Jean Laherrere 19 March Forecasts for US oil and gas production

Transcription

1 Jean Laherrere 9 March This paper was written in association with Charlie Hall in view of a presentation in New Orleans on March at ACS conference M. King Hubbert: Is He Relevant? Forecasts for US oil and gas production My assumptions for forecasting US oil and natural gas production are several: -Reserves estimates of LTO and shale gas, like conventional ones, are based on multiplying estimates of the volume in place (or the volume generated by the source rock) and a recovery factor. Such assessments are completely unreliable because the volume of the accumulation is fuzzy (no water plane) and the recovery factor unknown by lack of historical checks with LTO or shale gas fields produced until abandoned -Future production estimated by many (in particular EIA) by assuming a large number (tens of thousands) of future wells multiplied by the optimistic estimated EUR of past wells (in sweet spots) without bothering to check if there is enough room for drilling economical wells is for me very unreliable. EIA does not bother reporting the cumulative future production of their forecasts up to EIA publishes the LTO reserves at end ranked by proved reserves: Bakken is number one but in their AEO forecasts up to Permian basin is number one This EIA approach relying on number of wells and well ultimate reminds of the USGS estimate in 9 of 9 Gb for the US oil ultimate by V.McKelvey (based on Zapp s work by multiplying a large number of feet drilled ( E9 foot) by the recovery per foot op to 9 ( b/ft = Gb discovered for. Gft drilled = G.Bowden 9): Hubbert was strongly against such approach, saying that the oil recovery per foot was showing a roughly exponential decline (D.Strahan 7) and would be lower in the future. Present estimate of the US ultimate is far below Zapp s estimate (9 Gb), but higher than Hubbert s estimate (- Gb for conventional USL). It is the same with LTO. The present ultimate recovery per well (EUR) will be lower in the future and the location of future wells will be limited, in particular with lateral length increase. -I believe it is better to estimate the ultimate production of LTO and shale gas with HL (Hubbert linearization) of past production. But many plots display several declines and the last data with rise Hubbert Linearization is the plot of the ratio of annual production to cumulative production in percentage (Y axis) versus the cumulative production (Hubbert 9). Empirically once the field is moderately well developed, this gives one or several relatively straight lines and the last one can be extrapolated to the Y axis, giving the EUR when production will fall to zero (assumed to be the end of the production) -Past oil and gas data for the USLower displays cycles and most of the cycles display approximately symmetrical rises and declines, so the future cycle is assumed to be also symmetrical. This symmetry is explained by the fact that, in the USL, there are thousands of producers acting in random (law of large numbers). Random behavior is described as Brownian motion, displaying Gaussian symmetrical curve.

2 The past crude oil production 9-7 of the USL, Texas, California and Gulf of Mexico displays several cycles and each decline has the same slope as the last increase., USL crude oil + condensate annual production, Texas crude oil +condensate annual production, USL annual production Gb,,,, crude +cond annual production Gb,,9,, ap Gb,, Jean Laherrere Feb, Jean Laherrere February California annual crude oil production BOEM/MMS Gulf of Mexico annual production 97-7 crude oil prod DOGGR 9 total gas Gcf/. annual crude oil production Mb Mboe Gcf/. 7 oil +condensate Mb nfo/repcat/product/region.asp Jean Laherrere Feb Jean Laherrere Feb annual number of wells 9 7 US annual number of wells drilled all wells oil gas dry oil price $/b gas price $/MBtu 9 7 price $/b oil production b/d/w North Dakota: Bakken oil monthly production per well Antelope field 9 7 number of producing wells Jean Laherrere Feb Jean Laherrere March Our main forecasts are based on EIA monthly production data (unfortunately there are several different reports as weekly, monthly).

