Assessing Oil Resources in the Middle East and North Africa

Transcription

1 Assessing Oil Resources in the Middle East and North Africa Roberto F. Aguilera Program Officer and Research Scholar, Energy Program International Institute for Applied Systems Analysis (IIASA) Schlossplatz 1, A-2361, Laxenburg, Austria Tel: Fax: and Adjunct Professor, Department of Economics University of Vienna Hohenstaufengasse 9, A-1010, Vienna, Austria Abstract Some energy experts are concerned that the world will soon face a global crisis to dwindling oil resources and a peak in production. This paper analyzes the concern by estimating a cumulative supply curve for conventional oil in the Middle East and North Africa (MENA) region. It does so by attaching production costs to the endowment volumes of oil in the region, including volumes from provinces not previously assessed. A Variable Shape Distribution (VSD) model is used to estimate the volumes of the previously unassessed provinces. The findings show that MENA oil should last far longer than some concerned experts claim. In addition, the production costs are lower than current market oil prices, and significantly lower than prices observed in mid

2 Some factions of the energy industry believe that conventional oil depletion will produce significant scarcities in the coming years. Too address the concern, this paper proposes to construct a cumulative supply curve to assess the availability of conventional oil (more than 15 degrees API) in the Middle East and North Africa (MENA) region. This is not the first time that there has been serious concern over the depletion of oil. In the 19 th century, certain species of whales were hunted nearly to the point of extinction for the oil obtainable from their fat, which was used as fuel for lamps. Although this was devastating to the population of whales, the transition away from whale oil had little economic implication. In the early 20 th century, when the Ford Company was producing the automobile featuring the internal combustion engine, there was widespread concern that there would not be sufficient oil to power the vehicles. However, previous predictions on the longevity of oil have been consistently premature. In 1909, the United States Geological Survey (USGS) estimated that oil in the US would be exhausted by In 1916, they reported that the earlier assessment had been too optimistic, and that oil would run out in In 1919, the USGS revised their estimate, and predicted that the US would run out of oil in 1928 (Porter, 1995). More recently, USGS estimates of oil have nearly doubled since the early 1980s. The USGS World Petroleum Assessment (2000) estimates endowment volumes, which are equal to known plus undiscovered volumes (see Figure 1), for 32 MENA provinces. The study relies on various geological techniques combined with a probability 2

3 assessment to account for the uncertainty. They publicize the mean values, which are the volumes used in this paper. In total, the USGS states that there are 88 MENA provinces. Thus, the volumes in 56 of those provinces are not presented in USGS (2000), probably because they were not expected to be exploited within the adopted 30-year time horizon. As stated by USGS (2000), the assessed areas were those judged to be significant on a world scale in terms of known petroleum volumes, geologic potential for new petroleum discoveries, and political or societal importance. Furthermore, many of the unassessed provinces are in remote areas where oil may exist but due to location and other factors are likely to be high-cost and so presumed by the USGS to be of little commercial interest over its 30- year time horizon. In this paper, we estimate the endowment volumes for the unassessed provinces using a size distribution model, called the Variable Shape Distribution (VSD) model. Both the previously assessed and unassessed province volumes are then combined with estimates of production costs to construct cumulative supply curves. 3

4 Figure 1 The relationship between cumulative production, remaining reserves and undiscovered volumes. Terminology from USGS (2000). ENDOWMENT VOLUME KNOWN VOLUME Cumulative Production Remaining Reserves Undiscovered Volume FUTURE VOLUME 1. Earlier Methods Previous size distribution models used to estimate oil volumes of unassessed areas include the log-normal and the Pareto distributions. 1,2 Initial efforts to characterize the distribution of nature s oil resources led researchers to conclude that lognormal distributions provided the best fit of data available at the time (Kaufman, 1962). Over the ensuing years, several researchers at the USGS discovered that the lognormal distribution provided overly pessimistic results (Drew, 1997). They observed that, with additional exploration, there was an on-going discovery process that could better be modeled with a Pareto distribution. The difference between the two distributions can be seen in Figure 2, where they are shown as density distributions. It is generally acknowledged that the Pareto distribution tends to overestimate oil resources, while the lognormal distribution tends to underestimate them. 4

