How MRT Tunnel Ventilation and Airflow Work Using Mathematics

An underground MRT train does not merely travel through a tunnel. It moves the air in front of it, pulls air behind it, heats the space around it and changes the pressure field of the railway.

The tunnel is therefore not empty space around the train. It is part of the transport machine.

Imagine pushing a piston through a long cylinder.

The piston displaces the fluid in front of it.

An underground MRT train does something similar.

The tunnel is much larger than the train, so air can flow around the vehicle, through station passages, ventilation paths and other openings. But the train still occupies a large fraction of the available cross-section. As it moves, it creates a changing pressure and airflow field.

That airflow affects:

train aerodynamic drag,
tunnel temperature,
station cooling load,
air exchange,
passenger comfort,
platform-screen-door pressure loading,
ventilation-fan energy,
and—in an emergency—the movement of smoke and heat.

The same tunnel can therefore behave as:

an aerodynamic duct during normal train movement
+
a thermal system during everyday operation
+
a ventilation network during environmental control
+
a life-safety airflow system during an emergency

MRT tunnel ventilation is not one fan blowing air. It is a moving fluid network whose sources, resistances and objectives change whenever a train enters, stops, leaves or the railway changes operating state.

Singapore’s public engineering record makes this visible. LTA has described tunnel ventilation fans that can move air in either direction and tunnel airflow paths controlled through ventilation equipment. The current SCDF Code of Practice for Fire Precautions in Rapid Transit Systems requires smoke control in an affected underground trainway and a non-contaminated environment in the evacuation portion of the trainway. And the 2026 Rail Reliability Taskforce condition-monitoring baseline explicitly includes tunnel ventilation fans among the station assets to be monitored.

Normal operation and emergency operation therefore use some of the same physical infrastructure for completely different optimisation jobs.

This article continues the eduKateSG MRT mathematics cloud. The permanent whole-system owner is How MRT Works | It’s Mathematics. Closely connected pillars include How MRT Platform Screen Doors Work Using Mathematics, How MRT Noise and Vibration Work Using Mathematics, How MRT Power Supply Works Using Mathematics and How MRT Signalling and Train Regulation Work Using Mathematics.

The RFE — Why Does Tunnel Ventilation Exist?

The weakest possible objective is:

move as much air as possible

That would waste enormous energy and could create uncomfortable or undesirable pressure and airflow conditions.

The opposite objective—use no mechanical ventilation—would ignore heat, air quality and emergency smoke-control requirements.

The Reason for Existence changes with railway state.

Railway state	Ventilation RFE
Normal operation	Manage heat, air quality, pressure, airflow and energy while trains move through the underground network.
Hot or high-load operation	Remove enough heat and supply enough fresh air to preserve a usable station and tunnel environment.
Maintenance state	Provide the airflow conditions required for authorised engineering work.
Fire or smoke state	Control smoke movement and protect evacuation and response pathways according to the authorised life-safety design.
Degraded ventilation state	Preserve as much safe railway function as possible while ventilation capability is reduced.

The same fan can therefore have a different objective depending on state.

During normal service:

comfort + heat + air quality + energy

During a fire:

life safety + smoke control + evacuation tenability

The second objective dominates the first.

The RFE of tunnel ventilation is to put air where the railway needs it, move heat and contaminants away from where people need to be, and change objectives immediately when ordinary comfort becomes life safety.

Prompt 1 — Why Does a Train Behave Like a Piston?

Let tunnel cross-sectional area be A_t.

Let train frontal area be A_tr.

The blockage ratio is:

β = Atr/At

If β is small, the train occupies little of the tunnel.

If β is large, less space remains for air to pass around the vehicle.

The annular or bypass area is approximately:

Agap = At − Atr

A train moving at speed V displaces volume at an idealised rate:

Qdisplaced ≈ Atr V

If all of that displaced air had to move through the gap around the train, a very simple relative-gap velocity scale would be:

ugap ≈ Qdisplaced/Agap
     ≈ Atr V/(At−Atr)

Real underground railways are more complicated because air also moves through shafts, station volumes, leakage paths, cross passages and adjacent tunnel sections, and because the flow is unsteady and partly compressible.

