Environmental Quality Branch

VEHICLE EMISSIONS

B.C. Cleaner Gasolines Study

By Kilborn Inc. Toronto, Ontario
In Association with Mathpro Inc. Bethesda, Maryland

August 1995

Table of Contents

Summary
1.0  Introduction
2.0 Methodology
3.0 Results
4.0 Discussion
Appendix A: Initial Study Results
More Information

Summary

This study prepared by Kilborn Inc. in association with MathPro Inc. evaluates the impact of specifying various gasoline composition on the refineries which produce gasoline for sale in British Columbia. These refineries are:

• Chevron Burnaby, B.C.
• Husky Prince George, B.C.
• Esso Strathcona, Alberta
• Petro-Canada Edmonton, Alberta
• Shell Scotford, Alberta

In the study the refineries were modelled using a linear program model. A base case was developed to replicate the current refinery operations. The model was then adjusted to produce the various gasoline compositions and the incremental capital and operating costs from the base case were calculated. These costs were then used to develop three scenarios.

The five refinery scenario used the costs for five refineries and allocated all costs to the volumes assumed for supply into British Columbia.

The full volume scenario used the costs for the five refineries and applied them to the entire volumes produced to the B.C. gasoline specifications. This results in other market areas receiving B.C. quality gasoline.

The four refinery scenario assumed that the refiner which incurred the highest cost would obtain its B.C. volumes from the other refiners in exchange for providing equal volumes in other market areas. The models for the remaining refineries were adjusted for the increased volumes and new costs were obtained.

The following tables summarize to costs in ¢/L for each of these scenarios and indicates the range of costs incurred by the refiners.

While market forces will set the ultimate price paid for the cleaner gasolines, the costs will be specific for each refiner. As can be seen from the range of values, some refineries are better positioned to produce cleaner gasolines than others. Kilborn is not in a position to predict the final outcome.

1.0 Introduction

Gasoline for sale in B.C. as of July 1, 1995 is produced in five refineries. Two of these refineries are in B.C. - Chevron in Burnaby and Husky in Prince George. The other three are in the Edmonton area of Alberta - Esso Strathcona, Shell Scotford and Petro-Canada Edmonton. Shell and Petro-Canada had refineries in the Vancouver area which they have shutdown in the last few years; Esso ceased processing oil at its IOCO refinery on June 30, 1995.

Kilborn Inc. was retained by the British Colombia (B.C.) Ministry of Environment, Lands and Parks (MELP) to evaluate the implications of producing gasoline to 'cleaner fuel' specifications. Kilborn had previously, at MELP's request, prepared on analysis <1> of Chevron's Burnaby refinery based on public information <2> <3>. This study expands the analysis to include all five refineries. Baseline cases for the refineries which supply gasoline in B.C. were setup. From these cases, estimates of the changes in processing equipment and operations for each refinery, which would be required to meet each of the specifications, were prepared. Linear programming techniques were used to develop the costs related to each specification relative to the base cases.

Table 1.1 lists the various cases which were initially examined for the five refineries.

Table 1.1

B.C. Cleaner Gasoline Options: Total All Refineries

The results of these evaluations were reviewed on an individual basis with each refiner by MELP and to a limited extent by Kilborn. These results are reported in Appendix A. It is noted that, due to the numerous assumptions made, these results do not reflect the expected costs for the refiners. As a result of these reviews, it was determined that a more comprehensive evaluation could be prepared with input from the various refiners. Esso, Shell and Petro-Canada agreed to meet with Kilborn and provide detailed information on crude slates, operating parameters, operating constraints and capacities of the various units in their refineries. Husky discussed their operations with Kilborn and MELP in a conference call. Chevron declined to participate in a review.

Based on the information developed in the first evaluation, it was decided to focus the cases on narrower ranges of key parameters. The cases selected for more detailed review are listed in Table 1.2.

