COlXffiADO'S OXYGENATED FUELS PHOGSAM: ECONOMIC EVALQATIOff OF THE FIRST TEAR Prepared for: Michael Shelby U.S. Environmental Protection Agency OPPE Washington, DC Prepared by: josfaua B. Epel Ari H. Kichelsea Robert D. Rove RCG/Hagler, Bailly Inc. P.O. Drawer 0 Boulder, CO 80306 Work, performed under contract #68-01-7469 and vork assignment #88-27. Kention of trade oases does not constitute an endorsement of any product flra. [oms] ------- Table of Contents Page EXECUTIVE SUMMARY B-l I. Background and Objectives of the Analysis.... B-l II. Results of the Analysis E-4 1.0 INTRODUCTION AND OBJECTIVES 1-1 1.1 Introduction 1-1 1.2 Objectives of this Report 1-2 2.0 BACKGROUND ON THE COLORADO PROGRAM AND OXYGENATED FUELS 2-1 2.1 The Colorado Program 2-1 2.2 Vhat Are Oxygenated Fuels? 2-3 2.3 Selected Issues la Implementing An Oxygenated Fuels Program.. 2-5 3.0 THE COLORADO PETROLEUM INDUSTRY 3-1 3.1 Gasoline Production and Distribution 3-1 3.2 Historical Market Shares and Prices of Ethanol and HTBB Blends - 3-6 3.3 Market Share and Prices of Ethanol, MTBE and Clear Gasoline During the Mandatory Program . 3-11 4.0 ECONOMIC IMPACTS OF THE OXYGENATED FUELS PROGRAM 4-1 4.1 Engineering Costs Incurred By Refiners and Distributors 4-1 4.2 Oxygenate Costs and Values.... 4-5 4.3 Denver and Regional City Rack Prices 4-12 4.4 Denver Retail Prices 4-26 4.5 Regional Comparison of Retail Prices. 4-36 4.6 Fuel Bconoay Penalty 4-39 4.7 Other Social Impacts and Costs 4-40- 5.0 SUMMARY OF ECONOMIC IMPACTS 5-1 BIBLIOGRAPHY APPENDIX A APPENDIX B ------- LIST OF TABLES AND FIGURES Figure E-l Table B-l Executive Sunary Hap of Colorado Regulation 13 Program Area Summary of Colorado's Oxygenated Fuels Program Costs Figure 2-1 Chapter 2.0 Kap oE -Colorado Regulation IS Program Area Figure 3-1 Figure 3-2 Pigure 3-3 Figure 3-4 Table 3-1 Table 3-2 Chapter 3.0 Colorado Monthly Gasoline Consumption Front Range Colorado Physical Supply of Refined Products Monthly Blended Bthanol Gas Consumption Bthanol, MTBB and Crude Prices MTBB and Bthanol Prices (Cents Per Gallon) Gasoline Sales Summary Chapter 4.0 Table 4-1 Annualized Coat of MTBB Related Bquipaent Expenditures Table 4-2 First Tear Costs Related To Bthanol Table 4-3 MTBB Blending Costs and Octane Value Table 4-4 Bthanol Program Impacts, Costs to Nev and Previous Consumers Figure 4-1 Rack Gasoline and W7I Crude Oil Prices Figure 4-2 Regional Regular Gasoline Rack Prices Figure 4-3 Regional Unleaded Gasoline Rack Prices Figure 4-4 Regional Premium Gasoline Rack Prices Table 4-5 Regional Price Differences Compared Pith Denver Table 4-6 Range in Regional Unleaded Rack Prices Compared to Denver Table 4-7 Statistical Comparison of Rack. Price Fluctuations Table 4-8 Regional Comparison of Changes in Rack Prices Over Tine Figure 4-5 Denver Retail Gasoline Price Survey Figure 4-6 Rocky Mountain Nevs Gas Price Survey Table 4-9 Summary of the Denver Retail Price Survey Prices in Cents Per Gallon Figure 4-7 Denver Regular Retail Gasoline Prices Figure 4-8 Denver Unleaded Retail Gasoline Prices Figure 4-9 Denver Premium Retail Gasoline Prices Figure 4-10 Regional Unleaded Retail Gas Prices Table 5-1 Chapter 5.0 Summary of Colorado's Oxygenated Fuels Program Costs ------- Colorado's Oxygenated Fuels Prograa: Economic Evaluation of the First Year EXECUTIVE SUMMARY I. BACKGROUND AND OBJECTIVES OF THE ANALYSIS The Colorado oxygenated fuels program, Regulation 13, v&s enacted to reduce carbon monoxide emissions frost motor vehicles la the non-attainment areas of the Front Range of Colorado (see Figure E-l). Regulation 13 requires that all gasoline sold in carbon monoxide non-attainment areas during the high pollution months of November through February contain a specified minimum 2 percent oxygen content by weight. The minimum oxygen requirement can be met4 by blending 10 percent ethanol (3.52 oxygen), 111 MTBE (2.OX oxygen) or other oxygenates that have obtained an Environmental Protection Agency valver. In the first year of the program, the oxygen content standard vas reduced to 1.5 percent and the mandatory blending period shortened to January and February, 1988 to allow the petroleum and oxygenate marketing industries sufficient time to meet program requirements. The blending of oxygenated fuels in all gasoline during the high pollution months has been projected to reduce ambient carbon monoxide levels by 8 to 14 percent. The lower oxygen content of air at high altitude (18 X lower than sea level) causes most motor vehicles to burn fuel inefficiently and create excessive carbon monoxide emissions. Blending oxygenates with gasoline compensates for the lack of oxygen in the high altitude areas and reduces motor vehicle tailpipe emission. Throughout the analysis and Rulemaking process that led to the development of the program, numerous concerns about the consequences of the program vere raised. These included: the costs that vould be borne by consumers and Industry; potential motor vehicle driveability and maintenance problems, which could result in a lack of public acceptance; the market shares oxygenates vould control at different oxygen requirements; potential decreases in ROG/Hagler, Ballly Inc. B-l ------- Figure E-l Map of Colorado Regulation 13 Progran Area Air Program area within heavy box. ------- gasoline fuel economy; the cost effectiveness of the program as a strategy to reduce carbon monoxide exceedances and others. The Colorado Department of Health and the trade association of the petroleum industry had economic analyses conducted prior to the program. These studies provided a large range of cost per gallon estimates of $.005 to $.08. The range in cost estimates are based on assumptions of the market penetration of a particular oxygenate, and whether other states in the region vould adopt similar programs and constrain the availability of oxygenates. In addition to being computed in advance of the program, these estimates are based upon projected engineering costs that do not reflect market behavior. This analysis vas undertaken to track the program before and during implementation vlth the focus upon the costs of meeting the requirements. Specifically this vork: o Tracked rack and retail prices, o Tracked and compared average retail and rack prices in Denver and other cities, o Reexamined actual incremental engineering costs for production and distribution, o Examines octane benefit derived from blending vith oxygenated fuels, o Tracked and addressed market penetration by type, o Examines the cost per gallon and cost per ton of the program, and o Provides related information on other selected aspects of the program. The first four elements provide alternative methods to examine the cost per gallon incurred by citizens and the cost to Industry of the program. RCG/Hagler, Bailly Inc. K-3 ------- II. RESULTS OF THB ANALYSIS Market Penetration The major suppliers of gasoline into Colorado met the program oxygen requirement by blending 8 percent HTBE in approximately 95X of the gasoline sold in the program area. A special sub-octane grade of gasoline vas produced to permit the blending of ethanol during the program. In October of 1987, the ethanol blended gasoline market had decreased to 2.1 percent, dovn from 20 percent in 1986, due to the elimination of a state tax credit for ethanol and other factors. During the program, ethanol had a market share of 5.3 percent. During the Regulation 13 rulemaking, members of the Air Quality Control Commission assumed that ethanol might capture an equal share of the market under a standard requiring 1.5Z oxygen. The most probable reasons for the limited market share actually experienced for ethanol Includes the short term of the program in the first year; the concerns of major marketers, retailers and consumers about the suitability of ethanol; and the constrained quantity and quality of gasoline to blend vlth ethanol. Production And Distribution Cost Estimates One approach used to evaluate the incremental economic impacts of the program vas to compute the Incremental blending and distribution costs associated vlth the program. MTBB. Strictly blending higher priced MTBE into base gasoline vould Increase the price of gasoline In Denver betveen about 2.0 and 2.4 cents per gallon. However, blending MTBE Into gasoline increases the octane, and by adjusting other components that offset much of the price impacts a standard 85 octane regular gasoline can be produced. The net effect is estimated to Increase production costs by betveen 0.4 and 0.8 cents per gallon. Adding HTBE reduces RVP, which can also further offset blending costs by the use of cheaper RGG/Hagler, Bailly Inc. E-4 ------- butanes to gain octane. The price Impact of this butane substitution vas not computed. Capital costs incurred by individual producers and distributers of HTBE blended gasoline ranged from no costs (obtaining gasoline through exchanges) to building or purchasing rail off loading facilities, piping to storage, new storage tanks and in line blending equipment. These equipment purchases have productive use lives of multiple years, and, moreover, can be used for other purposes during non-program months. To account for these factors, equipment costs vere annualized and apportioned entirely to the program for an upper bound estimate of 0.005 cents per gallon. For a central case estimate, these equipment costs vere allocated for year round use for an estimate of 0.0005 cents per gallon. Bthanol. The distribution of ethanol can require the replacement of filters and cleaning of storage tanks. Assuming a maximum of 20 stations switching from regular gas to ethanol for the program (based on the increase in ethanol gasoline volume)) these costs vere estimated to range from 0.03 cents (as a central case estimate) to 0.09 cents (as an upper bound estimate). The central case allocates the cost over one year, the upper case allocates the entire cost to the tvo-month program. The estimated impact of svitchlng to ethanol includes the change in price of ethanol related to any changes in the dominant MTBB gasoline price (estimated to be up to 1.3 cents per gallon). Bthanol users, vho prior to the program purchased blended gasoline vlth an 87.5 octane, purchased ethanol blends during the program vlth 85 octane because only subgrade gasoline vas available for blending vlth ethanol. The value of the octane impact vas computed to be betveen 2 and 3.7 cents per gallon for those individuals vho, used ethanol prior to and during the program. Nev ethanol users vere estimated to have a 1 to 2 cent cost reduction per gallon. As an alternative method of determining vhether the program significantly impacted lover gasoline prices, relative rack prices in Denver and other cities vere compared. The rack price data can be interpreted to shov either BGG/Bagler, Bailly Inc. E-5 ------- that Denver's prices declined relative to select cities or that conversly, there vas a cost Impact. However, the intrinsic, relative short-term fluctuations in rack, prices across regional cities undermines the statistical validity of this approach for detecting small short-term price impacts from any one cause, and it cannot be defensably used to Independently calculate program costs at the rack level. A retail price survey vas also conducted in Denver and historic Denver retail prices vere compared to other cities. As vlth rack prices, no definite conclusions about the program price Impacts can he drawn from this analysis.' Fuel Bconomy Impacts Reductions or increases in gasoline consumption occur vlth oxygenated fuels,» depending on the type of pollution controls on a motor vehicle. Using Colorado Department of Health assumptions on fuel economy penalties by oxygenate, pollution control type and existing vehicle fleet mix resulted in calculated average fuel mileage penalties of up to 0.22 percent, or a $0,002 per gallon reduction in value. Clear Gas Some motorists in non-program areas used clear gasoline that bad been trucked in from out-of-state. Clear gasoline vas generally sold at a price equal to or exceeding HTBE blended gasoline. Assuming 4.4 million gallons of clear gas sold per month at a price penalty of up to $0,013 per gallon (maximum HTBE price increment) this cost increment during the program vould have been $112,463. Other Impacts Administrative and planning costs by industry and government did occur, but vere not quantified. Similarly, the ethanol industry is reported to have experienced loss of market share attributable to a shortage of sub-octane gasoline, but the costs are not identifiable vith any certainty. HOG/Hagler, Ballly Inc. B-6 ------- Total Cost Of The Program The range of costs Identified are summarized in Table B-l. Total costs statevlde ranged from $1,013,481 (central bound estimate) to no more than $3,559,604 (upper bound estimate). The lover bound estimates are zero. The costs vere largely Incurred by Colorado residents in the AIR area (722), although residents in non-program areas may have Incurred costs due to the Colorado petroleum distribution structure resulting In most of the state converting to oxygenated fuel. These costs are included in both the central and upper bound estimates. The central average price increase per gallon attributable to the program is $0.0045. Cost Per Ton of Pollutant Removal The Colorado Department of Health has estimated that a 942 market share of 82 HTBB and 62 market share ethanol (102) reduced ambient Carbon monoxide levels in the Denver Metropolitan area from 82 to 112* or from 160 to 220 tons per day. Using state-vide central and upper case cost estimates, and applying the carbon monoxide reductions to five days a veek, the dollar per ton cost of the program vould be $154.49/ton (Central estimate) to $542.62/ton (upper-bound estimate) for an 82 reduction and $112.36/ton to $394.63/ton for an 112 reduction. RCG/Hagler, Ballly Inc. B-7 ------- Table E-l Senary of Colorado's Oxygenated Fuels Prograa Costs ($ 1988) Cost Central Dollars Upper HTBB Capital Equipment HTBB Purchase Octane Value Added Total MTBK Cost Ethanol Cleaning Costs Market and Octane Costs Total Ethanol Costs Clear Gasoline Fuel Economy Penalty Total All Gasoline $ 110,044 4,198,100 -3.358.480 949,664 1,017 62.800 63,817 $1,013,481 $1,031,546 5,037,720 -3,358.480 2,710,786 29,260 282,575 311,835 112,463 424,520 $3,559,604 Cost Impacts By Location Dollars Cost Central Upper AIR AREA - Total $ $763,837 $2,615,536 - S/Gallon $0.0046 $0.0159 - $/Household $0,868 $2.97 REST OF STATE - Total $ $249,644 $943,968 - $/Gallon $0.0039 $0.0148 - $/Bousehold $0,729 $2.75 * Total costs statevlde* Sales volume during the tvo-month program: ethanol 9,419,000; HUE 209,905,000; and Clear Gasoline 8,651,000. i'l l~ii BOG/Hagler, Bailly Inc. E-8 ------- 1.0 DttBODQCnOW AMD OBJECTIVES 1.1 BJTRODOCTIOH The Denver metropolitan area has been listed by the EPA as having the vorst carbon monoxide pollution in the nation. In 1986, the Denver metropolitan area exceeded the CO NAAQS thirty-six times, and the single highest exceedance of the standard In the nation was recorded In dovntovn Denver (Metropolitan Air Quality Council, 1987). To meet the Pederal health standard, carbon monoxide emissions in the Denver metropolitan area vlll have to be reduced by 50 percent. Eighty-five percent of the carbon monoxide emissions are created by motor vehicles. Therefore, strategies designed to reduce carbon monoxide must either reduce motor vehicle exhaust emissions or reduce vehicle miles travelled (Metropolitan Air Quality Council, 1987). The Interim CO State Implementation Plan for the Denver Metropolitan Area identified oxygenated fuels as the most effective carbon monoxide reduction strategy available to the metropolitan area to help reduce ambient carbon monoxide levels to meet the Federal health standards (NAAQS) and to avoid up to $30 Billion annually in federal sanctions. Oxygenated fuels are gasolines blended vlth a component or components containing oxygen: either alcohols or ethers. These fuels are asserted to be effective In reducing motor vehicle exhaust emissions by up to 34 percent (Colorado Department of Health, 1985). The carbon monoxide emission reductions due to oxygenated fuels are dependent on the type and concentration of oxygenate used and the emission controls present on a motor vehicle. The tvo oxygenates that vere given the most consideration in developing the Colorado oxygenated fuels regulation vere ethanol and MTBB. Analysis of the effectiveness of oxygenated fuels by the Colorado Department of Health estimated that a mandated program could reduce ambient carbon monoxide levels by 8-16 percent (Oxygenated Fuels Task. Force, 1986). ROG/Hagler, Ballly Inc. 1-1 ------- BPA and local air quality officials throughout the country are also examining the use of oxygenated fuels to reduce carbon monoxide in other non-attainment areas. In Arizona, oxygenated fuels may soon be mandated either by EPA in response to a court order or through bills pending in the Arizona Legislature. In Washington, D.C., several proposed revisions to the Clean Air Act require the use of oxygenated fuels in carbon monoxide non-attainment areas. Other proposed Federal legislation would require the blending of ethanol (an oxygenate) in a large portion of gasoline sold in the United States for both energy and air quality reasons. 