But the equation explains the piston effect.

The train is a moving volume that forces air to rearrange.

A fictional blockage example

Suppose a tunnel has area:

At = 35 m²

and the train frontal area is:

Atr = 15 m²

Then:

β = 15/35
  ≈ 0.429

Gap area is:

Agap = 20 m²

If train speed is 18 m/s:

Qdisplaced ≈ 15×18
           = 270 m³/s

The naive gap-flow scale becomes:

ugap ≈ 270/20
     = 13.5 m/s

This is not a design calculation.

It shows why even a moderate-speed train can produce substantial tunnel air movement.

Pressure scale

A dynamic-pressure scale is:

q = ½ρu²

Using ρ≈1.2 kg/m³ and u=13.5 m/s in the fictional example:

q ≈ 0.5×1.2×13.5²
  ≈ 109 Pa

Again, actual transient tunnel pressures require aerodynamic simulation and measurement.

The scale tells us why platform doors, tunnel structures and ventilation paths experience repeated train-induced pressure changes.

The train creates airflow not because it has a fan on its nose, but because it continually occupies space that the air was using one moment earlier.

Prompt 2 — How Do Tunnel Pressure and Airflow Become a Network?

An underground railway is not one straight pipe.

It contains:

tunnel sections,
stations,
ventilation shafts,
ducts and dampers,
cross passages,
portals,
platform-screen-door leakage paths,
and fans.

This can be represented as a fluid network.

Nodes represent pressure points.

Edges represent airflow paths.

At a node, conservation of mass requires approximately:

Σ Qin − Σ Qout = dM/dt

For low-speed incompressible steady network approximation:

Σ Qin = Σ Qout

Pressure drop through an airflow path often scales approximately with flow squared:

Δp = K Q|Q|

where K represents geometry and resistance.

A fan adds pressure:

Δpfan = Ffan(Q,N)

where N represents fan rotational speed or operating state.

A moving train enters the network as a time-dependent pressure and flow source.

Network state at time t
= tunnels + stations + openings + fans + moving trains

Two trains can interfere

If two trains move through connected tunnel sections at similar times, their pressure fields interact.

One train may push air towards a station while another draws air away.

The resulting airflow is therefore not the sum of two isolated single-train cases if the network boundaries are coupled.

Recent field-tested research on train-induced airflow in subway station tunnels explicitly models train intervals, two-direction train arrival timing, piston-ventilation ducts, circulation paths and platform doors because each changes the network boundary condition.

Pressure waves

Air is compressible.

A pressure disturbance propagates through the tunnel at approximately the local speed of sound.

a = √(γRT)

where γ is the ratio of specific heats, R is the gas constant and T is absolute temperature.

Metro trains travel far below sonic speed, but the confined tunnel still produces transient pressure changes that can reflect at openings and changes in cross-section.

A dimensionless Mach number is:

M = V/a

Even at low M, unsteady pressure and piston effects can matter to ventilation design and passenger comfort.

The tunnel is not a passive container. It is an aerodynamic network whose pressure field is rewritten every time a train moves through it.

Prompt 3 — Why Does Tunnel Airflow Cost Train Energy?

A train must do work to push air.

A simple aerodynamic drag relationship is:

Fdrag = ½ρCd A V²

where C_d is an effective drag coefficient.

Inside a tunnel, effective aerodynamic resistance differs from open-air running because blockage, pressure waves, wall friction and train–tunnel interaction matter.

But the V² structure remains a useful intuition.

Power required to overcome drag is:

Pdrag = Fdrag V

So if drag scales approximately with V²:

Pdrag ∝ V³

This cubic power relationship is one reason high-speed tunnel aerodynamics become expensive rapidly.

A speed comparison

Suppose one aerodynamic state produces drag F at speed V.

Increase speed by 10 per cent:

V2 = 1.1V

If drag follows V²:

F2/F1 = 1.1²
      = 1.21

about 21 per cent more aerodynamic force.

