensitivity analysis was performed to address the different uncertainties such as the cost of the fuel. The result shows that if enough biomass resource would be available, a system based only on renewable energy technologies is economically profitable.

MSc ET 17003

Examinator: Joakim Widén, Associate professor at Uppsala Universitet Ämnesgranskare: Rasmus Luthander, Phd Candidate at Uppsala Universitet Handledare: Björn Laumert, Associate professor at KTH

(3)

KTH - ENTECH InnoEnergy MSc - Uppsala University

Acknowledgements

I wish my gratitude to the G.I.V.E. group behind the project for the interesting opportunity offered me and for the inspiring approach to life I became aware of.

I would like to express my gratitude to professor Björn Laumert, supervisor of my work, for keeping me on the good track during the entire development of the project and being always open for discussion and suggestions.

I’ also tha kful to Ras us Lutha de , su je t eade of thesis, ho as o sta tl the e to provide important feedbacks during the work, especially in the final part.

I would like to express my gratitude to Joakim Widén, program director, for the patience and support shown in this last year, when the stress to find an interesting thesis was approaching.

I feel the need to express my indebtedness to all my dearest friends at home for the vicinity shown in every possibility in these two years and the incredible moments spent together although the distance.

I want also to express my happiness for these two amazing years, in which I had the opportunity to meet inspiring and lovely people from all over the globe, who became sincere friends and with the whom I strongly believe we are going to share great experiences again together.

I’ tha kful to fa il , fo all the support and love expressed in every way possible during the entire path of my life and in particular in my studies, for the way they thought me to see the world and for the unconditional trust always given to me.

Last, I want to express my greatest gratitude to my girlfriend Rebecca, just met before leaving for this journey and now essential part of my life, for the extremely intense love experienced in every moment and place in these two years. Realizing that dreaming and living together, as if we always knew each other, give a wonderful taste to life.

(4)

KTH - ENTECH InnoEnergy MSc - Uppsala University 2

Nomenclature

Abbreviations

AC Alternating current AD Anaerobic Digestion CAPEX Capital Expenditure

DC Direct current

DER Distributed Energy Resources

GIVE Global Interactive Village Environment

HOMER Hybrid Optimization Model for Electric Renewable

HP Horse Power

IEA International Energy Agency

INR Indian Rupee

IRENA International Renewable Energy Agency LoadProGen Load Profile Generator

MATLAB Matrix Laboratory

MPPT Maximum Power Point Tracker MSW Municipal Solid Waste

NPC Net Present Cost

NREL National Renewable Energy Laboratory OPEX Operational Expenditure

O&M Operation & Maintenance

PV Photovoltaic

RES Renewable energy sources USD United States Dollar

WWTP Waste Water Treatment Plant

(7)

Simbols

C_ann,tot Annualized total cost [$/y]

CRF Capital Recovery Factor [-]

DM Dry Mass content in the substrates for anaerobic digestion [kg _{dry mass}/kg _{sub. input}] DOD Depth Of Discharge of a battery pack, in [%] of the total charge

E_prim,AC Total AC load served in a year by the hybrid system [kWh/y]

E_defer Total deferrable load served in a year by the hybrid system [kWh/y]

HRT Hydraulic Retention Time [days]

i Annual real interest rate [%]

LCOE Levelized Cost Of Energy (often just Electricity) [$/kWh]

N Number of years [-]

OLR Organic Load Ration for a biogas digester [kg_VDM/m³day]

R_h Uncertainty in [%] of the total functioning time of an appliance R_w Uncertainty in [%] of the operation windows of an appliance

R_prog Project lifetime [years]

SOC Battery State Of Charge in [%] of the maximum sub. Input Substrate input in the biogas digester [kg/day]

VDM Volatile Dry Mass content in the substrates for anaerobic digestion [kg volatile dry mass / kg _{dry mass}]

(8)

List of figures

Figure 1. (a) Map and (b) satellite view of Anand district (Google maps) ... 11

Figure 2. Water flow chart of the village under investigation ... 13

Figure 3. Elements of a generic hybrid energy system ... 14

Figure 4. Example of a PV field installation ... 14

Figure 5. Example of installation of a 10 kW Hummer wind turbine ... 15

Figure 6. Typical components of a battery storage system [9] ... 16

Figure 7. Series configuration of an hybrid energy system [18] ... 17

Figure 8. Switched configuration of an hybrid energy systems [18] ... 17

Figure 9. Parallel system configuration of an hybrid energy system, (a) AC coupling and (b) DC decoupling [18] ... 18

Figure 10. Example of installation of a household floating drum digester ... 19

Figure 11. Biogas end uses [19] ... 20

Figure 12. HOMER software - inputs and outputs ... 25

Figure 13. Scheme of a floating drum digester ... 26

Figure 14. Monthly Average daily irradiation Incident On A Horizontal Surface for the location31 Figure 15. Monthly average clearness index for the location ... 31

Figure 16. Monthly average wind speed data for the location at 10 m hub height ... 32

Figure 17. Typical electric load profile - week day - cooking demand not included ... 38

Figure 18. Typical electric load profile - weekend - cooking demand not included ... 39

Figure 19. Typical electric load profile - week day - cooking demand included ... 40

Figure 20. Typical electric load profile - weekend - cooking demand included ... 40

Figure 21. Electrical system architecture - configuration I... 42

Figure 22. Monthly average electric production - configuration I ... 43

Figure 23. System architecture - configuration II ... 43

Figure 24. Monthly average electric production - configuration II ... 44

Figure 5. Ele tri po er e olutio for o e day ith good at ospheri o ditio s - configuration I ... 45

Figure 26. Batteries SOC evolution for a day - April 2^nd - configuration I... 46

Figure 27. Yearly evolution of battery SOC - configuration I ... 46

Figure . Ele tri po er e olutio for o e day ith ad at ospheri o ditio s - configuration I ... 47