Table 1.2

B.C. Cleaner Gasoline Options: Total All Refineries: Revised Basis — 29-Jun-95

2.0 Methodology

To perform this study, each of the refineries which supply gasoline into the B.C. market were modelled using spreadsheet techniques. Refinery nameplate data taken from the Oil & Gas Journal <5> was used as a basis to approximate the refinery configurations. Assuming these configurations, the refineries were mass balanced to produce a product slate typical of a fuels refinery. Typical unit operating characteristics were taken from various sources <6> <7>. Crude slates were estimated and crude assays for typical Canadian crudes were used. The mass balance data was then used as a starting point for Linear Programming simulations using Mathpro's ARMS (Advanced Refinery Modelling System) LP.

In running the LP, a reference case was developed to simulate the current refinery operation. Gasoline quality as reported by CPPI was used to confirm the model. The various regulatory cases were then run to determine the changes in operating cost and capital additions which should be required relative to the reference case. This data was summarized and the results forwarded to MELP.

MELP forwarded the individual results to the refiners for review. The refiners expressed concern that the assumptions used to develop the models did not reflect the actual operating and market objectives of their refineries. As a result, they felt that the results did not accurately reflect their true costs. The refiners, with the exception of Chevron, agreed to meet with Kilborn to review the LP results and to provide insight into their operations. The refiners met with Kilborn and provided the following information:

• Crude slates and assays.
• Refinery block flow diagrams.
• Unit operating capacities and characteristics.
• Typical refinery gate gasoline quality.
• Minimum unit sizes for units which may be added.
• Driveability calculation method and specifications.
• Details on pipeline sulphur pickup and remedies at end of pipeline.
• Average product slate.
• Production constraints.

With this data, the models were rebuilt and run to confirm the current operation. Having confirmed that the models reflected the refinery operations and product slate, a reference case without MMT was run for each refinery. These cases were used as the reference case for calculating the cost changes to provide the gasoline quality of the CPPI offer and Cases 1 through 5. These cases were reviewed with each refiner and minor adjustments were made to correct product outputs. In addition to the costs identified by the LP, Kilborn estimated the costs for segregation of B.C. gasoline and the cost of process changes identified by the refiners which would not be identified in the LP. These costs included upgrades and additions to utility systems, tankage for purchased blend stocks, tankage for segregation of the B.C. gasoline, additions to distribution systems and capacity increases for individual components within process units which would not be identified by the LP.

The principle economic parameters used in the LP models are:

• The reference year is 1994.
• The target year for producing the RFG is 1998.
• 1 $ Canadian = 0.75¢ U.S.
• Refinery demands are based on actual 1994 demands.
• Refinery prices and costs are based on 1994, with no inflation to 1998.
• Marker crude, West Texas Intermediate $17.20 U.S./B.
• MtBE, Gulf Coast $39.90 U.S./B.
• Methanol, Gulf Coast $27.30 U.S./B.
• Natural Gas, Alberta Field Gate $ 1.25 U.S./B.

The following factors were used in computing annual capital charges for given capital investment:

• Location factor (relative to U.S. Gulf Coast): 1.18.
• Construction period: 1 year.
• Depreciation period: 7 years.
• Economic life: 15 years.
• After tax rate of return: 10% . <8>
• Future inflation rate: 0%/year.
• Property tax/insurance: 1% of (on-site + off-site investment).
• Land: 0.2% (on-site _ off-site investment).
• Combined Federal and Provincial corporation tax rate: 45.34%.

In preparing the base case analysis, the following assumptions were made:

1. All gasoline compositions are yearly average values.
2. RVP applies in Lower Fraser Valley (LFV) only, sulphur, aromatics, and benzene specifications apply to all B.C.
3. For parameters that are capped, hold at 115% of current values for each refinery after removal of MMT. <9>
4. For Alberta refineries, refinery gate gasoline is 40 ppm lower in sulphur due to sulphur pickup in the pipeline.
5. 100% of Chevron production is assumed to be sold in B.C. (2/3 in LFV).
6. 100% of Husky's production is sold outside of LFV.
7. Quality of gasoline sold by Alberta refineries in other market areas is not permitted to deteriorate.
8. Refineries will not be run in blocked mode to produce B.C. gasoline.
9. Current refinery gasoline is a weighted average of gasoline compositions from CPPI Report No. 94-5 and has been adjusted based on reported production for each refinery and test results for a survey performed with Environment Canada in 1994.
10. Current Refinery gasoline RVP is the LFV limit except for Husky.
11. All costs are applied to the B.C. volumes produced by each refiner.
    

