IN
DEGREE PROJECT INFORMATION AND COMMUNICATION TECHNOLOGY,
SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2021,
Energy Harvesting for Health Monitoring Balises
Analytical study
LEIRE CARRERAS OROBENGOA
KTH ROYAL INSTITUTE OF TECHNOLOGY
SCHOOL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCE
Energy Harvesting for Health Monitoring Balises
Analytical study
Leire Carreras Orobengoa
4/8/2021
Master’s Thesis
Examiner
Saúl Rodriguez Dueñas
Academic adviser Ana Rusu
Industrial adviser Gaël Chosson
KTH Royal Institute of Technology
School of Electrical Engineering and Computer Science (EECS) Department of Electrical Energy
SE-100 44 Stockholm, Sweden
Abstract
Abstract
Balises are transponders installed in railways. These devices are nowadays powered by means of a radiofrequency signal emitted by each running train that passes above them. It is only during this moment that the health state of the balises is checked. Hence, there is currently no way to check whether the balises are properly working before the train passes by them. With the aim of executing regular health checks to the balises, an additional source of energy to monitor the balises should be contemplated. Energy harvesting is observed as a suitable solution for this issue. However, a lack of suitability studies is contemplated which englobes the available energy harvesting solutions in railway environments. Therefore, this thesis presents an exploratory work that uses the health monitoring of the balises as a test case for the study of the compatibility of different energy harvesters in diverse railway environments. Hazardous and remote areas are identified as locations of interest for the implementation of the technology, as cabling in those areas is costly and health checks to balises that are not constantly active are of outmost interest. Thus, the addition of wireless communication networks is also studied, due to the need of sending the information obtained in the health checks to monitoring control units. After an initial research study is performed, requirements in railway environments are defined, and three railway scenarios are selected for a suitability study. Then, the investigated energy harvesters and wireless communication networks are compared analytically, and possible technologies for the storage of the harvested energy are presented. It is found that no energy harvester exists that suits all the environments and shows a sufficient power output to make constant checks in remote areas. Nonetheless, piezoelectric and wind harvesters are proposed, because of the commercial availability of the former and the potential of the latter. In terms of wireless communication networks, LoRaWAN shows a low power consumption, while it offers a wide communication range and global coverage. It is, therefore, proposed as the best framework for the wireless communication networks.
Keywords
Balise, Energy harvesting, Railway signaling, Wireless communication networks
Sammanfattning
Sammanfattning
Baliser är transpondrar installerade i järnvägar. Dessa enheter drivs numera med hjälp av en radiofrekvenssignal som sänds ut av varje tåg som passerar ovanför baliserna. Det är först i detta ögonblick som balises hälsotillstånd kontrolleras. Därför finns det för närvarande inget sätt att kontrollera om baliserna fungerar korrekt innan tåget passerar dem. I syfte att utföra regelbundna hälsokontroller på baliserna bör en ytterligare kraftkälla för att övervaka baliserna övervägas. Energy harvesting observeras som en lämplig lösning för denna fråga. Det råder dock brist på lämplighetsstudier som förenar de tillgängliga energy harvesting lösningarna i järnvägsmiljöer.
Därför presenterar denna avhandling ett undersökande arbete som använder hälsoövervakningen av baliserna som ett testfall för att studera kompatibiliteten hos olika energiskördare i olika järnvägsmiljöer. Farliga och avlägsna områden identifieras som platser av intresse för genomförandet av tekniken, eftersom kablar i dessa områden är kostsamma och hälsokontroller till baliser som inte ständigt är aktiva är av yttersta intresse. Således studeras också tillägget av trådlösa kommunikationsnätverk på grund av behovet av att skicka den information som erhållits vid hälsokontrollerna till övervakningskontrollområdena. Efter att en inledande forskningsstudie genomförts definieras krav i järnvägsmiljöer och tre järnvägsscenarier väljs ut för en lämplighetsstudie. Sedan jämförs de undersökta energiskördarna och trådlösa kommunikationsnätverk analytiskt, och eventuell teknik för lagring av den skördade energin presenteras. Det konstateras att det inte finns någon energiskördare som passar alla miljöer och visar en tillräckligt effekt för att göra konstanta kontroller i avlägsna områden. Ändå föreslås piezoelektriska och vindskördare på grund av den förstnämnda kommersiella tillgänglighet och den senare potentialen. När det gäller trådlösa kommunikationsnätverk visar LoRaWAN en låg strömförbrukning, medan det erbjuder ett brett kommunikationssortiment och global täckning. Det föreslås därför som den bästa ramen för de trådlösa kommunikationsnäten.
Nyckelord
Balise, Energy harvesting, Railway signaling, Wireless communication networks
Acknowledgments
Acknowledgments
I would like to thank Bombardier Transportation for giving me the opportunity to learn about railway signaling and for all the shown support and given motivation. Being supervised by Gaël Chosson was a very fruitful experience, as he has been of paramount help by the provision of knowledge and different perspectives. I would also like to give a special mention to Peter Herold and Julián García, who have always been keen to help and share knowledge and wise words. Finally, I would like to show my gratitude to the academic supervisor of this thesis, Ana Rusu, and the examiner, Saúl Rodriguez.
Your feedback has been very valuable for my work.
