The Physics of Breathing: How Your Lungs Move 10,000 Litres of Air Every Day

Every Four Seconds

You’re doing it right now. You’ve been doing it since the moment you were born, and you’ll keep doing it until you stop — at which point, everything else stops too. You breathe about 15 times per minute, roughly 900 times per hour, approximately 21,600 times per day. Over a lifetime, somewhere north of 600 million breaths.

And each one is a physics problem.

Not a simple one, either. Each breath involves fluid dynamics (air flowing through branching tubes), thermodynamics (gas mixtures at varying pressures and temperatures), surface tension (the force that would collapse your lungs without a molecular countermeasure), and diffusion (oxygen crossing a membrane so thin that light passes through it). The engineering is extraordinary. The fact that you never think about it is perhaps the most extraordinary thing of all.

Boyle’s Law in Your Chest

Breathing starts with a pressure difference. Not a big one — about 1–3 cmH₂O, which is roughly 100–300 pascals, or about 0.1–0.3% of atmospheric pressure. That’s all it takes.

The diaphragm — a dome-shaped sheet of skeletal muscle at the base of the thoracic cavity — does most of the work. When it contracts, it flattens and pulls downward, increasing the volume of the chest cavity by about 250–500 millilitres. The external intercostal muscles between the ribs assist by pulling the rib cage upward and outward, adding another 100–200 mL of volume expansion.

Here’s where Boyle’s law takes over. At constant temperature, the pressure of a trapped gas is inversely proportional to its volume: PV = constant. Expanding the thoracic volume reduces the pressure inside the lungs (the intrapulmonary pressure) to slightly below atmospheric. Air, being a fluid, flows from higher pressure to lower pressure. You inhale.

The flow rate during quiet breathing is about 0.5 litres per second. During exercise, it can reach 6–8 L/s. During a cough, momentary peak flows can hit 28 L/s — enough to expel mucus and foreign particles at velocities approaching 500 km/h in the trachea. That’s not far off the speed of sound.

Exhalation at rest is mostly passive. The diaphragm relaxes. The elastic recoil of the lung tissue — stretched during inhalation like a rubber band — pulls the lungs back to their resting volume. Intrapulmonary pressure rises slightly above atmospheric, and air flows out. No muscular effort required. You breathe out simply by letting go.

During exercise or forced exhalation, the internal intercostals and abdominal muscles actively compress the chest, generating pressures up to 40–100 cmH₂O. This is what powers shouting, singing, playing wind instruments, and the Valsalva manoeuvre.

The Airway Tree: 23 Generations of Branching

Air doesn’t just rush into a big empty sac. It flows through a branching network of airways — the tracheobronchial tree — that divides 23 times from the trachea to the alveoli.

The trachea (generation 0) is about 12 mm in diameter and 12 cm long. It divides into two main bronchi (generation 1), which divide into lobar bronchi (generation 2), then segmental bronchi, then progressively smaller bronchioles. By generation 16, the airways are about 0.5 mm in diameter — the terminal bronchioles, the smallest airways without alveoli. Generations 17–23 are the respiratory bronchioles and alveolar ducts, where gas exchange begins to occur.

The fluid dynamics of this branching network are fascinating. In the trachea, airflow is relatively fast (about 1.5 m/s during quiet breathing) and the Reynolds number is around 2,000 — right at the transition between laminar and turbulent flow. You can hear this turbulence: it’s the sound of breathing.

But as the air moves deeper into the lungs, something remarkable happens. Each generation roughly doubles the total cross-sectional area of the airways (one tube becomes two, but each daughter tube is only slightly smaller than the parent). By the time air reaches the alveolar ducts at generation 23, the total cross-sectional area has expanded from about 2.5 cm² (trachea) to approximately 12,000 cm² — a nearly 5,000-fold increase.

This means the airflow velocity plummets. In the terminal bronchioles, the velocity is about 1 cm/s. In the alveolar region, it’s essentially zero — the air is nearly stationary. At this point, gas transport is no longer driven by bulk flow (convection) but by molecular diffusion. Oxygen molecules simply diffuse down their concentration gradient from the airway into the alveolar gas. The transition from convective transport to diffusive transport happens around generation 17, in a zone called the diffusion front.