3 The following plots show for different plays: -left: annual & cumulative production as forecast for ultimate from HL -right: HL (Hubbert linearization) of past production extrapolated to ultimate (U) CP = cumulative production, RR = remaining reserves, CP+RR = initial reserves (discoveries) -Natural gas production -Shale gas: Bakken monthly production Gcf Bakken dry shale gas monthly production from EIA (FQ) Bakken FAQ U = Tcf U = Tcf CP FQ Bakken HL of dry shale gas production from EIA (FQ) jan-feb7 Jean Laherrere Feb Jean Laherrere Feb Mason Inman requested the EIA to get more details on the AEO and received the forecasts for each shale play for AEO to annual production Tcf,7,,,,, Bakken shale gas production from AEO & U = Tcf AEO AEO7 AEO AEO AEO AEO U= Tcf EIA FAQ Tcf Bakken cumulative shale gas production from AEO & U= Tcf AEO AEO7 AEO AEO AEO AEO U= Tcf hist FAQ,, Jean Laherrere March AEO annual production forecasts are chaotic with time! -Shale gas: Barnett Jean Laherrere March

4 7 Barnett Gcf/m Barnett dry shale gas monthly production from EIA Barnett HL of dry shale gas production from EIA monthly production Gcf 9 U = Tcf CP+RR CP Tcf U = Tcf jan-dec7 CP+RR =. Tcf =. Tcf =. Tcf 7 Jean Laherrere Feb 9 7 Jean Laherrere Feb From TxRRC annual data Barnett gas & liquids production from RRC U = Tcf HL gas Gcf/d RRC Barnett HL gas production RRC 99-7 gas ap/cp % - RRC Barnett gas decline -7 from RRC gas Gcf/d RRC -7 NG production Gcf/d RRC gas production Gcf/d Jean Laherrere March Jean Laherrere March Jean Laherrere March Mason Inman data AEO to and also from University of Texas UT Tcf per,,,,, Barnett shale gas production, 99 AEO AEO AEO AEO AEO7 AEO UT, AEO $ UT, AEO $ UT 7, $/Mcf hist (EIA) Tcf Jean Laherrere March AEO was much higher than AEO, missing the decline after the peak of,, UT 7 (University of Texas) is close to our forecast Barnett cumulative gas production from AEO & U= Tcf AEO AEO7 AEO AEO AEO AEO TxRRC U= Tcf hist AEO -Shale gas: Eagle Ford

5 Eagle Ford dry shale gas monthly production from EIA Eagle Ford HL of dry shale gas production from EIA monthly production Gcf Eagle Ford Gcf/m U = Tcf CP+RR CP Tcf U = Tcf dec-feb7 mar7-dec7 CP+RR =. Tcf =. TCF = 7. Tcf Jean Laherrere Feb Jean Laherrere Feb From TxRRC annual data, Eagle Ford NG production & forecast for U = Tcf Eagle Ford HL natural gas production from RRC, gas Tcf U = Tcf 9 annual production Tcf,, ap/cp % 7-7,, Jean Laherrere Feb Jean Laherrere Feb Mason Inman data AEO to, Eagle Ford shale gas annual production from AEO & U= Tcf Eagle Ford shale gas cumulative production from AEO & U= Tcf production Tcf,,,,, AEO AEO7 AEO AEO AEO AEO hist U = Tcf AEO AEO7 AEO AEO AEO AEO hist U = Tcf, Jean Laherrere March source: Mason Inman EIA Jean Laherrere March source: Mason Inman EIA -Shale gas: Fayetteville

