The potential benefits to balance power shortage in future mobility houses with hydrogen energy storages

UPTEC STS 19038

Examensarbete 30 hp Juni 2019

The potential benefits to balance power shortage in future mobility

houses with hydrogen energy storages

Melissa Eklund

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

Abstract

The potential benefits to balance power shortage in future mobility houses with hydrogen energy storages

Melissa Eklund

This master thesis investigated how a hydrogen energy storage could be used and

dimensioned to reduce the problem of power shortage in the local distribution

grid in Uppsala, Sweden. By implementing such a storage system in mobility

houses, which are parking garages with integrated charging stations for electric

vehicles and smart renewable energy solutions for power generation, the problem

with power shortage could be decreased. The results showed that by integrating a

hydrogen storage together with battery packs, it was possible to reduce power

peaks in mobility houses. Further, it was clear that more power peaks facilitated

the dimensioning of these type of systems. It was also shown that due to today's

initial cost of hydrogen storages, the total savings related to a limited purchase of

electricity from the grid were insignificant. It was therefore found that this type of

hydrogen storage would not reduce costs in the short term for the mobility houses

considered in this study. However, implementing a smaller kW storage could

generate and improve knowledge in the hydrogen/hybrid field, which could

facilitate the implementation of larger systems in the future. Furthermore, the

results showed that it could be interesting to implement hydrogen storages on a

bigger scale for municipalities or actors, who would want to reduce the power

shortage in the local distribution grid.

Sammanfattning

Dagens samhälle står inför växande utmaningar när det kommer till elnätets begräsningar med avseende på effektbrist och integration av förnybara energikällor. Det existerar redan stora problem gällande effekt- och kapacitetbrist i flera delar av Sverige. Detta försvårar utvecklingen av kommuners tillväxt och städernas urbanisering, samtidigt som det blir svårare att möta efterfrågan på den el som förväntas öka under de närmaste åren. Att bygga nya ledningar är dyrt och tar tid, och tillsammans med den ökande andelen el- bilar kommer problemet gällande effekt- och kapaicitetbrist också att öka (EI, 2019) (Energimyndigheten, 2019). Problemet blir även svårare att hantera till följd av att fler människor använder energin samtidigt. Det är därför intressant att undersöka hur ka- paciteten i elnätet till viss utsträckning, kan avlastas med smart teknik och genom att arbeta med flexibilitetsbehov på olika sätt. Det är således viktigt att undersöka möj- ligheterna att kontrollera och anpassa elförbrukningen för att säkerställa att Sverige kan hantera framtida effekt- och kapacitetsproblem.

Uppsala Parkerings AB (UPAB), har tillsammans med Stuns Energi, redan påbörjat ett arbete inom detta område med projektet, Morgondagens Mobilitetshus. Detta har lett till att UPAB vill undersöka olika metoder till energilagring som senare kan använ- das för att ladda elbilar och kapa effekttoppar. UPAB är i synnerhet intresserad av att undersöka hur denna typ av system tillsammans med elbilsladdning och säsongslagring av energi från vätgas, kan dimensioneras för att upprätthålla en elektriskt driven bilpool.

Införandet av vätgas i energisystemet är beroende av hur väl det kan mätas sig med andra metoder (NE, 2019b). Således är det intressant att förstå hur vätgas kan använ- das tillsammans med andra tekniska lösningar, för att skapa större värde för både lokala energisystem och för mobiliteshus. Mobiliteshus är tredimensionella byggnader som inte- grerar flera olika tekniska lösningar och metoder för att reducera effektbrist i det lokala distributionsnätet, samtidigt som de främjar hållbarhet och mobilitet i Uppsala stad.

Idag existerar det ingen studie om hur mobilitetshus med säsongslagring av energi, i form

av vätgaslager, kan bidra till detta komplexa och kommande problem. Projektet kommer

därför att bidra med ökad kunskap inom området för energisystemsteknik och hållbarhet

för parkeringsföretag, kommuner och akademin.

Acknowledgments

This master thesis was a collaboration between industry and academia during the spring of 2019. The opportunity to learn from highly experienced individuals from different dis- ciplines has highly motivated me during this process. I would like to thank everyone who has shown interested in my work and given me inspiration along the way. Especially, I would like to thank my supervisor, Rafael Waters at Uppsala University, for trusting me with my own ideas and giving me guidance along the way. I would also like to thank my subject reader, Valeria Castellucci at Uppsala University, for always contributing with enthusiasm while giving support with the development of the simulation model and this master thesis. It has been a pleasure working for the Division of Electricity at Uppsala University.

Melissa Eklund

July 2019,Uppsala.

Table of contents

1 Introduction 7

1.1 Aim of study . . . . 8

1.2 Limitations . . . . 8

1.3 Outline of the report . . . . 9

2 Background 9 2.1 The Swedish power grid . . . . 9

2.1.1 Power shortage and power peaks . . . . 10

2.1.2 Potential impact of electric vehicles on the power grid . . . . 11

2.2 Predictability and load forecasting . . . . 11

2.3 Mobility house - Dansmästaren & Brandmästaren . . . . 12

2.3.1 Previous research - Dansmästaren . . . . 13

3 Hydrogen Storage 13 3.1 Hydrogen . . . . 13

3.2 Production of hydrogen . . . . 14

3.3 Different electrolysers . . . . 15

3.3.1 Alkaline electrolysis . . . . 15

3.3.2 PEM-electrolysis . . . . 16

3.4 How is hydrogen stored? . . . . 17

3.4.1 Liquefied Hydrogen . . . . 17

3.4.2 Compressed Hydrogen . . . . 17

3.5 Fuel Cells . . . . 18

3.6 Batteries . . . . 19

3.7 Hybrid storage . . . . 19

3.8 Safety . . . . 20

4 Method 21 4.1 Overview of method and implementation . . . . 21

4.2 Data and assumptions . . . . 22

4.2.1 PV System and power consumption . . . . 22

4.2.2 Electricity price . . . . 23

4.2.3 Battery packs . . . . 24

4.2.4 Hydrogen Storage . . . . 24

4.5 Case 2 - H2 storage in the MW range . . . . 29

4.6 Summary of Cases and Scenarios . . . . 31

5 Results 32 5.1 Case 0 - A small H2 storage . . . . 32

5.2 Case 1 Simulations - A H2 storage in different dimensions . . . . 34

5.2.1 Simulation 1 - 30 power peaks . . . . 34

5.2.2 Simulation 2 - 60 power peaks . . . . 37

5.2.3 Simulation 3 - 90 power peaks . . . . 40

5.2.4 Simulation 4 - 120 power peaks . . . . 43

5.2.5 Simulation 5 - 150 power peaks . . . . 46

5.3 Case 2 - A H2 storage in MW range . . . . 49

5.3.1 Simulation 1 - Maximum H2 storage size . . . . 49

5.3.2 Simulation 2 - H2 storage with MW capacity . . . . 51

5.4 Summary of results . . . . 54

6 Discussion 55 6.1 Model limitations . . . . 55

6.2 The potential of H2 storages . . . . 56

6.3 Dimensions and requirements for H2 storages . . . . 58

6.4 Economical feasibility and estimates for H2 storage . . . . 59

6.5 Further scenarios and research . . . . 60

7 Conclusion 61 8 References 62 9 Appendix 1 - List of compontents used in all cases 66 9.1 Components for Case 0 . . . . 66

9.2 Components for Case 1 . . . . 66

9.3 Components for Case 2 . . . . 67

10 Appendix 2 - Technical design mobility houses 68 10.1 Brandmästaren . . . . 68

10.2 Dansmästaren . . . . 69

1 Introduction

Today’s society faces growing challenges related to the power grid’s limitations. In several parts of Sweden, there is already a great problem regarding power shortage and capacity of the power grid. It risks counteracting the integration of renewable energy and making it difficult to meet the increased demand for the electricity expected in the comping years.