This is elegant engineering. The branching network uses convection to move air rapidly through the large airways (where resistance would be high for diffusion alone) and then switches to diffusion for the final millimetres (where the enormous surface area makes diffusion extremely efficient). Two different transport mechanisms, seamlessly integrated by geometry.

The Alveoli: 300 Million Bubbles

At the end of the airway tree are the alveoli — tiny, thin-walled air sacs where gas exchange actually happens. There are roughly 300 million of them in an adult lung, and they provide a total surface area of about 70 square metres. That’s roughly the area of a singles tennis court, folded and packed into a space smaller than a football.

Each alveolus is roughly spherical, about 200 micrometres in diameter — visible under a basic microscope but too small to see with the naked eye. The alveolar wall is astonishingly thin: the air-blood barrier consists of the alveolar epithelium (a single cell layer, about 0.1–0.2 µm thick), a fused basement membrane, and the capillary endothelium (another single cell layer). Total thickness: about 0.2–0.5 micrometres.

To put that in perspective: a wavelength of visible light is about 0.4–0.7 µm. The membrane separating your blood from the air is thinner than the wavelength of red light. If it were any thinner, it would lose structural integrity. Evolution has pushed this membrane to approximately the physical limit of how thin a biological barrier can be while remaining mechanically stable.

Fick’s Law: Diffusion Across the Membrane

Oxygen crosses the alveolar-capillary membrane by passive diffusion — no pumps, no energy expenditure, no active transport. The driving force is a difference in partial pressure.

Dalton’s law tells us that the total pressure of a gas mixture equals the sum of the partial pressures of each component. Atmospheric air is roughly 78% nitrogen, 21% oxygen, 0.9% argon, and 0.04% carbon dioxide. At sea level (total pressure 101.3 kPa), the partial pressure of oxygen is about 21.2 kPa, or 160 mmHg.

By the time inhaled air reaches the alveoli, it’s been warmed to body temperature (37°C) and fully saturated with water vapour (vapour pressure 6.3 kPa at 37°C). It’s also mixed with residual gas from the previous breath. The result: alveolar oxygen partial pressure is about 13.3 kPa (100 mmHg), and alveolar CO₂ partial pressure is about 5.3 kPa (40 mmHg).

Deoxygenated blood arriving at the pulmonary capillaries has an oxygen partial pressure of about 5.3 kPa (40 mmHg) and a CO₂ partial pressure of about 6.1 kPa (46 mmHg).

The partial pressure gradient for oxygen is 100 − 40 = 60 mmHg, driving oxygen from alveolar air into the blood. The gradient for CO₂ is 46 − 40 = 6 mmHg, driving CO₂ from blood into alveolar air.

Fick’s law of diffusion governs the flux:

J = D × A × ΔP / T

where J is the rate of gas transfer, D is the diffusion coefficient (which depends on the gas’s molecular weight and solubility in the membrane), A is the surface area, ΔP is the partial pressure difference, and T is the membrane thickness.

The lungs are optimised on every term. Surface area A is enormous (70 m²). Thickness T is minimised (0.3 µm average). The pressure gradient ΔP is maintained by continuous ventilation (bringing fresh air) and perfusion (bringing deoxygenated blood). The only term the lungs can’t control is the diffusion coefficient D — but here CO₂ has a hidden advantage.

CO₂ diffuses about 20 times faster than O₂ across biological membranes, despite its larger molecular weight. The reason: CO₂ is far more soluble in water and tissue than O₂. Solubility dominates over molecular weight in membrane diffusion. This is why CO₂ can be efficiently eliminated with a pressure gradient of only 6 mmHg while oxygen requires 60 mmHg. It’s also why CO₂ retention is rarely a problem until lung disease is very advanced — the CO₂ diffusion capacity has an enormous safety margin.

The exchange is fast. Blood spends about 0.75 seconds in the pulmonary capillaries (at rest). Oxygen equilibration — the point where blood oxygen partial pressure matches alveolar oxygen partial pressure — occurs in about 0.25 seconds. That leaves a threefold safety margin. During intense exercise, cardiac output increases and blood transit time through the capillaries shortens to about 0.25 seconds, but equilibration still usually completes. It’s only at extreme altitude, during maximal exertion, that the system runs out of margin and blood leaves the capillaries not fully oxygenated.

Surfactant: The Molecule That Saves Every Breath

Here’s a problem that isn’t obvious until you think about the physics.

