BACHELOR'S THESIS
Fire Gas Ventilation Analysis of a Railroad Tunnel
Case Study: Hallsberg-Stenkumla Railroad Tunnel
Björn Grybäck Melin 2016
Fire Protection Engineering Fire Protection Engineer
Luleå University of Technology
Department of Civil, Environmental and Natural Resources Engineering
Luleå University of Technology
Bachelor Thesis
Fire gas ventilation analysis of a railroad tunnel
Case study: Hallsberg-Stenkumla railroad tunnel
Author:
Björn Grybäck Melin
External supervisor: Internal supervisor:
Nicolas Albornoz Dr. Adrianus Halim
Bachelor thesis - X7007B Fire Protection Engineer Fire Protection Engineering
The Department of Civil, Environmental and Natural Resources Engineering
I
Preface
This thesis would not have been made without the assistance and ideas from my supervisors Dr. Adrianus Halim at LTU and Nicolas Albornoz at WSP which I am and will continue to be grateful for.
I also am thankful for the support given by my colleagues at WSP and my family, especially my father Mats Melin.
Björn Grybäck Melin, Linköping 2016-06-13
II
Abstract
The Swedish Transport Administration is planning to expand the railroad network between Hallsberg and Stenkumla to increase the capacity and punctuality. The expansion includes a new track and also a new railroad tunnel.
A fire on a passenger train located in a tunnel could lead to an evacuation scenario.
Evacuation trough an underground tunnel during a fire is done with potentially great consequences. Several technical systems can be applied to minimize the consequences.
One example of such technical system is a fire gas ventilation system which also is going to be the focus for this thesis. Regulations, interviews, comparisons, calculations and simulations have been used to answer the thesis objectives.
• Which fire gas ventilation system is optimized for the railroad tunnel´s area of use?
• How is fire gas ventilation systems performed in other railroad tunnels compared with the current tunnel?
• What is the ongoing research work within the field about?
• Can the research lead to new solutions of fire gas ventilation systems in the near future?
After the identification of requirements, design parameters, system options and the indication by the calculations and simulations that risk reducing means might have to be applied. The application of a mechanical fire gas ventilation system was ruled out in the analysis due to other means are prioritized and more cost effective.
The comparison with other railroad tunnels confirmed that the approach was similar due to the fact that the majority of Sweden’s railroad tunnels longer than 1 km without stations are not equipped with a mechanical fire gas ventilation system.
Several limitations were discussed and they generated a continuous analysis consisting of profound information gathering and usage of more advanced simulation software.
No major ongoing research about fire gas ventilation systems in railroad tunnels has been identified.
III
Sammanfattning
Trafikverket planerar att utöka tågförbindelsen mellan Hallsberg och Stenkumla för att öka kapaciteten och punktligheten. Utökning inkluderar en ny dragning av tågspåret och en ny järnvägstunnel.
En brand ombord på ett persontåg i en tunnel kan leda till ett utrymningsscenario.
Utrymning genom en underjordisk tunnel medan det brinner görs med potentiellt höga konsekvenser. För att minska konsekvenserna kan flera tekniska system tillämpas. Ett exempel på ett tekniskt system är brandgasventilation som också är i fokus inom detta examensarbete. Regelverk, intervjuer, jämförelser, beräkningar och simuleringsprogram har tillämpats för att besvara examenarbetets mål.
• Vilket brandgasventilationssystem är optimerat för järnvägstunnels användningsområde?
• Hur är brandgasventilationssystem utförda i andra tunnlar jämfört med den aktuella tunneln?
• Vad handlar den pågående forskningen inom området om?
• Kan den leda till nya lösningar av brandgasventilationssystem inom en nära framtid?
Efter identifikation av krav, konstruktionsparametrar, systemalternativ och genomförda beräkningar och simuleringar indikerades det att risk reducerande medel kanske måste tillämpas. Tillämpning av ett mekaniskt brandgasventilationssystem uteslöts i analysen då andra medel är prioriterade och mer kostnadseffektiva.
Jämförelsen med andra järnvägstunnlar användes till att styrka detta tillvägagångsätt då majoriteten av Sveriges järnvägstunnlar längre än 1 km utan stationer inte är utrustade med mekanisk brandgasventilation.
Dock har flera osäkerheter stötts på som kräver vidare analyser i form av
informationssökande och användandet av mer avancerade simuleringsprogram. Ingen information om större pågående forskning inom området har identifierats.
IV
Contents
1 INTRODUCTION ... 1
1.1 Background... 1
1.2 Thesis objectives ... 1
1.3 Methodology ... 2
1.4 Boundaries ... 2
2 THEORY AND DATA ... 3
2.1 Laws, regulations and standards ... 3
2.1.1 Directive 2008/57/EG ... 4
2.1.2 SRT TSI ... 4
2.1.3 Construction regulation (2011:338) ... 5
2.1.4 Rescue service ... 5
2.1.5 TRVK Tunnel 11 ... 5
2.1.6 TRVR Tunnel 11 ... 6
2.1.7 BVH 585.30 ... 6
2.2 Risk analysis ... 7
2.3 Design parameters ... 8
2.3.1 Tunnel location ... 9
2.3.2 Tunnel configuration ... 11
2.3.3 Wind conditions around the tunnel ... 12
2.3.4 Characteristic of major fire source ... 13
2.3.5 Number of personnel that have to be evacuated ... 16
2.4 Fire gas ventilation systems ... 16
2.4.1 Non-mechanical (Natural) ... 16
2.4.2 Mechanical ... 17
2.5 Fire gas ventilation system in other tunnels in Sweden ... 21
3 RESULTS ... 23
3.1 Required Safe Egress Time (RSET) ... 23
3.2 Available Safe Escape Time (ASET) ... 24
V
3.2.1 Time and visibility ... 24
3.3 TuFT ... 26
3.3.1 Simulation ... 26
3.4 Ongoing research ... 28
4 ANALYSIS ... 30
4.1 Optimal fire gas ventilation design ... 30
4.2 Continuation of the design process ... 31
4.3 Limitations in this thesis ... 32
5 CONCLUSION ... 35
6 FURTHER RESEARCH ... 35
7 REFERENCE ... 36
APPENDIX A - SMHI PREDOMINANT WIND REPORT ... 39
APPENDIX B - FROUDE NUMBER CALCULATION ... 43
APPENDIX C - VISIBILITY CALCULATION ... 44
APPENDIX D - TUFT INPUT DATA FILE ... 45
APPENDIX E - CRITICAL VELOCITY CALCULATION ... 48
APPENDIX F - INTERVIEW WITH MR. HAUKUR INGASON FROM SP ... 49
1
1 Introduction
1.1 Background
A railroad tunnel is a confined space. Due to this nature, the consequence of a fire event in a railroad tunnel has the potential to be far greater than a fire event on the surface.
