1. NPRA Annual Meeting 2003 Effects of Feed Quality and Product Specification Changes on Refined Product Supply Joanne Shore John Hackworth Energy Information Administration NPRA Annual Meeting San Antonio, Texas March 2003 www.eia.doe.gov

2. NPRA Annual Meeting 2003 Petroleum Product Supply Affected By Numerous Factors Changing Product Specifications Changing Feed Quality Capacity Changes Petroleum Refined Product Supply Technology Imports

3. NPRA Annual Meeting 2003 U.S. Refining Capacity: Where We Are Today Source: EIA

4. NPRA Annual Meeting 2003 Domestic Capacity Is Expected To Grow Source: EIA Time Period Annual Average Shutdown Capacity (MB/CD) Average Annual Growth of Continuously Operating Capacity (MB/CD) 1990-199513976 1995-200096324 2000-200760346 Historical shutdowns are for years 1990-1994, 1995-1999.

5. NPRA Annual Meeting 2003 Gasoline Imports Are Important Source of Supply for East Coast Source: EIA

6. NPRA Annual Meeting 2003 Changing Feedstocks: Any Signs of Supply Impacts? Growing use of heavier crude oils impact on light product yields Historical changes in the use of C4’s, C5’s and oxygenates in gasoline

7. NPRA Annual Meeting 2003 Heavier Crude Oils Accounted for Import Growth Source: EIA

8. NPRA Annual Meeting 2003 U.S. Refinery Input Quality Source: EIA

9. NPRA Annual Meeting 2003 Rapid Growth in Bottoms Processing for Heavy Sour Crude Oils Source: EIA – Capacity as of January Each Year

10. NPRA Annual Meeting 2003 Changes in Light-Product Yields at Refineries with Coker and Heavy Crude Oils Projects Explored changes in refineries adding cokers between 1998 and 2001 Looked at two groups with different product yield impacts: –Group A : Little change in crude gravity, reduction in residual fuel oil yield –Group B - Large drop in crude gravity, little residual fuel oil to reduce

11. NPRA Annual Meeting 2003 Before & After Coker Additions Group AGroup BTotal BeforeAfterBeforeAfterBeforeAfter Crude API Gravity28.026.233.125.729.726.0 As Percent Crude & Unfinished FCC Input35.035.936.935.435.7 Hydrocracker Input5.65.08.113.56.58.2 Coker Input11.021.07.523.19.821.8 Pentanes Plus Input0.32.32.03.90.92.9 C4+MTBE Input4.83.45.24.65.03.9 Yield Adj Mogas: % C+U45.943.848.946.246.944.7 Yield Distillate: % C+U31.235.836.035.432.935.6 Yield Light Prod: % C+U78.881.288.584.182.382.2 Yield Residual: % C+U8.13.11.71.25.82.4 NOTE: Group A: 4 refineries, 140 MB/D coking capacity added; Group B: 4 refineries, 130 MB/D coking added. Gasoline yield is adjusted by removing contributions from C4, C5 and oxygenate blend components so that yields reflect only gasoline product derived from crude oil and unfinished feedstocks (C+U).

12. NPRA Annual Meeting 2003 Oxygenates, C4’s and C5’s Contribute Significant Volumes to Gasoline Production Source: EIA

13. NPRA Annual Meeting 2003 Flat Crude-Based Gasoline Yield Source: EIA

14. NPRA Annual Meeting 2003 New Regulatory Supply Impacts MTBE/Ether Bans MSAT with MTBE Bans ULSD

15. NPRA Annual Meeting 2003 Losing MTBE – Not Just MTBE Volume Loss MTBE represents over 10% RFG, 3% total gasoline supply. But physical & chemical properties are critical factors. No other hydrocarbon or alcohol can equal the emission and engine performance characteristics of MTBE and other ethers.

16. NPRA Annual Meeting 2003 MTBE and Ethanol Property Comparison Both good octane and clean relative to other gasoline components Issue is Ethanol relative to MTBE: –Higher oxygen content, so less volume needed for RFG oxygen requirement (5.8% v. 11.2% for MTBE) –Higher blending RVP than MTBE (VOCs problem) –Higher toxics than MTBE (MSAT issue) –Higher NOx in California model, not Federal model

17. NPRA Annual Meeting 2003 2007 Loss of Volumes When Moving from MTBE to Ethanol – Add Backs (Constant Inputs, MB/D) PADD 1 (1) PADD 2 PADD 3 PADD 5 Total U.S. Loss of MTBE Volumes -1000-93-113-306 Addition of Ethanol to RFG 7501858151 Refinery Increased Alkylate Production 9107117107 Commercial Alkylate or Iso-Octane Production 00251035 Net Before Other Losses -161021-28-13 (1)Assumes MSAT volume issues are resolved and rule does not hinder RFG production. Source: EIA

