Analysis Considering the Conversion of an Existing Road Tunnel Transverse Ventilation System to Transit Use

author list icon

The ventilation system designs for two existing road tunnels are analyzed to determine the feasibility of reusing them for a new LRT trainway.

model section Holland tunnel road New York ventilation 1927Figure 1 – Full-scale Model Section of the 1927 Holland Tunnel in New York

In 1927, the first fully transverse tunnel ventilation system was commissioned for the Holland Tunnel in New York (see Figure 1). Until that time, tunnels had been ventilated only by longitudinal airflow systems. 

Longitudinal ventilation systems provide airflow along the tunnel axis at a velocity sufficient to prevent smoke back-layering during fire emergencies, allowing safe egress in one direction (into the fresh airflow). Fully-transverse ventilation systems utilize separate supply and exhaust air ducts extending the length of the tunnel to provide fresh air into the tunnel at the roadway level and to extract heat, emissions, and smoke out of the tunnel near the ceiling. Transverse ventilation limits smoke spread to an area near a fire by extracting smoke generated through openings in the exhaust air duct, allowing safe egress away from the incident in both directions. 

Ole Singstad (Barclay, Parsons and Klapp 1917-1918) was the engineer responsible for developing the revolutionary Holland Tunnel ventilation system. At the time, ventilating such a congested vehicle tunnel was thought to be impossible, but the completed Holland Tunnel ventilation system would not only work, it would set the standard for fully transverse ventilation system design for many decades to come. 

Today, road tunnels throughout the United States and the world contain transverse ventilation systems. Traditionally, road tunnel ventilation systems have been designed to meet contaminant level criteria. In the decades since the Holland Tunnel ventilation system was designed, meaningful reductions in vehicle emissions have been realized due to advancements in automotive technology, and fire science has advanced significantly leading to a greater understanding of fire size and development. As a result, the design paradigm for tunnel ventilation systems has shifted over the past century from emissions control to a stronger focus on smoke control during fire emergencies. 

Considering Road to Rail Tunnel Conversion 

As rail transit systems continue to expand in an effort to offset heavily congested roadways, and with limited available right of way in populous metropolitan regions, some rail transit extensions are converting portions of existing roadways and road tunnels to passenger rail use (see Figure 2). In Seattle, Washington the Sound Transit East Link Extension project has proposed the conversion of existing Interstate I-90 center roadway and associated center tunnels into a light rail transit (LRT) trainway connecting Seattle to Bellevue (see Figure 3). 

In 2013, WSP | Parsons Brinkerhoff performed extensive ventilation analysis in support of the East Link project final design, investigating the ventilation system design for two existing road tunnels in the Seattle area, the Mount Baker Ridge Tunnel and the Mercer Island Lid Tunnel. The analysis specifically considered modifications to the existing fully-transverse road tunnel ventilation systems to determine the feasibility of reusing it for the new LRT trainway. 

tunnel interstate congested transit routeFigure 2– Transit Route in median of congested interstateroads transit extension I-90 Homer M Hadley floating bridge, fire life safetyFigure 3 – Simulation of the proposed East Link Extension along I-90 across the Homer M. Hadley floating bridge

Existing East Link Tunnel Ventilation 

The Mount Baker Ridge Tunnel and the Mercer Island Lid Tunnel each consist of a fully-transverse ventilation system with a central fan plant located above the tunnel near the mid-point. The Mount Baker Ridge tunnel is 3,478 feet (1060 meters) long and has air ducts along each side of the roadway, while the Mercer Island Lid tunnel is approximately 2,900 feet (884 meters) long with air ducts arranged above the roadway, separated from the roadway by a suspended ceiling. The ventilation airflow is evenly distributed along the tunnel length through small ports spaced at regular intervals. 

Investigating the Reuse of a Transverse Ventilation System 

In 1993, the Memorial Tunnel Fire Ventilation Test Program (MTFVTP) began conducting full-scale fire tests in an abandoned road tunnel to evaluate the ability of several ventilation system types to manage smoke and temperature. The tunnel ventilation systems were tested across a range of fire sizes. Sound Transit East Link project design criteria specified a medium t-squared (time squared) growth rate fire curve with a peak heat release rate (HRR) of 13.2 MW. Comparing this fire size and smoke generation rate with the ventilation performance for 10 and 20 MW fires from the findings of the MTFVTP, it would appear that some variation of extraction ventilation could control the smoke generated. However, road tunnel ventilation systems, like those considered for re-use in the East Link extension, were designed to the 100 cubic feet per minute per lane foot criteria. In MTFVTP findings, this criterion was not sufficient for many emergency fire scenarios. 