Aerodynamic power ratio becomes:

P2/P1 ≈ 1.1³
      ≈ 1.331

about 33 per cent more aerodynamic power.

Real train energy includes rolling resistance, traction efficiency, gradients, acceleration, auxiliaries and regeneration, so this is not a total-energy prediction.

It shows why tunnel aerodynamics belongs inside the MRT Energy pillar.

Headway changes the airflow field

More frequent trains mean the tunnel is disturbed more often.

At headway H, a simple repeated forcing frequency scale is:

ftrain ≈ 1/H

At H=120 s:

ftrain ≈ 0.00833 Hz

This is not an acoustic frequency. It represents the slow repeating cycle at which large piston-air events arrive.

Shorter headway changes the mean and transient ventilation problem because the tunnel may not fully return to the previous airflow and thermal state before the next train arrives.

Headway is also an aerodynamic variable: every extra train is another moving piston entering the same air network.

Prompt 4 — Where Does the Heat in an Underground MRT Come From?

An underground railway continuously generates heat.

Sources include:

traction and electrical losses,
mechanical losses,
friction braking when used,
train auxiliary systems,
air-conditioning heat rejection,
station equipment,
lighting and escalators,
passengers,
and heat stored in tunnel structures.

The tunnel and station environment therefore obey an energy balance.

Cth dT/dt
= Qgenerated
− ρcp Qvent(T−Tout)
− UA(T−Tground)

where:

C_th is effective thermal capacity,
Q_generated is heat generation,
Q_vent is ventilation airflow,
ρc_p converts airflow and temperature difference into sensible heat transport,
UA represents heat exchange with surrounding structures.

This is a lumped teaching model.

Real underground thermal modelling is spatial and transient.

Ventilation heat removal

Sensible heat carried away by air is approximately:

Q̇air = ρcp Qvent ΔT

Suppose a fictional tunnel generates 1.2 MW of net sensible heat and outside or replacement air is 8°C cooler.

With ρ=1.2 kg/m³ and c_p=1005 J/(kg·K), airflow needed if ventilation alone carried that heat would be:

Qvent
= 1,200,000 /(1.2×1005×8)
≈ 124 m³/s

This is purely illustrative. Actual station and tunnel cooling uses detailed environmental-control design, heat storage and other thermal mechanisms.

Platform screen doors change the thermal boundary

Full-height PSDs separate the conditioned platform from the hotter and dustier tunnel environment.

LTA says this reduces underground station air-conditioning energy consumption.

The train piston still changes tunnel pressure, but the PSD strongly changes how much of that airflow exchanges directly with the platform.

Recent research continues to study this interaction because piston wind can still drive air exchange through entrances, leakage paths and ventilation networks, thereby affecting station cooling load.

The PSD is therefore part of the airflow boundary condition:

train piston pressure
→ tunnel airflow
→ PSD/duct/shaft leakage and openings
→ station infiltration
→ cooling load

This connects directly to How MRT Platform Screen Doors Work Using Mathematics.

An underground station’s cooling problem begins partly on the track, because every train brings motion, heat and pressure into the tunnel beside it.

Prompt 5 — How Much Energy Does Mechanical Ventilation Use?

A fan does work by raising air pressure while moving a volume flow.

Ideal fluid power is:

Pfluid = Δp Q

If overall fan and drive efficiency is η:

Pelectrical = Δp Q/η

Suppose a fictional ventilation fan delivers:

Q = 120 m³/s
Δp = 800 Pa
η = 0.75

Then:

P ≈ 800×120/0.75
  ≈ 128 kW

A large ventilation system can therefore consume substantial electrical power.

Fan affinity laws

For geometrically similar fan operation near comparable system conditions:

Q ∝ N
Δp ∝ N²
P ∝ N³

where N is fan speed.

Increase fan speed by 10 per cent:

Q ratio ≈ 1.10
pressure ratio ≈ 1.21
power ratio ≈ 1.331

A modest increase in flow can therefore require a much larger increase in power.

