RFG IN TIGHT SUPPLY AT INTRODUCTION, THEN OPT OUTS REDUCED DEMAND

July 31, 1995
Aileen Bohn, Charles Dale, Tancred Lidderdale, John Zyren Energy Information Administration Washington, D.C. John Hackworth, Joanne Shore OnLocation Inc. Alexandria, Va. Mike Burdette Comsis Corp. Silver Spring, Md. The first reformulated gasoline (RFG) season went smoothly, meaning the industry was able to deliver product where it was needed. The absence of major refinery outages, along with unexpected reductions in demand resulting from several areas leaving the program at the last minute,

Aileen Bohn, Charles Dale, Tancred Lidderdale, John Zyren
Energy Information Administration
Washington, D.C.
John Hackworth, Joanne Shore
OnLocation Inc.
Alexandria, Va.
Mike Burdette
Comsis Corp.
Silver Spring, Md.

The first reformulated gasoline (RFG) season went smoothly, meaning the industry was able to deliver product where it was needed.

The absence of major refinery outages, along with unexpected reductions in demand resulting from several areas leaving the program at the last minute, contributed to this smooth transition.

But the RFG market remains in a state of transition related to unexpected events that have affected supply and demand. An analysis of the RFG supply/demand scenario highlights the effects these events have had on gasoline markets.

PRIOR EXPERIENCE

The RFG program is the latest in a series of environmental programs, dating back to the introduction of unleaded gasoline in 1973, designed to make petroleum products less polluting.

The RFG program, which started Dec. 1, 1994, at the wholesale level, and Jan. 1, 1995, at the retail level, is the most complex of these programs. This is one of the reasons the transition to RFG was watched closely by both industry and government.

The RFG transition was viewed with concern because of previous negative experiences with new product introductions. The introduction of oxygenated gasoline in the fall of 1992 went smoothly because refiners had built up large stocks of methyl tertiary butyl ether (MTBE) and other oxygenates prior to the start of the program.1

In contrast, the introduction of low-sulfur diesel in the fall of 1993 saw a number of brief difficulties. Problems occurred in California and the Midwest for different reasons, but with the same result-temporary product shortages with associated price increases.

The low-sulfur diesel experience illustrated that transition periods for new products are vulnerable to unexpected supply disruptions.

RFG COMPLEXITY

The complexity of the RFG program is derived from:

  • Changes in production

  • Service of noncontiguous areas

  • Requirements of the delivery system.

At the start of the program, RFG was required in extreme and severe ozone nonattainment areas, and in areas that opted in to the program, mostly concentrated in the Northeast, Southern California, and around Chicago (Fig. 1)(74947 bytes).

Although the areas shown in Fig. 1 (74947 bytes) look solid, in some regions, attainment areas are interspersed with nonattainment or RFG areas, making delivery complicated. Also, the Northeast and Midwest are not self-sufficient, in terms of RFG production, and must rely on the Gulf Coast and imports for supply- While the areas to be served add some complications, the main features that make the RFG program more complex than prior product changes pertain to production and diminished flexibility in the delivery system.

RFG is a totally new product-not just a slight change or addition to an existing product, as in the case of oxygenated gasoline. Major refinery changes were required to produce RFG.

Regarding diminished flexibility in the delivery system, the RFG program requires a large increase in the number of gasoline types that must be segregated. The pipeline and storage capacities used for deliveries, however, have not changed.

This diminished flexibility can result in numerous difficulties, including:

  • A slowdown in some pipeline flows, perhaps by as much as 10%

  • A reduced number of products being stored at some terminals because of tankage limitations resulting from reduced fungibility

  • Diminished exchange flexibility because terminals may no longer carry the particular products an exchange partner needs

  • Increased likelihood of running out of products at a terminal

  • Increased supply times (i.e., increased odds of local areas experiencing temporary shortages).

INDUSTRY SURVEY

Having experienced difficulties with introduction of low-sulfur diesel, and considering the complexity of the RFG program, the U.S. Congress commissioned the Department of Energy's Energy Information Administration (EIA) to review preparations for the introduction of RFG. The goal of the review was to identify potential problems and recommend appropriate solutions.

