Appendix B. MTBE Ban Effects on California

California refineries are affected differently by an MTBE ban than other refineries because they produce a different type of gasoline, CARB reformulated gasoline, and, as a result, they have different equipment than most other refineries. The subject of an MTBE ban has been a topic of discussion and analysis in California for a number of years and the volume and quality of analysis done on a California MTBE ban exceeds that done for other U.S. regions. Moreover, recent surveys by the California Energy Commission (CEC) of refiners’ production plans when MTBE is eliminated provide a check on past analyses. This appendix summarizes how individual California refineries are affected by an MTBE ban, and estimates the aggregate impact on the California supply-demand balance.

California Refineries

California refineries producing CARB reformulated gasoline have to meet more stringent gasoline requirements than refineries producing Federal RFG, and they are configured differently to produce this unique fuel. Furthermore, these refineries are not subject to the Mobile Source Air Toxics Rule, as indicated in the rule.

A study on the impact of an MTBE ban in California was done by Purvin & Gertz and Stratco (alkylation process experts).¹⁷ The results of this study are summarized in the following excerpts from the report:

West Coast Refinery

The West Coast refinery case evaluations are based on a coking facility that processes 200,000 barrels per day of Alaskan North Slope crude and produces 100 percent CARB Phase II gasoline (CaRFG II), jet fuel, 30 percent CARB-compliant diesel and 70 percent low-sulfur diesel. While most West Coast refineries do produce some conventional gasoline, the objective of the current study is to demonstrate the impact of producing compliant gasoline within the constraints of an MTBE ban and [CARB] Phase III specifications.

The study revealed that assuming a MTBE ban and constant crude throughput, gasoline production would decrease for the refiners studied. Additionally, blending ethanol will increase the required amount of pentane removal from the gasoline pool in order to meet RVP specifications. Determining an economical disposition for the excess pentanes will be an additional challenge for refiners. While there is no simple processing alternative that can eliminate the shortfall in gasoline production caused by an MTBE ban, several options can be used to minimize the loss.

The options studied are outlined below. Each option involved shutdown of the MTBE unit and increases in the amount of alkylate available for gasoline blending.

• [Case II] Send the incremental i-C ₄ (previously MTBE unit feed) to the alkylation unit
• [Case III] Add…catalyst to produce more olefins for alkylation
• [Case IV} Separate FCC C₅ olefins for additional alkylate production

Table B-1 taken from the Stratco/Purvin & Gertz study summarizes their California refinery analyses.

In summary, this study shows gasoline production losses of 7-17 percent when MTBE is replaced with ethanol.

EIA has done an analysis using a 100-thousand-barrel-per-day refinery with a process configuration similar to that used in the Stratco/Purvin & Gertz study. The main difference between the EIA and the Stratco/Purvin & Gertz study is that EIA used a higher percent of normal butane (n-butane). Since n-butane has a very high RVP, EIA’s estimated volume loss due to reduction of light, low-boiling-point material is less.

The EIA cases that analyzed gasoline production to meet CARB summer specifications are summarized in Table B-2. The cases were designed to analyze several issues and options facing California refiners. The first issue focused on the volume of ethanol California refiners might use to meet customers’ needs. The RVP impact of ethanol on the gasoline blend is nonlinear, with the greatest impact occurring with only a few percent of ethanol. By the time 5.8 volume percent is reached, the adverse RVP impact has been fully absorbed. That is, the RVP impact on a 5.8-percent blend is about the same as the impact on a 10-percent blend. So, why wouldn’t CARB reformulated gasoline producers add 10 percent to help make up for the volume loss when MTBE is eliminated? The answer is that it can’t be done in most cases and still meet emission constraints. The issue arose with the particular refinery gasoline pool used in EIA’s analysis. In developing EIA’s Case 2, ethanol is at 5.8 volume percent and could only be increased to 6.0 percent before this particular blend failed the CARB predictive model NO x emission test. Case 4 demonstrates that it is possible to get to 7.0 percent ethanol if the refiner can also purchase additional volumes of alkylate and iso-octane, but 10 percent was not practically achievable.