Figure . Ele tri po er e olutio for o e day ith good at ospheri o ditio s - configuration II ... 48

Figure 0. Ele tri po er e olutio for o e day ith ad at ospheri o ditio s - configuration II ... 48

Figure 31. Yearly evolution of battery SOC - configuration II ... 49

Figure 32. Cash flow summary by component and cost type - electrical system in configuration I ... 50

Figure 33. Total cost shares by component - configuration I ... 51

(9)

Figure 34. Total cost shares by component - configuration II ... 52

Figure 35. Cash flow summary by component and cost type - electrical system in configuration II ... 52

Figure 36. Optimal system type on varying of biomass resource quantity and diesel fuel price - configuration II ... 53

Figure 37. Total net present costs on varying of biomass resource quantity and diesel fuel price - configuration II ... 54

Figure 38. G.I.V.E Village site map ... 64

Figure 39. G.I.V.E. Village site map highlighting the position and parts of the water system .... 65

List of tables

Table 1. Expected total number of people and dwellings relative to the 4 different categories of inhabitants ... 11

Table 2. Type and number of facilities in the village ... 12

Table 3. Example of user class data input in LoadProGen ... 23

Table 4. Summary of water system electrical requirements, correspondent power and energy demand ... 30

Table 5: Biogas characteristics and yields for cow manure and organic fraction of MSW... 32

Table 6. PV plant - CAPEX breakdown and total ... 34

Table 7. PV plant - OPEX breakdown and total ... 34

Table 8. Cost components for biogas digester systems ... 36

Table 9. Mass of volatile total solids daily introduced in the digester ... 41

Table 10 Individual and total gas yield for different substrates ... 41

Table 11. Optimal electrical system seen from a NPC perspective - configuration I ... 42

Table 12. Optimal electrical system seen from a NPC perspective - configuration II ... 44

Table 13. Electrical system performances - comparison for the two configurations ... 44

Table 14. Capital cost components - 85 m³biogas digester ... 49

Table 15. Net present costs summary - electrical system configuration I ... 50

Table 16. Total cost resume - configuration I ... 51

Table 17. Net present cost summary - configuration II ... 52

Table 18. Costs and performances comparison with a conventional diesel generator ... 55

Table 19. Domestic load user classes input parameters in LoadProGen for the typical week day ... 66

Table 20. Community and infrastructure + water system user classes input parameters in ... 67

Table 21. Domestic load user classes input parameters in LoadProGen for the typical week end ... 68

Table 22. Community and infrastructure + water system user classes input parameters in ... 69

(10)

1. Introduction

According to the International Energy Agency (IEA), an estimated number of 1.2 billion people lack access to electricity, predominantly in rural areas of developing countries in Asia and sub- Saharan Africa [1]. In India alone, around 300 million of people are facing that issue [2]. Access to energy is a precondition for economic and social development, highlighting the need to eliminate energy poverty [3] [4]. Until now energy supply and relative development was obtained mostly through the deployment of fossil fuels as primary resources, in fact more than 81% of the total primary energy supply in 2014 (latest value) was covered by conventional fossil fuels [5]. The importance of achieving energy security, independency and reduction of CO₂ e issio s elated to the o ustio of the , o je ti e st e gthe ed the Pa is ag ee e t i , has pushed the focus of research, industry and policies into the direction of diverse sustainable and economically viable ways of generating electricity and energy. In fact, not only renewable energies such as wind and solar are increasing significantly their shares in the electricity production every year, but they also reached the fastest growth in the investments in 2015, faster than conventional fossil fuel technologies [6].

In the past, the development in the electricity and energy generation sector was attained by the construction of massive power plants (mostly based on fossil fuels) and long transmission networks capable to bring electricity around a country or region. During the last years, renewable energy technologies have changed this paradigm. The latter, together with the influence from liberalism principles and technological development, contributed to a gradual alteration of the energy policies and strategies towards the concept of distributed power generation [7]. Solar and wind energy have been and are indeed exploited through a staggering number of small, locally spread, installations, as well as through the construction of big power generation plants. However, because of their intermittent nature, they contributed to create great challenges in the power generation sector, related especially to the grid frequency management.

Inside this framework, microgrids composed by hybrid energy systems became a complementary and alternative solution to the expansion of the main grid or the implementation of conventional Diesel generators. As defi ed the Mi og id I stitute, a microgrid is a small energy system capable of balancing captive supply and demand resources to ai tai sta le se i e ithi a defi ed ou da [8]. An hybrid energy system relies on different distributed energy resources and components, which combined form a better system than its singular part. Renewable energy sources, generators and energy storage systems combined are capable to overcome the issue related to the intermittent nature of renewable resources and supply reliable energy. For rural development, but not only, they are considered a solution more convenient and sustainable than just conventional diesel generators, by achieving higher system efficiencies, lower costs and emissions thanks to the abundant use of renewable energy resources [9].

(11)

KTH - ENTECH InnoEnergy MSc - Uppsala University 9 The scope of the master thesis work is to investigate, design and compare different off-grid hybrid energy systems for a rural village in Gujarat, India. The energy system under investigation, part of the project Global Interactive Village Environment, aims at being as independent and sustainable as possible. Due to the very favourable natural conditions and the locally available resources, the scope of achieving a self-sustainable energy supply is a possible dream to be investigated.

1.1 Goals and objectives

The different systems aim at supplying reliable electrical power, as well as purify water and cover the cooking demand. Moreover, of outmost importance is the usage of the available natural resources as well as the waste flow of materials inside the village.

More detailed objectives to be accomplished are here introduced.