Stockholm, April 2021
Leire Carreras Orobengoa
Table of contents
Table of contents
1 Introduction ... 1
1.1 Background ... 1
1.2 Problem ... 1
1.3 Purpose ... 2
1.4 Goals ... 2
1.5 Research methodology ... 2
1.6 Delimitations ... 2
1.7 Structure of the thesis ... 3
2 Background ... 5
2.1 Railway signaling ... 5
2.1.1 Eurobalise ... 5
2.1.2 ERTMS application levels ... 6
2.2 Energy harvesting ... 8
2.2.1 Thermoelectric ... 8
2.2.2 Mechanical motion ... 9
2.2.3 Solar ... 11
2.2.4 Radio frequency ... 11
2.3 Energy harvesting in railway environments ... 12
2.3.1 Thermoelectric ... 12
2.3.2 Piezoelectric ... 12
2.3.3 Electromagnetic ... 13
2.3.4 Solar ... 14
2.4 Wireless communication networks ... 15
2.4.1 WPAN ... 16
2.4.2 WLAN ... 17
2.4.3 WMAN ... 17
2.4.4 WWAN ... 17
3 Research Methodology... 19
3.1 Research methodology ... 19
3.2 Process diagram ... 19
3.3 Data collection ... 20
3.4 Reliability and validity of the collected data ... 20
3.4.1 Validity ... 20
3.4.2 Dependability ... 20
3.4.3 Confirmability ... 20
3.4.4 Transferability ... 20
3.4.5 Ethics ... 21
3.5 Planned data analysis ... 21
3.6 Evaluation framework ... 21
4 Evaluation ... 22
4.1 Railway environment ... 22
4.2 Energy harvesting methods ... 24
4.2.1 Meteorological compatibility ... 25
4.2.2 Lifetime and maintenance needs ... 26
4.2.3 Development level ... 27
Table of contents | viii
4.2.4 Radio frequency ... 27
4.2.5 Cost ... 27
4.2.6 Comparison ... 28
4.3 Energy storage systems ... 28
4.4 Wireless communication networks ... 30
5 Results ... 33
5.1 Energy harvesting ... 33
5.2 Wireless communication networks ... 35
5.3 Reliability analysis ... 35
6 Conclusions and Future Work ... 37
6.1 Conclusions ... 37
6.2 Limitations ... 38
6.3 Future work ... 39
References ... 41
APPENDIX A : Interview questions ... 46
List of Figures
List of Figures
Figure 1: ERTMS levels. a) ERTMS level 1, b) ERTMS level 2, c) ERTMS level 3 [14]
is licensed under CC BY 3.0 ... 7
Figure 2: Seebeck effect on thermoelectric EH ... 8
Figure 3: Cantilever piezoelectric EHer ... 9
Figure 4: Electromagnetic EHer ... 10
Figure 5: Photovoltaic EHer ... 11
Figure 6. RF EHer ... 12
Figure 7: Wireless access geographic coverage. RFID, radio frequency identification; NFC, near field communication; NB-IoT, Narrowband Internet of Things; LTE, long term evolution; GSM, global system for mobile communications; MTC, machine type communication; SUN, smart utility network; LE, [67] is licensed under CC BY 4.0 ... 16
Figure 8: Process diagram of the thesis ... 19
Figure 9. Identified energy storage solutions, adapted from [82] ... 29
Figure 10: Sigfox global coverage map, adapted from [88] ... 31
Figure 11: LoRa global coverage map, adapted from [89]... 32
List of Tables
List of Tables
Table 1. Meteorological data from Mexico ... 24
Table 2. Meteorological data from Sweden ... 24
Table 3. Meteorological data from Spain ... 24
Table 4. EHers comparison table ... 29
List of Acronyms and Abbreviations
List of Acronyms and Abbreviations
ATP Automatic Train Protection BTM Balise Transmission Module
EH Energy Harvesting
ERTMS European Rail Traffic Management System ETCS European Train Control System
GSM-R Global System for Mobile Communications – Railways IEEE Institute of Electrical and Electronics Engineers IoT Internet of Things
IrDA Infrared Data Association ISM Industrial Scientific and Medical LEU Lineside Electronic Unit
LPWAN Low Power Wide-Area Networks LTE Long Term Evolution
LTE-M Long Term Evolution for Machines NB-IoT Narrowband Internet of Things
RF Radio Frequency
UWB Ultra-Wideband
WBAN Wireless Body-Area Networks WCN Wireless Communication Networks
WiMAX Worldwide Interoperability for Microwave Access WLAN Wireless Local-Area Networks
WMAN Wireless Metropolitan-Area Networks WPAN Wireless Personal-Area Networks WWAN Wireless Wide-Area Networks
Introduction | 1
1 Introduction
A balise is a transponder used as a beacon in numerous railways worldwide [1]. It is part of the railway signaling system which ensures a correct train traffic. Because of its passive element nature, it is only awake when a train containing a balise transmission module (BTM) with its corresponding antenna passes over it. One of the major characteristics of a balise is its need to be reliable, as human lives may be endangered if it malfunctions. Therefore, there is a constant research interest on the improvement of the robustness and reliability of balises in terms of availability, precision and durability.
1.1 Background
Balises are wayside products which are typically cable-free. They are part of the automatic train protection (ATP) system, which performs continuous supervision of the railways and spot communication. The European rail traffic management system (ERTMS) and the European train control system (ETCS) are projects that aim to make rail transport safer and more competitive. These projects include balises as railway signaling components.
Because of the small amount of information that needs to be sent to the train, the balises are passive and are powered via radiofrequency when a BTM’s antenna passes over them. However, recently different ways of powering railway signaling systems have been investigated. Energy harvesting (EH) is an alternative powering method which transforms ambient sources to useful energy.
1.2 Problem
It is crucial to ensure the reliability of systems which may endanger human lives. This is the case of transportation vehicles such as trains. In the case of these, wayside signaling directs railroad traffic so that they keep a safe distance from each other. One of the devices used for this purpose is the balise.
This product must unmistakably provide correct information to the train control system when a train passes on top of it. For this reason, balises are robust, redundant and maintenance free for 30 years [1].
