This paper was written in association with Charlie Hall in view of a presentation in New Orleans
on 21 March at ACS conference “M. King Hubbert: Is He Relevant?”
My assumptions for forecasting US oil and natural gas production are several:
1-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
2-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 2050
EIA publishes the LTO reserves at end 2016 ranked by proved reserves: Bakken is number one
but in their AEO forecasts up to 2050 Permian basin is number one
This EIA approach relying on number of wells and well ultimate reminds of the USGS estimate
in 1962 of 590 Gb for the US oil ultimate by V.McKelvey (based on Zapp’s work
https://pubs.usgs.gov/bul/1142h/report.pdf) by multiplying a large number of feet drilled (5 10E9
foot) by the recovery per foot op to 1961 (118 b/ft = 130 Gb discovered for 1.1 Gft drilled =
G.Bowden 1982): Hubbert was strongly against such approach, saying that the oil recovery per
foot was showing a roughly exponential decline (D.Strahan 2007) and would be lower in the
future. Present estimate of the US ultimate is far below Zapp’s estimate (590 Gb), but higher
than Hubbert’s estimate (150-200 Gb for conventional USL48).
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.
3-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 1980). 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)
4-Past oil and gas data for the USLower48 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 USL48, there are thousands of producers
acting in random (law of large numbers). Random behavior is described as “Brownian motion”,
displaying Gaussian symmetrical curve.
1
The past crude oil production 1859-2017 of the USL48, Texas, California and Gulf of Mexico
displays several cycles and each decline has the same slope as the last increase.
USL48 crude oil + condensate annual production Texas crude oil +condensate annual production
3,5
1,5
3,0 USL48
aP Gb
1,2
0,9
2,0
1,5
0,6
1,0
0,3
0,5
0,0
0,0 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020
1860 1880 1900 1920 1940 1960 1980 2000 2020 year
Jean Laherrere Feb 2018 year Jean Laherrere February 2018
700
300
Mboe Gcf/5.6
600
250
500
200
400
150
300
100
200
http://www.data.boem.gov/homepg/pubi
nfo/repcat/product/Region.asp
50 100
0 0
1875 1900 1925 1950 1975 2000 2025 2050 2075 1940 1950 1960 1970 1980 1990 2000 2010 2020
Jean Laherrere Feb 2018 year year
Jean Laherrere Feb 2018
US annual number of wells drilled North Dakota: Bakken oil monthly production per well
100 000 120 200 12000
all wells
90 000 108 180 10800
oil
80 000 gas 96 160 9600
dry
annual number of wells
40 000 48 80 4800
30 000 36 60 3600
20 000 24 40 2400
10 000 12 20 1200
0 0 0 0
1860 1880 1900 1920 1940 1960 1980 2000 2020 1950 1960 1970 1980 1990 2000 2010 2020
Jean Laherrere Feb 2018 year Jean Laherrere March 2018 year
Our main forecasts are based on EIA monthly production data (unfortunately there are several
different reports as weekly, monthly).
2
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
Bakken dry shale gas monthly production from EIA (FQ) Bakken HL of dry shale gas production from EIA (FQ)
4
50 6
Bakken FAQ
45
aP/CP%
U = 6 Tcf 5
40 jan2015-feb2017
U = 6 Tcf 3
monthly production Gcf
35 CP FQ
4
aP/CP%
25 3 2
20
2
15
1
10
1
5
0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 0 1 2 3 4 5 6
Jean Laherrere Feb 2018 Jean Laherrere Feb 2018 cumulative production Tcf
year
Mason Inman requested the EIA to get more details on the AEO and received the forecasts for
each shale play for AEO 2013 to 2018
Bakken shale gas production from AEO & U = 6 Tcf Bakken cumulative shale gas production from AEO & U=6 Tcf
25
0,7
AEO2018
AEO2018
AEO2017
AEO2017
0,6 AEO2016
AEO2016 20
AEO2015
AEO2015
AEO2014
annual production Tcf
0,5 AEO2014
AEO2013
AEO2013
U=6 Tcf
U=6 Tcf 15
0,4 hist FAQ
EIA FAQ
Tcf
0,3 10
0,2
5
0,1
0,0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Jean Laherrere March year Jean Laherrere March year
2018 2018
AEO annual production forecasts are chaotic with time!
