Post on 25-Jan-2023
ANADARKO PETROLEUM EXPLORATION LOOKBACK 2004-18:
LESSONS LEARNED FOR CONVENTIONAL EXPLORATION RISK AND UNCERTAINTY PREDICTIONS
R. BRENT BOYD - DIR. GEOPHYSICS, INT’L EXPLORATION
OCCIDENTAL PETROLEUM CORPORATION
CAUTIONARY STATEMENT
Cautionary Note to U.S. InvestorsThe SEC permits oil and gas companies, in their filings with the SEC, to disclose only proved, probable and possible reserves. Any reserve estimates provided in this presentation that are not specifically designated as being estimates of proved reserves may include “potential” reserves and/or other estimated reserves not necessarily calculated in accordance with, or contemplated by, the SEC’s latest reserve reporting guidelines. U.S. investors are urged to consider closely the oil and gas disclosures in our 2018 Form 10-K and other reports and filings with the SEC. Copies are available from the SEC and through our website www.oxy.com
CLARIFICATIONS
• Discovered volumes are based on the first post-drill estimate of the discovery resource size. For purposes of this lookback, the volumes are not updated after appraisal and development of the discovery.
• Risk and volume predictions should not involve external corporate and market influences. This lookback tries to focus on improving future geologic predictions without questioning the corporate decisions of strategy and funding.
15+ BILLION BOEDISCOVEREDFROM 173 WELLS DURING2004-2018
• Success is not due to chance.
• This was a concerted effort by Anadarko to add long-term corporate value
• Train exploration teams in proper risking methods
• Use analytical software built for risk and uncertainty predictions (Rose and Assoc MMRA, MZM and SAAM).
• Create a Risk Consistency Team to review all prospects pre- and post-drill
• Analyze well results and compare to predictions for any lessons learned.
• Share lessons with the Anadarko exploration community
• Participate in industry-wide risk consortia and workshops
LESSONS LEARNED FROM THIS LOOKBACK
• Successful exploration starts with a healthy prospect inventory. A healthy prospect inventory is:
• Deep
• Continually updated
• Regularly high-graded
• Keeps pace with drilling activity
• Drilling second-tier prospects from a stale inventory leads to under-performance.
• Drilling a small number of wells creates unpredictable success rates.
• There is a tendency to over-promise and under-deliver resource volumes in successful wells
MAJOR ANADARKO CONVENTIONAL EXPLORATION DISCOVERIES*
15+ Billion boe resources discovered from 173 conventional exploration wells (2004-18)* Resource Volume source: IHS
Heidelberg 330 MMboe
Lucius 331 MMboe
West Tonga 200MMboe
Ourhoud 1502 MMboe
El Merk 745 MMboe
Block 404 1400+ MMboe
Jubilee/TEN 961 MMboe
Golfinho/Atum 28 TCF
Prosperidade Complex 32.5 TCF
Mamba Complex 50 TCF
The Exploration Program: 2004-2018
4 16 20 24 28 33 35 35 36 40 44 47 49 51 5310
18 26 39 4357 71 80 90 103 113 116 120 120 120
712
2433
4152
6980
91100
108 117 117 117 122
0
50
100
150
200
250
300
350
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Nu
mbe
r of
Wel
ls/P
ilots
Cumulative Conventional and Unconventional Wells (2004 to 2018)
GOM Int'l/Frontier North America
Exploration Wells By Area
295 Total Wells • 122 Onshore US• 120 International• 53 GOM International
Onshore US
GOM
Algeria
Cote D’Ivoire
Ghana
Mozambique
Gulf Coast
Delaware Basin
Walker Ridge
Green Canyon
Lloyd Ridge
Mississippi Canyon
16
20
12
16
10
19 19
9
11
17
14
6 6
4
2
5 5
1211
6
11
14
11 11
98
9
0 0
5
0
5
10
15
20
25
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Num
ber
of W
ells
Annual Conventional and Unconventional Wells
Conventional Unconventional
The Exploration Program: The Past 15 Years
Years CONVENTIONAL UNCON.
2016-18 4 < 2
Years CONVENTIONAL UNCON.