This creates an energy optimisation problem:

use enough ventilation for temperature and air quality
but
avoid unnecessary fan flow and pressure

Demand-controlled ventilation

LTA says some newer and upgraded stations use carbon-dioxide sensors to adjust fresh-air supply so ventilation more closely matches demand.

A simple indoor concentration balance is:

V dC/dt
= Gpeople
+ Qfresh(Cout−C)

where G_people is pollutant or CO₂ generation by occupants.

If passenger density rises, G rises and more fresh airflow may be required.

The same philosophy applies broadly:

Ventilation should respond to the physical need, not run at maximum simply because maximum is available.

Prompt 6 — How Does Normal Ventilation Become Smoke Control?

This is the most important state transition in the article.

During normal operation, the optimisation target includes comfort and energy.

During a fire, the target changes to life safety.

SCDF’s current Code of Practice for Fire Precautions in Rapid Transit Systems states that fire safety is achieved through an integrated combination of facility design, equipment, procedures and software. For an underground or enclosed trainway, the code requires a non-contaminated environment in the portion used for evacuation and a ventilation system designed to control smoke in the affected trainway.

The public mathematical objective is therefore:

manage pressure and airflow
so smoke is kept away from the intended evacuation path
long enough for people to reach safety

Actual Singapore fan-control sequences, incident scenarios, velocities and emergency procedures are deliberately not reproduced here.

Smoke is an advected scalar

Let smoke concentration be C(x,t).

A simple one-dimensional transport equation is:

∂C/∂t
+ u ∂C/∂x
= D ∂²C/∂x²
+ S(x,t)

where:

u is longitudinal airflow,
D represents mixing or diffusion,
S represents smoke generation.

The ventilation system changes u.

That changes how smoke moves.

Buoyancy

Hot smoke is less dense than surrounding air.

For small temperature differences, an ideal-gas approximation gives:

Δρ/ρ ≈ −ΔT/T

Smoke therefore rises and tends to form a hot layer beneath the tunnel ceiling, while longitudinal airflow can transport it along the tunnel.

Too little directed flow can permit smoke to spread against the intended direction.

But “more airflow” is not automatically better in every scenario. Ventilation can alter flame behaviour, mixing, smoke temperatures and evacuation conditions.

Modern research therefore uses CFD, scaled experiments and network modelling to analyse tunnel-fire smoke under different train positions, fire states and ventilation conditions.

Piston wind complicates emergency airflow

If trains are still moving before a section is fully stabilised, their piston effect can interact with smoke-control airflow.

Recent research on moving-train tunnel fires explicitly studies this interaction.

The conceptual equation is:

utotal
= uventilation
+ upiston
+ ubuoyancy
+ uother disturbances

The emergency-control problem must consider the combined field rather than the fan alone.

Smoke control is not “turn the fans on”. It is shape the airflow field so the human escape path remains more tenable than the contaminated path.

Prompt 7 — How Do Engineers Model Something This Complicated?

No single mathematical model is best for every ventilation question.

Railway ventilation uses several levels of abstraction.

Model type	What it represents	Why it is useful
Lumped thermal model	Average temperature and heat balance	Fast intuition and system-level energy estimates
1D ventilation network	Pressures and flows through tunnel/station branches	Efficient whole-network transient simulation
Train aerodynamic model	Piston effect and tunnel drag	Connects speed, blockage and pressure
CFD	Three-dimensional unsteady air, heat and smoke fields	Detailed local geometry and emergency analysis
Reduced-order model	Approximation of a detailed model	Faster optimisation and real-time evaluation
Data-driven model	Relationships learned from sensors and historical operation	Forecasting, anomaly detection and calibration

The Navier–Stokes foundation

Detailed CFD ultimately solves conservation equations for fluid motion.

Mass conservation:

∂ρ/∂t + ∇·(ρu) = 0

Momentum:

ρ(∂u/∂t + u·∇u)
= −∇p + μ∇²u + ρg + other forces

Energy adds temperature and heat transfer.

Turbulence requires additional modelling or resolution.