However, for a superior performance, it is convenient to monitor the health of the balises to detect any possible fault in them before the communication with the train occurs. With the intention of the ERTMS/ETCS projects to make the European railways safer by 2030 [2], it is a good moment to increase the reliability of the balises by health monitoring them.
Moreover, cables are intended to be avoided, when possible, in railway environments because of their high installation and maintenance costs, and the attractiveness of their copper for thieves.
Additionally, maintenance or repairing operations of wired systems often require the railway traffic in the area to be stopped. For this matter, the utilization of ambient sources; that is to say, EH, is preferred, as in railway systems a large amount of potentially wasted energy is created [3], and this could instead be converted to useful electricity. Also, a high interest on the research of the exploitation of otherwise wasted energy has taken place during the past two decades [4].
Additionally, the communication of the balises with a control unit should also be wireless, and for this reason a wireless communication network (WCN) should be implemented. This WCN needs to be powered, and following the same reasoning from the previous paragraph, this should also be powered by ambient sources to avoid copper.
Consequently, the focus of this thesis is summarized in this question: is there an EH solution for monitoring the health of balises installed in railways?
2 | Introduction
1.3 Purpose
The thesis discusses the possibility for the incorporation of an energy harvester (EHer) in different railway environments for the health monitoring of balises. With this purpose, the thesis also describes the conditions related to the geographical location of the balises and it considers WCN as an addition that enables the transmission of the information from the health checks to a control unit.
Because of the current research interest on EH, this thesis aims to aggregate knowledge to the science. This will be achieved by the presentation of outcomes in this topic, obtained from the consideration of a specific test-case. Additionally, because WCN are considered as the complement for the communication, this thesis analyses the complete setup of the balise environment. Therefore, it benefits the researchers or engineers implicated in this topic, together with the railway industry.
Due to the sustainable nature of the concept of EH, the thesis commits to pursue the seventh goal from the United Nations Sustainable Development Goals, which aims to achieve affordable and clean energy by 2030 [5] by the consideration of the materials and procedures which are used for the manufacturing and installation of the harvesters.
1.4 Goals
The goal of this thesis is to analyze the available EH solutions and evaluate whether there is a possibility to include one for the health monitoring of the wayside balises. This has been divided into the following four sub-goals:
1. Conduct a research on currently available EH solutions in the market.
2. Determine the assumed requirements for the proposal of a WCN and EH combination.
3. Estimate the energy budget of standard WCN.
4. Perform an adaptability study of the EHers to different railway environments to understand whether there is a solution that would work at any railway equipped with balises.
Hence, the expected deliverable from this thesis is the proposal of a suitable WCN and EH combination related to each considered railway environment.
1.5 Research methodology
The analytical nature of this thesis makes it a qualitative research [6]. The proposed solution will be based on the analysis of the related work, and the determined requirements and assumptions.
Therefore, interpretivism is the philosophical assumption. The objectives are achieved by the exploration of existing research work and the study of existing products in different environments.
Hence, the research approach is inductive [7].
1.6 Delimitations
Because of the great effort and investment that is required to make tests in railways, this thesis does not include any field or laboratory test. Instead, available theoretical data and assumptions are adopted for the proposal of the solution. Also, as the focus of the research is to check the suitability of an EH technology in the railways, a complete cost and business analysis is not performed. Similarly, an extended research on available WCN is not completed. Production related aspects, such as the exact health monitoring setup, are not considered. Thus, the different EH solutions, together with some of the available WCN, are compared in a theoretical basis, in order to determine the suitability of the EH and WCN in the different environments chosen for analysis.
Introduction | 3
3 1.7 Structure of the thesis
• Chapter 2 introduces relevant background information about balises and railways, EH technologies in railways and WCN.
• Chapter 3 presents the selected research method and methodologies that are followed throughout the thesis in order to solve the identified problem.
• Chapter 4 consists of the explanation of the railway environments that are considered and the requirements they imply and the evaluation of the presented EH and WCN technologies in regards of workability in the chosen environments.
• Chapter 5 presents the obtained results, which are derived from the extensive analysis performed in the previous chapter.
• Chapter 6 concludes the report by the provision of an overall summary and conclusion of the thesis, which combines advantages and limitations and suggestions for future work.
Background | 5
2 Background
This section provides relevant information about railway signaling, WCN and EH, to familiarize the reader with the main concepts considered and discussed in this thesis. Furthermore, this chapter presents the related work on railway EH and the research approaches that have been adopted in previous literature.
2.1 Railway signaling
Continuous advancement towards the safety on the railways has been accomplished since the first passenger-carrying railway was inaugurated in 1830. Today, it takes about 2 km to stop a train that travels at a speed of 200 km/h. Even back in the 19th century, when trains could not reach higher speeds than 80 km/h, more than a kilometer was required to stop a train from its maximum velocity [8]. Several accidents occurred due to the lack of capability of the drivers to detect an upcoming obstacle in time. Hence, railway signaling was introduced in the form of a policeman who made a hand or flag signal based on a time interval since the last train passed. The signaling system has constantly developed ever since.
With the aim of ensuring the safe activity of the trains, railway signaling systems supervise and control the railway traffic [9]. The European Union adopted the ERTMS a standard signaling, communication and control system regulation so as to have a compatible rail frame throughout Europe. Moreover, many non-European countries have shown interest in the adoption of ERTMS.
Therefore, this railway safety system, which includes on-board and wayside systems, is considered in this thesis. The information received on-board may be intermittent, as it may be collected from the wayside products, which are located either in specific points of the railway, or received continuously, since the information can be exchanged by radio signals. A combination of both information transmissions is also applied [10]. For this reason, the ERTMS incorporates two main components, the global system for mobile communications – railways (GSM-R) and the ETCS.
GSM-R is a radio system for the communication between the train with traffic control points and devices located in the track. It uses frequencies that are reserved for railway application for the train and track to communicate by voice and data commands.