1.2 OBJECTIVES OF THIS REPORT The objectives of this analysis are to provide improved estimates of the costs of the Colorado oxygenated fuels program based upon the actual experience. Prior to the oxygenated fuels rulemaking, economic analyses of the proposed Colorado program vere conducted for the Colorado Department of Health (BBC, 1987) and for the trade association of the petroleum Industry (EAI, 1987). Both analyses took similar approaches: they projected expected engineering costs of complying vlth different oxygen content requirements and estimated per gallon cost increases attributable to different market share scenarios. Interestingly, the results of those tvo studies differed dramatically. The State estimates varied from $.005 to $.035 per gallon vhlle the Industry estimates ranged from $.042 to $.08 per gallon. The results of both prior studies are subject to question as substantial acknowledged uncertainty existed In many of the estimates. Engineering cost estimates also often ignore mitigating behavior on the part of producers and consumers, which may result in overstated cost estimates. Subsequent to the completion of the program, EAI (1988) and Amoco Oil have produced revised estimates of the program costs (Denver Post, March 29, 1988) based upon comparisons of changes in the wholesale price of gasoline in Denver and other cities, from which a $.022 to $.047 per gallon price Impact is estimated, plus BOG/Hagler, Bailly Inc. 1-2 ------- they added an assumed fut economy decre-jes of three percent per vehicle to yield an estimated total cost of $14 million in the first year of the program. Because the Colorado program is the only mandated oxygenated fuels program in the nation, the Environmental Protection Agency, Office of Policy and Planning Evaluation, contracted for this analysis of the economic impact of the Colorado program. This analysis vas conducted prior to and during the actual implementation of the program (January and February, 1988) to obtain the most accurate and timely information on actual program costs. This analysis considers: o Engineering and production costs considering actual facility and cleaning costs as veil as component substitution in the production of oxygenated gasoline, o Changes, and valuation, of octane in gasoline, o Market penetration of the oxygenates in the Front Kange and Statevide, o The ability to reveal market price impacts using comparisons of rack and retail prices, and o The cost of fuel economy losses based upon the Colorado Department of Health assumptions. This analysis also briefly touches upon the Issues and evidence concerning the other social costs of the program. The remainder of the report Is divided into four sections: o Chapter 2 introduces the Colorado program: vhat oxygenates are and selected Issues of concern in implementing an oxygenated fuels program. o Chapter 3 revlevs the Colorado petroleum Industry, and market shares and prices for ethanol, MTBB prior to and during the program. o Chapter 4 estimates the economic costs of the program;through review of engineering costs, market prices and other social impacts including reduced fuel economy. o Chapter 5 summarizes the results and relates them to estimated tons of carbon monoxide reduction. HCG/Bagler, Ballly Inc. 1-3 ------- 2.0 BACKGROUND OH THE COLORADO PROGRAM AMD OXYGENATED POKLS 2.1 THE COLORADO PROGRAM Colorado's oxygenated fuels program (Regulation 13, Appendix B) vas established in June, 1987 (Code of Colorado Regulatlonst 5 CCS 1001-16, 1987). In the first year of the program the implementation period vas January 1 through March 1,1988. The first year of the program established a shorter program duration (tvo months) and lover oxygen standard (1.5 percent) than the requirements of the second and subsequent years to allov the petroleum industry time to adjust to the production and distribution of oxygenated fuels. In the subsequent years of the program, Regulation 13 requires all gasoline to contain a minimum of tvo percent oxygen content (by velght) as the optimal oxygen level that vlll achieve maximum CO reduction and encourage healthy competition betveen ethanol and MTBB. Program implementation In subsequent years vlll be November 1 through March 1, vhich Is during the high pollution season. Pursuant to Regulation 13, the Air Pollution Control Division established a mechanism for spot testing the oxygen content of gasoline at retail stations. The Division employed five personnel to obtain fuel samples, and one staff member to maintain records. The Division conducted 556 inspections and Issued 45 Notice Of Violations to retail service stations found to have failed to meet minimum oxygen content requirements (Colorado Department of Health, 1988). Oxygenated fuels are mandated only in the areas of Colorado vhere the CO NAAQS are violated in the Front Range of Colorado (Figure 2-1). Hovever, because of the gasoline distribution system in Colorado, substantial portions of the State appear to also have been impacted by Regulation 13 during the first year of the program. ROG/Hagler, Bailly Inc. 2-1 ------- Figure 2-1 Map of Colorado Regulation 13 Program Area KOG/Hagler, Ballly Inc. 2-2 ------- In addition to regulating the distribution and oxygen content of gasoline, the oxygenated fuels progran vas designed vlth the intent of olnlolzing potential drlveablllty problems that could be caused by the use of oxygenates (see Section 2.3). Regulation 13 vas predicated on state lav that requires all gasoline meet fuel volatility standards (ASTM D-439) to minimize drlveablllty problems. State lav exempts gasoline blended vlth ethanol from meeting ASTM D-439; hovever the base gasoline must meet ASTM specifications (see Section 2.3). In accordance vlth State lav, the Oil and Gas Inspector of the Division of Labor Is required to test and ensure that all gasoline sold in Colorado meets appropriate ASTM and octane requirements. The Oil and Gas Inspector checks fuel sold at refineries and terminals and conducts spot checks of retail service stations. If gasoline does not meet the appropriate ASTM or octane requirements, the Inspector can require that the fuel be removed from the market. 2.2 VHAT ABE OXYGENATED FUELS?1 Oxygenated fuels vere first used in response to the oil embargoes of the 1970s. In an effort to reduce demand for imported oil, ethanol usage vas encouraged through federal tax Incentives to substitute ethanol for gasoline. Oxygenated fuels also came into demand as octane enhancers In the early 1980s. The demand for all oxygenates as octane enhancers accelerated vhen the Environmental Protection Agency required that the levels of lead in gasoline be reduced. Prior to the EPA's lead phasedovn policy, lead vas utilized extensively as an octane booster in gasoline (1988). Studies conducted by the EPA and papers presented by the Society of Automotive Engineers initiated the examination of using oxygenated fuels as a pollution ^References to exhibits refer to documents produced for the Colorado Air Quality Control Commission Rulemaking hearings, June 18 and 19, 1987. ROG/Hagler, Bailly Inc. 2-3 ------- reduction strategy. The State of Colorado began Investigating the possibility of utilizing oxygenated fuels as a carbon monoxide reduction strategy In 1978, and conducted a series of studies that demonstrated use of oxygenated fuels appears to significantly reduce carbon monoxide emissions (Colorado Department cf Health, 1987). The Clean Air Act (42 USC 7401 et sec.) established limitations on products that can be blended vith unleaded gasoline. The restrictions established in Section 211 of the Act reduces the pool of available oxygenates to those that have obtained EPA approval. Currently, there are 8 oxygenates that possess EPA approval: ethanol, MTBE, the DuPont Waiver, Oxlnol, Octamix, TAME, DIPB and TBA. The composition of the available oxygenates all differ significantly. However, the oxygenates can be divided Into tvo general categories: o Alcohols; (ethanol and methanol) These include ethanol, TBA and the methanol based oxygenates (trade names; the DuPont vaiver, Oxlnol and Octamix) o Ethers: These include MTBE, TAME and DIPE (Renevable Puels Foundation, Technician's Manual, 1987). Bthers are derived from blending various petroleum feedstocks. Currently, the EPA vaiver permits the blending of a maximum of 11X MTBB by volume vith gasoline. 112 MTBB is the equivalent of 2X oxygen content by velght. TAMB and DIPB are permitted up to a 22 oxygen content (Renevable Fuels Foundation, 1987). Ethanol is produced through the fermentation of agricultural products (corn, wheat, milo etc.) to produce a fuel grade alcohol. The EPA.vaiver allovs the blending of up to 10Z ethanol (90Z gasoline, 10Z ethanol), which produces an oxygen content of 3.5Z. Ethanol is almost always sold at 10Z by volume to take advantage of Federal tax credits (Renevable Fuels Association, 1987). Methanol can be produced from almost any carbon source, such as crude oil, coal, biomass, etc. Currently, the least expensive method of producing ROG/Hagler, Ballly Inc. 2-4 ------- methanol Is utilizing natural gas as a feedstock. Methanol Is always blended vith other alcohols (cosolvents). The Oxlnol vaiver permits the blending of 9.6X oxlnol vith gasoline and produces an oxygen content of 3.5%. The Dupont Vaiver permits a maximum 3.72 oxygen from blending up to 5X methanol plus a minimum 2.5Z cosolvent alcohol. 2.3 SELECTED ISSQBS HI SfFLBMERTIMG AH OXTGEHATHD MKLS PROCTAM Throughout the development of Regulation 13 certain key concerns vere raised as possible impediments to the implementation of the oxygenated fuels program. This section briefly describes concerns and costs of the oxygenated fuels program in Colorado. These include Issues that affect both oxygenates and gasoline such as volatility control, and concerns that are more pronounced vlth oxygenated fuels than gasoline, such as the solvent nature of alcohols, phase separation and materials compatibility. Volatility Gasoline is blended to achieve certain quality control specifications. One important fuel quality specification is the ability of a fuel to vaporize (change from a liquid to a vapor). The rate of vaporization (referred to as volatility) vlll determine how veil a fuel vlll perform under varying conditions. Volatility guidelines are established by the American Society of Testing and Materials (ASTM) and are generally followed by gasoline refiners and blenders. These volatility guidelines include standards for Reld Vapor Pressure (RVP), distillation profile and vapor/liquid ratio (ASTM, 1986). Meeting ASTM volatility standards is critical to ensuring proper motor vehicle performance. The ASTM standards reflect temperature during seasons, climate and altitude. Gasoline is olended to meet conditions governed by the applicable standard in a geographic area. A fuel vlth a low volatility in cold veather may have trouble starting. A fuel vith high volatility in hot weather may vapor lock, and stall. ROG/Hagler, Bailly Inc. 2-5 ------- Blending alcohols vlth gasoline raises the volatility of the gasoline (toqrgenated Fuels Task Force, 1986). Colorado lav requires that all finished gasoline or gasoline blends (except ethanol) meet ASTM volatility standards. State lav requires that ethanol be blended vlth a gasoline that meets ASTM standards. An issue throughout the development of Regulation 13 vas vhether the exemption o£ ethanol from ASTM standards vould increase the likelihood of vapor lock and create drlveability problems for Colorado aotorists. The basis for the concern vas the possibility that gasoline blended vith ethanol that exceeds ASTM standards may gain significant market penetration during the non-vlnter vara veather months and increase the possibility of vapor lock. Hovever, in approving the Regulation 13, the Commission concluded that increased vapor lock vould not likely be a problem because other states have substantial ethanol penetration year round, and the use of oxygenated fuels vas mandated only for the cold veather months of November through February. In addition to mitigating drlveability problems, RVP controls can reduce evaporative emissions (Oxygenated Fuels Task Force, 1986). Gasoline vlth a high RVP produces more evaporative emissions, which contribute to the formation of ozone, than a lov RVP gasoline. The addition of ethanol to a gasoline vlth a high RVP (permitted by ASTH and Colorado Lav) can increase the tendency of the gasoline to form evaporative emissions (NAP, 1987). Hovever, ozone formation in the Denver metropolitan area is a summertime phenomena, and, in approving Regulation 13, the Commission concluded that even if ethanol achieved substantial year-round market penetration, the increased evaporative emissions attributable to ethanol vould not significantly exacerbate Denver's ozone problem. Materials Compatibility and Solvent Nature of Alcohols Alcohol fuels have characteristics that require they be transported, stored and blended differently than gasoline. These distinctive handling require- ments create expenses not normally incurred by gasoline marketers. Additionally, some fuel systems in early model motor vehicles (1960 models and ROS/Hagler, Ballly lac. 2-6 ------- earlier) had elastpmers (rubber parts) that vere not fully compatible vith alcohols. Introduction of alcohols into a fuel system of an early model motor vehicle could cause the elastomer to svell and force the premature replacement of the elastomer. Alcohols also have the ability to remove deposits that have built up in the inside of vehicle gasoline tanks, storage tanks or pipelines. Vhen alcohol blends are Introduced into a storage tank, the deposits can be released. The release of deposits can result in clogging of a vehicles fuel filter or result in having impurities in a fuel. Alcohols are not shipped through pipelines because of the their ability to release impurities. The Commission heard testimony that older cars (frequently ovned by less affluent motorists) vere susceptible to experiencing the potential problems that alcohol fuels are alleged to create. At issue vas whether mandating oxygenated fuels vould result in videspread use of ethanol and the resulting possibility that owners of older cars may be required to incur expenses not associated vlth the use of unblended gasoline. The Commission concluded that the risk posed by ethanol vas acceptable, and consumers vould have the opportunity to use KTBE if they chose. Phase Separation Alcohols, unlike gasoline and ethers, are soluble vlth vater. If there is vater In a gasoline storage tank, and alcohol is introduced into that tank, the vater and alcohol can mix and form a distinct vater and alcohol layer. This is referred to as phase separation. If phase separation does occur in a gas tank, the engine vlll not be able to burn the vater-alcohol layer, and the vehicle vill have driveability problems. There are a number of maintenance steps taken by blenders and retailers of ethanol to avoid phase separation and the release of impurities created by the solvent nature of alcohols. The maintenance practices (and costs) range from simply changing filters on pumps, to pumping storage tanks dry prior to ROS/Hagler, Bailly Inc. 2-7 ------- Introduction of blended fuels to cleaning and drying the tanks. The costs of these procedures are discussed In Chapter 4.1. BOG/Bagler, Ballly Inc. 2-6 ------- 3.0 THE COLORADO PETROLEUM INDUSTRY This section provides an introduction to the production and distribution of gasoline and oxygenates in Colorado, vhich serves to highlight those companies affected by the program. Additional background on the Colorado petroleum and oxygenate industries can be found In BBC (1987) and KAI (1986, 1987). The chapter also addresses market penetration and prices of ethanol and MTBB prior to and during the program. 