Iinternational Space Station Chronicles

The ISS Chronicles is a riveting account of my communications with NASA concerning safety issues pertaining to the International Space Station following the Columbia accident. I have elected at this time to restrict complete elaboration on The ISS Chronicles according to a "need to know" basis. What follows are some examples.

I had considerable communications with Readdy about an air leak problem and probability of failure during early January, 2004. The following reflects my best available advice although the problem source was fortunately identified very soon after the astronaut realized that something didn't logically seem correct in the trouble shooting procedure and revisited for closer examination.

I've enclosed just a clip of one of our daily summaries 1/9 where we are in troubleshooting so you can appreciate the level of detail and complexity in tracking down something so minute -- so minute, in fact that fluctuations from thermal from our attitude changes could have accounted for most of retrospective trending. I appreciate your willingness to engage, but I believe our folks and the Russians are capable of dealing with the situation.

*******************************************************************************************

OK. I hope everything works out fine. It seems from reports that the daily leak rate is constant. I think there is a probability this leak could increase exponentially and then there could be a crisis. Be prepared.

*******************************************************************************************

Thanks for your cautionary commentary. Just curious -- what probability would you estimate for exponential leak increase, depressurization and crisis?

With respect to preparation, the crew is very prepared to deal with a variety of emergencies, including rapid station depress should it ever happen -- heaven forbid. They and the ground have periodically practiced those drills.

When averaged over a number of days you derive a rate and the appearance is of a steady 'leak'. Reality is the pressure has shown decreases and periods of stabilization. There are many modules and hatches an achieving a homogeneous gas O2 and N2 composition, temperature and humidity onboard it a very tough problem.

Remember the ISS has a habitable volume about twice that of Mir and about equivalent to a jumbo jet. We're also dealing with extreme thermal changes each orbital day/night cycle as well as the lack of convective circulation within the pressurized volume because of microgravity.

To give you a little perspective, nominal ISS cabin pressure is 760-735 mmHg (760 mmHg is sea level.) Current pressure is 731 mmHg, which is equivalent to an ISS cabin altitude of about 1000 ft. An airliner cabin altitude is typically from 5000 to 8000 ft.

The cabin pressure change is about 50 ft per day compared to an airliner which is about 500 ft per minute. We are methodically going through the list of possible suspects and are continuing a rigorous troubleshooting plan. We'll find out what's up.

Troubleshooting includes correlation of past events like attitude changes, thermal environment, O2 candles burned, etc. as well as a methodical elimination of suspect valves and components and isolation of larger elements such as Progress, Russian Docking Compartment, Soyuz, etc.

*******************************************************************************************

By definition, probability ranges from zero to the maximum value of one. I remember working for Union Carbide Linde Division decades ago as a young COOP student. I noticed an unusually strong odor somewhere at the plant so I made the comment during the lunch break that I thought it was chlorine gas. My supervisor jokingly remarked since no other person noticed the smell that if it were chlorine gas I'd probably know it for sure due to the nature of the hazard and others in the break room started to laugh at me. Well, considering all conditions, nobody else noticed the smell probably because it was a very windy day. That day the weather was so bad that a large tornado killed several people just a mile away while a passenger jet crashed during flight from Huntsville to Atlanta due to the same weather system. The weather had improved the next day and a fellow worker began to think about what was said in that break room. He revisited a chlorine gas cylinder he had prepared and noticed a leak at the needle valve. Within minutes, the plant was declaring an emergency and a special hazard procedure were implemented to mitigate the problem. Nobody was laughing now. Afterwards, fellow employees would sometimes ask me whether I smelled anything wrong just prior to their entrance into the gas cylinder areas. The point of this experience is that probability can be conditional and the true severity of a problem can be obscured by other events. Given that various activities aboard the ISS could actually mask accurate measurements of leak rate, the conditional probability of an exponential leak rate is greater although limited to the maximum value of one. Furthermore, you really don't know whether the total leak rate is not the cumulative effect of multiple leak sources so this factor influences probability calculation.