3
Barnett dry shale gas monthly production from EIA Barnett HL of dry shale gas production from EIA
4
170 50
Barnett Gcf/m
153 U = 25 Tcf 45
aP/CP%
CP+RR
136 40 jan2010-dec2017
CP Tcf 3
monthly production Gcf
U = 25 Tcf
119 35
aP/CP%
85 25 2
68 20
51 15
1 CP+RR
2016 = 33.8 Tcf
34 10 2015 = 32.8 Tcf
2014 = 38.8 Tcf
17 5
0 0 0
2000 2005 2010 2015 2020 2025 2030 0 3 6 9 12 15 18 21 24 27 30
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Tcf
21
aP/CP%
3 18 3
15
2 12
2
9
6
1 1
3
0
0 0 5 10 15 20 25 30 0
1990 1995 2000 2005 2010 2015 2020 2025 2030 0 5 10 15 20 25 30 35
year cumulative production Tcf
Jean Laherrere March 2018 Jean Laherrere March 2018 Jean Laherrere March 2018 cumulative production Tcf
Mason Inman data AEO 2013 to 2018 and also from University of Texas UT
Barnett shale gas production Barnett cumulative gas production from AEO & U= 25 Tcf
60
2,5 AEO2018
55 AEO2017
AEO2016
50
AEO2015
2,0 AEO2013 45 AEO2014
AEO2014 AEO2013
40
AEO2015 TxRRC
Tcf
0,5 10
0
0,0 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
1990 2000 2010 2020 2030 2040 2050 Jean Laherrere March 2018 year
AEO2013 was much higher than AEO2018, missing the decline after the peak of 2012, , UT
2017 (University of Texas) is close to our forecast
4
Eagle Ford HL of dry shale gas production from EIA
Eagle Ford dry shale gas monthly production from EIA
5
160 32
CP Tcf
U = 15 Tcf
aP/CP%
80 16
2
60 12
CP+RR
40 8 2016 = 30.1 Tcf
2015 = 25.4 TCF
1 2014 = 27.8 Tcf
20 4
0 0 0
2000 2005 2010 2015 2020 2025 2030 0 2 4 6 8 10 12 14 16 18 20
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Tcf
gas Tcf 90
U = 15 Tcf
2,0 80 aP/CP%
2013-2017
70
annual production Tcf
1,5 60
aP/CP %
50
1,0 40
30
0,5 20
10
0
0,0 0 2 4 6 8 10 12 14 16
2005 2010 2015 2020 2025
Jean Laherrere Feb 2018 http://www.rrc.state.tx.us/eagleford/index.php Jean Laherrere Feb 2018 cumulative production Tcf
year
AEO2018 55 AEO2018
AEO2017 AEO2017
2,5 50
AEO2016 AEO2016
AEO2015 45 AEO2015
AEO2014 AEO2014
2,0 40
cumulative production Tcf
production Tcf
AEO2013 AEO2013
hist 35 hist
U = 15 Tcf U = 15 Tcf
1,5 30
25
1,0 20
15
0,5 10
0,0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Jean Laherrere March 2018 year source: Mason Inman EIA Jean Laherrere March 2018 year source: Mason Inman EIA
5
Fayetteville dry shale gas monthly production from EIA Fayetteville HL of dry shale gas production from EIA
100 20 4
U = 10 Tcf
90 Fayetteville 18
70 CP 14 feb2013-dec2017
aP/CP%
50 10 2
40 8
CP+RR EIA
end 2016 = 13.5 Tcf
30 6 end 2015 = 13.6 Tcf
1 end 2014 = 17.3 Tcf
20 4
10 2
0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 0 1 2 3 4 5 6 7 8 9 10 11 12
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Tcf
1,5
AEO2013
AEO2016 30 UT 2017 $3/Mcf
U = 10 Tcf
AEO2017
Tcf per year
25 hist (EIA)
1,0 AEO2018
20
UT 2014,
AEO2014 $ 15
UT 2016,
0,5 AEO2014 $
10
UT 2017,
$3/Mcf
5
hist (EIA)
0,0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
2000 2010 2020 2030 2040 2050 Jean Laherrere March 2018 year
AEO2014 was completely unconnected with reality, when University of Texas UT2017
cumulative production up to 2050 is 18 Tcf against 42 Tcf for AEO2018 or 10 Tcf for my
forecast: UT is closer to my estimate than to AEO2018!
