Science Risk management and intuition 


Written for Canadian Community News by Mike Sterling

April 4, 2012

To Comment on this article Click Here

In this article we will define the concept of risk management and describe its use.  We will show that intuition or common sense sometimes fail even in simple cases.  We will note that the process of risk management in the planning stage has to reach a certain level to be meaningful.  To curtail the process early sometimes eliminates valid options.

Risk management is the identification, assessment, and prioritization of risks (defined in ISO 31000 as the effect of uncertainty on objectives, whether positive or negative) followed by coordinated and economical application of resources to minimize, monitor, and control the probability and/or impact of risky and/or dangerous events or to maximize the realization of opportunities.

This is a form of the standard definition of the field of Risk Management.  It is used extensively in business, engineering and science.  Often it is turned around and called Opportunity Management and is directed toward success  like financial strategies rather than avoidance of failures. 

The long, ongoing investigation into the need and location of Deep Geologic Repositories has to use risk management techniques to access solutions to a known problem and set of problems.

Risk Management uses extensive information gathering techniques and often creates models of processes. The models are loaded with carefully gathered information and run using stochastic methods sometimes known as Monte Carlo techniques.  This involves random samples from known or estimated distributions of risks. 

These samples perturb or play out in  the model to allow it to go through many different combinations of variables that may interact with each other in unforeseen and completely non-intuitive ways.

One of the most interesting results of such models is that they often point the observer to areas where they need much more information. 

The results tell them that they have not considered enough.  Aborting the Risk Management process too early often blinds the investigator to additional solutions and options.  That is why these studies take so long to complete.

So the process is:

  • lay out options

  • gather information

  • form and feed a model

  • go back and gather more information or options based upon what has been learned.

The above process is self feeding and if aborted too early may lead to inconclusive results.

Years ago I used to be called into a risk management type SWAT team when things were falling apart.  I learned then that intuition was the enemy and not a friend.

Let's give an example of how intuition can fool us. It's a simple one, but revealing. 

In the example I will give there is a way to figure out the optimal solution right away, but we could also use the Monte Carlo Method to get the result.  In this case the options are fixed. We could use Risk Management to get to the solution without the direct use of mathematics. 

This is particularly important in complicated problems, where the solution set is not clear.  The problem below is not a difficult one, but it has tripped up many very smart people.

The example is called "The Monte Hall" problem not to be confused with the Monte Carlo Method.

A guest on the TV show 'Let's Make a Deal' is asked the following:

Behind one door is a new car; behind the others, goats. You pick a door, say No. 1 [but the door is not opened], and the host, who knows what's behind the doors, opens another door, say No. 3, which has a goat. He then says to you, "Do you want to pick door No. 2?" Is it to your advantage to switch your choice?

It's fair to say that most people using their intuition would say it make no difference.  It's 50-50.  Guess what, they would be wrong.  See why in this video:

 Click here for the Video It will come up in a full new window.  When you are done, just delete the window.

Please play the video, it's well done

The video clearly explains the right strategy, but guess what?  You can use Monte Carlo Techniques  before you answer Monte Hall's question and not have to work out the solution.

Of course in Monte Carlo risk management models, you would be able to run tens of millions of simulations of this problem to get the right answer before you answered Hall's question.  You could run the model to drive your confidence in the right answer to 99.999999999....% without going directly to the reasoned solution.  If you decided it was an open and shut case and did not switch, you'd reduce your chances of success. 

It's clear that intuition is not something we can use in this simple problem and intuition does not help us in more complicated problems either.  There are not many like Einstein, who can dream the impossible dream and have others test it and find it to be true.  Experts can gather information and test via models that guide them to solutions.

Some problems are just too large for clear 100% certainty  results. People concerned with weather cannot model with enough certainty to get perfect predictions.  In the case of the DGR, the size and complexity of the model will be in the solvable range, if enough time and care is put into data gathering.

When faced with situations we often don't talk about intuition, but we do use a phrase that is more acceptable and less like looking at Tarot Cards.  We call it  'Common Sense'.  Common Sense can be uncommonly wrong.

The DGR problem is very difficult and intuition does not help us much.   Common sense can aid us in the beginning by allowing all the options we can think of including doing nothing to be included in any Risk Management Model.  The problem with intuition at the start is that it often shuts off options from consideration.   The thought crosses our minds:   Well, that can't be true!   I can ignore that. This happens all the time.  That temptation arises even in the Monte Hall problem.

Since DGR options have planned life cycles in the many thousands of years, most probably 100,000 years, it is clear that new options might be found over time. And new risks evolve out of careful research.

