Mars Orbiter Mission Update: Supplemental burn planned for today

EDITED Nov 11 3:46pm PST: ISRO now reports the supplemental orbit raise maneuver was successful. Next burn planned for November 16 IST (likely November 15 PST).

Yesterday, the Mars Orbiter Mission was commanded to perform a rocket burn that would have raised the apogee of its orbit around Earth from 71,623 kilometers to about 100,000 kilometers. The spacecraft began its burn, but underperformed, raising the apogee only to 78,276 kilometers. ISRO has planned a supplemental burn for November 12 at 05:00 IST (today at 15:30 PT / 23:30 UT) in order to make up the deficit.

ISRO issued a statement today explaining the anomaly. Here is the full text, with my comments interspersed:

Supplementary Orbit Raising Manoeuvre Planned for Mars Orbiter Spacecraft

In the fourth orbit-raising operation conducted this morning (Nov 11, 2013), the apogee (farthest point to Earth) of Mars Orbiter Spacecraft was raised from 71,623 km to 78,276 km by imparting an incremental velocity of 35 metres/second (as against 130 metres/second originally planned to raise apogee to about 100,000 [1 lakh]* km). The spacecraft is in normal health. A supplementary orbit-raising operation is planned tomorrow (November 12, 2013) at 0500 hrs IST to raise the apogee to nearly 1 lakh km.

"The spacecraft is in normal health" -- it is not in safe mode, a status confirmed today by NDTV reporter Pallava Bagla. The statement went on to talk about the successful testing and operation of major primary and backup spacecraft systems:

During the orbit-raising operations conducted since November 7, 2013, ISRO has been testing and exercising the autonomy functions progressively, that are essential for Trans-Mars Injection (TMI) and Mars Orbit Insertion (MOI).

During the first three orbit-raising operations, the prime and redundant chains of gyros, accelerometers, 22 Newton attitude control thrusters, attitude and orbit control electronics as well as the associated logics for their fault detection isolation, and reconfiguration have been exercised successfully. The prime and redundant star sensors have been functioning satisfactorily. The primary coil of the solenoid flow control valve was used successfully for the first three orbit-raising operations.

Mars Orbiter Mission has a much higher degree of redundancy and autonomy than past Indian spacecraft. When a spacecraft in Earth orbit runs into a snag, it can usually safely quit work and hang around orbiting Earth until engineers can figure out the problem. And it can be more cost-effective to plan and launch multiple missions than it is to build a lot of redundancy into a single mission -- it's like self-insuring rather than paying for the very costly insurance of redundant systems and the excess mass and complexity they bring.

But one-of-a-kind deep-space craft need to be able to respond appropriately to all kinds of problems and still know how to "phone home" for help; and it's so much trouble to get a spacecraft to Mars that your spacecraft is its own spare. The logic and autonomy required for a robot to not shut down in the face of a problem, identify the problem correctly, diagnose malfunctioning systems, shut them down safely, and switch over to appropriate backups, are complex and challenging to engineer. The Mars Orbiter Mission represents the most difficult challenge in spacecraft autonomy that India has faced yet. And this weekend, the artificial intelligence of their spacecraft was put to its first real test:

During the fourth orbit-raising operations held today (November 11, 2013), the redundancies built-in for the propulsion system were exercised, namely, (a) energising the primary and redundant coils of the solenoid flow control valve of 440 Newton Liquid Engine and (b) logic for thrust augmentation by the attitude control thrusters, when needed. However, when both primary and redundant coils were energised together, as one of the planned modes, the flow to the Liquid Engine stopped. The thrust level augmentation logic, as expected, came in and the operation continued using the attitude control thrusters. This sequence resulted in reduction of the incremental velocity.

I'll admit that I am having trouble parsing this paragraph. For instance, it's not clear to me whether both primary and backup valves to their main engine were supposed to be operated together, or if that was a mistake, or if it was a surprise that it didn't work to operate them together. "As one of the planned modes" makes it sound like it was intentional (or, at least, that it should have worked), but evidently it didn't work to operate them together. And why did the flow of fuel to the engine stop? Because the spacecraft detected a problem and stopped it, or because (for some reason) flow didn't proceed with both valves commanded to operate at the same time?

At least, when flow stopped, it sounds like the spacecraft successfully detected that and attempted "thrust augmentation" by using the attitude control thrusters to support the planned orbit maneuver. That is good logic to have in place when you are planning a must-not-fail event like a planetary orbit insertion maneuver. However, the final sentence is concerning, because it sounds like the augmentation with the attitude control thrusters did not accomplish what it should have. Rather than increasing the spacecraft's velocity, "This sequence resulted in reduction of the incremental velocity." That sounds bad for two reasons: it sounds like the autonomous function didn't accomplish what it should have, and it also means that precious fuel was wasted. If the orbit adjustment maneuver had simply been an underburn (the main rocket didn't fire for long enough), then there would have been no cost to fuel reserved. But if the attitude adjustment rockets actually reduced the spacecraft's velocity, that would have cost fuel.

Or maybe the final sentence doesn't mean that at all. Maybe they meant to say "This sequence resulted in reduction of the [planned] incremental velocity," explaining why they only got 35 m/s of increased velocity instead of the planned 130. I'm not sure. So I don't know whether we have two anomalies or just one. Regardless, ISRO is unconcerned enough to proceed with a supplemental burn immediately.

The statement ends:

While this parallel mode of operating the two coils is not possible for subsequent operations, they could be operated independently in sequence.

Again, I am left wondering if it was a surprise to mission operators that it did not work to operate both the primary and backup valves to the main rocket engines at the same time. Looking to media reports for quotes from officials makes it seem as though it was a surprise, but one that they say isn't a problem because they won't need to do it in the future. (If that's true, then why test it?) Here are a few statements:

• In an NDTV story, Pallava Bagla reports: "ISRO Chairman K Radhakrishnan told NDTV today, 'The space craft is healthy and it encountered a problem when a specific redundancy test was being conducted and it failed to reach the desired velocity it was to achieve.' A failure analysis committee will examine why this problem happened, he said, but added that crucially, not much fuel was wasted in the failed attempt.
• From the BBC, a story based on conversation with reporter Pallava Bagla: "In that redundancy test, two coils in the liquid engine were supposed to be energised simultaneously....Mr Bagla told BBC News that the attempt on Monday morning used up about 2kg of the craft's 852kg fuel load. But he added that the spacecraft's insertion into Earth orbit after launch on 5 November had been so precise, 6kg of liquid fuel had been saved. Even with Monday's glitch, the mission still had a fuel surplus of 4kg."
• An unnamed official spoke with Wall Street Journal blogger Joanna Sugden: "The fault was discovered when two coils in the satellite, which are responsible for speeding the satellite up and slowing it down, were turned on at the same time. This led to a blockage of fuel and oxidants to the main liquid-rocket engine. 'The satellite’s engine doesn’t work when both coils are simultaneously on,' a spokesman for the ISRO told The Wall Street Journal. 'This is not at all a setback, we got our redundancies [backup plans] checked by this process,' said the spokesman, who declined to be named. He said it was not necessary for both coils to be on at the same time during the rest of the flight plan."

According to the BBC story, assuming today's supplemental burn goes well, the final apogee raise maneuver -- which will take them to about 192,000 kilometers -- is to take place on November 16; the spacecraft will be sent to Mars on December 1.

*"Lakh" is a term used in India, even when communicating in English, to mean 100,000 (which is written "1,00,000"), while "crore" means 10,000,000 (written "1,00,00,000"). They don't typically use "million" or "billion" in India.