Emissions

This is where every earlier lesson comes back to present its bill. The aerodynamics set the mixing, the injector sets the initial fuel distribution, the cooling design allocates the air, and the stability loop limits how lean we dare go. All of it lands, measurably, in the exhaust, where it is compared with the ICAO CAEP limits for NOₓ, CO, unburned hydrocarbons and smoke.

NOₓ: a temperature problem

Nitrogen is the most inert thing in air, until it is hot enough. Almost all NOₓ from a gas turbine is thermal NO, made by the Zeldovich mechanismZeldovich mechanismThe chain O + N₂ → NO + N, N + O₂ → NO + O that makes thermal NO; its rate ∝ exp(−69 090/T).:

$\mathrm{O} + \mathrm{N_2} \;\rightleftharpoons\; \mathrm{NO} + \mathrm{N}$ $\mathrm{N} + \mathrm{O_2} \;\rightleftharpoons\; \mathrm{NO} + \mathrm{O}$ $\mathrm{N} + \mathrm{OH} \;\rightleftharpoons\; \mathrm{NO} + \mathrm{H}$

The first step, breaking the N≡N triple bond, is fiercely hard, with an activation temperature of about 38,000 K. Combined with the temperature-sensitive equilibrium supply of O atoms, the overall formation rate goes as

$\frac{d[\cnox{\mathrm{NO}}]}{dt} \;\propto\; [\mathrm{O_2}]^{1/2}\,[\mathrm{N_2}]\;\frac{1}{\sqrt{\cflame{T}}}\exp\!\left(-\frac{69\,090}{\cflame{T}}\right)$

That exponential is Lesson 2's Arrhenius tail in its most extreme form.

So thermal NOₓ needs three things: high temperature, time at high temperature, and available oxygen. Take any one away and it collapses.

Two minor routes complete the picture. Prompt NO (Fenimore) forms quickly in fuel-rich flame fronts, where hydrocarbon radicals attack N₂. Fuel NO comes from nitrogen bound in the fuel itself, which is negligible for kerosene but relevant for some alternative fuels, such as ammonia.

Higher engine pressure ratios make matters worse. Hotter compressor air lifts the whole flame-temperature curve (Lesson 2), and for conventional combustors the emission index grows roughly as $\text{EI}_{\mathrm{NO_x}} \propto P_3^{\,0.5}\, e^{T_3/\text{const}}$. The thermodynamic push for efficiency and the environmental push for low NOₓ are in direct conflict.

CO and unburned hydrocarbons: chemistry that didn't finish

CO and UHC are the opposite failure: incomplete combustion. The chemistry started but was frozen before it completed, because the gas was

• too cool: near the cooling films (Lesson 8), or in a lean, low-power primary zone;
• too quick: quenched by dilution air before CO could burn to CO₂;
• too rich: short of oxygen;
• too wet: large droplets at idle still evaporating as they leave (Lesson 6).

CO oxidation is itself strongly temperature-dependent, but in the other direction: it speeds up with temperature. Put the two pollutants on the same axes:

This is the most important plot in emissions. NOₓ climbs steeply to the right, CO climbs steeply to the left, and between them lies a window, typically around 1650–1850 K, where both are low. Lengthen the residence time and the window shifts: CO has more time to burn out, but NOₓ has more time to form.

Soot: formed rich, destroyed hot

Soot is born in fuel-rich, hot regions, where fuel molecules crack (pyrolysis), assemble into polycyclic aromatic hydrocarbons, nucleate into tiny particles, and grow by surface reactions and collisions. At the same time, soot is oxidised by OH and O₂ wherever it meets hot, oxygen-rich gas. The smoke leaving the engine is the small difference between two large numbers, formation minus burnout, which is why the intermediate zone, where rich products meet fresh air while still hot, is so critical. Aromatic content in the fuel is a strong soot precursor (Lesson 10).

Three routes through the danger zone

If the enemy is time spent near φ ≈ 1, then every low-NOₓ strategy is a plan for the path the gas takes through equivalence-ratio space as it moves through the combustor. That is the picture to keep:

Lean premixed (LPM / LDI)

Lean premixedLean premixedPremixing fuel and air lean before burning so no gas ever sees stoichiometric temperatures. combustion is the most direct attack.

The direct solution: mix fuel and air thoroughly before burning, at φ ≈ 0.5–0.6, so that no parcel ever sees stoichiometric conditions. NOₓ reductions of 60–90% relative to conventional combustors are possible, and there is almost no soot.

The price was foreshadowed in earlier lessons: flashback and autoignition in the premixer at high pressure and temperature; combustion instability (Lesson 7); and poor stability at low power, close to lean blowout (Lesson 5). Industrial "dry low-NOₓ" gas turbines, which spend most of their lives at one steady operating point, are built this way. In aero engines, lean-burn systems such as GE's TAPS (Twin Annular Premixing Swirler) in the GEnx and CFM LEAP pair a lean main stage with a pilot.

Rich-burn, Quick-quench, Lean-burn (RQL)

RQLRQLRich-burn, Quick-quench, Lean-burn: a low-NOₓ strategy that jumps across the stoichiometric band as fast as possible. is the cleverer path.

A cleverer path: jump over the stoichiometric band.

1. Rich primary zone ($\phi \approx 1.5$–$2$): very stable, and oxygen-starved, so little NOₓ forms. Soot does form, but it will be burned later.
2. Quick quench: a row of strong jets drives the mixture from rich to lean in the shortest possible distance, spending as little time as possible near φ = 1.
3. Lean burnout ($\phi \approx 0.5$): CO and soot are oxidised below the NOₓ threshold.

The whole design reduces to one question: how fast can the quench jets mix? Drag the quench length in the widget to see the benefit evaporate as the quench slows. Pratt & Whitney's TALON combustors are the best-known RQL family.

Staged combustion

Aircraft engines must run from idle to take-off, and no single fixed zone is ideal across that range. Staged combustors carry two personalities:

• At low power, only a pilot zone burns, richer and very stable, to guarantee ignition, stability and CO burnout.
• At high power, the lean main stage takes most of the fuel, delivering low NOₓ where the engine spends its highest-temperature time.

Staging can be radial (separate domes, as in some double-annular designs) or internal (pilot and main within each injector, as in TAPS). The fuel-split schedule between the stages is itself a major design and control problem.

Every lesson so far quietly assumed Jet-A. The final lesson asks what happens when the fuel itself changes.