6 9 U = Tcf Fayetteville Fayetteville dry shale gas monthly production from EIA Fayetteville HL of dry shale gas production from EIA monthly production Gcf 7 CP+RR EIA U = Tcf CP feb-dec7 CP+RR EIA end =. Tcf end =. Tcf end = 7. Tcf Jean Laherrere Feb 7 9 Jean Laherrere Feb Mason Inman data AEO to and UT 7 Tcf per,,,, Fayetteville, AEO was completely unconnected with reality, when University of Texas UT7 cumulative production up to is Tcf against Tcf for AEO or Tcf for my forecast: UT is closer to my estimate than to AEO! -Shale gas: Haynesville AEO AEO AEO AEO AEO7 AEO UT, AEO $ UT, AEO $ UT 7, $/Mcf hist (EIA) Fayetteville shale gas cumulative production from AEO & U = Tcf AEO AEO7 AEO AEO AEO AEO UT 7 $/Mcf U = Tcf hist (EIA) Jean Laherrere March monthly production Gcf 7 7 Haynesville dry shale gas monthly production from EIA U = 7 Tcf Haynesville CP+RR EIA U = 7 Tcf CP Tcf,,,, Haynesville HL of dry shale gas production from EIA oct-nov dec-dec7 CP+RR EIA end =. Tcf end =. Tcf end = 7. Tcf Jean Laherrere Feb, 9 7 Jean Laherrere Feb

7 Mason Inman data AEO to and UT 7 Haynesville Haynesville shale gas cumulative production from AEO - & UT7 AEO AEO7 AEO UT, AEO $ UT, AEO $ AEO Tcf per AEO AEO AEO AEO UT 7 $/Mcf 9 U = 7 Tcf hist (EIA) 7 UT 7, $/Mcf AEO AEO7 AEO AEO hist (EIA) Jean Laherrere March source: Mason Inman EIA -Shale gas: Marcellus My estimate for Marcellus gas is within the range - Tcf, when the cumulative production up to is forecasted by AEO to be over Tcf Marcellus HL of dry shale gas production from EIA Marcellus dry shale gas monthly production from EIA Marcellus Gcf/m U = Tcf prod (res) U = Tcf 9 U = Tcf CP+RR CP Tcf U = Tcf 7 Jun-Dec7 CP+RR EIA end = 9. Tcf end = 9. Tcf end = 97. Tcf monthly production Gcf Jean Laherrere March Jean Laherrere March 7 9 Marcellus basin is huge, but the sweet spots are concentrated in two areas and no drilling in between: its potential is likely to be poor! MCOR David Hughes 7

8 Mason Inman data AEO to & UT 7 Tcf per Marcellus AEO AEO AEO AEO AEO7 AEO UT, AEO $ Tcf Marcellus cumulative gas production from AEO & U= & Tcf AEO AEO7 AEO AEO AEO AEO U = Tcf U = Tcf hist AEO UT, AEO $ UT 7, $/Mcf hist (EIA) Jean Laherrere March The AEO forecast for Marcellus in is more than times the UT7 forecast -Shale gas: Utica monthly production Gcf Utica dry shale gas monthly production from EIA U = Tcf Utica CP+RR EIA U = Tcf CP 9 7 Utica HL of dry shale gas production from EIA CP+RR EIA end = Tcf end =.9 Tcf end = 7. Tcf jan7-dec7 Jean Laherrere Feb Jean Laherrere Feb -Shale gas: Woodford

9 9 U = Tcf Woodforfd Woodford dry shale gas monthly production from EIA 7 Woodford HL of dry shale gas production from EIA monthly production Gcf 7 CP+RR EIA U = Tcf CP 9 CP+RR EIA end =.7 Tcf end = Tcf end = Tcf sept-dec7 Jean Laherrere Feb Jean Laherrere Feb -Shale gas: all US U = Tcf US shale gas Gcf/d US dry shale gas monthly production from EIA ly/archivenew_ngwu/7/_/ 79 US HL of dry shale gas production from EIA jan-dec7 / CP+RR U = Tcf dry shale gas production Gcf/d cumulative Tcf EIA wet reserves EIA does not report any reserves for Permian and for Bakken 7 9 jan-feb7 Jean Laherrere Feb 7 7 Jean Laherrere Feb -Natural gas: US annual production Tcf U = Tcf marketed dry AEO AEO7 AEO AEO AEO U = Tcf cum prod mark EIA reserves US natural gas production & forecasts cumulative production & reservestcf HL of US marketed natural gas production 9-7 mark 9- AEO extrapolated towards = Tcf Jean Laherrere Feb Jean Laherrere %arch Summary for US shale gas plays 9