On the other hand, ETCS is a control standard for trains, which comprises on-board and wayside transmission units for a safe and trustable communication between the track and train [10], [12]. The on-board transmission unit consists of an antenna unit which is integrated with the transmission functionality of the overall on-board ATP equipment and a BTM function. The on-board system calculates and supervises the maximum train speed and direction in each section continuously based on the received information and observes the actions executed by the driver. The information is received uniquely from wayside balises or as a combination from balises and GSM-R, depending on the ERTMS level. In case the driver does not break in a dangerous situation, the emergency brakes actuate automatically [10]. The wayside transmission units and ERTMS levels are further discussed in the following subsections 2.1.1 and 2.1.2.
2.1.1 Eurobalise
Wayside units are signaling systems which are installed in the railway. These are composed of lineside electronic units (LEUs) and other external equipment, as light signals, track train detection units or balises. A LEU is the interface between the interlocking and the balises. It checks the state of the interlocking and sends the proper information in the form of telegrams to the balises in accordance with the lineside signaling. Balises are wayside ATP units based on the magnetic transponder
6 | Background
technology. More specifically, the Eurobalise is contemplated in this thesis, as it is the balise variant that complies the ETCS regulations, which are applied in the ERTMS.
The Eurobalise is a passive device, as it does not require any electric supply. Instead, it is energized by the BTM’s antenna located under the train which energizes the balise when it passes over it. This is achieved by means of the 27.095 MHz ±5 kHz magnetic field used for tele-powering.
On the other hand, a reserved frequency band of 4.234 MHz ±1 MHz is used for the information transmission from the balise to the train [13]. The balise awakens when the train passes on top of it and the distance between the BTM antenna and the balise is smaller than 1.3 m. If this condition is met and there are no failures, the transmission of the required information is accomplished. An information telegram in an Eurobalise is either 341 bits or 1023 bits long and it is transmitted cyclically while the BTM powers the Eurobalise. Balises must be installed at least in pairs, to ensure the proper direction of the train. If a big amount of data is shared, more balises may be installed nearby.
In respect of the size, a standard size Eurobalise measures 490×360 mm and can transmit a 1023- bit telegram for any train with a speed of up to 500 km/h. On the contrary, small sized Eurobalises measure 390×200 mm and can only send 1023 bit telegrams up to a velocity of 300 km/h. For a train that passes at higher speeds, and up to 500 km/h, short telegrams of 341 bits are transmitted [13].
The telegrams contain information about the infrastructure such as location or speed limits. They provide fixed information, such as location, or controlled data, which is modified by the wayside signaling system.
With regard to the information provided by the Eurobalises, this depends on the type of balise.
There are two types of Eurobalises, fixed and controlled. The fixed balise is programmed to contain steady information which is transmitted to every train in the form of a telegram. This information typically includes the location, any existing speed restriction, and the structure of the line, as approaching curves or gradients. The controlled balise, for its part, is reconfigurable by the LEU and it superimposes the conventional signaling system when needed with the dynamic feedback that the LEU gets either from the signaling control tower or the lineside railway signal [11].
The ambitious mechanical and electronic reliability requisites that the Eurobalises need to obey makes them robust and reliable enough to have an operational lifetime of 20 years in the case of the controlled balises and 30 years in the case of fixed balises [11]. This means that these devices are not health monitored during prolonged periods of time, other than the checks compassed when a train passes over them.
2.1.2 ERTMS application levels
The ERTMS application levels are defined to differentiate the particular ways of interacting between the railway and the train. They are associated to the way the wayside information reaches the on- board units, to the utilized wayside equipment and to which functions are processed on-board and which in the wayside. ERTMS/ETCS is configured to operate in one of the listed levels [14]. Levels 1, 2 and 3 are depicted in a simplified image in Figure 1:
• Level STM is used when trains are equipped with ETCS but run in lines equipped with national train control and speed supervision systems. The national train control system produces the train control information and transmits it to the train by the national communication system. No ETCS is used in this level, except for the announcement of a level modification. Thus, balises are still read if they are located in the railway, but only specific commands are decoded.
Background | 7
7
• Level 0 is used when trains are equipped with ETCS, but the lines are either not equipped with any train control system or they cannot use this supervision for the operation at the moment.
• Level 1 is used when the railway information is sent to the train from a fixed or controlled Eurobalise, which means that intermittent communication is supplied. The LEU included in this level obtains information about the traffic and converts it to the telegram that is later transmitted to the onboard system by the balises. The balises have a maximum separation of 5 km from each other, with 3 km as the nominal distance. The aim of the information provided by the balises is to inform about the conditions of the railway ahead, in order to ensure a safe velocity and operation. The onboard system is in charge of calculating the required speed in accordance with the received information. There may also be an in-fill message that is sent by the GSM-R communication protocol so as to inform the train prior to the arrival at the signal. This way, the train can accelerate before it receives the information from the signal, as it is enough to receive it from the in-fill system, which permits a continuous information system prior to the reception of the signal. Lineside signals are necessary and train detection is performed by trackside equipment out of the scope of ERTMS.
• Level 2 implies uninterrupted supervision of train movement and constant communication.
This communication is provided by GSM-R and lineside signals are optional in this case.
Train detection is also performed by trackside equipment out of the scope of ERTMS. No trackside signaling is required. Train position is known by the balise information, while GSM- R is utilized for providing movement authorities. Train detection is performed by trackside systems complying ERTMS.
• Level 3 is also used for a continuous train supervision and continuous communication between train and railway. For this to happen, bi-directional radio communication is used to transfer all the information. However, as a differentiation from level 2, the train location and integrity are performed by devices in the scope of ERTMS, so there is no need for wayside equipment.