5 Figure 2 Density distributions of number of provinces versus province size (adapted from Barton, 1995). 2. The Pareto Distribution Mandelbrot (1982) also indicates that the size distribution of oil resources could be modeled with the Pareto (i.e. fractal) distribution. 3 The Pareto distribution specifies that a log-log plot of the cumulative number of discovered oil fields versus the size of the fields could result in an approximate straight line with a constant negative slope. The straight line is observed for the larger fields, as shown by the solid circles in Figure 3 (Barton, 1995). The dashed straight line has a constant slope known as the shape parameter. The clear circles represent discovered volumes of oil and gas below levels that are currently economic. The volume of discovered oil and gas is given by the area under the curve. The straight line is extrapolated to an arbitrary minimum volume to 5

6 calculate the undiscovered oil and gas, which is equal to the area between the straight line and the open circles. Figure 3 Cumulative number of discovered oil and gas fields versus size of the field (Barton, 1995). 1995): A Pareto distribution is provided by a power law of the form (Barton and Scholz, a p N( V) CV (1) where: C - constant of proportionality. V - specified oil volume. 6

7 N(V) - number of provinces with a volume greater than or equal to V. a p - constant (scale invariable) shape parameter, which is also known as Pareto exponent, Pareto constant, or fractal dimension. Taking logarithms of both sides of Equation 1 leads to: log N ( V ) log( C ) a log p V (2) Equation 2 indicates that a plot of N(V) versus V, on log-log coordinates, should result in a straight line with a slope equal to a p and an intercept, at V = 1, equal to C. Historically, all the methods used to forecast oil volumes have been based on an assumed form of the size-frequency distribution of the natural population of oil and gas accumulations (Barton, 1995). The Variable Shape Distribution (VSD) is different in that we start by observing the curvature (on a log-log plot) given by the size and number of provinces from USGS (2000). We then develop the VSD model which allows the data to determine the specified relationship between the size and number of provinces. 3. The Variable Shape Distribution (VSD) Model This section begins by describing the VSD model (which is a size distribution model), and then estimates and validates its parameters. 4 Finally, the model is used to estimate endowment volumes of conventional oil for those provinces that the USGS has not assessed. Other models commonly used to forecast oil supply are life cycle models (e.g. Hubbert s logistic curves), rate of effort models, geologic-volumetric models, subjective probability models, discovery process models and econometric models. As 7

8 stated in Adelman et al. (1983, p. 90), the concept of deposit size distribution is an essential component of models of petroleum supply designed to reflect industry behavior in a logical way. 3.1 Description of the VSD Model The VSD method starts by ranking, in decreasing order by volume, the assessed MENA provinces from USGS (2000). When the data is plotted on a log-log scale, the vertical axis shows the rank of a province according to its volume, while the horizontal axis shows the volume of the province. Assuming most of the larger provinces have already been assessed, the VSD calculates volumes for unassessed provinces. Thus, the slope of the approximate straight line given by the assessed, larger provinces remains constant as we include unassessed provinces in the ranking. As mentioned earlier, all the earlier methods used to forecast oil volumes have been based on an assumed form of the size-frequency distribution of the natural population of oil accumulations. In this paper, we start by observing the curvature given by the USGS (2000) data points on a log-log plot. We then develop the VSD model (see Equations 3 and 4) which allows the data to determine the specified relationship between the size and number of provinces. As with all size distribution models, the original sample used to estimate the parameters contains most of the largest and promising data. This allows one to estimate the slope and intercepts, on log-log coordinates, of the straight line given by the largest 8

9 data (these parameters remain constant during the forecasting stage). The previously unassessed data will then generally contain smaller volumes than the assessed. Furthermore, many of the unassessed provinces are in areas where oil may exist but due to location and other factors are likely to be higher cost resources. 3.2 An Example of the VSD Using World Data (from Aguilera et al., 2009) As an illustration, USGS (2000) provides estimates of the world oil and natural gas liquids (NGL) endowment for 129 provinces, excluding the provinces of the United States. We then use non-linear regression to estimate the parameters of the VSD model that provides the best fit of the USGS (2000) data. As Figure 4 shows, the dotted curve generated by the estimated VSD model provides a very good fit of the actual data. Volumes of the oil and NGL endowment for provinces in the United States are assessed in the USGS National Oil and Gas assessment (1995) and Minerals Management Service Outer Continental Shelf assessment (1996). The VSD model is next employed to assess the size distribution relationship among all assessed provinces, those in the United States as well as those elsewhere. When combined with the provinces of the rest of the world, a total of 202 assessed provinces are available. The parameter values estimated with the previous sample are used. Figure 4 shows that the estimated VSD curve again provides a very good fit of the actual data. In addition, the volumes for individual provinces generated by the estimated VSD model compare very well with the values estimated by geological methods. 9

