- 9 - AN INVESTIGATION OF LONGITUDINAL VENTILATION FOR SHORT ROAD TUNNELS WITH HIGH FIRE HRR O Gorman S., Nuttall R., Purchase A. Parsons Brinckerhoff, Australia ABSTRACT Recent fire tests for tunnels have suggested high fire heat release rates (HRRs), of the order of to MW, may be appropriate under certain conditions. The actual design fire used for a given tunnel will likely be determined from a quantitative risk assessment, an assessment of the maximum HRR expected in the tunnel, or be specified in a project requirement. This paper does not focus on the selection of a design fire HRR, but rather its implications on the ventilation system in short tunnels. For short road tunnels there are practical limitations to the size of a longitudinal ventilation system, which can include space limitations, portal wind conditions and noise issues. This paper investigates the practical limitations of jet fan longitudinal ventilation for direction of travel smoke management schemes. It identifies that there is a maximum limit to the fire HRR that a jet fan ventilated tunnel of a given length can achieve. Beyond this limit a different smoke control scheme should be investigated or methods to reduce the anticipated design fire size (e.g. risk assessment). The paper also provides some guidance on the impact of tunnel grade, area and noise criteria on the maximum limit. These results can be used by the designer to establish if a jet fan based longitudinal ventilation system is appropriate for a given tunnel and fire HRR. Keywords: ventilation design, fire HRR, incident ventilation, short tunnels. INTRODUCTION The design of a longitudinal ventilation system for short road tunnels provides unique challenges compared to longer road tunnels. Due to the length it is likely the design of the ventilation system will be driven primarily by emergency scenarios rather than the need to dilute vehicle emissions. For short tunnels the dominant impacts on the ventilation system are tunnel area, tunnel grade, tunnel air temperature (as it affects buoyancy and jet fan de-rating), portal wind effects and tunnel acoustic criteria. However, the size of the ventilation system that can be installed is dependent on the length of the tunnel within the bounds of commercially available jet fan sizes. As a result, there exists a limit based on a tunnel s length and characteristics where a longitudinal smoke control system may become impractical. These considerations can become more critical for high fire HRRs. The purpose of this paper is to investigate the effect of high fire HRR on the size of the ventilation system for short tunnels and identify an indicative limit to the size of the HRR that can be accommodated by a longitudinally ventilated tunnel of a given length. Beyond this limit alternative smoke control or egress strategies may be required or a re-evaluation of the design fire HRR (e.g. risk assessment). The investigation is based on simulating the performance of a tunnel ventilation system for tunnels of various lengths, grades and fire HRRs. 6 th International Conference Tunnel Safety and Ventilation, Graz
- 5 -. MECHANICAL VENTILATION OF SHORT ROAD TUNNELS.. Short road tunnels The requirement to mechanically ventilate short road tunnels varies around the world. In Australia, the New South Wales requirements are for any tunnels over 6 m in length to be mechanically ventilated (RTA 6, []). Tunnels less than 6 m require a performance assessment. However, there are various examples within Australia of short tunnels being mechanical ventilated ranging from 8 m to 5 m. In Europe a review of design guidelines suggests that the requirement for mechanical ventilation varies between countries from 5 m long to 5 m long (Fire in Tunnels 6, [8]). In the United States, NFPA 5 requires mechanical ventilation for tunnels greater than m in length and should be considered for tunnels greater than m in length (NFPA 5, []). Based on the above it is likely that tunnel lengths from m and longer may require mechanical ventilation. This paper has focussed on tunnel lengths from m to 6 m... Fire HRR and smoke management strategies The fire HRR used for the design of a mechanical ventilation system is heavily dependent on the expected vehicle usage. Typical fire HRRs vary from 5 MW for passenger cars, up to MW for some forms of large vehicles