This also causes difficulties for the growth of municipalities and future urbanization.

Building new lines is expensive and takes time, and together with the rapidly increasing proportion of electric vehicles (EVs), the problem with power- and capacity shortage will also increase (EI, 2019)(Energimyndigheten, 2019). Because of the fact that more and more people are also using energy at the same time, the problem becomes more difficult to manage. It is therefore interesting to investigate how the capacity in the power grids can to some extent be revealed through smart technology and by working with flexibility demand in different ways. Thus, it is important to examine the possibilities of controlling and adapting the consumption of electricity to ensure that Sweden can cope with future power and capacity problems.

Uppsala Parkerings AB (UPAB) is, together with Stuns Energi, already investigating this area with a project, The Mobility House of Tomorrow. This has lead to UPAB want- ing to examine different ways to store energy, that later can be used for EV charging.

In particular, UPAB is interested in investigating how this type of system together with

EV charging and seasonal storage of hydrogen (H2) can be dimensioned to sustain an

electrical powered car pool. The introduction of H2 in the energy system depends on

how well H2 can cope with other methods and technologies (NE, 2019b). Thus, it is also

interesting to understand how H2 can be used together with other technologies to create

a bigger value for both the local energy system and the mobility houses. A mobility house

is a building which will integrate multiple new technologies and methods to reduce power

shortage in the local distribution grid, while increasing sustainability and mobility in the

city. There is no study today that examines how a mobility house with a seasonal H2

energy storage can contribute to this complex and upcoming problem. This project will

therefore contribute to increase knowledge in this area of energy system engineering and

sustainability for parking companies, municipalities and academia.

1.1 Aim of study

The purpose of the study is to examine how H2 energy storage systems can be used and dimensioned for peak shaving and what opportunities it can create for planned mobility houses in Uppsala. Further, it is interesting to investigate in what ways different dimen- sions and functions of such a system can contribute with diverse potential and difficulties.

The aim of the thesis is to answer the following research questions:

• What is the potential of a H2 energy storage system to meet different power needs during winter?

• How can the system be dimensioned to contribute to mobility houses like UPAB’s Brandmästaren and Dansmästaren?

• What are the economical estimates of a H2 storage system based on prices of avail- able systems and components on the market?

I have given an answer to these questions by developing a computing model in MATLAB and Simulink, together with analyzing the market of available off-the-shelf systems for H2 storages.

1.2 Limitations

The focus of the study is to investigate the aspect of H2 energy storages in relations to the mobility house. It is therefore interesting to examine the aspects of reliability, cost, size, efficiencies and other relevant parameters that will be important for the mobility houses. Because the mobility houses are under design and construction, all specifications regarding the energy system and dimensioning are not finished. This contributes to the fact that only the two first planned mobility houses will be used and modelled for this report. The focus will be on the second planned mobility house, Brandmästaren, due to the first planned mobility house, Dansmästaren, already being under construction. How- ever, Dansmästaren will be used for comparison in order to get a broader understanding of different dimensions of both H2 storages and components related to the system.

The data used for the developed computing model is mainly stochastic and based on

previous research. These limitations will influence the results, for example when choosing

off-the-shelf H2 storage components. The most extensive limitations in the simulations of

the H2 energy storage are the low resolution and uncertainty in the existing data. Both

solar radiation data and energy profile data have a resolution of one hour, which affects

losses in the system are approximations of real conditions.

Since the use of renewable energy is the main focus in this report, the production of H2 is only considered to be through water-electrolysis when modelling for a whole stand- alone system. The production of H2 can be through different methods in other cases, but often with the use of fossil fuels (Hydrogen Europe, 2019a). Further, due to time constraints only main components for a H2 storage will be considered and investigated.

1.3 Outline of the report

The report is structured as follows. Chapter 1 introduces the subject, the topic of the report and the aim of the study. Chapter 2 presents the background with focus on the Swedish power grid, power shortages and the predictability of power peaks. It also provides a background of the mobility house project and previous research. Chapter 3 is mainly for readers with basic or little knowledge about H2 energy storages. It gives a brief overview of technologies relevant to or used in this study. Technical definitions will be described as well. Chapter 4 presents the method for the project, where the MATLAB and Simulink-model which is used for the dimensioning and analysis of the storages will be explained. It also presents the data and assumptions used for the dimensioning and design of the storages. Chapter 5 presents the results from the simulations done in the MATLAB and Simulink model. Chapter 6 discusses the restrictions of the model and therefore the feasibility of this type of H2 storage solution for mobility houses in Sweden.

The results from all simulation cases are also discussed together with suggested further research. Lastly, Chapter 7 presents the conclusion.

2 Background

This chapter presents information with focus on the the Swedish power grid, power short- age, predictability of power peaks and how the increase of EVs can affect the power consumption. It also provides a background of the mobility house project and previous research.

2.1 The Swedish power grid

The Swedish power grid is divided into three categories: transmission grid, regional and

local distribution grid. The regional and local power grid is owned by approximately a

power plants to all local an regional distribution grids. The regional distribution grid then branches into local distribution grids where electricity is transferred and distributed to smaller industries, households and other users. Together, they form a single connected system together with all power generation plants and all places where electricity is used (IVA, 2016).

2.1.1 Power shortage and power peaks

Power shortage occurs when the demand for electricity at a certain time is greater than the production. Not enough electricity is delivered during certain time windows, even locally. Situations which contribute to power shortage are cold weather, not enough wind for wind turbines (usually when it is cold), a dry year, less power from nuclear, re- striction of electricity import and reserve power, for example, gas turbines, are unable to deliver sufficient power. We have an increased risk of power shortage in our growing cities:

Uppsala is one of the cities which is in danger of this problematic situation (Ellevio, 2019).

Much like power shortage, power peaks occur in the power grid when the power de- mand is at its highest, usually when many electrical appliances are used simultaneously, or during dry years and cold winters (Swedish Smartgrid, 2019)(Energimarknadsinspek- tionen, 2014). Cutting power peaks is desirable for both economic and technical reasons.

A more even load on the power grid, under the same total energy consumption, enables the connection of more outlet customers or more electricity products without the capac- ity of the power grid having to increase. This results in the possibility to connect more renewable energy sources, micro production and plug-in electric vehicles (Copenhagen economics, 2017).

