Gas turbines typically bleed air from the late compressor stages. Some of this air bypasses the combustion chamber and is expanded through the turbine to prevent the blades from becoming too hot. I believe bleed air is also used to provide hydraulic control of the control systems and also for utilities such as air conditioning and electricity generation.

I am interested to know just how much of the air intake to the gas turbine compressor becomes bleed air. I'm also interested in knowing how much of this bleed air goes to each application e.g. Utilities vs. turbine cooling vs. hydraulic control.

I suspect it varies a fair bit so I am mostly interested in how this occurs on larger airliners that would typically fly long international flights.


Indeed about 1-2% of compressed air is bled off for cooling turbine disks and blade roots. When turbine inlet temperatures are higher than 1350K, the turbine blades must be internally cooled as well, and up to 15% of compressor air may be required as bleed air for cooling. Source

enter image description here

Bleed air is also used for aircraft systems. If we take a typical airliner such as the B737NG, compressed air can be supplied by either the engines, the APU or an external air chart, and is used for the following purposes:

  • Air conditioning & cabin pressurisation, which are both subject to regulation quality standards.
  • Wing & engine thermal anti-icing
  • Engine starting
  • Hydraulic reservoir pressurisation. The reservoirs are pressurised to ensure that hydraulic fuel reaches the hydraulic pumps.
  • Water tank pressurisation

Compressed air is tapped from the engine 5th and/or 9th compressor stage. Air supply is regulated according to demand and varies according to flight stage - during flight stages with relatively high engine power (take-off, climb, most cruise), 5th stage compressed air only is used. During low engine power conditions such as approach and landing, 9th stage compressed air is used for the pneumatic systems as well. This air is under higher pressure, and hotter, than the 5th stage air.

enter image description here

Exactly how much bleed air is used for the systems is hard to get a reference for, but here is an estimate for the CFM56-7B18:

  • Bypass ratio = 5.5:1, Air mass flow = 307 kg/s => hot flow = 307 / 5.5 = 55.8 kg/s
  • Cabin air systems refresh the whole cabin volume every 2-3 mins. Let's take 2 minutes, without recirculation.
  • Internal volume of a 737NG is about 400 $m^3$ (from fuselage width and length of an -800).
  • Air density = 0.94 kg/$m^3$ (we'll take cruise conditions, cabin air density @ 2,000 m)
  • Aircon flow = 400 * 0.94 / 120 = 3.1 kg/s

There are two engines, so the percentage of flow used for the aircon and pressurisation would be 3.1 / (2 * 55.8) = 2.8% maximum. Please again note that this is a ROM value which I will gladly give up for an official reference.

As a magnitude check: the APU brochure mentions 154 pounds of air per minute = 1.16 kg/s for environmental control and engine start. The APU can deliver air conditioning on the ground without the engines running so that would be the right order of magnitude.

Final estimate: bleed air for air conditioning is between 1 - 3 kg/s is about 1 - 3% per engine.

Bleed air is not used for:

  • Hydraulic power generation. The hydraulic pumps are engine driven pumps (via a gearbox) and electrical pumps, powered by the AC system
  • Electrical power generation. This is provided by Integrated Drive Generators, from a generator in the APU, and from the batteries through an inverter.

Source: own work from open source on-line pilot training courses.

| improve this answer | |
  • $\begingroup$ You didn't say how much bleed air for a/c, etc $\endgroup$ – TomMcW Aug 16 '17 at 18:49
  • $\begingroup$ @TomMcW You're right, I made an estimate. $\endgroup$ – Koyovis Aug 17 '17 at 13:59
  • $\begingroup$ Cool. +1 I've always wondered this myself. I knew when they needed some extra thrust the pilots could cut bleed air, but I never knew how much they could squeeze out $\endgroup$ – TomMcW Aug 17 '17 at 14:14
  • $\begingroup$ I like the approach for a ROM, however, isn't there a large degree of recirculation of the cabin air? aerospace.pall.com/en/commercial-fixed-wing/… says 50%. Does that needs to be allowed for in the calc? $\endgroup$ – Penguin Aug 18 '17 at 12:56
  • $\begingroup$ @Richard yes there is, there is quite some variation with number of passengers, set temperature, recirculation etc. My ROM was for a maximum demand scenario, it seems to be on the high side judging from the APU capacity. $\endgroup$ – Koyovis Aug 18 '17 at 13:44

Internal cooling air might be better termed bypass air as it is not actually bled out of the engine so the energy of compressing this bypass air remains within the overall cycle rather than draining substantial energy from the system, and because turbine combustion is normally oxygen rich this bypassed air does not reduce the total fuel that can be burned.

The pilots can shutoff all accessory bleed air for improved engine performance on takeoff and low altitude climbs. So 0% is the lower limit for bleed air.

Accessory bleed air is used mainly for cabin pressure, engine starting, and deicing. Pressurizing reservoirs is an extremely minor use and only needed at higher altitudes to reduce the chance of pump cavitation and foaming from dissolved gasses.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.