Thread: FN Curves and Risk Analysis

Hi all

I am trying to assess the risk for a project and must create an FN curve from a similar (but much larger and more complex) set of data below.

Using the potential frequency of fatalities (as F) and the equivalent fatalities (as N) I need to make the FN Curve. I think it will be logarythmic but don't know whether it is a stepped graph, line graph, area graph or whatever. I am unsure also whether I simply add up all of the frequencies of the same no of equiv. fatalities or whether I use some sort of complex cumulative frequency (which I am unsure how to use properly in this context).

I have extensively seached the net but no simple answer is forthcoming.

Many thanks

Yogi

Using the scheme outlined in http://home.att.net/~d.c.hendershot/...s/riskland.pdf, you should add up the frequencies of any outcome for which the equivalent fatalities are greater than or equal to a given value of N. Then plot the frequency sums as the ordinate against N as the abscissa. In the small data set you gave in the attached Excel file, you would get the following:

N F log F
0.1 12.31737 1.090517987
1 12.21406 1.086860049
10 12.20203 1.086432088

The third column is the base-10 logs of the frequency sums. I usually see F-N curves with a line connecting the points, but there are many ways one might display the information. The attached Excel file displays an F-N curve based on your data, with the frequency sums expressed in logs.