Haynesville dry shale gas monthly production from EIA Haynesville HL of dry shale gas production from EIA
250 40 2,0
U = 27 Tcf aP/CP%
225 36 oct2014-nov2016
Haynesville
dec2016-dec2017
CP+RR EIA
200 32
U = 27 Tcf 1,5
monthly production Gcf
175 CP Tcf 28
CP+RR EIA
cumulative production Tcf
1,0
125 20
100 16
75 12
0,5
50 8
25 4
0,0
0 0 0 3 6 9 12 15 18 21 24 27 30
2000 2005 2010 2015 2020 2025 2030 2035
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Tcf
6
Mason Inman data AEO 2013 to 2018 and UT 2017
Haynesville Haynesville shale gas cumulative production from AEO 2013-2018 & UT2017
130
5
AEO2018
120
AEO2013 AEO2017
110 AEO2016
AEO2014 AEO2015
4 100
AEO2014
2 UT 2014, 50
AEO2014 $
UT 2016, 40
AEO2014 $
UT 2017,
30
1 $3/Mcf
20
hist (EIA)
10
0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
2000 2010 2020 2030 2040 2050 Jean Laherrere March 2018 year source: Mason Inman EIA
U = 80 Tcf 80
450 CP+RR
cumulative production Tcf
CP Tcf 70
400 3
U = 100 Tcf
aP/CP%
CP+RR EIA
350 60 end 2016 = 109.1 Tcf
end 2015 = 91.4 Tcf
300 50 end 2014 = 97.4 Tcf
2
250
40
200
30
150
1
20
100
50 10
0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 0 10 20 30 40 50 60 70 80 90 100
Jean Laherrere March 2018 year Jean Laherrere March 2018 cumulative production Tcf
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 2018
7
Mason Inman data AEO 2013 to 2018 & UT 2017
Marcellus cumulative gas production from AEO & U= 80 & 100 Tcf
Marcellus
400
14 AEO2018
AEO2017
350
AEO2016
12 AEO2013
AEO2015
AEO2014 300 AEO2014
10 AEO2013
AEO2015
U = 100 Tcf
250
Tcf
AEO2016 U = 80 Tcf
8
Tcf per year
hist AEO
AEO2017
200
AEO2018
6
UT 2014, AEO2014 $ 150
UT 2016, AEO2014 $
4 100
UT 2017, $3/Mcf
hist (EIA)
2 50
0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
2000 2010 2020 2030 2040 2050 Jean Laherrere March 2018 year
The AEO2018 forecast for Marcellus in 2045 is more than 5 times the UT2017 forecast
Utica dry shale gas monthly production from EIA Utica HL of dry shale gas production from EIA
160 20 14
U = 10 Tcf 13
Utica 18
140 12
CP+RR EIA aP/CP%
16 11
U = 12 Tcf
monthly production Gcf
120
10
CP 14 jan2017-dec2017
cumulative production Tcf
9 CP+RR EIA
100 end 2016 = 18 Tcf
12 end 2015 = 13.9 Tcf
8
aP/CP%
6
8
60 5
6 4
40
3
4
2
20
2
1
0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 0 2 4 6 8 10 12 14 16 18 20
Jean Laherrere Feb year Jean Laherrere Feb 2018 cumulative production Tcf
2018
8
Woodford dry shale gas monthly production from EIA Woodford HL of dry shale gas production from EIA
100 30 4
U = 18 Tcf
90 Woodforfd 27
70 CP 21 sept2011-dec2017
aP/CP%
end 2014 = 20 Tcf
50 15 2
40 12
30 9
1
20 6
10 3
0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 0 5 10 15 20
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Tcf
cumulative Tcf
35 217 jan2013-feb2017
EIA wet reserves
2
cumulative production Tcf
30 186
aP/CP%
20 124
1
15 93
10 62
5 31
0 0 0
2000 2005 2010 2015 2020 2025 2030 2035
0 25 50 75 100 125 150 175 200 225 250
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Tcf
-Natural gas: US
US natural gas production & forecasts HL of US marketed natural gas production 1900-2017
45 2400
6
2200 mark aP/CP%
40 U = 2400 Tcf
1983-2006
marketed 2000 5
dry
35
cumulative production & reservesTcf
AEO2018 1800
AEO2017
annual production Tcf
30 1600 4
AEO2016
AEO2013 1400
25
aP/CP%
AEO2010
U = 2400 Tcf 1200 3
20 cum prod mark
1000
EIA reserves
15 800 2
0 0 0
1900 1920 1940 1960 1980 2000 2020 2040 2060 2080 2100 0 400 800 1200 1600 2000 2400 2800 3200 3600 4000
Jean Laherrere 20 Feb 2018 year Jean Laherrere 16 %arch2018 cumulative production Tcf
9
gas play Tcf ultimate cum prod 2017 remaining 2017
Barnett 25 18 7
Eagle Ford 15 9 6
Fayetteville 10 8 2
Haynesville 27 15 12
Marcellus 80 31 49
Utica 10 4,5 5,5
Woodford 18 6 12
shale gas 250 110 140
natural gas 2400 1400 1000
-Other forecasts
EIA/AEO forecasts from 1979 to 2018 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!
