Version: March 19, 2009
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: 43 2236 807 262 Fax: 43 2236 807 488 Email:
[email protected] and Adjunct Professor, Department of Economics University of Vienna Hohenstaufengasse 9, A-1010, Vienna, Austria Email:
[email protected] 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-2008.
1
Version: March 19, 2009
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 19th 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 20th 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 1935. In 1916, they reported that the earlier assessment had been too optimistic, and that oil would run out in 1921. 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
Version: March 19, 2009
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 30year 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
Version: March 19, 2009
Figure 1 The relationship between cumulative production, remaining reserves and undiscovered volumes. Terminology from USGS (2000). Cumulative Production KNOWN VOLUME ENDOWMENT VOLUME
Remaining Reserves
FUTURE VOLUME
Undiscovered 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
Version: March 19, 2009
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
Version: March 19, 2009
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).
A Pareto distribution is provided by a power law of the form (Barton and Scholz, 1995): N (V ) CV
where: C - constant of proportionality. V - specified oil volume.
6
a p
(1)
Version: March 19, 2009
N(V) - number of provinces with a volume greater than or equal to V. ap - 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 p log 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 –ap 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
Version: March 19, 2009
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
Version: March 19, 2009
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
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 R2 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 10000 Pareto Distribution USGS (2000) data for 129 provinces
Cumulative Number of Provinces
VSD for 129 provinces (R2 = 0.99)
1000
USGS (1995, 2000) and MMS (1996) data for 202 provinces VSD for 202 provinces (R2 = 0.99) VSD for 937 provinces
100
10
1 0
1
10
100
1000
10000
100000
Size of Oil and NGL Provinces (MMBOE)
10
1000000
10000000
Version: March 19, 2009
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) R2 coefficient of determination =
0.99
ap =
0.55
Nx =
rm =
1.62879E-05
Nm =
1
Vx =
1,320,000
Vs =
84800
Vm =
22
Ψ=
0.30
am =
0.31435
severity =
Province Code
Province Name
Cumulative # of Oil World Provinces Endowment Nt (MMBOE)
TOTAL 2024 2030 2021 2019 2043 2023 2054 2020 2071 2056 2022 2016 2048 2004 2009 2014 2075 2058 2010 2006 2074 2011 2015 2089 2033 2047 2025 2028 2061 2031 2013 2017
Mesopotamian Foredeep Basin Zagros Fold Belt Greater Ghawar Uplift Rub Al Khali Basin Sirte Basin Widyan Basin-Interior Platform Trias/Ghadames Basin Interior Homocline-Central Arch Red Sea Basin Illizi Basin Qatar Arch Fahud Salt Basin Pelagian Basin Ma'Rib-Al Jawf/Masila Basin Masila-Jeza Basin Ghaba Salt Basin Euphrates/Mardin Grand Erg/Ahnet Basin Ghudun-Khasfeh Flank Province Shabwah Basin Khleisha Uplift South Oman Salt Basin Central Oman Platform Anah Graben Sinai Basin Hamra Basin Mukalla Rift Basin Rutbah Uplift Ougarta Uplift Zagros Thrust Zone Huqf-Haushi Uplift Oman Mountains
(rm)
52
353,576 167,060 155,530 126,474 43,927 38,650 21,680 21,195 15,047 6,617 6,176 5,493 3,508 2,146 1,728 1,715 1,343 1,093 926 798 762 597 495 181 105 89 53 53 41 38 30 22
11
am
=
0.03125