The alveoli are lined with a thin film of water — they have to be, because the alveolar cells are living tissue that must remain moist. But water has a surface tension of about 72 mN/m at 37°C. And the Law of Laplace tells us that the inward pressure generated by surface tension in a sphere is:

P = 2γ / r

For a 200 µm alveolus lined with pure water, this gives P = 2 × 0.072 / 0.0001 = 1,440 Pa, or about 14.7 cmH₂O. That’s a substantial inward pressure trying to collapse the alveolus. And it gets worse: the Laplace pressure is inversely proportional to radius. Smaller alveoli have higher internal pressure than larger ones. If two alveoli of different sizes are connected (they are — they share common airways), the smaller one would empty into the larger one, the larger one would grow, and the smaller one would collapse. This positive feedback would cause progressive alveolar collapse throughout the lung.

This doesn’t happen because of pulmonary surfactant.

Surfactant is produced by type II alveolar cells (which make up about 5% of the alveolar surface area but are metabolically critical). It’s a complex mixture: about 90% lipids (mainly dipalmitoylphosphatidylcholine, DPPC) and 10% proteins (surfactant proteins A, B, C, and D, each with specific functions).

DPPC molecules are amphiphilic — they have a hydrophilic phospholipid head and two hydrophobic fatty acid tails. At the air-water interface, they arrange with heads in the water and tails pointing into the air, forming a molecular monolayer. This monolayer reduces the surface tension from 72 mN/m to about 25 mN/m during normal breathing.

But the really clever part is what happens during expiration. As the alveolus shrinks, the surfactant molecules are compressed together at the interface. Their concentration per unit area increases. At high compression, the surface tension drops to near zero — some measurements give values below 2 mN/m. This is critical because it’s during expiration, when the alveolus is at its smallest radius, that the Laplace pressure is highest. By driving surface tension to near zero precisely when r is smallest, surfactant neutralises the Laplace pressure at exactly the moment it’s most dangerous.

Surfactant also solves the small-alveolus-empties-into-large-alveolus instability. Because smaller alveoli compress their surfactant more (smaller surface area at a given compression), they achieve lower surface tension. This automatically equalises the Laplace pressure between alveoli of different sizes, stabilising the system.

When Surfactant Is Missing

Neonatal respiratory distress syndrome (NRDS) is what happens when surfactant isn’t there. Premature infants born before about 34 weeks of gestation often have immature type II cells that don’t produce enough surfactant. Without it, the alveoli tend to collapse with each expiration, and enormous muscular effort is required to reinflate them. Each breath is a battle against Laplace pressure. Without treatment, the infant exhausts quickly.

The treatment — exogenous surfactant delivered directly into the airways — was developed in the 1980s and is one of the great successes of biophysics-informed medicine. Understanding the physics of surface tension in curved interfaces directly saved millions of premature babies’ lives.

The Mechanics: Compliance and Resistance

The mechanical work of breathing is spent overcoming two forces: elastic recoil (the tendency of the lungs and chest wall to return to their resting position) and airway resistance (the frictional resistance to airflow in the branching tubes).

Compliance

Lung compliance is the change in volume per unit change in pressure: C = ΔV / ΔP. A highly compliant lung is easy to inflate; a stiff lung requires more pressure for the same volume change. Normal lung compliance is about 200 mL/cmH₂O — meaning a pressure change of 1 cmH₂O expands the lung by 200 mL.

Compliance depends on two factors: the elastic properties of the lung tissue (collagen and elastin fibres in the alveolar walls) and the surface tension forces in the alveoli. In fact, surface tension accounts for about two-thirds of the lung’s elastic recoil at normal volumes. Surfactant, by reducing surface tension, increases compliance — making breathing easier. Without surfactant, compliance drops dramatically, and the work of breathing increases by a factor of 3–5.

The compliance of the chest wall also matters. The chest wall has its own elastic properties — it naturally springs outward. At the resting lung volume (functional residual capacity), the inward recoil of the lung exactly balances the outward recoil of the chest wall, and the respiratory muscles can relax completely. This is the volume your lungs settle to at the end of a passive expiration.

