: In the last few years, serious accidents including many casualties due to the explosion of gushing natural gas and fire in tunnelling work occurred one after the other. As this type of accident has been uncommon to tunnelling work in Japan up to the present, there has been insufficient attention given to planning for safety in case of gas explosion and fire.
    Due to the increasing trend of construction of tunnels, it may be thought that accidents will continue to occur if preventive countermeasures against gas explosion and fire are not implemented.
    Thus, at the request of the labour administration, the Research Institute of Industrial Safety has planned to perform a special research project in order to establish the countermeasures against gas explosion and fire in tunnelling work.
    The basic countermeasures to prevent these types of accidents are considered as follows ;

1. Ventilation which ensures a safe concentration of gas in a tunnel
2. Methane detectors to check on the gas concentration and gas alarms which sound a warning if the gas concentration exceeds a certain risk threshold in a tunnel
3. Effective management of ignition resources to avoid explosion even if the gas concentration exceeds a risk threshold in a tunnel
4. Emergency countermeasures to ensure the security of workers when the risk of explosion is imminent in a tunnel
5. Detection of strata containing natural gas and the monitoring of the gas which enable workers to execute the emergency countermeasures quickly in a tunnel

    This special research project has been conducted in cooperation with all divisions of the research institute for the purpose of providing information about the methodology and way of thinking which are necessary for executing these countermeasures.
    The results would be though to be very useful to the personnel concerned with tunnel construction work. Also, in order to perform these countermeasures most efficiently, it is necessary to execute them synthetically and comprehensively as a system, since the above mentioned countermeasures are not independent of each other.
    The following are the details of each research work.

: To improve the ventilation systems used by the tunneling works where methane gushes out, the experiments in the diffusion of methane with model tunnels and the examinations of the systems are practiced.
    In this study, model rules are set up as follows : (i) Geometric similarity. Two model tunnels, which have same cross sections shaped such as the combination of the half-circle whose radius is H/2 and the rectangular form (H x H/2), and whose length is 9H - 12H, are used. The value of H in the large model is 2 [m], in small one it is 0.5 [m]. H is taken as the representative length in this model. (ii) Main π numbers of the physical quantities. As the temperature in the tunnels is assumed to be constant, so main π numbers that must be agree between the models and real things are Ret (turbulent Reynolds number), Fr (Froud number), and Pet (turbulent Peclet number). (iii) Another dimentionless numbers rekted to the representative velocity of the ventilations and the gas gush, density of the gushing gas, and the representative concentration of methane in it are considered.
    The experiment is mainly assumed that methane gushes out from the working face, the bottom or the ceiling near the face in the tunnels, and the distributions of methane concentration are examined in several conditions. These conditions are the representative concentration of methane, the distance A between working face and the tip of the wind duct, the effect of the supports, the representative velocity of the ventilations, the positions of the gas gush points and wind ducts, the ventilation type that is discharge or exhaust, and the presence of the obstacles in it.
    In each cases, the dangerous areas where the concentration of methane by vol. is above 5 [%] are grasped. It is confirmed that under the conditions that the distance A is under 4H the dilution of methane is good, if the distance A is above 5H the dilution of methane is bad. It is also recognized that when the representative concentration of methane is under 1.5 [%] the methane gushing from the place near the working face is completely diluted at the place whose distance from the working face is equal to A, although some areas above 5 [%] exist between the tip of the wind duct and the gas gush points.
    It is reported that the π number Ret and Pet are satisfied automatically if the air current in the space is well disturbed. And the experiments are performed based on this assumption, but it is not clear if the air current near the working face is well disturbed or not, so it is necessary to check up the realization of the model rules. Then if the large model is assumed to be a real thing, the small one is thought to be a quarter model of large one, some pilot experiments are performed by using these two models. And the result of the experiments shows that the similarity is nearly satisfied from the case of small velocity of the ventilations to the case of relatively high velocity, the adequacy of the model rules is confirmed.
    On the basis of these consequences of the experiments and the data of the field studies, some examinations are discussed about the ventilation systems used in the tunnels where methane gushes out.