The average B.C. gasoline quality was provided by MELP and represents the weighted average for all grades of gasoline sold in the province in 1993. <10>

In addition to the base case scenario defined above, B.C. MELP requested two additional scenarios be evaluated. These scenarios were identical to the base case except:>

Allocate the costs of producing the B.C. Cleaner gasolines across the total volumes produced at that B.C. specification.

Assume that the Refiner, which incurred the highest costs in the base case, would exchange its B.C. volumes with other refiners for equal volumes in other market areas.
The results of these later two scenarios were not reviewed with the refiners.

3.0 Results

In this section the results of the three analyses are described. For simplicity, these analyses are termed:

• Five Refinery Scenario.
• Full Volume Scenario.
• Four Refinery Scenario.

Each of these scenarios is discussed independently.

Five Refinery Scenario
The results of the modelling and discussions determined that the cost of producing reformulated gasoline to the specifications listed in Table 1.2 range from 0.11¢/L for the CPPI case to 2.6¢/L for Case 5.

The cost incurred to attain the CPPI offer are mostly due to RVP and benzene reduction. Two of the refiners also incur costs to reduce sulphur.

Of the costs incurred to attain Regulatory Case 1 composition sulphur reduction costs represent the largest cost while RVP reduction in LFV is the second and benzene reduction is third. The other parameters that were capped did not represent constraints as the LP was able to provide a solution without reaching these upper boundaries.

Case 2 represents a reduction of benzene to 0.8% from the Case 1. The cost of this is an additional 0.2¢/L.

Case 3 addresses a reduction of sulphur to 100 ppm compared with the reduction to 150 ppm indicated in Case 1. The cost to go directly to 100 ppm is 0.4 ¢/l more than the cost to go to 150 ppm as indicated in Case 1. The differential cost would be higher if the refiners were to go to 150 ppm and then to 100 ppm.

Case 4 combines the reduction of sulphur to 100 ppm and the reduction of benzene to 0.8 ppm. Case 4 relative to Case 3 shows that the benzene reduction cost is 0.1 ¢/l when sulphur is limited to 100 ppm.

Case 5 limits sulphur levels to 100 ppm, benzene at 0.8%, and aromatics to 22%. Comparison of this case with case 4 indicates that the additional cost to include aromatics reduction to 22% is 1.4 ¢/l. Table 3.1 summarizes these results.

Table 3.1

B.C. Cleaner Gasoline Options: Total For Five Refineries

Cost Estimates in $M Cdn

Full Volume Scenario
In this scenario the capital and operating costs were spread across the entire volume of gasoline produced in each refinery to B.C. specifications. The total volume used in the evaluation increased from 4001 million litres per year to 6656 million litres. The cost of the reformulated gasoline in this scenario ranges from 0.5¢/litre for the CPPI offer to 2.1¢/litre for B.C. case 5. Since the only change in this scenario is the volume of gasoline used the analysis of the previous scenario applies in terms of the reasons behind the cost for each case. Table 3.2 summarizes the results of this analysis.

Table 3.2

B.C. Cleaner Gasoline Options: Total for Five Refineries: Full Volume Basis

Cost Estimates in $M Cdn

Four Refinery Scenario
In the Five Refinery scenario above the analysis indicated that one refinery would incur substantially greater costs than the average for the industry. This analysis was undertaken to determine a possible cost if this refinery did not supply into B.C. but obtained its supply from other refineries in exchange for supplying those refineries in other market areas. The costs under this scenario are estimated to be approximately one third of the costs in the Five Refinery case. The costs range from 0.3¢/litre for the CPPI proposal to 1.3¢/litre for B.C. Case 5.

The Costs incurred for the CPPI offer are largely for RVP and Benzene reduction. The cost for these reductions is estimated as 0.3¢/L.