Vi rv
rt
at
977,148 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
32
function
(MMBOE) 978,595
0.94391 0.25218 0.11344 0.06308 0.03935 0.02638 0.01856 0.01352 0.01010 0.00769 0.00594 0.00464 0.00366 0.00290 0.00230 0.00183 0.00146 0.00116 0.00092 0.00072 0.00057 0.00044 0.00033 0.00025 0.00018 0.00013 0.00009 0.00005 0.00003 0.00001 0.00000 0.00000
0.94393 0.25220 0.11346 0.06310 0.03937 0.02639 0.01857 0.01353 0.01011 0.00771 0.00596 0.00466 0.00367 0.00291 0.00232 0.00185 0.00148 0.00118 0.00094 0.00074 0.00058 0.00045 0.00035 0.00027 0.00020 0.00014 0.00010 0.00007 0.00005 0.00003 0.00002 0.00002
0.00000 0.50318 0.50480 0.50172 0.49755 0.49297 0.48819 0.48329 0.47830 0.47323 0.46807 0.46282 0.45746 0.45199 0.44638 0.44062 0.43469 0.42856 0.42222 0.41562 0.40873 0.40152 0.39393 0.38590 0.37737 0.36826 0.35851 0.34809 0.33712 0.32619 0.31716 0.31435
9.99978E-01 373800.299 3.54828E-01 182118.284 5.81962E-05 149745.317 2.53622E-11 83289.762 2.36783E-18 51968.758 2.69359E-25 34838.870 6.17379E-32 24518.044 3.07740E-38 17862.977 3.13313E-44 13351.189 5.88055E-50 10172.862 1.82656E-55 7865.389 8.47618E-61 6149.221 5.34820E-66 4847.559 4.20708E-71 3844.305 3.80451E-76 3060.773 3.65978E-81 2442.181 3.46941E-86 1949.493 2.99893E-91 1554.327 2.17854E-96 1235.679 1.21741E-101 977.761 4.74034E-107 768.540 1.14776E-112 598.733 1.51101E-118 461.097 9.20312E-125 349.931 2.13051E-131 260.705 1.47681E-138 189.798 2.32081E-146 134.301 6.25953E-155 91.871 2.60385E-164 60.631 3.25992E-174 39.103 2.97949E-183 26.203 1.01707E-187 21.500
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. 1000
Cumulative Number of Provinces
Pareto Distribution
USGS (2000) data for 32 provinces = 977 BBOE
100
10
1 10
100
1000
10000
100000
1000000
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
min {Vx ,a p ,Vs , , S } (Vi V i ) 2 i 1
12
(3)
Version: March 19, 2009
Subject to:
1 log N x log N m ap ( ) log V log V x m Vm 1 Vm V x Vx N t V x
Vi
1 log N x log N m ( ) ap log V log V x m Vm 1 Vm ( ) [1 ( )] 1 exp V x Vs N V Vx t x
(4) S
where: ap - slope of straight line approximated from USGS sample points with larger province volumes (same as slope of Pareto distribution). Nm - minimum number of USGS provinces (= 1). Nt - cumulative number of provinces. Nx - 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). Vm - minimum USGS province volume (BOE). Vs - 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).
V i - estimated volume of a province (BOE).
13
Version: March 19, 2009
Vx - maximum volume (BOE) given by the Pareto straight line (at Nm = 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 - Vx, ap, Vs, 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 (R2). 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 (Vx) at Nm equal to 1 = 1,320,000 MMBOE Pareto shape exponent (ap) = 0.55 Volume of separation (Vs) = 84,800 MMBOE Separation ratio (ψ) = 0.30 Severity exponent (S) = 52
3.4 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
Version: March 19, 2009
Figure 6 VSD estimate for 32 MENA provinces. USGS (2000) data shows endowment volumes for 32 MENA provinces. 1000
Cumulative Number of Provinces
Pareto Distribution
USGS (2000) data for 32 provinces = 977 BBOE
100
VSD for 32 provinces = 979 BBOE (R2 = 0.99)
10
1 10
100
1000
10000
100000
1000000
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 R2 coefficient of determination equal to 0.99. The good fit provides an initial validation of the VSD model.
15
Version: March 19, 2009
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 (R2) 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, email
[email protected]).