The EIA results found that the volume of total gasoline production from this refinery could easily be restored to the MTBE volume levels with purchased alkylate (Case 5).

Purchased alkylate was increased enough to match the prior MTBE-blended gasoline volumes and no emissions constraints occurred. Furthermore, Cases 3 and 5 show that, with increased C 4 alkylate in the gasoline pool, more heavy ends could be added back into the mixture. The troublesome issue with the alkylate solution is whether adequate volumes of alkylate will be available for purchase from other U.S. refinery regions or internationally at a reasonable price during the startup of the MTBE bans.

California Aggregate Supply Impacts of an MTBE Ban in 2004

Using the information learned from the individual refinery analyses, estimates about refinery adjustments, volumes losses and means of making up those losses on an aggregate basis can be made. This section explores total losses to California production capability and compares EIA results to those of a recent analysis commissioned by the CEC.

California produces about 40 percent of the reformulated gasoline produced in the U.S. and consumes about 40 percent of the MTBE. As such, a California MTBE ban has ramifications that extend beyond that region. Not only will significant volumes of ethanol be needed, significant volumes of gasoline components such as alkylate will have to come from the Gulf Coast or import sources, which is described below.

The California Energy Commission has developed California supply-demand balances and described some of their analyses in a presentation in July 2001.¹⁸ In that presentation, a Commission representative stated, “ethanol provides little supply benefit during the majority of the year.” The presentation discussed how volume losses occur both due to less ethanol being used than MTBE and due to removal of pentanes to balance ethanol’s high RVP during the low-RVP season (8 months per year in California). The gap would need to be filled by importing key blendstocks, primarily alkylate. In discussing the role of alkylate for both volume and specification purposes, another California Energy Commission presentation indicated the alkylate volume needs from outside the California refineries might top 50 thousand barrels per day.¹⁹

In 2001, CEC also commissioned Stillwater Associates²⁰ to analyze the California supply situation under an MTBE ban. Stillwater results are shown in Table B-3. When 102 thousand barrels per day of MTBE is removed and replaced with 55 thousand barrels per day of ethanol, the volume of light, high RVP products (C 4’s and C 5’s) removed is 46 thousand barrels per day. In addition, 10 thousand barrels per day of heavy, low RVP components must be removed. Given the same level of crude oil throughputs (which is a reasonable assumption for California), the net result is a loss of volume of 103 thousand barrels per day, or a loss of 10 percent over what had been produced when MTBE was used.

CEC has not only commissioned studies on the subject, but has been monitoring refiner’s plans and actions to eliminate MTBE and find additional supply. CEC's recent assessment of the supply situation with an MTBE ban was that the Stillwater results are a very good reflection of the yield impacts that refiners are experiencing. CEC's current estimate is that, after refiners’ process adjustments, expansions, planned blendstock imports and shift from conventional to CARB production, there still remains a potential CARB supply shortfall of 3-5 percent.²¹

In order for EIA to compare its refinery yield impacts to the Stillwater study volumes, EIA annualized its results, using the Stillwater production estimates. The comparison is shown in Table B-4. The EIA estimates are slightly higher because, with the components available in the blending model, less light-ends adjustment had to be made, but a larger heavy-ends’ adjustment more than offset the light-ends’ adjustment, resulting in a slightly higher total. For EIA’s regional supply-demand balance, the Stillwater estimates are used because Stillwater, through the auspices of the CEC, had better information on actual refinery plans.

Using the loss of production capability on existing refinery capability as described in this section, results were extended to 2007 under both the High and Low Capacity Scenarios to analyze the U.S. supply-demand balances covered in the body of the report.

Appendix B: MTBE Ban Effects on California - Tables
Notes and Sources