Power grids will need to deal with power peaks to a greater extent due to the fact that fu-

ture infrastructure will become more electrically powered. It is therefore important from

a socio-economical perspective, to provide incentives to the companies which are willing

to implement measures for a more efficient use of electricity (Copenhagen economics,

2017).

2.1.2 Potential impact of electric vehicles on the power grid

The electrification of the transportation that requires charging infrastructure will accord- ingly to Elforsk (2014) Future demand on the electricity grid, mean higher power peaks and thus a capacity limit which is exceeded in certain segments of the power grid (Elforsk, 2014). This is because charging of EVs without incentives or control will probably coin- cide with Sweden’s power peaks, contributing to an increasing electricity demand during already high consumption time windows. However, if EVs can be charged with surplus energy from renewable power generation, it can instead have a positive effect on the power grid (IVA, 2016).

2.2 Predictability and load forecasting

The power grid needs to support any level of load to avoid power outages and in order to control the power peaks, it is important to predict a peak prior to its occurrence (Karlsen

& Goodwin, 2014). In order to reduce the power peaks, it is valuable to be able to pre- dict them, in other words, to forecast the consumption load of the building, in case of a mobility house for example. The consumption load of a building is often in line with the load in the power grid (Copenhagen economics, 2017).

Another factor which causes power peaks is cold weather. During winter, the days are shorter and the temperature is usually below 0 degrees Celsius. Due to it being darker and colder outside, people turn more lights on and turn up the heat in buildings. In winter, electricity consumption in Sweden doubles. According to (SvKs) forecasts, the ability to be self-sufficient with electricity decreases when the electricity is needed the most (Svenska Kraftnät, 2019).

There are various studies regarding load forecasting where different parameters, methods

and models are used. A research group from the University of Adger shows that with a

stochastic model that uses neural networks, power peaks of hour size can be predicted

up to one week in advance and with 80 % accuracy. The study is done by mapping the

prediction activities and solve on previous consumption data (Goodwin & Yazidi, 2014).

2.3 Mobility house - Dansmästaren & Brandmästaren

Uppsala Parkerings AB (UPAB) is a municipal-owned company with the mission to build mobility houses for 1.8 billion SEK over the next ten years. A mobility house is a parking garage with integrated charging stations for EVs and smart renewable energy solutions for power generation. The first two mobility houses are being designed and build to be used as test beds for cutting-edge technology and system solutions in mobility. A problem generated by the increasing number of electric charging vehicles and by urban growth is the power grid’s limitation with regard to power. The mobility houses will, therefore, be part of the problem because of the large number of charging poles for EVs, and therefore a high power requirement. However, there is a number of technical and social systems that together can operate so that the mobility houses can contribute to balancing and strengthening the power grid instead. Because of this, UPAB wants to investigate the possibility of moving electricity consumption and cutting power peaks using different technical solutions, such as energy storages, as mentioned earlier, and Vehicle-to-Grid (V2G). Thus, the role of the mobility house in urban planning has the potential for inno- vation, such as the opportunities in cutting power peaks to balance the local distribution grid (Naturvårdsverket, 2019).

The first mobility house, Dansmästaren, is under construction and estimated to be fin-

ished by 2020. It will be a three-dimensional property, consisting of a car park, grocery

store and student housing. It is located in the district of Rosendal, Uppsala. Here, 500

parking spaces are planned, of which 108 with charging poles with a maximum power

of 3.7 kW each. The house will also have a bio-roof together with PV panels with an

efficiency of 19.3 % and covering a roof area of about 400 m

. The PV panels will be used

to produce power for the charging infrastructure for EVs when the need exists. When

the PV production is greater than the consumption of the mobility house, the excess

power can be stored in battery packs over shorter time intervals. The construction of the

second mobility house, Brandmästaren, is planned to start in the spring of 2020 and will

be located in the district of Rosendal as well. Because Dansmästaren is already under

construction, the design, technical drawings and planning of Dansmästaren will be used

as a reference in this thesis when modelling and dimensioning the H2 storage for Brand-

mästaren. The technical drawings of Dansmästaren and Brandmästaren can be viewed

in Appendix 2. To clarify, these drawings are not completely finished and changes can

therefore be made later on.

2.3.1 Previous research - Dansmästaren

Uppsala University together with, UPAB, Stuns Energi and university students have studied the design and suggested the technologies needed for Dansmästare in order to reduce power peaks that arise in the mobility house. This has been done by modelling in MATLAB and Simulink as well by investigating new technologies and components which satisfy the needs of the mobility house. The MATLAB and Simulink model integrates the consumption need for the building, the power production from the PV system, charging and discharging of battery packs together with an economical model. The economical model calculates the total savings and earnings of the electricity sold back to the power grid and the difference in electricity consumption when using energy storages and when only purchasing electricity from the power grid. The consumption in the house is assumed to be dependent on lightning for all levels of the parking garage and the EV charging poles. This study also resulted in the choice of the battery packs for storing excess PV power: Nilar and Greenrock battery packs having capacity of 23 and 60 kWh respectively.

3 Hydrogen Storage

This chapter presents the theory and information about hydrogen, hydrogen storage meth- ods together with battery packs, conversion methods and relevant components.

3.1 Hydrogen

Hydrogen is the lightest and most common of all elements. Hydrogen gas is about 14.3 times lighter than air with the same volume, pressure and temperature. It is a color-, flavour-, odorless gas and has the highest diffusion and effusion ability of all elements, due to its poor molecular mass. Hydrogen is not toxic or dangerous by itself, but explo- sive in a mixture with air or pure oxygen. Mixtures with air and 5-75 % hydrogen are explosive as well as mixtures with pure oxygen and 4.7-94 % hydrogen. The reaction is particularly violent if the gas mixture consists of the same proportions as in water, two parts of hydrogen and one part of oxygen. The explosion can be initiated by e.g electric sparks or heating (NE, 2019b).

Already in the 19th century, information and knowledge regarding hydrogen as a fuel

in engines or for other energy applications were developed. During the time 1920-30,

many scientists described methods and techniques for production, storage and use of hy-

becoming as an important future energy carrier in different applications. Even though hydrogen is available in almost inexhaustible amount and has several properties that make it interesting in an energy context, it is not an energy source. Hydrogen must be produced by a energy, but unlike electricity hydrogen can be stored directly (NE, 2019b).

Hydrogen produced with renewable energy can become an important element of the energy system. In the near future, it is reasonable that the interest is largely focused on the good environmental properties of hydrogen. The biggest environmental benefit of hydrogen in an energy context, is that the emission mainly consists of water vapor, depending on the production method. This also means that combustion of hydrogen do not increase the greenhouse effect, due to its emissions does not contain carbon dioxide (NE, 2019b).

The amount of energy per unit weight of hydrogen is greater than in any other fuel, almost three times the size of fuel oil or petrol. However, its energy per unit volume is low. Pressurized hydrogen takes a greater space than liquefied hydrogen, and the methods of storage are presented further down in this section. The technology is still expensive when comparing to other energy carriers or energy storage alternatives (NE, 2019b).