One component to ensure the safety of evacuating passengers and rescue personnel in a fire event could be a fire gas ventilation system to evacuate smoke and fire gases and thereby extending the available safe escape time and possibly aid the rescue service in their work. It is therefore important to analyze the design of a fire gas ventilation system in a railroad tunnel.
The Swedish Transport Administration (Trafikverket) is planning to expand the railroad network between Hallsberg and Stenkumla to increase the capacity and punctuality.
Several locations of the railroad tunnel within the chosen corridor have been considered and the final decision was made in February 2016. The construction start of the tunnel will begin in 2021. The railroad tunnel will approximately be 2000 meters long and is going to consist of two tubes with one railroad track in each tube. (Trafikverket, 2016)
1.2 Thesis objectives
The main goal of this thesis is to establish a preliminary design of the Hallsberg-
Stenkumla railroad tunnel´s fire gas ventilation system. The second goal is to inform the reader about the ongoing and upcoming research works and its associated effect on designing fire gas ventilation systems in railroad tunnels.
• Which fire gas ventilation system is optimized for the railroad tunnel´s area of use?
• How is fire gas ventilation systems performed in other railroad tunnels compared with the current tunnel?
• What is the ongoing research work within the field about?
• Can the research lead to new solutions of fire gas ventilation systems in the near future?
2
1.3 Methodology
The work in this thesis was carried out in cooperation with the department Fire & Risk at WSP Sverige AB in Linköping. WSP (William Sale Partnership) is a consulting and management company housing more than 32000 employees globally. One of the main service areas is Transport and Infrastructure, which include underground structures of various kinds.
The first thesis objective was approached by first identifying relevant requirements, design parameters, fire gas ventilation system options and then verifying which system is needed using manual calculations and simulation software called Tunnel Fire Tools.
The simulation software Tunnel Fire Tools (TuFT) was used for the simulations as a complement to the manual calculations made. TuFT simulates fire scenarios in railroad or road tunnels and its effect on evacuating passengers and rescue personal. It uses a simplified form of the energy equation which means a one-dimensional calculation procedure is applied.
The last three thesis objectives were approached by performing a literature study and the last two thesis objectives were complemented with an interview with Haukur Ingason from SP Technical Research Institute of Sweden.
1.4 Boundaries
The boundaries for this thesis are defined by the title “Fire gas ventilation analysis of a railroad tunnel. Case study: Hallsberg-Stenkumla railroad tunnel.” The level of detail is according to the time available to answer the thesis objectives.
3
2 Theory and data
The answer to the defined question “Which fire gas ventilation system is optimized for the railroad tunnel´s area of use?” stated in the introduction, begins with identifying the purposes of such system in a railroad tunnel in order to proceed.
Any fire gas ventilation system has four main purposes: Ensuring the possibility to self- evacuate, ensuring work conditions for the rescue service, reducing fire and smoke spread and enabling salvage at an early stage. These all four can be summarized by stating that fire gas ventilation systems’ purpose is to change pressure conditions to ensure the safety of lives and protection of property. However, since all buildings are not the same, the purposes a of fire gas ventilation system has different importance from building to building. (Svensson, 2000)
The important general purposes of a fire gas ventilation system in railroad tunnels are to ensure the possibility to self-evacuate and to ensure safe working conditions for the rescue service. The reason is because railroad tunnels are confined spaces with limited fire- and smoke detection. (Ingason, 2015)
To verify when the purposes of the fire gas ventilation system are achieved, the capacity of the system and the effects generated by that capacity are compared to the
requirements that are stated by laws, regulations and standards.
2.1 Laws, regulations and standards
The European Union wants to create a uniform railroad system. By doing that the railroad network in Europe becomes more effective and competitive against other transport solutions. EU is enforcing directives due to the problem of different technical solutions and national regulations. (European Union, 2008)
An overview of the safety regulation documents are shown in Figure 1. The practical usable requirements regarding a fire gas ventilation system are found at the bottom of the hierarchy ladder but it is important to know where these are derived from.
Therefore the investigation for this thesis starts at the top and proceeds down in the hierarchy ladder. Note that only requirements and information regarding fire gas
4
ventilation systems of the design of such system is presented.
Figure 1 The hierarchy of the safety regulating documents for railroad tunnels regarding fire gas ventilation system and its associated name. (BeFo,2015)
2.1.1 Directive 2008/57/EG
The directive 2008/57/EG establishes how a TSI (Technical specification for
interoperability) is produced and what it should contain. (European Union, 2008)
2.1.2 SRT TSI
The TSI relating to SRT (Safety in railway tunnels) presents the lowest acceptable
requirements on the safety in railroad tunnels. Its purpose is to create an optimal safety level in railway tunnels in the most cost efficient way. The requirement for fire gas ventilation systems are as followed: (EU, 2014)
• The need to provide installations to aid the self-rescue and evacuation of passengers and personal, and the possibility for the rescue service to save people at an accident in a railroad tunnel must be considered when designing a railroad tunnel.
EU-directives and regulations Directive 2008/57/EG and SRT TSI
Swedish law/ordinance Construction regulation (2011:338)
Government regulations & general advice -
Local regulations Rescue service
Handbooks, guildelines & builder requierments TRVK tunnel 11, TRVR tunnel 11 and BVH 585.30
5
2.1.3 Construction regulation (2011:338)
According to Swedish regulation a railroad tunnel is defined as a construction and must therefore fulfill chapter 3 paragraph 8; Characteristics required for safety in case of fire.
Point 4 and 5 are relevant for fire gas ventilation systems and its purpose.
(Näringsdepartementet, 2011)
• 4. People who are in the construction in case of fire can leave it or be rescued by other means
• 5. Consideration has been given to the safety of rescue teams in case of fire.
However, the Construction regulation (2011:338) has already been incorporated through SRT TSI since the 1th of January 2015 because it is an EU regulation and applies directly in Sweden.
2.1.4 Rescue service
Rescue services in railroad tunnels are carried out by fire departments and the primary responsibility for the tunnel between Hallsberg and Stenkumla is the Fire department of Nerike. A dialog between the afflicted fire department and the project members are established to determine a sustainable safety design regarding primarily rescue services in the railroad tunnel, which could conclude the necessity of a fire gas ventilation
system.