3.1 GASOLENE PRODUCTION AND DISTRIBUTION Bight petroleum companies are the primary suppliers of an average of 101t340 barrels (42 gallons per barrel) per day of gasoline into Colorado in 1986 (1987). Figure 3-1 shovs the actual monthly sales of gasoline from 1986 through January 1988. In January 1988, the first month of the program, 2,750,000 barrels of gasoline were sold in Colorado. The Regulation 13 program area consumes approximately 72 percent of the gasoline sold in Colorado (BBC, 1987). Gasoline sold in Colorado is supplied either by tvo local refineries or is shipped through one of the four pipelines that have access to Colorado's Front Range (see Figure 3-2). The tvo local refineries are Conoco and Total, each of vhich is located in Commerce City. These tvo refineries produce 33,300 barrels per day (BPD) of the gasoline sold in the State. Both refineries have approximately 15 percent of the gasoline market in Colorado. Each of the four pipelines serves different refineries and move varying amounts of gasoline into Colorado (1986 estimates from EAI, 1987): o The Vyco pipeline serves the Amoco and Frontier Refineries and moves approximately 24,600 BPD into Colorado. o The Phillips pipeline serves Phillips Petroleum and Diamond Shamrock and moves 15,300 BPD. o The Chase pipeline serves Texaco and others and moves 14,400 BPD. RCG/Hagler, Ballly, Inc. 3-1 ------- Figure 3-1 Colorado Monthly Gasoline Consumption C <0 O X) > ID ofc o 170 160 - 150 - 140 - 130 - 120 - 110 - 100 - 90 7/85 10/85 1/86 4/86 7/86 10/86 1/87 4/87 7/87 10/87 1/88 Month/Year Source: Colorado Department of Revenue, 1988 ------- Figure 3-2 FRONT RANGE COLORADO PHYSICAL SUPPLY OF REFINED PRODUCTS 0 1986 0 UNITS: ANNUAL AVERAGE BPD 0 PRIMARY SUPPLIERS SHOWN SINCLAIR PL MOGAS 13500 JET 600 OIST 2700 SINCLAIR WYCO PL MOGAS 24600 JET 2250 DIST 6100 AMOCO, LARCO, FRONTIER MISC TRUCKS- MOGAS 200 JET 0 D1ST 400 FRONTIER, CRA, SINCLAIR TOTAL SUPPLY MOGAS 10 1340 JET 29450 DIST" 2 5440 TEXAS PANHANDLE PHILLIPS PL TRUCKS MOGAS 15300 40 JET 5700 0 DIST 3500p 40 Source: Energy Analysts International Inc., 1987 ROG/Bagler, Ballljr Inc. 3-3 ------- o The Sinclair pipeline Is a proprietary pipeline and moves 13,500 BPD. o Trucks aove only 600 BPO into the Front Range. Kthanol Production and Distribution The Coors Brewery in Golden, Colorado is the only existing source of ethanol production in Colorado. Coors can produce up to 120,000 gallons of ethanol a month from its beer vaste stream and yeast production process. To ensure product quality vhich meets industry standards, Coors adds a detergent and corrosion inhibitor package to their ethanol. The volume of ethanol production is seasonal, vith the least production occurring in December, January and February, the time frame of the mandatory program. Assuming its maximum production capacity, Coors vas capable of producing 25.7 percent of the ethanol used during the program, or 1.4 percent of the total oxygenate needed to meet the requirements of a 100 percent ethanol program. Of the approximately 65.4 thousand barrel per day ethanol production in the United States in 1986, 63 percent of the production capacity is located in the Midwest (EAI, 1987), with Archer Daniels Midland of Illinois having 50 percent of the nations' production capacity (Oxygenated Fuels Task Force, 1986). The three primary ethanol marketers in the program area (Bthanol Managment Company, Spruce Oil and Vestern Refining) obtain their ethanol either from Coors Brevery In Golden, Colorado, or by rail from Archer Daniels Midland. A lOOt ethanol penetration during the mandatory program vould have required approximately 6,387.0 barrels per day for the program area and 8,870 barrels per day Statevlde or 9.8Z and 13.6Z respectively of the national dally ethanol production capacity.* *Analysis of the volume of oxygenate necessary to meet the requiremen of the program must be based on the gasoline sold in the Front Range ar? However, blended fuel was sold throughout the entire State. RCG/Hagler, Bailly, Inc. 3-4 ------- MTBK Production, Pricing and Distribution MTBB is a petroleum based chemical produced from tvo refinery products, isobutylene and methanol. The price of MTBB is directly related to the price of crude oil, the price and availability of feedstocks, and the price of substitute octane blend stocks such as toluene. Vhlle HTBE prices generally track vith the price of crude petroleum, supply-demand imbalances of MTBB and required feedstocks result in price fluctuations. The overall demand for octane blend stocks has increased vith the phasedovn of lead In gasoline. Toluene, a petroleum based product that is not an oxygenate, Is another commonly used octane enhancer and a substitute for MTBB. Because of Its substltutability, the price of toluene constrains the price of MTBB. In the last fev years the demand for toluene as a chemical feedstock has increased its relative price giving HTBE an economic edge In enhancing gasoline octane (BBC 1987). MTBB is used as an octane enhancer in other areas of the country because of its high octane value (blending octane of 108-112) and lov volatility (RVP 8). In Colorado, small amounts of HTBB in lov concentrations (2-4 percent) vere blended prior to the program to boost gasoline octane. To meet the Colorado Regulation 13 oxygen requirement, MTBB vas blended vith gasoline In an.8 percent mixture. U.S. production capacity for MTBB in 1987 vas approximately 81.7 thousand barrels per day nationally. Approximately 91 percent of the production capacity Is in Texas, vith tvo companies, ARCO Chemical and Texas Petrochemical, possessing 68 percent of domestic MTBB production capacity (Colorado Department of Health, 1987). MTBB production^ vas projected to increase to 100,000 BPD by the beginning of 1988 (BBC, 1987). HTBB is also imported into the U.S. MTBE can be transported into Colorado either by rail or, when it is blended in gasoline, through a pipeline. Based on January, 1988 sales, a 1002 penetration of 8 percent HTBE during the mandatory program vould have required approximately 5,100 barrels per day for RCG/Hagler, Ballly, Inc. 3-5 ------- the progran area, and 7,096 barrels per day Statevide. A 100X MTBB program vould represent 6.22 and 8.7X respectively of the dally national MTBB production capacity. A 100X MTBB program (using January, 1988 gasoline volumes) using 112 MTBB vould represent 8.6Z and 11.9Z of the 1987 dally MTBB production. Retail Betvork There are approximately 3,210 service stations in Colorado, vith 1,589 located on the Front Range. Non-branded suppliers and bulk consumers account for-one third of total gasoline sales In Colorado. Among branded retailers in Colorado, Amoco had 13Z of the market, Conoco 12.8X, Texaco 10.3X, Phillips 6.72 and the rest Is divided among numerous other branded retailers (BBC, 1987). In the program area, ethanol vas sold in sixty non-branded retail stations. 3.2 HISTORICAL MARKET SHARKS AND PRICKS OF ETHANOL AND MTBB BLENDS gfhannl Of the tvo oxygenates available during the mandatory program, only ethanol has a history of substantial market penetration in Colorado. Historically, the periods of greatest ethanol sales growth and decline correspond vith the adoption and elimination of a Colorado five cent per gallon ethanol tax credit. The State of Colorado provided a five cent per gallon tax credit for ethanol from July of 1978 to June of 1986. Figure 3-3 shovs historical Colorado Department of Revenue Statevide sales of ethanol prior to and during the mandatory program and industry estimates of ethanol sales. During the program ethanol use represented 0.6 percent of U.S. production. Federal tax incentives lover the gross cost of ethanol so that it can be economically blended vith clear gasoline. Ethanol is generally priced so that the net tax cost of the blended product vill be tvo cents per gallon belov the rack price of gasoline. The tvo cent difference may be used to pay for some ROG/Hagler, Bailly, Inc. 3-6 ------- Figure 3-3 Monthly Blended Ethano Colorado ~1—I—I—I—I—I—I—1—I—I—I—1—I—I—I—T 7/85 10/85 1/86 4/86 7/86 10/86 Month/Year E Colorado Department of Revenue Gas Consumption data not available T—I—I—I—!—I—I—1—I—I—I—I 1/87 4/87 7/87 10/87 + Ethanol Marketing Sources ------- of the costs of selling ethanol, such as tank cleaning, to Increase the retailers' margin or passed along to the consumer. In the Denver area retail ethanol blended gasoline prices average about a cent less than other gasoline, but varies by retailer. Ethanol prices are site specific depending on the supplier, market, method of transport (rail or truck) and incentive programs. The delivered price of ethanol does not include the federal tax Incentive. OTBK There is virtually no history of MTBB sales in Colorado prior to the mandatory program. Small concentrations of MTBB (blended at 3-4Z) vere sold in the premium gasoline of some major retailers In the State prior to the mandate. Independent marketers blended MTBE in gasoline prior to the mandate. For proprietary reasons, concentrations and volumes of MTBE sold in the State are not available. MTBE is sold at the Gulf Coast price plus transportation. Figure 3-4 and Table 3-1 shov the Gulf Coast spot market price of MTBB. Transportation costs vary by company, shipment size and other factors. On average, approximately 8 to 10 cents per gallon must be added to the Gulf Coast price for rail transportation costs to Denver (Colorado Department of Health, 1987). Prior to the program, concern vas expressed about increases in oxygenate prices resulting from the Colorado program demand (EAI, 1987). Colorado demand for MTBE and ethanol vas small relative to production capacity and, according to industry sources (ARCO Chemical, the primary supplier of MTBB), did not have an impact on oxygenate prices. Statewide Colorado use of MTBB in January vas 8.0 percent of daily U.S. production. ROG/Bagler, Ballly, Inc. 3-8 ------- Figure 3-4 Ethanol, MTBE and Crude Prices (Cents per Gallon) Date 1987-1988 ~ Ethanol + MTBE ^ Crude ------- Table 3-1 MTBK and Kthanol Prices (Cents Per Gallon) Date Bthanoll MTBK2 November 24 85.00 72.50 November 30 85.00 72.50 December 8 85.00 67.50 December 15 85.00 65.00 December 22 86.00 64.00 January 4 86.00 62.00 January 11 87.00 58.00 January 18 87.00 63.00 January 25 86.00 70.00 February 1 87.00 70.00 February 8 87.00 70.00 February 15 87.00 69.00 February 22 87.00 68.00 February 29 90.00 69.00 1 Delivered Price to Denver excluding the federal tax incentive of 10 percent or 60 cents per gallon. 2 Gulf Coast Spot Market Price. 8 to 10 cents per gallon must be added for rail delivery to Denver. RCG/Hagler, Bailly, Inc. 3-10 ------- 3.3 HABKBT SHAKE AND PRICKS OF ETHANOL, MTBB AMD CLEAR GASOLINE DURING THB MANDATORY FBOfSAM The volume of a particular oxygenate used during the mandatory program is dependent on the percentage of the market the oxygenate possesses and the oxygen content requirement of Regulation 13. In the first year of the program, 8 percent MTBE vas used to meet the 1.5 percent oxygen standard. Ten percent ethanol (3.5X oxygen) vas blended vith gasoline to take advantage of federal tax credits. The sales of gasoline during the program months of January and February 1988 are summarized in Table 3-2. In preparation for and during the mandatory program, the four pipelines shipped gasoline blended vlth MTBB into Colorado for most of December and all of January and February.2 The practical result of MTBB being blended vith all gasoline transported into, and refined in, the State vas the virtual elimination of any gasoline suitable for blending vlth ethanol, as is normally done. To ensure the availability of a blending stock for ethanol, the Governor met vlth representatives of the refineries that service Colorado and arranged for the production of gasoline suitable for blending vith ethanol. Consequently, a special non-oxygenated "sub-octane" gasoline (82.5 octane unleaded regular, 84 octane regular) vas produced to allov the blending of ethanol. Three refineries, Conoco, Total and Sinclair, vere the only suppliers of sub-octane gasoline. The sub-octane gasoline sold for one cent less per gallon than unleaded gas. The rack prices of MTBB and ethanol just prior to and during the program are summarized in Figure 3-4 and Table 3-1. 2Several refiners and blenders began to produce and distribute oxygenated fuels by mid-December. The earlier start vas necessary to ensure that retail stations vould have oxygenated fuels by January 1, 1988. Regulation 13 required that the type of oxygenate blend be posted on pumps in the program area from January 1 through February 29. Gasoline vith ethanol vas already required to be posted. Prior to the program some stations may have carried HTBB blended gasoline vlthout knowing it. Less than three percent of the stations reported selling MTBB blended gasoline in the December 13, 1987 RCG/Hagler, Bailly Inc. retail price survey. RCG/Hagler, Bailly, Inc. 3-11 ------- Table 3-2 Gasoline Sales Sunary 1988 January/February Progiaa Area Statewide Area Gallons Z Gallons Z Bthanol 9,419,000 5.8 9,419,000 4.1 Clear Gasoline — -- 8,651,000 3.8 KT6E (1U) 3,672,000 2.3 3,672,000 1.6 HTBE (8Z) 148.488.000 91.9 206.233,000 90.5 Total Sales 161,579,000 100 227,975,000 100 Sources: Colorado Department of Revenue, Colorado Department of Health ROG/Hagler, Bailly, Inc. 3-12 ------- Ethanol Using Department of Revenue estimates, ethanol blended gasoline sales (In the Regulation 13 program area) of approximately 3,318,000 gallons in December and 4,530,500 gallons in January represent 4.1 and 5.8 percent, respectively, of the oxygenate market. This vould represent an ethanol sales Increased 36Z from December to January. However, ethanol marketing industry sources, contrary to Colorado Department of Revenue determinations, have estimated that ethanol sales for the program period actually decreased slightly based upon their sales records. No explanation for these different estimates has been determined by either source. An issue of Interest to air quality officials is vhy there vas not greater market penetration of ethanol during the mandatory program, and whether ethanol penetration vlll increase next year. There are a number of apparent explanations for the veak sales of ethanol during the program. Bthanol vas carried only by independent stations because the major gasoline marketers vere not prepared to make the adjustments necessary to accommodate ethanol. Ethanol, unlike gasoline or MTBB, is not fungible (it cannot be shipped via pipeline, and requires special handling) and the short period betveen adoption of Regulation 13 and implementation of the program vould have required changes in operating procedures. Some marketers also argue that consumers do not accept ethanol as a product and are unvllllng to risk consumer rejection of the oxygenate they market. Hovever, it is interesting to note that some of the