10 gas play Tcf ultimate cum prod 7 remaining 7 Barnett 7 Eagle Ford 9 Fayetteville Haynesville 7 Marcellus 9 Utica,, Woodford shale gas natural gas -Other forecasts EIA/AEO forecasts from 979 to display a huge range of uncertainty or poor practice (often change of the person doing the estimate): it means that the last forecast is likely to be poor too! dry production Tcf dry gas production Tcf US natural gas dry production forecasts from EIA/AEO 979- AEO 979 AEO 9 AEO 9 AEO9 AEO99 AEO99 AEO99 AEO997 AEO99 AEO999 AEO AEO AEO AEO AEO AEO AEO AEO7 AEO AEO9 AEO AEO AEO AEO AEO AEO AEO AEO AEO Jean Laherrere Feb actual dry US natural gas dry production forecasts from USDOE/EIA/AEO Jean Laherrere March? AEO 979 AEO 9 AEO 9 AEO9 AEO99 AEO99 AEO99 AEO99 AEO99 AEO997 AEO99 AEO999 AEO AEO AEO AEO AEO AEO AEO AEO7 AEO AEO9 AEO AEO AEO AEO AEO AEO AEO AEO7 AEO actual dry dry production Tcf for for for for for US natural gas dry production forecasts from EIA/AEO Jean Laherrere Feb of forecast cumulative dry gas production Tcf US cumulative dry gas production forecasts from USDOE/EIA/AEO AEO Jean Laherrere March? AEO 9 AEO 9 AEO9 AEO99 AEO99 AEO99 AEO99 AEO99 AEO997 AEO99 AEO999 AEO AEO AEO AEO AEO AEO AEO AEO7 AEO AEO9 AEO AEO AEO AEO AEO AEO AEO AEO7 AEO cum actual cum + P

11 cumulative dry gas production Tcf 7 9 US cumulative dry gas production forecasts from USDOE/EIA/AEO for for for for for ultimates JL mark Jean Laherrere March of forecast cumulative dry gas production Tcf US cumulative dry gas production forecasts from AEO & and JL ultimates AEO AEO JL ultimate JL ultimate cum actual Jean Laherrere March? My own estimates for the ultimates of total US NG marketed has increased from Tcf in to Tcf in In contrast, the official EIA/AEO forecast is for dry gas production, which is about 7% less than marketed gas. Our forecast U= Tcf is for marketed gas. The difference of forecasts between AEO dry gas and U= Tcf/,7 is Tcf for, Tcf in and Tcf in ; But AEO NG consumption is well below AEO production by about 7 Tcf, but exceeds my forecast after and in Tcf is missing. The plot of AEO US NG consumption shows that US in, instead of exporting Tcf, will be forced to import Tcf: quite a change compared to the official statements US natural gas production, consumption & forecasts AEO & U= Tcf annual production Tcf U = Tcf marketed AEO dry consumption AEO AEO-U= Tcf/, Jean Laherrere Feb Europe counts on the US shale gas to import LNG, but for that to occur the US must be able to produce more than they consume, which is unlikely All other analysts beside the US EIA give results that disagree with the EIA estimates, specifically: David Hughes displays US NG production - from EIA/AEO 7 (but not his own forecast) and Art Berman displays US NG monthly production -7 (production starts at Gcf/d)

12 DNV GL (Det Norske Veritas Germanischer Lloyd in Norway is assumed to tell the truth!) report Oil and gas forecast to - Energy Transition outlook 7 displays a peak of unconventional onshore gas production in for North America at G.m = Tcf DNV forecasts that North America unconventional onshore gas in will be 7% of value (going down), against % (going up) for AEO forecasts for the US unconventional gas (only onshore) as shown in the above graph. It means that DNV disagrees with the EIA forecasts for US unconventional gas. However DNV adds: unconventional gas will be the primary source for North American LNG exports. ExxonMobil outlook forecasts that the North America unconventional gas should be in % of its value, in agreement with AEO -Gas price, oil/gas price and flaring Gas price has sharply peaked in and and today back to pre price despite that the heat content has increased.

