Contents
The King's Cross fire was a fire at King's Cross Underground station in London on 18 November 1987. The NFPA handbook describes unexpectedly rapid spread along an inclined wooden escalator and dangerous conditions in the ticket hall above. The case became significant for understanding how fuel geometry can alter flame behaviour. [1]
Cox and Kumar's SFPE Handbook chapter discusses the computational investigation and subsequent physical testing. Flames were predicted to lie along the escalator channel under particular flow conditions instead of rising as a separate vertical plume. The investigation therefore linked an unusual observation with a physical mechanism and tests of the proposed explanation. Its engineering significance concerns the configuration of the fire, the supply of entrained air and the careful interpretation of a model's unexpected result. [2][3]
Inclination and channel geometry
The Trench effect concerns fire behaviour in an inclined, laterally confined arrangement. The handbook explains that attachment of flames along an inclined burning surface can increase heating ahead of the fire. That mechanism differs from an assumption that a localised fire spreads only slowly sideways. The escalator's configuration was therefore part of the fire problem, not merely a support for the burning material. [4]
Cox and Kumar describe the importance of the width over which burning occurred. When the fire occupied the channel width, air could be drawn from above and below without the same lateral supply. The predicted flow then differed from that around a fire restricted to one side. The explanation connects the flame's orientation with the air available to it. [5]
This is a configuration-dependent mechanism. It does not establish that all inclined surfaces or every escalator fire will produce identical behaviour. The handbook's wider chapter distinguishes pools, walls, ceilings and confined surface arrangements. The relevant question is which physical configuration and burning conditions the evidence represents. [6]
What modelling and experiments established
Fire investigation can combine observations, calculations and physical experiments. The SFPE chapter explains that an unexpected numerical result may reflect real physics or a numerical artefact. For King's Cross, sensitivity to fire-source conditions was examined and physical tests were subsequently used. The agreement increased confidence in the proposed flow mechanism. [5]
The distinction is methodological as well as physical. A visually plausible calculation does not establish that the geometry, fire source and boundary conditions are correct. Conversely, an initially surprising prediction is not necessarily wrong. Cox and Kumar use the case to illustrate the need for independent checks and an understanding of the represented process. [5]
Flashover is a broader fire-development term and should not replace the specific explanation of flame spread along the escalator. The handbook's discussion centres on inclination, attachment and surface heating. Keeping those mechanisms explicit avoids reducing the case to a single dramatic label. The explanation concerns why the fire grew and spread in the observed configuration. [1]
Relevance to transport fire protection
Escalator fire protection concerns equipment, combustible deposits, materials and the surrounding building arrangement. The handbook's building-transport chapter discusses cleaning machinery and truss spaces and recognising the movement of burning items. The King's Cross case adds evidence about the effect of a particular combustible channel geometry. These are connected concerns but do not describe identical mechanisms. [7]
A Rail transit system can contain interconnected routes and occupied spaces beyond the fire's immediate location. The handbook describes people being threatened in the ticketing area above the escalator. The assessment therefore has to connect fire development with the spaces and movement routes it can affect. The location of ignition alone does not define the full exposed area. [1]
As a Catastrophic multiple-death fire, the event is relevant to the consequences of an inadequately understood mechanism. Its value for later analysis lies in the relationship between configuration, evidence and protective assumptions. A model for another station still needs its own inputs and checks. The historical result does not supply a universal growth rate for escalators or transport buildings. [2][8]
What the assessment has to establish
An assessment starts by identifying what could be harmed and how a fire could produce that harm. Life, property, continuity of operations, heritage and the environment can all be relevant. These objectives overlap, but they do not describe the same outcome. A building that allows everyone to escape can still suffer a loss that stops its activities. Equally, preserving the structure does not necessarily preserve its contents or services. Meacham's account of building fire risk analysis makes these distinctions part of defining the problem. [9]
A hazard assessment then identifies possible ignition sources, combustible materials and the conditions through which fire could spread. The arrangement matters as well as the quantity of material. The description includes the building, its compartments and the protection systems that may affect the event. A hazard assessment asks what could happen under specified conditions. A risk assessment adds consideration of how likely those conditions and consequences are. A calculated consequence therefore does not, by itself, establish a probability of loss. [9]
The people affected need to be described with similar care. An occupancy name provides a useful starting point, but does not fully describe an individual's ability to respond. The NFPA handbook treats occupancy classification as a substitute description for a collection of life safety characteristics. Detailed assessment needs more information when those characteristics determine the result. Age, mobility, awareness, familiarity and available assistance can change the protection that an occupant needs. A category suitable for organising code requirements is therefore not a complete prediction of an evacuation. [10]