Capitalizing on industry cooperation, EIA conducted a voluntary survey in September 1994 to serve as a cornerstone for this study.2 3 Out of 152 companies surveyed, 119 responded. A smaller number provided updates in November.

The EIA survey gathered data on industry plans for RFG production, imports, stock builds, and information on potential problems. EIA used the results of the special survey, along with its own estimates of demand and transportation capability, to develop a supply/demand balance for December and January, as shown in Table 1(61039 bytes).

December demand for total motor gasoline was projected at 7.683 million b/d, which was higher than December 1993's record demand of 7.661 million b/d, and consistent with the strong gasoline demand experienced throughout 1994. While the planned production and imports information provided by the survey respondents was adequate to meet this demand level, little excess capacity was reported.

The imports shown in the table were derived from the survey and, along with conventional imports, were in line with historical data. As indicated, the industry expected that less than 175,000 b/d of RFG would be imported, beginning in January. When demand, production, and imports were taken into consideration, the resulting stock builds were small.

In November, RFG production increased rapidly. About 77,000 b/d of RFG were imported in November in preparation for the Dec. 1 wholesale deadline. Of the imported RFG, 43% came from Canada, 43% from the Virgin Islands, and 14% from Venezuela.

Conventional gasoline stocks had been drawn down in October to make room for RFG. Although gasoline stocks, including RFG, were building in November, RFG stock levels at the end of November appeared relatively low.

Low November RFG stocks and projections showing little stock building during December or January left EIA to conclude that the supply/demand balance was tight, exposing the system to potential shortages.

It seemed that little excess RFG production capacity was available-even that known supply might not be available quickly in the event of an unexpected need. The balance was close enough that the loss of a single refinery in the Northeast or West Coast could create regional supply shortages.

The more likely problem was that the loss of flexibility in the delivery system would increase the potential for temporary local shortages during the transition.

A court-ordered delay and subsequent overturning in April of the Renewable Oxygenate Standard, the Environmental Protection Agency's requirement that as much as 30% of the oxygenates in RFG come from renewable sources, helped relieve an element of complexity for the transition.

THE ACTUAL EXPERIENCE

Much happened after the completion of EIA's survey of the RFG marketplace at the beginning of December.

The transition from the consumer perspective was indeed smooth, in the sense that the industry did not have difficulty delivering product to where it was needed. The RFG market, however, remains in a state of transition because of unexpected events that affected both supply and demand.

DEMAND

Finished gasoline demand in December was higher than anticipated, at 7.902 million b/d, and January demand was 7.157 million b/d. The relatively mild winter and stronger-than-expected economy boosted demand over what was projected prior to the RFG start-up.

RFG demand for this period, as represented by product supplied, is shown in Fig. 2 (46218 bytes).

Before Dec. 1, RFG production generally was used for converting conventional gasoline tanks to RFG service, and for some secondary stocking. After Dec. 1, regulations allowed only RFG deliveries in ozone nonattainment areas, so motorists were pumping mostly RFG after the first week in December.

Although overall gasoline demand was high, RFG demand decreased to less than the EIA study estimates in December and January. The most obvious reason for the lower-than-expected demand was an unforeseen number of regions that opted out of the program.

In December, 28 counties in Pennsylvania and 11 in New York and Maine, all of which had voluntarily opted into the RFG program, unexpectedly withdrew. This action reduced demand for RFG in January by an estimated 216,000 b/d. Additionally, EIA had estimated "spillover," or RFG sold into areas where it was not required, at 5%. Actual spillover was much less.

Thus, although RFG accounted for 26% of estimated gasoline demand in January, it averaged 1.9 million b/d, considerably less than the 2.5 million b/d projected prior to the RFG start-up.

Gasoline demand for first quarter 1995, at 7.5 million b/d, was 4.1% higher than demand in first quarter 1994. RFG represented 26% of gasoline demand, or 1.9 million b/d.

PRODUCTION, IMPORTS

In order to meet high gasoline demand this past winter, gasoline production ran higher than it had the previous winter, and much higher than the 5-year historical average. However, RFG production shifted as demand expectations and price incentives changed.