Airway Resistance

Airflow through the airways follows the principles of fluid dynamics. For laminar flow in a tube, the Poiseuille equation gives:

R = 8ηL / πr⁴

where R is resistance, η is the viscosity of air, L is the tube length, and r is the radius. The critical insight is the fourth-power dependence on radius: halving the radius increases resistance sixteenfold. This is why even modest narrowing of the airways — from asthma, bronchitis, or mucus — dramatically increases the work of breathing.

In practice, about 80% of total airway resistance occurs in the first seven generations of airways (trachea through medium bronchi), where the tubes are relatively narrow and flow is turbulent or transitional. The tiny bronchioles, despite their individually small radius, contribute little resistance because there are so many of them in parallel — and resistances in parallel add reciprocally.

Asthma illustrates the physics vividly. Bronchospasm (smooth muscle contraction around the bronchioles) narrows the airway radius. If the radius decreases by just 20%, the resistance increases by (1/0.8)⁴ ≈ 2.4-fold. A 50% narrowing increases resistance by (1/0.5)⁴ = 16-fold. The wheeze of asthma is the sound of air forced at high velocity through narrowed tubes — the acoustic signature of Poiseuille’s r⁴ law in action.

Altitude: When the Physics Changes

At sea level, the atmosphere pushes on you with 101.3 kPa of pressure, and every breath delivers oxygen at a partial pressure of about 21 kPa. The system works beautifully.

At altitude, the same system faces a different set of numbers.

Atmospheric pressure drops roughly exponentially with altitude, halving approximately every 5,500 metres. The oxygen fraction stays the same (20.9%), but the partial pressure drops proportionally:

At 2,500 m (many ski resorts): P_O₂ ≈ 15.5 kPa. Mild effects — slightly increased breathing rate, reduced exercise capacity.

At 3,500 m (La Paz, Bolivia): P_O₂ ≈ 13.7 kPa. Noticeable breathlessness on exertion. Altitude sickness possible.

At 5,500 m (Everest base camp): P_O₂ ≈ 10.5 kPa. Half the sea-level value. Severe exercise limitation without acclimatisation.

At 8,849 m (Everest summit): P_O₂ ≈ 7.1 kPa. One-third of sea-level. Without supplemental oxygen, the partial pressure gradient driving diffusion across the alveolar membrane is so small that blood cannot be fully oxygenated, even with maximal hyperventilation. The climbers who reach the summit without oxygen operate on the ragged edge of the system’s physical limits.

The body acclimatises through several mechanisms, all of which the physics predicts: increased ventilation (raise the alveolar O₂ partial pressure by breathing out more CO₂), increased red blood cell production (more haemoglobin to carry oxygen despite lower partial pressure), increased capillary density in tissues (shorter diffusion distances), and increased 2,3-DPG in red blood cells (shifting the oxygen-haemoglobin dissociation curve to release oxygen more readily to tissues).

Every one of these adaptations is the body adjusting one of the variables in Fick’s law or the oxygen dissociation equation to compensate for the reduced driving pressure that altitude imposes.

Oxygen and Haemoglobin: A Cooperative Binding Curve

Oxygen doesn’t just dissolve in blood plasma — that would be hopelessly insufficient. At body temperature and a partial pressure of 100 mmHg, only about 3 mL of oxygen dissolves per litre of plasma. The body’s resting oxygen consumption is about 250 mL per minute. With a cardiac output of 5 litres per minute, dissolved oxygen could supply only 15 mL/min — about 6% of the requirement.

Haemoglobin solves this. Each haemoglobin molecule (four protein subunits, each carrying one haem group with an iron atom at its centre) can bind four oxygen molecules. With about 150 grams of haemoglobin per litre of blood, fully saturated blood carries about 200 mL of oxygen per litre — nearly 70 times more than dissolved oxygen alone.

The binding follows a sigmoidal (S-shaped) dissociation curve, not a simple linear relationship. This cooperativity is a consequence of the physics of protein conformational change: when the first O₂ binds to one haem group, it changes the protein’s conformation in a way that makes the second O₂ bind more easily, and so on. The Hill coefficient for haemoglobin is about 2.8 (maximum possible: 4), indicating strong cooperativity.

