Canadian Consulting Engineer

Transit Station Fires

May 1, 2007
By Ray Sinclair, Ph.D & Aim

Fire in a transit tunnel or station can lead to the rapid spread of hot toxic smoke within minutes. The primary goals for emergency planning are to evacuate the occupants safely and provide conditions...

Fire in a transit tunnel or station can lead to the rapid spread of hot toxic smoke within minutes. The primary goals for emergency planning are to evacuate the occupants safely and provide conditions for emergency services people to respond effectively.

Many transit systems have emergency ventilation systems to assist with these goals. Exhaust fans are meant to remove smoke and also draw fresh air into the station. The systems are designed to keep exit pathways relatively clear of smoke. Often, smoke is exhausted to the atmosphere at grade-level vents or at elevated discharges through sidewall vents. Make-up air is drawn down into the station through pedestrian entranceways, emergency exits and mechanical intakes.

The design performance of these emergency ventilation systems assumes that the make-up air will be fresh and clean. However, the smoke exhaust locations are often located too close to the make-up air locations. Therefore, there can be a relatively high risk of re-entrainment of dense smoke back into the station and exit pathways, obstructing exits and delaying emergency response measures. Even at street level, dense smoke can cause visibility problems for emergency services.

It is important, therefore, to understand the impact of emergency ventilation systems on emergency response measures and, where possible, to have an adequate separation between the exhaust and make-up air locations.

Complex wind flows can lead to re-entrainment

With the increasing need for sustainable design, many new building developments are being located close to mass transit stations. Similarly, stations are being located in high density areas to encourage the use of public transit. This situation creates a challenge for the design of emergency smoke systems due to the complex wind flow patterns created in a built-up urban environment (see figure 1). These wind flows create an added complication in locating the emergency ventilation exhausts and intakes because of the risk that the systems may not perform as desired in a fire.

Quantitative risk assessment techniques provide a method to evaluate different stations. A risk ranking matrix can be used to evaluate the likelihood, consequences and severity of an emergency. The evaluation can lead to balanced risk/cost decisions to help prioritize capital spending on stations. Locations with a higher level of probability of an event or higher consequences may be selected for screening or a more detailed analysis to assess the performance of the emergency ventilation system.

What are tenable conditions?

Codes and design standards vary in their requirements for the separation of exhausts and intakes. NFPA 130 (2007) is a commonly referenced standard for transit system design, from which Section 7.5 states that vent shafts “… shall be positioned or protected to prevent recirculation of smoke into the system through surface openings.” The section further states that “adjacent structures and property uses also shall be considered.”

These goals are well defined and clearly important, but in practice under different conditions they may be difficult to achieve. Therefore, RWDI suggests a process of design and analysis that seeks to reduce impacts to acceptable levels of probability, while fitting in with safety goals and cost constraints.

At a minimum, smoke exhaust vents and make-up air intakes should be separated sufficiently to ensure that re-entrained smoke is adequately diluted and tenable conditions are maintained for people to exit safely.

Inside the station or tunnel environment, tenable conditions are defined, for example, by the criteria in NFPA 130 (2007). Annex B.2.1.3 states that doors and walls should be discernible at 33 feet (10 m) and illuminated signs should be discernible at 100 feet (30.5 m). Temperature and toxicity criteria are also provided, but for many fire events these parameters are acceptable if the smoke is diluted sufficiently to meet acceptable visibility levels.

Outside the station and tunnel, smoke exhausted during a fire scenario should be diluted to similar tenable conditions as those required inside the station. The dilution must occur before the smoke reaches the intakes that supply “clean” make-up air to the emergency ventilation system. Achieving this level of dilution is particularly important at pedestrian entranceways that patrons would use as familiar exit paths during an emergency.

Modelling the dispersion of a smoke plume

Once discharged into the atmosphere, the dispersion of a smoke plume will be determined by the exhaust parameters, which include volume flow rate, discharge velocity, location, temperature, etc. Another factor in the dispersion is the ambient conditions, most notably wind speed and wind direction, and the complexity of the surrounding environment.

Dispersion modeling can assess the level of external dilution for various conditions and the probability of adverse impacts. These models can be used to determine the level of dispersion that occurs as the exhaust plume travels from the discharge point to locations such as the make-up air intakes and pedestrian exits.

Dispersion modeling techniques include numerical calculations on a computer, or physical scale modeling in a boundary layer wind tunnel. Numerical models can be useful for a screening level assessment. However, wind tunnel modeling should be used in complex settings, and for more detailed assessments.

Locating the exhaust and intake

Factors that can be controlled within the design include:

* exhaust location and configuration

* exhaust parameters (volume flow rate, discharge velocity and temperature)

* separation distance from exhaust to make-up air location.

The interaction between, and optimization of, these parameters will determine the performance of different design options. For example, a strategy of increasing the discharge velocity will not improve the dilution sufficiently from a grade-level discharge point, but could substantially improve the dispersion from a rooftop smoke exhaust. Similarly, the buoyancy dispersion benefit from a heated smoke exhaust will not be as evident for a grade-level discharge due to the dominant effect of the mechanical mixing created from wind flow around buildings. If a grade-level discharge is necessary, the exhaust and intake should not be on the same side of a building and, in addition, a large separation distance will be required.

In general, it is preferable to locate the intake at grade and the discharge point at an elevation above grade and out of the influence of building effects. A slightly elevated horizontal louvre on a building sidewall will not provide significant improvement over a street-level exhaust because the dispersion will still be limited by the building. A well placed vertical rooftop exhaust will provide more dilution with smaller separation distances and will show the benefit of more optimal exhaust parameters, such as increased stack exit velocity (see figure 2).

Steps at different design stages

As a reference to designers and decision makers, the following steps should be considered through the planning and design stages of transit stations and surrounding developments:

(1) Conceptual design stage. Review the complexity of the surroundings of existing and proposed transit stations; get expert advice on wind and exhaust vent dispersion issues in order to identify concerns and solution options as early as possible.

(2) Schematic design stage. A more detailed review should include screening-level dispersion modeling to help rank issues and solution options. The scope and benefits of wind tunnel dispersion modeling would be decided.

(3) Design development stage. Conduct detailed wind tunnel dispersion modeling to quantify vent shaft and make-up air vent performance. Use data in a risk assessment to ran
k the consequence versus the likelihood of different events.

The engineering reports from the above assessments will be useful for supporting capital cost decisions and also for review by authorities.

Ray Sinclair, Ph.D., is a principal and Aime Smith, P.Eng. is an associate with Rowan Williams Davies & Irwin (RWDI) of Guelph, Ont.

Advertisement

Stories continue below

Print this page

Related Stories