The rescue service estimates its need of safety installments on the basis of the Construction regulation (2011:338). The level of capability and experience with
firefighting in railroad tunnels differentiate between fire departments. (Gehandler, et al., 2013)
2.1.5 TRVK Tunnel 11
TRVK stands for Trafikverket’s technical requirements and TRVK Tunnel 11 is Trafikverket’s added requirements to SRT TSI. It should be applied with a certain reservation since TRVK tunnel 11 is not updated according to the latest SRT TSI.
However, this version is still valid since an updated version of TRVK is not yet available. The requirements could affect the design of a fire gas ventilation system.
• The spread of fire gases to evacuation routes, emergency exits and extra spaces must be limited.
6
• Smoke in an emergency exit cannot spread to another emergency exit.
• Smoke spread between the track tunnel and emergency exits are prevented.
It also states if the requirements are not fulfilled with a non-mechanical ventilation system must a mechanical ventilation system be installed. The design of mechanical ventilation must take in account that fans near the fire can be destroyed and that
aerodynamic conditions affect the ventilation system. The whole associated calculation process must be presented and motivated. (Trafikverket, 2011)
2.1.6 TRVR Tunnel 11
TRVR stands for Trafikverket’s technical advice and TRVR Tunnel 11 is the official advice to TRVK Tunnel 11. It states that verification of the system must be done according to good engineering if the advices are not followed and with good reason.
(Trafikverket, 2011)
• Installations necessary for the evacuation must withstand the fire impact to ensure the self- evacuating possibility and work conditions for the rescue personal.
• Safety analyses of personal safety may be performed according to BVH 585.30 depending on the length of the tunnel. Tunnels with a length between 300-1000 meters is evaluated case-to- case if a safety analyses is supposed to be performed and tunnels with a length over 1000 meters means that a safety analyses must be performed.
2.1.7 BVH 585.30
BVH stands for the Swedish Rail Administration Handbook and BVH 585.30 is the required handbook to use when analyzing and evaluating the personal safety since the tunnel between Hallsberg and Stenkumla tunnel is longer than 1000 m. The handbook was created to guide the design process of railroad tunnels and to verify the safety level according to Trafikverket´s level of ambition.
“Railroad traffic per kilometer in tunnels shall be as safe as railroad traffic per kilometer on land tracks, exclusive crossings” (Banverket, 2007).
BVH 585.30 provides limits on time and visibility, toxic gases, heat radiation and temperature which could affect the design of the fire gas ventilation system. It is normally fire gases which affect the time and visibility and toxic gases that exceed the limits first. (Banverket, 2007)
7
Time and visibility: The evacuating passengers cannot stay in areas more than 15 minutes with sight lengths less than 3 meters.
Toxic gases: The last evacuating group of passengers can reach a safe place before the toxic gases cause unconsciousness.
Heat radiation: The limits are 2,5 kW/m2 for long duration and 10 kW/m2 for short duration or a maximal total radiation of 60 KJ/m2 except the radiation of 1 kW/m2. Temperature: Accumulating fraction dose Ftemp < 1
2.2 Risk analysis
The risk analysis of the Hallsberg-Stenkumla tunnel was carried out within the project and defines an acceptable risk level, as followed: “For the absolute majority (<90 %) of possible fire scenarios larger than 1 MW on a standstill passenger train in the tunnel shall at least 99 % of the passengers be able to evacuate to a safe place before critical conditions prevail in the tunnel”, see Figure 2 (WSP Sverige AB, 2016). It is based on Trafikverket´s ambition presented in the previous subsection 2.1.7.
Figure 2 Illustration of the acceptable risk level.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Number of surviviours in percentage
Scenarios which represent a percentage of the worst possible scenario Death of personnel is acceptable
Death of personnel is not acceptable
8
2.3 Design parameters
An identification of the design parameters which could affect the design of a fire gas ventilation system in a railroad tunnel must be done in order to verify that the acceptable limits, presented in the previous section 2.1, are achieved. The design parameters have been identified via qualitative discussion within the project, in comparison with other ongoing railroad tunnel projects and based on which data is needed for the calculations and simulations.
The data presented in the design parameters represents a worst case scenario which according to the level of ambition represents 100 % of the fire scenarios. A quick
sensitivity analysis was performed on some of the design parameters to identify which data generates a worse condition in the fire scenario and thus to represent the worst case scenario.
One could define a fire scenario that would represent 90 % of the possible fire scenarios and design the tunnel´s level of security accordingly. However, that would mean the design parameters must include multiple options with a corresponding probability.
Unfortunately that is too large of a scope for this thesis.
The design parameters are as followed
• Tunnel location
• Tunnel configuration
• Wind conditions around the tunnel
• Characteristic of major fire source
• Number of personnel that have to be evacuated
It must be noted that the project is still ongoing and the design parameters might change and thus could affect the design of the fire gas ventilation system.
The main structural system of the railroad tunnel is going to be designed to achieve a technical longevity of 120 years and after 120 years the design parameters might have altered. A long term perspective is therefore important to have when investigating the design parameters. (Trafikverket, 2011)
9
2.3.1 Tunnel location
The location of the expansion is south of Örebro, between Hallsberg and Degerön. The tunnel itself is located between Hallsberg and Stenkumla, see Figure 3 and Figure 4. The exact starting and ending point of the tunnel is not yet determined. (WSP, 2016)
Figure 3 The freight line through Bergslagen (the red route) passing the western main line (the blue route) within the pink circled area is the part of Trafikverket’s expansion which include the construction of a railroad tunnel.
(Trafikverket, 2014)
10
Figure 4 Aerial view of the chosen corridor and between the black dots on the purple line is approximately where the tunnel portals will be. (Trafikverket, 2014)
11
2.3.2 Tunnel configuration
The tunnel is two parallel tubes with a center to center distance of roughly 20 meters with a connecting cross tunnel every 500 meters. The normal section of both tunnels has a height of 7,4 meters and an area of roughly 55,5 square meters. The two tunnels are as good as identical, hence referred to as only one tunnel, see Figure 5. (WSP, 2015)
Figure 5 The normal section of the tunnel. (WSP, 2015)
The length of tunnel is around 2000 meters. The grade of the tunnel is at its steepest points no more than 10 per mille. The northern tunnel opening, closest to Hallsberg, is the tunnel’s highest point, see Figure 6. (WSP, 2015)
Figure 6 The railroad track (red line) and the ground level (black line). (WSP, 2015)
12
2.3.3 Wind conditions around the tunnel
Any ventilation system is aiming towards either creating or eliminating pressure differences. It is therefore important to investigate the factors that could have an effect on the wind conditions in a railroad tunnel. These possible factors are;
• Wind speed against the tunnel’s openings,
• heat release rate,
• the tunnel’s grade,
• shape and surface roughness of the tunnel´s interior walls,
• objects in the tunnel,
• temperature differences
• and train movements within the tunnel.
