NASA plan for pre-launch tank trouble

by Chris Bergin

With most of the media attention focusing on how the External Tank performs during ascent on STS-121, NASA is putting into place plans to counter two major problems that occurred ahead of last year’s STS-114.

A series of documents show NASA is concerned about the LH2 vent value registering too many cycles during pre-press – as well as the potential for a leak, while an ECO (Engine Cut Off) troubleshooting plan has been put in place.

L2, for expanded documents and insider information used in our articles. To join L2, click the advert.

The following is collated from four NASA documents – published in full on L2.


While the ECO issue has been well documented, with a solution likely found by the replacement of the sensors in Discovery’s tank (ET-119), the other issue – which nearly caused a late scrub during STS-114’s countdown – is less publicised. Thursday’s PRCB meeting discussed the troubleshooting plan (with an article to follow on their findings).

The Marshall Space Flight Center (MSFC) designed duplex screen in the LH2 diffuser was blamed as the cause for nearly double the normal amount of cycles through the LH2 pressurization relief valve (hydrogen vent valve), breaching Launch Commit Criteria (LCC) – on both tanking tests, involving ET-120. The new design diffuser will never be used again.

‘The diffusers will only use the single-shute-wire weave screen – the duplex screen will never be used again,’ said a Michoud Assembly Facility (MAF) source. ‘The problem with (the duplex screen) is that it almost over-diffused the GHe (Helium) entering the tank.’

However, following the rollback and replacement of ET-120 with ET-121 – due to the ECO troubles with the former, again the cycle count was high, so much so sources insist they were one more cycle away from scrubbing the launch (now confirmed in documentation) – which eventually proceeded as planned. ET-121 used the old style diffuser.

The diffuser, which is used to inject helium into the tank, a process that aids the correct pressure and temperature of the ET prior to launch, needs to be working within parameters for a number of reasons. Notably with STS-121, the effects on the tank’s foam will be crucial, especially if a scrub forces another ‘cryo cycle’ – which would not aid the prospects of reducing foam liberation on ascent due to the stresses involved in filling the tank full of supercold propellant.

MSFC, unhappy that their new diffuser was being blamed, asked for another tanking test with ET-121. That was rejected, despite an MSFC protest that there was something seriously wrong with the count – which was noted again after the launch of Discovery last July, with a total of 11 cycles observed, while the LCC is set at 12 cycles. Breaching LCC equates to a scrub being called.

Another problem NASA is concerned about is a leak from the tank – relating to the prepress. NASA deem small leaks in this scenario as ‘acceptable risk’ for the shuttle and crew, given the lack of an ‘ignition source.’ This is because the leak would be at the opposite end of the shuttle’s engines (the major ignition source) – with only static electricity and lightning a potential danger.

STS-121 Mission T-Shirt. Use discount code ‘nasaspaceflight’ when ordering from Countdown Creations & receive 5% off!

However, how to treat an observed leak is a topic that has varying opinions within NASA ranks.

‘STS-114’s higher than expected LH2 prepress cycle count exposed sensitivity to ground prepress system performance (LCC ET-04),’ noted a NASA document presented to last week’s PRCB. ‘Leak detection and preventing an inadvertent scrub are in conflict – protection needs to be balanced.’

For STS-121, 11 cycles won’t cause a scrub, with the note that: ‘for STS-121,10.5 cycles are expected (nominal – no variation), but variance stack-up could result in 13.3 cycles (LCC violation occurs at 14 cycles),’ which shows NASA has moved the allowable amount of cycles before a scrub to be 14, as opposed to 11.

The calculation NASA has to balance is how many cycles are needed, how much pressure is going into the tank and if there’s a leak affecting both readings.

‘This is why the cycle count is so crucial, because originally the LCC was developed to protect the tank against leakage,’ added the MAF source. ‘With the over-diffusion of helium (on ET-120 and ET-121), we weren’t getting the pressure increase from each 0.5-sec burst, so extra cycles were feed into the tank.’

Those ‘bursts’ or ‘pulses’ are also the subject of a difference of opinion, based on ensuring there is enough – and not too much – pressure going into the tank.

‘SICB/PSIG recommendation – accept current valve timing (0.476 seconds) with decreased LCC margin. Dissenting opinion – return valve timing to MLP-1&3 historical performance (0.484 to 0.539 seconds).’

As far as what is deemed as a leak, the range between safe and unsafe is classed: ‘GH2 leak of 0.005 lbm/sec is flammable, but will not cause structural damage in flight. GH2 leak of 0.065 lbm/sec is flammable and can cause structural damage in flight. The transition point from safe to unsafe is unknown.’

‘Predicted pulse count for STS-121 is 10.5. If LCC limit is approached, but not broken, T-0 will take place with a leak (or performance dispersion equal to a leak) of roughly 888,000 SCIM (or 3,500 times specification).

‘One pulse is equal to roughly 250,000 SCIM (or 1,000 times specification). Unable to determine if a leak detected on the ground will remain constant in flight.’

While most opinions on the risks associated with a leak remain within ‘acceptable risk (catastrophic/remote)’ – NASA Propulsion Systems Engineering & Integration Office disagree.

‘Presidential Commission Report identified leakage concern, they noted in their presentation, ‘GH2 leakage greater than 0.005 lb/s has potential for ignition. Hazard – Catastrophic failure due to structural overheating as a result of undetected GH2 leakage in the presence of an ignition source.’

PSE&I recommend that ‘Valve timing (should) be retested ~7 days prior to launch.’ It is not clear if this was rejected, due to the ‘dissenting’ opinion of USA, ET Project and SCIM.

The real test will come in the actual countdown for the July 1 launch of Discovery, with a possible scrub due to any problem coming before T-43 seconds, from which point any error is – according to the documents – undetectable via the current procedures in at the Kennedy Space Center’s Launch Control Center.

**NASASpaceflight.com Job Opportunities** 


              LIVE UPDATE PAGES

Related Articles