10 Cumulative Number of Provinces Version: March 19, 2009 According to the USGS, the world can be divided into 937 provinces, of which they have provided endowment volumes for 202 provinces. Thus, the VSD model is then used to estimate the size distribution relationship for all of the 937 provinces of the world, including those not previously assessed. Again, the model is run using the estimated parameter values from the first sample. The size distribution relationship is shown in Figure 4 and allows us to estimate the endowment volumes of oil and NGL in the previously unassessed provinces, assuming most of the larger provinces have already been assessed. The good fit for the previous two samples (including a high R 2 and comparable volumes) provides some confidence that the estimated volumes for 937 provinces are reasonable. Figure 4 Endowment of Oil and NGL Number of Provinces versus Province Size Pareto Distribution USGS (2000) data for 129 provinces 1000 VSD for 129 provinces (R2 = 0.99) USGS (1995, 2000) and MMS (1996) data for 202 provinces VSD for 202 provinces (R2 = 0.99) VSD for 937 provinces Size of Oil and NGL Provinces (MMBOE) 10

11 3.3 Parameter Estimation and Validation for MENA This section estimates and validates the VSD model for the MENA region, using data from USGS (2000). The oil endowment volumes for 32 MENA provinces assessed by USGS (2000), shown in the second column of Table 1, have been used to estimate the parameters of the VSD model using non-linear regression. Table 1 Oil Endowment Volumes, 32 MENA Provinces Assessed by USGS (2000) R 2 coefficient of a p = 0.55 N x = 32 determination = 0.99 r m = E-05 N m = 1 V x = 1,320,000 V s = V m = 22 Ψ = 0.30 a m = severity = 52 (r m ) am = Cumulative # of Oil World Provinces Endowment V i Province Code Province Name N t (MMBOE) r v r t a t function (MMBOE) TOTAL 977, , Mesopotamian Foredeep Basin 1 353, E Zagros Fold Belt 2 167, E Greater Ghawar Uplift 3 155, E Rub Al Khali Basin 4 126, E Sirte Basin 5 43, E Widyan Basin-Interior Platform 6 38, E Trias/Ghadames Basin 7 21, E Interior Homocline-Central Arch 8 21, E Red Sea Basin 9 15, E Illizi Basin 10 6, E Qatar Arch 11 6, E Fahud Salt Basin 12 5, E Pelagian Basin 13 3, E Ma'Rib-Al Jawf/Masila Basin 14 2, E Masila-Jeza Basin 15 1, E Ghaba Salt Basin 16 1, E Euphrates/Mardin 17 1, E Grand Erg/Ahnet Basin 18 1, E Ghudun-Khasfeh Flank Province E Shabwah Basin E Khleisha Uplift E South Oman Salt Basin E Central Oman Platform E Anah Graben E Sinai Basin E Hamra Basin E Mukalla Rift Basin E Rutbah Uplift E Ougarta Uplift E Zagros Thrust Zone E Huqf-Haushi Uplift E Oman Mountains E

12 Cumulative Number of Provinces Version: March 19, 2009 Figure 5 shows this data on a log-log plot of the cumulative number (rank) of provinces versus the size of the provinces. These data points are represented by solid diamonds. Note that the data shows leftward curvature as the volumes become smaller Figure 5 USGS (2000) data shows endowment volumes for 32 MENA provinces. Pareto Distribution 100 USGS (2000) data for 32 provinces = 977 BBOE Size of Provinces (MMBOE) The next step is to use the VSD model to provide the best possible fit of the USGS (2000) data in Figure 5. In Equation 3, we present the VSD model as a non-linear least squares (NLS) model. In particular, the problem is: n 2 { V a p V s S i x,,,, } ( V V i ) i1 min (3) 12

13 13 Subject to: S s x x m a p Vm Vx Nm Nx x m t x x m a p Vm Vx Nm Nx x m t i V V V V V V N V V V V V N V 1 ) log log log log ( 1 ) log log log log ( 1 exp 1 )] ( [1 ) ( 1 (4) where: a p - slope of straight line approximated from USGS sample points with larger province volumes (same as slope of Pareto distribution). N m - minimum number of USGS provinces (= 1). N t - cumulative number of provinces. N x - maximum number of provinces. S - severity exponent that controls the steepness of the slope of the estimated VSD curve where it separates from the Pareto straight line (on the right tail of the distribution, near the largest volumes). V m - minimum USGS province volume (BOE). V s - approximate volume (BOE) at which the USGS data begins to deviate from the Pareto straight line (on the right tail of the distribution, near the largest volumes). i V - estimated volume of a province (BOE).