The wind speed against the tunnel’s openings will be further investigated due to it is one method to try and identify a likely initial air velocity of the fire scenario in the tunnel. Some of the other factors are design parameters as well but their effect on the wind conditions will not be elaborated further. (Banverket, 2007)
The wind climate differs from year to year, and particularly the wind speed can vary significantly between different time periods. It is difficult to identify any clear trends of the wind climate in the future based on the information gathered. (Asp, 2015)
Results presented below are based on observations every hour at Örebro airport under 19 years; 1996-2014. The wind speed is measured on a 10-minute average wind speed at 10 meters above the ground. The predominant wind speed is in 91,6% of the time in the interval 0-6,5 m/s. The prevailing wind directions are generally between southwest and west. See Figure 7 and the rest of the wind report in the Appendix A.
13
Figure 7 Wind rose showing the percentage wind directions with associated percentage wind speeds.
Based on the wind profile power law, where the atmosphere always is neutrally layered, wind speeds at 2-3 meters above the ground are between 80-84% of the wind speed at 10 meters above the ground. The interval becomes roughly 0-5,3 m/s at heights 2-3 meters above the ground. (Asp, 2015)
Even though a wind speed against the tunnel´s openings can be approximated with predominant winds and directions it does not generate enough information to set a likely wind speed in the tunnel. Factors mentioned above also need to be fully
investigated. To further the work with the thesis, a wind speed of 2 m/s in the tunnel is used based on another project. (Hägglund, 2012)
2.3.4 Characteristic of major fire source
The characteristic of the major fire source is based on combustible materials inside the tunnel. A railroad tunnel’s building components and installations, without stations,
14
generally generate a small amount of combustible materials and therefore do not significantly contribute to the fire effect curve. Therefore, the source of combustible materials is represented by trains operating in the tunnel.
The passenger trains that will operate the new stretch between Hallsberg and Stenkumla are one model of a regional train, namely the Regina X51-2 by Bombardier, see Figure 8.
It is 54 meters long and has 4 doors on each side with a width of 1,4 meters. The traffic prognosis for the year 2030 states that the passenger trains operating the same stretch will be a similar train model (Hedlund, 2015).
Figure 8 Overview of a Regina X51-2. (Trafikverket, 2015)
Freight trains operate in the tunnel as well and those trains could have a fire effect curve to stagnate around 100 MW. Freight trains will not be taken in consideration, since a fire of that size is impossible to control with a fire gas ventilation system. The purpose of a fire ventilation system is questionable in a fire scenario with a freight train due to the small amount of people working on a freight train. It will, however, be a design
parameter for example in the design of the tunnel´s structure and interior walls, so the tunnel does not collapse.
The characteristics of a fire source are mainly represented by the Heat Release Rate (HRR), and the HRR is different from train to train. Since only one type of passenger train will operate in the tunnel, the HRR is based on that one train. A possible worst case scenario is with fires whose growth rate is of 0,012 kW/m2, up to 20 MW, see Figure 9.
(Trafikverket, 2015)
15
When comparing to other transportation vehicles the train develops a fire as four cars, 5 MW each. (Lönnermark & Ingason, 2013)
Figure 9 Diagram showing the heat release rate over time produced by TuFT.
The following Table 1 presents data regarding the fire for the upcoming calculations and simulations. The values were taken from exercise assignments in TuFT´s manual.
(Fridolf & Frantzich, 2014)
Table 1 Values regarding the train fire characteristic.
Definition Value and unit
Mass optical density 304 m2/kg
Heat of combustion 15500 kJ/kg
CO2 Yield 1,5 kg/kg
CO Yield 0,027 kg/kg
0 5000 10000 15000 20000 25000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
HRR [kW]
Time [s]
Heat Release Rate
16
2.3.5 Number of personnel that have to be evacuated
Previous sub-section 2.3.4 presented that there are only one kind of passenger train used on the stretch between Hallsberg and Stenkumla. The Regina X51-2 can carry a total of 200 sitting and standing passengers including working personal. According to the traffic prognosis of 2030 the passenger train, Regina X51-2, might drive double connected meaning the maximum amount of passengers add up to 400 passengers. (Frisk, 2015)
2.4 Fire gas ventilation systems
Fire gas ventilation systems must not be confused with normal ventilation systems. A normal ventilation system’s purpose is to lower the concentration of contaminants and lower the temperature in tunnels. If both ventilation systems are required they could and should be combined. However, this is common in longer road tunnels and not in railroad tunnels due to all trains being electrical except some diesel-powered repair locomotives. (Ingason, et al., 2015)
2.4.1 Non-mechanical (Natural)
A non-mechanical system does not include any fans or similar machines. It is dependent on the predominant wind conditions around the tunnel. Together the factors presented in 2.3.3 create pressure difference between tunnel’s portals which defines the direction and velocity of the airflow. Non-mechanical fire gas ventilation systems in railroad tunnels can therefore only extract smoke longitudinally. Vertical extraction without fans has been tested and was yielded not effective. (Ingason, 2015)
A railroad tunnel without fans or similar machines is not automatically classified as a railroad tunnel with non-mechanical fire gas ventilation. To meet this classification the critical air velocity must be achieved and maintained to minimize the back-layering smoke, see Figure 10. (Ingason, et al., 2015)
17
Figure 10 Non-mechanical longitudinal fire gas ventilation system principle.
If the pressure difference does not produce adequate airflow the smoke will disperse to both directions, see Figure 11. The smoke produced by the fire will be cooled by the tunnel´s interior walls and will descend to the floor and flow back to the fire with some fresh air. (Ingason, et al., 2015)
Figure 11 Fire in a tunnel with a low total pressure differences between the tunnel openings. (Ingason, et al., 2015)
2.4.2 Mechanical
A mechanical system creates pressure difference between tunnel portals using a mechanical device, usually a fan. There are several mechanical systems, which are as followed.