: Explosions of methane are the most serious disasters encountered in tunnel works carried out in such strata containing natural gases. One of the effective methods to avoid such explosion hazard is air ventilation through a tunnel under construction. Although there have been used several types of air ventilating system, none of them is said to be satisfactory from the viewpoint of preventing methane accumulation. A suitable gas alarm system, therefore, should be introduced into every tunnel works in order that methane concentration in air flow is monitored continuously and that warning alarms are given, if methane concentration in the tunnel exceeds a certain threshold, so that any adequate precautions are to be taken successively. The gas alarm system for such temporary works as tunnelling should be not only that to ensure safety, but also be economical. This report describes the method of constituting a gas alarm system for practical use in tunnel works.
    In the first part of the report, response characteristics of several commercial gas alarm systems are studied to obtain informations on the effect of structural components of the methane detector on response time of the system. Experiments are made using a rectangular test chamber in which a known mixture of methane and air is prepared, and into which a methane detector of catalytic combustion-type (Fig.3.1) is thrown vertically. Output voltages from an amplifier have been calibrated by known mixtures. Two kinds of response time are used as indices of response characteristics (Fig.3.2): one is the time for the detected concentration to attain a certain absolute value (e.g. 1.5% methane), and the other is the time to attain a certain relative value (e.g. 90% of the methane concentration of the mixture in the test chamber). The former response times increase with methane concentration detected, but the latter ones are, in so far as relative concentrations are less than 60%, independent of methane mixtures in the test chamber (Fig.3.3). This fact means that the response time might be largely increased if the concentration for alarm is set up inadequately (Fig.3.4). Thus, in determining the alarm level of concentration, response characteristics of the gas alarm system in use should be taken into account. Other experiments are made to know effects of such structural components of a detector as flame arrester, protective guard and waterproof cover on response time (Fig.3.5). Flame arrester, made of sintered metal, and waterproof cover are shown to have ill effect on response time. Some structural improvements of commercial detectors are suggested.
    Even when an adequate gas alarm system is introduced into the tunnel, it is of no use if the system could not detect the existence of gushed out methane. Here, the most important factor is how to install methane detectors in the tunnel.
    The second part of this report describes the experiments carried out, using a simulative tunnel of nearly full-scale (Fig.3.6), for the purpose of obtaining informations on optimum installation of detectors. Methane of constant rates, up to 0.2 m^3/min., is flown into the tunnel through either of three openings on the terminal plate which is simulating a face of tunnel, the other end of the tunnel being opened to the atmosphere. Distributions of methane in the tunnel are measured by the use of the commercial gas alarm system composed of 17 units of gas alarm and detector. Types of air ventilation are as follows: air blowing by either of main duct or sub-duct, air suction by main duct, suction by main duct combined with air blowing by sub-duct, and with no ventilation. In each experiment about 10 minutes are to be passed until a nearly steady state of ventilation is attained, even though there exist some fluctuations of methane concentration. The maximum value recorded in this period is used as an index of methane concentration for each point of measurement in the tunnel. Parameters of experiments are type of ventilation, flow rate of ventilating air (Q_AB or Q_AS), flow rate of methane (Q_M), position of the opening for methane inlet, and initial velocity of methane flown-in (V_M). From those experiments, distributions of methane layers are drawn as "contour maps of concentration" for vertical and horizontal directions. In most of those maps boundary lines for 1.5% methane are drawn for reference by thick line, because such a concentration (i.e. 30% of the lower explosion limit of methane in air) have been considered as a marginal value for the prevention of explosions. Figures 3.8 - 3.15 and Tables 3.2 - 3.4 are results for air blowing ventilation by main duct. Figures 3.16 - 3.18 are drawn from results of experiment for air suction ventilation by main duct. In Table 3.5, effects of air suction ventilation combined with eccentric air blowing by sub-duct are compared with those of air blowing ventilation by sub-duct only. With no ventilation in the tunnel, results shown in Figures 3.21 - 3.25 are obtained, including an interesting discussion on the stretching velocity of methane layer. Because of its lower density than air, methane tends to accumulate near the roof of tunnel, and, therefore, the qualitative behaviors of methane in the tunnel are understood rather easily. However, the thickness and the length (i.e. vertical and horizontal gradient of methane concentration) are largely affected by the type of air ventilation as same as by flow rate of air and methane and other factors. It is concluded, then, that the systematic installations of detectors could not be unified quantitatively over every tunnel works, but should be determined for individual tunnel, taking into account such factors as type and efficiency of air ventilation, total number of available detectors, number of spots at where methane gushes out constantly, and alarm level of methane concentration for each detector.
    Discussions are also made on the application of results of our experiments in the simulative tunnel to other scaled-up tunnels, methods of constituting an adequate gas alarm system, precautions for the use and maintenance of the system, and on merits or demerits of each type of air ventilating system.