13 US natural gas price: wellhead, citygate & Henry Hub from EIA & heat content, 7 US oil over gas price MBtu ratio from EIA and flared over marketed percentage 9, annual oil/gas ratio oil/gas ratio AEO NG price $/kcf & $/MBtu 7 wellhead $/MBtu citygate $/kcf Henry Hub spot $/Mbtu heat content mark MBtu/kcf heat content dry MBtu/kcf,,,,,, heat content dry gas MBtu/kcf oil/gas ratio & flared/marketed % oil/gas ratio AEO flared/marketed % US flared/sold % ND flared/sold % North Dakota Jean Laherrere March Jean Laherrere March If in Europe some gas contracts are indexed to oil price. The US ratio of oil to gas price per Joule (MBtu) varies largely being over in 9 decreasing slowly to in, increasing to in and is presently around. EIA forecast that it will increase but stay below in. The problem is that the cost of transporting gas internationally is much higher than transporting oil and when gas is cheap and in excess associated to oil it is flared. The oil over gas price correlates with flared over marketed percentage in the US and since in North Dakota. I am surprised that the 9- trend for equality between oil and gas price is not the goal of EIA. I doubt that the future increase of this ratio will occur.

14 Oil production -Tight oil: Bakken North Dakota & Montana Mb/d,,,,,9,,7,,,,,, US Bakken monthly production from EIA play & forecast U=. Gb Bakken (ND & MT) U =. Gb CP+RR EIA P U =, Gb cum prod Mb, Jean Laherrere Feb 9 7 cumulative production Mb US HL of Bakken LTO monthly production from EIA play 9-Sept7 CP = Mb -7 Jan-Feb7 WTI Jean Laherrere Feb cumulative production Mb WTI $/b Bakken ND from ND.gov Ultimate Gb oil production kb/d 9 7 North Dakota: Bakken oil monthly production U = Gb Jean Laherrere March From Mason Inman: AEO to, and UT7 forecast kb/d ap/cp % Jean Laherrere March Bakken North Dakota monthly oil production Hubbert linearization cumulative production Mb 9- Nov-Dec Jan -Feb 7 Feb7-Dec7 WTI $/b 7 WTI $/b

15 Mb/d,,,,, Bakken oil production Gb US Bakken annual cumulative production from EIA AEO & U=. Gb AEO AEO7 AEO AEO AEO AEO U=. Gb past EIA/FAQ,, 99 AEO Bakken AEO Bakken AEO Bakken AEO Bakken AEO7 Bakken AEO Bakken UT7 ($/bbl case) Bakken UT7 ($/bbl case) Bakken UT7 (AEO7 price case) Bakken historical (EIA) Bakken Jean Laherrere March -Tight oil: Eagle Ford Mb/d,,,,,,,, US Eagle Ford monthly production & forecast U = Gb Eagle Ford play U = Gb CP+RR EIA P U = Gb CP Mb 7 cumulative production Mb 7 US HL of Eagle Ford LTO monthly production from EIA play Oct-Nov WTI WTI $/b,, Jean Laherrere Feb Jean Laherrere Feb cumulative production Mb From TxRRC annual production Eagle Ford RRC oil & condensate production & forecast for U = Gb U = Gb oil + cond Mb oil condensate annual production Mboe, Jean Laherrere March Mb/d ap/cp % 9 7 Jean Laherrere Jan Eagle Ford HL RRC crude oil +condensate production - -7 WTI $/b cumulative production Mb 9 7 WTI $/b From Mason Inman data: AEO to