As I mentioned, I don't have access to the necessary data and other resources to provide a qualified, specific probability number. I might be able to predict something if given all the daily readings and perform an analysis of the trend. In general, an exponential leak rate occurs when the flow cross-sectional area increases given the same pressure differential conditions. Potential causes include secondary degradation of material due to stress, corrosion, etc. For example, a spring in a valve controlling closure or throttling could initially fail to the extent that initially it is dislodged into a position causing a relatively constant flow rate.However, the dislodged position could now encounter new stresses or environmental conditions that accelerate additional failure, in particularif parts now experience abnormal motion characteristics that could lead to a fatigue failure.

Has there ever been a fire aboard the ISS? Probability calculation can be significantly impacted if a fire caused any damage to components.However, the following is a guess for the probability of exponential leak rate given existing condition of a constant leak rate:

Let's assume this conditional probability of an exponential leak rate is the same as failure of a mechanical seal. Then the probability of an exponential leak rate is 0.85 based on the following reference:

Data from ultrasonic leak testing seems non-applicable to this probability assessment since the measurement essentially does not quantitatively identify the degree of degradation as a function of time ( i.e. Either it has failed or it has not failed ) based on the following reference:

Other types of failures, such as relief valves, impose other probability problems. I don't have any data, but perhaps you can talk the boys and girls at Stennis into providing the records of a problem regarding relief valves for an oxygen system based on the following reference:

*******************************************************************************************

The following was extracted from a news report after the source of the leak was found:

"Sasha [Kaleri] and I wonder why we didn’t come to this conclusion a day or two earlier," he told Houston. He explained that they had checked the window a few days earlier, but that background noise from an operating science instrument had overwhelmed the signal then. Kaleri had still been suspicious about the window and had tried various techniques to isolate the noise, but in the end both men had concluded at that time that what they were hearing was from the science instrument.

The following was extracted from Expedition 8 Crew STATUS REPORT: ISS04-3 2 p.m. CST, Monday, Jan. 12, 2004 after the source of the leak was found:

Foale said he couldn’t hear any hissing noise from the flex hose during a previous leak check last week because of other ambient noise generated by operating payload racks in Destiny. Sunday, those racks were shut down for a short time and the hissing noise was obvious. Foale reported that as soon as the flex hose was disconnected, the noise stopped. While additional evaluation is needed for confirmation, the pressure in the Station appears to have stabilized since the removal of the flex hose.

*******************************************************************************************

Well now, it doesn't take a rocket scientist to see the comparative similarity in my COOP experience and the background noise. Furthermore, a popular answer or a majority answer should never be assumed as the correct answer.

*******************************************************************************************

The Destiny Lab window has been a serious concern due to leaks. It is possible the seals of the Primary and Redundant panes enclosing the Volume D space are experiencing hidden degradation with potential loss of pressure boundary integrity. During early 2004, this problem was exacerbated by a broken flex hose connected to Volume D.

Astronaut Fincke performed a leak check using an ultrasonic instrument on July 21, 2004 and could not detect a problem. Although this would seem good news, the instrument was limited by the background noise. Therefore, the rate of window seal degradation could be deceptive without better diagnostic devices and procedures.

I submitted a concept about a week later to Howell, JSC Director. This concept was forwarded to his engineering director for further consideration as follows:

How is leakage to outboard side of Destiny Lab Window detected and/or measured?

Have your boys and girls in the design department used any imagination like the following concept?

See conceptual sketch leakdet.gif.

Leak sensing apparatus would be comprised of a metallic ring mount fitted similarly to outboard protective cover. Attached to metallic ring mount would be one or more oxygen sensors and with corresponding oxygen data processor and transmitter. Wireless signals would be transmitted locally to an inboard receiver and display to monitor leakage. As an option, the receiver could be designed to retransmit data to Mission Control Center for further evaluation or annunciate locally if measurements become off-scale or indicative of a possibly dangerous leak rate. I think that the size of sensors, processor/transmitter and receiver/display could be relatively small. If for no other reason, this concept could be considered as an experiment to determine feasibility of external leak detection for future designs such as a mission to Mars.