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
Version: March 19, 2009
Table 2 Assessed and Previously Unassessed Provinces in the MENA region
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Region
Assessed Provinces
Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa
Mesopotamian Foredeep Basin Zagros Fold Belt Greater Ghawar Uplift Rub Al Khali Basin Sirte Basin Widyan Basin-Interior Platform Trias/Ghadames Basin Interior Homocline-Central Arch Red Sea Basin Illizi Basin Qatar Arch Fahud Salt Basin Pelagian Basin Ma'Rib-Al Jawf/Masila Basin Masila-Jeza Basin Ghaba Salt Basin Euphrates/Mardin Grand Erg/Ahnet Basin Ghudun-Khasfeh Flank Province Shabwah Basin Khleisha Uplift South Oman Salt Basin Central Oman Platform Anah Graben Sinai Basin Hamra Basin Mukalla Rift Basin Rutbah Uplift Ougarta Uplift Zagros Thrust Zone Huqf-Haushi Uplift Oman Mountains
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88
Region
Unassessed Provinces
Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa Middle East and North Africa
Palmyra Zone Mediterranean Basin Nile Delta Basin Lesser Caucasus Thrace/Samsun Jafr-Tabuk Basin Levantine Basin Haleb Atlas Uplift Rif Basin Essaouni Basin Arabian Shield Sharmah Rift Basin Adana/Sivas Tellian Foredeep South Harrah Volcanics Tellian Uplift Aaiun-Tarfaya Basin Reggane Basin Cyrenaica Uplift Beirut Wadi-Surhan Basin Yemen Volcanic Basin (North) Yemen Volcanic Basin (South) Hays Structural Belt Mukalla Rift Basin Masirah Trough Gulf of Oman Basin North Harrah Volcanics Syrian Arch Upper Egypt Basin Cyrenacia Basin Nubian Uplift Nefusa Uplift Hauts Basin Thiemboka Uplift Atlas Basin Rabat Basin Canary Islands Tindouf Basin Reguibate Uplift Guercif Basin North Red Sea Shield Araks Tuz/Corum Kardiff/Menders Massif Central Iranian Basins Central Iranian Microcontinents Lut Block and Depression Alborz Fold Belt Mirbat Precambrian Basement Fezzan Uplift Abu Gharadiq Basin Murzuk Basin East Flank Oman Sub-basin 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
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). 1000 Pareto Distribution
Cumulative Number of Provinces
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
10
1 10
100
1000
10000
100000
1000000
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
Version: March 19, 2009
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 Proved Reserves (BBOE) (Source: BP 2008)
Country
Iran Iraq Kuwait Oman Qatar Saudi Arabia Syria United Arab Emirates Yemen Other Middle East Algeria Egypt Libya Tunisia TOTAL
% of Total
Oil Endmt from VSD Model (BBOE)
Oil Endmt plus Reserve Growth (BBOE)
138.4 115.0 101.5 5.6 27.4 264.2 2.5 97.8 2.8 0.1 12.3 4.1 41.5 0.6
17.0 14.1 12.5 0.7 3.4 32.5 0.3 12.0 0.3 0.0 1.5 0.5 5.1 0.1
196.6 163.4 144.2 7.9 39.0 375.3 3.6 138.9 3.9 0.2 17.5 5.8 58.9 0.9
281.1 233.5 206.1 11.3 55.7 536.6 5.1 198.6 5.6 0.3 25.0 8.3 84.3 1.2
813.8
100.0
1156
1653
4. 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
Version: March 19, 2009
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 + 891 BBOE). Table 4 Calculation of Reserve Growth Percentages - based on data from USGS (2000) Mean Oil (Billion barrels) World (excluding USA) Undiscovered conventional Reserve growth (conventional) Remaining reserves Cumulative production Total
649 612 859 539 2659
Mean Gas (BBOE)
Mean NGL (Billion barrels)
Total Mean Petroleum (Billion barrels)
778 551 770 150 2249
207 42 68 7 324
1634 1205 1697 696 5232
Known volumes
1398
920
75
2393
Reserve growth based on known volumes (%)
43.78
59.89
56.00
50.36
Mean Gas (BBOE)
Mean NGL (Billion barrels)
Total Mean Petroleum (Billion barrels)
88 59 29 142 318
combined with oil combined with oil combined with oil combined with oil combined with oil