Longitudinal
There are several layouts of longitudinal systems. The most common layout is placing jet fans along the celling or walls of the tunnel, see Figure 12. The jet fans could be reversible to enable a more strategic smoke evacuation. In this layout, the smoke is blown out via one of tunnel portals.
18
Figure 12 Groups of two jet fans mounted to the tunnel’s celling. (Fan factories,2010)
Longitudinal system is more suitable for road tunnels regarding self-evacuation. The drivers and passengers who are stuck in traffic upstream of the fire escape by foot.
Downstream of the fire they simply drive out of the tunnel, see Figure 13.
Figure 13 A longitudinal fire gas ventilation principles in a roadway tunnel. (TunnelTalk,2012)
In a railroad tunnel the part of the train that is downstream of the fire is unable to drive out since it is one vehicle. Downstream of the fire the conditions are harsh for
passengers to evacuate by foot. Hence, longitudinal fire gas ventilation in a railroad is
19
mostly useful after the evacuation is completed and an attack point for the rescue service is needed to be made. (Ingason, 2015)
A possible layout for mechanical longitudinal fire gas ventilation in the tunnel between Hallsberg and Stenkumla could be as followed. Jet fans are placed in the following positions counted in meters from the lowest placed exit at 100, 350, 600, 850, 1100, 1350 and 1600. (Lilliengren, 2015) A total of 7 jet fans are needed with given redundancy which means that only 5 impulse fans are required to work at the same time to produce a sufficient air flow. Note that the tunnel is designed as two separate tubes which require one ventilation system each. (Lilliengren, 2015)
Transverse
Under normal working conditions a transverse ventilation systems supplies a tunnel with fresh air and extracts contaminated air. The extract ducts are normally placed in the celling, since the contaminated air is hot and rises to the celling, while the supply ducts could be found at the floor, walls and ceiling depending on the design, see Figure 14. (Ingason, et al., 2015)
Figure 14 Transverse ventilation which supplies and extracts air. (Ingason, et al., 2015)
When a fire is detected in the tunnel a transverse fire gas ventilation system should either shut off or reverse the supply of fresh air depending on the position of the ducts to not worsen the fire scenario. The ducts closest to the fire should be fully activated for a maximal smoke control. It requires an extensive detections system which determines the selection of which ducts to activate. Since the smoke extraction takes place closer to the fire the smoke spread can be better controlled compared with a longitudinal
ventilation system making it the safer option during an evacuation. (Ingason, et al., 2015)
20
Semi-transverse
The word semi means “half, partial or incomplete” which corresponds well with the definition of a semi-transverse ventilation system compared with a transverse
ventilation system. There are namely two variants of a semi-transverse ventilation system; One which only supplies and the other which only extracts air, see Figure 15.
Figure 15 Normal sections of both variants of a semi-transverse ventilation system. (Ingason, et al., 2015)
The supply variant must have its ducts placed at the celling and be reversible to be able to function as a fire gas ventilation system. There is no significant function difference between semi-transverse and transverse fire gas ventilation systems during a fire scenario. Semi-transverse fire gas ventilation systems extract air vertically as well, making it safer for the evacuating personnel, see Figure 16.
Figure 16 A semi transversal fire gas ventilation principle in a roadway tunnel. (TunnelTalk,2012)
21
2.5 Fire gas ventilation system in other tunnels in Sweden
Only the tunnel´s configuration regarding the layout and length has been used as comparable design parameters. The comparison only covers 15 of all about 30 Swedish railroad tunnels longer than 1 km. (Lundman, et al., 2009)
Table 2 Comparisons between Sweden´s railroad tunnels sorted by the length of the tunnels.
Name and the year of commissioning
Layout Length (meters) Ventilation system
Hallandsåstunneln 2015
Two single track tunnels
West tunnel 8722 East tunnel 8710
Non-mechanical
Namntalltunneln 2009
Single track tunnel with a service tunnel
6004 Non-mechanical
Citytunneln 2010
Two single track tunnels with two stations
5889 Only mechanical at
stations
Björnböletunneln 2009
Single track tunnel with a service tunnel
5099 Non-mechanical
Kroksbergstunneln 2012
Single track tunnel with a service tunnel
4541 Non-mechanical
Bjässholmstunneln 2012
Single track tunnel with a service tunnel
3490 Non-mechanical
Åskottstunneln 2009
Single track tunnel with a service tunnel
3276 Non-mechanical
22 Nygårdstunneln
2008
Double track tunnel with a service tunnel
3030 Non-mechanical
Snarabergstunneln 2012
Single track tunnel with a service tunnel
2411 Non-mechanical
Varvbergstunneln 2008
Single track tunnel with a service tunnel
2087 Non-mechanical
Murbergstunneln 2012
Single track tunnel with a service tunnel
1635 Non-mechanical
Strannetunneln 2008
Single track tunnel with a service tunnel
1436 Non-mechanical
Hjältatunneln 2008
Single track tunnel with a service tunnel
1254 Non-mechanical
Kalldalstunneln 2008
Single track tunnel 1115 Non-mechanical
Åsbergstunneln 2008
Single track tunnel with a service tunnel
1004 Non-mechanical
23
3 Results
The limits presented by BVH 585.30 relate more or less to either the fire or the smoke produced by the fire. Heat and radiation are directly connected to the fire and for the evacuating passengers it might only be a problem in the early stages of a fire scenario.
Toxic gases are directly connected to the smoke and are more prone to be the cause of deaths during an evacuation in a tunnel. Visibility is also directly connected to the smoke but only affect the time needed to evacuate. (Ingason, et al., 2015)
A comparison between Available safe escape time (ASET) and Required Safe Egress Time (RSET) is one of the few ways to confirm that the requirements are upheld.
(Fridolf, 2015)
3.1 Required Safe Egress Time (RSET)
RSET can be divided into three parts; Time to detection of the fire, time to initiate evacuation and the evacuation itself. The first two are completely dependable on the safety installations on the train, the train operators’ actions and the behavior of the personnel. To truly determine those would exceed the time available for the thesis.
Hence, the total time until the passengers start to evacuate is based on experiences from other projects. Time until evacuation begins is approximated to be between 2-10 min.
The evacuating itself consists of the time to egress the train and walking to the nearest cross tunnel or tunnel opening. An estimated time to egress the train per person and door was set to 0,2 p/s, with reservations. The walking speed of 0,2-0,8 m/s in the tunnel is based on laboratory experiments in road tunnels. (Fridolf, 2015)
Since there are many uncertainties, RSET has been set to be between 13-50 minutes. It is according to a worst possible scenario of 400 evacuating passengers and the fire
blocking the one of the exits making the distance to walk 450 meters. Since there are uncertainties with this determination of the RSET one should apply an additional minute to be one the safe side.
