Verification Testing of Evacuationz, a Coarse Network Evacuation Model

Abstract

Several evacuation models exist in the market. It is an issue for both the user and developer to evaluate the ease of use and the quality of the model results. Verification is the process of determining that the results of a calculation method match with the developer's conceptual modelling representation expectations, i.e., the modelling implementation is accurate. Different testing procedures are available for the verification of evacuation models/simulators. Recently, International Organization for Standardization (ISO) released ISO:20414, a standard procedure including a detailed list of verification tests. It includes testing of 4 components – Basic components, Behavioural components, Fire-people interaction Components, and Building-specific Components.
This thesis focuses on applying the verification tests included in ISO:20414 to the fire evacuation model Evacuationz. This software is a coarse network model using pseudo-random sampling from distributions and it can be used in the context of fire risk assessment. The scope of this work is dual, namely 1) to perform a systematic verification of the Evacuationz model using ISO:20414, 2) to evaluate the applicability of ISO:20414 to coarse network models, being this the first time, this testing procedure document is applied for such type of models.
In this thesis, a total of 21 verification tests of ISO standard are considered. Nevertheless, 16 verification tests were successfully conducted and achieved desirable results. The majority of the tests were performed explicitly, but movement around a corner was performed implicitly because corners cannot be represented directly in coarse network models. Since the version under consideration of the software does not support horizontal counterflows, one test did not yield the expected result. Also, four tests could not be conducted because of the software limitations, i.e., features such as disabling exits, showing occupant incapacitation using FED, or the inclusion of modern building evacuation features such as Lifts and Escalators were not present in the version of the software in use. Overall, it can be concluded that the ISO:20414 procedure concerning verification testing is considered applicable to coarse network models.

Popular Abstract

Evacuation models are tools used by fire engineers to evaluate and check the suitability of evacuation design on the building layouts. The models are used to estimate the time required for people to escape from a building to a safe location in the event of an emergency. There are numerous models on the market to calculate the required time. It is difficult to judge whether the results of these models are accurate or not. Therefore, to establish the accuracy of the results the models must be tested. Verification is the process to check that the results of calculation methods match with the implementation of the conceptual representation – that the model does what was intended. Over a period of time, different verification test methods have been developed. At this time, there are four verification test procedures, and a document released by the International Organization for Standardization is the latest procedure for verification released (ISO:20414). The focal point of the thesis is to implement the ISO verification test procedures in a coarse network model. This is the first time that the ISO test procedure has been applied to this type of model. A coarse network model uses a node system to represent simple and complex structures of buildings. People/occupants are considered as points in the node system (people representation does not have any real characteristics but it is shown as a dot), and the model user directs each movement of the occupant in the node. Evacuationz software which is a coarse network model was chosen for this thesis. In this model there are six input files to run a simulation. The input files are named as Map, Populate, Agent type, Exit Behaviour, Simulation and Scenario. Every file has a unique attribute for the simulation. In the ISO publication, there are a total of twenty-one verification tests. These tests are broken down into four categories: Basic components, Behavioural components, Fire-People Interaction components, and Building-Specific components. Basic components consist of 13 tests that primarily focus on the testing of basic features that most models possess, such as maintaining the walking speed of occupants in a corridor and stairs checking whether occupants can walk to specified exits when directed to them, and many more. The second component is the Behaviour component which consists
of four tests and focuses on the behaviour of the building's occupants in the event of an emergency. The third component, Fire-People Interaction, consists of two verification tests. Its primary goal is to assess the impact of the fire and smoke on those who are trying to escape the premises. The effects include a decrease in the walking speed and the amount of smoke inhaled by people. The last component is the Building-Specific component whose goal is to determine whether the model has the capability of evacuation via lifts and escalators.
After running all the verification tests on the Evacuationz model, it was found that only seventeen out of twenty-one tests could even be performed. The remaining four tests were not performed given the absence of certain features. These tests will be successful if the specific features such as addition of ramp, lift, escalators, representation of Fractional Effective Dose (FED), and disabling an exit can be
added to the model. A sub-set of the ISO:20414 tests are applicable given the nature of this type of network model. The outcome of performed tests were compared to the expected results specified in the ISO document and desirable results were obtained. Hence, it can be concluded that the ISO:20414 procedure for verification testing is applicable to coarse network models.

• author: Gupta, Antariksh
• supervisor: Enrico Ronchi
• organization: Division of Fire Safety Engineering
• course: VBRM05 20221
• year: 2022
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering
• publication/series: LUTVDG/TVBB
• report number: 5676
• other publication id: LUTVDG/TVBB-5676-SE
• language: English
• id: 9098785
• date added to LUP: 2022-09-01 09:05:28
• date last changed: 2022-09-01 09:05:28