14 V x - maximum volume (BOE) given by the Pareto straight line (at N m = 1). ψ - separation ratio that controls the amount of separation between the Pareto straight line and the estimated VSD curve (on the right tail of the distribution, near the largest volumes). As seen in Equation 3, there are five parameters being estimated in the VSD equation - V x, a p, V s, ψ, and S that cause the equation to best fit the volumes from USGS (2000). The parameters are estimated based on visual inspection of the curves, comparison of volumes, and inspection of the coefficient of determination (R 2 ). The VSD model is run on the 32 provinces for which oil endowment data from USGS (2000) exists. The following estimates of the five parameters give the best fit: Maximum volume given by Pareto straight line (V x ) at N m equal to 1 = 1,320,000 MMBOE Pareto shape exponent (a p ) = 0.55 Volume of separation (V s ) = 84,800 MMBOE Separation ratio (ψ) = 0.30 Severity exponent (S) = VSD Validation Figure 6 is the same as Figure 5, but now shows a continuous solid line, which is the estimated curve generated by the VSD model, using the above parameters. 14

15 Cumulative Number of Provinces Version: March 19, 2009 Figure 6 VSD estimate for 32 MENA provinces. USGS (2000) data shows endowment volumes for 32 MENA provinces Pareto Distribution USGS (2000) data for 32 provinces = 977 BBOE 100 VSD for 32 provinces = 979 BBOE (R2 = 0.99) Size of Provinces (MMBOE) Visual inspection of the USGS (2000) data points and the estimated VSD curve shows a good fit, even for the largest provinces that do not lie on the Pareto straight line. In addition, the VSD calculated oil endowment volume of 979 billion BOE, shown at the top of the ninth column ( V i ) in Table 1, compares well with the 977 billion BOE published in USGS (2000). This is supported mathematically by an R 2 coefficient of determination equal to The good fit provides an initial validation of the VSD model. 15

16 Estimated volumes ( V i ) generated by the VSD model for each province (presented in the ninth column of Table 1) compare favorably with the actual USGS (2000) estimates (shown in the fourth column of Table 1). In addition, the VSD model has been further validated using the cases of (1) known oil, gas, and NGL, (2) future oil, gas, and NGL, and (3) gas and NGL endowment. In all cases, the estimated volumes are very close to the actual USGS volumes. Also, the coefficients of determination (R 2 ) are always either 0.98 or 0.99 (results other than oil endowment are not shown in this paper; for more information about these cases, aguilera@iiasa.ac.at). 3.5 Application of VSD to Unassessed Provinces Given the validations in the previous sections, the VSD model can now be used to forecast oil endowment volumes in previously unassessed MENA provinces. The USGS has indicated that the region can be divided into 88 provinces, of which they have presented volumes for 32. Table 2 provides a list of the previously assessed and unassessed provinces in the MENA region. 16