US natural gas dry production forecasts from EIA/AEO 1979-2018 AEO 1979 US natural gas dry production forecasts from EIA/AEO
45 AEO 1982 45
AEO 1983
AEO1985 for 2050
40 AEO1990 40 for 2040
AEO1995 for 2030
AEO1996 for 2020
35 AEO1997 35 for 2010
AEO1998
AEO1999
30 AEO2000 30
dry production Tcf
AEO2001
AEO2002
25 AEO2003 25
AEO2004
AEO2005
20 AEO2006 20
AEO2007
AEO2008
15 AEO2009 15
AEO2010
AEO2011
10 AEO2012
10
AEO2013
AEO2014
5 5
AEO2015
AEO2016
0 AEO2017 0
1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 AEO2018 1990 1995 2000 2005 2010 2015 2020
actual dry
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 year of forecast
AEO 1979
US natural gas dry production forecasts from USDOE/EIA/AEO AEO 1982 US cumulative dry gas production forecasts from USDOE/EIA/AEO AEO 1979
50
AEO 1983 4000 AEO 1982
AEO1985 AEO 1983
AEO1990
3750 AEO1985
45
? AEO1993
3500
?
AEO1990
AEO1994 AEO1993
40 AEO1995 3250 AEO1994
AEO1996 AEO1995
AEO1997 3000 AEO1996
35
dry gas production Tcf
AEO1997
cumulative dry gas production Tcf
AEO1998
2750 AEO1998
AEO1999
AEO1999
30 AEO2000 2500 AEO2000
AEO2001 AEO2001
AEO2002 2250 AEO2002
25 AEO2003 AEO2003
2000
AEO2004 AEO2004
AEO2005 1750 AEO2005
20 AEO2006 AEO2006
AEO2007 1500 AEO2007
AEO2008 AEO2008
15 1250 AEO2009
AEO2009
AEO2010 AEO2010
1000 AEO2011
10 AEO2011
AEO2012
AEO2012 750
AEO2013
AEO2013
500 AEO2014
5 AEO2014
AEO2015
AEO2015
250 AEO2016
AEO2016 AEO2017
0 AEO2017
1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 0 AEO2018
AEO2018 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 cum actual
Jean Laherrere March 2018 year actual dry Jean Laherrere March 2018 year cum + 1P
10
US cumulative dry gas production forecasts from USDOE/EIA/AEO US cumulative dry gas production forecasts from AEO 2008 & 2018
3000
and JL ultimates
for 2010 4000
900 1250
1000
600 750
500
300
250
0 0
1990 1995 2000 2005 2010 2015 2020 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
Jean Laherrere March 2018 year of forecast Jean Laherrere March 2018 year
My own estimates for the ultimates of total US NG marketed has increased from 1600 Tcf in
2008 to 2400 Tcf in 2018
In contrast, the official EIA/AEO forecast is for dry gas production, which is about 7% less than
marketed gas. Our forecast U=2400 Tcf is for marketed gas. The difference of forecasts between
AEO 2108 dry gas and U=2400 Tcf/1,07 is 13 Tcf for 2030, 20 Tcf in 2040 and 30 Tcf in 2050;
But AEO2018 NG consumption is well below AEO production by about 7 Tcf, but exceeds my
forecast after 2023 and in 2050 20 Tcf is missing. The plot of AEO 2018 US NG consumption
shows that US in 2050, instead of exporting 8 Tcf, will be forced to import 20 Tcf: quite a
change compared to the official statements
US natural gas production, consumption & forecasts AEO2018 & U=2400 Tcf
45
40 U = 2400 Tcf
marketed
AEO2018
35
dry
consumption AEO 2018
30
annual production Tcf
AEO2018-U=2400 Tcf/1,07
25
20
15
10
0
1980 1990 2000 2010 2020 2030 2040 2050 2060
Jean Laherrere 20 Feb 2018 year
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 2018 displays US NG production 2012-2050 from EIA/AEO 2017 (but not his
own forecast) and Art Berman displays US NG monthly production 2000-2017 (production starts
at 20 Gcf/d)
11
DNV GL (Det Norske Veritas Germanischer Lloyd in Norway is assumed to tell the truth!)
