[[ Check out my Wordpress blog Context/Earth for environmental and energy topics tied together in a semantic web framework ]]

Wednesday, January 03, 2007

Censored Data

The lower 48 reserve growth analysis from Attanasi & Root likely suffers from the curse of censored data. If you look at the data posted on TOD recently, you can see the window gets clipped in a ~15 year timeframe from 1977 to 1991. Yet the authors decide to show the data as if the growth looks continuous up to 90 years.

This leads to charts like the following where the authors deem it alright to extrapolate backwards to a much larger window:

Now, most analysts admit that many of the fields included in the analysis become "shut-in" or dormant or plugged during their lifetime, and then when the economics become reasonable, the field can produce oil with a renewed vigor. Unfortunately this means that reserve growth has to drop somewhat reasonably to a near-to-zero value during the dormancy stage. The field remains essentially abandoned so you would think that this approximation would make some sense. Yet you would never notice this from a cursory glance of the sliding data of the first chart transforming into the larger aperture of the second chart. Properly supplying reasonable censored data for a much-larger-than-15 year window would in fact suppress the slope of the profile in the second chart considerably.

Doing censored data correctly falls into the class of statistical problems called survivability analysis. Analyze censored data incorrectly and you will face results fraught with errors. A classical case frequently referenced concerns survival studies of tobacco smokers. These studies would go on for 10 years or so, and the statisticians would collect enough data to project death rates in the future. The results become problematic when the extrapolator assumes a linear projection whereas in reality the actual data could vary non-linearly. Without the uncensored data available, i.e. a longer time scale, one could come to the wrong conclusion. Big Tobacco has used this statistical sleight-of-hand in the past to justify any conclusion they wanted to, preferably to stress that tobacco does not cause premature death (or addiction). I fear that Big Oil can practice the same sleight-of-hand until people start to catch on.

The censoring in the particular case of reserve growth matches the typical tobacco survival in a general fashion only; for reserve growth the censoring of data happens in the previous years to the 1977-1991 window. Let me give you a better analogy as to what this means. Say, instead of oil, you looked at growth data of people (e.g. say their height) placed in suspended animation at age 15, and you added that to data 100 years from now after reviving the subjects from their hibernating stupor. Assuming that suspended animation prevented aging, you would have a completely misguided view of how people grew as a function of time. Without understanding that cold storage prevented growth, you would think that humans continue to increase in height well past their 90's. To top it off, these humans would also have to consist of a range of sizes from tiny dwarfs to giants so as to match the range in oil field sizes. That basically amounts to what R&A did with their data. They completely neglected any part of the data that didn't show an increase in value. A conservative read of the data would only justify looking at the first 15 years of the second chart.

To summarize, for all we know, many of those regions outside the 15 year data collection window became shut-in or dormant which meant they had zero reserve growth for a long period. But then to resurrect the data and show growth 90 years after discovery displays an incredible amount of deception or rank ignorance on how censored data works. If Attanasi & Root had an agenda of deception, then I consider them equally as evil as the tobacco "scientists" who jiggered the survival statistics over the years leading to the premature deaths of millions of people worldwide.

If on the other hand, A&R simply chose to bathe in their own mathematical ignorance, these two should forever inhabit a special state of suspended animation, and we shouldn't wake them up to do more harm.

It took me about a year to actually understand what this weird reserve growth meant. Now that the extra insight hit me like a ton of bricks, I have become more than a bit peaved. No wonder they call it censored data, Big Oil doesn't want the truth to come out. You'd think a company would fire somebody who makes this bad a mistake, but think about the agenda of the employer and they would deem the dynamic duo of A&R as vital to their long term best interests. Exactly. They want to present this optimistic reserve growth to the public and if they can find tools (i.e. useful idiots) to keep parading around their bad statistics, they would just as soon keep said employee on staff than to fire his sorry behind. Welcome to how the Tobacco industry worked in the 20th century and the way Big Oil works today.


Professor Anonymous Sanka said...

Your final paragraph about sums up the pathology inherent in "modern life". If survival is dependant on creating wealth for oneself (as meager as it might be), then it will be clear to all but the most dense that survival equates to *ideological dependancy* upon those systems (monetary/capital) providing for our "survival". The epitome of this monetary/capital system as understood by the public at large are the financial markets (the stock market etc). It is often projected to the public via the media that the health of their/it's economy is directly correlated to these "markets"; when this is often very much *not* the case. Low unemployment might *sound* great to John Q Public, but Wall Street hates it, since wages will tend to rise and hence, inflation. It's far more insiduous and beneficial to manufacture "cheap" credit than raise wages. How else to make sure the hook is well set?

My very long winded point here is that our favourite "neutral lubricant" allows anyone the defense - "I've got a mortgage to pay".

1:10 AM  
Professor Anonymous Anonymous said...

Perhaps what Attanasi & Root are trying to get across is that reserves depend on price to a large extent. Rather than graph reserve growth vs time, graph the growth vs (projected) price. The price of finite oil/gas must increase over time unless obviated by some miraculous energy discovery.

1:48 PM  
Professor Blogger WHT said...

I don't know if they have or really should display any interest in price. A&R claim to be pure geologists but they do appear to have some sort of hidden agenda. With that said, you are indeed correct that the effective cost of a barrel of oil certainly has increased that much.

6:01 PM  

Post a Comment

<< Home

"Like strange bulldogs sniffing each other's butts, you could sense wariness from both sides"