【正文】 某些殘余可能性因子可由統(tǒng)計(jì)數(shù)據(jù)獲得，如果沒有可用的統(tǒng)計(jì)數(shù)據(jù)，則需要進(jìn)行主觀判斷。 ? 比如沒有“在公寓中使用阻燃材料”對減少火災(zāi)發(fā)生影響的統(tǒng)計(jì)數(shù)據(jù)。 ? This allows the fire protection engineers and regulators to assess the impact of these fire protection measures based on their assessments of the reduction of the probability of fire occurrence. Some of these residual probability multipliers can be obtained from statistics, if they are available. If no such information is available, then subjective judgment may be required. ? 火災(zāi)安全工程師基于火災(zāi)安全措施對火災(zāi)發(fā)生可能性的減縮情況來確定火災(zāi)安全措施的影響。 ? 我們只能對此進(jìn)行假設(shè)，在本例中假設(shè)“在公寓中使用阻燃材料” 與“在建筑中禁煙”對火災(zāi)發(fā)生可能性的影響相同，其殘余可能性因子為 (見 Table )。如圖 ，每一種火災(zāi)保護(hù)措施對火災(zāi)發(fā)生固有頻率的影響通過一個(gè)殘余可能性因子來體現(xiàn)（ residual probability multiplier ）。即在“在公寓中使用阻燃材料”成功，其結(jié)果是將火災(zāi)發(fā)生的頻率下降到其固有頻率的。 plan) (failure probability of ‘sprinkler system’) (failure probability of implementing a ‘regular evacuation drills’ plan). ? In the eventtree method, the probability of fire occurrence for each fire scenario is assessed based on the inherent rate of fire occurrence and the impact of various fire prevention measures to minimize this inherent rate of fire occurrence. In Figure , the impact of each of the fire protection measures on the inherent rate of fire occurrence is assessed using a residual probability multiplier. ? 在事件樹方法中，每一個(gè)火災(zāi)場景出現(xiàn)的可能性基于火災(zāi)發(fā)生的固有頻率和各種火災(zāi)安全措施對降低火災(zāi)發(fā)生固有頻率的影響來確定。 ? For example, there is no statistical information that can be easily found on the reduction of fire occurrence of implementing a ‘no smoking material’ plan for apartment buildings. ? Without such statistical information, we have to make an assumption. Let us assume for this example that the residual probability multiplier of a ‘no smoking material’ plan in apartment building is , the same as that for house fires (see Table ). ? That is, the consequence of a successful ‘no smoking material’ plan is the reduction of the rate of fire occurrence to of its inherent value. ? This residual probability multiplier for each fire scenario is shown in Figure . For example, Scenarios A, B, C, D, all with a failure of implementing the ‘no smoking material’ plan, have a residual probability multiplier of 1 (no reduction)。 whereas the measures of ‘sprinklers and‘regular evacuation drills’ would have an impact on lowering the consequence of a fire occurrence by suppressing or controlling the fire or by allowing the occupants to evacuate more quickly. ? As was in the discussion of the checklist method, the event tree in Figure is only an example to show how an event tree can be used for quantitative fire risk assessment. ? The description of each event is the judgment for this example only. Other applications may have different descriptions. The descriptions allow more transparent discussions and agreements among stakeholders. ? In an eventtree method, the probability of each fire scenario is calculated using the probability values of success or failure of implementing the fire protection measures that are associated with the scenario. ? Some of these probability values can be obtained from statistics, if they are available. ? For example, NFPA statistics show that, based on the 1999–2023 data, sprinklers in apartment buildings have a reliability of 96% of activating and controlling large fires that should activate sprinklers (Kimberly and Hall, 2023). Note that smouldering fires and small fires may not activate sprinklers. ? If no such information is available, then subjective judgment may be required. For example, there is no statistical information that can be easily found on the probability of success or failure of implementing a ‘no smoking material’ plan so that there will be a lower rate of fire occurrence. Without such statistical information, we have to make an assumption. ? Let us assume for this example that the probability of success of implementing a ‘no smoking material’ plan is 70 %. ? Similarly, there is no statistical information that can be easily found on the probability of success or failure of implementing a ‘regular evacuation drills’ plan in apartment buildings so that the occupants would know what to do in case of a fire alarm and would therefore evacuate more quickly than without such drills. Let us assume for this example that the probability of success of implementing a ‘regular evacuation drills’ plan is 80 %. ? In real risk assessments, these values need to be carefully analysed and agreed upon by fire safety engineers and regulators. Success is defined as that the fire protection plan actually works. ? The scenario probabilities are shown in Figure . ? For example, Scenario A has a probability of 10?03, which is the product of (failure probability of implementing a ‘no smoking material39。 whereas Scenario E, F, G, H, all with a success of implementing the ‘no smoking material’ plan, have a residual probability multiplier of . ? 每一個(gè)火災(zāi)場景的這個(gè)殘余可能性因子在圖 顯示出來。 ? whereas the measures of ‘sprinklers’ and ‘regular evacuation drills’ would have an impact on lowering the consequence of a fire occurrence by suppressing or controlling the fire or by allowing the occupants to evacuate more quickly. ? It should be emphasized that this is just an example to show how quantitative fire risk assessment can be carried out using a checklist method. There are no standard checklist methods in fire risk assessment. ? In Table , the inherent fire risk values (without the help of any fire prote