24
3.2 Available Safe Escape Time (ASET)
It has not been possible to identify a specific ASET using manual calculation. It does only show that after 10 minutes after the fire has started the conditions for the
evacuating personnel are bad. This is mainly due to too many calculation simplifications and the number of calculations needed to be made. The choice to calculate the
conditions at a certain time were set to 10 minutes because it is highly likely, if not guaranteed, that personnel are evacuating in the tunnel at that time.
3.2.1 Time and visibility
Gas layer height calculations are complex and not applicable due to too many
uncertainties. A simple model of stratification, which was developed by Ingason et al (2015) was used instead. The model is based on the identification of where and when region 1 ends and region 2 starts , see Figure 17. Region 2 includes disturbance in the stratification between the smoke and fresh air and it results in evacuating passengers to walk in smoke.
Figure 17 Fire scenario in a tunnel which includes three regions of stratification. (Ingason, et al., 2015)
Region 2 starts when Froude number is greater than 0,9 and the time of interest depends on when the evacuating passengers is walking downstream of the fire. (Ingason, et al., 2015)
According to the calculations performed the Froude number is above 0,9 at 10 minutes (600 seconds) close to the fire thus the evacuating passengers walk in smoke almost the entire time, see Figure 18. Froude number increases further downstream of the fire due to the smoke and fire gases are cooled along the tunnel´s interior walls. A clear
stratification can established with low air velocities and a greater HRR. The model,
25
which was developed in an Excel spreadsheet, can be seen in Appendix B. (Ingason, 2016)
Figure 18 Froude’s number along the tunnel at 600 seconds.
Visibility was calculated using a model developed by Beard and Carvel (2005), as shown in Appendix C. Due to no stratification, the visibility at 600 seconds along the tunnel is low, see Figure 19.
Figure 19 Visibility along the tunnel at 600 seconds.
0 5 10 15 20
0 20 40 60 80 100 120 140 160 180 200 220 240
Froudes number (-)
Distance from fire (m)
Froudes number along the tunnel at 600 s
Fr(x)
0 0,5 1 1,5 2
0 20 40 60 80 100 120 140 160 180 200 220 240
Visibility (m)
Distance from fire (m)
Visibility along the tunnel at 600 s
Visibility
26
3.3 TuFT
The simulation software called TuFT is applied since the comparison between ASET and RSET, only using manual calculations, is insufficient due to no exact times could be set.
TuFT stands for Tunnel Fire Tools and it is a text based calculation model to simulate the fire dynamic consequences in a tunnel and its effect on evacuating passengers and rescue personnel. Its main purpose is to support the planning of rescue operations in tunnels, but the rescue personnel and evacuating passengers do not interact with each other. The rescue personnel are programmed to progress in the tunnel to the fire and put it out. Therefore only the evacuating passengers are of interest in this simulation.
(Fridolf & Frantzich, 2014)
TuFT is primarily focused on calculations and the results from TuFT only represents as comma based text files and the visualization consist of diagrams made by an Excel macro. It only takes a few seconds to simulate a scenario. This has been used to vary some design parameters used as input data, also known as a sensibility analysis. For the full input data file see Appendix D. (Fridolf & Frantzich, 2014)
3.3.1 Simulation
The tunnel length is set to 2000 meters and the train is located in between of the tunnel openings, at 1000 meters. The fire is located at one of the train´s end and the fire is blocking an exit. The evacuating personnel are programmed to walk away from the fire and in this case downstream of the fire to the next cross tunnel or so called exit 500 meters from the blocked exit. The scenario constructed is partly shown in Figure 20.
Figure 20 Rough sketch of the simulation made in TuFT.
27
When evacuation occurs it is assumed the train is evacuated through doors on both side of the train. Since TuFT only simulates evacuation through doors on one side of the train the evacuating personnel is set to 200 in the input file for TuFT. Hence, the consequences are supposed to be viewed as doubled regarding the number of affected evacuating personnel but the percentage of affected personnel stays the same. To sum up, TuFT is set to evacuate 200 personnel through doors on one side of the train and track but it represents an evacuation with 400 personnel through doors on both sides of the train.
The results are shown in percentage to not misinterpret the consequences. The worst case scenario is summarized in Table 3.
Table 3 Summary of the design parameters and respective values defining the worst case scenario.
Design parameter Value and unit
Fire source 20 MW with a medium fire growth
Number of evacuating personnel 400 personnel Time until evacuation begins 10 minutes Distance between cross tunnels 500 meters
Wind conditions 2 m/s
The result for the worst case scenario reveals that only 7 % of the evacuating personnel reach an exit. It has been found that the safety level, without any safety installations, is not enough to handle a worst case scenario. However as presented in section 2.2 the worst case scenario which represents 100 % of the possible fire scenarios is not required to be managed but instead 90 % of the possible fire scenarios. Since a determination of the 90 % fire scenario is too much work for this thesis, an attempt has been made to evaluate the effect of some of the major alterable design parameters by changing them by 30 % separately, see Table 4.
28
Table 4 The effect some of the alterable design parameters have on the percentage of evacuated personnel sorted smallest to highest effect.
Description Changed value with 30% and
unit
Percentage of evacuated personnel
Time walking in the tunnel by the last personnel to
reach an exit
Fulfillment of BVH 585.30´s requirements
Worst case scenario - 7 % - No
Scenario with a decreased number of evacuating personnel
280 personnel 10 % - No
Scenario with a decreased maximum
heat release
14 MW 100 % 2345 No
Scenario with an
increased air velocity 2,6 m/s 100 % 2103 No
Scenario with a decreased time until
evacuation begins
420 seconds 100 % 1684 Maybe
Scenario with a decreased walking
distance in the tunnel
350 meters 100 % 1541 Maybe
Longitudinal fire gas ventilation systems, both mechanical and natural, must reach and maintain the critical air velocity to function as a dependable fire gas ventilation system.
The air velocity 3,5 m/s has been calculated to be the critical velocity, see Appendix E for calculations. When the worst case scenario is run with TuFT but with an air velocity set to 3,5 m/s, 100 % of the evacuating personnel make it to an exit within the requirements of BVH 585.30.
