Reserve Growth Curve
Creaming Curve
Several of these papers including Attanasi and Root (quote below) point to the difficulty in separating reserve growth from new discoveries that occur in close proximity to the original fields in some finite or slowly expanding area/volume. So in one case you have backdating and in the other case you just have additional discoveries -- which get counted conventionally, i.e. not backdated, as a creaming curve.
"At the basis of all statistical extrapolations used to project field growth is the assumption that fields are well defined and that patterns of the past will continue into the future. The enigma of field growth is due, in part, to changing and ad hoc field definitions. In individual cases, fields are defined on the basis of convenience by regulatory agencies, by operators, or simply as artifacts of the historical discovery process. For example, the 1990 EIA OGIFF file had several hundred more fields than the 1991 file. Nearly all of the entries not in the 1991 file had been combined with other, larger fields. As new pools, reservoirs, and fields are developed, they are frequently included in older fields for geologic reasons and perhaps for the convenience of unit operators and regulators. When this happens, the discovery date of these more recently discovered hydrocarbons are backdated, and growth of the older fields is extended."This brings up the fine distinction between reserve growth curves and creaming curves describing a region. These two curves have similar shapes and it seems a matter of taste which term to use when you compile data. Both creaming curves and reserve growth curves have steep slopes that turn into a creep upward.
My comment prompted this reply from rockdoc in a peakoil.com thread.
strictly speaking this should not be a problem....increase in area would refer to a single field, if not a single reservoir. It is not a separate discovery which would mean it is not in hydrodynamic communication with the rest of the field....ie. a separate structural closure or a separate trap. What is being referred to is that often quite a bit of time after discovery additional work (3D seismic, outpost drilling etc.) identifies that initial assumptions regarding oil/water contact, depth conversion, location of bounding faults etc. were incorrect....as a consequence the area of closure in the field/pool can be increased. The unfortunate problem in all of this is that often the folks who play with the "numbers" have no idea about the reservoir dynamics. As an example people still like to speak of Ghawar as if it were a single field when in fact it is made up of a number of separate closures, and several reservoirs, some in communication some not. The geologists and reservoir engineers working on those particular projects know what is going on and report the information correctly....the "powers that be" decide they can group and lump that data. Hence folks simply grabbing that data from various sources will be led astray. Using Ghawar as a whole in any kind of statistical analysis skews the results beyond belief....for instance look at the backdated discovery curve with the huge peak in the sixties.Claiming it as "magic" that the ordinary layman can't grasp sure seems a condescending attitude, especially since we know this doesn't involve quantum mechanics or relativity theory or some other non-intuitive concepts. I would rather attribute this lack of desire to impart wisdom to a combination of factors including laziness, arcane lingo, messy data, and double-talk politics that the business of oil production seems to thrive in.
You are right about the shape of the curves, but as I have pointed out on another thread it is pretty easy to get mislead by a particular creaming curve when you don't understand the underlying geology of the basin. Hence Hubbert missing new discoveries in the US, and creaming curves for Algeria created in 1990 missing out on about 6 billion barrels of new discoveries in the following decade. As to understanding the geology, Hubbert was a brilliant geologist (see my thread regarding his contributions) but the limit of our understanding of the subsurface at any point in time is often a product of the sampling density of that "universe", which, when Hubbert was ruminating on the problem, was good but by no means stellar.
I guess all I am trying to get across is that the idea of reserve growth makes perfect sense to us folks who deal with the discovery-development-production history of fields. The unfortunate thing is it just comes across as magic to those that haven't been directly involved in it. Not sure if there is a better way of explaining it.
I still claim that nothing prevents us from nailing a good macro understanding using nothing more than an expected value for a typical reserve growth model or creaming curve. I assert that the two curves basically amount to the same thing; in particular when we apply the numbers in a statistically valid way to the oil shock model, either one fits nicely into the maturation phase.
Mike Malloy on AirAmericaRadio.com came on fire today (more than his blazing norm), with an interesting rant relating to the fact that the U.S. guvmint and oil companies understand peak oil all too well, but hide that fact expertly ... and endorse just the figurehead they need:
By gosh, they will support Bush because he is a jibbering moron