Wednesday, April 6, 2011

World Oil Reserves

An Armchair Audit of World Oil Reserves

It is well known that the lack of international standards on reserves estimation methodologies has led to disparities in the oil reserves reported by the more than 90 oil producing countries. For instance, most of the countries with large oil reserves, notably, Iran, Iraq, Kuwait, Saudi Arabia and Venezuela are known to report up to five times more reserves to sustain a similar level of production capacity when compared with other major oil producing countries such as Mexico, Norway, Russia, UK and the USA. In the case of the North Sea, Norway reports twice as many proven reserves as the UK although the petroleum geology of the basins is common in size, and the number of fields to both countries. Additionally, because of the importance of oil in their balance of payments and in the economy, many oil-producing developing countries outside OPEC have developed a policy of withholding technical information on reserves and production. Although production statistics are temporarily held confidential by some countries, it is generally felt that because oil is such an important commodity in world trade, surveyed production data by reputable international firms is fairly reliable. Usually when the production data is released it compares satisfactorily with the surveyed data. In the case of reserves, however, there is no source for comparison.

The objective of this study is to use decline curve analysis of production data to independently corroborate the crude oil reserves of ten important oil producing countries over a period of twenty years, from 1980 to 2000. Five are members of OPEC - Saudi Arabia, Kuwait, Iran, Venezuela and Nigeria; the others are the USA, Russia, Mexico, Norway and the UK. Together these ten countries account for almost two-thirds of the reported world’s crude oil reserves and all are net exporters of oil with the exception of the USA.

The Reserves Estimation Issue
The determination of proven, probable, and possible reserves in an oilfield is a difficult task primarily because reserves cannot be measured directly. Oil reservoirs may be buried several miles below the surface or located underneath tough environments such as deep water. Additionally the oilfield may consist of a few massive (thousands of feet in thickness) reservoir sands, typical of the geology in the Persian Gulf, or multiple stacked thin sands (~25 feet or more) as in Venezuela. In any case, the only source of direct measurements of rock and fluid properties are the wells drilled, but these can be sparsely located - as few as one for every 25 square miles - in an oilfield which can be as extensive as 2,800 square miles (178 mi. x 16 mi.) like the famous super giant Ghawar field in Saudi Arabia.

The combination of sparce areal coverage and the thickness distribution of the oil column, among others, limits the number of samples of reservoir data that can be economically retrieved from an oilfield. The bottom line is that reservoir data is often insufficient and therefore the determination of reserves is subject to judgmental
factors. To stay within a 5 percent range of error which is reasonable, the key is consistency in the reserves estimates using different tools over the life of the field. Prior to start-up of production, volumetric estimates of reserves, either deterministic or stochastic methods, are the only option. As production data become available
several methods based on reservoir performance, such as production decline analysis and reservoir simulation, provide more accuracy in the reserves estimates. The decline curve approach is generally regarded as one of the primary tools for reserves calculations, both for its simplicity and minimizing of arbitrary assumptions on geologic and raw engineering data that go into the volumetric formula. Decline analysis is based entirely on measured production data and the logistic equation has proven to be a powerful and versatile decline tool. It is applicable not only to individual oilfields but also to conglomerates of oilfields, typical of oil producing countries.


The logistic decline method was applied to the production data of the ten sample countries mentioned previously. Three periods were chosen to evaluate the reserves: 1980, 1990 and 2000. These years represent milestones in recent world oil production history which must be factored into the decline analysis.
The OPEC countries experienced a robust production growth through 1980.
Thereafter, production rates were severely curtailed, picking up again at the end of the decade and continuing through 2000. Saudi Arabia’s production history is a typical example. Its production grew continuously from 1.3 million barrels per day (b/d) at the start of the 1960s to an all time high of 10 million b/d in 1980. Production subsequently dropped to a low of 3.2 million b/d in 1985, rebounding to 8 million b/d
in 1991 and essentially staying around this level through 2003. In contrast, none of the five non-OPEC countries in this study has experienced any major oil production fluctuations over the same time period.
Straight line portions of the decline plot were constructed using as the end point the cumulative production data through each of the three years, 1980, 1990 and 2000.

This theoretically could generate three different best fitting straight lines, each with a different ultimate recovery value in time as more data became available. In order to minimize data fluctuations caused by short-term events, the production rate values were smoothed over 5-year intervals. This did not mask, however, any major production changes due to long-term market variations as occurred during the 1980s, new IOR programs and any significant new discoveries made during the course of the country’s production history.
Figs. 1, 2 and 3 illustrate the graphical techniques used to establish the ultimate recovery values for the USA, Russia and Saudi Arabia, respectively, over the 20 year evaluation period. In the case of the USA and Russia, a single straight line was valid for the three periods, 1980, 1990 and 2000. This implies that their ultimate recovery values were stable - no major changes due to market events, new discoveries or IOR
programs - over the entire 20 year interval. For Saudi Arabia, however, the decline history through 1980 is not useful for extrapolation because steady state conditions did not prevail at the time. Steady state conditions generally set in after more that 20 percent of the initial reserves have been produced.

Likewise, the production data during the 1980s also has to be filtered because of severe market events. The valid straight line portion of the decline curve, after 1990, indicates an ultimate recovery of 160 Gb. Similar detailed analyses were made for each of the countries in this study.

It should be emphasized that with the explicit exclusion of the unsteady state and the non-production mechanically related data values such as those pertaining to extreme market events, the decline lines generated in this study are all data-driven. In the event that the decline plot had not stabilized by a specific year for example 1980, the ultimate recovery of the following period (1990) was assumed applicable to the previous period. These back extrapolated values are marked with an asterisk.This situation affected six countries. Norway and the UK only began producing in the mid 1970s, not long enough a period to establish steady state by 1980. Others like Iran that began production around 1915, Mexico (1900s), Saudi Arabia (1940s) and Nigeria (1960s) still had not reached their steady state production threshold by 1980. Remaining reserves are equal to the ultimate recoverable reserves minus the cumulative production. This relationship is used to calculate the ‘decline’ remaining reserves value once the ultimate recovery is established. They were then compared with the remaining reserves reported for the different countries at the end of each
of the three years (1980, 1990, and 2000) evaluated.

No comments:

Post a Comment