18. NPRA Annual Meeting 2003 2007 Loss of Volumes When Moving from MTBE to Ethanol – Further Losses (Constant Inputs, MB/D) PADD 1 (1) PADD 2 PADD 3 PADD 5 Total U.S. Net Before Other Losses -161021-28-13 Reduction in Light Ends for RVP -250-38-50-113 Reduce Heavy Ends for Distillation Points -140-9-17-40 Added Ethanol in Conventional 0-2000 Total -55-10-26-95-186 (1) Assume MSAT volume issues are resolved and rule does not hinder RFG production Source: EIA

19. NPRA Annual Meeting 2003 The MSAT Issue Concerning Ethanol (Complex Model) Blend Component Share (Percent) Emission Reduction from Industry Baseline (Percent Reduction) Volume Percent Oxygen- ate Hydro- carbons MTBEEthanolVOCToxicsNO x Benzene MTBE88.811.2025.933.48.10.3 Ethanol 1 94.205.815.528.37.20.4 Ethanol 2 90.0010.018.628.87.90.3 Note: Fractions of specific components in the hydrocarbon blend are the same in all cases. Emissions reductions are based on comparisons to 1990 industry average baseline fuel composition. Source: EIA

20. NPRA Annual Meeting 2003 ULSD - The Lower the Sulfur and the Higher the Volume the Greater the Cost Source: EIA

21. NPRA Annual Meeting 2003 Potential Regulatory Changes More ULSD (off-road) More diesel light duty vehicles (diesel quality and volume issues) Increased ethanol use (mandated or otherwise) Other product clean up (sulfur reduction jet fuel, home heating oil)

22. NPRA Annual Meeting 2003 Increased U.S. Light Duty Diesel Fuel Use? 2001 New Retail Sales Light Duty Vehicles Sources: US Chart -- ORNL Investigation of Class 2bTrucks, March 2002 & EIA Estimates; Europe Chart -- ACEA

23. NPRA Annual Meeting 2003 Proposed Renewable Fuel Standard Source: EIA, Monthly Oxygenate Survey and H.R. 4

24. NPRA Annual Meeting 2003 Elimination of RVP Waiver Diminishes Ethanol Supply Contribution Volume and Energy Effects to Adjust for RVP When Adding Ethanol to Make 10% Blend RVP (psi) Volume Increase (Decrease) By Adding Ethanol Energy Increase (Decrease) By Adding Ethanol With RVP Waiver Without RVP Waiver With RVP Waiver Without RVP Waiver 9.011.1%7.5% 4.5% 7.810.3%2.2%6.8%-0.2% 7.08.3%-0.9%5.1%-2.8% Source: EIA

25. NPRA Annual Meeting 2003 Results of Tighter Product Specs and Greater Refinery Complexity Increased likelihood of outages –Single unit outage has greater impact on product production –Decrease in maximum achievable utilization Diminished yields of prime fuels per barrel of crude oil Greater “stay-in-business” investment hurdle for smaller refiners

26. NPRA Annual Meeting 2003 Summary: Impacts of Feedstock Changes Fewer heavy crude projects, but will result in some light-product yield loss Reduced ability to use light components (C4’s & C5’s) and MTBE will reduce gasoline yield per barrel of crude input

27. NPRA Annual Meeting 2003 Summary: Potential Regulatory Impacts on Gasoline Ether Bans  Largest supply challenge with significant RFG production loss, exacerbated by MSAT (Loss of MTBE, C4’s C5’s exceeds ethanol additions) RFS  Adds some volume, but loss of 1-lb RVP waiver reverses some of that gain Import streams meeting RFG specifications likely to become harder to find and more expensive More distinct gasoline types adding to distribution system stress

28. NPRA Annual Meeting 2003 Summary: Potential Regulatory Impacts on Diesel ULSD  High and widely varying costs of production and new technology may cause some refiners to delay implementation & create shortages initially Possible increase in light duty diesel vehicles  More ULSD volume and near-term quality improvements that increase refining challenges Distribution losses with ultra-clean products is a challenging unknown

29. NPRA Annual Meeting 2003 Conclusion: Challenging Environment Implies Tight Market Years Ahead Light Product Yield Feedstock changes – Loss of light product yield Product spec changes – Net loss of light product yield Capacity Need more capacity to meet demand, yield loss, potential lower utilizations Pressures to increase: –Demand growth –Higher import prices Pressures hindering expansion –Resources limited to spec changes –Margins –Shutdowns