A Titanic Failure of Disaster Planning

MP900407559As we mark the 100th anniversary of the sinking of the Titanic, questions that were raised then are being repeated now; how did this happen?  Obviously, I'm not going to cite all of the opinions, but as one can imagine, blame is assigned virtually everywhere.  Regulation (or lack, thereof) is to blame.  Management is to blame.  You know what I think?  They all miss the point, entirely. 

Why?  Because 1,500 people are dead, that's why.  Titanic was doomed before it ever left port.  One thing we can state with certainty; it was known that if the ship sank in an isolated area, then there wouldn't be enough lifeboats on-hand to save all of the lives on board.

Analysts point to a comedy of errors that, if they hadn't occurred in sequence, then Titanic wouldn't have sunk.  There's that word, if, again.  But this flies in the face of competent disaster planning.  We already know that if events occur as expected, then there won't be a disaster.

Let's enter the equation at the real-time departure of Titanic in its actual condition (meaning, not enough lifeboats).  Here are some of the assumptions that might have been made:

  • The ship will not sink
  • The number of lifeboats are an acceptable risk versus the unlikely possibility the ship might sink
  • Other disasters may occur (e.g. boiler explosion), but the ship won't sink and there are sufficient lifeboats
  • Even if the ship were to sink, it will sink very slowly – or close to land – allowing landed resources or other ships to respond in sufficient time
  • The experts on board – and in the surrounding area (such as the California, or wireless operators) – will act predictably (i.e. not make any mistakes), thereby avoiding errors
  • Icebergs will be exactly where we expect them to be

Here's a question I haven't seen (although I'm sure someone has probably raised it):  Even if the ship had sufficient lifeboats on board, would the crew have been able to launch all of them in the rapid time that titanic sunk (estimated at 2 hours, 40 minutes)?

A good risk management team understands the first rule of disaster planning – follow Murphy's Law:  If anything can go wrong, it will.  The second rule?  Follow-up with O'Toole's Commentary:  Murphy was an optimist.

  1. No plan will ever be correctly analyzed unless it begins with an honest assessment of the following question:  What can go wrong?
  2. The rest becomes a matter of probability.  What is the likelihood that anything (and everything in-between) on the list created by question one might go wrong.

Then, it's a matter of prioritizing between several factors, such as time needed to address & correct, manpower required, cost vs. budget, insurance, politics, etc.

I wish we could go with something more simplistic:  The chance of the bread landing with the buttered-side down is directly proportional to the cost of the carpet…

Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedInPin on PinterestShare on RedditEmail this to someone