 For example low cost high speed breeder reactors might be used to reduce the amount of waste on site, if the economics improve.  The economical use of new technology can come quickly or over many hundreds of years.  Whatever solution is implemented, it must be adaptable to new technology.

Continued Click the orange arrow to go to the 2nd column

A site in North America geared at 100,000 year life would have to consider the next ice age in any solution.  This would be important for all the options.  What would the result of above ground storage be with ice building up to  a thickness of 2 miles?  Is this improbable?  Of course not, we've had them before.  So it you have a vision, it has to be viewed through the lens of 100,000 years.  We just are not used to thinking in eons.

Do we care about events 20,000 years from now?  Does our vision  include that time span and more?

Other sources of danger could come from natural storms.  In the Dallas Fort Worth area, on Tuesday, April 3, 2012 30,000 pound trailer tractor trucks were tossed hundreds of feet in the air.  What is the effect of an event like that on structures and storage containers for waste?  Can a solution be found that takes into account types of natural disasters that we have not seen in our lifetime and will not see?.

What is the effect of movement on sedimentary rock vs. crystalline structures.  All these situations are being considered by NWMO and will be fed into models with their corresponding risk distribution.

So, how do we select options and see risks?  In our last article we outlined what NWMO has done so far.

Technology NWMO's options for storing nuclear waste Read More

Are there more possible options?  Most probably the answer is yes.  What are they?  I don't know.

We've been talking about risk.  What exactly is it and how is it measured and used?

We have all become familiar with the Bell Curve in school.  If you studied hard, you were over on the right on the standard graph.  This is called the Gaussian Distribution after the famous German Mathematical Wizard Gauss.

It you have a series of events or sequences that combine, the resultant will be a bell shaped curve.  For example, I have an old house with a back stairs.  The stairs are worn down.  If you look at the treads and measure them, they form perfect shallow and inverted Gaussian Bell Curves.

The combined feet and paths of countless people over 134+ years have followed the expected distribution, but some feet have not.  They are not exceptional because without them, we would not have the Bell Curve, we would have a narrow groove.  Some have followed the path least taken.

Risk as shown in 3D

So, if one occupant constantly went up the right side of the stairs, the preponderance of people followed a path closer to the centre of gravity of the Bell.  It is sometimes viewed in 3 dimensions as the figure above shows.

What Risk Management does is to reduce risk to the far blue areas and control the processes to avoid any proximity to the 'hot areas'. 

If you can keep the probability of a failure of your process and solution matrix forever in the blue area, you are ok and in a low risk zone.  The risk cannot be reduced to zero with any solution that is so complicated, however, but it can be controlled.  In our case no matter how far you get from the hot zone, there is still risk.  This is not geographic distance necessarily, but it is the combinations of the risks properly controlled.

You have to acquire as much data on the risks as possible and also their individual distributions.  For example, a risk like physical distance diminishes radiation and high risk  is skewed toward close proximity, but the combinations of all the risks taken randomly together will form the Bell curve.  We want the total to be pushed way, way out independent of distance of the site from our current location.

The problem of waste management is pervasive and is somewhat distance insensitive due to the effects of wind currents.  It is distance sensitive in another sense too.  The further the distance to transport, the higher the risk of transport accident or attack.

There are questions to be asked of the experts as the process plays out. What exactly are the risks?

We have to pay attention to other areas too.  Every country in the world is concerned with the decision matrix, methods and risks associated with all the options.  If the experts come up with low risk solutions for Bruce County, Ontario, then Canada, North America and more, it will be a triumph of process, but still not immune to risk.  All the options have to be accessed with risk in mind including the present methods of storage and shipment.

There is a huge list of risks and processes to control them.  We don't know all of them yet and the experts don't either, but they have a process to find out what they are and what is the control process associated with them.  They also have the advantage of years of study of the problem

It will be interesting to see the sophistication and depth of the questions grow.  I'm sure they will improve over time.  It will be interesting if all of us can say with certainty that they have studied what is available, attended public meetings and can say that they are convinced of some solution that comes from the many that will be presented including staying with the status quo.

Next time we will take a look at what Finland has done with their waste management problem. 

Scrolling stops when you move your mouse inside the scroll area.  You can click on the ads for more

Survey Participate in our latest Kincardine Times survey Read More Survey Participate in our latest Saugeen Times survey Read More Survey Participate in our latest Walkerton News survey Read More
for world news, books, sports, movies ...

Wednesday, April 04, 2012