A moving train means the computational geometry itself can change with time, which is why dynamic-mesh and moving-boundary methods appear in subway piston-effect research.

Validation

A beautiful CFD colour plot is not proof.

The model must be compared with:

field pressure measurements,
air velocity measurements,
temperature data,
scaled physical experiments where appropriate,
or full-scale system tests.

Let predicted air velocity be û and observed velocity be u.

eu = u − û

Let predicted pressure be p̂ and measured pressure p.

ep = p − p̂

A systematic residual means the model is missing something:

leakage area,
flow resistance,
train geometry,
fan performance,
boundary conditions,
or thermal effects.

The ventilation model is valuable only after real tunnel air has had the chance to disagree with it.

Prompt 8 — How Does the MRT Know Its Ventilation System Is Healthy?

A tunnel ventilation fan is itself a rotating machine.

It has bearings, motors, blades, electrical equipment, dampers, sensors and control interfaces.

It can degrade.

The 2026 Rail Reliability Taskforce condition-monitoring baseline explicitly lists tunnel ventilation fans under station health.

That means ventilation reliability is being pulled into the same predictive-maintenance architecture as lifts, platform screen doors, communications and other critical station assets.

Fan condition vector

A generic monitoring vector might include:

x(t) = [
vibration,
bearing temperature,
motor current,
fan speed,
pressure rise,
airflow,
damper state
]

Expected fan relationship includes:

Δp ≈ F(Q,N)

If measured airflow falls while speed and pressure demand remain similar, the system can investigate whether:

flow resistance changed,
a damper state is incorrect,
fan performance deteriorated,
or a sensor is wrong.

An anomaly should not become automatic diagnosis.

It is evidence for engineering inspection.

Availability

For a repairable fan system:

A ≈ MTBF/(MTBF+MTTR)

High reliability matters because emergency ventilation equipment must not discover its failure only when the emergency begins.

Testing, monitoring and maintenance therefore exist partly for a state that may rarely occur.

A tunnel fan can spend most of its life supporting ordinary air and still be judged by the extraordinary day when its life-safety function is needed.

A Complete Fictional Tunnel-Ventilation Example

Consider a fictional underground MRT section.

Every value below is invented for teaching and does not represent Singapore MRT tunnel dimensions, fan capacities or emergency settings.

Step 1 — Train piston volume

Tunnel area At = 36 m²
Train area Atr = 15 m²
Train speed V = 16 m/s

Blockage ratio:

β = 15/36
  ≈ 0.417

Displaced volume-flow scale:

Qdisplaced ≈ 15×16
           = 240 m³/s

Step 2 — Simplified bypass velocity

Gap area:

Agap = 36−15
     = 21 m²

If all displaced air used the gap:

ugap ≈ 240/21
     ≈ 11.4 m/s

In a real tunnel, some air would be routed through other connected spaces and the flow would be transient.

Step 3 — Pressure scale

With ρ=1.2 kg/m³:

q = ½ρu²
  ≈ 0.5×1.2×11.4²
  ≈ 78 Pa

That pressure scale acts repeatedly as trains pass.

Step 4 — Normal ventilation heat balance

Suppose net tunnel sensible heat to be removed is:

Qheat = 0.9 MW

Replacement air is 6°C cooler.

If ventilation alone carried that heat:

Qvent
= 900,000 /(1.2×1005×6)
≈ 124 m³/s

Real heat transfer to walls, station cooling and transient storage would alter the result.

Step 5 — Fan power

Suppose delivering 124 m³/s requires 700 Pa and overall efficiency is 0.78.

Pfan
= 700×124/0.78
≈ 111 kW

Ventilation is therefore itself an electrical load on the railway.

Step 6 — Increase airflow by 20 per cent

If the same fan family follows affinity-law scaling and airflow rises 20 per cent:

N2/N1 ≈ 1.20
P2/P1 ≈ 1.20³
      ≈ 1.728

Fan power could rise about 73 per cent under this idealised relationship.

This explains why “more air” is an expensive normal-operation strategy.