16 Mb/d Eagle Ford oil production,,,,,,,,,, 99 AEO Eagle Ford AEO Eagle Ford AEO Eagle Ford AEO Eagle Ford AEO7 Eagle Ford AEO Eagle Ford Gb US Eagle Ford cumulative production fromaeo & forecast U = Gb AEO AEO7 AEO AEO AEO AEO U = Gb past Jean Laherrere March source Mason Inman -Tight oil: Permian Basin Mb/d,,,,,,,, US LTO Permian monthly oil production for U = Gb Permian play real LTO? U = Gb CP+RR EIA P U = Gb CP Mb source: EIA/FAQ 9 7 cumulative production Mb kb/d HL of Permian LTO monthly oil production - sept 7 March-Feb7 WTI source: EIA/FAQ 9 7 WTI $/b,,, Jean Laherrere Feb Jean Laherrere Feb cumulative production Mb All Permian basin annual production displays a decline since the peak of 97 which is constant at %/a until before the burst of LTO, this % decline is extrapolated down to 7 and the difference with all oil production is assumed to be the LTO, in agreement with EIA values, except for the period - where the EIA confuses LTO and horizontal wells annual production Mb 9 7 Permian Basin oil production & forecasts from ultimates decline% + U= Gb U = Gb Tx+NM = conv decline % U LTO = Gb LTO =all-decline% LTO EIA/FAQ all PB EIA,,,9,,,, Mb/d Permian Basin LTO oil production Hubbert linearization -7 -,,, Jean Laherrere Feb source TxRRC & NMemnrd 7 9 Jean Laherrere Feb cumulative oil production Gb

17 From Mason Inman data: AEO to Mb/d,,,,,,, Permian oil production Gb Permian cumulative oil production from AEO & U = Gb AEO AEO7 AEO AEO AEO AEO U = Gb past,,, 99 AEO Permian Basin AEO Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] AEO Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] AEO Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] AEO7 Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] AEO Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] Jean Laherrere March -Tight oil: all less Permian, Bakken & Eagle Ford Mb/d,7,,,,, all LTO-Permian-Bakken-Eagle Ford monthly production LTO-Permian-Bakken-Eagle U = Gb real LTO? CP+RR EIA P U = Gb CP Mb cumulative production Mb US HL of all LTO-Permian-Bakken-Eagle Ford monthly production Aug-Jul7 WTI,, Jean Laherrere Feb Jean Laherrere Feb cumulative production Mb - Tight oil: all US Mb/d US LTO monthly production from EIA play & forecast U= Gb all plays U = Gb CP Mb EIA P CP+RR U = Gb cumulative production Mb US HL of LTO production from EIA play 9-Dec7 -Sept7 Dec-Feb7 WTI ultimate in Gb Permian = Bakken =. Eagle Ford = rest = total =. WTI $/b Jean Laherrere Feb CP = Mb Jean Laherrere Feb cumulative production Mb 7

18 From Mason Inman data: AEO to Total LTO production 9, US LTO cumulative production from AEO & U= Gb, 7,,,,,, Gb 9 7 AEO AEO7 AEO AEO AEO AEO U = Gb past,, 99 AEO Total AEO Total AEO Total AEO Total AEO7 Total AEO Total Jean Laherrere March -Crude oil: USL without Alaska, USL crude oil (incl condensate) oil production 9 US Lower HL of oil production 9-7 annual production Gb,,,,, USL USL less LTO U = Gb U = Gb U = Gb U less LTO = Gb cum prod + reserves cum prod cum USL less LTO cumulative production Gb annual/cumulative % before LTO 9-97 at Hubbert's 9 paper time,, Jean Laherrere Feb Jean Laherrere Feb cumulative production Gb -Crude oil: US annual production crude +condensate Gb,,,,,,,,,, US crude oil + condensate production from EIA & forecasts from U= Gb & AEO AEO AEO7 AEO AEO AEO U= Gb past cum + P U = Gb cum prod, Jean Laherrere Feb 7 9 cumulative production & reserves Gb annual/cumulative % 9 7 US HL of crude oil production AEO extrapolated towards = Gb Jean Laherrere March cumulative production Gb