2004-15 14 9
CONVENTIONAL EXPLORATION RISK PREDICTION2004-18
Definitions:Pg = Well Chance of SuccessAverage Predrill Pg (Predicted Pg): The average Pg value for all the drilled prospects in a yearActual Success Rate: The percentage of drilled prospects that are successful in a year.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Average Pre-drill Pg vs Actual Success Rate
Predicted Actual
Conventional Pg Prediction vs Actual by Year
2004-18Predicted: 57%
Actual: 53%
Industry average: approx. 49%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Average Pre-drill Pg vs Actual Success Rate
Predicted Actual
2004-15Predicted: 53%
Actual: 52%
Conventional Pg Prediction vs Actual by Year
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Average Pre-drill Pg vs Actual Success Rate
Predicted Actual
Attempting to only drill the ‘sure bets’
2004-15Predicted: 53%
Actual: 52%
Conventional Pg Prediction vs Actual by Year
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Average Pre-drill Pg vs Actual Success Rate
Predicted Actual
Attempting to only drill the ‘sure bets’
2004-15Predicted: 53%
Actual: 52%
Conventional Pg Prediction vs Actual by Year
Gives unpredictable results with the small sample sizes
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Actu
al S
ucce
ss R
ate
Average Pre-drill Pg
Annual Average Pre-Drill Pg vs Actual Pg2016
2018
2017
Conventional Pg Prediction vs Actual by Year
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95
Act
ual S
ucce
ss R
ate
Predrill Pg (binned)
Predicted vs Actual Success
N = 6
Conventional Pg Prediction vs Actual by Pg Decile Bins
N = 17N = 25
N = 34
N = 27N = 28
N = 23
N = 9
N = 13
0-0.19 0.20-0.29 0.30-0.39 0.40-0.49 0.50-0.59 0.60-0.69 0.70-0.79 0.80-0.89 0.90-1.0
Predrill Pg Observations:
• Under 20% is a dry hole, but sample size is small
• 20-29% - the results are really 40%
• 80-100% - over-confident
GOM Exploration Program: Pg Predicted vs Actual
61% Pre-drill
65% Actual
PgW
ells
International Exploration Program: Pg Predicted vs ActualPgW
ells
50% Pre-drill
48% Actual
RISK PREDICTION LESSONS
• With enough samples, the Pg predictions were very good.
• Be honest with the Pg, especially if only drilling one or two wells for the year.
• Be wary of over-confidence when assessing prospects where perceived Pg >80%
• Perceived Pg of 20-29% may be better than you think.
CONVENTIONAL EXPLORATION RESOURCE PREDICTION2004-18
Definitions:P10Volume: 10% of the range of resource volume estimates are larger than this amount.P90Volume: 90% of the range of resource volume estimates are larger than this amount.Predrill Mean Resource Volume: The sum of all the risked mean resource volumes for all prospects drilled in a yearDiscovered Mean Volume: The sum of the mean resource volumes for all the successful prospects in a year.
Cumulative Conventional Exploration Volumes Predicted vs Success
Predicted mean
Success mean
Predicted P10
Predicted P90
Cumulative Conventional Exploration Volumes Predicted vs Success
Predicted P10
Predicted P90
Predicted mean
Success mean
Two successes with large overestimations of volume
Cumulative Conventional Exploration Volumes Predicted vs Success without two wells
Predicted mean
Success mean
Predicted P10
Predicted P90
Cumulative Conventional Exploration Volumes Predicted vs Success without 2004
Predicted mean
Success mean
Predicted P10
Predicted P90
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)*(2004 to 2018)
Pre-Drill Risked Post-Drill
Conventional Exploration Volumes by Year Predicted vs Success
*Includes appraisals
Conventional Exploration Volumes by Year Predicted vs Success
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)(2004 to 2018)
Pre-Drill Risked Post-Drill
Good prediction
2004-09
Pre-drill mean resource volume Discovered mean resource volume
Conventional Exploration Volumes by Year Predicted vs Success
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)(2004 to 2018)
Pre-Drill Risked Post-Drill
2004-11
Pre-drill mean resource volume Discovered mean resource volume
Surprise! Mozambique!
Underestimate Resources
Good prediction
Conventional Exploration Volumes by Year Predicted vs Success
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)(2004 to 2018)
Pre-Drill Risked Post-Drill
Surprise! Mozambique!
Underestimate Resources
Good prediction
2004-13
Pre-drill mean resource volume Discovered mean resource volume
Adjust:Good
prediction
Conventional Exploration Volumes by Year Predicted vs Success
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)(2004 to 2018)
Pre-Drill Risked Post-Drill
2004-15
Pre-drill mean resource volume Discovered mean resource volume
Surprise!Overestimate
Resources
Surprise! Mozambique!
Underestimate Resources
Good prediction
Adjust:Good
prediction
Conventional Exploration Volumes by Year Predicted vs Success
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)(2004 to 2018)
Pre-Drill Risked Post-Drill
2004-18
Pre-drill mean resource volume Discovered mean resource volume
Surprise!Overestimate
Resources
Surprise! Mozambique!