17 Table 2 Assessed and Previously Unassessed Provinces in the MENA region Region Assessed Provinces Region Unassessed Provinces 1 Middle East and North Africa Mesopotamian Foredeep Basin 33 Middle East and North Africa Palmyra Zone 2 Middle East and North Africa Zagros Fold Belt 34 Middle East and North Africa Mediterranean Basin 3 Middle East and North Africa Greater Ghawar Uplift 35 Middle East and North Africa Nile Delta Basin 4 Middle East and North Africa Rub Al Khali Basin 36 Middle East and North Africa Lesser Caucasus 5 Middle East and North Africa Sirte Basin 37 Middle East and North Africa Thrace/Samsun 6 Middle East and North Africa Widyan Basin-Interior Platform 38 Middle East and North Africa Jafr-Tabuk Basin 7 Middle East and North Africa Trias/Ghadames Basin 39 Middle East and North Africa Levantine Basin 8 Middle East and North Africa Interior Homocline-Central Arch 40 Middle East and North Africa Haleb 9 Middle East and North Africa Red Sea Basin 41 Middle East and North Africa Atlas Uplift 10 Middle East and North Africa Illizi Basin 42 Middle East and North Africa Rif Basin 11 Middle East and North Africa Qatar Arch 43 Middle East and North Africa Essaouni Basin 12 Middle East and North Africa Fahud Salt Basin 44 Middle East and North Africa Arabian Shield 13 Middle East and North Africa Pelagian Basin 45 Middle East and North Africa Sharmah Rift Basin 14 Middle East and North Africa Ma'Rib-Al Jawf/Masila Basin 46 Middle East and North Africa Adana/Sivas 15 Middle East and North Africa Masila-Jeza Basin 47 Middle East and North Africa Tellian Foredeep 16 Middle East and North Africa Ghaba Salt Basin 48 Middle East and North Africa South Harrah Volcanics 17 Middle East and North Africa Euphrates/Mardin 49 Middle East and North Africa Tellian Uplift 18 Middle East and North Africa Grand Erg/Ahnet Basin 50 Middle East and North Africa Aaiun-Tarfaya Basin 19 Middle East and North Africa Ghudun-Khasfeh Flank Province 51 Middle East and North Africa Reggane Basin 20 Middle East and North Africa Shabwah Basin 52 Middle East and North Africa Cyrenaica Uplift 21 Middle East and North Africa Khleisha Uplift 53 Middle East and North Africa Beirut 22 Middle East and North Africa South Oman Salt Basin 54 Middle East and North Africa Wadi-Surhan Basin 23 Middle East and North Africa Central Oman Platform 55 Middle East and North Africa Yemen Volcanic Basin (North) 24 Middle East and North Africa Anah Graben 56 Middle East and North Africa Yemen Volcanic Basin (South) 25 Middle East and North Africa Sinai Basin 57 Middle East and North Africa Hays Structural Belt 26 Middle East and North Africa Hamra Basin 58 Middle East and North Africa Mukalla Rift Basin 27 Middle East and North Africa Mukalla Rift Basin 59 Middle East and North Africa Masirah Trough 28 Middle East and North Africa Rutbah Uplift 60 Middle East and North Africa Gulf of Oman Basin 29 Middle East and North Africa Ougarta Uplift 61 Middle East and North Africa North Harrah Volcanics 30 Middle East and North Africa Zagros Thrust Zone 62 Middle East and North Africa Syrian Arch 31 Middle East and North Africa Huqf-Haushi Uplift 63 Middle East and North Africa Upper Egypt Basin 32 Middle East and North Africa Oman Mountains 64 Middle East and North Africa Cyrenacia Basin 65 Middle East and North Africa Nubian Uplift 66 Middle East and North Africa Nefusa Uplift 67 Middle East and North Africa Hauts Basin 68 Middle East and North Africa Thiemboka Uplift 69 Middle East and North Africa Atlas Basin 70 Middle East and North Africa Rabat Basin 71 Middle East and North Africa Canary Islands 72 Middle East and North Africa Tindouf Basin 73 Middle East and North Africa Reguibate Uplift 74 Middle East and North Africa Guercif Basin 75 Middle East and North Africa North Red Sea Shield 76 Middle East and North Africa Araks 77 Middle East and North Africa Tuz/Corum 78 Middle East and North Africa Kardiff/Menders Massif 79 Middle East and North Africa Central Iranian Basins 80 Middle East and North Africa Central Iranian Microcontinents 81 Middle East and North Africa Lut Block and Depression 82 Middle East and North Africa Alborz Fold Belt 83 Middle East and North Africa Mirbat Precambrian Basement 84 Middle East and North Africa Fezzan Uplift 85 Middle East and North Africa Abu Gharadiq Basin 86 Middle East and North Africa Murzuk Basin 87 Middle East and North Africa East Flank Oman Sub-basin 88 Middle East and North Africa North Egypt Basin The estimated VSD curve for all 88 provinces is shown in Figure 7. The oil endowment volume is calculated to be 1,156 billion BOE. Again, the same control parameters as before have been used to generate the curve and volumetric estimates. The 17

