The comparison is divided into two tests. In Test 1, no behaviour
is performed in order to check that the simulation of movement is
working satisfactorily. No pre-movement is considered and a fixed
unimpeded walking speed of 1 m/s is assigned to all of the tunnel
users. In Test 2, behavioural parameters were implemented. In this
test, the scenario was run 100 times by the evacuation models to
capture stochastic variations in the results. GridFlow, STEPS and
PathFinder do not represent the ‘‘domino effect’’ of the pre-movement time used in EvacTunnel. In order to implement the same evacuation characteristics closely, the pre-movement time was assigned in these models by increasing the mean and standard deviation every 20 m (the tunnel is divided into 13 zones). The same unimpeded walking speed distribution has been assigned for all tunnel´s users in all models. This is a normal distribution with a mean value of 1.20 m/s and a standard deviation of 0.20 m/s