Figure 1: ERTMS levels. a) ERTMS level 1, b) ERTMS level 2, c) ERTMS level 3 [14] is licensed under CC BY 3.0
8 | Background
2.2 Energy harvesting
There is a need for environmentally friendly sources to fill the increasing demand for energy.
Moreover, it is sometimes a very arduous work to get power from a grid in remote areas. Hence, the interest in EH during the last decade has grown largely. By definition, EH is the action of obtaining the desired form of energy from ambient sources that would otherwise be wasted [15]. This is achieved by the transformation of energy in the surroundings of the system to be powered. In the following paragraphs, different approaches with EH derived from heat or thermal gradients, mechanical motion, light and radio frequency (RF) are introduced.
2.2.1 Thermoelectric
An electrical current is produced following the Seebeck effect, which means that the temperature difference between two electrical conductors creates a voltage difference [16], [17]. Electrons from the conductor with a higher temperature, flow towards the one with a lower temperature when heat is applied to one of the conductors. Therefore, there is a flow of direct current in the circuit. The effect can also be inversed, as a current may be used to produce a gradient in temperature, as can be seen in Figure 2.
The thermocoupler is composed by two opposite Seebeck coefficient materials, which are united in their ends. (1) shows how the Seebeck coefficients 𝛼𝐴and 𝛼𝐵 and the highest and lowest temperatures 𝑇ℎ and 𝑇𝑐 respectively, lead to a voltage difference 𝑉. Also, the obtained power is defined by the external load resistance 𝑅𝐿, the internal resistance 𝑅𝑇𝐸𝐺, and 𝑉, as shown in (2) [18].
𝑉 = (𝛼𝐴− 𝛼𝐵) ∙ (𝑇ℎ− 𝑇𝑐) (1)
𝑃 = 𝑉2 𝑅𝐿 (𝑅𝐿+ 𝑅𝑇𝐸𝐺)2
(2)
Thermoelectric devices are compact, which means their weight and size is reduced, they do not contain any part that moves, they are highly reliable, and are electrically safe as they are not exposed to electrical noises. However, their conversion efficiency is not high, as only 5-10 % of successful energy conversion is achieved, with a common power density of 60 µW/cm2. The highest recorded harvest in the previous research work is 460 µW with 17 mm long thermocouples and a temperature variation of 193 ºC [19]. However, even if it is small in comparison to traditional powering systems, this energy may be sufficient for small systems, which have a low power budget. Moreover, there is potential in this EH solution, as recent studies focus on the design and fabrication of thermocouples with segmented materials in order to obtain the highest impedance at the available heat source temperature [20].
Figure 2: Seebeck effect on thermoelectric EH
Background | 9
9 2.2.2 Mechanical motion
Vibration or mechanical motion is another source that has been exploited for the obtention of power.
This occurs by means of piezoelectric, electromagnetic or electrostatic materials.
2.2.2.1 Piezoelectric and electrostatic
A variation in a crystalline material is what generates the power in this case. Specifically, a piezoelectric material undergoes a mechanical strain which produces a charge and an electric field, as the material contains dipoles. The most typical application is in environments with vibrations. The bigger the vibration intensity, the larger the produced field [21], [22].
Figure 3 shows a typical piezoelectric cantilever with proof mass at its free end. The values of the components on the circuit depend on the characteristics of the piezoelectric material.
The vibrations are caused by several factors, which include an unbalance in a rigid body or the wear and tear of it. The nature of each material is defined by its damping factor and natural frequency. The study of the dynamic behavior of a piezoelectric harvester is usually summarized by a lumped spring-mass system with a single degree of freedom.
The maximum generated power occurs when the system is excited at its natural frequency, and it is expressed by the (3), where 𝑚 is the oscillating mass, Υ the displacement of the amplitude, 𝜔𝑛 the natural frequency of the system and 𝜁𝑇the damping factor.
𝑃𝑚𝑎𝑥=𝑚Υ2𝜔𝑛3
4𝜁𝑇
(3)
It is needed to note that the excitation frequency may be altered due to a change in temperature or other environmental circumstances. Moreover, if the amplitude of the oscillation is too high, the behavior of the system may be non-linear and if this happens the resonance state is not possible.
This EH method has been extensively researched during the past decades [21]. For this reason, new materials with enhanced properties have been introduced, but it is still of interest in research, due to the possibilities of this solution in different fields. A research study showed a piezoelectric EHer that showed an output power of 13.5 mW at 15 Hz [23].
The electrostatic principle is similar to the one of the piezoelectric EHer, as it also converts mechanical energy into electricity [15]. However, the conversion in this case is accomplished by a variable capacitor structure. The relative motion between two plates of a variable capacitor in this structure is what obtains the energy.
There are two different conversion principles for electrostatic EHers. On the one hand, electret- based converters transform mechanical power into electricity directly, by the use of electrets, which are pieces of dielectric materials that carry a permanent electrical charge. On the other hand, there
Figure 3: Cantilever piezoelectric EHer
10 | Background
are electret-free converters which need conversion cycles which consist of charges and discharges of the capacitor [24].
A previous research study showed up to 25 mW output power in a EHer which had a surface of 730 cm2 [25]. On the contrary, considering less spacious solutions, a research study [26] showed a power output of 38 µW for a 4 cm2 EHer. Similar to piezoelectric materials, these have been researched since the 2000s. However, no commercial solutions exist as of today. Moreover, electret stability is strongly linked to environmental conditions as temperature and humidity [24].
2.2.2.2 Electromagnetic
By the electromagnetic induction EH, electrical power is obtained due to an electromagnetic force induced by a conductor which breaks magnetic flux lines [27]. This is based on Faraday’s law, and the effect is used for the generation of large amounts of power in applications such as wind and wave power plants. However, the principle is employed in low-power scenarios, as a coil can generate power from a flux fluctuation.