3.4 Ongoing research
The thesis’ objective second goal is to answer what the ongoing research work within the field is about and if that can result in new solutions of fire ventilation systems in railroad tunnels. Hence an interview with Haukur Ingason from SP Technical Research
29
Institute of Sweden, was carried out to further attain knowledge about ongoing research. See Appendix F for full interview.
During the literature study and the interview with Mr. Ingason, it became clear that there are no major ongoing researches regarding fire gas ventilation systems in railroad tunnels. Fire gas ventilation research is some way or another involved in international collaboration projects which are a lot harder to get money for nowadays. Focus is instead on the fire development in trains and fire gas ventilation systems in underground stations.
30
4 Analysis
4.1 Optimal fire gas ventilation design
An initial fire gas ventilation system selection suggests that a mechanical system or an equivalent non-mechanical system might is needed. This is due to the result generated by the manual calculations and especially the TuFT simulations compared with the requirements by BVH 585.30. A longitudinal fire gas ventilation system which achieves the critical velocity enables for all evacuating passengers to survive and for rescue personal to reach upstream of the fire with ease. However, a cost analysis of the risk reducing options might suggest that a mechanical fire gas ventilation system is one of the least cost effective options since it is an apparatus which need regular maintenance during the tunnel´s life span compared to cross tunnel which basically is a one-time cost.
A fire gas ventilation system is also placed on the last steps in the prioritize order when it comes to risk reducing measures, see Figure 21.
Figure 21 Prioritizing amongst risk reducing means. Fire gas ventilation is on the last two steps. (Banverket, 2007)
There are other means to achieve an acceptable level of safety. The comparison with other tunnels matches the statement since no mechanical ventilation systems have been installed in the majority of Swedish railroad tunnels longer than 1 km without stations.
Hence the optimal fire gas ventilation design for the Hallsberg-Stenkumla tunnel is natural ventilation with no requirements for it to reach and maintain a critical air velocity during a fire scenario.
31
If mechanical fire gas ventilation would be installed, despite of all odds against it, the associated fire detection system must be able to detect a fire and activate it early enough to affect the safety of the evacuating passengers. As in road tunnels, general lightning must be installed for CCTVs to work in order for the control center to choose which direction the air flow should go.
If the wind conditions generate an air velocity similar to the critical velocity, as a
mechanical longitudinal fire gas ventilation system, for the majority of the year it could be classified as a fire gas ventilation system. However, it is not recommended regarding the amount of uncertainties it brings and how the level of safety is reached by other means. During a fire scenario focus should instead be on identifying the actual wind speed and direction inside the tunnel and make information based decisions. An
anemometer linked to the control center could be installed to indicate wind speeds and direction. However, if some kind of fire gas ventilation system is needed by the rescue service to put out smaller fires in a railroad tunnel a mobile fan might suffice.
4.2 Continuation of the design process
A setback in this thesis was the complexity to defining a fire scenario which represents 90 % of all the major fire scenarios. In the comparison between the alterable design parameters it shows that a change in walking distance and time until evacuation yields a notable effect on the time spent walking in the tunnel by the last personnel. However a cost analyses might present different options.
One could set probabilities for of all the alterable design parameters and multiple values to further define all possible fire scenarios larger than 1 MW with a certain probability.
After a probability has been set, a comparison between all the fire scenarios´ risk levels must be performed to be able to rank the fire scenarios. After that, the fire scenario which represents 90 % of all possible fire scenarios larger than 1 MW can be identified and used to test different designs of the tunnel.
The probability for a train to stop in the tunnel due to a fire is 2 % meaning 98 % of the evacuations happen outside of the tunnel (WSP Sverige AB, 2016). The 20 MW fire, 400 passengers, maximum walking distance and time until evacuation begin, all have a
32
certain probability. Basically, the probability for a worst case scenario to happen is almost 0%.
The thought behind the second goal is linked to the risk when designing a level of safety by looking too much at other projects and copying without much thought. Tunnels are not built every day and every tunnel has different design parameters. Since there are no ongoing research regarding fire gas ventilation in railroad tunnels one could study some of the ongoing research carried out for road tunnels for inspiration. Road tunnels are generally equipped with more safety installations.
4.3 Limitations in this thesis
There are several limitations in this thesis which have to be noted:
Ongoing project: The design process of the tunnel is an active process which faces constant changes. The conditions presented in this thesis will most likely be outdated the longer the project goes on. Hence the results from this thesis must be treated accordingly.
Wind conditions: It truly is a design parameter when treated as a variable in the TuFT simulations. It affected the outcome of the survival rate immensely. Hence the usage of 2 m/s as a constant air velocity in the tunnel is needed to be profoundly motivated which it is not. All the affecting factors are needed to be taken into account and somehow calculate a most possible air velocity perhaps depending on the season.
However, the effects on the fire by the air velocity has and cannot been taken into
account in TuFT. A higher wind velocity could potentially generate worse conditions for the personnel downstream in the tunnel since the fire is supplied with more fresh air.
Comparison: It is uncertain if the rescue service depends on the non-mechanical ventilations systems to create a certain air velocity to enable acceptable working conditions during a fire scenario. It is something to keep in mind as it may equal to those tunnels being equipped with a fire gas ventilation system.
The comparison does not cover all railroad tunnels especially those that are being designed or built at this time. Information about railroad tunnels that are being
33
designed or built must be gathered to make a thorough comparison with similar properties.
TuFT: TuFT is based on a one dimensional calculation method which entails that the conditions are uniform across the whole normal section and only varies with the
distance to the fire. The calculation series are carried out every second. These two factors combined makes TuFT´s simulation time very short and enable many scenarios to be simulated, as performed in the results.
Unfortunately the simplifications applied in TuFT´s calculation method generate a too unreliable result and should only be used as an indication in preparation of CFD simulations. For example is the grade of the tunnel not used which both affects the evacuation and the air velocity. CFD simulations are time consuming and must be well thought through and TuFT aides that process.
Using TuFT, the evacuating passengers have a walking speed solely based on the visibility. It has been stated in other reports that, amongst other factors, the width of a walking path affects the walking speed. (Lundström, 2012) A more accurate
representation of the evacuating passengers’ movement during a fire scenario is provided, in combination with the results from CFD, by using STEPS, Pathfinder etc.
instead of TuFT.
Train: This thesis is based on the condition that only one train at a time is in each tunnel.