and - MW for dangerous goods vehicles (PIARC 999, [6]). Some recent fire tests for tunnels have suggested fire HRRs in the order of MW may be possible under certain conditions. NFPA 5 nominates 7- MW for HGV s and - MW for tankers. The choice of a design fire HRR will likely be determined from a quantitative risk assessment, an assessment of the maximum HRR expected in the tunnel, or be specified in a project requirement. This paper does not focus on the selection of a design fire HRR, but rather its implications on the ventilation system in short tunnels. For a longitudinal ventilation system the smoke management strategy has been described in various documents (including PIARC 999, [6]). The fundamental goal of longitudinal ventilation is to provide airflow in the direction of travel to prevent smoke back-layering upstream of the fire location. This provides tenable conditions for occupants upstream of the fire and an access path for fire service intervention... Longitudinal ventilation issues The momentum equation for a jet fan based longitudinal ventilation system has been developed previously and documented various times (Armstrong et al. 99 [], PIARC 995 [5]). A simplified form is reproduced in equation. Where:, = + + -() ΔP JF = Total pressure rise provided by the jet fans (Pa) ΔP Tunnel = The sum of the pressure loss in the tunnel (e.g. hydraulic losses including tunnel friction, vehicle drag resistance and tunnel losses such as entry contraction losses) (Pa) ΔP wind = Pressure loss from wind forces acting on the portals (Pa) ΔP buoyancy = Buoyancy induced pressure loss due to high temperature smoke (Pa) 6 th International Conference Tunnel Safety and Ventilation, Graz
- 5 - For a short tunnel, as distinct from a longer tunnel, the relative contribution of the different pressure terms can vary significantly. In particular, the adverse portal wind pressure (ΔP wind ) dominates the contribution from the tunnel resistance (ΔP Tunnel ) at shorter tunnel lengths. At high grades the buoyancy force of the hot smoke is also a considerable driver of the total tunnel thrust requirement (Reiss et al., [7]). The tunnel air temperature will be significantly hotter in the first 5- m of a fire incident and as a result the buoyancy force for short tunnels will likely be a larger proportion of the total required thrust compared to longer tunnels. The pressure developed in a tunnel section by the action of a jet fan has been described in several references (Armstrong et al. 99, []) and is shown in equation., = -() Where: Q f = Flow rate through jet fan (m /s) v f = Jet fan discharge velocity (m/s) v T = Tunnel velocity (m/s) β = Jet fan installation factor (-) ρ = Air density (kg/m ) A T = Tunnel cross-sectional area (m ) Downstream of a fire incident the tunnel air density ρ and tunnel velocity v T change with temperature. The effective result is a significant reduction in jet fan thrust downstream of a fire incident. It is also common practice to assume that jet fans located in the vicinity of a fire incident have failed due to the impact of the fire and cannot be relied upon to provide additional tunnel thrust. The combined effect is that for various scenarios it can be difficult to provide sufficient jet fan thrust in a short tunnel.. METHODOLOGY A generic tunnel was analysed to investigate the effect of HRR on the ventilation system performance for a given tunnel length. The performance of a generic tunnel was simulated for various tunnel lengths, tunnel grades, tunnel areas and fire HRRs to determine the required longitudinal thrust to achieve critical velocity. The simulations were undertaken using SES and the critical velocity was calculated using the critical velocity equation (Kennedy 996, []). The results presented in this paper are based on SES simulations, although numerous iterations were required to estimate the required thrust. A scripted approach was used to generate, run and post-process the models. A generic tunnel shape was assumed for a and lane tunnel and is shown in Figure. Other simulation inputs are shown in Table. 6 th International Conference Tunnel Safety and Ventilation, Graz
