Monday, September 13, 2010

When Catastrophic Failure Hits Home


I've wanted, and circumstances have kind of forced me, to take a break from the blog and contemplate on something that happened in my community over this past week. As many of you will see when you look at my profile, I live on the San Francisco Peninsula. More specifically, I live in the small town of San Bruno, CA. San Bruno became a headline news story last Friday and over the weekend all over the country. This was due to a natural gas line that exploded in a residential neighborhood. The explosion caused the confirmed fatalities for four people, destroyed several dozen homes, displaced hundreds of families, and required the concerted efforts of many cities fire crews and emergency personnel to maintain order and prevent the fire from spreading into other neighborhoods.

At 6:15 PM on Thursday evening, September 9, 2010, reports came in of a loud explosion in San Bruno near Crestmoor Canyon. Due to the proximity of San Bruno to the San Francisco International Airport, many thought that a jet airliner crashed in the canyon, and that the fire we were seeing was from such a crash. Shortly afterwards, we heard a report that all planes were accounted for, and that the odds of it being an airplane were non-existant. Then what? What was causing a massive fireball to shoot up in the air to heights of 100 feet and higher?

We didn't get to have the time to speculate on exactly what was happening, because our first motivation was to pack up our things and get out of the area as quickly as possible. Actually, this was more my wife and kids decision, because I had not returned home yet. I first learned of this situation when I got off of the Bay Area Rapid Transit (BART) train that I take every day to come home from work. As I drove up the street heading for my home, I saw the smoke and I saw the fireball, and it looked from my vantage point to be in the canyon. This scared me in a big way; Crestmoor Canyon is filled with Eucalyptus trees. Eucalyptus trees were what fueled the Berkeley Hills fire of 1991, in which 3000 homes and other buildings were destroyed. If Crestmoor Canyon were indeed on fire, and those Eucalyptus trees were to catch, then a large part of western San Bruno had the potential to go up in flames.

As I fought my way home through traffic and a police barricade and being told I could not drive into my neighborhood because an evacuation was in effect, I parked my car on the onramp to Southbound 280 and made my way by foot up to my wife's parent's place, where she and the kids were staying. From there, I was able to get the rest of the story. The fire was not in the canyon, but was in the neighborhood just to the west of the canyon, often referred to as "Crestmoor 2" by real estate agents, or Glen View/Claremont by two of the main streets. The cause was determined to be a 30 inch gas line that had ruptured, and it was the high pressure in this gas line that made the explosion so severe that it created a crater 40 feet across and 40 feet deep, while registering a 1.3 quake on the Richter scale as reported by the US Geological Service. The explosion took place very near the intersection of Glen View and Claremont, which is a low lying section of road locals call "the gulley" and where storm and water runoff heads into Crestmoor Canyon. The fireball that resulted from this explosion quickly incinerated the houses immediately surrounding it, and then the fire spread to other houses. In addition, the blast also damaged a water main, which meant there was no immediate water they could use to fight the fire; they had to either fill trucks some distance away or run hoses from hydrants more than 1/4 mile away.

Firefighters from several nearby cities were called in to help fight the blaze, work with the city and county to shut down the gas line that was feeding the fireball (and the fact that it had to be shut down slowly so as to not cause ruptures further down the pipe) plus a committed team of firefighters that were told, in no uncertain terms, "do whatever you have to do to make sure that fire does not hit the canyon!" They knew that, if the fire did spread to the canyon, many other neighborhoods would be in danger.

My family spent the night on the floor huddled around a battery powered radio (since the power was out for the greater western part of town), listening to the reports. With the smoke and the confusion, we were hearing estimates of 50 + houses destroyed and 120+ houses damaged. Knowing how many of our friends lived in that neighborhood, we knew this was not a good sign. With the power being out, we also had a hard time communicating with our friends and family to let them know we were safe. What I later found to be cool was the fact that a friend we did contact made a point to post to both my and my wife's Facebook pages that we were safe, had been evacuated, and that we were dealing with no power and an overloaded cell phone network and otherwise couldn't call or email to let people know we were OK. We later commented that this was the first "Facebook" enabled disaster, and that Facebook played a major role in helping get the word out to people about the event (expect me to do a post about that another time).

As the evening wore on, I decided that I needed to get my car off the freeway, and so I drove and found a place to park. As I was coming back up to where my wife's parents lived, I was stopped by the police and told I couldn't continue. I pulled out my wallet and said "look, I live here; my wife and children are just up the street. You may escort me up if you wish, you may arrest me if you wish, but one way or another, I'm getting back to my wife and kids!" With that, they nodded and let me pass and get back to my in-law's home.