Fig. 2 (46218 bytes) shows that RFG production began in September, then increased to build stocks before the program start-up.

RFG production was expected to begin in earnest earlier than it did. One of the reasons for the apparent delay was the Colonial Pipeline break in October, which, by raising conventional gasoline prices relative to RFG, decreased the incentive to switch to RFG; production. Once the break was repaired, higher differentials returned, and RFG production surged.

The peak production level of 2.2 million b/d was reached during the first 3 weeks of December. Then opt-outs began to take their toll. Refiners began to cut output in late December because of reduced demand, decreased financial incentives, and uncertainty as to what gasolines would eventually be required in many areas.

Preliminary data indicate that total gasoline production in first quarter 1995 averaged 7.3 million b/d, 5.9% higher than first quarter 1994. RFG production averaged 25% of finished gasoline production.

Normally, the industry performs refinery turnarounds in March, relying on the January stock build to meet demand as production is temporarily reduced. High production figures for March, however, indicate that refinery turnarounds may have been delayed, or were very modest this year. In addition, stock levels stabilized, preventing stronger-than-normal price increases in the early summer.

Imports, as the source used to satisfy short-term supply and demand imbalances, are much more variable than production.

Total imports were low in 1993, as they generally have been since the Persian Gulf war, because of changes in world trade patterns. Imports increased to higher-than-usual levels during the 1994 driving season and, as a result of high demand, sufficient price differentials developed with Europe.

From November through February, except for one brief week in early December, imports were below the seasonal average. The decline in imports occurred, in part, because European refiners did not have sufficient incentive to make RFG, and supplies of conventional and oxygenated gasoline were ample.

Total gasoline imports have been modest in the first quarter of 1995, averaging 263,000 b/d, compared with 302,000 b/d in the first quarter of 1994. RFG accounted for 42% of total imports. With exports running slightly greater than last year's rates, net imports in the first quarter averaged 160,000 b/d.

STOCKS

Fig. 3 (14013 bytes) shows that total gasoline stocks have been well below the 5-year average since January 1994, except in November.

In August 1994, total gasoline stocks dropped well below the 5-year average as refiners prepared for oxygenated and reformulated gasolines. Stocks peaked at the end of November, when RFG had to be in place at bulk terminals.

Except during November, RFG did not increase total stocks, but simply replaced a portion of the conventional and oxygenated inventories. Total gasoline stock levels dropped again in December, as production cutbacks more than offset demand declines.

In January, conventional gasoline stocks began to build more rapidly than RFG stocks, perhaps to serve areas that had switched back to conventional gasoline.

Normally, stocks increase substantially in January because demand is low during this month and because refiners are producing high levels of heating oil for winter and simultaneously producing high levels of gasoline. But total gasoline stocks remained low at the end of the month, and through March.

RFG stocks increased fairly quickly once production began, attaining 40 million bbl by the end of November, when RFG supply from terminals was required. Stocks peaked at about 44 million bbl in mid-December, or about 24 days of supply at January-demand levels.

After that, RFG stocks gradually declined, as falling demand was more than offset by production cuts. RFG stocks stabilized in late February at about 40 million bbl, and have since fluctuated between 40 and 45 million bbl.

PRICE VOLATILITY

Similar to demand and production, prices also fluctuated during the RFG introduction, reflecting uncertainties and changing expectations.

Before Dec. 1, futures prices were available to monitor the market's reaction to the RFG start-up. Futures and spot prices provide a barometer of the pressures in the marketplace and, as such, play an important role in flagging potential concerns.

On a cost basis alone, RFG was expected to be 4-6/gal greater than conventional gasoline, mainly because of increased use of oxygenates-particularly MTBE. MTBE is sometimes used in conventional gasoline to improve the octane rating.)

As a result of simultaneous RFG and oxygenated gasoline sales, MTBE demand was expected to reach new peaks during the winter. During the third quarter of 1994, spot MTBE prices climbed to more than $1/gal, double what they were the previous year, as shown in Fig. 4 (48630 bytes). Ethanol prices also were on the increase in the fall.

Methanol, a major feedstock for MTBE production, was in tight supply last summer and was driving the increase in MTBE prices. Methanol prices increased sharply in June, rising from about 75/gal to about $1.20/gal by the beginning of August-triple what they were the previous year.