- 5 - Figure : Generic and lane tunnel shape Table : Model Inputs Parameter Value Tunnel friction factor.5 Tunnel area ( lane) 66 m Tunnel area ( lane) 9 m Tunnel grade (constant along length of tunnel) to -5% Physical limit (max) jet fan catalogue thrust 5 N (per jet fan) Acoustic limit jet fan catalogue thrust 8 N (per jet fan) Installation factor.85 Adverse portal wind velocity (pressure) 5 m/s (5 Pa) The maximum available (or installed) thrust for a generic tunnel configuration was based on applying best practice for the design of a jet fan based longitudinal ventilation system. This included a separation equivalent to hydraulic diameters between jet fan banks, the loss of one bank of fans in a fire and commonly used installation factors for fan location relative to tunnel ceilings and walls. The maximum jet fan thrust was estimated from manufacturer data assuming no-special high thrust fans (i.e. catalogue available jet fans only). The size of jet fans that can be installed at a given tunnel location is primarily dependent on the size of the tunnel and the project s acoustic criteria (if any). For a project with no acoustic criteria, a maximum jet fan thrust of 5 N was assumed. For projects with an acoustic criteria (e.g. 85 dba at.5 m above the walking surface for emergency scenarios) a maximum jet fan thrust of 8 N was assumed. The selection of 8 N thrust was based on simple acoustic analysis and verified by recent commissioning experience on a lane 5 m longitudinally ventilated tunnel. Direction of traffic / smoke control Direction of egress Figure : Jet fan installation and fire scenario 6 th International Conference Tunnel Safety and Ventilation, Graz
- 5 - It should be noted that this investigation used idealised inputs and best practice installations. It is possible that more jet fans could be installed at a given location, the separation between jet fans relaxed or larger jet fans installed. This would need to be considered on a project specific basis and appropriate analysis undertaken.. RESULTS The results of this investigation are shown in Figure. The results are plotted separately for the different fire HRR s of MW, 5 MW and MW, and for two different tunnel areas. The graphs are normalised by the maximum available jet fan thrust that can be achieved for a given tunnel length with an acoustic criteria (left axis) and without an acoustic criteria (right axis). If a data point sits above the horizontal criteria line (i.e. greater than unity for a given criteria) then the available thrust is sufficient for the given tunnel length, grade and fire HRR. Conversely, if the data point sits below the horizontal criteria line (i.e. less than unity for a given criteria) then the available thrust is insufficient. If a data point sits between the acoustic criteria and physical limit lines then a jet fan based longitudinal ventilation system will only be suitable if there is no acoustic criteria set. Panels A and B of figure show that, for the tunnel studied, critical velocity can be achieved for a MW fire for the majority of grades and tunnel lengths. However, for the m long, two lane tunnel at high grades (-5%), there is insufficient thrust to achieve critical velocity for the case with an acoustic limit. For a 5 MW fire, Panel C shows that, for the two lane tunnel studied, there was insufficient available thrust for a m long tunnel at grades of -5%. The inclusion of an acoustic limit made it difficult at % grade. Panel D shows that for the lane case there is sufficient available thrust for all cases unless an acoustic limit is applied. In the case of an acoustic limit for a m long tunnel and -5% there is insufficient available thrust. For a MW fire, Panel E shows insufficient available thrust in a two lane tunnel at high grades for all tunnel lengths. For lengths of -6 m the results show that critical velocity can be achieved at low (-%) grades; however, introducing an acoustic limit results in achieving critical velocity at low grade (-%) for tunnels of 5-6 m in length. For the lane tunnel (Panel F) the outcome is improved slightly due to the additional jet fan per bank. However, there is still insufficient thrust for a m tunnel at all grades, albeit and % grades can be ventilated without acoustic criteria. 6 th International Conference Tunnel Safety and Ventilation, Graz