The consequences could be devastating if one train catches on fire, stops and another train enters the same tunnel. If the train that gets caught on fire is a freight train, which fire could be five times worse than a passenger train fire, and the other train is a
passenger train it is most certain that many personnel will perish. Surely it is a more dangerous situation but will it affect the choice of a fire gas ventilation system. Probably not due to the probability of such event happening is insignificant to motivate a fire gas ventilation system.
Detailed information about the train’s fire properties and detection configuration is needed to truly determine all possible scenarios and associated probabilities and consequences.
34
Evacuating passengers: It is mentioned in the previous section that 99 % of the
evacuating passengers must make it to an exit. This is refereeing to passengers unable to perform an evacuation. For example a wheelchair that is creating a queue or evacuating passengers’ behavior has not been taken into account. Hence a more advanced
evacuation simulation is needed.
Ongoing research: A more in depth search regarding ongoing research is needed since the method in this thesis is too shallow to find smaller ongoing research within the field.
As mentioned earlier one could study some of the ongoing research carried out for road tunnels.
35
5 Conclusion
The optimal fire gas ventilation design for the Hallsberg-Stenkumla tunnel is natural ventilation with no requirements for it to function as a fire gas ventilation system due to the level of safety will be reached by other means. The proposed design is same as the design in the majority of Sweden´s railroad tunnels that are longer than 1 km without stations since it is assumed for that design process to come to the same conclusion.
A continuous analyze consisting of in depth information gathering and usage of more advanced simulation software is required to eliminate or reduce some of the most serious uncertainties.
No major ongoing research about fire gas ventilation systems in railroad tunnels has been identified.
6 Further research
Mr. Ingason expressed that there is a need of a study where the safety in railroad tunnels is analyzed and compared globally with other railroad tunnels, which could be associated with the selected fire gas ventilation system.
There is also a need for research on strategic use of fire ventilation during the early stage of the fire development in railroad tunnels, and in relation to firefighting and use of new innovative solutions such as IR-cameras, robots etc. Also evacuation studies from trains inside a tunnel in relation to the early stage of the fire and when the fire services arrive.
The process to estimate the wind conditions in a railroad tunnel is one of the limitations.
Future research should try to design a standard method to identify the initial wind conditions in a railroad tunnel.
It is difficult to motivate why a risk reducing option is better than the other. A cost analyses regarding risk reducing measures in railroad tunnels are needed.
36
Reference
Asp, M., 2015. Mail conversation: Vindros Örebro. s.l.:SMHI.
Banverket, 2007. Handbok BVH 585.30, Borlänge: Banverket.
Beard, A. & Carvel, R., 2005. The handbook of tunnel fire safety. 1 ed. London: Thomas Telford.
EU, 2014. COMMISSION REGULATION (EU) No 1303/2014 concerning the technical specification for interoperability relationg to 'safety in railway tunnels' of the rail system of the European Union, Brussels: Official Journal of the European Union.
European Union, 2008. DIRECTIVE 2008/57/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL.
Fridolf, K., 2015. Rail tunnel evacuation. Lund: Lunds universitet.
Fridolf, K. & Frantzich, H., 2014. TuFT: Tunnel Fire Tools - Teknisk dokumentation, Lund:
Lunds universitet.
Frisk, G., 2015. Mail conversation: Information tågtyp, s.l.: Tågkompaniet.
Hedlund, T., 2015. Mail conversation: FW: Hallsberg-Stenkumla - Tågtyper. s.l.:Trafikverket.
Hägglund, B., 2012. CFD-beräkningar med CFX, s.l.: Trafikverket.
Ingason, H., 2015. Telephone [Interview] (12 August 2015).
Ingason, H., 2016. Mail conversation: SV: Exjobb brandgasventilation jvgtunnel. s.l.:s.n.
Ingason, H., Lönnermark, A. & Yi, Y. Z., 2015. Tunnel Fire Dynamics. Springer New York:
Springer.
Lilliengren, U., 2015. Mail conversation: VB Hall-sten. Stockholm: WSP.
Lundman, P., Wilhelmsson, A., Andrén, A. & Olofsson, O., 2009. Personsäkerhet i tunnlar Inventering. s.l.:Banverket.
37
Lundström, J. A. F. V., 2012. Tunnelutrymning - Effekten av gångbanans bredd på förflyttningshastighet vid utrymning i en spårtunnel, Lund: Lund University.
Lönnermark, A. & Ingason, H., 2013. Recent Achievements Regarding Heat Release and Temperatures during Fires in Tunnels, Borås: SP Swedish National Testing and Research Institute.
Nordwall, E., 2015. Telephone [Interview] (17 Augusti 2015).
Näringsdepartementet, 2011. Plan- och byggförordningen, s.l.: s.n.
pixabay, 2015. pixabay. [Online]
Available at: https://pixabay.com/sv/tunnel-sp%C3%A5r-1041120/
[Accessed 26 5 2016].
Svensson, S., 2000. Fire Ventilation. s.l.:Davidsons Tryckeri AB, ISBN: 91-7253-279-3.
Trafikverket, 2011. TRVK Tunnel 11 Trafikverkets tekniska krav Tunnel TRV publ nr 2011:087, s.l.: Trafikverket.
Trafikverket, 2011. TRVR Tunnel 11 Trafikverkets tekniska råd Tunnel TRV publ nr 2011:088, s.l.: Trafikverket.
Trafikverket, 2014. Järnvägsplan - val av lokaliseringsalternativ, s.l.: Trafikverket.
Trafikverket, 2014. Trafikverket. [Online]
Available at:
http://www.trafikverket.se/contentassets/d60e630b0cd54363a130fd0b81415461/flygfoto_
vald_korridor_hallsberg_stenkumla.pdf [Accessed 31 10 2015].
Trafikverket, 2015. Järnvägsplan-Utredning Hallsberg-Stenkumla PM Dimensionerande brandeffektkurvor, s.l.: Trafikverket.
Trafikverket, 2016. Trafikverket. [Online]
Available at: http://www.trafikverket.se/nara-dig/Orebro/projekt-i-orebros-
38 lan/hallsberg-degeron/hallsberg-stenkumla/
[Accessed 26 5 2016].
WSP Sverige AB, 2016. Riskbedömning tunnel, Eskilstuna: Trafikverket.
WSP, 2015. Normalsektioner, s.l.: WSP Berg.
WSP, 2015. Spårlinje Alt U2, s.l.: WSP Spår.
WSP, 2016. Hallsberg-Stenkumla. s.l.:Powerpoint presentation.
39
Appendix A - SMHI predominant wind report
40
41