Different forms of evidence answer different questions. Inspection can identify an obstruction or an unprotected opening. A calculation can explore a fire scenario under stated assumptions. Historical records can indicate recurring causes or consequences. None of these sources removes the need to check whether the evidence applies to the particular building and activity. Meacham identifies changing materials, uncertain data and future changes in use as limitations when past experience informs a new assessment. [8]
Assessment also includes the behaviour of safeguards when conditions are unfavourable. A review can examine a failed component, an unavailable protection system or an action that does not occur as intended. Fault trees work backwards from an unwanted event towards possible causes. Event trees work forwards from an initiating event through possible outcomes. Their structure can clarify dependencies before reliable numerical probabilities are available. Assigning a number to each branch is a separate evidential step. [11]
The resulting judgement depends on the purpose of the assessment and the acceptance criteria. A result that answers a narrow equipment question may leave wider building risks unresolved. Stakeholders can also value consequences differently, especially where life safety, economic loss and irreplaceable heritage are involved. Recording the scope makes the meaning of an acceptable result visible. It also identifies which changes in use, population or protection would require the assessment to be reconsidered. [9][12]
Evidence, review and the limits of a decision
A fire safety decision needs a defined question before calculations or inspections can resolve it. The International Fire Engineering Guidelines describe a fire engineering brief as the agreed basis for an analysis. It records the scope, relevant stakeholders, building and occupant characteristics, objectives and methods. Acceptance criteria and proposed designs are considered as part of that process. The brief is a technical basis for the assessment, rather than simply the commercial agreement for the work. Its detail varies with the complexity of the problem. [13]
The distinction between an objective and a test of that objective is significant. Protecting occupants states an aim. A criterion identifies a condition against which the design can be evaluated. The calculation method then estimates the quantities needed for that comparison. These steps cannot be replaced by selecting software and accepting its output. The NFPA handbook states that verification methods help evaluate a solution; the design and approval decisions remain with the responsible people. [14]
An analysis depends on its assumptions even when the output appears precise. Input data may be incomplete, the model may simplify a physical process, and actual conditions may vary. The handbook distinguishes uncertainty caused by limited knowledge from variation inherent in the conditions being represented. Additional information can reduce some uncertainties, but cannot remove all variation. Identifying the source of uncertainty makes it possible to judge which further evidence would change the decision. [15]
Sensitivity analysis explores how changing an input changes the result. It can reveal that an apparently minor assumption controls a conclusion. A separate assessment can examine the point at which a result changes from acceptable to unacceptable. These methods help explain the margin between a prediction and the acceptance criterion. They do not create evidence for an input that has no defensible basis. The scale of the uncertainty and the significance of the outcome remain relevant to interpreting the margin. [16]
Review also has an organisational dimension. A reviewer needs enough information to understand the proposed arrangement and the reasons offered for accepting it. The handbook places the authority having jurisdiction within the performance design process from an early stage. The IFEG similarly describes consultation and revision of the brief as the work develops. Early agreement on scope can identify disagreements before they become hidden within a finished calculation. It also makes departures from the original assumptions visible. [16][17]
Professional competence and independence affect the quality of this evidence. Van Rossum explains that complex analyses can be difficult for a reviewing body to check fully. Commercial or organisational interests may also favour a particular result. A technically convincing report therefore depends on competent work and an honest account of assumptions and limitations. These issues concern the integrity of the assessment, alongside the correctness of individual calculations. They cannot be resolved merely by making a report longer or its numerical output more detailed. [18]
Notes
- ↑ 1.0 1.1 1.2 Cote et al. (2008), §2.2, p. 2-25.
- ↑ 2.0 2.1 Cox and Kumar (2002), pp. 3-207–3-208.
- ↑ Cote et al. (2008), §2.2, pp. 2-25–2-26.
- ↑ Cote et al. (2008), §2.2, pp. 2-21, 2-25.
- ↑ 5.0 5.1 5.2 Cox and Kumar (2002), p. 3-207.
- ↑ Cote et al. (2008), §2.2, pp. 2-21, 2-25–2-26.
- ↑ Cote et al. (2008), §10.3, pp. 10-57, 10-67; §2.2, p. 2-25.
- ↑ 8.0 8.1 Meacham (2002), pp. 5-154–5-156.
- ↑ 9.0 9.1 9.2 Meacham (2002), pp. 5-153–5-155.
- ↑ Cote et al. (2008), §20.1, p. 20-9.
- ↑ Meacham (2002), p. 5-156.
- ↑ Australian Building Codes Board (2005), §1.2.4, pp. 1.2-6–1.2-7.
- ↑ Australian Building Codes Board (2005), §1.2, pp. 1.2-2–1.2-3.
- ↑ Cote et al. (2008), §3.10, pp. 3-167, 3-173–3-174.
- ↑ Cote et al. (2008), §3.10, p. 3-173.
- ↑ 16.0 16.1 Cote et al. (2008), §3.10, p. 3-174.
- ↑ Australian Building Codes Board (2005), §1.2, p. 1.2-2.
- ↑ Van Rossum (2027), §§2.1–2.3.
References
- Australian Building Codes Board. (2005). International fire engineering guidelines.
- Cote, A. E., Grant, C. C., Hall, J. R., Jr., & Solomon, R. E. (Eds.). (2008). Fire protection handbook (20th ed.). National Fire Protection Association.
- Cox, G., & Kumar, S. (2002). Modeling enclosure fires using CFD. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., Section 3, Chapter 8, pp. 3-194–3-218). National Fire Protection Association.
- Meacham, B. J. (2002). Building fire risk analysis. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., Section 5, Chapter 12, pp. 5-153–5-175). National Fire Protection Association.
- Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.