Engineers pumped a half-million gallons of supercold rocket fuel into the shuttle Discovery's external tank today, repeating a test that uncovered two unexpected problems in April. This time around, suspect fuel depletion sensors in the tank worked normally while a hydrogen pressure relief valve continued to cycle more often than normal.

Bill Parsons, shuttle program manager at the Johnson Space Center in Houston, said it will take engineers additional time to fully analyze the data. But it would appear whatever caused the fuel depletion sensor problem in April was corrected during earlier troubleshooting.

As for the hydrogen vent valve issue, today's test showed no difference in operation regardless of whether a new heater, installed as part of NASA's post-Columbia safety upgrade program, was turned on or off. That would appear to exonerate the heater.

Parsons said he was optimistic a thorough review of data from today's test will allow engineers to pin down the cause of the problem and clear the way for a launch attempt July 13 as currently planned.

"We had just a perfect test as far as I'm concerned," he said.

The three-hour fueling process began at 5:36 a.m. and was complete by around 8:30 a.m. At that point, just like in a real countdown, a small team of engineers headed to the launch pad for a detailed external ice inspection.

During the first tanking test April 14, humidity was relatively low and ice formation was minimal. Engineers were hopeful today's test would provide additional insights into where ice might build up on the tank. Contrary to initial predictions, however, the humidity was not much higher - 67 percent - and with a breeze to keep condensation down, ice and frost formation were only slightly higher.

Today's test was run in two stages. The first ended around 1:40 p.m. with a countdown cutoff at the T-minus 31-second mark. For that run, the bipod heaters at the bases of the two struts connecting the nose of the orbiter to the external tank were turned on to prevent ice buildups.

The shuttle Columbia's destruction was triggered by the loss of foam insulation around the left-side strut. In the wake of the accident, the foam in question was removed and the bipod heaters were installed to do the same thing: Prevent ice formation.

For the second part of today's test run, the countdown was recycled to the T-minus 20-minute mark and the heaters were turned off. The countdown resumed and continued to the T-minus 31-second mark a second time to find out if the bipod heaters played any role in how often the hydrogen pressure relief valve operated.

The pressure relief valve works during the final two minutes of the countdown when gas is being pumped into the tank to maintain the proper pressure levels for flight.

During the April 14 tanking test, the pressure relief valve cycled 13 times compared to a normal cycle rate of eight or nine. During today's test, with the bipod heaters were powered on, the valve cycled 13 times just like it did in April. During the second part of the test, when the bipod heater was off, it cycled 13 times once again.

Special instrumentation installed for today's test showed the valve was not leaking, focusing attention on a screen-like "diffuser" that engineers already planned to replace. The diffuser disperses the gas used to pressurize the tank just before launch. If that system was damaged or clogged, it could affect the operation of the vent valve.

Another problem seen in April involved two of four hydrogen depletion sensors in the tank that failed to operate properly. The sensors are used to make sure the shuttle's main engines shut down properly after reaching space and all four must be operational for a countdown to proceed.

During today's test, all four sensors operated normally. Engineers suspect cable harness replacements and other troubleshooting in the orbiter's engine compartment corrected whatever the problem was in April. But that remains to be confirmed.

And in any case, it's likely a moot point for Discovery. NASA managers decided earlier to move the shuttle to a different tank and boosters for its launch attempt in mid July. The new tank will feature an additional heater to prevent ice buildups around a flexible propellant line bellows, which prompted concern during Discovery's initial launch pad processing flow.

In addition, the new tank's engine cutoff sensors presumably will work properly and the tank features a diffuser with a known track record. But Parsons said engineers might ask for a third tanking test before Discovery's eventual launch just to make sure. As it now stands, he said, a third tanking test is not planned.

The shuttle will be hauled back to the Vehicle Assembly Building next Tuesday. If all goes well, Discovery will be moved back to pad 39B around June 14 for a launch attempt on July 13. If an additional tanking test is ordered, the flight could slip another week or so.