report “Oil and gas forecast to 2050- Energy Transition outlook 2017” displays a peak of
unconventional onshore gas production in 2016 for North America at 880 G.m3 = 31 Tcf
DNV forecasts that North America unconventional onshore gas in 2050 will be 70% of 2016
value (going down), against 160 % (going up) for AEO2018 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 2018 outlook forecasts that the North America unconventional gas should be in
2040 150% of its 2016 value, in agreement with AEO 2018
12
US natural gas price: wellhead, citygate & Henry Hub from EIA & heat content US oil over gas price MBtu ratio from EIA and flared over marketed percentage
10 1,16 7 56
annual oil/gas ratio
9
1,14 6 oil/gas ratio AEO2018 48
flared/sold % ND
1,1
6
4 32
wellhead $/MBtu
5 1,08
citygate $/kcf
3 24
4 Henry Hub spot $/Mbtu
1,06
heat content mark MBtu/kcf
3 2 16
heat content dry MBtu/kcf
1,04
2
1 8
1,02
1
0 1 0 0
1950 1960 1970 1980 1990 2000 2010 2020 2030 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050
Jean Laherrere March 2018 year Jean Laherrere March 2018 year
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 6 in 1950 decreasing slowly to 1 in 2004, increasing to 6 in
2013 and is presently around 3. EIA forecast that it will increase but stay below 4 in 2050. 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 2004 in North Dakota. I am surprised
that the 1950-2005 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.
13
Oil production
US Bakken monthly production from EIA play & forecast U= 4.5 Gb US HL of Bakken LTO monthly production from EIA play 1953-Sept2017
1,3 9 000 4 120
1,2 Bakken (ND & MT) 2000-2017
U = 4.5 Gb 8 000 Jan2014-Feb2017
1,1
CP+RR WTI 100
1,0 EIA 1P 7 000
3
cumulative production Mb
U = 4,5 Gb
0,9
cum prod Mb 6 000 80
0,8
5 000
aP/CP%
WTI $/b
0,7
Mb/d
2 60
0,6 4 000
0,5
3 000 40
0,4
0,3 1
2 000
0,2 20
1 000
0,1
CP 1953-1999 = 42 Mb
0,0 0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Mb
WTI $/b
800
3 75
700
aP/CP %
WTI $/b
600
500 2 50
400
300
1 25
200
100
0 0 0
2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 0 500 1000 1500 2000 2500 3000 3500 4000
14
Bakken oil production US Bakken annual cumulative production from EIA AEO & U= 4.5 Gb
3,0 26
24 AEO2018
2,5
AEO2017
22 AEO2016
20 AEO2015
2,0
AEO2014
18
AEO2013
Mb/d
1,5 16 U=4.5 Gb
past EIA/FAQ
14
Gb
1,0
12
10
0,5
8
0,0 6
1990 2000 2010 2020 2030 2040 2050
US Eagle Ford monthly production & forecast U = 3 Gb US HL of Eagle Ford LTO monthly production from EIA play
1,8 7 000 8 120
Eagle Ford play
aP/CP%
1,6 U = 3 Gb
6 000 7
CP+RR Oct2013-Nov2016 100
1,4 EIA 1P
WTI
cumulative production Mb
6
U = 3 Gb 5 000
1,2 CP Mb 80
5
1,0 4 000
aP/CP%
WTI $/b
Mb/d
4 60
0,8 3 000
3
0,6
40
2 000
0,4 2
1 000 20
0,2 1
0,0 0
2005 2010 2015 2020 2025 2030 0 0
0 500 1000 1500 2000 2500 3000 3500 4000
Jean Laherrere Feb 2018 year https://www.eia.gov/petroleum/data.php#crude Jean Laherrere Feb 2018 cumulative production Mb
70
350 1,0 60
60
aP/CP %
WTI $/b
300 0,8 50
Mb/d
50
250 0,7 40
40
200 0,5
30
30
150 0,4
20
20
100 0,3
10 10
50 0,1
0 0
0 0,0 0 500 1000 1500 2000 2500 3000 3500
2005 2010 2015 2020 2025
Jean Laherrere March 2018 http://www.rrc.state.tx.us/eagleford/index.php Jean Laherrere Jan 2018 cumulative production Mb
year
15
Eagle Ford oil production US Eagle Ford cumulative production fromAEO & forecast U = 3 Gb
20
1,8
AEO2018
18 AEO2017
1,6
AEO2016
16
1,4
AEO2015
AEO2014
14
1,2 AEO2013
U = 3 Gb
12
1,0 past
Mb/d
Gb
10
0,8
8
0,6
6
0,4
4
0,2
2
0,0
1990 2000 2010 2020 2030 2040 2050 0