Hydrogen can be used in a variety of energy applications in both transport, industrial and housing sectors. It can be used for heat production, fuel for vehicles and energy storage applications, on both large and small scale (NE, 2019b). Due the inherent high mass energy density and insignificant leakage of hydrogen when using it to store energy, it is suited for long term and seasonal storage applications (Agbossou et alt, 2004). The Swedish efforts in the field of hydrogen are limited, many have basic research character and are aimed at the production of hydrogen by photochemical and photobiological meth- ods and for the development of hydrogen storage in metal hydrides. However, Sweden is participating partly in cooperation within the EU framework program to the research in hydrogen and hydrogen applications (NE, 2019b).

3.2 Production of hydrogen

Electrolysis of water is one method for the production of hydrogen, which enables the possibility to produce green hydrogen with electricity from renewable energy sources.

However, it is more expensive than other common production methods, especially if the

electrical need for the production is bought from the power grid. This method also enables

3.3 Different electrolysers

An electrolyser uses electrolysis which is the process of using electricity to split water into hydrogen an oxygen, which can be seen in Figure 1 (Hydrogen Europe, 2019b). The developed electrolysers that exist on the market today are Alkaline (AEL) and Polymer Electrolyte Membrane (PEM) electrolysers. There are also other methods and designs, Soild Oxide Electrolysis Cell (SOEC)-electrolysis, High-temperature electrolysis, Photo- electrolysis and Photo-biological production. However these methods and designs are not ready for implementation due to the stage of the development process and therefore only AEL and PEM-electrolysers are being analyzed further in this report (ÅF, 2015).

Both AEL and PEM electrolysers consists of the same main components, electrodes, electrolyte and a membrane, as well as consistent flow of electricity and deionized wa- ter when operating. Electrolysers are differentiated by the temperature at which they operate and the material of the elctrolyte. AEL electrolysis is the easiest, most applied and mature electrolysis technology which is commercially available. It is the basis for the development of PEM electrolysis (ÅF, 2015). Both AEL and PEM electrolysers are explained further in the next sections.

Figure 1. Illustration of the chemical process of an electrolyser (Shell,2017) .

3.3.1 Alkaline electrolysis

The alkaline electrolysis (AEL) is a mature and robust technology, with relatively low

for decades. The electrolysis takes place at relatively low temperatures, between 70-80

◦

C and works either atmospherically or under elevated pressure. The pressurized AEL has a lower efficiency than atmospheric AEL and produces a lower purity product. The foremost advantage of using pressurized AEL is that it produces compressed hydrogen with less additional energy input. This implies that the reduction in electric efficiency of the AEL with increased pressure is lower when comparing to the energy needed to compress the already produced hydrogen (Götz et al. 2015).

Anode and cathode materials are usually made of steel or nickel-plated steel. Due to the electrolyte being corrosive, the electrodes must be replaced or re-activated after 2 to 5 years. The technology has been developed for continuous operation and is therefore not sufficiently flexible to meet varied electricity production. One problem is the restart time of the system after a following shutdown. This means that this type of technology is not the optimal choice when considering using electricity from renewable production.

The disadvantage is also that it has a relatively low efficiency (50-70 percent) when com- paring against other techniques. Nor it can go into reversible operations which means that it is not possible to produce electricity from hydrogen gas, which is possible with PEM-electrolysis (ÅF, 2015).

3.3.2 PEM-electrolysis

The PEM electrolysis is a further development of the AEL design. The electrodes and

electrolyte together form a membrane electrode assembly (MEA). The life of electrodes

and electrolyte is between 5 and 10 years where the worn is mostly caused due to thermal

and mechanical stress. Compared to an AEL electrolyser, the PEM design has about the

same working temperature, about 80

C, but a higher efficiency (67-82 %) (DNV KEMA

2013). Due to the fast response time of the PEM technology, it has better conditions for

varied operation, compared to the AEL design. It also has a reversible function which

means it can be used both for electrolysis and as a fuel cell for the production of electricity

(ÅF 2015). Furthermore, the purity of the produced hydrogen is very high with PEM

electrolysis and the minimum load is reported to be 5 %. PEM electrolysers can therefore

handle a very quick start-up and shutdown at the same time as it is more compact than

alkaline electrolytes (Götz et al. 2015).

3.4 How is hydrogen stored?

There are different methods for storing and transporting hydrogen, and each method has its advantages and disadvantages. H2 can be stored as liquid, gas or bound to other substances in hydrogen rich chemical compounds, such as methanol and ammonia, or to metals in the form of metal hydrides. H2 gas in pressure vessels is bulky, while H2 in liquid form requires condensation. Meta hydrides are relatively heavy, but safer than other methods of storage. Transport of H2 can take place in pipelines, train or ship (NE, 2019b). This thesis will further on only highlight the methods for liquid and compressed H2.

3.4.1 Liquefied Hydrogen

Liquefied hydrogen (LH2) has a higher energy density as an energy carrier, than gaseous H2. It is thus important to implicate that LH2 requires liquefaction at -253

C, which involves a complex technical plant and extra economic cost. However, due to higher energy density the storage can become less bulky due to less volume needed for the same amount of energy. When storing LH2, the tanks and storage facilities need to be insulated in order to check if evaporation occurs, if heat is carried over into the stored content.

This can happen due to conduction, radiation or convection. Today, tanks for LH2 are used primarily in space travel (Hydrogen Europe, 2019c).

3.4.2 Compressed Hydrogen

The easiest way to decrease H2 gas volume while keeping a constant temperature, is to increase its pressure. Compressed H2 is a method of storage where the shape of the vessels is generally cylindrical and the pressure is usually between 300 and 700 bar, usually a compressor system is used for this processes. Compressing H2 to 350 bar decreases the required storage volume by 99.6 %, which can be derived from Equation 1. If higher pressure is used, the required storage volume is lowered, but the compression work input and safety concerns increase (Dagdougui et alt. 2018).

p

· v

= p

· v

(1)

where p

and v

represent the pressure and volume before the H2 gas is compressed, and

p

and v

represent the pressure and volume under compression.

3.5 Fuel Cells

Much like an electrolyser, a fuel cell consists of two electrodes, and anode and a cathode, with an electrolyte between them. Unlike the electrolyser, the fuel cell produces electricity from H2 and oxygen, with only water and heat as by products. H2 reacts with a catalyst at the anode, which creates a positively charged ion and a negatively charged electron.

The proton passes through the electrolyte at the same time as the electron travels through the circuit and creates a electrical current. At the cathode, water and heat are formed due to the reaction between oxygen, ion and electron. This process can be seen in Figure 2 (Hydrogen Europe, 2019b).

Figure 2. Illustration of the chemical process of a PEM fuel cell (PowerCell, 2019b.