18 Cumulative Number of Provinces Version: March 19, 2009 only change is the number of provinces being evaluated, equal to 88. The estimate of 1,156 billion BOE cannot be compared with any volumes estimated by the USGS or other organizations, since all 88 provinces have not been previously assessed. However, the validations discussed earlier provide some confidence that the estimate is reasonable. Figure 7 VSD estimate for 88 MENA provinces (including unassessed provinces) Pareto Distribution USGS (2000) data for 32 provinces = 977 BBOE 100 VSD for 32 provinces = 979 BBOE (R2 = 0.99) VSD for 88 provinces = 1,156 BBOE Size of Provinces (MMBOE) While the VSD model provides volumes for all 88 provinces, it does not indicate which volumes correspond to which provinces. As Table 3 shows, we have allocated the volumes from assessed and unassessed provinces among MENA countries on the basis of each country s share of proved reserves (BP, 2008). 18

19 The largest volumes are allocated to Saudi Arabia, and to a lesser extent Iran and Iraq. These countries account for almost 65% of the oil endowment volumes. Table 3 Allocation by Country of MENA Oil Endowment Volumes Oil Endowment from VSD model for 88 Provinces (BBOE) = 1,156 Oil Oil Endmt Oil Endmt Proved % of from VSD plus Reserve Country Reserves Total Model Growth (BBOE) (BBOE) (BBOE) (Source: BP 2008) Iran Iraq Kuwait Oman Qatar Saudi Arabia Syria United Arab Emirates Yemen Other Middle East Algeria Egypt Libya Tunisia TOTAL Reserve Growth Reserve growth is a factor defined by the USGS as the increase in reserves of a previously discovered field through time. Reserve growth provides a very significant increase to oil, gas and NGL volumes. The pessimists, for various reasons, do not generally give consideration to reserve growth. However, in spite of the complexity of estimating reserve growth, it is essential to account for this factor when assessing the availability of oil volumes. 19

20 Reserve growth, as classified by the USGS, applies to known volumes (cumulative production plus remaining reserves). In this study, reserve growth also applies to endowment volumes. It is estimated by calculating a percentage for reserve growth, based on known volumes from USGS (2000), and applying it to estimated oil endowment volumes of both assessed and unassessed provinces. We have calculated reserve growth percentages using values from Table 4. In the world total section, for instance, we calculate the reserve growth percentage of known oil, which amounts to 42.97%. This comes from dividing reserve growth (688 BBOE) by the summation of cumulative production plus remaining reserves (710 BBOE BBOE). Table 4 Calculation of Reserve Growth Percentages - based on data from USGS (2000) Mean Oil Mean Gas Mean NGL Total Mean Petroleum (Billion barrels) (BBOE) (Billion barrels) (Billion barrels) World (excluding USA) Undiscovered conventional Reserve growth (conventional) Remaining reserves Cumulative production Total Known volumes Reserve growth based on known volumes (%) Mean Oil Mean Gas Mean NGL Total Mean Petroleum (Billion barrels) (BBOE) (Billion barrels) (Billion barrels) USA Undiscovered conventional combined with oil 171 Reserve growth (conventional) combined with oil 135 Remaining reserves combined with oil 61 Cumulative production combined with oil 313 Total combined with oil 680 Known volumes combined with oil 374 Reserve growth based on known volumes (%) combined with oil Mean Oil Mean Gas Mean NGL Total Mean Petroleum (Billion barrels) (BBOE) (Billion barrels) (Billion barrels) World Total Undiscovered conventional Reserve growth (conventional) Remaining reserves Cumulative production Total Known volumes Reserve growth based on known volumes (%)

21 4.1 Discussion of Application The calculated percentages of world reserve growth are used in the estimation of endowment volumes plus reserve growth of both the previously assessed and unassessed MENA provinces. The basic assumptions in the application of reserve growth are that (i) the reserve growth percentages based on known volumes will be the same for endowment volumes, (ii) the reserve growth percentage for the world is applicable to the MENA region, and (iii) the ranking of endowment volumes plus reserve growth, by size, will be the same as the ranking of the endowment volumes without reserve growth. The estimated VSD curve for oil endowment plus reserve growth (42.97%), for 88 MENA provinces, is presented in Figure 8. This figure is same as Figure 7, with the only difference being the addition of the oil endowment plus reserve growth curve. It lies to the right of the original curve showing the VSD estimate of 88 provinces without reserve growth. It shows that at each province, each volume is now 42.97% greater than the original estimate. In other words, the original curve has been shifted to the right by 42.97%. The estimated oil endowment volume plus reserve growth, for 88 provinces, is 1,653 billion BOE. 21

