Research on human traffic flows parameters at metro stations gates

. The country's largest cities are working intensively on the development of underground space in order to utilise their territory more fully and efficiently. Increase of fire hazard for people in underground space is determined both by the difficulties of evacuation of large masses of people of different mobility upwards, requiring, in particular, the use of means of vertical transport, and the peculiarities of the dynamics of CPF in a confined space. For designing of underground space objects with floor depth below 15m special technical specifications (STS) are developed. However, the authors of STS have no substantiated data for their development, as the scientific research necessary for their obtaining has not been carried out. Therefore, the supervisory authorities have no opportunity to analyse the degree of validity of the proposed solutions and the corresponding fire risk assessment. The article presents the results of research on the movement of people at metro stations through the "underground" type exit doors, allowing to assess the probability of crush formation during evacuation and to propose solutions to ensure unimpeded movement.


Introduction
Subways do not belong to any class of functional fire safety and according to part 1 of Art.48. 1 of the Town Planning Code refers to particularly dangerous, technically complex objects.Consequently, the safety of people in pedestrian underground structures (stations and interchange hubs) should, according to Federal Law No. 123 [1], be determined by a calculation based on the assessment of individual fire risk.This concept also corresponds to the requirements of paragraph 5.16.6.17 of BC 120.13330.2012[2]: "The sufficiency of design solutions to ensure safe evacuation of people at the station should be assessed by calculation.... When calculating the evacuation time, all protected evacuation routes are taken into account".
At metro stations, evacuation routes include platforms, escalators, turnstiles, as well as the area in front of the turnstiles, which create an obstacle when people exit.It is not prohibited to install turnstiles and other devices on evacuation routes, if these devices are duplicated by ordinary swing doors or are provided with devices (systems) allowing to unlock (open) the doors in case of fire automatically, remotely and manually.The safety of people in case of fire is determined by the criteria for ensuring timely and unimpeded evacuation [1].To assess these criteria, various models of human flows and fire development are used [3].At the same time, the choice of the model of human movement and the list of input data determine the reliability of the obtained results and their adequacy to the real picture of the process of human movement.As initial data for the assessment of human safety criteria are used parameters of the movement of mixed human flows of the main contingent of the object under consideration.But BC 120.13330.2012 [2], nor the Methodology [3] do not give the calculated values of the parameters of the movement of people through the doorways of "Metro" type, necessary for an adequate assessment of people safety.

Methods and materials
According to para.5.4.1.22 of BC 120.13330.2012[2], when the fire alarm system is activated, for safe evacuation of people, the doors of lobbies leading outside should have devices for fixing in the open position.The estimated number of evacuees should be determined on the basis of the maximum load of the train during rush hour.
As it is known, metro entrance doors are glass and wooden.They differ in their dimensions: wooden doors are larger than glass doors.
To investigate the capacity of the doors it is necessary to take measurements at maximum density of people.For the experiment purity the stations were chosen, which are located in different districts of Moscow city, having different nearby infrastructure, depth of laying and years of construction.The stations "Tsaritsyno" (as from this station there is a transition to D2 -second diameter) and "Komsomolskaya" (as this station leads to the area of three railway stations Yaroslavsky, Kazansky and Leningradsky) were chosen.These stations are among the busiest in Moscow.
The experiment was conducted by the visual method of research, which was described in the works of Professor S.V. Belyaev [4] and Professor A.I. Milinsky [5].
The methodology of this observation is as follows: an observer takes his/her place near the doorway a few minutes before the passenger flow leaves the metro station.During the human flow, the observer selects one door and sets a stopwatch for 1 minute.Then he selects one person in the flow and, when this person reaches the boundary of the beginning of the experimental area, the stopwatch is switched on.In 1 minute, the number of people who have passed through the doorway must be counted.When the time expires, the stopwatch is switched off and the value of the number of people is entered into the table.This is the end of the first measurement procedure.Then the person in the flow is selected again and the procedure is repeated.
It is possible to determine the capacity of the doorway at maximum flow density D=9 person/m2 by this experiment.
In order to construct the dependence graphs necessary for the capacity study, field observations were made at other density values.For this purpose, the method of video recording was used [6].A camera is installed above the doors and video of the human flow is recorded.
The next step is the construction of a volumetric scale grid on the video, the mesh sizes of which are equal to 1x1m (Fig. 1).Square metres in front of the doors were measured in advance to determine the intensity of people's movement at different densities of the human flow in front of the doors.The results of the survey were processed as follows.A grid square was selected and the time for each person to traverse a distance of 1 m was determined.If there was one person in the grid square, the time was recorded for the density D=1person/m 2 .This is how the human movement time is calculated for each interval of human flow density.Based on the time of human movement, the speed of human movement and further the intensity of movement at different densities is determined by multiplying the density of human flow by its speed.

Gathering and systematising empirical data
The results of the conducted studies are shown in Tables 1 and 2. To assess the adequacy of the obtained research results, statistical processing of the empirical data was carried out.The textbook [7] shows that the approximation coefficient R describes the general tendency of the influence of the human flow density on its speed.
The empirical values of R for each density interval in all series of field observations were determined by the formula: , here  0 is a random value of free movement speed (in the absence of influence of surrounding people), depending on the type of path and emotional state of people, m/min;  defines human flow speed at a certain density, m/min.The results of approximation are presented in Fig. 3, 4.

Results
The graphs of dependence of human traffic intensity on density are presented in Fig. 5 and Fig. 6 for wooden and glass doorways, respectively.The process modelling of people evacuation through the underground doorways was carried out for comparative analysis of the obtained results with the existing data.Modelling of the evacuation process was carried out in the "Pathfinder" programme, which allows to calculate the evacuation time and the time of existence of crowds according to the individualflow model of movement.The area of horizontal projection of people was taken as f=0.125m 2 .The number of people accepted was 500.
When modelling the evacuation, it was assumed that all exit doors were open.Four variants of evacuation of people from the underground lobby were modelled: 1) According to the capacity of wooden doors obtained from the results of the study 2) By the capacity of glass doors obtained from the study results 3) According to the capacity from the Methodology

Modelling results and comparative analysis of the obtained data
In the course of numerical modelling, the estimated values of evacuation times for four scenarios were obtained for different capacities of doorways.The results are shown in Table 1.The calculations are estimates, and do not take into account the emotional state of people and other processes that complicate the evacuation of people from such structures.

Conclusions
1) Full-scale observations by means of visual method and video shooting method allowed to establish more than 600 values of human flow speed in front of pendulum doorways of "metro" type in density intervals from 1 to 9 persons/m2.
2) Statistical analysis of the results obtained by means of video shooting allowed establishing the dependence between the parameters of human traffic flow, which allowed establishing the intensity at maximum density of 9 persons/m2 for wooden doors -5.14 persons/m2*min, for glass doors -6.19 persons/m2*min.
3) Modelling of evacuation of people in the "Pathfinder" programme for different scenarios allowed carrying out a comparative analysis of the obtained results of field observations with the existing data in the normative documents.The results of the modelling showed that the longest evacuation is more than 10 minutes through wooden pendulum doorways, which in turn affects the safety of people when assessing the probability of evacuation at the stages of design and operation of metro stations.

Fig. 3
Fig.3 Approximating theoretical dependence of R on density D (for wooden doorways)

Table 1 .
Results of calculating velocities at different densities (wooden doorways)

Table 2 .
Results of calculating velocities at different densities (glass doorways)

Table 1 .
Comparison of evacuation time for different scenarios