When Lyondell Petrochemical Co.'s Channelview, Tex., methanol plant closed for maintenance in August, prices increased further to $1.35/gal. Prices then declined slightly, but subsequently shot up to $1.78/gal following a fire in October at an Enron Corp. facility in Pasadena, Tex., which accounts for about 8% of U.S. methanol supply. Prices began to ease in November, then plummeted, starting in late December, in reaction to falling MTBE demand resulting from counties opting out of the RFG program.

MTBE prices followed the rise and fall of methanol prices, although the changes were less dramatic. MTBE prices began to recover in late March, however, as its blending value increased with rising gasoline prices.

Gasoline prices, as shown in Fig. 5 (58386 bytes), declined through the beginning of December, with the exception of a temporary decrease on Dec. 2 in response to Pennsylvania's opt-out.

Conventional and RFG prices increased in stages through mid-January, amid falling RFG stocks, a growing shortfall of finished motor gasoline stock levels, and increasing uncertainty regarding potential and actual opt-outs.

During this period, crude oil prices also increased, but with less volatility. Thus, the fluctuation in gasoline prices was driven mainly by product markets, not crude markets.

Prices generally eased after mid-January, as the opt-out situation stabilized. Opt-outs negatively impacted suppliers who built inventories for requirements that were later eliminated, but loosened what could have been a tight balance, resulting in generally steady prices through mid-March.

More recently, gasoline prices have risen sharply because of higher crude oil prices and concerns about low gasoline stocks and high demand during the driving season.

Fig. 6 (45698 bytes) shows the price differential between conventional gasoline and RFG. After reaching a peak of 8 on the futures markets just prior to the Colonial Pipeline break, the premium for RFG receded to 4-6, very close to EIA's earlier assessment of the additional cost of producing RFG.

The differential reached a low of slightly more than 3 at the end of December, following the opt-outs of nine additional counties in New York and two in Maine. After Jan. 1, the differential briefly strengthened on falling RFG stocks and production, until declining oxygenate costs reduced it to current levels of 0-2/gal.

OUTLOOK

In late April, refiners still faced some interesting challenges. Stocks were low, with finished gasoline stocks at the end of March representing 22 days of supply vs. the historical 5-year average of 25 days.4

Motor gasoline demand is predicted to be stronger this summer than in 1994. The high demand will be met by increased production, imports, and stock draws of conventional gasoline.

Refiners that can keep their plants running at high utilization through the summer will have to look at the tradeoff between gasoline profits and producing the accompanying distillate. This choice becomes more cumbersome in light of the fact that distillate will likely be selling at lower prices because of off-season demand.

During the last 2 years, the price spread between New York Harbor spot unleaded gasoline and West Texas Intermediate crude oil increased seasonally in March and April to 10 or 12/gal, dropping back several cents through the middle of the summer. The same spread in March of this year was still only a little more than 6/gal. However, by Apr. 21 the spread had widened to 14/gal.

In addition to normal market uncertainties, regulatory uncertainties are complicating refiners' and marketers' plans. The RFG opt-outs are creating large uncertainties in industry planning of production, delivery, and storage.

REFERENCES

  1. Dale, C., Hackworth, J., Shore, J. and Ostrich, J., "First oxygenated gasoline season shakes out differently than expected," OGJ, Oct. 25, 1993, pp. 66-75.

  2. Energy Information Administration, "The Energy Information Administration's Assessment of Reformulated Gasoline," Vols. 1 and 2, SR/OOG/94-02/1 and SR/OOG/94-02/2, October 1994, Washington, D.C., U.S. Government Printing Office.

  3. Energy Information Administration, "The Energy Information Administration's Assessment of Reformulated Gasoline: An Update," SR/OOG/94-03, December 1994, Washington, D.C., U.S. Government Printing Office.

  4. Energy Information Administration, "1995 Motor Gasoline Assessment," Petroleum Marketing Monthly, DOE/EIA-038005/05), May 1995, Washington, D.C., U.S. Government Printing Office.

Copyright 1995 Oil & Gas Journal. All Rights Reserved.