A permanent magnet generates the magnetic flux, and a mass which oscillates attached to a coil or another magnet generates a fluctuation in the magnetic flux, which induces current in the coil, as seen in Figure 4. The frequency of the magnetic flux linkage determines the voltage induced in the coil [28]. This voltage, 𝑉, is calculated as in the (4), where 𝑁 represents the number of turns of the coil, while ∅ is the flux change and 𝑡 is time [29].
𝑉 = −𝑁𝑑∅
𝑑𝑡
(4)
This equation is adapted for the cases of linear vibration or time-varying magnetic field. In the case of linear motion, a relative movement occurs between the magnet and the coil, whereas in the time-varying magnetic field case, the voltage depends on the angle between the direction of the flux density and the area of the coil.
A load resistance, 𝑅𝐿, is connected to the coil so as to harvest power. To acquire the maximum electric power output, the electromagnetic damping and fluctuation speed need to be maximized. For its part, the magnetic material determines the strength of the magnetic field. A maximum of 10 mW was registered in a frequency range of 20-40 Hz [30].
Figure 4: Electromagnetic EHer
Background | 11
11 2.2.3 Solar
Light, either artificial or natural, is the resource from which the energy is obtained. Three methods exist for the collection, which are solar thermal collectors, concentrating solar power generators and photovoltaic harvesters. The first two are not considered harvesters, due to their size and fields of application. Thus, this section refers to photovoltaic power as the solar EH method [31], [32].
Regarding the EH principle of this solution, materials are doped so as to create a positive-negative (p-n) structure. The n-type silicon gets the electrons while the p-type gives them, making holes. When the sunlight hits the panel, the photons from the light excite the electrons and these become electron- hole pairs, which are then induced to separate because of the internal electric field. Therefore, the holes move to the positive electrode and the electrons to the negative one. A conductor connects the negative and positive electrode and a load in the form of a serial circuit. Thus, an electric current is generated to power the external load, as can be seen in Figure 5.
In an open circuit, the voltage difference appears in the junctions because of the separation of the carriers, and in ideal conditions, when the photo-generated current 𝐼𝑝ℎ is equal to the short-circuit current, open circuit voltage 𝑉𝑂𝐶 is described by the (5).
Here, 𝐾 represents the Boltzmann constant, 𝑇 is the absolute temperature, 𝑞 describes the charge of the electron and 𝐼0 is the current generated by the cell in limited-light conditions.
Their main advantage over other types of EHers is their large efficiency when exposed to direct sunlight with a brightness of up to 120,000 lux, as the values go up to 40 % of the outdoor solar conversion efficiency in this case. This implies a typical power density of 100 mW/cm2 during daytime, which makes it a high-power harvester [33]. However, in low light environments or during nighttime, when the brightness is lower than 1 lux, the efficiency drops, as the photovoltaic harvesters are optimized for the brightness of the sun.Still, this EH method has a bigger potential, as the photons from short and long waves in the entire light spectrum are not absorbed yet. For that, research is performed currently to get nanotechnology which allows to get ultraviolet light spectrum photons. As a main opposing point, the energy is only available when the light source is available, so the application possibilities is limited by this factor, which is not totally predictable.
2.2.4 Radio frequency
This technology obtains energy from wireless microwaves and RF. This is obtained from the radiation emitted by both, dynamic transmitters which work periodically such as Wi-Fi access points, or static sources, which emit radiation in a stable and predictable manner, such as television towers [34]. In
𝑉𝑂𝐶=𝐾𝑇
𝑞 ln (1 +𝐼𝑝ℎ
𝐼0
) (5)
Figure 5: Photovoltaic EHer
12 | Background
this case, RF waves are in the ranges of 3 kHz to 300 GHz, and the power density depends on how far the receiver from the transmitter is, the incident power density, the antenna size and the power conversion efficiency of the device [35]. The typical RF EHer is formed by an antenna which captures the RF signal, an impedance matching circuit to obtain the maximum power from it, a rectifier to convert this power into direct current and a voltage multiplier, which increases the DC voltage level, as seen in Figure 6.
The RF that arrives to the harvester is evaluated in terms of electric field strength 𝐸, and the obtained power density 𝑆, is calculated by the (6), where 𝑍0 is the characteristic free space impedance [35].
𝑆 =𝐸2 𝑍0
(6)
Typically, the efficiency of these devices is 50-70 % and the power density goes as high as 0.1 µW/cm2 in the case of GSM signals and 0.01 µW/cm2 for application such as in Wi-Fi access points.
2.3 Energy harvesting in railway environments
An extensive number of research work has been performed related to EH with the purpose of a safer railway environment together with a reduced maintenance cost in the past decades. While some solutions have been researched to be applied in the vehicles for powering on-board systems, the following subsections introduce the most relevant research studies related to EH in railways.
2.3.1 Thermoelectric
Only one thermoelectric EHer research study for railway applications has been detected [36].
Therefore, the obtained values purely depend on this test environment. The thermal gradient between the soil or underneath the track and the bottom of the rail was identified as a possibility for harvesting.
Two high conductivity thermal pads were used to maintain the thermal gradient, and a temperature variation of up to 30 ºC was calculated by numerical simulations. However, on-field tests showed a difference of only 8 ºC, at which the power output was 5.8 mW with a 7 Ω load. Still, in lab tests showed an output of 316.8 mW due to a temperature contrast of 29.2 ºC and with a 9.6 Ω load, which caused the authors to be optimistic about this technology.