Sulphur and RVP reduction are required to meet B.C. Case 1. This would cost an additional 0.2¢/litre over the CPPI proposal.

To achieve the benzene reduction to 0.8% B.C. Case 2 is estimated to cost an additional 0.1¢/L.

Case 3 requires a further reduction in sulphur from Case 1. To achieve this reduction would cost 0.4¢/litre more than Case 1. It is noted that this cost is based on going directly from the present sulphur levels to 100 ppm. To achieve 150 ppm and then to go to 100 ppm would be more expensive.

Case 4 is a reduction of benzene to 0.8% from the 1.0% benzene level of Case 3. The cost to go directly to this level is estimated to be 0.1¢/litre more than Case 3.

The cost to go to 22% Aromatics is shown in Case 5. The additional cost for Aromatics reduction is estimated to be 0.3¢/litre relative to Case 4. Table 3.3 summarizes the results of this analysis.

Table 3.3

B.C. Cleaner Gasoline Options: Total For Five Refineries

Cost Estimates in $M Cdn

4.0 Discussion

The results of the study indicate that the average cost to produce a reformulated gasoline for the B.C. market could vary anywhere between 0.3¢/litre and 3.4¢/litre depending on the gasoline composition selected. As can be seen in the summary tables, there is a wide range in the costs for individual refiners for any given scenario. This results in competitive advantages for low cost refiners. In the Five Refinery scenario, the costs for four of the refineries have been reviewed with the refiners. The fifth refinery has not reviewed the evaluation of its refinery. These costs have therefore not received the same level of vetting as the other costs. This refinery represents approximately one quarter of the gasoline produced for the B.C. market. A change in the cost for this refinery could produce a significant change in the overall costs.

In the full volume case, the cost of producing B.C. gasoline is allocated to the volume of cleaner gasoline which is produced. Analysis to determine whether these costs can be recovered from B.C. markets or markets outside of B.C., where this cleaner gasoline is sold, is outside of the scope of this report.

In the Four Refinery Case, the underlying assumption is that the refiner incurring the highest cost can exchange volumes with other refiners and that these refiners will not extract a premium. While the high cost refinery's output quality would not change, the exchange partners quality will change as no two refineries produce identical gasolines. This could cause a change of the overall gasoline pool in other market areas.

Appendix A: Initial Study Results

B.C. Cleaner Gasoline Options: Total All Refineries

Cost Estimates in $M CDN

Notes:

[* 4398.78]

1. All limits are yearly average values.
2. RVP applies in Lower Fraser Valley (LFV) only; sulphur, aromatics, and benzene apply to all B.C..
3. For parameters that are capped, hold at 115% of current values for each refinery. (LP runs indicate that this limit was not reached).
4. For Alberta refineries gate gasoline is 40 ppm lower in sulphur due to sulphur pickup in the pipeline.
100% of Chevron production is assumed to be sold in the LFV.

Document Footnotes

<1> Implications of producing CARB Phase 2 at Chevron's Burnaby Refinery, March 1995, Kilborn Inc. with MathPro Inc.

<2> Oil & Gas Journal Dec. 19, 1994 pg. 61.

<3> CPPI Report No. 94-5.

<4> Overall average.

<5> O&G Dec. 19, 1994 pg. 59-61.

<6> Gilman R.H., Oil & Gas Journal, Sept. 3, 1990.

<7> Petroleum Refinery Process Economics, Maples R.E. Penn Well Books 1993.

<8> LP runs were performed with a 13% r.o.r., the results were then adjusted to 10% to be consistent with the CCME report.

<9> Bill C93 an act to regulate interprovincial trade in and the importation for commercial purposes of certain manganese-based substances. May 19, 1995.

<10> Using the study volumes and gasoline compositions, the current B.C. gasoline is estimated to be lower in sulphur, benzene, aromatics and olefines.

For More Information:

Environmental Quality Branch
Ministry of Environment
Government of British Columbia

PO Box 9341
Stn Prov Govt
Victoria, British Columbia
Canada V8W 9M1
http://www.env.gov.bc.ca/air