major petroleum marketers in Colorado (Amoco, Conoco, Sinclair and Texaco) currently sell ethanol in other states. Another proposed explanation for ethanol's small market penetration Is the lack of an assured supply of gasoline suitable for blending vith ethanol, vhich may have constrained sales. Ethanol marketers vere restricted in the amount and quality of gasoline that vas available for blending during the months of December, January and February. Bthanol Is traditionally blended vith gasoline to produce a product vhlch has a higher octane number (tvo to ROG/Hagler, Ballly, Inc. 3-13 ------- three octane points) than gasoline. Selling a high octane product Is a marketing tactic of retailers of ethanol blends. The unavailability of clear gasoline during the program forced ethanol marketers to either use sub-octane gasoline or ship gasoline in from Vyoming or Kansas If they wanted to use clear gasoline to produce a final high octane product. At least one ethanol retailer discontinued selling ethanol blended product because of the unavailability of clear gasoline. Additionally* the sub-octane gasoline vas reportedly not alvays available in sufficient quantities for blending vith ethanol. Any of, or a combination of all of the above reasons may explain vhy ethanol did not gain a larger share of the oxygenate market. MTBB MTBB fulfilled most of the oxygenate demand for the months of December, January and February. During the program, an estimated 210 million gallons of HTBB blended gasoline vas sold In Colorado. Of the MTBB sales, approximately 3.7 million gallons contained MX MTBB, the remainder vas blended at the 82 level (Colorado Department of Health, May, 1988). Clear Gasoline Although most of the gasoline sold in the State is transported into the Pront Range, some non-oxygenated "clear gasoline" (not sub-octane) vas transported into the Western Slope and the eastern plains (geographic areas outside the mandatory program area) via truck during the program months. Product vas moved into the Vestern Slope from Utah and Nev Mexico. Approximately 3,000,000 gallons of clear gasoline vas sold on the Vestern Slope in January 1988. On the Front Range and eastern plains approximately 1,470,000 gallons of clear gasoline, 1.7 percent of statevlde gasoline sales vas brought in from Vyoming HCG/Hagler, Ballly, Inc. 3-14 ------- and Kansas by truck. Some of the clear gasoline vas used to blend vlth ethanol, and the rest vas marketed as oxygenate free gasoline. ROG/Hagler, Ballly, Inc. 3-15 ------- 4.0 BOOHOMIC IMPACTS OF THE OXYGEHATRD PPKLS PB0G8AM Videly varying economic Impacts of Colorado Regulation vere Bade prior to program Implementation. This study uses several methods and alternative data sources to quantify the actual industry and consumers economic Impacts of the program. This section presents the analyses of: o Capital equipment costs incurred by refineries and distributors, o Oxygenate costs and values, o Denver and regional city rack prices, o Denver and regional city retail prices, and o Changes In fuel economy. 4.1 ENGIHKERING COSTS IHCURRED BY REFINERS AND DISTRIBUTORS The estimated costs of Implementing Regulation 13 vere based, in part, on the Infrastructure costs Incurred by the producers and distributors of gasoline and oxygenates. These costs Included: o Building or refurbishing storage facilities, o Constructing rail unloading facilities, o Piping from unloading facilities to storage tanks and o Purchasing In-line blending equipment to enable refineries and terminals to blend and distribute oxygenated fuels. The suppliers of gasoline and blenders of oxygenates vere contacted to determine the actual costs of implementing the program. The information provided for this study will be presented in aggregated form to ensure the confidentiality of the proprietary information. All major, participants in the Colorado petroleum industry vere contacted and vere generally very cooperative in providing information on a confidential basis (See Appendix 1 for a list of contacts). RCG/Hagler, Bailly Inc. 4-1 ------- Costa To Refiners, Distributors and Retailers Related To MTBB Rail Off Loading and Blending Equipment. The costs Incurred by Individual suppliers for the purchasing and building of rail off loading and in-line blending equipment varied from no cost (suppliers who obtained blended gasoline through exchanges) to $750,000 (for an unloading facility, piping to a storage tank and In-line blending equipment). The aggregated cost for all identified rail off-loading and blending facilities totaled $2,925,000 (Conversations vlth gasoline and oxygenate distributors). Storage Tanks. The refineries and terminals that supply product into Colorado possessed different storage capacity. Some suppliers utilized available tanks* others vere required to rededicate existing storage tanks, and some suppliers vere required to construct nev tanks (the upper bound expense for constructing a nev tank vas $750,000). The aggregate cost for providing tanks to store KTBB and sub-octane gasoline totalled $2,100,000 (Conversations vlth gasoline and oxygenate distributors). Annualized Cost of Equipment. The above equlpaent Installed to meet the requirements of Regulation 13 generally has a life expectancy of 10 to 30 years. Therefore, the cost of the equipment should be annualized, and not charged exclusively to the first year of the program. further, such equipment can be used by a refinery or terminal year round for purposes other than blending oxygenates during the four months of the program. For the purpose of this analysis, ve have calculated the annual cost of complying vlth Regulation 13 in tvo vays (Summarized in Table 4-1): 1. As an upper bound on costs, amortizing equipment at 10 percent vlth a tax life seven years (Tax Reform Act of 1986) and assuming the equipment is useful only during the program; and ~ 2. As a central estimate, amortizing the equipment at 10 percent over 15 years and assuming the equipment can be utilized the entire year and therefore, allocating the cost of the equipment to the program in proportion to the months used for the program. ROG/Hagler, Bailly Inc. 4-2 ------- Table 4-1 Aimuallzed Cost of HIBB Related Equipment Expenditures Cost Of Equlpaent Upper Central Rail, Piping Blending $2,925,000 Storage $2,100,000 TOTAL: $5,025,000 $1,031,546/year $110,044/year Sources: Amoco Oil, Conoco, Dlanond Shamrock, Exxon, Prontler Oil, Phillips Petroleum, Sinclair Oil, Spruce Oil, Texaco Oil, Total Petroleum, Vestern Refining. Utilizing the upper bounds of the engineering costs attributable to the first year, and attributing the entire annual cost to only the program months (January and February) gasoline sales, the engineering cost per gallon of MTBS mixed gasoline is $0.0049. This upper bound cost vould be half this amount in a four-month program. Using the central case estimates, the per gallon engineering cost for tvo months vould be $0.0005. Of tbe tva prior studies, BBC, (1987) projected a capital cost related to using HTBB of $0,006 per gallon and BAI (1967} estimated HTBB capital costs (la total dollars on a seasonal basis) of $4.4 million. Meeting the 2.0 percent oxygen standard in the second year of the program could require some refineries and terminals to incur additional engineering costs. To avoid the logistical problems created by the program (see Section 4.7) the refineries and terminals that vlll be blending MTBB shipped by railcar vlll have to dedicate an additional 50,000 barrel tank for storage of HTBB. In addition to providing the additional storage capacity, the blender will have to take shipment of the MTBE in August of 1968, and not receive the benefit before November 1, 1988. Refinery operators have estimated that the cost of providing the storage for the next years program will exceed the amount spent in the first program year. Assuming all eight refineries are required to build new HTBB storage capacity, at the upper case of $750,000 each (for a total of $6 million), the RGG/Hagler, Bailly Inc. 4-3 ------- annualized central case estimate of per gallon engineering costs attributable to the 1988-1939 program vould be $.0017 (assuming statevlde monthly consumption of 115,500,000 gallons per month). Cleaning Of Tanks Related To Kthanol The standard procedure for the handling and distribution of ethanol is to ensure that storage tanks are clean and contain no water to minimize the possibility of phase separation and the release of deposits (because of the solvent nature of alcohols). The cost estimates conducted prior to the rulemaking established a range of service station maintenance costs of $25 to $4000. The actual costs of preparing a service station to handle ethanol tended towards the lover end cost estimates (summarized in Table 4-2). Stations Incurred costs of $20 to $49 per pump hose for changing filters and puap castings. There were approximately 60 service stations on the Pront Range that distributed ethanol blended gasoline during the program. Less than 20 of the stations started blending ethanol during the program. The stations that were blending vlth ethanol prior to the program therefore did not Incur added preparation expenses due to the rulemaking. The stations that incurred the greatest expense are those that required that vater/gasoline mixture be removed from their tanks. Vater from a tank must be treated as a hazardous vaste, and disposal of the liquid is the greatest cost in the cleaning process. The expense of drying a storage tank ranged from $90 to $250 per tank. Approximately 50 to 70 storage tanks were dried and cleaned vith a total cost range from $4,500 (50 x $90) to $17,500 (70 X $250) (Conversations vith ethanol blenders and tank cleaning firms). ROG/Hagler, Ballly Inc. 4-4 ------- Table 4-2 First Tear Costs Related To Ethanol Total Estimates Lover Upper Itea $ Per Unit # of Stations Bound Bound Filters $20-49 average 20 (4-12 $1,600 - 11,760 (per pump hose) pump hoses/Station) Tank Cleaning $90-250 Tank 50-70 $4,500 - 17,500 TOTAL: - — $6,100 -$29,260 Sources: Bthanol Management Company, Kubat Equipment and Service Company, Spruce Oil, Western Refining. Assuming that twenty service stations Incurred the range of cleaning costs and sold equivalent amount of blended gasoline as all stations selling ethanol blends, the per gallon cost of cleaning the stations vould have been an upper case cost of $.01 and the central case of $.002. Hovever, assuming that the same stations continue to sell ethanol year round, the costs of preparing a station are spread out over the entire year, and the per gallon cost Is reduced to $.0017 and $.0003 respectively. Of the tvo prior studies, one estimated that the velghted average cost of preparing a service station vould be $735.00. The other study estimated that service station preparation vould cost retailers $.01 per gallon. 4.2 OXYGEHAIE COSTS AND VALUES This section examines the production costs and octane values of blending MTBE and ethanol into base gasolines to meet the 1.5 percent oxygen requirement of the Colorado program. The gross costs of MTBE and ethanol are higher than the cost of the base gasolines they replace. This by Itself vould indicate higher production costs for gasoline blended vlth these oxygenates. Hovever, because they boost the octane rating, a higher octane gas, vlth a higher market value, ROG/Hagler, Bailly Inc. 4-5 ------- may be sold or other lover price octane components may be substituted to offset some or all of their cost. 4.2.1 MTBK Blending Costs MTBE delivered to Colorado during the program averaged between 70 and 80 cents per gallon. In relative terms, this vas 20 to 30 cents per gallon above gasoline during the program, or 140 to 160 percent of the cost of base gasoline. Relative prices are Important in analyzing costs because gasoline and HTBE prices change over time. The program price differential between MTBK and gasoline is in the same range as historical relative prices shown by BBC (1987). A study by Energy and Environmental Analysis, Inc. (1987), using new plant economics at current petroleum prices, projects MTBB prices to remain about 25 cents more per gallon than base gasoline, thus maintaining the same price differential. The change in cost from adding HTBE can be calculated from the prices and amounts of base gasoline and MTBE added per gallon of blended gasoline. Multiplying the price per gallon times the percent of that component in the finished product and then summing the costs vill yield the cost of the product. The cost of blended gasoline can then be compared to the cost of unblended gasoline. For example, in a gallon of blended gasoline, 92 percent, is base gasoline and 8 percent is MTBB. Assuming a price of 50 cents per gallon for base gasoline the component cost is $0.46 ($0.50 X .92) per gallon. Assuming the higher program price of 80 cents per gallon of MTBB, the cost is $0,064 ($0.80 X .08) per gallon. The blended gasoline cost per gallon vould therefore be $0,524 or 2.4 cents more per gallon than base gasoline if no other considerations are made. Assuming the program average price of $0.75 per gallon of MTBE, the blended gasoline cost vould be $0,520 or 2.0 cents per gallon more than base gasoline (see Table 4-3). These estimates of 2.0 to 2.4 cents/gallon serve as the central and upper bound estimates o£ program changes in production costs. The addition of MTBE increases the octane of the base gasoline by approximately tvo numbers, vhlch has a value to producers and consumers. One RCG/Hagler, Ballly Inc. 4-6 ------- Table 4-3 MTBE Blending Costs and Octane Value MTBE Blending Costa Blending Price Per Percent Of Cost Per Gallon Component Gallon Gallon I. MTBE Blending Costs Base Gasoline $0.50 x 0.92 MTBB $0.75 - 0.80 x 0.08 Total 1.00 Change in cost per gallon (Blended cost - $0.50) H. Octane Substitution Value Change in Octane tines Octane Value (2 x $0,008) TTT. Total MTBE Blending Cost and Octane Value Ipact $0,004 - 0.008 (added cost) $0.46 $0,060 - 0.064 $0.520 - 0.524 $0,020 - 0.024 (added cost) -$0,016 (Saving) Note: Assumes 50 cents per gallon for base gasoline and actual MTBE program prices o£ 75 and 80 cents per gallon for central and upper cost estimates. Octane substitution value is based on oil Industry figures of 0.8 cents per octane gallon. This is a conservative estimate of octane value and does not include additional octane value obtained through the increased use of butane. Actual octane value varies by refinery feedstock, equipment, and operation. RCG/Hagler, Ballly Inc. 4-7 ------- method that can be used to estimate producer octane value is to compare the difference betveen unleaded regular and premium gasoline. Their rack prices differ by approximately 6 cents vhlle their octane differs by 4 points (89 vs 85). Using this method, the octane value Is approximately 1.5 cents per octane number per gallon. Although this estimate is probably on the high side, because it Includes wholesale profits as veil as production costs. It demonstrates that octane has value. As illustrated, increased octane gasoline available from strictly blending HTBB could have been sold at a higher price or, as actually occurred, producers could maintain the octane at normal preprogram levels (85 unleaded, 86.5 regular and 89 premium). By substituting HTBB octane for other octane components, the cost of producing gasoline Is decreased by the amount of octane components substituted. This savings is the production value of octane. Octane costs vary by refinery feedstock and equipment. Refineries with higher octane costs vlll receive larger benefits of octane enhancers than refineries vith lover octane costs. The production cost to Increase octane by one number has been estimated at 0.8 cents per gallon (BBC 1987) and 1.0 cent per gallon (George Yogis, ARCO Chemical, AQCC testimony June 4, 1987). HTBB increases the octane of blended gasoline by approximately tvo numbers. Using a value of 0.8 cents per octane number, the octane value of HTBB through substitution of other inputs is 1.6 cents per gallon of gasoline. The 1.6 cents per gallon offsets the purchase cost of HTBB (see Table 4-3). Therefore, the net change in production cost of blending KIBE including octane value ranges from 0.4 to 0.8 cents per gallon. Confidential industry sources have confirmed these are reasonable estimates of the production costs of blending KTBE. Octane substitution benefits estimated here may be conservative as evidenced by the economic use of HTBB in other areas of the country and by the possible continued use of HTBB after the program by three companies in the Denver market. Another form of HTBE octane substitution Is also possible that vould Increase the blending valve. Additional octane value, other than the direct boost of HTBB, can be derived through butane substitution. Butanes are an abundant, ROG/Hagler, Ballly Inc. 4-8 ------- lov cost octane enhancer, but only limited amounts can be added to most gasolines because of their high RVP (55). Because HTBB lovers the vapor pressure o£ gasoline (RVP of KTBK =¦ 8, RVP of gasoline ° 14) vhen blended, additional lov cost butane can also be blended to Increase octane Instead of using more expensive octane components.