Step 7 — Emergency state

Now suppose a fire state is declared.

The normal objective:

minimise energy while controlling heat

is replaced by:

establish the authorised smoke-control airflow field
that protects the evacuation route

The fan-energy optimum is no longer the governing objective.

Life safety has changed the cost function.

Step 8 — World Return

Suppose predicted normal airflow is 124 m³/s but measured airflow is 112 m³/s.

eQ = 112−124
   = −12 m³/s

That 9.7 per cent deficit becomes evidence.

The investigation asks whether the cause is:

fan degradation,
unexpected system resistance,
damper position,
sensor error,
or changed network boundary conditions.

The model is corrected only after the physical airflow answers.

The Tunnel-Ventilation Deletion Tests

Remove blockage ratio

The train is treated as though it occupies no tunnel volume and creates no piston effect.

Remove pressure

Airflow has no driving force through tunnels, ducts or leaks.

Remove train speed

Piston airflow and aerodynamic drag stop responding to how fast the train moves.

Remove network topology

Stations, shafts and cross connections are assumed unable to redirect air.

Remove platform screen doors

Tunnel and conditioned-platform environments are assumed to exchange air freely regardless of the real boundary.

Remove heat

The underground railway can operate indefinitely without temperature rising or cooling energy.

Remove fan energy

Unlimited ventilation becomes free.

Remove smoke buoyancy

Fire smoke behaves like a passive cold tracer rather than a hot, density-driven flow.

Remove evacuation receiver

Smoke control optimises airflow without asking whether people can still move through a tenable route.

Remove condition monitoring

A fan needed in an emergency can fail silently until the day it is required.

The Tunnel-Ventilation Paradoxes

Paradox 1 — The train is both the transport object and a ventilation actuator

Every moving train mechanically pumps air through the underground network.

Paradox 2 — More ventilation can cost dramatically more energy

Fan power can scale approximately with the cube of fan speed.

Paradox 3 — Platform screen doors can reduce station cooling load while increasing the importance of tunnel ventilation

Separating the platform from the tunnel protects conditioned station air but leaves more train and equipment heat on the tunnel side to be managed there.

Paradox 4 — Piston wind can save ventilation energy or increase cooling load

Train-induced airflow can provide useful fresh-air exchange, but in a hot humid climate it can also bring outdoor heat and moisture into conditioned spaces.

Paradox 5 — Faster trains can need disproportionately more aerodynamic power

Drag can scale with speed squared and aerodynamic power with speed cubed.

Paradox 6 — The best everyday airflow is not the best emergency airflow

Normal operation optimises comfort and energy; a fire state optimises life-safety smoke control.

Paradox 7 — More airflow is not automatically safer during a fire

Ventilation changes smoke, mixing and thermal fields, so the correct emergency airflow depends on the validated incident scenario and evacuation design.

Paradox 8 — A rarely used emergency fan can be one of the most important assets in the station

Low operating hours do not imply low criticality.

The Tunnel Airflow Audit

What railway state are we in: normal, degraded, maintenance or emergency?
What is the ventilation RFE in that state?
What tunnel and train cross-sectional areas matter?
What is the blockage ratio?
What train speed and direction?
How many trains are in the connected airflow network?
What pressure and airflow measurements are available?
Where can air enter or leave?
What role do platform screen doors play in the boundary?
What pressure-drop coefficients govern each path?
What fan pressure and flow are available?
What fan energy follows?
What heat sources are active?
What heat is stored in the tunnel structure?
What fresh-air requirement is driven by passenger demand?
What does piston airflow contribute naturally?
What undesirable cooling load can that piston airflow create?
What aerodynamic drag is the train paying for?
What happens when headway shortens?
What if one ventilation fan becomes unavailable?
Is fan condition being monitored?
In an emergency, which evacuation zone must remain tenable?
How does train-induced airflow interact with emergency smoke flow?
What measurements validate the airflow model?
What result would force the ventilation model to change?

How the Mathematics Grows from School to Research

Primary Mathematics and Science

area and volume,
speed,
temperature,
air as matter that occupies space.