19 The difference between AEO and my forecast for U = Gb is over Gb in and. Gb in. AEO crude oil consumption is forecasted flat around. Gb/a from to ; But in percentage of AEO my forecast is % in but only. % in annual production crude +condensate Gb,,,,,,,,,,,,, US crude oil production & consumption AEO & U= Gb AEO cons AEO prod U= Gb AEO-UGb past % UGb/AEO 9 7 % UGb/AEO, Jean Laherrere March Summary for the oil plays oil play Gb ultimate cum prod 7 remaining 7 Bakken,,, Eagle Ford,, Permian,, LTO, 9, USL US -EIA/AEO forecasts evolution EIA/AEO forecasts from 979 to display a huge and chaotic range of uncertainty: it means that the last forecast is likely to be poor too! For example AEO forecasted Mb/d for the US in when over 9 Mb/d was reached in! crude oil production Mb/d 9 7 US crude oil production forecasts from USDOE/EIA/AEO Jean Laherrere March? AEO 979 AEO 9 AEO 9 AEO 9 AEO 99 AEO 99 AEO 99 AEO 99 AEO 997 AEO 99 AEO 999 AEO AEO AEO AEO AEO AEO AEO AEO 7 AEO AEO 9 AEO AEO AEO AEO AEO AEO AEO AEO 7 AEO actual crude oil production Mb/d 9 7 US crude oil production forecasts from USDOE/EIA/AEO for for for for for Jean Laherrere Feb of forecast 9

20 cumulative crude oil production Gb US cumulative crude oil production forecasts from USDOE/EIA/AEO AEO Jean Laherrere March? AEO 9 AEO 9 AEO 9 AEO 99 AEO 99 AEO 99 AEO 99 AEO 997 AEO 99 AEO 999 AEO AEO AEO AEO AEO AEO AEO AEO 7 AEO AEO 9 AEO AEO AEO AEO AEO AEO AEO AEO 7 AEO cum actual cum + P cumulative crude oil production Gb Jean Laherrere Feb of forecast Rystad (a Norwegian data seller) has published a paper on world reserves in. Laherrere J.H. «World, US, Saudi Arabia, Russia & UK oil production & reserves -Comments on Rystad world reserves» August Bowden reported in 9 the US crude oil resources (ultimate?) showing an evolution with time: ultimates less than Gb from 9 to 9 (Hubbert in 9 was using & Gb for USL), between and Gb (with Zapp s estimate in 9 = 9 Gb from unrealistic recovery per feet drilled) from 9 to 97 and between and Gb from 97 to 9: the range was wild and it is still! US cumulative crude oil production forecasts from USDOE/EIA/AEO for for for for for The future US crude oil (+condensate) production is modeled with the likely ultimate of Gb and the unlikely Rystad Gb and the EIA/AEO forecasts from to. It appears that Rystad ultimate of Gb is related to AEO forecast of more than Mb/d in. There is a wild change in LTO forecasts between AEO and AEO

21 -Other forecasts David Hughes displays US crude oil production from AEO peaking in, because of steady tight oil but not his own forecast. As for unconventional gas, DNV forecasts unconventional onshore oil production for North America, but if for gas Canada production is small compared with the US, it is not the same for oil. CAPP forecasts the oilsands to increase by Mb/d from to for Canada, when AEO forecasts an increase of Mb/d for the US, when DNV forecasts an increase of Mb/d for North America: it means that DNV is more optimistic than EIA ExxonMobil Outlook forecasts that tight oil in will be times the value of