^ The value of the substitution is termed the butane credit. BBC (1967) estimated the butane credit at 0.2 cents per gallon for 11 percent HTBB gasoline. The amount of credit depends on the price and availability of butane and the composition of the base gasoline. Because of the complexity and variability of refinery operations, butane octane credit vas not Included in the estimate of HTBB octane value. The Inclusion of butane credit vould increase the substitution value and lover the total cost of blending HTBB. Based on the amount of MTBE blended gasoline produced during the tvo-aonth program without acknowledging the value of octane, the oxygenate purchase coat of the program ranged betveen $4,198,100 and $5,037,720 or $0,020 to $0,024 per gallon of blended gasoline. This serves as an upper bound estimate on the blending cost components. Vhen adjusted for the value of octane, the cost of blending MTBE ranged from 339,620 to $1,679,240 or from $0,008 to a central estimate of $0,004 per gallon of gasoline. 4.2.2 Kthanol Costs Although ethanol vas marketed economically before the program, some consumers incurred program Induced costs as a result of the ethanol pricing strategy and base gasoline substitutions. Because of the fixed margin pricing strategy of ethanol blends, the price of ethanol follows the price of the dominant gasoline. If the price of other gasoline increases because of the cost of blending HTBB, the price of ethanol blend vould also increase. Gasohol consumers vould pay more even if the cost of producing ethanol blend had decreased. *The vlnter RVP of gasoline in Denver averages 14 and must be below 15 to meet ASTH requirements (Staff Statement to the AQCC Concerning Proposed Regulation 13, June, 1987). BOG/Hagler, Ballly Inc. 4-9 ------- Previous ethanol users vho stayed vith ethanol nay have Incurred Increased prices and reduced product value because the octane of ethanol blend during the program vas reduced as compared to before the program. Ethanol blend prior to the program had an octane rating tvo to three points higher than the same grade of non-ethanol gasoline. This vas considered a consumer selling point of ethanol blends. During the program, ethanol vas blended vith lover octane "subgrade" gasoline. The finished blend using subgrade vas the same octane as other gasoline. The program induced cost vas the 2 to 3 point reduction In octane. The subgrade vas provided at a slightly lover cost, from one half to one cent per gallon less than MTBE blended gasoline, and vas the only gasoline available for blending vith ethanol. The ethanol blend gasoline to unblended gasoline price margin before the program vas approximately 1.5 to 2.0 cents per gallon. The reduction in cost of the subgrade (relative to gasoline) alloved blenders of ethanol to maintain the price margin betveen gasoline and ethanol blend, thus, offsetting program costs and/or Increasing their profit margin. Maintaining the ethanol blend to gasoline price margin results in a cost to the consumer by the save $/gallon amount of the MTBB gasoline price increase. (The price increase Is in relative terms. For example, the retail price of gasoline declined during the program( but ethanol blend may not have declined as much as it vould have vlthout the program.) The consumer cost Incurred depends on the type of gasoline that vas purchased before and during the program (see Table 4-4). For example, assuming up to a $0.01 per gallon increase in the cost of gasoline because of blending HTBE, the retail prices of both HTBE and ethanol blends of gasoline vould Increase by up to $0.01 per gallon. Consumers that purchased ethanol blends before the program and during the program paid up to $0.01 due to the MTBE effect (line 4 - line 2 in Table 4-4). Consumers that svitched from gasoline to gasohol during the program paid $0.01 less per gallon (line 4 - line-1) due to the ethanol retail pricing strategy. If the change in price of blending MTBE vas zero, then the consumer change in price vould be zero. Previous ethanol consumers also experienced a reduction in octane. For the central estimate of this impact, a production cost of 0.8 cents per octane gallon (2.5 octane reduction) vas used. The producer value of 1.5 cents per ROG/Hagler, Ballly Inc. 4-10 ------- Table 4-4 Bthanol Program Impairs, Costs to Rev and Previous Consumers Fuel Costs $ per Gallon Octane Before Program 1. Gasoline 0.85 85 2. Ethanol Blend 0.83 87.5 Poring Program 3. HTBB Gasoline 0.85-0.86 85 4. Ethanol Blend 0.33-0.84 85 Program Induced Change in Costs 5. Ethanol Blend to Ethanol -2.5 Blend (line 4 - line 2) - Price Impact 0.0 to +0.01 - Octane Value Reduction 0.02 to 0.037 - Total Impacts 0.02 to 0.047 (added costs) 6. Switching from gasoline to Ethanol Blend (line 4-line 1) - Price Impact -0.01 to -0.02 - Total Impacts -0.01 to -0.02 (added savings) Assumes, Cor the purpose of sample calculation, preprogram unleaded prices of $0.85 and $0.83 vith an ethanol octane boost of 2.5 and maintaining $0.02 retailer price margin. Costs of the program are from changes in base gasoline costs, changes in MTBE gasoline prices, and loss of octane. The cost to previous consumers is the total of the cost increase in other gasoline (0.0 to 0.01) plus the loss in octane value. The cost (benefit) to new consumers is the cost reduction in ethanol blends. ROG/Hagler, Ballly Inc. 4-11 ------- octane gallon (2.5 octane reduction) vas used. The producer value of 1.5 cents per octane gallon vas used for the upper estimate of this Impact. The value of reduced octane ranged front $0.02 to $0,037 per gallon for ethanol blends as compared to the previous use of ethanol. Assuming an upper bound cost increase of HTBE OF $0,013 per gallon (calculated figure from equipment and blending costs), the total program cost for previous ethanol blend consumers ranged from $0.02 to $0.05 per gallon. The total program cost (benefit) to nev ethanol blend consumers ranged from $-0.01 to $-0.02 per gallon. 4.3 DENVER AND REGIONAL CITY HACK PRICES Gasoline rack (vholesale) price behavior in Denver vas analyzed for Impacts of the oxygenated fuels program. Rack prices reflect crude oil purchase costs, refining and blending costs, cost of the oxygenate, and vholesale marketing and distribution costs and profits. During the program period, Denver rack prices Include the cost of HTBE, except for sub-octane gasoline. Based on the actual engineering and other costs Incurred ve vould expect only a negligible or small change in rack prices attributable to the program (approximately zero to 0.013 cents per gallon). Estimates of price Increases from $0,034 to $0.06 vere made prior to the program by BBC and EAI). This section analyzes the actual price behavior of Denver and other cities to see If these price changes can be verified. 4.3.1 Denver Back Prices To distinguish program related price changes, adjusted for etude oil, other market factors and seasonal trends, previous and current program period price movements and changes in margins were examined. Figure 4-1 shows Denver monthly average rack prices by grade. A two-year time span was selected to examine historical trends in Denver rack prices. Prior to and during the oxygenated fuels program Denver rack prices declined and then held steady. RGG/Hagler, Ballly Inc. 4-12 ------- Crude oil Is at substantial component of the cost of gasoline. Vest Texas Intermediate (VTI) crude oil spot prices from January 1986 through February 1988 are also shovn in Figure 4-1. VTI vas recommended by the oil Industry as being representative of U.S. crude oil price movements. Gasoline prices basically follov crude oil costs. Market imbalances rather than changes in component costs are attributed to the difference in rack and crude oil price trends. Based on seasonal gasoline demand patterns, adjusted for crude oil price changes, one might expect Denver prices to rise In the summer vlth the hlgfi tourist demand in the rocky mountains and decline in the vinter. Denver seasonal rack prices do not consistently follov this trend. Prices during the vinter of 1986 vere falling sharply; prices during the vinter of 1987 vere rising; and prices during the program vinter of 1988 vere steady. The previous seasonal price trends do not explain program price trends. Despite large overall price fluctuations, the margins betveen grades in Denver have remained fairly stable. Regular leaded and unleaded prices remain very close vlth leaded gasoline slightly higher since January of 1987. Premium has been and continues to be priced 5 to 6 cents above regular and unleaded gasoline. Sub-octane gasoline used for blending vlth ethanol only became available in December 1987, in preparation of the program. The sub-octane vas 1.5 to 2.S points lover than standard octane or MTBB blended gasoline during the program and sold for half a cent to a cent less (the octane varied by source and time; the price by brand name). Based on octane value, the sub- octane vould have been expected to cost even less than it did. Increased production, storage and distribution costs of separate clear grades of gasoline in addition to the full line of MTBB blends vere cited for the small difference in sub-octane gasoline prices. 4.3.2 Comparison Cities Based on the changes in Denver rack prices there vas little or no cost increase of the program. Hovever, because of other concurrent market changes ROG/Hagler, BaiUy Inc. 4-13 ------- Figure 4-1 Rack Gasoline and WTI Crude Oil Prices Denver Month/Year ~ PRE + UNL ~ REG A SUB X WTI Source: Rock prices from Petroscan. Crude oil west Texas intermediate prices from the Oil and Gas Journal Energy Data Base. ------- (including crude oil and inventory changes) Denver rack prices should not be analyzed in isolation. Relative price analysis is necessary to determine if Denver prices significantly increased or decreased relative to other non- oxygenated fuel regions. Other city rack prices are compared vlth Denver for changes in relative prices, historical relationships, previous program period trends and actual program period price relationships. To have meaningful and accurate comparisons, other cities should track closely to Denver prices. Back price changes in Denver vere compared vlth rack price changes in Kansas City, KS; Salt Lake City, UT; and Tulsa, OR to determine if program price changes can be separately identified and if so, hov much prices before and during the oxygenated fuels program Increased or decreased relative to other markets. These cities vere selected for comparison based on gasoline supply and market similarities vlth Denver and based upon Industry recommendations. A partial list of the attributes sought for comparison regions Includes: a combination of refinery and pipeline supply; delivery from the same pipelines or refineries; similar market size; geographic locations in or Influencing the rocky mountain region; and regions vlth little or no current oxygenate use. Representatives from the oxygenated fuels and oil industries indicated that no other city is directly comparable vith Denver because of its unique market structure. Hovever, at the Colorado Department of Health, February 4, 1988 Oxygenated Fuels Program meeting, Industry representatives did agree that Denver prices should not be vieved In Isolation and recommended that Kansas City, Salt Lake City and a mid-vest petroleum producing city,such as Tulsa,vould be the most suitable comparison cities. Kansas City, Salt Lake City and Tulsa are connected by pipelines and/or the same refineries that supply Denver. Kansas City, Salt Lake City and Tulsa are roughly comparable in market size. The refinery at Eldorado supplies Kansas City by pipeline, ties into other mid-vest pipelines and supplies Denver through the Chase pipeline. Tulsa, which is in Petroleum Administration for Defense District 3 and Is part of "the Group" of interlinked oil producing cities, is considered representative of industry conditions because of its central location, supply of crude oil and extensive pipeline distribution system vhich Interconnects Tulsa vith other refineries and markets. Salt Lake ROG/Hagler, Bailly Inc. 4-15 ------- City Is In the rocky nountain region and is served by the same refinery (Amoco in Caspert Vyomlng) that supplies a substantial share of the Denver market via the VIGO pipeline. All of the comparison regions have little or no current oxygenate use. Other cities, such as Casper, Wyoming, are also connected to the Denver market, but vere not used for comparison because they differ significantly in market supply, size and/or structure. Each of the cities selected for comparison satisfies many of the criteria attributes. Hovever, no city is exactly comparable with the Denver market. Gasoline markets are unique, and prices can and do change Independently of other markets. Reasons for differential price changes, other than costs of the oxygenated fuel program, Include: differences In refining and distribution costs, changes in inventories, changing price margins (profits), market competition and supply and demand imbalances. One attribute that distinguishes Denver from most other markets of the same size Is its limitation in responding to supply-demand imbalances. Kansas City, Salt Lake City and Tulsa have the ability to import and export gasoline, vhile Denver does not. Denver is at the end of the pipelines and only imports gasoline. Octane numbers in Denver and the rocky mountain region are tvo numbers lover for the same grades of gas than for the rest of the country. Octane numbers In the rocky mountains are 85 for unleaded, 87 for regular and 89 for premium. This precludes effectively comparing rack prices in Denver to those in other cities outside of the region. 