Production data are extracted from public domain sources, and so it is not possible to fully identify those factors which make each property unique. We are thus limited in our ability to understand why structures produce in a particular way without detailed, site-specific information.
Our ability to infer production trends from curve fitting exercises is similarly limited, and hence our results are only indicative of general trends, and should only be interpreted in this manner.
Inventory may change
We only considered those structures that were destroyed in the 2004 and 2005 hurricane seasons. An additional 76 structures were severely damaged, and some of these may not return to producing status.
Conversely, early producers and those structures with a sufficient amount of remaining reserves may be redeveloped in the future, which would subsequently reduce the quantity and value of “lost” production reported. We assumed that all idle structures will not return to producing status. This is believed to be a reasonably good assumption, but we recognize that in special cases some idle structures may have returned to production status if they were not destroyed.
The impact of damaged or destroyed infrastructure that served as an active conduit, link, or hub for other producing structures was not assessed.
Stable conditions
Decline curve analysis is an empirical technique that relates production data with one or more attributes, such as time, cumulative production, reservoir pressure, etc. Empirical equations fit data to assumed model forms, and do not include most of the factors that affect past, present, or future performance.
Extrapolating the results of an empirically-derived equation to the future assumes that all the factors affecting performance in the past have exactly the same cumulative effect in the future. This is a strong, and certainly, questionable assumption. The use of decline curves necessitates assumptions regarding operating policy, field investment, mechanical problems, marketing issues, and the occurrence of exogenous events.
The collective set of all these conditions are assumed constant for all future time and are referred to as “stable reservoir and investment conditions.” Changes to any of the above-named factors have the potential to dramatically change both the production rate and reserves, which will impact the forecast results.
Decline curve reliability?
The production rate of a well, group of wells, structure, lease, field, or other aggregation unit can be fit to any functional form or curve type. The function may fit the observed data so well that the user may consider it to be an accurate and reliable predictor of the future.
Such a conclusion would, of course, be a serious mistake, since many other factors that we cannot control or directly account for will impact production levels. The ability to forecast is severely restricted by conditions that are both unobservable and unpredictable.
Stage of production
The stage of production of an asset is often used as a rough indication of the amount of uncertainty that can be expected in forecast models, but such indicators are often ambiguous and should be used with caution. Early in the life of a field, little is known about the extent, quality, and drive mechanisms of the reservoir, and at this time, reserves estimates and production forecasting is the most uncertain.
As a field is developed and production performance accumulates, the range of error often decreases, assuming no change in investment strategy. For mature assets, forecasting is considered less uncertain because a smaller number of factors influence production and dwindling remaining reserves are likely to be irreversible.
“Less uncertain” is not certainty, however, and there is still a large degree of unpredictability that can occur near the end of the life cycle of production. Offshore production has different economic characteristics than onshore production, where fixed costs are smaller and production profiles tend to decline at smoother rates.
In offshore operations, production levels tend to be chaotic throughout the life of the field, and smoothly declining profiles near the end of production are not common, exacerbating the difficulty associated with forecasting abandonment timing.
