The Endeavour delay: Complexity, the APU, and the Load Control Assembly

The last launch of the Endeavour space shuttle has been delayed 48 hours (update: indefinitely) due to a problem with the APU heater and the Load Control Assembly. I wanted to find out what exactly these troublesome components are, so I did some investigation. There's a lot of extremely detailed information on the Space Shuttle available online, but it is very hard to find. I've summarized the information here in case anyone else wants to know the specifics.

Space Shuttle APU locations

The Space Shuttle has three independent hydraulic systems to operate engine valves, actuators, landing gear, and so forth during launch and landing. The hydraulic pumps are powered by three Auxiliary Power Units (or APUs), which are hydrazine-powered turbines. Each APU is 88 pounds and produces 135 horsepower (which is about the same horsepower as a Honda Accord).

Hydrazine is a highly-toxic rocket fuel; when exposed to a catalyst, it energetically decomposes into hot gases at 1700°F. It is convenient for applications such as the APU, since it doesn't need oxygen, and the decomposition can be easily started and stopped.

Space Shuttle Auxiliary Power Unit
(Click on the image for tons of detailed information.)

Apparently the fuel heaters in APU 1 are not working. Since the hydrazine fuel will freeze at 34°F, each APU has redundant heaters to keep the system above 45°F. Since the heaters are redundant, the Space Shuttle would still be able to operate with the current problem, but would not be able to handle another failure. If the second fuel heater failed, then the fuel would freeze and the APU would not be able to work. Since there are three APUs, even this failure would not be a major problem. But still, you wouldn't want to take off with the heater not working, because losing hydraulic pressure would be a very bad thing.

According to articles, the fuel heater problem is due to a lack of power from the Aft Cabin Load Control Assembly, a switchbox that powers a heater circuit for Auxiliary Power Unit 1. There are three Aft Load Controller Assemblies, as well as many other Load Controller Assemblies. (Sources are inconsistent about whether it is called a Control vs Controller.)

A complex Electrical Power System provides power to all parts of the Shuttle. Three fuel cells (10 kW each) generate 28-volt direct current. The fuel cells feed three main DC power buses, as well as powering AC inverters to feed three AC buses with three-phase, 117-volt, 400-hertz AC power. From the fuel cell, power goes to a Distribution Assembly (DA), then to the aft Power Controller Assemblies, and then to the Aft Load Controller Assemblies.

The Load Control Assemblies contain solid-state switching devices for loads up to 20 amps, and relays for loads up to 135 amps. These switching devices are internally fused.

Reportedly there is a short or other electrical fault in the Aft Load Controller Assembly 2, which is causing the APU heater to fail to operate. The fuel is being drained from Endeavour so technicians can access the assembly and resolve the problem. If I'm interpreting everything correctly, it seems like they'll need to replace one of the internal fuses in the Load Control Assembly.

Space Shuttle Power Distribution

Complexity and the Space Shuttle

One amazing thing about the Space Shuttle is the layers and layers of complexity. The APU system is just one example of this. For instance, each APU has as lube oil system to keep it lubricated. This requires a lube oil pump, which requires a nitrogen pressurization system to start the pump in zero gravity. The oil also requires a 181-pound water spray boiler system, which sprays cooling water onto the oil pipes; the water boils into steam and is vented into space. The boiler requires controllers, panel switches, and status displays, as well as yet another nitrogen pressurization tank, and yet another system of heaters to keep the water from freezing.

Space Shuttle Water Spray Boiler

The water spray boiler doesn't have anything to do with the current launch delay, other than being part of the APU, but it provides an interesting example of the complexity of the systems involved. To summarize the complexity along just this one path, the engines require hydraulic pressure, which requires APUs to power the hydraulic pumps, which require a lubricating oil system, which requires a complex boiler system, which requires a own control and monitoring system. And this is just one small sub-path! I'm ignoring equally complex systems such as the APU injector cooling system (more water and pressurized nitrogen), or the APU fuel pump (which for instance, has a catch bottle in case its seals leak, a drain port if the catch bottle overflows, and associated monitoring system).


Given the level of complexity of the Space Shuttle, I'm not surprised by the launch delays, and wish NASA the best of luck in resolving the problem promptly. My opinion is while the Space Shuttle is a marvel of engineering, simpler rocket systems such as the SpaceX Falcon will turn out to be more practical in the long run.

The images and much of the information above is from the 1988 Shuttle Reference Manual at This manual goes into extreme detail and is very interesting (if you find this sort of thing interesting). I should probably make it clear that this posting is based on what I've read; I have no connection with the space program.

P.S. I've found extensive details on the LCA and launch issues are available at, e.g. Endeavour receives her new LCA – Blown driver examined.


Lefty McStinky said...

The whole system was unnecessarily complex. You only need hydraulic power possibly on launch, and definitely on reentry to activate the elevons, the rudder (which IIRC also splits into a speed brake) and to deploy the landing gear. You do not need to retract the gear, since you only get one shot at landing.

A simple one shot pneumatic system or a Coffman starter like gunpowder charge could have done that. The control surfaces could have been powered by one APU with a wind turbine as a backup. If that fails you ride it down to fifty thousand feet and bail out.

However, redundant heaters on the single hydrazine tank would have been a Good Idea.

It's otiose now. Like the SR, Concorde, and a dozen other superb extensions of current tech the Shuttle is done.

Lauren Adams said...
This comment has been removed by the author.
burritolikethesun said...

The solutions offered by Lefty are not really viable. You can't stick a wind turbine out into the air stream at Mach 20. If the hydraulic pressure fails at launch you lose total control of the SSME engine gimbals, which would be an insane situation that would mean almost guaranteed death of the crew and possibly dangerous to those below.

The APU/hydraulic system is one of the most criticised areas of STS by both Shuttle designers and operations people--including Wayne Hale, Chris Kraft, etc. Watch the MIT lecture series 16.885j for more info. It's pretty unanimous they should have gone with an electromechanical or electrohydraulic system, but it was a timing issue--during Apollo they did not have very reliable or robust fuel cells, but by the mid 70s they had ones that I've heard one Shuttle designer describe as "so awesome you could arc weld with them." However, by this time the contractor had already hired people to design a hyrdrazine/hydraulic system, so that's what got built. Kind of a shame since it probably was one of the main obstacles to cheaper and more rapid turnaround.