Fuel cells can vary from devices producing only a few watts of electricity to large power

plants which produces megawatts. The type of the fuel cell is generally classified according

to the type of electrolyte they use. Each type is suitable for different applications and

requires different materials and fuels. The PEM technology is the most commonly used

technology today when it comes to fuel cells. The technology is reliable with dynamic

characteristics which allows for full power output within seconds. It also has the capability

for extensive starts and stops (PoweCell, 2019). In this project, PEM-fuel cells are being

considered due to its application and features fits with the integration of renewable energy.

3.6 Batteries

Batteries are included in the category of electrochemical energy storages, and can be categorized further into primary or secondary battery cells. Primary battery cells are non rechargeable while secondary battery cells are rechargeable and can be used in an energy storage perspective (MIT Electric Vehicle, 2008). A battery cell is similar to a fuel cell and include a cathode, anode, electrolyte and a separator. Battery cells can also be ordered into sets, in both parallel and series, then called battery packs (NE, 2019a).

An important characteristic of a battery is the State of Charge (SOC), which describes how much capacity (in Ah) is available of the battery. If a battery is fully charged for example, it has a SOC of 100 % (Murnane et alt, 2019).

The energy stored in battery packs will discharge due to the characteristics of the battery called self-discharge (NE, 2019a). The discharge rate describes how fast a battery can be discharged, which is different for different battery types. The discharge rate can be described in effect and is called E-rate. For example if a battery has a discharge rate of 0.25, it means that it takes 4 hours to discharge the battery due to 1 hour / 0.25 = 4 hours (MIT Electric Vehicle, 2008).

3.7 Hybrid storage

In real life energy storage application where H2 is a key component, the system includes

both short-term and long-term energy storage. The short-term storage is often based on

battery packs due to its high round-trip efficiency, ability to take care of instantaneous

power peaks and the convenience of charging and discharging. However, the battery packs

low energy density, self-discharge and leakage-, make the technology not suitable for usage

as a long-term energy storage. Because of this, H2 is used as a solution for the long-term

energy storage. By combining battery packs and H2, an improved storage solution is

created and both technologies can complement each other to increase the performance

of a renewable energy system (Agbossou et alt, 2004). Different energy storage methods

and their optimal area of use is shown in Figure 3.

Figure 3. Illustrates the capacity and discharge duration for different energy storage methods (Hydrogen Euorpe, 2019c)

.

There are both research and case studies, as well as real life implementations of H2 storages in hybrid energy storage systems. One actor who is well known for driving much of the research of H2 storage systems in Sweden is Hans-Olof Nilsson who has built his own off-grid house. Hans-Olof has founded a company which provides complete commercial solutions for off-grid projects and support. One of the main projects is his own house which has been operating since 2015. The off-grid system is power by PV power and uses a hybrid energy storage consisting of H2 gas and battery packs. This project has showed that the household can be self-sufficient during summer as well as winter (Nilsson Energy, 2019).

3.8 Safety

H2 safety is a compelling matter due to H2 gas being highly flammable when in con-

tact with oxygen. However, it is important to emphasize that H2 has been used in vast

quantities as an industrial chemical in industries for over 40 years. Because of this, in-

frastructures which produce, store, transport and utilize H2 safely have been developed

over a long period of time (Hydrogen Europe, 2019a)(Vätgas Sverige, 2019e).

H2 is the lightest element in the world and therefore it ascends rapidly into the atmo- sphere when a leakage occurs. Like any other fuel and due to being a highly compressed gas, H2 requires clear rules of usage (Hydrogen Europe, 2019e). If H2 leaks out of fuel taps and would be mixed with oxygen in normal air, the results could be production of rock gas that may explode. To prevent such a risk, indoor ventilation is sufficient in most cases. If the H2 is stored outside of the building, the risk are very small, due to H2 gas being fourteen times lighter than air and therefore quickly draws on weather and dilutes.

But the handling of H2 still requires a safe technique to prevent these risk despite them being low (Energimyndigheten, 2019a)

4 Method

The work is carried out in collaboration with Stuns Energi and is also a case study of the mobility houses Dansmästaren and Brandmästaren in Uppsala. The case study will partly consist of simulations, energy calculations, economical calculations and scouting of commercial solutions to produce one or more concrete proposals for system solutions.

4.1 Overview of method and implementation

A case study with the mobility house Dansmästaren and its power production and load consumption is examined through simulations to determine the effects of a seasonal H2 storage. Data for the consumption and production from PV systems are provided from the department of engineering sciences, division of electricity at Uppsala University. The model created in MATLAB and Simulink for this master thesis, is integrated in the model from the research group which makes it possible to simulate the H2 storage together with battery packs and other technologies already included in the used model.

In order to examine different dimensions and options with H2 storages, three different

cases were modelled, where two of them were simulated. In order to have an understand-

ing of how to dimension the H2 storages, the power shortage in Uppsala municipality’s

low voltage grid has been used as a parameter. It is estimated that today’s power short-

age is in the range of 30-150 hours per year, and it is predicted that this number will only

increase in the future (Kristina Starborg, 2019). It is therefore interesting to use 30-150

hours to investigate different case studies for the mobility house with a H2 storage and

its capacity. All three cases are explained further in the section.

4.2 Data and assumptions

In order to implement a model where the H2 storage is being simulated against the 30- 150 largest power peaks for Dansmästaren and Brandmästaren, it was assumed that it was possible to predict these peaks 12 hours in advanced at least. As mentioned in the Background, where load forecasting is explained, short-time forecasting is used as a valid argument for the choice of model design.

4.2.1 PV System and power consumption

Brandmästaren is expected to have a larger PV system, car park and energy storage than Dansmästaren. However, the MATLAB and Simulink model from previous research is developed and used for Dansmästaren, which means that all components and their ca- pacities will need to be scaled up in order for it to be used for Brandmästaren. The size of the PV system for Brandmästaren is estimated from the PV system of Dansmästaren and from the expected consumption load. Due to Brandmästaren still being under de- sign, the scaling of the PV system is a rough estimate and the choice is to scale the PV system two times the size of the PV system for Dansmästaren. It is done because the dimensioning of the H2 storages for the mobility houses will result in a clearer difference and give more information about dimensions for different power needs. The installed capacity for the PV system will therefore be 302 kW. The number of parking spots are expected to be around 733 which is mentioned in the Background and can be seen in Appendix 2. The number of charging poles for EVs is also estimated from the previous research of Dansmästaren. The calculation and prediction of increasing EVs from Dans- mästaren are used and results in an estimate of 158 EV charging poles for Brandmästaren.

Figure 4 below shows the estimated net production and overproduction from the PV panels for Brandmästaren. The net production gives an indication of the power con- sumption and production for Brandmästaren, and therefore how often the mobility house needs to purchase electricity from the power grid to cover their power consumption need.

It also gives an understanding of the power peaks and when they occur. The overpro-

duction indicates how much surplus PV power is available after the power consumption

need of the mobility houses is covered and both the Nilar and Greenrock battery packs

are fully charged. It can therefore be used for an implication when dimensioning the H2

storage.

Figure 4. Illustrates the estimated net production and overproduction of the PV-system for year 2023. The PV power production minus power consumption of Brandmästaren equals the data for the net production. The overproduction represents all available surplus power from the PV system when both the power consumption of Brandsmästaren is met and both battery packs are fully charged.