22 Cumulative Number of Provinces Version: March 19, 2009 Figure 8 VSD estimate for 88 MENA provinces (including unassessed provinces and reserve growth) Pareto Distribution USGS (2000) data for 32 provinces = 977 BBOE VSD for 32 provinces = 979 BBOE (R2 = 0.99) 100 VSD for 88 provinces = 1,156 BBOE VSD for 88 provinces, plus Reserve Growth (42.97%) = 1,653 BBOE Size of Provinces (MMBOE) 5. Cumulative Supply Curves Figure 9 shows cumulative supply curves for conventional oil in the MENA region. They are constructed by graphing the volumes, by country, that can be produced economically at various average costs of production. Table 5 shows average capital costs in column 4, while average operating costs are shown in column 5. Capital costs are mainly composed of expenditures for development drilling, processing equipment, production facilities, pipelines and abandonment. Operating costs include mainly field operating costs and transportation costs. 22

23 However, there is much uncertainty associated with the estimation of production costs. Also, they vary significantly depending on which expenditures are considered. For example, the costs presented in Table 5 and Figure 9 include a rate of return on invested capital but do not include taxes and royalties. External costs, it is important to note, are also not considered, mostly because there is no agreement on what these costs could be. The grey curve in Figure 9 represents oil volumes without reserve growth and shows a volume of 1,156 BBOE. For each country, two estimates of production costs are provided, reflecting an approximate range of costs per country. The data for these costs form various sources (see bottom of Table 5). For more details about cost estimation methods, refer to Aguilera et al. (2009). Figure 9 also shows a supply curve that take into account estimated reserve growth. As can be seen, this curve reflects the same costs as in the grey curve (which does not consider reserve growth), but each block or step now represents a larger volume, due to the expected reserve growth. Incorporating reserve growth increases volumes for oil to 1,653 BBOE. 23

24 Average Total Production Cost (2008 US$/BOE) Version: March 19, 2009 Figure 9 Cumulative Long-Run Supply Curves for MENA Conventional Oil without reserve growth = 1,156 BBOE with reserve growth (42.97%) = 1,653 BBOE ,000 1,200 1,400 1,600 Oil Endowment Volume (BBOE) Table 5 Oil Endowment Volumes and Average Production Costs by Country Location Oil Oil Endmt plus Average Average Average Total Endowment Future Reserve Capital Cost Operating Production Cost (BBOE) Growth (BBOE) (2008US$/BOE) Cost (2008US$/BOE) (2008US$/BOE) Oil Growth = 42.97% Iraq Saudi Arabia Oman Libya Iran Syria Egypt Tunisia Kuwait Yemen Qatar Algeria United Arab Emirates TOTAL OIL ENDOWMENT VOLUME (BBOE): 1,156 1,653 Sources Compania Espanola de Petroleos Sociedad Anonima, CEPSA (2005) United States Geological Survey (2000) Variable Shape Distribution (VSD) model Wood Mackenzie ( ) 24

25 6. Oil Life Expectancies The life expectancies for any particular depletable resource depends on three factors its future volumes, its current production, and the growth over time of its production. 5 Table 6 reports life expectancies for MENA conventional oil. The rows indicate how many years the future volumes for these resources would last assuming production grows in the future at 0%, 1%, 2% or 3% a year. The table indicates that with production growth of 1% a year (which is above the MENA average annual growth in production over the past several decades) future volumes from previously assessed provinces assuming no reserve growth would last for 53 years. Adding in future volumes from unassessed provinces increases this figure to 62 years and considering reserve growth pushes it to 84 years. Table 6 Life Expectancies MENA MENA Future MENA a Life Expectancy in Years, at Various b MENA Average Annual c Conventional Volumes Average Annual Growth Rates in Production Growth in Production, Oil (BOE) Production (BOE) (%) 0% 1% 2% 3% From USGS (2000) 7.58E Including Unassessed Provinces 9.37E E Including Unassessed Provinces and Reserve Growth 1.434E Notes: a. Average annual production comes from British Petroleum (2008) b. Life expectancies estimated by this study c. Average annual growth in production calculated from British Petroleum (2008) 7. Conclusion The Variable Shape Distribution (VSD), a size distribution model, is unique in that it allows actual oil resource data to determine the relation between the size and 25