- 5 - Available Thrust (8 N / fan) / Required Thrust 5 5 5 55 6 Available Thrust (5 N / fan) / Required Thrust Available Thrust (8 N / fan) / Required Thrust 5 5 5 55 6 Available Thrust (5 N / fan) / Required Thrust (A) MW design fire, lane tunnel (CSA = 66 m ) (B) MW design fire, lane tunnel (CSA = 9 m ) Available Thrust (8 N / fan) / Required Thrust 5 5 5 55 6 Available Thrust (5 N / fan) / Required Thrust Available Thrust (8 N / fan) / Required Thrust 5 5 5 55 6 Available Thrust (5 N / fan) / Required Thrust (C) 5 MW design fire, lane tunnel (CSA = 66 m ) (D) 5 MW design fire, lane tunnel (CSA = 9 m ) Available Thrust (8 N / fan)/ Required Thrust 5 5 5 55 6 Available Thrust (5 N / fan)/ Required Thrust Available Thrust (8 N / fan)/ Required Thrust 5 5 5 55 6 Available Thrust (5 N / fan)/ Required Thrust Key: (E) MW design fire, lane tunnel (CSA = 66 m ) (F) MW design fire, lane tunnel (CSA = 9 m ) Grade = % Grade = % Grade = % Grade = % Grade = % Grade = 5% Acoustic Limit Physical Limit Notes:. Acoustic limit jet fans (8 N) based on achieving 85 dba at.5 m above the walking surface. Refer to left y axis for acoustic limit analysis.. Physical limit jet fans (5 N) based on maximum catalogue fan for the nominal tunnel. Refer to right y axis for physical limit analysis.. Best practice jet fans installations at separations of tunnel hydraulic diameters. The first jet fan is at 5 m within the inlet portal.. Fire located at 5 m within the inlet portal. First jet fan bank assumed to be destroyed by fire. Downstream fans are temperature derated. 5. If the ratio of required thrust to available thrust is greater than unity (for a given y axis) then longitudinal ventilation is suitable. Figure : Investigation results 6 th International Conference Tunnel Safety and Ventilation, Graz
- 55-5. CONCLUSION Short tunnels provide a unique challenge for smoke management by longitudinal ventilation systems. This is especially the case for tunnels with high fire HRRs. A generic tunnel has been investigated and the required thrust to achieve critical velocity has been simulated for various fire HRRs, tunnel grades, tunnel areas and tunnel lengths. The required thrust has been compared against the maximum available thrust for a given tunnel length with a longitudinal ventilation system designed with best practice parameters. The results of this investigation indicate that: For short tunnel lengths there is a practical limit to the fire HRR that can be accommodated by a jet fan based longitudinal ventilation system. The HRR limit for a jet fan based longitudinal ventilation system for a given tunnel length varies significantly based on tunnel grade and cross sectional area (Likely driven by the increase in space for additional jet fans in a larger cross section) The addition of an acoustic limit reduces the available thrust (i.e. lower noise generally equates to lower thrust per fan) and further limits the HRR that can be accommodated. The ventilation system for a given tunnel should be designed on a case by case basis. However, this paper indicates there is a limit to the HRR that can be longitudinally ventilated. Beyond this limit, alternative smoke management strategies (e.g. Saccardo nozzles), egress strategies (e.g. longitudinal egress passages and / or closer exit door spacing) may need to be adopted or a re-assessment of the design fire HRR. 6. REFERENCES [] Armstrong J., Bennett, E.C., Matthews, R.D. (99); Three-dimensional Flows in a Circular Section Tunnel due to Jet Fans; in proceedings of the 8 th International Symposium on Aerodynamics and Ventilation of Vehicle Tunnels, Mechanical Engineering Publications, London, pp 7-756. [] Kennedy W.D. (996); Critical Velocity-Past, Present and Future; in Proceedings of the Seminar on Smoke and Critical Velocity in Tunnels, London, England. [] NFPA: 5 (8) Standard for Road Tunnels, Bridges, and Other Limited access Highways [] Roads and Traffic Authority (RTA) (6) NSW: Road tunnel design guide, Fire Safety Design, RTA Pub. 6.57A [5] World Road Association (PIARC) (995): Vehicle Emissions Air Demand Environment Longitudinal Ventilation (5..B) [6] World Road Association (PIARC) (999): Fire and Smoke Control in Road Tunnels (5.5.B) [7] Riess I., Bettelini M.,Brandt R. (); Smoke Extraction in Tunnels with Considerable Slope; in proceedings of th International Conference Safety in Road and Rail Tunnels, Madrid. [8] Fire in Tunnels, European Thematic Network (6); Technical Report Part, Fire Safe Design Road Tunnels 6 th International Conference Tunnel Safety and Ventilation, Graz