I. To explore the literature and understand the state of the matter, as well as the technological solutions implemented in similar cases;

II. To investigate and define the most realistic load profile for the village studied, relying on the proje t o e ’s e pe tatio a d the u e t situatio i I dia;

III. To evaluate the natural resources available in the location, as well as all the end use resources available inside the village, in order to exploit most of the renewable resources that are locally available;

IV. To design, optimize and simulate the behavior of the different systems;

V. To compare the results from a techno-economic point of view, dealing with uncertainties o the odel’s h pothesis when relevant;

VI. To estimate emissions and social benefits related to the project;

The main goal would be to develop the new sustainable, economically feasible and scalable technological solution for rural development and electrification in an Indian village, but also to create a methodology suitable for other case of studies.

(12)

2. Background

2.1 Project background: G.I.V.E. Centre of Excellence

2.1.1 Vision

The Scandinavian foundation behind the project is formed by a growing group of Indians and Scandinavians with a common vision: build new, and transform existing villages for a sustainable lifestyle as a competitive alternative to cities and megacities, in full balance with nature. The first step to reach the ambitious objective is set by the establishment of a small village in India, named Global Interactive Village Environment (G.I.V.E.) Center of Excellence, where sustainable technologies for village development are adapted, applied and showcased, in order to serve as a odel fo a sustai a le illage so iet .

The project specifically revolves around three specific goals:

1. Create a center of excellence for education, development and sharing of knowledge;

2. Develop a business hub for Scandinavian clean tech, facilitating sharing of knowhow about products and markets with Indian actors;

3. Determine the marriage between the traditional and technological excellence in sustainable lifestyle infrastructure with embedded technologies for climate and economic resilience;

Inside this framework, the thesis work takes place, designing and outlining realistically the most self sustainable and feasible energy system for the center of excellence.

2.1.2 Location

Characterized by a population of 63 million, a density of roughly 310 inhabitants per km² and the longest coastal length among the Indian states, Gujarat State is classified for more than 70% of the country as a semi arid to arid climatic area. Essentially two defined seasons can be distinguished in the area, summer (four months from June to September) which is the moonsoonal season, and the rest of the year, dominated by severe dry weather. Significant d ops i the te pe atu e do ’t o u , hi h o o thl a e ages is al a s a o e °C during the entire year [10].

Even if the precise location has not yet been decided, as mentioned in section 1, the area of i te est fo the illage’s o st u tio is situated in the rural part of Anand district, between the cities of Anand and Borsad [22.46 °N, 72.93 °E], as presented in figure 1.

(13)

KTH - ENTECH InnoEnergy MSc - Uppsala University 11 (a) (b)

2.1.3 Village boundary conditions

Expectations regarding the village design and features are reported to give an understanding of the framework. The number of people expected to live in the village is estimated to be around 200 - 250. The Inhabitants composition, because of the particular functions intended to take place inside the village, ai l elated to edu atio a d fa i g, a ’t e dedu ed f o the average village existing now in Gujarat. Therefore, after discussion and planning, the most

edi le g oup’s e pe tatio is reported in table 1.

Table 1. Expected total number of people and dwellings relative to the 4 different categories of inhabitants

Category % Among the total inhabitants

Average number of family components

Total expected number

Equivalent number of dwellings

Farmers 40 5-7 100-120 18

Skilled workers

and academics 40 4-6 100-120 20

Students 10 Alone 18 6

(3 per dwelling)

Foreign Businessmen and

academics

10 Alone 16 8

(2 per dwelling) The illage’s desig has been discussed in the early stages of the project and the site map is presented in Appendix I. With the objective to enhance the human dimension of living, houses and the other buildings will be constructed with a particular dome shape, capable to harvest water on the rooftop and to reduce as much as possible the need for cooling (achieved trough natural ventilation and other natural techniques) [11]. This is an important assumption that

Figure 1. (a) Map and (b) satellite view of Anand district

(14)

KTH - ENTECH InnoEnergy MSc - Uppsala University 12 should be verified in the implementation phase and could shape significantly the future energy demand. Moreover in order to better optimize sewage water collection and relative purification of it, common toilets will be implemented. Also the village common infrastructure, as listed below in table 2, vary significantly compared to the usual case in developing countries.

Table 2. Type and number of facilities in the village

Type of facility Number of facilities Common toilets 5 Administration 1

Kiosk 1

Nursery 1

Primary school 1

Computer lab 1

Workshop 1

Art exhibition 1

Restaurant 1

The water system design plays a fundamental part in achieving a sustainable and environmental friendly lifestyle. The idea is to harvest rain water in a central reservoir, built with height difference in order to permit the water to reach the centre only by gravity force.

Once harvested, the water will be purified through sand filters before serving any request, and after utilisation, the gray water coming from domestic usages will be purified again through sand filters, before reaching the reservoir another time. The purified water is intended to be pumped up to a water tower through the help of a pumping station when convenient during the day, and then be redistributed only by gravity to the consumers. Anyway water has to pass a second time through the sand filters before reaching the households, in order to assure the complete clean up of bacteria or any other biological life self generated in the water reservoir.

Slow sand filtration is a natural method which takes more time than distillation membranes, common technologies adopted in poly-generation projects [12], but i deed it does ’t e ui e any additional electrical energy [13].

The water used in toilets (addressed as black water) will reach the waste water treatment plant through the help of several pumping stations (one at each common toilet). There, it will be treated and after sludge removal achieved with a drying process, the remaining part will be disposed in a water reservoir to serve irrigation for agriculture.

In order to visualized what just described, the water system design is added in Appendix I, while the water flow chart is represented in figure 2.

(15)

KTH - ENTECH InnoEnergy MSc - Uppsala University 13 Figure 2. Water flow chart of the village under investigation

As above mentioned, also farming activities are meant to be conducted inside the village, agriculture is projected to be done through permaculture and aquaponics techniques, innovative and much less water intensive. Breeding will be as well an important activity and resource. According to the livestock census [14], and witnessed by the presence of the biggest Indian cooperative for dairy production (AMUL cooperative), Gujarat is characterized by intense dairy production and a high number of cattles. Based on these facts and o the g oup’s experience and knowledge about Indian conditions of living, an estimate of one cow per person for a village of this size.