2.3.2 Piezoelectric
With regards to piezoelectric devices aimed to be installed on the track and rail bone, a solution that harvested energy by means of a piezoelectric layer which was glued to the bottom of the track was proposed and developed [37]. The performed field tests of the prototype showed a maximum power output of 53 μW with a 387 kΩ load resistor. An output power potential of up to 1.1 mW was then
Figure 6. RF EHer
Background | 13
13 predicted by the authors in case higher axle-load trains passed above the harvester. This means that energy is only harvested when a vehicle runs on the track section where the piezoelectric material is installed, but it does not need to be tuned to the natural frequency of the oscillator, as it is based on the rail deformation.
Similarly, adhesive piezoelectric patches installed on bridges were analyzed. Although no field tests were performed, simulation work showed that piezoelectric harvesters could obtain 588 µW from a train running at a speed of 120 km/h [38].This research work showed that the output power is related to the train type, speed and axle-load.
With regards to solutions based on a cantilever beam, a cantilever beam fastened to the rail, which harvested 4.9 mW at low frequencies of 5-7 Hz, was proposed [39]. Also, a solution that consisted of a cantilever beam fixed to an external frame from both sides and which included a seismic mass on the middle was developed [40]. The size of the solution was 36 mm × 21 mm, which allowed multiple devices to be placed together for a higher power output. The excitation frequencies in this case were 380-500 Hz, and 380 μW were harvested with a 20 kΩ load resistor and a 1.45 seismic mass.
On the other hand, [41] followed a very different approach, which consists on a stack of piezoelectric modules, connected in parallel in electrical terms, and in series mechanically, to get larger bends. This research work presents a numerical study that calculates the dependency of power, optimal resistor, and output voltage on the vehicle speed. A maximum rms power of 150 μW rms power when the train runs at 320 km/h with an optimal load resistor of 50 kΩ was then predicted.
Instead, for a vehicle running at 110 km/h, the maximum harvested rms power was of 27 μW, with a 180 kΩ optimal load resistor, which means that the same resistor should not be used for optimal results at different vehicle speeds. Moreover, with the same idea of parallel mechanical connection for an increased deflection, [42] presented an array of piezoelectric diaphragms. This study showed 21.4 mW harvested with a 110 kΩ optimal load resistor and excitation frequency of 150 Hz.
2.3.3 Electromagnetic
Related to motion based electromagnetic harvesters, research studies can be divided in geared and vibration harvesters. The former ones convert the linear motion of the vibration from the track and vehicle into a continuous rotative motion of a shaft, by the use of gear drives and mechanisms. The latter is based on the relative linear motion of a magnet and a coil, since one of the elements is typically attached to a vibrating body, whereas the other one is located in a stable ground.
Various similar research studies [43], [44] focused on a geared harvester that did not fulfil the specifications to power the railway lighting system, which required 40 W. With a complex mechanical design based on bearing clutches and rack-pinion transmission, the solution was designed to be mounted on a transversal plate in two sleepers, so as to utilize sleeper and rail vibrations. Numeric simulations show a prediction of 300 W power output when a loaded train passes by at a speed of 96.5 km/h, but the studies show steps that still need to be researched, as the power is not enough in the case of an unloaded train. With the aim of reaching the minimum needed power, even when the vehicle is unloaded, a hydraulic system concept was presented [45], but was not further investigated.
Furthermore, an anchorless system was proposed [46] as an improvement of a previously researched mechanically complex system, which only harvested 1.4 W in the best conditions [47]. In the old version of the system, a rigid bar connects both racks and these are anchored to the ballast and have pinions in contact with them. This way, the vibrations originated by a passing train cause a relative motion between the two racks and the pinions. This causes high friction losses and hence a low efficiency of the system. In the newer proposed version, the deficiencies related to the anchoring of the racks to the ballast are avoided. Instead, the threaded load is integrated in the ballast, connected to a base plate. On field tests for this system show peaks up to 56W and an average power output of 7 W with a 16.7 Ω load.
14 | Background
A higher power output has not been identified, as [48] proposed a system with a nut and ball screws to convert the linear motion of the track into a rotation in the generator, but this showed a maximum output power of 2.24 W with a 8 Ω load. Also, another study contemplated the conversion of the kinetic energy into electricity by means of a brushless DC motor [49]. Experimental results show a peak electromotive force of 6.45 V, but the authors did not specify the power, as they uniquely stated that this is strongly related to the electromotive force.
On the other hand, the research on vibration harvesters in railways and trains was mostly performed for the application in trains. Nonetheless, some research work has been recognized for the application of these EHers in railways. A simple solution was firstly presented, which included a simple inductive coil attached to the rail and a permanent magnet located on the track soil [37]. The deflection of the rail created a relative motion between the coil and magnet, and on-field tests showed average power outputs of 12 mW when a loaded train travelling at 21 km/h passed by. On the other hand, the output was 4 mW in the case of an unloaded vehicle at 18 km/h, both with a load of 7.5 Ω.
Additionally, a resonant device which consists of two fixed magnets and an oscillator with springs, copper coil and a mass block was developed [50]. The system was designed for a specific environment, which is segmented prefabricated assembly bridges, and all the simulations and optimizations were done for this environment. Finally, the calculated peak power was 35.3 W each time a metro-train ran in the rails.
With respect to wind EH, many solutions have been researched and developed for the application in low power autonomous systems. Nevertheless, little impact has been identified in research work for the utilization of wind turbines EH in railway environments.
Although [51] focused on wind energy, the study was mostly concentrated in piezoelectric materials vibrating due to the effect of the wind. Withal, this study performed wind tunnel tests with different configurations of wind turbines, so as to understand which solution was a better fit indoors.