4.3.3 Back Price Changes in Comparison Cities The reason for examining rack prices in comparison cities is to separate price changes due to natural market forces from program related price changes. Seasonal movements, relative city price relationships and monthly price variations are analyzed. Identifying these relationships and changes in the price are necessary to be able to verify estimates of program price impacts. Over the past tvo years, vith notable exceptions, overall rack prices In the comparison regions have tended to move together, obviously follovlng major ROG/Hagler, Bailly Inc. 4-16 ------- movements In crude oil prices and widespread market Impacts. Figures 4-2, 4-3 and 4-4 shov the monthly average rack prices for regular, unleaded and premium gasoline In Denver and the three comparison regions from January 1986 through February 1988. Rack prices of the different grades of gasoline change in a closely related manner. Because of this, and in order to simplify the discussion, only regular unleaded gasoline, which comprises the largest share of the market (53 percent; BBC 1987) is addressed in the subsequent discussion. Tvo exceptions to the general trend of parallel movements in regional prices vere vhen prices diverged in the summer of 1986 and diverged even more dramatically In the fall of 1987. Regional prices converged in the vlnter of 1986-1987 and appear to be converging again in the vlnter of 1987-1988. In discussions vith the oil industry, no specific reasons could be given for the changes in these spreads or relative prices. Seasonal changes in prices and gasoline demand vere examined to see If they could explain variations in relative city prices. The volume of gasoline demanded in Denver Increases vith tourism in the summer and decreases In the vlnter. Prices vould normally be expected to track vith demand, hovever, rack prices have not consistently folloved this seasonal pattern in previous program periods. Prices fell rapidly during the vlnter of 1986 and generally rose over the vlnter of 1987. In the fall of 1987, prices fell sharply, but, during the program, prices in Denver vere steady and prices in tvo of the three comparison regions (Kansas City and Tulsa) vere rising. Only in Salt Lake City, vhere prices had peaked one to three months after the other regions, vere prices still falling. Seasonal price trends vere not consistent in explaining relative rack price variation. Considerable change has occurred in the relative rack prices among cities in the region. For example, during 1986 and the first half of 1987, Denver's rack price generally ranked in the middle of the comparison cities, but also ranged from the lovest prices (twice) to the highest prices (twice). In August of 1988, five months before the program, Denver rack prices were higher RCG/Hagler, Bailly Inc. 4-17 ------- Figure 4-2 Regional Regular Gasoline Rack Prices (Cents per Gallon) &A /' \ X V' n-—l—r—j—r —r "i i"—r ! T ..r-T__r 1/86 3/86 5/86 7/86 9/86 11/86 1/87 3/87 5/87 7/87 9/87 11/87 DEN Month/Year + KC O SLC TUL Source: Petroscan ------- c o G \ m +i c o u © u u Q. Figure 4-3 Regional Unleaded Gasoline Rack Prices (Cents per Gallon) A ' ^ V ^ .*-%? ,-V \ /p^ ^ y W ^ 'I 1 1 r—I 1—T —! 1 ,— f- , , , ! j 1/86 3/86 5/86 7/86 9/86 11/86 1/87 3/87 5/87 7/87 9/87 11/87 1/88 DEN Month/Year KC O SLC TUL Soun. : Pc L cos can ------- Figure 4-4 Regional Premium Gasoline Rack Prices (Cents per Gallon) Month/Year ~ DEN + KC O SLC A TUL Source: Pe t roscan ------- than all of the comparison cities. After this peak, Denver prices fell rapidly and then leveled off. Historically, the above type of relative price fluctuation is not unusual when comparing cities. This therefore Halts the ability to detect relatively small price changes and assign the difference to any one source of Impacts. In fact, to verify or quantify specific program Impacts, the regional and temporal variability in prices must be smaller than the price Impact to be Identified. Table 4-5 shows the inter-city difference In price as compared to Denver by month for 1986 through February 1988. Figures 4-2 to 4-4 and Table 4-5, show there are limited consistent relative price trends. Prices in each comparison city fluctuate from below and to above Denver prices over the course of a few months, apparently due to variability In local market pressures. The relative price changes can be quite large and do not follow simplistic patterns. The rack price spread between the comparison cities is highly variable and has ranged from a high of 13.86 cents per gallon (1/1986, Tulsa-lov and SLC-hlgh) to a low of 2.6 cents per gallon (8/1986, SLC-lov and KC-hlgh) for unleadc gasoline. Immediately prior to the program, the price spread was 12.1 cei per gallon (12/87, Tulsa and SLC). In the last month of the oxygenated fuels program, Denver prices were just a quarter of a cent higher than SLC and the spread between the four cities had narrowed to only 3.26 cents per gallon, a change In the spread of almost 9 cents per gallon. Month to month fluctuations in rack price in comparison cities as compared to Denver, are large. Historical variations in comparison city prices are much larger than the estimated impacts of the program. A comparison of the relative price differences in January of the program shovs that they are not consistent with previous January price spreads but, are veil within the range of observed historical price differences. It appears impossible to identify small (even the 3 to 6 cent pre-program estimates) short-term program related changes in rack price when there are larger non-program related price changes. Table 4-6 summarizes the variability in rack prices relative to Denver from KCG/Hagler, Bailly Inc. 4-21 ------- Table 4-5 Regional Price Differences Compared VI th Denver (Unleaded Gasoline, Cents per Gallon) Price Relative To Denver Date Denver Kansas City Salt Lake City Tulsa 1/86 71.15 2.82 9.20 -4.30 2/86 55.56 -0.72 11.12 -2.68 3/86 45.06 -0.46 4.01 -2.02 4/86 46.85 2.31 0.09 0.28 5/86 51.57 6.98 - 2.37 4.39 6/86 54.64 -4.18 4.52 -5.69 7/86 41.74 -2.54 11.53 -3.72 8/86 47.69 1.72 - 0.92 -0.40 9/86 48.02 -1.59 - 1.15 -3.14 10/86 43.85 0.46 1.21 -1.51 11/86 44.70 0.97 - 0.29 -0.50 12/86 43.38 1.31 - 0.45 -0.12 1/87 49.68 2.44 - 2.73 0.49 2/87 50.58 -0.24 1.96 -1.77 3/87 51.05 2.60 1.94 0.61 4/87 53.04 1.89 - 2.12 0.33 5/87 54.66 2.63 - 2.46 1.09 6/87 57.77 1.53 - 2.04 -0.19 7/87 61.56 0.72 - 3.01 -1.16 8/87 68.11 -6.17 - 2.53 -7.29 9/87 64.45 -8.51 3.92 -9.44 10/87 58.18 -0.58 6.59 -2.34 11/87 55.64 0.12 5.45 -0.99 12/87 50.91 -4.25 6.81 -5.29 1/88 50.04 -3.57 2.84 -4.86 2/88 50.03 -1.95 - 0.24 -3.26 Source: Petroscan for monthly prices. Positive prices are above Denver, negative price belov. ROG/Hagler, BalUy Inc. 4-22 ------- Table 4-6 Range In Regional Unleaded Rack Prices Compared to Denver (Price Difference In Cents per Gallon) Range in Prices Kansas City Salt Lake City Tulsa Jan 1986 - Feb 1988 Above Denver 7.0 11.5 4.4 Belov Denver 8.5 3.0 9.4 Range 15.5 14.5 13.8 Program Period Rack Price Relative To Denver Jan 1986 -2.8 9.2 -4.3 Jan 1987 2.4 -2.7 0.5 The range In prices represents Petroscan monthly price difference between Denver and each city. ROG/Hagler, Ballly Inc. 4-23 ------- January 1986 through February 1988. During this time, monthly price differences in the comparison cities have ranged from 9.4 cents per gallon below Denver to 11.5 cents per gallon above Denver. The one-month fluctuation in Denver prices during the period 1/86-2/88 averaged 4.2 cents per gallon and ranged from 0.01 to 15.69 cents per gallon. Comparison cities one month price changes relative to Denver averaged 2.5 to 2.7 cents per gallon and ranged from 0.02 to 12.5 cents per gallon. The tvo- month fluctuation in relative prices averaged from 3.4 to 4.6 cents per gallon with a range of from 0.03 to 13.9 cents per gallon (see Table 4-7). Again, this highlights the fact that large (relative to the projected program impacts) relative price impacts over short periods of time are common due to natural market forces suggesting that small relative price changes are not likely to be defensibly attributed to any one factor such as Regulation 13. Because short-term price changes in comparison city prices do not track exactly with Denver, evaluation of program Impacts based upon relative changes in prices are completely dependent on the time frame selected. During the program, Denver rack prices held steady in absolute terms and actually declined relative to tvo out of the three comparison cities. Kansas City and Tulsa prices Increased by approximately 1.6 cents per gallon relative to Denver, while Salt Lake City prices declined by 3.1 cents per gallon relative to Denver. An analysis based only on rack price changes in these four cities during only the program months would indicate program benefits (relative price decreases in Denver) from blending with oxygenated fuels rather than costs, but again this is likely to be a spurious unsubstantiated result. Comparing regional price changes over other time periods associated vlth the program yields a variety of different results and program costs (see Table 4- 8). A comparison of regional price changes between December 1987, when fuel blenders began program implementation, and February 1988, the last month of the initial program, indicates that Denver prices declined relative to two out the three comparison regions. Between November 1987 and February 1988 rack RCG/Hagler, Ballly Inc. 4-24 ------- Table 4-7 Statistical Comparison of Rack Price Fluctuations (January 1986 - February 1988) Kansas City Salt Lake City Tulsa One-Honth Average Change 4.7 4.4 4.4 Standard Deviation 4.0 4.1 3.9 One-Konth Relative Average Price Change 2.7 3.1 2.5 Standard Deviation 2.7 3.0 2.4 Tvo-Honth Relative Average Change 3.7 4.6 3.4 Standard Deviation 3.1 3.9 2.7 Three-Month Relative Average Change 3.6 4.7 3.3 Standard Deviation 2.6 3.9 2.5 KOTES: One-Nonth figure = changes In price spread between comparison city and Denver over a one-month period in absolute value. Two and three month price changes are calculated the same vay. ROG/Hagler, fiallly Inc. 4-25 ------- Table 4-8 Regional Comparison of Changes in Back Prices Over Time Change In Price Comparison Kansas Salt Lake Comparison Period Driver City City Tulsa City Average Jan-Feb 88 Regular 0.06 1.83 -3.13 1.71 + .14 Unleaded -0.01 1.61 -3.09 1.59 + .04 Premium -0.02 1.64 -3.14 1.65 + .05 Dec 87-Feb 88 Regular -0.82 1.76 -7.96 1.34 -1.62 Unleaded -0.90 1.42 -7.93 1.15 -1.79 Premium -0.88 1.45 -7.98 1.15 -1.79 Nov 87-Feb 88 Regular -5.52 -7.34 -11.33 -7.70 -8.79 Unleaded -5.61 -7.69 -11.30 -7.88 -8.96 Premium -5.59 -7.76 -11.48 -7.88 -9.04 August 87-Feb 88 Regular -17.99 -13.51 -15.82 -13.85 -14.43 Unleaded -18.08 -13.86 -15.79 -14.05 -14.57 Premium -18.15 -13.87 -16.05 -14.05 -14.66 Source: Petroscan, 1988. ROG/Eagler, Bailly Inc. 4-26 ------- prices In the three comparison regions declined approximately 1 to 3 cents more per gallon than the change In Denver prices. A comparison of regional price changes between August 1987, when market prices peaked, and February 1988 shovs that Denver prices fell by 2.5 to 4.0 cents per gallon more than the prices In the other comparison cities. 4.4 DENVER RKTATL IBICES Changes In costs that are common to producers, distributors or retailers are likely in large part to be passed on to the consumer in competitive markets through changes in retail prices. Although some companies may incur substantial individual .costs that cannot be passed through to consumers, retail prices often reflect changes in costs to the industry as a vhole. Changes in Denver retail prices are of Interest In evaluating industry costs and consumer price Impacts of program. Retail prices may also Indicate if there vere different Impacts by type of blend. Unfortunately, it is difficult to obtain timely, reliable retail price information and impossible to find retail price Information by blend. One of the fev available sources of Denver retail prices Is the Rocky Mountain News, vhlch conducts a weekly price survey of ten Denver stations. However, blends are not identified in the survey and the sample size is too small to provide reliable price comparisons. For these reasons, Hagler, Ballly Initiated its own retail price survey to track gasoline prices by blend during the program. The Denver Retail Price Survey RCG/Hagler, Ballly Inc. conducted a bi-monthly survey in the Denver-metro area to track changes in retail gasoline prices, by grade and by blend, prior to and during the oxygenated fuels program. The Denver-metro area is the single largest retail market affected by the program with approximately 60 percent of the affected population (Denver Regional Council of Governments). A stratified sample of 106 retail stations vas selected to be surveyed based upon the Colorado Oil and Gas Inspectors list of 760 retail stations in the RCG/Hagler, Ballly Inc. 4-27 ------- Metro-Denver area. The survey includes retailers in Denver, Aurora, Commerce City, Littleton, Bnglevood, Lakevood, Wheat Ridge, Arvada, Federal Heights, Thornton, Westminster, Northglenn and Boulder. Retail prices by gasoline grade (regular, unleaded and premium) and by identified oxygenate blend (MTBE or ethanol) vere collected every tvo veeks from December 13, 1987 through February 28, 1988. Only self service prices vere collected to eliminate the vide variations in full service profit margins. The final survey captured approximately 14 percent of the Denver- metro market. Results of the Denver Price Survey Denver retail prices for both blends and all grades of gasoline declined during the oxygenated fuels program. Figure 4-5 shovs the average price of MTBB and ethanol blends for regular, unleaded and premium gasoline by survey period. Prices fell as oxygenated blends vere phased in during December, Jumped by approximately of 3 cents per gallon in old-January and then fell to belov preprogram price levels. A comparison vlth movements in rack prices does not explain the sudden Increase then decrease of Denver retail prices in January. Similar retail price movements are also observed in the Rocky Mountain Nevs survey, as summarized in Figure 4-6, although the Rocky Mountain Nevs Survey did not examine price movements by blend and only sampled ten stations in the Metro-area. The Rocky Mountain Nevs retail survey prices differed from the Denver retail price survey by from +2.5 to -1.5 cents per gallon. The Denver retail survey revealed a high correlation betveen station location and price. Stations located in close proximity almost alvays had comparable prices, regardless of the price level or oxygenate used, vhile prices betveen locations varied videly. The existence of many local competitive markets underlines the importance of having a large sample size and a videly distributed survey. RCG/Hagler, Ballly Inc. 4-28 ------- c o o cn 0 +» c o u © L. Q. n o O Figure 4-5 Denver Retail Gasoline Price Survey MTBE and Ethanot Averaged Prices 12/13/87 1/3/88 1 /16/88 1/30/88 Survey Date Unleaded ~ Regular + Source : RCG/Hagler, Bailly, Denver Retail Price Survey 2/14/88 Premium 2/28/88 ------- Denver retail prices declined prior to and overall during the oxygenated fuels program. During the program period (January 1 through March 1), the average price of both oxygenate blends declined by 1.92 cents per gallon for regular, 2.58 cents per gallon for unleaded and 2.29 cents per gallon for premium (see Table 4-9). A decline of from 1.6 to 2.7 cents per gallon vas also observed immediately prior to the program. To measure consumer price impacts of the program vith retail prices, changes in retail prices should be directly linked vlth changes in rack (industry) prices. Over the December through February period, rack prices only declined by about tvo cents per gallon, less than the 4 to 5 cent drop in retail prices. However, this smaller change in rack prices follovs a sharp drop in rack prices throughout the fall. Hypotheses about the difference betveen program period retail and rack price movements include: that retail prices lag behind changes in rack prices; that inventories built up at the retail level; and that retail competition increased, resulting In reduced retailer profit margins. In either case, changes (decreased) in retail prices are larger than the changes (decrease) in rack prices and therefore cannot be used to reject above engineering costs estimates of the relative small program costs. Other retail market changes occurred during the program, vhich also made it difficult to distinguish program related impacts. Regular gas (vithout oxygenated added), vhich has traditionally been priced belov unleaded gasoline at the retail level, vas at some retail stations priced the same as or above unleaded gas during the program. Regular gas has been and Is priced slightly higher at the wholesale level. The pricing change brings the regular-unleaded retail price spread in line vith the rack price spread. Changes in gasoline grades have also ocurred. Texaco has discontinued distributing regular leaded gasoline and is selling a mid-grade (87 octane in