AEO2013 Eagle Ford AEO2014 Eagle Ford AEO2015 Eagle Ford 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
AEO2016 Eagle Ford AEO2017 Eagle Ford AEO2018 Eagle Ford year source Mason Inman
Jean Laherrere March2018
US LTO Permian monthly oil production for U = 10 Gb HL of Permian LTO monthly oil production 2000- sept 2017
2,0 10 000 3 120
aP/CP%
1,8 9 000 March2015-Feb2017 108
Permian play
1,6 real LTO ? 8 000 WTI 96
U = 10 Gb
cumulative production Mb
WTI $/b
CP Mb
Mb/d
kb/d
1,0 5 000 60
source: EIA/FAQ
0,8 4 000 48
1
0,6 source: EIA/FAQ 3 000 36
0,4 2 000 24
0,2 1 000 12
0,0 0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050 0 2 000 4 000 6 000 8 000 10 000 12 000
Jean Laherrere 10 Feb 2018 year Jean Laherrere 10 Feb 2018 cumulative production Mb
All Permian basin annual production displays a decline since the peak of 1974 which is constant
at 3%/a until 2005 before the burst of LTO, this 3% decline is extrapolated down to 2017 and
the difference with all oil production is assumed to be the LTO, in agreement with EIA values,
except for the period 2005-2010 where the EIA confuses LTO and horizontal wells
Permian Basin oil production & forecasts from ultimates Permian Basin LTO oil production Hubbert linearization 2004-2017
900 decline3% + U=10 Gb 2,5 35
U = 50 Gb
Tx+NM = conv
800 2,2 aP/CP%
decline 3% 30
2012-2016
U LTO = 10 Gb
700 1,9
LTO =all-decline3%
LTO EIA/FAQ 25
600 1,6
annual production Mb
all PB EIA
500 1,4 20
aP/CP%
Mb/d
400 1,1
15
300 0,8
10
200 0,5
100 0,3 5
0 0,0
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 0
year source TxRRC & NMemnrd 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Jean Laherrere Feb 2018 Jean Laherrere Feb 2018 cumulative oil production Gb
16
From Mason Inman data: AEO 2013 to 2018
Permian oil production Permian cumulative oil production from AEO & U = 10 Gb
4,5 50
AEO2018
4,0
45 AEO2017
3,5 AEO2016
40
AEO2015
3,0
35 AEO2014
2,5 AEO2013
Mb/d
2,0 30 U = 10 Gb
past
Gb
1,5 25
1,0
20
0,5
15
0,0
1990 2000 2010 2020 2030 2040 2050
10
AEO2013 Permian Basin
AEO2014 Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp]
5
AEO2015 Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp]
AEO2016 Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp]
0
AEO2017 Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
AEO2018 Permian? [Austin Chalk + Avalon/Bone Springs + Spraberry + Wolfcamp] Jean Laherrere March 2018 year
all LTO-Permian-Bakken-Eagle Ford monthly production US HL of all LTO-Permian-Bakken-Eagle Ford monthly production
0,7 3 500 2 120
LTO-Permian-Bakken-Eagle aP/CP%
U = 3 Gb Aug2015-Jul2017
0,6 real LTO? 3 000 100
WTI
CP+RR
cumulative production Mb
0,4 2 000
Mb/d
1 60
0,3 1 500
40
0,2 1 000
20
0,1 500
0 0
0,0 0 0 500 1000 1500 2000 2500 3000 3500 4000
2000 2005 2010 2015 2020 2025 2030
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Mb
US LTO monthly production from EIA play & forecast U= 20 Gb US HL of LTO production from EIA play 1953-Dec2017
5 25 000 4 120
all plays 2000-Sept2017
U = 20 Gb Dec2012-Feb2017
CP Mb WTI 100
4 20 000
EIA 1P
3
cumulative production Mb
CP+RR
U = 20 Gb 80
3 15 000
aP/CP%
WTI $/b
Mb/d
2 60
2 10 000
ultimate in Gb
Permian = 10 40
Bakken = 4.5
1 Eagle Ford = 3
1 5 000 rest = 3
total = 20.5 20
CP 1953-1999 = 42 Mb
0 0 0 0
2000 2005 2010 2015 2020 2025 2030 2035 0 2 000 4 000 6 000 8 000 10 000 12 000 14 000 16 000 18 000 20 000 22 000 24 000
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Mb
17
From Mason Inman data: AEO 2013 to 2018