Secondary Mathematics and Physics

pressure,
density,
flow rate,
energy and power,
quadratic drag relationships,
heat transfer.

Junior College

calculus,
differential equations,
thermodynamics,
fluid continuity,
ideal-gas relationships,
oscillatory pressure waves.

University and Research

fluid mechanics,
computational fluid dynamics,
turbulence,
thermodynamics and HVAC,
transient ventilation-network modelling,
fire dynamics,
smoke transport,
fan and duct optimisation,
reliability engineering,
data-driven surrogate modelling.

A Primary student learns that pushing an object into water displaces water.

The underground-rail engineer asks what happens when a 100-metre moving object continuously displaces air through a kilometres-long network while another train, a station, a fan, a platform barrier and a possible smoke plume are all changing the boundary conditions at once.

The World Return — When Tunnel Air Answers the Simulation

The model predicts pressure.

Sensors return pressure.

The model predicts airflow.

Anemometers or system measurements return airflow.

The model predicts tunnel temperature.

Temperature sensors return the real thermal field.

The model predicts fan performance.

Current, speed, pressure and vibration reveal the fan’s actual condition.

eQ = Qobserved − Qpredicted

ep = pobserved − ppredicted

eT = Tobserved − Tpredicted

If piston airflow is consistently higher than predicted, the tunnel or opening resistance may be wrong.

If station cooling load is larger than predicted, infiltration or heat-source assumptions may be incomplete.

If a fan produces less airflow at the expected speed, asset condition or network resistance may have changed.

The World Return loop is:

model tunnel and train
→ predict pressure, flow and heat
→ operate
→ measure
→ compare
→ inspect model and equipment
→ recalibrate or maintain
→ operate again

Emergency models require the same discipline through approved testing, commissioning and scenario validation.

The tunnel is the final examiner. If the measured airflow disagrees with the beautiful simulation, the simulation loses.

RFE Return — What Does Good Tunnel Ventilation Owe the Passenger?

The passenger should usually notice almost none of it.

They should not need to know the tunnel pressure field.

They should not need to think about fan curves.

They should not need to know whether one train is pushing air towards a station while another is pulling it away.

The public outcome should be:

manageable underground temperature
+
adequate fresh air
+
controlled pressure and drafts
+
efficient use of fan and cooling energy
+
reliable ventilation equipment
+
validated emergency smoke-control capability

The ventilation system also owes the passenger restraint.

It should not use maximum energy when a smaller airflow is enough.

It should not optimise ordinary comfort in a way that compromises emergency readiness.

And it should not treat a rare emergency function as unimportant merely because it is rarely called upon.

The RFE of MRT tunnel ventilation is to make an underground railway feel like usable human space even though every train is continually disturbing the air—and to be ready for the moment when controlling that air becomes part of saving life.

Conclusion — Every Underground Train Moves Two Things

An MRT train moves passengers.

It also moves air.

The first movement is obvious.

The second is hidden.

As the train enters a tunnel, pressure rises ahead.

Air accelerates through the remaining gap.

Some air moves towards shafts and stations.

A low-pressure region develops behind the vehicle.

Another train changes the field.

Platform screen doors alter the boundary.

Fans add pressure where natural piston flow is insufficient or undesirable.

Heat accumulates from trains and equipment.

Ventilation and cooling remove it.

Energy is consumed to move the air.

Condition monitoring watches the fans that make the system dependable.

And if the railway enters a fire state, the objective changes immediately from comfort and efficiency to the controlled movement of smoke around a human evacuation path.

train movement
→ piston airflow
→ tunnel pressure network
→ aerodynamic drag
→ heat and air exchange
→ ventilation fans
→ station cooling and energy
→ monitored equipment
→ emergency smoke-control state
→ safe human receiver

The passenger sees a train disappear into a tunnel.

The fluid dynamicist sees a moving boundary.

The HVAC engineer sees heat and airflow.

The energy engineer sees fan power.

The fire-safety engineer sees tenability and smoke movement.

The maintenance engineer sees tunnel ventilation fan condition.