4.2.2 Electricity price

All simulations for both Dansmästaren and Brandmästaren will be executed for the year

2023. This year is chosen due to it being expected that Brandmästaren will have the

constructing start of 2020, and therefore it should be finished around 2023. It is also

because of Dansmästaren already being under construction. Furthermore, this study is

based on the previous research and existing model of Dansmästaren, some assumption

and limitations occur. The economical part of the existing model of Dansmästaren is

based on the average electricity price for the year 2017 and 2018. Each hour will have

the same electricity price. However, the capacity subscription is included which is based

on spot price + marginal cost, without tax or VAT. So the time of the day and year will

4.2.3 Battery packs

The battery packs in the MATLAB and Simulink model was based on the battery packs which are being implemented in Dansmästaren. These are the GreenRock battery pack and Nilar battery pack wich both have 95% discharge depth. Because Brandmästaren is expected to have more EV charging poles and therefore a higher consumption, the battery packs will be scaled to have a higher capacity, as mentioned with the PV system.

The battery packs in Brandmästaren will therefore be scaled two times the size of the battery packs for Dansmästaren, which can be seen in Table 1.

Table 1. The dimensions of the battery packs used for Brandmästaren.

Parameter Battery 1 Battery 2

Capacity 2 · 23kWh 2 · 60kWh

E-rate 3 0.25

4.2.4 Hydrogen Storage

The dimensions of the H2 storage will change for each simulation case, which also means that no component is fixed. However, the components used are available on the mar- ket. Because prices for most components are not specified, the information of Hans-Olof Nilsson’s H2 energy storage system is used as a reference system to compare with for components without exact price information. This system has an electrolyser with an estimated cost of 500 000 SEK, fuel cell of 410 000 SEK and H2 gas cylinders for 367 SEK/m3. Nilsson’s H2 storage is approximately of the size of 3000m3 (Nilsson, 2019).

Because only the key components, electrolyser, gas cylinders and fuel cell are used in this study, the price information will be a rough estimate because of a real system having much more components and other costs. Another factor which affects the economical calculation is the scaling of the initial price of each components based on their scaling of capacities and sizes. This results in a higher calculated initial price due to not considering that prices do not change linearly. However, a real system will have other costs due to more components are needed, together with the installation costs. Lastly, the develop- ment of the pricing for components and H2 storages is not included. Thus using Nilsson’s system as a reference will not give an exact realization of the cost of this type of system.

All information about the components which will be used can be found in Appendix 1.

4.3 Case 0 - A small H2 storage

Case 0 is called 0 because its purpose is to give price indication which facilities the knowledge and understating of the subject and the system technologies. It will give price indication without any simulations being preformed. This means that no calculations with battery packs are being preformed. Case 0 only takes small kW H2 storage systems available on the market into account, and how that would affect the budget for the mo- bility house. Its purpose is also to bring knowledge of the components and what would be a good initial implementation. This small system will be interesting to UPAB when implemented due to measurements can be preformed and the company can learn how the system behave together with rest of the technologies in the mobility house. It means that the system itself does not need to create value in terms of cutting power peaks, but instead contribute with knowledge.

In order to give a price indication for a small H2 system, 30 power peaks, of the 30- 150 hours of power shortage, are being used as a reference as well as the capacity of the chosen fuel cells. This will give some information regarding the size of such a system and what price range it entails. Four different storage systems, together with two different storage sizes will be examined. Both the AEL and PEM technolgy will be used for the different H2 storages. This is because the purpose again of this case is not to find the best dimension or suitable system for peak shaving. Two of the systems will be stand alone systems consisting of an electrolyser, gas cylinders and a fuel cell, while the other two systems will purchase H2 directly from a distributor and therefore will not need an electrolyser. When purchasing H2 directly from a H2 gas distributor, it is possible to rent each cylinder for each day or year. It is also possible for the distributor to deliver and pick up the cylinders when needed. The size and capacities of each component used in this case, is chosen by considering costs, availability and efficiency. This means that the components used may not be the optimal choice, but they serve the purpose of the study.

4.4 Case 1 - H2 storage of different dimensions

In this case, simulations were executed with different size and capacity of each component

to provide information regarding different system dimensions. This case was therefore

designed to create a model of the system where the H2 storage is only used for the 30-150

largest power peaks. This will result in different dimensions of the storage system and

dimensioning the H2 storages. Each simulation scenario will be based on different power needs, ranging between 30-150 power peaks, for H2 storages in both Dansmästaren and Brandmästaren. This is because it gives a better understanding of different dimensions and the affect of components and technology sizes. This case is also based on the H2 storage being a part of a hybrid storage together with battery packs, as seen in the system design description. The H2 storage will only have the function of supporting the battery packs, which are used due to their capacity and fast response time. The system design can be seen in Figure 5.

Figure 5. Schematic of the simplified system design and its power flow.

The system is designed to be charged during the summer period of March to October,

and discharged during the rest of the year when most of the power peaks are arising. It

is based on that the overproduction from the PV production is firstly directed to both

battery packs, and if any surplus energy is remaining, this energy will go directly to the

electrolyser which will produce H2 to be stored in the H2 storage. For the discharge part,

when one of the 30-150 largest power peaks arrives and if both battery packs are not

fully charged, the fuel cell starts to produce power to charge both battery packs. This is

because the capacity of the fuel cell is not efficient to handle the biggest power peaks by

itself and because battery packs are better at handling fast consumption fluctuations. To

meet the consumption need for the mobility house during the time both battery packs

are charged with power from the H2 storage, power directly from the power grid will be

used during these hours. The fuel cell will continuously run on max power output while

charging the battery packs. In order for the battery packs to be fully charged before one

of the biggest power peaks arrives, calculations of the battery packs capacity and E-rate

are considered together with the fuel cell’s capacity. The charge time in MATLAB can

chargetime = f loor((BC

+ BC

· ERate

)/F C

) (2) where BC

and BC

are the capacities of the battery packs, F C

is the power production of the fuel cell and ERate

is the E-rate of the Greenrock battery pack.

The E-rate for the Nilar battery pack is not included due to it being high enough to have the ability to be fully discharged in less than one hour. The calculation of the charge time results in how much time is needed to fully charge both battery packs with only the H2 storage. As mentioned earlier, to use this information, it is also necessary to predict the biggest power peaks in order to know when the H2 storage should kick in and charge the battery packs. This is done with a part of the Simulink model which simulates the power flow of PV production and power consumption of the mobility houses for each hour. The power production and power consumption can then be extracted and analyzed in order to predict when the net production is at its lowest, which will correspond to power peaks in the mobility houses. This part of the Simulink model can be seen in Figure 6.

Figure 6. The Simulink Model which provides the net production for Dansmästaren. The model uses calculated solar production for each hour of a year and stochastic data for the consumption. The net production is later used for Case 1 and Case 2.