Total LTO production US LTO cumulative production from AEO & U=20 Gb
9,0 110
Gb
4,0 past
50
3,0 40
2,0 30
20
1,0
10
0,0
1990 2000 2010 2020 2030 2040 2050
0
AEO2013 Total AEO2014 Total AEO2015 Total 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
AEO2016 Total AEO2017 Total AEO2018 Total
Jean Laherrere March 2018 year
USL48 crude oil (incl condensate) oil production US Lower 48 HL of oil production 1859-2017
3,5 280
9
USL48
3,0 240 8 1859-2017
USL48 less LTO
U = 235 Gb 1978-2008 before LTO
7
U = 215 Gb
cumulative production Gb
6
annual production Gb
3
1,0 80
2
0,5 40
1
0,0 0 0
1860 1880 1900 1920 1940 1960 1980 2000 2020 2040 2060 2080 0 20 40 60 80 100 120 140 160 180 200 220 240 260
Jean Laherrere Feb 2018 year Jean Laherrere Feb 2018 cumulative production Gb
-Crude oil: US
US crude oil + condensate production from EIA & forecasts from U=280 Gb & AEO US HL of crude oil production 1859-2017
5,0 300 9
AEO2018
4,5 AEO2017 270 8 aP/CP%
AEO2016
AEO2015 1931-2007
annual production crude +condensate Gb
4,0 240 7
AEO2014
cumulative production & reserves Gb
U=280 Gb
annual/cumulative %
3,5 210
past 6
cum + 1P
3,0 U = 280 Gb 180
cum prod
5
2,5 150
4
2,0 120
3 AEO2018 extrapolated
1,5 90 towards 2100 = 500 Gb
2
1,0 60
1
0,5 30
0
0,0 0 0 50 100 150 200 250 300 350 400 450 500
1860 1880 1900 1920 1940 1960 1980 2000 2020 2040 2060
Jean Laherrere Feb 2018 year Jean Laherrere March 2018 cumulative production Gb
18
The difference between AEO 2018 and my forecast for U = 280 Gb is over 2 Gb in 2030 and 3.8
Gb in 2050. AEO 2018 crude oil consumption is forecasted flat around 6.2 Gb/a from 2016 to
2050; But in percentage of AEO 2018 my forecast is 50% in 2030 but only 6.5 % in 2050
US crude oil production & consumption AEO2018 & U=280 Gb
6,5 100
6,0
90
5,5 AEO2018 cons
AEO2018 prod
annual production crude +condensate Gb
80
5,0 U=280 Gb
AEO2018-U280Gb
4,5 70
past
% U280Gb/AEO2018
4,0 % U280Gb/AEO2018
60
3,5
50
3,0
40
2,5
2,0 30
1,5
20
1,0
10
0,5
0,0 0
2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Jean Laherrere March 2018 year
19
US cumulative crude oil production forecasts from USDOE/EIA/AEO AEO 1979 US cumulative crude oil production forecasts from USDOE/EIA/AEO
500 AEO 1982
AEO 1983
400
AEO 1985
450 AEO 1990
? AEO 1994 350
350
AEO 1999
AEO 2000
AEO 2001
250
300
AEO 2002
AEO 2003
250 AEO 2004 200
AEO 2005
AEO 2006
200 AEO 2007
150
AEO 2008 for 2010
AEO 2009
150 for 2020
AEO 2010
AEO 2011 100
AEO 2012
for 2030
100
AEO 2013 for 2040
AEO 2014 50
50 AEO 2015 for 2050
AEO 2016
AEO 2017
0 AEO 2018
0
1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 cum actual
1990 1995 2000 2005 2010 2015 2020
Jean Laherrere March 2018 year cum + 1P
Jean Laherrere Feb year of forecast
2018
Rystad (a Norwegian data seller) has published a paper on world reserves in 2016.
Laherrere J.H. 2016 « World, US, Saudi Arabia, Russia & UK oil production & reserves -Comments on
Rystad 2016 world reserves » August http://aspofrance.org/files/reservesUS_SA_%20Ru_UK-JL2016.pdf
https://aspofrance.org/2016/08/11/world-us-saudi-arabia-russia-uk-oil-production-reserves-august-2016-jean-
laherrere/
Bowden reported in 1982 the US crude oil resources (ultimate?) showing an evolution with time:
ultimates less than 200 Gb from 1940 to 1956 (Hubbert in 1956 was using 150 & 200 Gb for
USL48), between 150 and 600 Gb (with Zapp’s estimate in 1962 = 590 Gb from unrealistic
recovery per feet drilled) from 1956 to 1974 and between 200 and 280 Gb from 1974 to 1980:
the range was wild and it is still!
