Recover the network after a failure

In January 1994, Anik E1 and Anik E2 suffered control-system failures that interrupted communications services. NASA’s later case history attributes damage to spacecraft charging and electrical discharge. The case connects the space environment with effects experienced by broadcasters, data users and telephone customers.

The report describes E1 recovery using a backup system, while E2’s backup did not restore its operation. It also records rerouting services, adjusting ground antennas and using other available links. Recovering some customer traffic and recovering the original spacecraft were different tasks.

A fleet can contain more than one satellite and still share exposure to an environmental disturbance. Redundancy should therefore be evaluated by the failures it can actually tolerate. Counting duplicate boxes does not by itself prove that the backups remain usable under the same event.

NASA reports that engineers eventually restored E2 service using a ground-controlled approach that fired its thrusters to maintain orientation. This brought useful operation back, but increased fuel use. The case does not describe physically replacing the failed electronics in orbit.

For a fictional comparison, suppose a remaining fuel budget is 120 units. A simplified mode consuming 2 units per month would last 60 months if nothing else used fuel. A recovery mode consuming 3 units per month would last 40 months under the same assumption. These are invented values, not estimates of Anik E2’s propellant or lifetime.

The calculation illustrates why a recovered capability can still carry a cost. Real fuel budgeting includes several maneuvers and uncertain demands; a straight division is a teaching model. Engineering decisions need the actual operating requirements and remaining reserves.

What has recovered after an outage?

Imagine an outage board with three entries: original spacecraft control restored, alternate customer route working, full planned capacity available. An alternate route might restore essential traffic while the original satellite remains unavailable and some capacity is still missing.

Each entry needs its own test. A successful spacecraft command does not prove that a subscriber can exchange data. A successful subscriber test through another route does not prove the original spacecraft has recovered. Recording the route used makes these conclusions distinguishable.

Anik’s history thus offers more than a succession of launches. It shows how orbital design, control, ground infrastructure and operating choices combine into a service. Evaluating the whole chain makes the achievements—and the consequences of interruptions—more understandable.