2.2 Literary review

In this section, the different technological solutions adopted in the project are briefly introduced and described in their main features and components. In conclusion, a brief literature survey presents several projects of main relevance for similarity in the topic, context and methodology.

2.2.1 Micro grids and hybrid energy systems

Micro grids, a terminology that refers to all the systems composed by a hybrid energy system and electric grid, are further divided in four main categories: off-grid, campus, community and nano grid [8]. An off grid system, which is the focus of the project, operates in a island mode and therefore requires to be self-sufficient, since there is no connection to the main electrical grid which could operate as an energy reserve. This could be seen as both disadvantage and advantage. In fact, maintaining a power balance to ensure stable operation results in a more complex system than a system that is grid tied. On the other side, the system is totally independent from any kind of power faults occurring in the main grid.

Hybrid energy systems combine distributed energy resources (DER), like renewable energy technologies and generators, with storage components, like the battery system (illustrated in figure 3). Power and energy management of the system is achieved through the

(16)

KTH - ENTECH InnoEnergy MSc - Uppsala University 14 implementation of a regulation and conversion unit, and according to different strategies of operation.

Figure 3. Elements of a generic hybrid energy system

The main components, the two different case studies and the dispatch strategies are described in the sub sections of this chapter.

2.2.1.1 Renewable energy technologies

Wind and solar energy technologies achieved an incredible development and growth in the recent years. Thanks to the favorable policies, economic conditions and technological development, they reached a high share in the electrcity sector, becoming also from an economic perspective, competitive alternatives to fossil fuel plants [15]. In hybrid energy systems, solar photovoltaic and wind generators are the two most deployted renewable electricity generation technologies in terms of modularity, flexibiity in design and low costs.

Solar photovoltaic cells are p-n junctions capable to convert solar energy into electricity through the photovoltaic effect. When a photon reaches the PV cell with an energy greater than the band gap of the semiconductor, an electron is released and removed through the aid of the p-n junction of the material. The electron is then free to flow as current due to the electric field created between the n-type and p-type semiconductors [16]. Several cells are assembled together to manufacture a photovoltaic module, which is then connected in series and/or parallels to others to form an array and eventually a photovoltaic field, as depicted in figure 4.

Figure 4. Example of a PV field installation

(17)

KTH - ENTECH InnoEnergy MSc - Uppsala University 15 In order to work at the proper conditions of current and voltage, PV arrays are combined with the power unit (usually addressed as inverter). In grid connected systems the inverter covers both the function of MPPT (which ensure the optimal operational conditions of the PV modules) and inverter, transforming the DC current generated by the modules into AC to be injected in the grid. In off grid systems, depending on the configuration, different solution can be adopted. Often, a solar charge controller does the function of MPPT, and then the bi- directional inverter (which is capable to convert the power between AC and DC in both directions) does the conversion and regulation of the system electrical production, including PV electricity.

Wind turbines are capable to convert the kinetic energy of the wind into mechanical energy.

Due to the high velocity of the wind and the aerodynamics of the turbine blades, the kinetic energy is transformed into rotational energy which is converted into electric energy through the electrical generator placed in the nacelle. Wind turbines are available in different order of sizes, from W to MW and types, mainly vertical or horizontal axis. In hybrid energy systems for communities or villages, horizontal axis generators with a nominal power of about 10 kW, as the one represented in figure 5, are more commonly selected.

Figure 5. Example of installation of a 10 kW Hummer wind turbine

2.2.1.2 Energy storage system

Because of the decreasing trend in battery prices and the technological development which has occured in the past years, batteries are now considered among the most appealing solutions to serve different energy storage purposes, such as short term regulation, peak shaving, energy supply shift and island off grid operations [9]. In off grid applications, batteries increase renewable penetration while enhancing the reliability of the system. The concept behind the adoption of batteries in off grid hybrid energy systems is simple, to store energy when excess electricity is convenient to be generated, as according to the selected dispatch strategy, and release it in other times, instead of turning on for example the diesel generator.

Ho e e atte s ste s alo e a e ’t apa le of providing a reliable back up system for the renewable sources in a long time. Because of the limited energy intensity, batteries still a e ’t convenient and feasible to achieve long term energy storage (like seasonal). Furthermore, the lifetime of battery packs is highly influenced by the number of charge-recharge cycles fulfilled.

(18)

KTH - ENTECH InnoEnergy MSc - Uppsala University 16 In order to enhance the cycle lifetime, complementary generators are sometimes preferred for power back up and regulation.

Anyway, the battery system is not only composed by the storing units but also a set of different parts as depicted in figure 6.

Figure 6. Typical components of a battery storage system [9]

2.2.1.3 Dispatch strategies

As defined in [17], Dispatch refers to the aspect of control strategy that pertains to energy flo s a o g the ajo o po e ts . In off grid applications comprising both batteries and generators, dispatch strategies aim especially at the prolongation of the equipment lifetime by controlling the operation mode of the generator, which is the most flexible part of the system.

Four defined dispatch strategies can be adopted, according to [17], but only the two most interesting strategies for the project are described: load following strategy and cycle charging strategy (or often addressed as state of charge set point strategy).

In cycle charging dispatch strategy, whenever a generator needs to operate to serve the primary load, it operates at full output power. Furthermore, if a set point state of charge of the battery bank is applied, then the generator will likely serve the load and produce excess electricity to charge the battery bank until reaching the before mentioned set point, if

e e a le ele t i it p odu tio a ’t a age itself.

In load following strategy instead, whenever a generator operates, it produces only enough power to meet the primary load. The sources are generally dispatched in a way that the load is served by the combination of components with the least cost.