171 135 61 313 680
Mean Oil (Billion barrels) USA Undiscovered conventional Reserve growth (conventional) Remaining reserves Cumulative production Total Known volumes Reserve growth based on known volumes (%)
83 76 32 171 362 203
171
combined with oil
374
37.44
34.50
combined with oil
36.10
Mean Gas (BBOE)
Mean NGL (Billion barrels)
Total Mean Petroleum (Billion barrels)
866 610 799 292 2567
207 42 68 7 324
1805 1340 1758 1009 5912
Mean Oil (Billion barrels) World Total Undiscovered conventional Reserve growth (conventional) Remaining reserves Cumulative production Total
732 688 891 710 3021
Known volumes
1601
1091
75
2767
Reserve growth based on known volumes (%)
42.97
55.91
56.00
48.43
20
Version: March 19, 2009
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
Version: March 19, 2009
Figure 8 VSD estimate for 88 MENA provinces (including unassessed provinces and reserve growth). 1000 Pareto Distribution
Cumulative Number of Provinces
USGS (2000) data for 32 provinces = 977 BBOE
VSD for 32 provinces = 979 BBOE (R2 = 0.99)
VSD for 88 provinces = 1,156 BBOE
100
VSD for 88 provinces, plus Reserve Growth (42.97%) = 1,653 BBOE
10
1 10
100
1000
10000
100000
1000000
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
Version: March 19, 2009
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
Version: March 19, 2009
Figure 9 Cumulative Long-Run Supply Curves for MENA Conventional Oil
Average Total Production Cost (2008 US$/BOE)
14.00 without reserve growth = 1,156 BBOE
12.00
with reserve growth (42.97%) = 1,653 BBOE
10.00
8.00
6.00
4.00
2.00
0.00 0
200
400
600
800
1,000
1,200
1,400
1,600
Oil Endowment Volume (BBOE)
Table 5 Oil Endowment Volumes and Average Production Costs by Country 1
2
3
4
5
6
Location
Oil Endowment (BBOE)
Oil Endmt plus Future Reserve Growth (BBOE)
Average Average Average Total Capital Cost Operating Production Cost (2008US$/BOE) Cost (2008US$/BOE) (2008US$/BOE)
Oil Growth = 42.97%
Iraq
163.4
233.6
Saudi Arabia
375.3
536.6
Oman
7.9
11.3
Libya
58.9
84.2
Iran
196.6
281.1
Syria
3.6
5.1
Egypt
5.8
8.3
Tunisia
0.9
1.3
Kuwait
144.2
206.2
Yemen
3.9
5.6
Qatar
39.0
55.8
Algeria
17.5
25.0
United Arab Emirates
138.9
198.6
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 (2004-2006)
24
1.02 2.04 2.83 2.10 1.24 1.58 2.60 3.06 7.14 2.83 8.75 2.26 4.75 2.26 1.92 3.15 2.04 2.15
2.83 2.49 1.81 1.58 3.51 3.39 2.60 1.22 3.84 2.72 5.66 3.39 1.13 3.85 4.53 4.33 2.49 4.75
1.13 3.85 2.10 4.53 2.50 4.64 3.68 4.75 1.40 4.98 2.30 5.20 4.28 10.98 5.54 14.41 1.13 5.66 4.20 5.88 2.10 6.11 6.45 7.49 4.53 6.90
Version: March 19, 2009
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 1
2
MENA Conventional Oil
MENA Future Volumes (BOE)
From USGS (2000)
7.58E+11 9.37E+11 1.434E+12
Including Unassessed Provinces Including Unassessed Provinces and Reserve Growth
3
4 a
2005-2007 MENA Average Annual Production (BOE)
1.09E+10
5
Life Expectancy in Years, at Various Growth Rates in Production
b
0%
1%
2%
3%
70 86 132
53 62 84
44 50 64
38 43 53
MENA Average Annual Growth in Production, 1977-2007 (%)
0.73
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
c
Version: March 19, 2009
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
Version: March 19, 2009 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, 161-194. Barton, C.C. (1995), “A New Approach to Estimating Hydrocarbon Resources”, United States Geological Survey Fact Sheet, http://energy.usgs.gov/factsheets/HydroRes/estimat.html 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 2008. 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 96-0034. 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 (2004-2006), Upstream Asset Analysis Reports.
Version: March 19, 2009
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