What duty cycle is and why 85% is the number everyone quotes
An injector is a solenoid valve that is either open or shut. The ECU controls fuelling by how long it holds it open, and duty cycle expresses that on-time as a percentage of the time available in one engine cycle. At 50% duty the injector is open half the cycle; at 100% it never closes.
The 85% convention exists because an injector is not linear near either end of its range. Opening and closing each take a finite time — typically around a millisecond in total for a saturated injector — during which the flow is neither zero nor the rated figure. As duty cycle climbs, the closing event starts to overlap the next opening, and the delivered fuel stops tracking the commanded pulse width. The ECU asks for 3% more fuel and gets 1%.
The failure mode this produces is dangerous specifically because it is progressive. The mixture leans out gradually as the engine climbs through its power band, which is exactly where cylinder pressure and detonation risk are highest. An engine that made a clean pull to 6,000 rpm and then leaned out and dropped a piston at 6,800 has very often simply run its injectors out of time.
The three calculations behind the number
Fuel demand. An engine's fuel consumption is described by brake specific fuel consumption — pounds of fuel per horsepower per hour. Multiply BSFC by horsepower and you have the fuel mass flow rate the engine needs, in the same lb/hr units that injectors are rated in. A 500 hp engine at 0.50 lb/hp·hr wants 250 lb/hr of fuel.
Duty cycle. Divide that demand by the total capacity of the injectors: number of injectors times the flow of each. 250 ÷ (8 × 42) = 250 ÷ 336 = 74.4%. Notice what is not in this expression: rpm. Duty cycle is a ratio of mass flows, so it depends only on power, BSFC and injector capacity.
The injection window. This is where rpm enters. A four-stroke engine completes one cycle every two crank revolutions, so at N rpm the cycle takes 2 × 60,000 ÷ N milliseconds, which is 120,000 ÷ N. At 6,500 rpm that is 18.46 ms. Multiply by duty cycle and you have the actual pulse width: 0.744 × 18.46 = 13.74 ms.
Pressure scaling. An injector is an orifice, so flow follows the square root of the pressure differential across it, exactly as it does on a flow bench. Raising base pressure from 43.5 to 58 psi multiplies flow by √(58/43.5) = 1.1547, so a 42 lb/hr injector becomes 48.50 lb/hr. This is also why the differential matters rather than the gauge reading: a boost-referenced regulator raises fuel pressure with manifold pressure specifically so that the differential across the injector stays constant and the flow does not fall away under boost.
Worked example: 500 hp on eight 42 lb/hr injectors
A naturally aspirated V8 targeting 500 hp, with eight 42 lb/hr injectors rated at 43.5 psi and running at that pressure, checked at 6,500 rpm with a BSFC of 0.50.
- Fuel demand. 500 × 0.50 = 250 lb/hr.
- Injector capacity. 8 × 42 = 336 lb/hr.
- Duty cycle. 250 ÷ 336 × 100 = 74.40%.
- Injection window. 120,000 ÷ 6,500 = 18.462 ms.
- Pulse width. 0.7440 × 18.462 = 13.736 ms.
- Ceiling at 85% duty. 0.85 × 336 ÷ 0.50 = 285.6 ÷ 0.50 = 571 hp.
So there is 71 hp of headroom on this fuel system. Now boost the same engine to 650 hp and assume a worse BSFC of 0.60 because it is now forced induction: demand becomes 650 × 0.60 = 390 lb/hr, duty becomes 390 ÷ 336 = 116% — static and 54 lb/hr short. The 85% ceiling at that BSFC is 0.85 × 336 ÷ 0.60 = 476 hp, so the injectors were never going to reach the target.
Raise the pressure to 58 psi instead of changing injectors: capacity becomes 8 × 48.50 = 388.0 lb/hr, and duty at 390 lb/hr is 100.5% — still static. Pressure alone cannot fix it, because flow only rises with the square root of pressure: getting the 16% more flow you need takes 1.16² = 1.35 times the pressure. Bigger injectors are the answer.
Reading your duty cycle and choosing a BSFC
Design to 80–85% at the highest power the engine will ever see, not at its rated peak. Fuel pressure sags, pumps get hot, voltage drops, and a hot summer track day makes more demand than the dyno cell did. The margin between your calculated figure and 100% is not spare capacity; it is what absorbs all of that.
Very low duty cycles carry the opposite problem. Below about 15% at full power, the pulse width at idle drops into the region where the injector's opening delay is a large fraction of the total on-time, and fuelling becomes non-linear and hard to calibrate. That is why oversizing injectors by a huge margin makes an engine idle badly, and why staged systems exist.