2.2.1.4 Types of configuration

Depending on the application, the ways in which the different parts of a hybrid energy system are connected can be multiple. The three main configurations are [18]: series, switched and parallel hybrid energy systems. In the first of the mentioned configurations (depicted in figure 7), all the components are connected through a DC bus in order to recharge the battery bank and serve the AC load only after the inverter has converted the power into AC again. While the RES are equipped with an individual charge controller, which regulates the power production flowing to the batteries, the generator is equipped with a rectifier acting also as a charge controller. This leads to significant conversion losses since the electrical energy produced by

(19)

KTH - ENTECH InnoEnergy MSc - Uppsala University 17 the generator is first converted into DC and then again into AC to serve the load. Furthermore in this configuration both the inverter and diesel generator have to be sized in order to meet the peak demand, which could rise the total system prices significantly.

Figure 7. Series configuration of an hybrid energy system [18]

The switched system configuration (illustrated in figure 8) does ’t pe it contemporary operation of the inverter and the generator. Thus, they are both designed to meet the peak load as in the series configuration. Additionally, when switching the AC source, the power to the load is momentarily interrupted. However, the conversion efficiency reaches higher values since the load can be directly served by the generator without passing through any conversion unit.

Figure 8. Switched configuration of an hybrid energy systems [18]

Parallel system configuration is further classified in AC and DC couplings (as shown in figure 9).

The bi-directional inverter linking the batteries with the AC source, can act as rectifier to recharge the batteries when excess electricity is available or act as inverter to discharge the batteries and serve the load. Even if this configuration is more complex and might require higher costs in the design and operation phase, it presents two big advantages. By the implementation of only one component (bi-directional inverter) acting as rectifier and inverter, the number of system components and related wiring is minimized. Furthermore, the maximum peak to serve needs to be met by the sum of the inverter and generator power, instead of one individually.

(20)

KTH - ENTECH InnoEnergy MSc - Uppsala University 18 (a) (b)

Figure 9. Parallel system configuration of an hybrid energy system, (a) AC coupling and (b) DC decoupling [18]

2.2.2 Biogas

Citing [19], One of the main advantages of biogas production is the ability to transform waste material into a valuable resource, usi g it as su st ate fo a ae o i digestio . Espe iall i India where waste disposal is a big problem, biogas production can definitely be a part of the solution. Moreover, the manufacture and maintenance of the system can provide several job opportunities with additional benefits to the society. The flexibility of its usages, which can vary from direct combustion in gas stoves for cooking purposes to combustion in micro gas turbines, fuel cells, Otto and Diesel engines to produce electricity, makes it suitable for several applications.

2.2.2.1 Anaerobic digestion

Anaerobic digestion is a microbiological process of decomposition of organic matter in absence of oxygen. The initial material is continuously broken down into smaller pieces in a process composed by several linked steps, of which the main four are: hydrolysis, acidogenesis, acetogenesis and methanogenesis [19]. The two main products of the process are a combustible gas, mainly composed by methane and carbon dioxide, and a nutritious digestate, which is the residual part of the decomposed substrate. The gas, called biogas, can serve different usages, while the digestate has the potential to become a good soil fertilizer. The reactions take place in a air and water tight chamber defined as bio digester. Biogas yields of the process depend on several factors, mainly the temperature of the reaction and the C/N (carbon versus nitrogen) ratio of the substrates.

2.2.2.3 Types of substrate

In general, all types of biomass containing carbohydrates, proteins, fats, cellulose and hemicelluloses as main components can be adopted as substrates for AD. However when selecting, some guidelines need to be taken into consideration. For instance, the content of organic substance should be appropriate for the selected fermentation process, the nutritional value of the organic substance, and hence the potential for gas formation, should be as high as

(21)

KTH - ENTECH InnoEnergy MSc - Uppsala University 19 possible and finally the substrate should be free of pathogens and other organisms which would need to be made innocuous prior to the fermentation process [20]. According to [19], some of the most common and appropriate substrates are:

 Animal manure (cow, pig, poultry) and slurries

 Agricultural residues and by-products

 Digestible organic wastes from food and agro industries (vegetable and animal origin)

 Organic fraction of municipal waste and from catering (vegetable and animal origin)

 Sewage sludge in waste water treatment plants

 Dedicated energy crops (e.g. maize, miscanthus, sorghum, clover).

In some cases, also the co-digestion of different substrates is adopted to achieve higher biogas yields , as in the case studied in [21] and [22].

2.2.2.3 Digester technologies

The bio digester is the physical structure where the process of anaerobic digestion is happening. The main function of this structure is to provide anaerobic conditions within it, and as a chamber, it should be air and water tight. It can be made of various construction materials and in different shapes and sizes [23]. They are categorized according to the feedstock input and output in batch and continuous types. While the batch type digesters are fed with a portion of fresh substrate which after degradation is removed and substituted by fresh one, the continuous one are fed constantly and the pressure of the incoming substrate remove the digested one [19]. Among the continuous fed type digesters, the plug flow and complete mixed should be mentioned for their numerous applications. For a smaller scale plant design such as community or household, other categories emerge like the floating drum (depicted in figure 10) and fixed dome digester. The selection between the numerous systems is done depending on the substrates dry and volatile dry matter content.

Figure 10. Example of installation of a household floating drum digester

2.2.2.4 Utilisations

The numerous final end uses of biogas are described in the scheme represented in figure 11.

The simplest use of biogas, especially at household and community plant scales in developing countries, is the direct combustion in boilers or gas stoves [19]. Otherwise, the crude gas can be either upgraded to biofuel or go to the scrubbing process. If upgraded, it can be either injected in the public grid or used as vehicle fuel. When instead it undergoes the process of

(22)

KTH - ENTECH InnoEnergy MSc - Uppsala University 20 scrubbing, necessary to remove especially H₂S formed in the bioreaction, can serve generation purposes. In fact it is commonly utilised in CHP (combined heat and power) plants, or adapted to Diesel or Otto combustion engines to produce electricity and useful heat.