In order to predict the H2 volume that needs to be stored in order for the storage to handle

the 30-150 larges power peaks, calculations with the battery packs and the specification

needed for the whole 30-150 that represent the power shortage in the local distribution grid. The calculation in MATLAB of the H2 storage size can be expressed as

storagesize = (BC

· ERate

+ BC

) · (F C

/F C

)) · peaks · scale (3) where BC

and BC

are the capacities of the battery packs, ERate

is the E-rate of the Greenrock battery pack, F C

is the power production of the fuel cell, F C

is the H2 consumption of the fuel cell and scale is the percentage which scales the size of the storage to have the right dimension. The scale parameter is needed due to the model being based on off-the-shelf components, which cannot be customized for these specific systems. The scale parameter is chosen through multiple simulations where the parameter is given different percentage values. Different storage sizes and number of peaks were used to find the optimal percentage to use for the scaling and selected the parameter value. The percentage which resulted in the least amount of H2 left during the end of the year, which means that almost all H2 has been used is the percentage used for the rest of the simulations which produced the results for this report. The E-rate for the Nilar battery pack is again high enough to not be included in the calculation. It is also assumed that both the electrolyser and fuel cell use the PEM-technology and can handle extensive start and stops without being affected. All parameteres except scale, can be seen in Figure 7, which illustrates the power flow in the Simulink model for the H2 storage. The scale parameter is included in the MATLAB model and is therefore not visible in Figure 7.

Figure 7. Shows the part of the H2 storage in the Simulink-Model. The feedback loop

represents the initial state of charge of the H2 storage, where the H2 volume from the

previous hour is used as a reference to the next hour.

In order to obtain an understanding of how the system affects the economy of Brand- mästaren, the economical analysis is based on the Simulink model from the the previous research made for Dansmästaren. This model uses existing and stochastic predicted elec- tricity prices and simulates the gross cost of all consumption, which is if no energy storage or PV system exists, gross earnings which is the excess PV power sold to the grid and the gross cost from the grid which is the actually purchased power from the power grid.

The total savings and earnings in SEK are then calculated by adding the gross cost of all consumption together with the gross sales earnings before subtracting the gross cost from the grid. The economical model is not based on the cost of the system, only electricity prices are included. The economical model can be seen below in Figure 8.

Figure 8. Illustrates the economical model in Simulink that uses the time step, over- production, electricity purchased from the power grid and the electricity consumption as input parameters to simulate and calculate the savings and earnings of the system at a given time.

4.5 Case 2 - H2 storage in the MW range

In this case the intention is to investigate what the outcome would be if all surplus energy from the PV production during summer would go directly to fill up the H2 storage. This means that the battery packs will be charged with power directly from the grid during the night. The 150 hours which represents the biggest consumption peaks will be used for simulation 1, and if the storage is to small, more peaks will be used for simulation 2.

The system design can be seen in Figure 10 below.

Figure 9. Illustrates a schematic and simplified system design and its power flow.

The reason why Case 2 only has two scenarios compared to Case 1 is because the purpose is not to find the best dimension of a H2 storage. It is to get an understanding of how a large H2 storage system can operate and what value it can bring. This is to distinguish what benefits would be created and then added to the mobility houses and to the city of Uppsala. However in this case compared to Case 1, both battery packs and the H2 storage will be used simultaneously to cut power peaks. It means that the H2 storage will not be used to charge the battery packs as in Case 1.

Possibilities that are interesting in this case are to check for how long the mobility house manages to be off-grid if Uppsala municipality would have a power outage. If Uppsala reaches the capability limit, could the mobility house offer flexibility in terms of discon- necting itself to release a certain amount of space from the power grid. What potential does this type of H2 system have for island operation? To understand how big the storage needs to be, the capacity of the fuel cell and the number of power peaks are used. This means that the scale parameter for dimensioning the H2 storage is not needed, due to the storage dimension not being based on battery packs.

The model is also based on the fact that the fuel cell can run on maximum or with

50 % of its rated power. This is based on both theory and the size of the system. It

is reasonable that the fuel cell does not have to run on maximum power output every

hour of its operation. Mainly because of what is previously stated, but also due to the

net production which is shown in Figure 6, indicates that the size of the power peaks

are estimated to have the maximum size of 200 kW. The H2 storage will also be used

cell can be to high at certain times if the production is not regulated. In Case 1, the fuel cells have less capacities and therefore it is not necessary or interesting to include.

Further, it is important to mention that the economical model for this cases is based on the same method as Case 1. This means that if the only existing energy storage is based on battery packs, this storage will be charged with overproduction from the PV system, and not from the power grid. However, when considering a hybrid storage, consisting of H2 storage and battery packs, overproduction will charge the H2 storage and only electricity from the power grid will be used to charge the battery packs, as previously mentioned.

4.6 Summary of Cases and Scenarios

A summary of the all three cases are presented in Table 2, to clarify what each case entails, and what they will contribute with.

Table 2. Summary of cases used in this report and their method.

Case 0 Case 1 Case 2

No simulations 5 simulation scenarios 2 simulation scenarios No battery packs Battery packs charged with

PV and H2 power

Battery packs charged with power from power grid Suggests 4 systems for

Brandmästaren, where two systems (System 1 and Sys- tem 3) are stand-alone sys- tems and the other two (System 2 and System 4) are based on H2 being pur- chased

Suggests 5 systems with 5 different dimensions for Brandmästaren, and compares these dimen- sions with corresponding systems in Dansmästaren

Suggests two systems with

one maximized H2 storage

in the kW dimension and

the other one being in the

MW dimension for Brand-

mästaren

5 Results

This chapter presents the results generated from each simulation case done in the MAT- LAB and Simulink model. Each case is presented separately where all their relevant results. Case 0 gives a summarized result regarding small systems and the economical aspect. Case 1 and Case 2 present components used, the capacity for the H2 storage, the change of the H2 storage for one year together with the economical aspects.

5.1 Case 0 - A small H2 storage

Table 3 to 6 presents a summarised version of the most relevant parameters and compo- nents which are used to form a small H2 energy storage. Table 3 and Table 5 present information about stand alone H2 storage systems, while Table 4 and Table 6 present information about H2 storage systems where H2 is purchased directly from distributors.

More information about components and technical specifications can be found in Ap- pendix 1.

Table 3. Information about components and parameters for System 1.

Parameter Electrolyser

AEM

Steal Cylinders Fuel Cell

Fuel Consumption 2.4 kW 10-55 l/cylinder 1.56 Nm

/h

Production 0.2 Nm

/h - 2 kW

# 1 5 1

Cost 130 000 SEK 367 SEK / m3 75 000 SEK

Volume - 47 Nm

-

Total Price: < 300 000 SEK - -

Table 4. Information about components and parameters for System 2.

Parameter Steal Cylinders Fuel Cell

Fuel Consumption 20 l/cylinder 1.56 Nm

/h

Production - 2 kW

# 12 1

Cost 588 SEK/cylinder 75 000 SEK

Rent per cylinder 2225 SEK / year -

Volume 47 Nm

-

System 1 and System 2 use the same fuel cell and therefore contribute to the same amount of power. However, the price difference of both systems contributes to different costs of same amount of power being produced and used for peak shaving. System 1 has a initial price of maxium 300 000 SEK and produces 2 kW per hour, which results in 60 kWh for 30 peaks. This results in one kWh having the price of 5000 SEK, while System 2 has a kWh price of 2167 SEK.