The passenger feels a small rush of air and thinks almost nothing of it.

An underground MRT tunnel works because mathematics turns the air displaced by every train from an invisible disturbance into a controlled part of the railway.

Key Equations

β = Atr/At
Tunnel blockage ratio

Agap = At−Atr
Bypass area around train

Qdisplaced ≈ Atr V
Piston-displacement flow scale

ugap ≈ AtrV/(At−Atr)
Simplified bypass velocity

q = ½ρu²
Dynamic-pressure scale

ΣQin−ΣQout = dM/dt
Air-mass conservation at a node

Δp = KQ|Q|
Simplified ventilation-network resistance

a = √(γRT)
Speed of sound

M = V/a
Mach number

Fdrag = ½ρCdAV²
Aerodynamic drag scale

Pdrag = FdragV
Aerodynamic power

Cth dT/dt = Qgenerated−ρcpQvent(T−Tout)−UA(T−Tground)
Simplified tunnel thermal balance

Q̇air = ρcpQventΔT
Sensible heat carried by ventilation

Pfan = ΔpQ/η
Fan electrical power

Q∝N, Δp∝N², P∝N³
Fan affinity laws

V dC/dt = G + Q(Cout−C)
Simple fresh-air concentration balance

∂C/∂t + u∂C/∂x = D∂²C/∂x² + S
Simplified smoke-transport equation

Δρ/ρ ≈ −ΔT/T
Thermal buoyancy approximation

∂ρ/∂t + ∇·(ρu)=0
Fluid mass conservation

ρ(∂u/∂t+u·∇u)=−∇p+μ∇²u+ρg+...
Fluid momentum equation

A ≈ MTBF/(MTBF+MTTR)
Generic fan-system availability

eQ = Qobserved−Qpredicted
ep = pobserved−ppredicted
eT = Tobserved−Tpredicted
World Return residuals

Reader-safety note: All fictional tunnel areas, fan flows, pressure values, heat loads and operating examples are educational abstractions. This article does not reproduce Singapore MRT tunnel-ventilation topology, actual fan capacities, emergency fan-control sequences, smoke-control velocities, fire-design parameters, damper logic, evacuation procedures or other security- or life-safety-sensitive operating details.

Continue the MRT Mathematics Cloud

How MRT Works | It’s Mathematics — the permanent whole-system hub.
How MRT Timing Works Using Mathematics — the system of seconds.
How MRT Braking Works Using Mathematics — stopping distance and braking.
How MRT Headway Works Using Mathematics — train spacing and throughput.
How MRT Station Dwell Time Works Using Mathematics — passenger exchange.
How MRT Delays Propagate and Recover Using Mathematics — instability and recovery.
How an MRT Timetable Is Built Using Mathematics — service planning.
How MRT Energy Use Is Optimised Using Mathematics — traction energy and coasting.
How MRT Passenger Capacity Is Calculated Using Mathematics — delivered line capacity.
How MRT Routes and Transfers Are Optimised Using Mathematics — graph routing.
How MRT Network Resilience Is Measured Using Mathematics — system resilience.
How MRT Predictive Maintenance Works Using Mathematics — condition prediction.
How MRT Track Inspection Works Using Mathematics — continuous track health.
How MRT Wheel–Rail Contact Works Using Mathematics — contact forces.
How MRT Noise and Vibration Work Using Mathematics — waves and resonance.
How MRT Power Supply Works Using Mathematics — traction electricity.
How MRT Signalling and Train Regulation Work Using Mathematics — CBTC and regulation.
How MRT Platform Screen Doors Work Using Mathematics — precision interface and passenger flow.

The next natural pillar is How MRT Depots and Fleet Operations Work Using Mathematics: fleet availability, train assignment, maintenance windows, stabling, dispatch, spare ratio, depot queues, cleaning, inspections and how the trains passengers see every morning have to be manufactured again as an operating fleet every night.

How MRT Tunnel Ventilation and Airflow Work Using Mathematics: Why an Underground Train Behaves Like a Moving Piston