Figure 11. Biogas end uses [19]

2.3 Literature survey on similar projects

Several authors have already performed techno-economical feasibility analyses of hybrid energy systems in rural electrification projects.

In [24], a small hydro/PV/Wind hybrid system for off-grid rural electrification in Ethiopia is investigated and HOMER energy software is used for the optimization and sensitivity analysis about the system components costs. The relative cost of energy is found to be close to 0.16

$/kWh.

In [25], the techno-economic sizing of an off-grid hybrid renewable energy system for rural electrification in Sri Lanka has been performed. HOMER energy is again the software adopted for the optimization and the results obtained show that a system comprising of wind turbines, photovoltaic system, a battery bank and a diesel generator of 40 kW, 30 kW, 222 kWh, and 25 kW respectively, supplies electricity with a LCOE of approximately 0.3 $/kWh.

Hybrid application of biogas and solar resources to fulfil household energy needs have been studied in [26], where HOMER software was adopted to identify the optimal system architecture matching different estimated household energy needs.

Also in [27], a feasibility study of using a biogas engine as backup in a decentralized hybrid (PV/wind/battery) power generation system in Kenya has been investigated, bringing to the conclusion that biogas generator can replace diesel ones by lowering the levelized cost of energy.

Even though many authors have been choosing HOMER energy to perform techno-economic feasibility analyses of hybrid energy systems, in [28], by using the genetic algorithm, the optimal design of a PV-diesel hybrid system for electrification of an isolated island in Bangladesh was achieved.

(23)

KTH - ENTECH InnoEnergy MSc - Uppsala University 21 A very interesting approach regarding the techno-economic analysis of a small scale biogas based poly generation systems, using Bangladesh as case study has been performed in [29].

The results indicate that the poly-generation system is much more competitive and promising (in terms of LCOE) than other available technologies when attempting to solve the energy and arsenic related problems in Bangladesh.

Several other authors with different approaches have investigated the feasibility of systems composed by different energy sources for poly-generation or hybrid energy system applications [30] [31] [32] [33].

(24)

3. Methodology

The scope of this section is to describe the process and procedure followed for the accomplishment of the work. The definition of a credible daily load profile, essential to perform the optimization and to achieve reasonable results, was an important part of it. In order to obtain it, an accurate procedure implemented in a software tool named LoadProGen was used, developed by the Energy4Growing team, research group at Politecnico di Milano.

Contemporary, the assessment of the available resources was performed. In fact, depending on the assets and the energy requirements to be covered, two plausible energy system configurations capable to serve the needs were identified, studied and optimized from a techno-economic point of view through the help of HOMER software. In particular, once assumed the economic and technical parameters of the system components, obtained after a long and extensive market and literature research, the size of the electrical components of the energy system was optimized and the costs of it were aggregated with the ones of any other additional energy system (i.e. biogas system for cooking purposes), when part of the total system. Finally the performances and total costs of the different options were compared, and a brief sensitivity analysis was conducted for the most sensible parameters.

3.1 Case studies

Two system configurations serving the energy needs with two different energy means were considered worth to be analyzed.

In the first one, hybrid electrical system and a biogas digester system are designed independently to meet two different loads. Electrical load requirements (including water purification) are met by an autonomous electrical system, composed by PV, wind, diesel generator and a battery system, while thermal cooking demand is met by the biogas production by direct burn into gas cooking stoves.

First it was considered to analyze the implementation of a unique biogas plant, designed to work as a generator for the hybrid electrical system and as source of biogas to serve the thermal cooking demand. But in this case, as consequence of the estimated small amount of locally producible biogas, the capability of the system to serve both the loads is reduced and a ’t e a feasi le optio . Ho e e , it is possi le to agg egate the the al ooki g de a d into the electrical one, by utilizing electric stoves instead of gas ones. Even if unfavorable from an efficiency point of view and traditionally not adopted in India, it might prove to be a feasible and cheaper option. This is due to the economic advantage that might occur when the two needs are combined into one electrical load. As it will be shown, electricity production costs of the hybrid system are low thanks to the extremely low costs of PV systems, especially in India.

In this way a second more compact configuration, producing only one kind of energy is designed to cover the two different loads. PV, wind and battery system are kept, while the

(25)

KTH - ENTECH InnoEnergy MSc - Uppsala University 23 diesel generator and the biogas generator are compared with each other in terms of performances and costs.

3.2 Load profile generation

At this design phase of the project development, a numerical series defining the average hourly constant power load required by the village within a given time-step would be enough.

Anyway, data of that atu e a e ot a aila le fo illages i u al a eas, a d the also ould ’t be suitable to be used as reference for such a particular case as the one under investigation. In the literature, when it comes to define the load profile, it is noticeable a lack of a accurate procedure to follow. In fact several authors are introducing them without explanation about their origin [33] [34], others introduce a typical set of appliances with correspondent rated power, estimate the time of functioning and compute the total daily energy requirements [35]

[24]. Then the load profile is built in a way that fits the expectation of low request during the night, medium during the day and peak during the evening.

After an intensive literature research, a reliable and accurate novel mathematical procedure to formulate load profiles for off-grid rural areas was found in [36]. LoadProGen, software tool implementing the procedure (open source at [37]), was adopted for the case. The MATLAB algorithm, based on the new mathematical procedure, receives a given set of input data and formulates n load profiles requested by the user. On different levels, the input data (of which a sample is depicted in table 3) needs o be organized and introduced. Firstly, it is necessary to divide the type and number of consumers (nature of the electrical load) into user classes, associated by the common set of appliances and behaviour. For instance, households with a similar social status or buildings with the same function to provide are clustered together.