Friday morning, we received the news that the evacuation order for our neighborhood had been removed, and that we could go back to our homes. We were overjoyed to find out that the fire had been contained to the Claremont area, that the fire did not make it down into the canyon and that our house was unaffected. With this brief joy, however, came the stark realization that many of our neighbors had lost everything. With power restored, we could see on the television and the Internet the pictures and video of the damage. Several dozen houses were gone. Not just burned down, but vaporized! Nothing left but chimneys and foundations. To this we received word that several people lost their lives, one of the fatalities being the older sister of a boy in the Cub Scout Pack that I used to lead, and that hundreds of families were displaced, for how long was anyone's guess.

As a software tester, I can't help but look at this in the lens of what I do for a living. I heard reports from people that said that due to a "glitch", the pipe ruptured. When a pipe ruptures and close to 50 houses are destroyed, and four people are confirmed dead, that is not a "glitch", that is a catastrophic failure! Many things have come to light from this incident that I had not known about my neighborhood, such as the fact that the Claremont area was sitting atop this 30 inch gas line running right through a densely populated neighborhood. I do not think any of the residents knew that. I certainly didn't, and I've lived here for 11 years. I did do some research to find out where the rest of these large scale pipes were located, and thankfully, the one that services my neighborhood is out by the main road that leads up to our neighborhood, but no houses are near it. Questions will certainly be asked, most of them pointed at Pacific Gas and Electric (P.G. & E.). During the reports, there were comments that residents' smelled gas for three weeks prior to the explosion, but little was known if anything was done about it. If indeed P.G. & E. did know there was an issue and did not do anything about it, then this rests squarely with them. Bigger questions will rise, such as "was the pipe decayed because of its location?" or "did running groundwater weaken the integrity of the pipe?", but I think one of the questions needs to be "who decided that building a densely populated community right over such a large gas main was a good idea?". Perhaps this is common, and I just wasn't aware of it before. Knowing what I know how, had I been in the market for a house, I'd think twice about buying a house where a primary 30 inch gas main lies just under my development.

Again, as a tester, I often think about these things in the abstract; a software failure that is deemed catastrophic is one where the system cannot perform its function. Even in such circumstances, the outcome rarely consists of more than lost time or loss of revenue (important in their own spheres, to be sure). In this case, a catastrophic failure has resulted in the deaths of four people (perhaps more), the destruction of dozens of homes and the displacement of hundreds of families. One thing is for certain, it's unlikely I'll be forgetting about this catastrophic failure any time soon!

6 comments:

Eusebiu Blindu said...

In cases of a big catastrophe is not easy to say always that testing could have prevented though.

Michael Larsen said...

Eusebiu, thank you for the comment. You are absolutely right. Sometimes, things just happen, and they happen in ways that no one will expect. I'm too close to the situation to make a rational and calculated observation about this situation. Many of my friends and neighbors are directly affected. I will, however, be watching with great interest to see what happens in the coming days, weeks and months regarding this incident. If testing and inspections can help make sure that another neighborhood doesn't have to go through something like this again, I'd consider it a success.

Unknown said...

Michael,
I'm glad you and your family got away unscathed.
As you said, this is likely to stay with you for the rest of your life, long after the immediate effects on the neighbourhood and friends have gone.
It goes to show that no system is fool-proof and, more importantly, that even if you do reasonable research/testing before buying a house that you will miss things that you hadn't considered before.
As you observed, in the middle of the situation, under stress, it's hard to make a rational decision. It's easy to see the parallel with our day jobs there.
I hope you and your family get on well with the aftermath of it.

Take care,

Thomas

Joe said...

Wow!

Glad to hear that you and your family are OK.

Zeger Van Hese said...

Michael, good to hear you and your family are ok.
With regards to software, I would say that "catastrophic" failures are the ones that seriously harmed or killed people, which is not too far-fetched when - let's say - working in the medical devices industry. Wrong measurements, too heavy radiation doses etc.
In a way, software can be as deadly as a gas pipe. Software testers like us can only try to take the edge off, point out and help reduce risks, try to prevent catastrophies. But in the end, we are bound to be frustrated because we can never be 100% sure. Or can we?

DiscoveredTester said...

Michael,

I'm glad to hear that you and your family are okay. We had something similar happen in a town called Ghent just up the mountain from where I live. They had fire crews from I believe 3 or 4 counties trying to contain, control, and extinguish the blaze it caused, and it destroyed numerous homes, businesses, and 1 fire house too.

Its things like this that reminds us to remember what is important. I'm glad you guys are safe. I hope PG&E will do some inspections on their lines nearby to make sure nothing else is damaged from the explosion though.