: It may be considered that the probability of occurrence of accidents due to fire in tunnelling work has been increasing since the introduction of the increased use of combustible materials and the increment of the activities using fire in the construction process. In fact, not a few serious accidents due to fire in tunnelling work have taken place recently.
    However, to date, there has been insufficient fire safety consideration in tunnelling work ; therefore this paper deals with preventive countermeasures against fire in tunnels, especially in the construction stage.
    Firstly, since a series of countermeasures against fire can be considered as a system, a plan for a fire safety system in tunnelling work was drawn up. The framed fire safety system in tunnelling work includes the following three kinds of sub-systems as a whole.
  S₁ : prevention of occurrence of fire in tunnels
  S₂ : prevention of fire spread, fire extinguishing, and smoke control
  S₃ : workers' emergency escape and rescue of workers
    Secondly, in order to pursue the objectives of the fire safety system in tunnelling work, basic requirements for each sub-system were pointed out.
    Finally, for the purpose of preventing fire in tunnelling work, it may be concluded that various types of safety programs related to fire safety system should be carried out simultaneously and comprehensively in accordance with individual tunnel characteristics. And, though accidents due to fire are occurring in the construction process stage, these countermeasures should be prepared for not only the construction process stage but also the design, cost estimate, and execution planning stage.

: Since the increment of construction of tunnels, the necessity of excavation of tunnels where the stratums contain natural gas, is increasing.
    And in the last few years, there were grave accidents caused by the explosion of gushing natural gas.
    At the sites of tunnelling works, there are many factors concerning phenomena of gushing natural gas, and it is difficult to explain or estimate these phenomena sufficiently.
    Therefore in order to establish preventive countermeasures against the accidents caused by gushing natural gas, it is necessary to investigate the actual condition of tunnelling works.
    In this study, firstly we investigated by means of field investigations about the actual condition of many tunnelling works in which the risk of gushing natural gas existed.
    And secondly we discussed mainly about the geological characteristics of stratums containing natural gas and finally the usefulness of the pilot-boring as one of the method of detection of gushing natural gas was discussed.
    The results of this investigation are as follows ;

• ⅰ) There are two kinds of natural gas with which the tunnelling work has a possible to encounter; the first is the natural gas dissolved in water and the second is the oil field gas.
• ⅱ) According to the consideration about the geologic condition, we can specify, to a certain extent, the stratums which may cause the gushing of natural gas.
• ⅲ) In the sites of tunnelling work, there are three principal countermeasures against the gushing of natural gas ; the efficient management of fire, the previous detection of natural gas and the sufficient ventilation of the site. And the execution of pilot-boring is thought to be very useful for previous detecting of natural gas.
• ⅳ) The method of the execution of pilot-boring should be determined by taking into considerations ; the situation of working site, the condition of the geology and so on.

: In tunnelling workplaces where methane or other combustible gas/vapor wells out, it is possible that explosion or fire may be caused by electrical spark or heating of electrical equipments. One of the prevention methods for those hazards is to use the appropriate explosion-protected apparatus in the tunnel construction.
    Several criterions for explosion-protected electrical apparatus have been already proposed and applied to electrical equipments used in general industries. But those criterions are not necessarily to apply to electrical equipments for the tunnel construction. Consequently, it is the purpose of this research to develop out new explosion-protected electrical apparatus and to draft criterions for electrical equipments for use in the tunnelling workplaces.
    Following research works on developments of explosion-protected electrical apparatus for the tunnel construction were made ;

1. Investigations about the real circumstances in tunnels, the construction methods of tunnels and the kinds of electrical equipments used in the tunnelling workplaces.
2. Consideration of the specifications on explosion-protected electrical apparatus (flame-proof and increased safety types) for the tunnel construction.
3. Designs and trial make-up of explosion-protected electrical apparatus (incandescent lamp and motor) for use in the tunnelling workplaces.

    After many tests and investigations on the trial make-up, it has been confirmed that those electrical equipments have faculty of explosion-protected electrical apparatus for use in the tunnel sites. From the results obtained above, the manufacturing specifications and the testing methods required as explosion-protected electrical apparatus for the tunnel construction have been established.