Subsequently, a set of typical appliances and the relative rated power corresponding to each of them is introduced. Also the total daily functioning time and the windows of operation (up to three) for each of the electrical devices are entered. Finally, the software tool, with a stochastic approach and equipped with correlations between the different load profile parameters (i.e.

load factor, coincidence factor and number of consumers) computes a load profile for each of the appliances separately, before aggregating all the individual contributions. In this way, it accounts for use s’ coincidence behaviour. More details about the mathematical procedure and software principle of operation are described at [36].

Table 3. Example of user class data input in LoadProGen

Specific user class

Numb er of users in class

Type of electrical appliance

Nominal power rate [W]

Number of appliance

s in class

functioni ng cycle [min]

functioni ng time [min]

Rh Rw

Number of functioning windows

Startin g time Win 1 [min]

Ending time Win 1 [min]

Farmers 18 TV 100 1 30 480 0 0 1 480 1440

Radio 5 1 30 120 0 0 2 360 480

Refrigerat

or 250 1 5 480 0 0 1 1 1440

Lighting 9 3 10 300 0 0 3 1 60

Phone

charger 5 2 30 180 0 0 2 1 360

(26)

KTH - ENTECH InnoEnergy MSc - Uppsala University 24 Even though in LoadProGen, it is possible to introduce uncertainties regarding both total functioning time and windows of operation for each of the appliances (R_h and R_w), uncertainty was added on the total load in the software adopted for the simulation.

Furthermore, the tool computes power values with a minute step resolution, more accurate than what is needed for this phase of the project and also not a possible input for the software used in the optimization of the systems. Thus hourly averages have been computed, even if the quality of the load profile was reduced and spikes in the power demand might have been neglected.

3.2.1 Main assumptions

According to the village boundary conditions (chapter 2.1.3), the main electrical usages have been caegorized as following:

1. Domestic

2. Commercial and common infrastructures 3. Water system

The decisions upon the user classes and relative appliances belonging to each of them, as well as all the specifications (such as rated power and windows of operations), relied on databases, values adopted in similar project and the work group's experience on the field. Everything aimed at achieving an energy profile as much realistic as possible, and all the assumptions passed through validation within the project group.

When computing the load profile, it is important to account for the variability due to the particular time of the year (season) and of the week (week end or week day).

The season might affect especially the time in which lights are switched on because of the different day lengths. But in Gujarat, situated in the vicinity of the equator, the day length does ’t a sig ifi a tl [38], thus no differences in lighting are accounted. Even if irrigation requirements differ from one season to the other, as it will be shown in chapter 4.1.3, they account for only 3 kWh/day (less than 1% of the daily total energy consumption) and therefore considered negligible.

If on one side seaso a ia ilit does ’t eall have a big impact on the usage pattern of appliances, the variation between weekends and weekdays does. In fact, most of the community facilities are closed during the weekends and also the residential usage pattern is shaped by more randomness and a higher consumption due to the free day schedule.

In conclusion, for each of the case studies the variability along the year was accounted with the computation of two different load profiles (one for the week days and one for the weekends).

3.3 Hybrid optimization model for electric renewable

Usually addressed as HOMER energy, the software developed by NREL is a very powerful tool capable to perform precise optimization and simulation of performances for hybrid energy systems. The methodology adopted by the software is simple and hereafter illustrated with the main inputs and outputs as in figure 12.

(27)

KTH - ENTECH InnoEnergy MSc - Uppsala University 25 Figure 12. HOMER software - inputs and outputs

On one side, six main categories of input data are required. Load profile, availability of energy resources, choice of the system architecture (including technical parameters), life cycle costs of the components, constraints for the system operation mode and economical parameters such as interest rate and lifetime of the project. Typical load profiles for each month of the year are entered with an hourly resolution as mentioned earlier in the report. Random variability between days and time steps can be accounted as percent of maximum deviation from the typical ones introduced (5% for both was set). The availability of primary resources refers to both meteorological data (i.e. wind speed and solar irradiation) and fuel quantities. In case of fossil fuel deployment, only if any restriction or limit about the yearly quantities consumed needs to be considered, instead when alternative sources such as biogas are considered, the daily expected quantity needs to be expressed. Furthermore, technical parameters such as efficiency curves are required to run simulation of the system. Essential to perform the optimization, interest rate and project lifetime were set at 8% and 25 years respectively. Also lifetime for each component and life cycle costs are entered in four different main cost components: initial capital, replacement, operation and maintenance and fuel. In the system constraints, a value regarding the allowable capacity shortage of the system needs to be established. Expressed as percentage of the total electrical load to be covered in a year, 1%

was assigned to the parameter. Additionally, sensitivity variables and emissions related to the fuel, if adopted any, can be introduced, if intention of the user. The software, according to all the introduced parameters and two different electrical control strategies (cycle charging and load following), performs a simulation of the electrical behavior along the time, varying the

(28)

KTH - ENTECH InnoEnergy MSc - Uppsala University 26 system components sizes and combinations, as chosen by the user. Different simulated systems, presented with relative technical performances and economic calculations, are ranked in order of increasing net present cost, computed as following:

(1) Where CRF is the Capital Recovery Factor, defined in equation (2) dependent on the interest rate (i) and the lifetime of the project (R_prog), and is the total annualized cost, computed as the sum of capital and replacement annualized cost and annual O&M cost.

(2) HOMER also computes the levelized cost of electricity (LCOE), defined in equation (3), which is

an important parameter to compare two different systems.

(3) Where and , respectively the boiler marginal cost and the yearly total thermal

load served, are considered obviously only if a boiler is present. The parameters at the denominator are the different types of electrical energy load served in a year.

