CRUISESHIP TECHNOLOGY
The population used in these simulations consists of 1300 passengers representing 75% of the maximum capacity of the vessel as specified by the IMO guidelines. The primary purpose of this investigation is
to examine the impact that the response time distribution has on the overall evacuation performance. Thus several scenarios are considered which are identical with the exception of the response time distribution imposed on the passenger population. The base scenario investigated in this paper is the IMO ‘day’ scenario. Here, all passengers are initially distributed throughout the service areas of the vessel; i.e. around the bars, dining areas etc and passengers are not located in their cabins. The population is divided amongst the three vertical fire zones as shown in Table 1. The sections of the vessel with no passengers are typically cabin sections or parts of the vessel that passengers cannot occupy. At the start of the simulation, once the passengers' assigned response time has expired they move off to the assembly deck (Deck 8) via the shortest route. Several different response time distributions
were investigated. The first was the standard IMO specified response time distribution. This consists of a uniform random distribution of response times with a lower limit of 210 sec and an upper limit of 390 seconds. This response time distribution is referred to as Response Time Distribution 1 or RTD1. To demonstrate the consequence of using
more realistic non-uniform distributions several other response time curves derived from the building industry were also implemented. While the actual response times may very
well be different for sea based applications, it is possible (and indeed very likely) that the shape of the distribution may be similar for land and sea based applications. Response times varied
from 4 seconds to 110 seconds. As can be seen most people respond within 60 seconds of hearing the alarm. The range of response times is very different to that found in the IMO day case distribution. To gauge the impact of the shape of the distribution on the evacuation performance, the land based response time distribution was modified to fit the response time range specified in the IMO day case. Effectively, 210 seconds was added to all of the times recorded in the land curve and an extra point was then appended to the curve in order to make the curve fit exactly the same range as the IMO response distribution. This point extended the curve from 320 seconds to 390 seconds. The modified land based curve is positively
skewed. This means that the majority of people move during the first 50% of the range of response values, whereas in the IMO distribution, an equal number of people move in each time interval. As set out in the MSC 1033, this and other
scenarios were simulated 50 times in order to generate a distribution of results. After every five simulations the passengers' starting locations were swapped. The Cumulative Congestion time for each
person in a particular simulation can be averaged to produce the Average Cumulative Congestion time. This then is a measure of the average amount of time wasted in congestion for a particular simulation. When the simulation is repeated a number of times, the average of the Average Cumulative Congestion time can be determined. In the first scenario, the average congestion
experienced by an occupant is significantly greater for all the cases examined when compared against the IMO day case. On average, the response time (298 seconds) accounts for approximately 78% of
a person’s personal assembly time (382 seconds) while the time spent in congestion (12 seconds) accounts for approximately 3% of the assembly time. Thus on average, only 19% of a person’s assembly time is spent in free walking (i.e. travelling freely to their destination). It is noted that most of the time spent in
congestion results from the congested areas around the stairs. Analysis of the four main congestion areas reveals that there are no areas of congestion considered significant i.e. areas with 4 persons/m2
for more than 10% of the total
simulation time. Using the artificial IMO response time
distribution the hypothetical ship design is deemed to pass the IMO day scenario. However, using the more realistic response time distributions, the same vessel is deemed to fail the IMO day scenario due to a number of regions of unacceptable local congestion. This brings into question the suitability of the IMO response time distribution. In the example cases examined using the
maritimeEXODUS software and the standard methodology and response time distribution specified in MSC 1033, a hypothetical passenger ship design was deemed to have satisfied both day scenario acceptance conditions relating to time to assemble and congestion. However, when the vessel was tested using two more realistic response time distributions, the same vessel was deemed to fail the benchmark test on both occasions due to excessive levels of congestion, even though similar overall assembly times to that found in the standard IMO case were produced. The benchmark evacuation simulations
specified in MSC 1033 to certify the evacuation performance of passenger ships utilise an unrealistic mathematical form to describe the passenger response time distribution, a key evacuation parameter. From land-based evacuation experience, the response time distribution associated with a multitude of day time evacuation scenarios assumes a characteristic skewed or log-normal distribution. The IMO specified response time distribution is artificially set as a uniform random distribution. This unrealistic mathematical form can lead to serious congestion issues being overlooked in the evacuation analysis. In light of these results, it is vital that
IMO undertake research to generate passenger response time data suitable for use in evacuation analysis of passenger ships. Ideally, these experiments should be conducted using real passenger ships under realistic conditions, preferably at sea. The authors, as part of the European Union Framework V project FIRE EXIT (grant: G3RD-CT-2002-00824) have started to collect such data. Until this type of data becomes readily
available, it is strongly recommended that rather than continuing to use the artificial and unrepresentative uniform random distribution, IMO should adopt plausible and more realistic response time data from land-based applications.
Table 2: Average results (over 50 repeat simulations) for each Scenario.
THE NAVAL ARCHITECT FEBRUARY 2007 45
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96 |
Page 97 |
Page 98 |
Page 99 |
Page 100 |
Page 101 |
Page 102 |
Page 103 |
Page 104 |
Page 105