Table 5. Information about components and parameters for System 3.

Parameter Electrolyser

PEM

Steal Cylinders Fuel Cell

Fuel Consumption 1.7 kW 10-55 l/Cylinder 3.9 Nm

/h

Production 0.27 Nm

/h - 5 kW

# 1 11 1

Cost < 300 000 SEK 367 SEK / m3 180 000 SEK

Volume - 117 Nm

-

Total Price: 500 000 SEK - -

Table 6. Information about components and parameters for System 4.

Parameter Steal Cylinders Fuel Cell

Fuel Consumption 50 l/cylinder 3.9 Nm

/h

Production - 5 kW

# 12 1

Cost 864 SEK/cylinder 180 000 SEK

Rent per cylinder 2225 SEK /year -

Volume 117 Nm

-

Total Price: 220 000 SEK -

System 3 and 4 uses a fuel cell with a capacity of 5kW, which results in 150 kWh for 30

power peaks. This type of system contributes to a higher power output, but also requires

more H2 and therefore a bigger H2 energy storage, than System 1 and 2. The power

output is the same for System 3 and 4, but the initial price is not. System 3 has a initial

price of 500 000 SEK, while System 4 has a initial price of 220 000 SEK. This contributes

to System 3 having a kWh price of 4 274 SEK and System 4 has a kWh price of 1 880

5.2 Case 1 Simulations - A H2 storage in different dimensions

Figure 10, 12, 14, 16 and 18 illustrate different dimensions of H2 storages for Dans- mästaren and Brandmästaren, together with their curve of both charge and discharge of H2. Each figure is simulated for the year 2023 and the horizontal axis represents each hour of the year from January to December. The vertical axis represents the volume of the H2 stored in the storage during each hour. Each result of the H2 storage is based on two simulations, where the storage is initially empty and being charged during the summer period (late March to the beginning of October). After one simulation, the H2 volume left in the storage is used as the initial volume for the next simulation which results in a year representation of the H2 storage. The results of each simulation scenario is also presented with figures that illustrate the amount of overproduction from the PV system used to charged the H2 storage in Brandmästaren, together with tables that show dimensions of key components used and finally a economical calculation table. These fig- ures are Figure 11, 13, 15, 17 and 19. The vertical axis represents the amount of available energy and the energy used for the H2 storage. Lastly, overproudction is also represented by the parameter H2 power in tables for Brandmästaren. It represent the amount of PV power needed to produce the H2 volume in the H2 storages.

Table 7. The dimensions of both PV system and battery packs for Dansmästaren and Brandmästaren which are used in all simulations for Case 1.

Parameter Dansmästaren Brandmästaren

Nilar Battery Pack Capacity = 23 kWh E-rate = 3

Capacity = 46 kWh E-rate = 3 Greenrock Battery Pack Capacity = 60 kWh

E-rate = 0.25

Capacity = 120 kWh E-rate = 0.25 PV System Installed power = 76 kW Installed power = 150 kW

5.2.1 Simulation 1 - 30 power peaks

Table 8. Parameters for Dansmästaren.

Parameter Value

Electrolyser 1 Nm

/h & 5.5 kW

Fuel Cell 4.2 Nm

/h & 5 kW

H2 storage size 670 Nm

Table 9. Parameters for Brandmästaren.

Parameter Value

Electrolyser 2 Nm

/h & 11 kW

Fuel Cell 4.2 Nm

/h & 5 kW

H2 storage size 1341 Nm

Scale 0.7

H2 power 6.89 MW

Figure 10. H2 storages where the positive incline shows the storages being charged during summer and the negative incline shows the discharge due to charging of battery packs. The 30 biggest power peaks together with their size and the time they occur is also showed.

As seen in Figure 10, one of the 30 biggest power peaks occurs around hour 2200 which

is during summer and when the H2 storage is being charged with power from the PV

over dimensioned compared to the H2 storage of Dansmästaren. Furthermode the size of each power peaks varies between 165 and 205 kW, and the battery packs can contribute with around 65 kW for one hour. The battery packs contribute with less power than their initial capacity indicate, due to the E-rate of the Greenrock battery pack being less than one and the existing energy losses of both battery packs. This means that the H2 storages are contributing with shaving 32-39% of the 30 largest power peaks together with the battery packs.

Figure 11. Illustrates the portion of all overproduction from PV, which is used for the H2 storage in Brandmästaren, or sold back to the power grid. The overproduction is all surplus energy available after the power consumption is met in Brandmästaren and both battery packs are fully charged.

In Figure 11, it is shown that the H2 storage is fully charged before the summer period

has ended. This can also be seen in Figure 12, where no overproduction is used for the

H2 storage after July (Month 7). The amount of overproduction used to produce H2,

compare to the amount of overproduction available is relatively low. This is self explained

Table 10. The economical results from two simulations, one with H2 storage and battery packs (Hybrid) and one with only battery packs.

Storage type Gross cost of all cons

Gross cost from grid

Gross cost earn- ings

Hybrid 316 800 SEK 195 400 SEK 23 700 SEK

Battery packs 316 800 SEK 203 000 SEK 27 950 SEK

Total savings 3400 SEK

The economical results extracted from the economical model in Simulink can be seen in Table 10. The result indicate that it is a small difference between the simulation with only battery packs compared to the one with both H2 storage and battery packs. When not considering the initial cost for the H2 storage system, the total savings, given from the simulation of the hybrid storage, are 3400 SEK. The inital cost for the H2 storage system is approximately 1 900 000 SEK, which makes the total savings from the economical model insignificant for this scenario.

5.2.2 Simulation 2 - 60 power peaks

Table 11. Parameters for Dansmästaren.

Parameter Value

Electrolyser 2 Nm

/h & 11 kW

Fuel Cell 4.2 Nm

/h & 5 kW

H2 storage size 1149 Nm

Scale 0.6

Table 12. Parameters for Brandmästaren.

Parameter Value

Electrolyser 3 Nm

/h & 16.5 kW

Fuel Cell 4.2 Nm

/h & 5 kW

H2 storage size 2298 Nm

Scale 0.6

Figure 12. H2 storages where the positive incline shows the storages being charged during summer and the negative incline shows the discharge due to charging of battery packs. The 60 biggest power peaks together with their size and the time they occur is also showed.

Figure 12 indicates that one of the 60 biggest power peaks occurs during summer (time

2600) when the H2 storage is being charged with power from the PV system. This

means that one of 60 peaks are not being cut. It is also shown that the H2 storage for

Brandmästaren is fully charged around hour 5300, which is 900 hours before the summer

period has ended and makes the storage slightly over dimensioned compared to the H2

storage of Dansmästaren. Further, the size of the power peaks varies between 158 and

212kW, which means that the H2 storages is contributing with shaving 31-41% of the 60

largest power peaks together with the battery packs.