3.4 Design principles of the biogas digester

If on one side, the design and optimization of the electrical energy systems are attained through the help of HOMER, the biogas digester design is achieved through a different approach. The target is to exploit the biogas resource as much as possible, therefore the design aims at the optimal size to digest all the available organic resource with reasonable costs. Among the different types of digester, the floating drum digester (illustrated in its main components in figure 13) is selected as a suitable alternative for the reduced complexity necessary in developing countries [39].

Figure 13. Scheme of a floating drum digester

(29)

KTH - ENTECH InnoEnergy MSc - Uppsala University 27 Once computed the available mass of organic input, the volume of the digester, which depends mainly on the daily intake, is consequentially set.

It is important to ensure a right retention time for the substrate, sufficient for the degradation processes but not so high to cause overfeeding. Following the design principles suggested in [40], two parameters are used to calculate the proper digester volume, the organic loading rate (OLR) and the hydraulic retention time (HRT). Where OLR represents the amount of volatile dry mass (VDM) introduced daily di ided the digeste ’s olu e, and HRT represents the theoretical time period that the substrate spends in the digester.

They are respectively computed as described in equation (4) and (5):

(4) (5)

Where Sub. input represents the digester daily mass intake, DM and VDM are the percentages of dry mass and volatile dry mass on the total and the digester volume is expressed in cubic meters.

The two parameters have a limit depending on the kind of digester. Thus by reversing the formula and introducing these limit values, two different lower boundaries for the digester volumes are found. The most restrictive limit sets the dimensions.

Biogas yields for the available kind of resources are then acquired from references found in the literature and the total amount producible is achieved summing all the contributions.

The total costs of the system are then computed following the same methodology implemented in HOMER, in order to obtain coherent results for the comparison.

(30)

4. Data

The important data and assumptions regarding energy consumption, which are necessary to determine the electrical load profile and the cooking demand, are presented and described in section 4.1. Availability of resources and components choice with relative costs are introduced in chapter 4.2 and 4.3 respectively. None of the data presented has been directly measured.

They were assumed according to different methodologies and different references/datasets, described case by case in the subsections. Because of the very specific situation and the early stages of the project development, this was a necessary choice, which obviously affects the results with uncertainties. However every decision was taken in accordance with the project group in order to meet the requirements of the project owners and achieve results with enough accuracy.

4.1 Energy consumption

The data regarding the electrical consumptions are described in the following chapters divided in the three main categories of electrical load defined in chapter 3.1.1. Thermal cooking demand is then described separately in chapter 4.1.4.

4.1.1 Domestic

One user class is assigned to each of the four main categories expected to inhabitate the village, as listed in section 2.1.3. A guideline furnished by the project group is to consider efficient and energy savings appliances when considered appropriate (i.e. LED was selected for lighting technology). The electrical appliances' rated power is difficult to determine and real surveys were not conducted due to time constraints. Therefore, the values computed in projects based on the same load estimation procedure as in real surveys in similar socio- economic context [41] [42] have been chosen and compared with the group expectations and the residential power consumption census for India [43]. When values were not available, an average among two different dataset [44] [45] was calculated. If necessary for the important contribution in terms of power, specific datasheet of appliances were accounted (i.e. LED lighting technology [46]). Also the definition of the windows of operation followed the same p o edu e, o i i g the g oup’s e pe ie e a d the intervals found in the above mentioned projects.

The resulting set of typical appliances, usage pattern and windows of operations is shown in appendix II.

4.1.2 Commercial and common infrastructure

The appliances, technologies and usage patterns, presented in appendix II, were assumed and dimensioned following the same procedure and references described for the domestic load estimation.

(31)

KTH - ENTECH InnoEnergy MSc - Uppsala University 29 4.1.3 Water system

The daily water requirement is indeed one of the most important parameter to design the system and components, in order to estimate then the electrical needs. According to the literature, the minimum requirement of water quantity per person to live and fulfil the basic activities is around 50 l/day [47]. According to a field survey provided by the stakeholders, the per capita domestic water consumption in several cities of India stay around 90-100 l/day.

Considering the number of inhabitants and the water requirement of 100 l/pp/day, in order to be conservative, the daily amount of water to be pumped for domestic purposes is about 25 m³/day. With the defined water requirements and the technical support from the expertise in the field represented by GÖran Spaxes (founder of Spaxes miljÖteknik AB), the technical specifications and relative electrical energy requirements were defined.

To exploit as most renewable energy technologies as possible, domestic drinkable water is going to be pumped during the most convenient hours of the day (with higher irradiation) up to a water tower, through the help of a 5.5 HP pump capable to cover the requested quantity within 3 hours maximum at 30 meters height, which is the assumed gross head between the two reservoir. At the WWTP, the electrical requirement consists in one pump of 1.5 HP and two air blowers of 0.12 kW, necessary to dry the sludge. The pump operates for 10 minutes every 2.5 hours and each of the blowers works for a total of 15 hours per day (3 hours in each bath of sludge, 5 bathes per day). Furthermore, seven pump stations collecting sludge from different points of the village are needed. At each of them, a single pump with 1.5 HP will operate three times per day for a total of one hour, in intervals of 20 minutes each. It is essential to empty the stations before other sludge is arriving to not overload the system when peak times occur, hence it is important to make them work at least once during the afternoon before the evening peaks. In respect to agricultural water needs, the estimates were made by the expertise inside the group. Not only the agricultural land of 20 m²/pp, but also citrus and mango planted with a decorative function require considerable watering. The combination of them results in a total request of 3,136 m³/year, meaning 11.6 m³/day, which is possible to be covered just with a small pump of 0.75 HP.

Different pumps found in the catalogue of Shakti Pumps Ltd, I dia’s leading manufacturer of energy efficient submersible pumps [48] were used as references in order to assess the pu ps’

relative electrical power request.

In table 4, the technical requirements and relative energy loads corresponding to each purpose are summarized.