Underestimate Resources
Good prediction
Adjust:Good
prediction
Low well count/challenging
Conventional Exploration Volumes by Year Predicted vs Success
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Tot
al R
esou
rce
(MM
BO
E)
Exploration Performance by Year (MMBOE)(2004 to 2018)
Pre-Drill Risked Post-Drill
Results:1544 MMBOE/yr14 wells drilled/yr2004-15 average
Results:337 MMBOE/yr4 wells drilled/yr2016-18 average
Pre-drill mean resource volume Discovered mean resource volume
Successful Wells: Reserve Prediction – Int’l Exploration
0.1
1
10
100
1000
10000
0.1 1 10 100 1000 10000
Post
-Dril
l Mea
n Re
serv
es (M
MBO
E)
Pre-Drill Mean Reserves (MMBOE)
Successful Exploration Drilling 2004-18
30:100Pre-drill 30 MMBOE = Post-
drill 100 MMBOE
30:10Pre-drill 30 MMBOE = Post-
drill 10 MMBOE
Successful Wells: Reserve Prediction – GOM Exploration
1
10
100
1000
1 10 100 1000
Post
-Dril
l Mea
n Re
serv
es (M
MBO
E)
Pre-Drill Mean Reserves (MMBOE)
Successful Exploration Drilling 2004-18
30:10Pre-drill 30 MMBOE = Post-
drill 10 MMBOE
30:100Pre-drill 30 MMBOE = Post-
drill 100 MMBOE
RESOURCE PREDICTION LESSONS
• Pre-drill prospect resource estimates tend to be larger than the actual resources discovered.
• Be vigilant to keep the P99 resource volume low to prevent overestimation.
• Keep P01 resource volume within geologic parameters
Conventional Exploration Prospect Inventory2007-18
0
10
20
30
40
50
60
70
80
90
100
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Drilled Drillable Prospects
Prospects
Ready-to-DrillProspects
Drilled
Conventional Exploration Prospect Inventory2007-18
0
10
20
30
40
50
60
70
80
90
100
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Drilled Drillable Prospects
Prospects
Ready-to-DrillProspects
Drilled
Conventional Exploration Prospect Inventory2007-18
0
10
20
30
40
50
60
70
80
90
100
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Drilled Drillable Prospects
Prospects
Ready-to-DrillProspects
Drilled
Conventional Exploration Prospect Inventory2007-18
0
10
20
30
40
50
60
70
80
90
100
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Drilled Drillable Prospects
Prospects
Ready-to-DrillProspects
Drilled
Stagnant
Conventional Exploration Performance by Year2004-18
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Res
ourc
e (M
MB
OE
)Exploration Performance by Year (MMBOE)
(2004 to 2018)
Pre-Drill Risked Post-Drill
20
40
60
0
Inven
tory Size
Conventional Exploration Performance by Year2004-18
0
1,000
2,000
3,000
4,000
5,000
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Res
ourc
e (M
MB
OE
)Exploration Performance by Year (MMBOE)
(2004 to 2018)
Pre-Drill Risked Post-Drill
Last two years of the harvest –The first tier prospects had been drilled and these were the second tier prospects
Disappointing results
20
40
60
0
Inven
tory Size
PROSPECT INVENTORY LESSONS
• A healthy prospect inventory is as vital to exploration success as good Pg and volume prediction.
• The effects of a dwindling prospect inventory may not be seen for many years.
• Prospects have an “expiration date”. They need to be constantly competing with new prospects in the inventory.
• The number of new prospects being delivered to the inventory needs to keep ahead of the drilling rate.
• Without replenishment and competition, second-tier prospects appear better than they really are.
LESSONS LEARNED FROM THIS LOOKBACK
• Successful exploration starts with a healthy prospect inventory. A healthy prospect inventory is:
• Deep
• Continually updated
• Regularly high-graded
• Keeps pace with drilling activity
• Drilling second-tier prospects from a stale inventory leads to under-performance.
• Drilling a small number of wells creates unpredictable success rates.
• There is a tendency to over-promise and under-deliver resource volumes in successful wells
FINAL LESSON LEARNED:
ANADARKO WAS PRETTY GOOD AT EXPLORATION
Acknowledgements: - Occidental Petroleum Corporation- Dr. Alexei Milkov, Colorado School of Mines - Matt Morris, Anadarko- Diane Williams, Anadarko- The many past Anadarko employees who were a part of the Risk Consistency Team
and maintained the Risk Consistency database
A PERFECT PG PREDICTION (+/- 10%) FOR VARIOUS PORTFOLIO SIZES
With more than 10 wells, a bad Pg portfolio outcome is more likely due to poor prediction than bad luck.With 10 wells or fewer, luck determines the outcome and not predictive ability