Colorado) unleaded gasoline In its place. The midgrade gasoline vas excluded from the survey. Differences In Blended Gasoline Prices There vere significant differences betveen the prices of oxygenate blends. Figures 4-7, 4-8 and 4-9 and Table 4-9 shov the average prices of MTBE and ethanol blends by grade of gasoline. Bthanol blends vere lover in price for RGG/Hagler, Bailly Inc. 4-30 ------- Figure 4-6 Rocky Mountain News Gas Price Survey Blends Not Identified 94 93 92 91 ~-v C 90 0 "a o> 89 \ ¦3 c 88 ® u 87 0 u 'lI a. 86 0) o e> 85 84 83 82 81 Source: ~ Regular Denver, Rocky Mountain News Survey Date ------- c o o o» N a c o u N_X o u u Q_ o o o 86 Figure 4-7 Denver Regular Retail Gasoline Prices By Blend 85 84 - 83 82 - 81 - 80 1/3/88 1/16/88 1/30/88 Survey Dote + Ethanol 2/14/88 2/28/88 ~ MTBE Source: RCG/Hagler, Bailly, Inc., Denver Retail Price Survey ------- c o *0 a> B) +> C o u © u *C 0. a o O 90 89 - 88 - 87 - 86 - 85 - 84 - 83 - 82 Figure 4-8 Denver Unleaded Retail Gasoline Prices By Blend 1/3/88 1/16/88 1/30/88 2/14/88 2/28/88 Survey Date ~ MTBE + Ltnonol Source: RCG/Hagler, Bailly, Denver Retail Price Survey ------- c 0 0 u> \ n +> c o u © u r £L n o 0 101 100 99 -1 98 -I 97 - 96 - 95 - 94 H 93 - 92 Figure 4-9 Denver Premium Retail Gasoline Prices By Blend 1/3/88 1/16/88 1/30/88 Survey Dote ~ MTBE + Ethanol 2/14/88 2/2B/88 Sourc RCC/llagler, Bailly, Denver Retail Survey ------- Table 4-9 Snuy of the Denver Retail Price Survey Prices In Cents Per flallnn Pre-Pragran Program Average Price By Survey Date Price Price Grade Change Change and Blend Dec. 13 Jan. 3 Jan. 16 Jan. 30 Feb. 14 Feb. 28 Dec. 13-Jan.3 Jan.3-Feb.28 REGULAR 84.28 82.66 85.28 83.93 82.14 80.74 -1.62 -1.92 MTBE * 82.90 85.46 84.07 82.10 80.84 * -2.06 Ethanol 83.90 81.21 84.21 83.10 81.97 80.17 -2.69 -1.04 Observations 85 86 88 102 103 103 unleaded 87.71 86.13 89.56 87.40 85.29 83.55 -1.58 -2.58 MTBB * 86.22 89.90 87.54 85.30 83.67 * -2.55 Ethanol 87.26 85.59 87.52 86.43 85.17 82.83 -1.67 -2.76 Observations 89 89 91 106 106 106 PREMIUM 101.07 98.34 100.26 99.10 97.21 96.05 -2.73 -2.29 MTBE * 98.41 100.46 99.26 97.34 96.31 * -2.10 Bthanol 98.50 97.40 97.57 97.04 95.50 92.70 -1.10 -4.70 Observations 58 57 59 69 69 69 Source: ROG/Eagler, Ballly Inc. Denver Retail Price Source, The gasoline grade average is for all blends in the sample * Less than 2X of the stations reported having KIBE in the December 13 price survey. BOG/Hagler, Bailly Inc. 4-35 ------- all grades. Premium gasoline had the largest price spread betveen blends ranging from 1.8 to 3.6 cents per gallon. The price spread betveen HTBB and ethanol blends vas about the same for regular and unleaded gasoline. The spread betveen unleaded gas blends Increased In mid-January to about 2.4 cents per gallon along vlth the Increase in prices, decreased to about 0.1 cents per gallon in mid-February as retail prices fell and then increased to about 1.0 cents per gallon by the end of February. While a difference in blend prices vas to be expected because of the vay ethanol blends are priced, reasons for the changes in the price spreads are uncertain. Ethanol blends are generally priced approximately 2.0 cents per gallon less than other gasoline at the wholesale level. Retailers have the option of passing on this savings, minus additional ethanol related costs such as tank preparation, or increasing their retailer profit margin. One consequence of this pricing policy is that the rack price of ethanol blends vlll follov the rack price of other gasoline. If HTBB gasoline prices increase so vlll rack prices of ethanol blends. Increased competition In the retail market may pressure prices closer together, reducing profit margins and narrowing the spread. Ethanol blend retailers have indicated that they price in competition vlth the stations In their local area, maintaining their profit margin If possible but, If competition demands It, they vlll lover prices passing some of the savings on to the consumer. The conpetltive local pricing observed In the Denver retail survey provides evidence of this type of behavior. 4.5 REGIONAL COMPARISON OF RETAIL PRICKS Retail prices in Denver vere compared vlth retail prices in Kansas City, Salt Lake City and Tulsa to determine If prices before and during the oxygenated fuels program Increased or decreased relative to other cities. These cities vere selected for comparison based on industry recommendations of comparability and similarities in market structure, a combination of refinery and pipeline supply, service by the same pipelines supplying Denver, RCG/Hagler, Ballly Inc. 4-36 ------- geographic location and market size (see Section 4.3 for sore detail on comparison city selection). There are very fev sources of timely retail price information available. Retail information collected by the U.S. Department of Bnergy and published in the Monthly Refinery Report is not available until three to five months after the survey period. Retail unleaded gasoline prices for the four comparison cities for January 1986 through February 1988 vere available from the Oil and Gas Journal. This data provides a good indication of price trends, but it Is not as reliable in providing precise prices because the data is adjusted using changes from base retail price levels. More than tvo years of data vere used to evaluate historical retail price relationships between cities. Figure 4-10 shows the monthly unleaded retail- prices for Denver and the three comparison cities. Retail prices over the past tvo years have followed the same general trends in all four regions reflecting major fluctuations in the price of crude oil. Kansas City has usually had the lovest retail prices, followed by Tulsa, Denver and Salt Lake City. Prices in Denver and Salt Lake City Increased relative to Kansas City and Tulsa during the summers of 1986 and 1987. After the 1986 increase, Denver prices tracked at a level more closely with Salt Lake City than with Kansas or Tulsa. This can be attributed in part to a July 1, 1986 six cent per gallon tax Increase in Colorado that brought Denver retail prices nearer to the level of Salt Lake City. After tracking more closely vlth Salt Lake City prices in August of 1987, Denver prices started falling more rapidly, diverging from Salt Lake City prices. Salt Lake City prices were still falling during the program but, remained above Denver prices. Kansas City and Tulsa price levels compared to Denver also fluctuate over time. The fluctuations in intercity retail prices make it impossible to quantify impacts that may be attributable to the oxygenated fuels program. Retail price changes do not consistently follow the wholesale cost of gasoline on a month-to-month basis. Comparisons of retail and rack prices in each region yielded widely fluctuating price margins that did not follow a ROG/Hagler, Ballly Inc. 4-37 ------- Figure 4-10 Regional Unleaded Retail Gas Prices c o o o> » c © u © t,/ 0) a O t—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—t~ n—i—i—i—r 1/86 3/86 5/86 7/86 9/86 11/86 1/87 3/87 5/87 7/87 9/87 11/87 1/88 3/88 ~ DEN + Source: Oil and Gas Journal Energy Data Base Date SLC O TUL KC ------- consistent pattern over time. The fluctuations In retail-rack price margins and changes in the relative city retail prices make It virtually impossible to identify program related retail costs with this data. 4.6 FUEL ECONOMY PENALTY Throughout the Regulation 13 rulemaking there has been concern that the program could Increase gasoline consumption and create additional expenses for consumers. As recently as March 29, 1988 Denver newspapers reported a Q ^trj y petroleum Industry analysis vhich estimated a mileage penalty of 55.6 million, y «,j ks An analysis of the literature on oxygenated fuels and gasoline consumption by the Colorado Department of Health concluded that the type of pollution controls on an automobile determines the effect oxygenated fuels has on gasoline consumption (Colorado Department of Health, 1987). Using the Health Department figures, fuel economy losses are calculated in relation to fleet composition, the mileage penalty associated vlth a particular control- technology and market share of oxygenates during the program. As Table 4-10 illustrates, the fleet mix and Colorado Department of Health assumptions on fuel economy changes shovs an average fuel economy change of less than one percent. Using the January and February statevlde gasoline consumption figures of 329.4 million gallons, an average price of about $0.87 per gallon and the vorst case fuel penalty scenario of 0.22X, the cost to consumers vas $420,000. Equlvalently, one could consider each gallon of gas as, on average, having 0.22* or a $0,002 reduction In value during the first year of the program. BCG/Hagler, Bailly Inc. 4-39 ------- Table 4-10 Fuel Econoiiy Changes X Changes In Fuel Bconoay Control Technology Z of Fleet MTBB Kthanol Non-Catalyst Catalyst Closed Loop Veighted Average: 34Z 41Z 25* 100X o.ox +1.3X -IX-3X -.22X to + .28X -0.3X +0.3Z -0.3S -.051 Source: Estimates of percent change in fuel economy from the Colorado Department of Health, 1987. 4.7 OTHER SOCIAL IMPACTS AND COSTS Industry Administrative Costs In January, 1988 approximately 8.4 million gallons of MTBE vas transported into Colorado either by rail or, when blended vith gasoline, via pipeline. Gasoline containing MTBE vas transported through the Chase and Phillips pipelines (30 percent of the Colorado market). The remaining 5.88 million gallons (70 percent of total MTBE sales) vas transported to Colorado or Wyoming for blending in approximately 200 rail cars (30,000 gallons per car). The Incremental rail costs are incorporated into the HTBE production cost estimates in Section 4.1. Hovever, gasoline marketers vho obtained MTBE by rail vere inconvenienced by the logistical difficulties of obtaining shipment o£ MTBE. The gasoline distributors vere confronted vith the unavailability of MTBE shipments and vere occasionally forced to borrov MTBE from other distributors until a shipment arrived. ROG/Hagler, Ballly Inc. 4-40 ------- Govermental Coats The Colorado Department of Health and the Oil and Gas Section of the Department of Labor Incurred costs enforcing Regulation 13. The Health department purchased nev equipment to test for oxygen content and hired staff take fuel samples and to respond to inquiries on their telephone Hotline. Costs Outside The Program Area Almost all of the gasoline sold in Colorado contained oxygenates* Any costs associated vith the program vere borne by all Coloradans. However, soae motorists in non-program areas used clear gasoline, vhlch had been trucked In from out-of-state (see Section 3.3) adding transportation costs over the normal delivery procedures. Clear gasoline generally sold at a price equal or exceeding HTBE blended gasoline (Conversations vith industry representatives), rather than belov HTBE blended gasoline, as might have occurred vithout the program. Assuming 8.6 million gallons sold during the program at a price penalty of up to $0,013 (maximum HTBE price Increase) results In an added cost during the first program year of $112,000. If there is a strong demand for clear gasoline in future years, petroleum marketers should be able to provide the product. Maintenance Costs Testimony given prior to and during the rulemaking established the possibility that use of oxygenated fuels would result in some drlveabillty problems and maintenance expenses (clogged fuel filters, deterioration of rubber parts). The Department of Health Hotline, and an Informal tracking of fuel related repairs by the Automobile Association of America did not reveal any quantifiable increase In maintenance costs caused by the use of oxygenated fuels. HOG/Hagler, Bailly Inc. 4-41 ------- Potential Increases In Ozone Testimony was presented prior to and during the rulemaking that a large year- round penetration of ethanol could increase hydrocarbon emissions, and exacerbate the existing summer time ozone problem. Analysis by the Colorado Department of Health has concluded that even vith a 100 percent penetration of ethanol (exempted from ASTM) during the non-vlnter months, the overall Increase of evaporative emissions vould increase ozone from the current highest recorded levels of 0.13-0.15 parts per million to 0.15-0.17 PPM (Colorado Departaent of Health, 1985). BOG/Hagler, Ballly Inc. 4-42 ------- 5.0 Sumaary of Economic Impacts Table 5-1 summarizes the range of costs estimates of Colorado's first year Regulation 13 program. Total costs statewide ranged from $1,013,481 (central bound estimate) to no more than $3,559,604 (upper bound estimate). The lover bound estimates are zero. Program costs varied by oxygenate blend. MTBE blends comprised about 95X of the gasoline sold in the program area. The costs were largely incurred by Colorado residents in the AIR area (72X), although residents in non-program areas may have Incurred costs due to the Colorado petroleum distribution structure resulting In most of the state converting to oxygenated fuel. These costs are Included in both the centred and upper bound estimates. The central average price increase per gallon attributable to the program is $0.0046, the cost per program area household Is $0.87. The economic impacts estimated in this costing analysis, which are based upon Incremental capital equipment and blending costs are not rejected by careful examination of rack prices in Denver and other cities. No substantive Increase In Denver rack prices relative to comparison cities can be defended. In fact, if price trends are analyzed in the cities recommended by the petroleum and oxygenate industries (Section 4.3.2), Denver's prices during the program months appear to decline relative to those cities. However, it is Important to note that the intrinsic, relative short-term fluctuation in rack prices across regional cities due to local market forces is of such magnitude that small short-term impacts due to a program like Regulation 13 are unlikely to be defensibly quantified using this approach. Therefore, the recent estimates by EAI (1988) are likely to have little or no statistical validity. In fact, alternative analysis of their data indicates numerous alternative conclusions, all with little or no statistical power. ROG/Hagler, Bailly Inc. 5-1 ------- Table 5-1 Su—niy of Colorado's Oxygenated Fuels Prograa Costs <$ 1968) Dollars Dollars/Gallon Cost Category* Central Cost Kstlaate Upper Bound Estimates Central Estimates Upper Estimates HTBB Capital Equipment MTBE Purchase Octane Value Added Total MTBE Cost $ 110,044 4,198,100 -3,358,480 949,664 $1,031,546 5,037,720 -3,358,480 2,710,786 0.0005 0.020 -0.016 0.0045 0.0049 0.024 -0.016 0.0129 Bthanol Cleaning Costs Market and Octane Costs Total Ethanol Costs 1,017 62,800 63,817 29,260 282,575 311,835 0.0003 0.0066 0.0069 0.009 0.03 0.039 Clear Gasoline 0 112,463 0 0.013 Fuel Economy Penalty 0 424,520 0 0.002 Total All Gasoline* $ 1,013,481 $3,559,604 0.0045 0.016 Cost Impacts By Location Dollars Area and Iapact Central Case Estlntes Upper Bound Estimates AIR AREA - Total $ - ^/Gallon - $/Bousehold $763,837 $0.0046 $0,868 $2,615,536 $0.0159 $2.97 REST OF STATB - Total $ - $/Gallon - $/Household $249,644 $0.0039 $0,729 $943,968 $.0148 $2.75 * Total costs statevide. Sales volume during the tvo-month program: ethanol 9,419,000; HTBB 209,905,000; and Clear Gasoline 8,651,000. RCG/Hagler, Bailly Inc. 5-2 ------- Cost Per Ton of Pollutant Keaczal The Colorado Department of Health has estimated that a 942 market share of 82 HTBE and 62 market share ethanol (102) reduced ambient Carbon monoxide levels in the Denver Metropolitan area from 82 to 112, or from 160 to 220 tons per day. Using state-vide central and upper case cost estimates, and applying the carbon monoxide reductions to five days a week, the dollar per ton cost of the program vould be $154.49/ton (Central estimate) to $542.62/ton (upper-bound estimate) for an 82 reduction and $112.36/ton to $394.63/ton for an 112 reduction. RCG/Bagler, Ballly Inc. 5-3 ------- BIBLIOGRAPHY American Society of Testing and Materials (ASTM) 1986. Annual Book Of ASTM Standards, 1986, Philadelphia. ARCO Chemical. 