BSFC is the assumption that moves the answer most, so choose it honestly. Naturally aspirated gasoline engines at full load sit around 0.45–0.50 lb/hp·hr, and boosted engines are conventionally sized at 0.55–0.65 because they are run richer for detonation margin. These are working figures used for sizing, not measurements of your engine.
Alcohol fuels change everything. E85 needs about 1.55 times the fuel mass of gasoline for the same air, because its stoichiometric ratio is around 9.5:1 against gasoline’s 14.7:1, and 14.7 ÷ 9.5 = 1.55 — the air fuel ratio and lambda calculator works that factor out for any blend. Multiply your gasoline BSFC by the same ratio and re-run this page before ordering hardware. Methanol is worse again, at over twice the gasoline fuel mass.
If the duty cycle comes out too high, the fix is bigger injectors. Size them from the power target rather than from a percentage increase with the fuel injector size calculator, and confirm the airflow the engine will actually move with the compressor airflow calculator, since fuel demand only makes sense against the air the engine can pass.
Maximum horsepower at 85% duty cycle by injector size
| Injector (lb/hr) | Approx. cc/min | 4 cylinders | 6 cylinders | 8 cylinders |
|---|---|---|---|---|
| 24 | 252 | 163 | 245 | 326 |
| 30 | 315 | 204 | 306 | 408 |
| 36 | 378 | 245 | 367 | 490 |
| 42 | 441 | 286 | 428 | 571 |
| 55 | 578 | 374 | 561 | 748 |
| 80 | 840 | 544 | 816 | 1,088 |
The cc/min column uses the industry conversion of 10.5 cc/min per lb/hr, which assumes a gasoline specific gravity of 0.72: 453.592 g/lb ÷ (60 min × 0.72) = 10.50.
Duty cycle does not depend on rpm — pulse width does
This trips people up constantly. Duty cycle is a ratio of fuel mass flows, so at a given power and BSFC it is the same at 3,000 rpm as at 7,000 rpm. What changes with rpm is the injection window: 40 ms at 3,000 rpm and 17.1 ms at 7,000 rpm. The pulse width scales with the window, so the same 74% duty is 29.8 ms at 3,000 rpm and 12.8 ms at 7,000 rpm. If a logged pulse width looks alarming, check it against the window at that rpm before concluding anything — and remember that batch-fire systems firing twice per cycle halve both the fuel per event and the window per event, leaving duty cycle unchanged.
Mistakes that hide a fuel-system limit
- Using wheel horsepower. The engine burns fuel to make crank power, and the drivetrain takes its share afterwards. Convert first with the drivetrain loss calculator or you will undersize the fuel system by 10 to 20%.
- Assuming rated flow at your pressure. Published flow is at a stated differential, usually 43.5 psi. Running lower pressure, or losing differential under boost with a non-referenced regulator, reduces flow by the square root of the pressure ratio.
- Sizing on gasoline BSFC and then switching to E85. The alcohol needs about 1.55 times the fuel mass for the same air, which turns a comfortable 75% duty into a static injector.
- Forgetting the pump and the lines. Injectors that can flow 400 lb/hr are useless behind a pump that delivers 250 lb/hr at the pressure you need. Check pump flow at pressure, not free flow.
- Checking duty cycle only at peak power. The highest duty cycle occurs where fuel demand is highest relative to capacity, which is usually at peak torque rather than peak power on a boosted engine. Log duty cycle across the whole pull.
Where duty cycle fits in a fuel system
Duty cycle is the last of four constraints, and a fuel system is only as good as the tightest one. The pump has to deliver the mass flow at the working pressure, the lines and filter have to pass it without dropping pressure, the regulator has to hold the differential across the injectors, and only then does injector capacity matter.
Check them in that order. A pump curve is quoted at a pressure and falls off steeply above it, and a common failure is a pump rated at 340 lph free-flow that delivers far less at 58 psi with a hot tank and 13.2 volts. Line size and filter restriction show up as a pressure drop between the rail and the regulator, which reduces the differential across the injector and therefore its flow.
When you do change injectors, expect to retune rather than to scale. Larger injectors have different opening delays, different non-linear regions at short pulse widths, and different spray characteristics, so the ECU's dead-time and small-pulse compensation tables need updating along with the main fuel table. Simply scaling the fuel map by the ratio of injector sizes gets full load approximately right and idle noticeably wrong.
Finally, remember that fuel demand exists to match air. Work out the airflow the engine will actually move at your target, then the fuel needed to burn it at your target lambda, and only then the hardware. Starting from a horsepower number and a BSFC assumption is a shortcut through that chain, and it is why the BSFC you pick deserves more thought than it usually gets.