26 number of provinces. In the past, all methods were based on an assumed form of the size distribution of resources. The VSD model is estimated and validated utilizing data from USGS (2000). The study presents volumes for 32 provinces out of a MENA total of 88, so there are 56 provinces left unassessed. Given the validations of the VSD demonstrated in this paper, the model can be used to estimate reasonable endowment volumes of all 88 provinces. For the case of oil endowment which is used as an illustrative example in this analysis the VSD estimates a total of 1,156 billion BOE in 88 provinces. If reserve growth is accounted for, the oil endowment volume increases to 1,653 billion BOE. Thus, the quantity of available conventional oil in the MENA region is greater than often assumed, since there is a tendency to overlook unassessed provinces and reserve growth. An important implication is that MENA conventional oil is likely to last far longer than some concerned experts claim. In addition, the costs of production are lower than current market oil prices, and substantially lower than mid-2008 prices. 26

27 References Adelman, M. A. et al. (1983), Energy Resources in an Uncertain Future: Coal,Gas, Oil, and Uranium Supply Forecasting, Cambridge, MA: Ballinger Publishing Company. Aguilera, R.F. (2006), Assessing the Long Run Availability of Global Fossil Energy Resources, PhD Dissertation, Colorado School of Mines, Golden, CO. Aguilera, R.F., Eggert, R.G., Lagos, G., and Tilton, J.E. (2009), Depletion and the Future Availability of Petroleum Resources, Energy Journal 30, 1, Barton, C.C. (1995), A New Approach to Estimating Hydrocarbon Resources, United States Geological Survey Fact Sheet, Barton, C.C. and Scholz, C.H. (1995), The Fractal Size and Spatial Distribution of Hydrocarbon Accumulations, Fractals in Petroleum Geology and Earth Processes. Edited by Barton, C.C. and LaPointe, P.R. New York: Plenum Press. British Petroleum (2008), Statistical Review of World Energy London: British Petroleum. Compañía Española de Petróleo Sociedad Anónima CEPSA (2005). Personal Communication with H. Quiroga. Drew, L. J. (1997), Undiscovered Petroleum and Mineral Resources, Assessment and Controversy, New York and London: Plenum Press. Kaufman, G. M. (1962), Statistical Decision and Related Techniques in Oil and Gas Exploration, Englewood Cliffs, NJ: Prentice-Hall. Mandelbrot, B. B. (1982), The Fractal Geometry of Nature, San Francisco, CA: Freeman. Minerals Management Service (1996), An Assessment of the Undiscovered Hydrocarbon Potential of the Nation s Outer Continental Shelf, Report MMS Porter, E.D. (1995), Are We Running Out of Oil? American Petroleum Institute Policy Analysis and Strategic Planning Department. Discussion Paper 81, December. United States Geological Survey (1995), National Oil and Gas Assessment. CD-ROM. United States Geological Survey (2000), World Petroleum Assessment. CD-ROM. Wood Mackenzie ( ), Upstream Asset Analysis Reports.

28 Footnotes 1 The log-normal distribution is a continuous distribution in which the logarithm of a random variable is normally distributed. 2 The Pareto distribution is also known as the power law, Bradford, hyperbolic, fractal, scaling, Zipf (when the slope is 1.0), log-geometric, and J-shape distributions. The Pareto distribution, named after the Italian economist Vilfredo Pareto, is a power law distribution, where the exponent of the power law is constant. 3 The Pareto distribution is the probability distribution characteristic of fractals. Mandelbrot (1982) defines a fractal as a structure that contains an organized arrangement of repeating patterns over many ranges of scale. In fractals, the part is reminiscent of the whole. Some fundamental properties of fractals are selfsimilarity, self-affinity and scale invariance. 4 Aguilera (2006) provides the full mathematical development of the VSD model. 5 Future volumes are equal to endowment volumes minus cumulative production. Thus, MENA cumulative production of 219 BBOE (presented in USGS, 2000) is subtracted from the MENA endowment volumes presented earlier, resulting in the following future volumes: From USGS (2000): 977 BBOE 219 BBOE = 758 BBOE Including Unasessed Provinces: 1156 BBOE 219 BBOE = 937 BBOE Including Unassessed Provinces and Future Reserve Growth: 1653BBOE 219 BBOE = 1434 BBOE