The future US crude oil (+condensate) production is modeled with the likely ultimate of 300 Gb
and the unlikely Rystad 480 Gb and the EIA/AEO forecasts from 2013 to 2016. It appears that
Rystad ultimate of 480 Gb is related to AEO 2016 forecast of more than 11 Mb/d in 2040. There
is a wild change in LTO forecasts between AEO 2015 and AEO 2016
20
-Other forecasts
David Hughes 2018 displays US crude oil production from AEO2018 peaking in 2042, 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 1 Mb/d from 2015 to 2030 for Canada, when
AEO2018 forecasts an increase of 2 Mb/d for the US, when DNV forecasts an increase of 4
Mb/d for North America: it means that DNV is more optimistic than EIA
ExxonMobil Outlook 2018 forecasts that tight oil in 2040 will be 3 times the value of 2016
21
-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 220 to 300 Gb between 2002 and
2018, when the extrapolation of AEO 2002 cumulative production to 2100 (assumed close to
ultimate) is 350 Gb and AEO2018 500 Gb. My change is less than EIA
US all USL48
site year forecast ultimate Gb site year forecast ultimate Gb
Uppsala 2002 220 Enc Energy 2003 200
Enc Energy 2003 230 ENSMP 2004 200
Lisbon 2005 230 MIT 2014 225
Beijing 2006 230 2018 235
Luxembourg 2006 250
comRystad 2016 300
2018 280
US cumulative crude oil production forecasts from USDOE/EIA/AEO US cumulative crude oil production forecasts from USDOE/EIA/AEO 2018 & 2002
500 and JL ultimate
400
450
AEO 2018
350 ?
AEO 2002
cumulative crude oil production Gb
400
JL ultimate 2018
300
JL ultimate 2002
cumulative crude oil production Gb
350
cum actual
250 300
200 250
200
150 for 2010
for 2020 150
100 for 2030
for 2040 100
for 2050
50
ultimates JL 50
0 0
1990 1995 2000 2005 2010 2015 2020 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100
Jean Laherrere Feb 2018 year of forecast Jean Laherrere March 2018 year
120 -65 60
-88
105 -70
-89
WTI oil price $/b
90 -75 50
-90
75 -80 -91
60 -85 -92
40
-93
45 -90
-94
30 -95 30
-95
15 -100 WTI $/b
-96
dollar shift 30 days
0 -105 20 -97
1/1/00 12/31/04 1/1/10 1/1/15 1/2/20 Jan-15 Jan-16 Jan-17 Jan-18 Jan-19
Jean Laherrere March 2018 date Jean Laherrere March 2018 date
22
Some believe that the oil price follows the private stocks of crude oil, but since 2014 it appears
that the stocks follows the WTI 13 weeks before, except few months in 2017 after the OPEC
Russia deal for reducing production
US weekly private stocks crude oil & WTI *-1 shifted 13 weeks US weekly private stocks crude oil & WTI *-1 shifted 13 weeks
540 -15 560 0
520 -25
510 -20
500
-35
-45 460
Weekly U.S. Ending Stocks excluding SPR
460 of Crude Oil "excluding lease stock" Mb
-60
-55 WTI *-1 13 weeks before
440 410
400 360
-75 -100
380
-85 310
360 Weekly U.S. Ending Stocks excluding SPR of -120
Crude Oil "excluding lease stock" Mb -95
340 260
WTI *-1 13 weeks before -140
320 -105
Conclusions
The USDOE/EIA believes that the US oil and gas production will be higher in 2050 (and in
every year in between) than in 2017, 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 USL48 many cycles of past
production and drilling were symmetrical that the shale plays will also display symmetrical
future production curves.
AEO2018 forecasts US production in 2040 for oil at 12 Mb/d, my forecast is 4 times less at 3
Mb/d, for gas at 40 Tcf, my forecast is half at 20 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 2P (proved +
probable) reserves. The EIA has to recognize the fact reported by the SPE/PRMS that the
aggregation of proved (1P) reserves is incorrect, only the addition of 2P is correct. (Laherrère
J.H. 2008 «Advice from an old geologist-geophysicist on how to understand Nature»
presentation Statoil Oslo 14 August http://aspofrance.viabloga.com/files/JL_Statoil08_long.pdf)
USDOE/EIA should report the 2P US reserves like the USDOI/BOEM for the reserves of the
Gulf of Mexico. Reporting 2P and backdating discovery allows to estimate ultimate using
creaming curve.
EIA once reported as an open file backdated established reserves with USDOE/EIA-0534 1990
"US oil and gas reserves by year 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 10 years ago were reporting proved and non-proved reserves (2P), but today
they report only 1P to follow the SEC rule. Today only Gazprom is reporting 1P and 2P, as also
their queer ABC1.
23
All scouting agencies report the confidential 2P reserves used internally by the operators to
decide the development of their fields.
24