Therefore, lab tests were performed emulating a tunnel environment and with different positions for the turbines, including small and large wind turbines, a four-blades turbine, a three-blades turbine, and a vertical-axis turbine for two shapes of trains, which are bluff or streamlined bodies. The research concluded that tracks where bluff bodies run should have the wind turbines installed under the bottom of the vehicle, whereas high-speed tracks, where streamlined trains pass, should install the turbines in the sides of the track for optimal EH. Due to a stronger boundary effect, the three- blade turbine showed the best results. Yet, the provided data in terms of small rage turbines is that of the vertical-axis turbine, which could generate 110 mW with a load of 56 Ω. For large scale turbines, 5W were achieved with a load of 470 Ω by the horizontal-axis turbine. Both tests were performed with wind blowing at 36 km/h.
Also, recently a track-borne concept for wind EH to provide electrical supply to tunnel health monitoring systems in remote areas was introduced [52]. The vertical-axis turbine was preferred because of its capability to capture wind from any direction and its lower installation costs, with respect to the horizontal-axis one. Different rotations were divided by two one-way bearings, by maintaining or stopping the rotation of one of the rotors. That is, if there were no trains running in the tunnel, natural wind made only the S-to spin. Differently, the H-rotor started to spin faster than the S-rotator due to the piston effect. In this test, a maximum power output of 107.76 mW was achieved, which represents an efficiency of 23.2 %.
2.3.4 Solar
Related to energy obtained from sunlight, several efforts have been made lately for the increase of energy produced by this natural source, and this also applies for its utilization in railways aimed for low power applications.
Background | 15
15 Firstly, solar panels that charged lithium ion batteries with the purpose of detecting anomalies or defects in the soil strength were presented [53]. The authors stated that 500 mW should be sufficient for the proper working of the monitoring wireless system. The eccentricity of this system is that the solar panels were designed to be mounted both, on the track and on the train. This way, the RF transmitter and sensor nodes from the railway could be powered, but also the microcontroller and RF transceiver located on the train. Influenced by this, another research study used this EH approach to detect obstacles laying on the railway 2km in advance [54].
Besides, the strain on critical parts was controlled by a solar powered sensor network in a study [55]. Here, 12 V car batteries powered the system, and these were recharged by solar panels installed on top of a bridge. The study was performed for a limited time, but the authors stressed the potential for the implementation of solar EHers for safety purposes such as health monitoring in railways.
Despite the possibilities, contraindications were also shown, as dust and dirt highly affected the results, and weather conditions were not stable.
2.4 Wireless communication networks
Wireless networks use radio waves so as not to use cables for the communication between devices.
The functioning of these networks is analogous to the conventional wired networks, but the information signals need to be modified into a pattern which enables transmission through the air.
This form of transmission facilitates devices to be installed apart and connect remotely. In this thesis, a WCN is contemplated as an enabler of the communication needed between the balises and a control unit in order to monitor the health of each balise. Therefore, the following paragraphs summarize the available forms of WCN.
With regard to wireless networks, this communication method does not need cabling of any type, and instead uses radio waves to connect devices to each other [56]. Legal issues need to be considered, as many countries have different legislation in terms of the allowed frequency ranges for commercial communication. These networks are commonly classified according to their signal range. This way, the ascending range order is wireless body-area networks (WBAN), wireless personal-area networks (WPAN), wireless local-area networks (WLAN), wireless metropolitan-area networks (WMAN) and wireless wide-area networks (WWAN), as shown in Figure 7. The former three are categorized as short-range, which mostly operate over the unlicensed frequency spectrum, that is reserved for industrial, scientific, and medical applications. Although the frequency ranges vary depending on the country, 2.4 GHz and 5 GHz are the most used frequency bands globally. On the other hand, WMAN and WWAN are long-range, which are normally provided as a communication service by telecommunication companies. These networks aim for a global coverage, and hence they may imply the use of satellites.
An overview of the presented communications is performed in the upcoming paragraphs. WBAN is not considered in this thesis, as it is used in human bodies and it is out of the scope of this research work.
16 | Background
2.4.1 WPAN
WPAN is based in the institute of electrical and electronics engineers (IEEE) 802.15 standard. This communication requires a limited amount or no infrastructure, nor a connectivity outside the link and its typical communication range is 10m. Therefore, power-efficient, it allows the construction of low-cost networks [57].
Typical technologies include Bluetooth, ZigBee, infrared data association (IrDA) or ultra- wideband (UWB), which are distinguished because of their low bit rate as well as a low power demand.
With regard to Bluetooth, this was firstly introduced for short range point to multipoint communication and inexpensive equipment, although late development included Bluetooth-enabled components to amplify the field applications of it [58]. It corresponds to the IEEE 802.15.1 standard, which is comprised of four classes, with a typical range of 100 m, 10 m, 1 m, and 0.5 m in an ascending class order. The permitted power of Bluetooth classes is proportional to their range, with 100 mW or 20 dBm for class 1, and 0.5 mW or -3 dBm for class 4. When different devices are inside each other’s coverage ranges, they share information at up to 720Kbps in the 2.4 GHz band.
With reference to ZigBee, it is built on the IEEE 802.15.4 standard and was aimed to be an inexpensive, easy to implement, and reliable open global standard for low data rate and low-power device networks [59]. Same as Bluetooth, it operates on the 2.4 GHz band, and it also uses the 868 MHz and 900 MHz unlicensed bands, with a superlative transfer rate of 250 Kbps, a range of 10-20 m, and a power consumption of 10-100 mW. With star, cluster tree, and mesh as supported topologies, ZigBee allows the creation of larger WCN which do not involve high data throughput.
Figure 7: Wireless access geographic coverage. RFID, radio frequency identification; NFC, near field communication; NB-IoT, Narrowband Internet of Things; LTE, long term evolution;
GSM, global system for mobile communications; MTC, machine type communication;
SUN, smart utility network; LE, [67] is licensed under CC BY 4.0