22 -My forecasts If the performance of the EIA is not good in forecasting US crude oil production, what about my forecasting? My modeling of the future is based on the estimate of the ultimates My estimate of the US crude oil ultimate has changed from to Gb between and, when the extrapolation of AEO cumulative production to (assumed close to ultimate) is Gb and AEO Gb. My change is less than EIA US all USL site forecast ultimate Gb site forecast ultimate Gb Uppsala Enc Energy Enc Energy ENSMP Lisbon MIT Beijing Luxembourg comrystad US cumulative crude oil production forecasts from USDOE/EIA/AEO US cumulative crude oil production forecasts from USDOE/EIA/AEO & and JL ultimate cumulative crude oil production Gb for for for for for ultimates JL cumulative crude oil production Gb AEO AEO JL ultimate JL ultimate cum actual? Jean Laherrere Feb of forecast Jean Laherrere March -Oil price and dollar value Oil price influences the US oil production, it is then important to know how the oil price changes. The best correlation for the WTI is the dollar value as I presented in many papers WTI oil price $/b 9 7 WTI $/b US WTI crude oil daily price and dollar value * - Dollar Index (major currencies)*- mary/indexn9_b.htm hx?n=pet&s=rwtc&f=d - // // // // // Jean Laherrere March date nominal major currencies dollar index *- WTI oil price $/b 7 US WTI crude oil daily price and dollar value * - shifted days h/summary/indexn9_b.htm es/dcoilwtico/downloaddata WTI $/b dollar shift days -97 Jan- Jan- Jan-7 Jan- Jan-9 Jean Laherrere March date nominal major currencies dollar index *- shifted days -US private crude oil stocks and WTI months before

23 Some believe that the oil price follows the private stocks of crude oil, but since it appears that the stocks follows the WTI weeks before, except few months in 7 after the OPEC Russia deal for reducing production US weekly private stocks crude oil & WTI *- shifted weeks - US weekly private stocks crude oil & WTI *- shifted weeks US private stocks crude oil Mb Weekly U.S. Ending Stocks excluding SPR of Crude Oil "excluding lease stock" Mb WTI *- weeks before Jean Laherrere March WTI *- weeks before US private stocks crude oil Mb Weekly U.S. Ending Stocks excluding SPR of Crude Oil "excluding lease stock" Mb WTI *- weeks before Jean Laherrere March WTI *- weeks before Conclusions The USDOE/EIA believes that the US oil and gas production will be higher in (and in every in between) than in 7, based on the growth of shale plays coming from the drilling of a huge number of wells. They do so without bothering to check if there is enough room in the sweet spots to do so or if the yield in the non sweet spots is enough to generate that much oil ; we believe the contrary, based on the estimate of the ultimate of shale plays from the Hubbert linearization of past production and on the assumption that in the USL many cycles of past production and drilling were symmetrical that the shale plays will also display symmetrical future production curves. AEO forecasts US production in for oil at Mb/d, my forecast is times less at Mb/d, for gas at Tcf, my forecast is half at Tcf. It is not a small difference, but a huge one. I can be wrong, but EIA has to prove that their future drilling is possible economically and geologically, but up to now I cannot find any view from EIA on this problem. The EIA has also to estimate the ultimate of each play, but, in order to do so, they have to reject the stupid rule of the SEC (to please the bankers) forbidding the reporting of P (proved + probable) reserves. The EIA has to recognize the fact reported by the SPE/PRMS that the aggregation of proved (P) reserves is incorrect, only the addition of P is correct. (Laherrère J.H. «Advice from an old geologist-geophysicist on how to understand Nature» presentation Statoil Oslo August USDOE/EIA should report the P US reserves like the USDOI/BOEM for the reserves of the Gulf of Mexico. Reporting P and backdating discovery allows to estimate ultimate using creaming curve. EIA once reported as an open file backdated established reserves with USDOE/EIA- 99 "US oil and gas reserves by of field discovery» Aug: it was supposed to be the first of a series, but in fact later it was censured and never updated. It is time to restart such good practice. Exxon and Total s ago were reporting proved and non-proved reserves (P), but today they report only P to follow the SEC rule. Today only Gazprom is reporting P and P, as also their queer ABC.

24 All scouting agencies report the confidential P reserves used internally by the operators to decide the development of their fields.