The shape of this curve is exquisitely suited to the body’s needs. At alveolar P_O₂ (100 mmHg), haemoglobin is about 98% saturated — it loads up almost completely in the lungs. At tissue P_O₂ (about 40 mmHg at rest), saturation drops to about 75% — so haemoglobin releases about 25% of its oxygen to the tissues. During exercise, tissue P_O₂ falls further (to about 20 mmHg), saturation drops to about 35%, and haemoglobin releases about 65% of its oxygen. The steep part of the sigmoid curve falls exactly in the physiological range of tissue P_O₂, maximising oxygen delivery where it’s needed most.

This isn’t luck. It’s the result of billions of years of evolutionary optimisation of protein structure to match the physical constraints of gas exchange.

The Control Loop: Why You Can’t Hold Your Breath Forever

Breathing is automatic — controlled by the respiratory centre in the brainstem (medulla oblongata and pons). But the control variable might surprise you.

You’d think the body monitors oxygen levels and increases breathing when oxygen drops. It does, but this is not the primary driver. The main signal that regulates breathing rate is carbon dioxide concentration — specifically, the pH of cerebrospinal fluid, which changes as dissolved CO₂ reacts with water to form carbonic acid:

CO₂ + H₂O ⇌ H₂CO₃ ⇌ H⁺ + HCO₃⁻

Central chemoreceptors in the medulla detect the resulting change in H⁺ concentration (pH) with extreme sensitivity. A rise in arterial CO₂ partial pressure of just 1–2 mmHg (from a resting level of about 40 mmHg) triggers a noticeable increase in breathing depth and rate. A rise of 5 mmHg feels distinctly uncomfortable. At arterial CO₂ of about 60–70 mmHg, the sensation becomes unbearable.

This is why you can’t hold your breath indefinitely. It’s not oxygen deprivation that forces you to breathe — it’s CO₂ accumulation. The urge to breathe is driven by rising CO₂ and falling pH, not by falling O₂. Peripheral chemoreceptors in the carotid bodies do sense low oxygen, but they kick in mainly when O₂ drops below about 60 mmHg — a level that represents significant hypoxia and a much later warning signal.

The control loop is a classic negative feedback system: increased CO₂ → increased H⁺ → chemoreceptor stimulation → increased ventilation → increased CO₂ exhalation → CO₂ drops back to setpoint. The response time is about 20–30 seconds, fast enough to track changes in metabolic rate during exercise with minimal lag.

The Work of Breathing

Breathing requires energy. At rest, the respiratory muscles consume about 3–5% of total body oxygen consumption — roughly 1–2 watts of mechanical power. This is remarkably efficient for a system processing 6–10 litres of air per minute against elastic and resistive forces.

During heavy exercise, ventilation can increase to 100–150 L/min, and the energy cost of breathing rises to 10–15% of total oxygen consumption. At extremely high ventilation rates, the respiratory muscles themselves become a significant metabolic load — consuming oxygen that could otherwise go to locomotor muscles. This is one of the factors that limits maximal exercise capacity: at some point, breathing harder costs more oxygen than it delivers.

The total work of one breath at rest is about 0.3–0.5 joules — roughly the kinetic energy of a small coin dropped from table height. Three hundred million alveoli, 70 square metres of surface area, 300 millilitres of gas exchange — all for half a joule. The most important physics in your body runs on pocket change.

What Breathing Teaches Us

The physics of breathing is the physics of optimisation under constraint. The airway tree minimises dead space while maximising surface area. The alveolar membrane is as thin as structural integrity allows. Surfactant reduces surface tension dynamically, solving a stability problem that pure physics would make insurmountable. Haemoglobin’s cooperative binding curve is tuned to the exact partial pressure range that physiology requires.

Every component makes sense only when you understand the physics it evolved to satisfy. The branching geometry minimises the total work of breathing — Murray’s law applies to airways just as it does to blood vessels. The transition from convective to diffusive transport happens exactly where the cross-sectional area makes diffusion more efficient than bulk flow. The surfactant system neutralises the Laplace pressure at the exact moment — end-expiration, minimum radius — when it’s most dangerous.

You’ve taken roughly a dozen breaths while reading this article. Each one moved about half a litre of gas through a fractal tree of tubes, delivered it to 300 million microscopic bubbles lined with a surface tension reducer, diffused oxygen across a membrane thinner than light, loaded it onto a cooperatively binding protein, and sent it off to fuel every cell in your body. The whole process took about four seconds and cost less energy than lifting a pencil.

Physics doesn’t get more personal than this.