The existing road tunnel ventilation system for each East Link tunnel consists of three fans for supply and three fans for exhaust ventilation. WSP | Parsons Brinckerhoff’s ventilation analysis specifically considered the re-use of the exhaust ventilation fans for the new LRT trainway. The analysis utilized computational fluid dynamics (CFD) to investigate the performance of several system types: 

  • Single Zone Exhaust System – Steady State Run: use the existing system to exhaust from the entire length of tunnel (supply fans off). The existing system failed due to smoke spread (see Figure 4). The tunnel width and existing fan capacity limited the effectiveness of the extaction system. 
single zone exhaust system ventilation tunnels systemFigure 4 – Single Zone Exhaust System
  • Two-Zone Exhaust System – Transient Run: use the existing system to exhaust half of the tunnel by closing isolation dampers at the tunnel midpoint (supply fans off). The existing system failed to control smoke spread at 6 minutes (see Figure 5). Smoke spread was limited, but not contained to the immediate area at the fire car due to wind forces and extraction port locations. 
double zone exhaust system ventilation tunnels systemFigure 5 – Two-Zone Exhaust System
  • Point Extraction System - Close all existing exhaust ports and install new openings, 160 square feet (14.9 square meters) with motorized dampers every 250 feet (76.2 meters) on-center creating a point extract system where the three closest dampers to the fire are opened, and the exhaust fans extract the smoke through the open dampers (see Figure 6). This system controlled smoke for the duration of egress (8 minutes) with increased fan capacity, but was rejected due to the rigorous structural analysis required and potential seismic retrofit work. 
point extract ventilation tunnels systemFigure 6 – Point Extraction System
  • Four-Zone Extraction System - Close all existing exhaust and supply port openings. Convert the supply duct into a second exhaust duct and effectively subdivide the system into 4 zones. Install isolation dampers to direct all exhaust ventilation to a single incident zone. Create large openings, 240 square feet (22.3 square meters) in the suspended ceiling to extract smoke and hot gases from the incident fire zone (see Figure 7). This system controlled smoke spread for the duration of egress (12 minutes) with increased fan capacity, but the option was rejected due to concerns about the structural impact on the existing tunnel. 
four zone extraction ventilation tunnels systemFigure 7 – Four-Zone Extraction System
  • Longitudinal System (Jet Fans) – install 14 jet fans (12 fans operating - analysis considered 1 fan out of service and 1 fan out due to fire) along the tunnel walls providing longitudinal ventilation and protecting the egress path indefinitely (see Figure 8). 
longitudinal ventilation tunnels system Figure 8 – Longitudinal System


Converting an existing fully-transverse road tunnel ventilation system to transit use is feasible but presents many challenges to the designer. Based on CFD analysis of the East Link tunnels for the stated fire size, the following parameters significantly affect the ventilation system performance: 

  • Wind speed (normal to the portals) – This was the primary parameter contributing to smoke spread along tunnel length for the extraction system analysis; 
  • Extraction points – The existing ventilation port size and spacing significantly limited the extraction capacity nearest the fire car; 
  • Tunnel Width – Extraction airflow velocities across the tunnel cross-section were non-uniform and diminished opposite the ports; and 
  • Fan Capacity – Fan size was limited by maximum allowable airflow velocities in the ducts and at the egress walkway. 

Fully-transverse road tunnel ventilation systems were designed to provide distributed transverse air flow evenly along the length of the tunnel. Such a design works well for controlling air quality during times of congested traffic, where automobiles fill the tunnel from portal to portal and contaminates are evenly distributed. In road tunnels, parameters such as wind speed improve air quality by purging the tunnel pollutants. 

For transit tunnels, where emergency fire conditions drive the design, extraction ventilation systems must effectively control smoke and heat to a very limited area near the fire car under worst-case conditions. Existing fully-transverse ventilation systems can be modified, allowing the extraction capacity to be directed nearest the fire source, but such modifications require structural analysis of the tunnel and may not be able to control smoke spread for fire sizes greater than 10MW. The East Link tunnel ventilation analysis led Parsons Brinckerhoff (now part of WSP | Parsons Brinckerhoff) to recommend a longitudinal system that resulted in significant cost savings over the preliminary design. 

back to top