Measuring the Quality of the Crude Oil Supply and its Impact on Basis Differentials

Transcription

1 Measuring the Quality of the Crude Oil Supply and its Impact on Basis Differentials Quality Flexibility Relationship Innovation Information = Advantage Ali AL-SUMAITI PhD Candidate, Colorado School of Mines L. G. Chorn PhD Chief Economist, Platts 19 SEPTEMBER 2007

2 Introduction and AGENDA The global crude oil supply quality is slowly declining (~ 0.1 API per year) in API gravity and growing (~0.03% per year) in sulfur content. These metrics quickly lead to the realization that the refining industry must invest heavily to reconfigure capacity to meet clean fuels requirements with a declining quality feed stock. This presentation describes a work-inprogress to forecast the decline and estimate its impact on quality-driven basis differential prices. 1

3 AGENDA Project Goals Analysis Methodologies Statistics of Producing Fields Decline analysis assumptions Basis Differential model to be used Preliminary Results Forecast of conventional oil quality change through 2030 Expectation of basis differential change Observations and Plans 2

4 The Larger GOAL In the dynamic supply/demand environment foreseen through 2030 by IEA, among others, the upstream and downstream segments of the industry must be increasingly aware of supply and its quality. The downstream must invest to accommodate quality changes and meet refined product demands. The upstream must have a clear perspective of quality-driven basis differentials as fewer world-scale fields remain to be developed and capital costs rise at an alarming rate, in order to optimize investment programs. Platts has undertaken an effort to bring together analytics capabilities with market price assessments and global news reporting to provide this insight. This presentation describes one of the first deliverables. 3

5 Analysis Methodologies Present Statistics of Producing Fields Coverage 90 producing countries with 4,800 producing fields 10 years of production history Field-by-Field metrics average daily production rate API gravity sulfur content (partial set) TAN (to be gathered) Traded crudes and blends ~110 regularly traded crudes with price assessments API gravity, sulfur content, TAN and metals content Linked to source fields (in progress) 10 years of price history (quality-driven basis differentials relative to benchmarks) Present coverage is ~ 95% of daily crude oil production with some production not yet associated with quality metrics 4

6 Analysis Methodologies Decline Curve Assumptions We assume fields behave in this general manner 2.5 daily production rate (MMbbls) Plateau period Decline period Abandonment period

7 Analysis Methodologies Decline Curve Assumptions Fit empirical production rate data to exponential decline curve, IF field is not facility limited, cartel-allocated or new and still on plateau Production Rate (BOPD) y = 5E+116e x R 2 =

8 Analysis Methodologies Missing or Incomplete Data Problems (and work-arounds ) Problem Data are not reported every year Annual production data may be aggregated by reporting agency into a basin, blend or sometimes even a total country rate. Small production rate fields suffer from wide rate variability, year over year Work-around If intermediate year is missing, then estimate, otherwise delete from database until (or if) another source can be obtained First, try to match production with a Plattstracked crude (blend) and use its properties. OR, determine if production is exported or used domestically. If domestic, its not in the traded supply. If exported, then seek additional data source these are not included in forecast at this time! Average available production rate data and decline with global-average decline rate. 7

9 Analysis Methodologies Missing or Incomplete Data Problems (and work-arounds ) CONT D Problem New field, still on plateau New field, just beginning production, added to database Work-around If large, use Platts OilGram archive to find reserves estimate and begin global-average decline* after 40% produced If small, assume global-average decline beginning immediately Use Platts OilGram archive to find reserves estimate and begin global-average decline after 40% produced. * Weighted global-average decline rate (non-opec) is 7.4%/year. 8

10 Analysis Methodologies Quality-Driven Basis Differentials Model Use model form developed by Bacon and Tordo Crude Oil Price Differentials and Differences in Oil Qualities: A Statistical Analysis,ESMAP Technical Paper 081. differential t = a + b + b + b API sulfur TAN ( Benchmark * ΔAPI) ( Benchmark * Δsulfur) ( Benchmark * ΔTAN) differential a b i Benchmark Δ quality metric t ($/bbl) at a date t intercept ($/bbl) regression coefficients price ($/bbl) of reference crude difference in gravity, sulfur %, or total acid number from reference crude 9

11 Analysis Methodologies Quality-Driven Basis Differentials Model Update regression coefficients quarterly using Platts market-assessed crude oils ($/bbl FOB), because differentials are not constant over time, as shown below for Maya (22 ; 3.3%) vs. Brent (38 ; 0.45%) Brent Price ($/bbl) Differential ($/bbl) Jan 07, 2000 Jun 30, 2000 Dec 22, 2000 Jun 15, 2001 Dec 07, 2001 May 31, 2002 Nov 22, 2002 May 16, 2003 Nov 07, 2003 Apr 30, 2004 Oct 22, 2004 Apr 15, 2005 Oct 07, 2005 Mar 31, 2006 Sep 22, 2006 Mar 16, 2007 Brent Price Differential 10

12 Preliminary Results API metrics only Recent volume changes in OPEC crude supply vs. API 10,000 9,000 8,000 Saudi Arabia Production (1,000 BOD) 7,000 6,000 5,000 4,000 3,000 2,000 1,000 Venezuela Iraq Iran Kuwait Angola Qatar UAE Nigeria Indonesia Libya Algeria

13 Preliminary Results API metrics only Forecasted volume (BOPD) change in API ranges ( ) - existing, conventional, non-opec oil fields only 1.4E+07 Production (BOD) 1.2E E E E E E E

14 Preliminary Results API metrics only Forecasted volume (percent) change in API ranges ( ) - existing, conventional, non-opec oil fields only 40% 35% 30% 25% 20% 15% % 5% 0%

15 Observations and Plans Even at this intermediate work stage, it is possible to foresee the decline in average API gravity for existing conventional oil production. Incorporating non-conventional oil production, such as Canadian syncrude and ultra-heavy crudes, into the database and forecast will indicate an even more dramatic downward trend of API gravity in the future crude oil supply. Basis differential modeling will link the crude quality to refinery margins, reflecting the pace of the refining industry to increase upgrading capacity. When differentials grow, the crude quality is declining faster than the refinery upgrading effort. When differentials shrink, the upgrading effort is moving faster than the quality is declining. 14