by Yasmine Hifi, Safety at Sea Ltd
[Originally published in The Naval Architect, October 2019]
IMO rules for evacuation analysis have, until recently, only applied to Ro-Pax ships. However, from January 2020, all passenger ships (in addition to Ro-Ro passenger ships) carrying more than 36 people and with keel laid on or after the same date will have to prove adequate evacuation times through an evacuation analysis. Shipyards or designers will need to show that their design meets the IMO performance standard which is the time taken to both muster and abandon ship. This performance standard is defined as 60 minutes for Ro-Pax ships or ships with three main vertical zones or less, and 80 minutes for ships with more than three main vertical zones.
The main purpose of an evacuation analysis is to verify that a specific ship design will allow for people onboard to get to assembly stations in an orderly and timely manner and hence estimating the time required to abandon ship. Furthermore, an evacuation analysis can be integrated into the design cycle by highlighting areas of congestion or possible bottlenecks that can then be addressed.
The IMO mandates that the analysis to estimate the travel duration to reach muster or assembly areas onboard the ship can be performed using either a simplified or an advanced method. It is assumed that the preparation and launching of the lifesaving appliances would overlap with the assembly process and is taken to be 30 minutes unless data from full-scale trials, manufactures or simulations are available and can be used instead.
Simplified analysis
The simplified analysis provides a basic approach for calculating the required evacuation performance parameters required by the IMO and is most suitable at the early stages of ship design or for low complexity ship layouts. ‘Simplified’, however, does not necessarily mean less time consuming.
This methodology is based on defining the escape routes within the vessel as a hydraulic network, in which the public spaces are the tanks, the corridors and stairs are the pipes and doors and other possible restrictions are defined as the valves. Several parameters such as ‘clear width’ and ‘specific flows’ are to be considered when defining the hydraulic network. Calculations for the simplified method could, hence, be performed within a spreadsheet.
Although this methodology is acceptable, it may not provide a time saving as expected, the calculation output may be difficult to validate, and modelling people movements as hydraulic flows can only offer a coarse view of what may happen during an evacuation. The benefits and the availability of tools on the market to perform the advanced evacuation simulation make it a much more attractive option.
Advanced analysis
The advanced analysis is based on performing computer simulations of the evacuation process: the ship geometry is modelled as built and passengers onboard are represented as individuals, referred to as ‘agents’ within the simulation.
The model of the ship can be defined with a high level of detail and can easily be maintained and updated during the design cycle, from the very early design stage with a basic layout all the way to a very detailed final layout. People onboard, modelled as individuals, have unique characteristics (mainly represented by probability distributions) and can move towards specifics spaces whilst avoiding each other and other obstacles along the way.
Unlike a hydraulic model, a digital model of the ship offers a representation close to reality where not only escape routes are modelled but all spaces from passengers and crew cabins, to crew service and public areas are represented and are easily recognisable when viewing the ship model in 2D or 3D.
Performing an evacuation analysis using this methodology offers a host of additional benefits. Firstly, identifying areas of congestions and bottlenecks, which is an important part of the regulation, as well as location of cross and counterflow becomes much easier and more intuitive. Furthermore, as more simulations are performed, specific issues or areas of the layout can be investigated and modified to further optimise the design.
For any software to be accepted for use for as an evacuation analysis by the IMO, it needs, as a minimum, to demonstrate that it passes 12 specific benchmark cases defined in the regulation. These tests are designed to verify that differing components of the program are performing as intended, thus ensuring that the software is capable of modelling some key characteristics and behaviours of passenger dynamics.
Grid-based or continuous space methodologies
There are two main approaches on the market today to modelling a ship using simulations: grid-based and continuous space modelling. Within the grid-based approach, the ship layout is simplified to a series of uniform grids for each deck. Each agent occupies one grid cell and makes their way to their destination by moving from cell to cell. Some variations also exist where an agent would occupy more than one cell and others where multiple agents can occupy a larger grid cell. Cells can be free, occupied or not allowed, representing walls, furniture or any other obstacles. An agent can only move to a free adjacent cell.
The main advantage of this approach is in the simplification of the overall simulation and hence the overall calculations, with, for example, simple transition and collision avoidance rules as valid or allowable movement is limited by the number of available free cells.
The major disadvantage, however, is the accumulation of errors introduced by the grid representation of the model: as the grid cell size is fixed (usually 0.4×0.4m2 or 0.5×0.5 m2), the model of the ship geometry would need to be distorted to fit the grid. Additionally, the grid system also digitises the speed of motion as it can only vary (increase or decrease) in discrete steps governed by the size of the cell.
The continuous space approach instead produces a representation closer to reality as the geometry does not have to fit a predetermined grid. People can occupy any point in space as long as the space is not already used by another person and can move freely in any direction with a speed of motion only affected by the simulation time step. The range and direction of motion are restricted when people are in the close vicinity of the moving person to avoid collision. Collision avoidance calculation is more complex in this approach, but this does not present a challenge considering the computing capability available today.
How is an evacuation assessment performed?
To perform an evacuation analysis a number of specific scenarios have to be assessed, each of which must be repeated multiple times (known as ‘batch runs’) in order to ensure statistical significance due, in part, to the fact that agents are defined through probability distributions. Ultimately, the 95th percentile of all the estimated travel times is used to calculate the total evacuation duration.
There are four scenarios that must be simulated, each with a specific distribution of initial location of passengers and crew, reaction time and walking speed. They are split into night and day scenarios, each with a primary and a secondary case. The primary cases assume all escape routes to be available. The secondary cases are focused on the assessment of impaired escape routes and they are identified based on the results of the primary cases.
Two further additional scenarios are to be assessed depending on the type of vessel and design:
- The open deck scenario which is applicable only if the vessel has an open deck area of more than 400m2 for use by passengers. The open deck scenario would be assessed as a day case.
- The embarkation scenario, if the design locates the assembly stations away from the embarkation stations. If this scenario is simulated, then the estimated times to the embarkation stations should be used in the overall calculated evacuation duration instead of the default 30 minutes.
Meeting regulatory requirements and beyond
Brookes Bell welcomes these changes in regulation, and not purely as a producer of evacuation analysis software, EVI™, but because widening the applicability of evacuation analysis and increasing the focus given to evacuation considerations on the design and layout of passenger vessels can only improve the overall safety of more and more passenger vessels.
This article has focused on providing a high-level description of the impact of the regulations, the differing options available for designers in terms of tools available on the market, and the process of performing an evacuation analysis. Brookes Bell has been involved in passenger vessel design since the late 1990s and provides the market with its advanced evacuation analysis tool EVI™.
The benefits of pedestrian dynamics software are not limited to evacuation analysis only but can be used in assessing ‘what if’ scenarios and simulating situations that involve movements of people. Evi™ has indeed been used in a great variety of applications, from investigating ferries turnaround times while in port, escape planning from a dockyard during shipbuilding, to helping develop crew procedures during mustering and to improving passenger comfort through assessing service levels in public areas. These benefits go beyond meeting regulations to looking at different aspects of the customer experience.