1987. "Testimony of George Yogis Before the Air Quality Control Commission, June 18 and 19." Denver, CO. Alberta Gas Chemicals, Inc. 1987. "Exhibit Before the Air Quality Control Commission, June 18 and 19, 1987." Denver, CO. Brovne, Bortz and Coddington, Inc. (BBC). 1987. Economic Impact Analysis and Market Assessment of A Colorado Oxygenated Fuels Program. Report to the Colorado Department of Health. January, Denver, CO. Colorado Code of Regulations, 1987, Denver, CO. Colorado Air Quality Control Commission. 1987. Transcript of Hearings, In the Matter of Proposed Regulation 13. June 18 and 19, Denver, CO. Colorado Department of Health, 1983. Effects of Bthanol-Blended Fuel on Motor Vehicles at High Altitude. September, Denver, CO. Colorado Department of Health. 1985. Bthanol Blended Fuel as a CO Reduction Strategy at High Altitude. August, Denver, CO. Colorado Department of Health. 1987. Staff Statement to the Air Quality Control Commission Concerning Proposed Regulation 13. June 18 and 19, Denver, CO. Colorado Department of Health. 1988. Briefing to the Air Quality Control Commission. March 17. Energy Analysts International, Inc. (RAI). 1987. Impact of A Mandated Blended Fuel Program on Consumer Costs, Product Supply and Economics. June 4, Denver, CO. Metropolitan Air Quality Council. 1987. "Interim Carbon Monoxide State Implementation Plan for the Metropolitan Denver Non-Attainment Region." June, Denver, CO. Renevable Fuels Association. 1987. Testimony Before the Air Quality Control Commission. June 18 and 19, Denver, CO. Renevable Fuels Foundation. 1987. "Changes in Gasoline and the Automobile Service Technician.11 Washington, DC. Oxygenated Fuels Task Force. 1986. "Report and Recommendations of the Oxygenated Fuels Task Force." October 29, Denver, CO. RCG/Hagler, Bailly, Inc. -1- ------- National Advisory Panel on Cost Effectiveness of Fuel Bthanol Production (NAP). 1987. Puel Bthanol Cost Effectiveness Study. Pinal Report. October, Washington, DC. RGG/Hagler, Bailly, Inc. -2- ------- APPENDIX A ------- APPENDIX A List of Persons Interviewed RCG, Hagler, Bailly vould like to express its appreciation to the following people vho generously provided us vith invaluable Information. Individual Company Mike Barvig Spruce Oil Barbar Bauerle Automobile Association of America Ken Buckler Total Petroleum Inc. Pete Coggeshall Amoco Corporation Vern Combs Pennzoll Products Company Dick Coven Sinclair Oil Company Jan Cool Exxon Dennis Creamer Conoco Inc. Tom Dunn Colorado Department of Revenue Dick Ervin Texaco Jerry Gallagher Colorado Department of Health Ron Hagmier Ventra, Inc. Ted Holman Colorado Department of Health Lance Hoboy Vestec Petroleum Loren Hoboy Vestern Refining Jim Kaiser Sinclair Oil Company Jeff Kramer Frontier Oil and Refining Tom Lareau American Petroleum Institute Jerry Levine Amoco Corporation Kim Levo Colorado Department of Health Sandra Nobbe Oil and Gas Journal Stan Lomax Texaco Bob MeHall Diamond Shamrock Dave Meyers Conoco Inc. Bill Piel ARCO Chemical Company Dick Piper Phillips 66 Company Hike Povell Colorado Department of Labor, Oil Joe Scott Chase Transportation John Snodgrass Diamond Shamrock Patty Stolp Bthanol Managment Company Jim Suttle Chase Transportation Rod Voight Archer Daniels Midland Jack Vilkins Kubat Equipment Ron Villiams Gary Refining George Wright Coors Darcy Void Total Petroleum Inc. George Yogis ARCO Chemical Company ------- APPENDIX B ------- e I9C7 THE PUBLIC RECORD CORPORATION AU. worn RESERVED 10CR7 7-87 p*0*1 RE6ULATI0N NO. 13 "The Reduction of Carbon Monoxide Emissions from Gasoline Powered Motor Vehicles through the use of Oxygenated Fuels" OXYGENATED FUELS PROGRAM I. Statement of Intent, Area of Application, and Definitions A. Statement of Purpose The purpose of this regulation Is to reduce carbon monoxide and particulate emissions from gasoline powered motor vehicles 1n the AIR Program area through the winter time use of oxygenated gasolines. The attached Statement of Basis, Statutory Authority and Purpose, and Fiscal Statement are Incorporated herein, for the purpose of reference only, as Sections IV. and V., respectively. B. Area of Application This regulation shall apply to the AIR Program area as defined 1n C.R.S. 42-4-307 (11). C. Definitions The following terms shall have the following meaning when used In this regulation: 1. "AIR Program" means those parts or all of Colorado's Front Range counties as defined In C.R.S 42-4-307 (11). 2. "Commission" means the Colorado Air Quality Control Commission. 3. "Division* means the A1r Pollution Control Division of the Colorado Department of Health (CDH). 4. "Motor Vehicle" means any self-propelled vehicle which 1s designed primarily for travel on the public highways and which Is generally and commonly used to transport persons and property over the public highways. For the purpose of this regulation, motor vehicles shall refer to spark Ignition motor vehicles which use on a part or full time basis, gasoline or gasoline-type products. 5. "MTBE" means methyl-tert-butyl-ether. 6. "Oxygenated Fuels" means gasolines blended with a component or components containing oxygen, generally an alcohol or ether. THE COOE OF COLORADO REGULATIONS 5CCR 1001-16 ------- Pigtl 7. "Class A Motor Fuel" means any gasoline type product as defined In C.R.S. 8-20-202. II. REQUIREMENTS OF THE OXYGENATED FUELS PROGRAM A. Class A Fuel Requirements 1. Beginning January 1, 1988, to March 1, 1988, no Class A motor, fuel shall be supplied or sold by any person Intended as a final product for fueling of motor vehicles within the AIR Program Area, or sold at retail, .or sold to a private fleet for consumption, or Introduced Into a motor vehicle In the AIR Prograa area by any person, unless the fuel has at least a 1.51 oxygen content by weight. Oxygenated fuel containing 81 by volume MT8E shall be considered equivalent to 1.5X oxygen content by weight. 2. Beginning November 1, 1988, to March 1, 1989 and for each period of November 1 to March 1 thereafter, no Class A aotor fuel shall be supplied or sold by any person Intended as a final product for fueling of motor vehicles within the AIR Program area or sold at retail, or sold to a private fleet for consumption, or introduced Into a motor vehicle in the AIR Program area by any person unless the fuel has at least a 2.OX oxygen content by weight. Oxygenated fuel containing IIS by volume MTBE shall be considered equivalent to 2.0X oxygen content by weight. 3. All oxygenated motor fuel shall be labeled at the pump .during the periods stated In Sections II.A.I. and II.A.2., Identifying the type and amount of oxygenate contained In the motor fuel, 1n accordance with labeling criteria developed by the Division consistent with any applicable law. B. Reporting and Review Requirements 1. The A1r Pollution Control 01v1s1on, In consultation with the State 011 and Gas Inspection Section, the Environmental Protection Agency, the fuel refiners, oxygenate manufacturers, marketers and retailers, the Colorado Auto Dealers, lead air quality planning agenlces, motorists and environmental organizations, shall prepare a report, to be filed with the Commission, on April 15 of each year regarding the results of the Oxygenated Fuels Program, with particular attention to a cost/benefit analysis, to include such factors as air pollution reductions obtained from the program, drlveabllity problems, 1f any, the cost of the program to motor vehicle owners, refiners, marketers and retailers of the fuel, and other Information which 1s relevant to whether the oxygenated fuels program should be continued. The Division shall also work with all appropriate entitles to develop and Implement a public education program. / S CCR1001-16 THECOOEOFCOLOflAOO REGULATIONS ------- C 1917 THE PUBLIC RECORD CORPORATION all ughti reserved 10CR7. 7-87 Pag* 3 2. At its May meeting of each year, the Commission shall consider, in light of the report and other available Information, whether the oxygenated fuels program should be modified, expanded, or terminated by rule, and the Commission shall transmit to the General assembly the report of the Division and the results of the Commission's consideration. C. Enforcement/Penalties for Non-Compliance Compliance with the requirements of this regulation shall be monitored and enforced by the Division. Tolerance for measurements of fuels defined In Section 1I.A.1. and 2. shall be determined by the Division and shall be consistent with reasonable practices. Pursuant to Section 25-7-lll(f), the Olvlslon may designate any appropriate agency of the State to assist 1n the monitoring and enforcement of this regulation. The Division shaTl make every effort to coordinate monitoring and enforcement of this regulation with the current duties of the State Inspector of 011s, conducted pursuant to C.R.S. 8-20-101 et seq. D. Severability The provisions of this, regulation are severable, and If any provisions, or the application of the provisions to any circumstances, 1s held Invalid, the application of such provision to other circumstances and the remainder of this regulation shall not be affected. 111. REFERENCED MATERIAL Pursuant to C.R.S. 24-4-103 (12.5), the following materials referenced 1n this regulation are available for public Inspection during normal working hours, or copies are available upon request at •cost, from the Technical Secretary of the A1r Quality Control Commission, Ptarmigan Place, 3773 Cherry Creek Drive North, 3rd floor, (303 ) 331-8597: C.Tt.'S. 42-4-307 (11), C.R.S. 8-20-202, C.R.S. 8-20-204, C.R.S. 8-20-229, C.R'.S. 24-4-103(12.5), C.R.S.' 6-4-101 et seq, C.R.S. 25-7-lll(f), C.R.S. 8-20-101 et seq. IY, STATEMENT OF BASIS, SPECIFIC STATUTORY AUTHORITY AND PURPOSE The primary purpose of Regulation No. 13 is to reduce ambient levels of carbon monoxide along the Front Range of Colorado. To achieve this reduction, Regulation 13 will institute an oxygenated fuels program throughout the AIR Program area during the period of January 1 to March 1 of 1988, and for each period of November 1 to March 1, thereafter. The Commission has determined that a voluntary program from November 1, 1987 through December 31, 1987 would be of great benefit to the public and will result In air quality benefits by reducing carbon monoxide, and directs the Division to proceed with coordinating and implementing such a voluntary program. An oxygenated fuels program is a necessary step for Colorado to attain the National Ambient Air Quality Standard (NAAQS) for carbon monoxide. Of primary concern is the eight hour, long term carbon monoxide standard, which provides for an eight hour carbon monoxide limit of 9 PPM. Having more than THE COOE OF COLORADO REGULATIONS 5 CCR 1001-16 ------- *Q»« one exceedance of this standard per year constitutes a violation of the carbon monoxide NAAQS. Areas along Colorado's Front Range, especially 1n Denver, have consistently failed to Beet this standard. Use of oxygenated fuels reduces carbon monoxide emissions from gasoline powered notor vehicles. The technical basis for this determination 1s as follows: Most carbon monoxide emissions along Colorado's Front Range are from motor vehicles. It Is estimated that 1n 1987, 77X of the areas CO emissions are from motor vehicles. Oxygenated fuels, containing oxygen via an alcohol or ether blended with gasollone, have been shown, through testing by CDH and others, to be effective at lowering carbon monoxide emissions from motor vehicles. Reductions are directly attributable to the oxygen contained 1n these fuels, by leaning the air/fuel ratio. Host gasoline powered motor vehicles are set up to run slightly rich at sea level. Unless these vehicles are altitude compensating, they will be further enriched as the altitude Increases, since there Is less oxygen present. The amount of leaning Is directly proportional to the level of oxygen contained 1n the fuel. The higher the percent of oxygen In the fuel, the leaner the effective air to fuel ratio will be. As the air to fuel ratio becomes leaner, CO emissions are reduced. Thus, at Colorado's altitudes, most vehicles run richer than at sea level. This excess fuel results 1n less complete combustion and thus Increased carbon monoxide emissions. Oxygentated fuels counter-act these factors. Since vehicles are running rich at high altitudes, the enleanlng effect of oxygenated fuels Is not anticipated to result In vehicle drlveablllty degradation. Current State law prohibits the Commission from requiring ASTM Reed Vapor Pressure (RVP) standards for ethanol blends. There 1s no undue concern with the lack of RVP standards on ethanol blends for the following reasons. First, Regulation No. 13 Is being proposed as a wintertime only program. Volatility Induced vapor lock Is primarily a summertime problem. Marketing ethanol blended gasoline during the other eight months of the year would be the decision of Individual fuel marketers. Second, data provided by the State Inspector of Oils Indicates that the vast majority of gasoline sold In Colorado during 1986 was significantly below the applicable maximum RVP standard. In terms of ethanol blending, this Implies the majority of gasoline could have been blended with ethanol and remain within RVP limits. State law C.R.S. 8-20-204, as revised July 1, 1986 1s largely responsible for this lower RVP gasoline. The Commission would prefer to have appropriate ASTM RVP standards apply to all final oxygenated fuels but recognizes that the Colorado General Assembly has this prerogative S CCR1001-16 ------- S ran THE PUBLIC RfcCORD COAPOHATION ALL RIGHTS RESEXVED 10CR7, 7-87 PigiS The Commission encourages and directs the Division to wort with non-retail rotor fuel suppliers to keep adequate records and Inform purchasers of non-retail fuel regarding the amount and type of oxygenate In the fuel supplied to and 1n the AIR Program area. The oxygenated fuels program provided In Regulation No. 13 will be fn effect for the period of January 1 to March 1 of 1988 and for each period of November 1 to March 1, thereafter. This tine period is being used for the following reasons: High levels of carbon monoxide are experienced fn Denver and Colorado's Front Range during- the winter months. For the most part exceedances of the CO standard occur between November 1 and March 1 of each winter. There are several environmental, climatic and geographic reasons for these high winter-time concentrations. These Include a large motor vehicle fleet, traffic congestion, high altitude, cold weather and atmospheric temperature Inversions. High CO concentrations In Colorado are brought about In part by Colorado's altitude and cold weather. At higher altitudes and colder temperatures, motor vehicles tend to have less efficient fuel combustion, resulting in Increased levels of CO. Another cause of high carbon monoxide concentrations is the incidence of winter-time temperature inversions which can develop during the evening. An inversion will trap pollutants such as carbon monoxide near the surface. In these conditions of stagnate air, extremely high concentrations of carbon monoxide build and collect. If the marketplace for Class A motor fuel operates 1n an open and competitive manner, a number of oxygenates should be available for use In the AIR Program area which, when blended with a base fuel, will produce oxygenated fuels that meet this regulation. Any attempt to limit the choice of oxygenates available in the AIR Program area between December 1 to March 1 of 1988 and for each period of November 1 to March 1, thereafter, while this regulation is 1n effect, through a combination, conspiracy, trust, pool, agreement or contract intended to restrain or prevent competition in the supply or price of base fuels suitable for blending with an oxygenate to produce.oxygenated fuels which comply with this regulation and applicable EPA requirements, may constitute an illegal restraint of trade in violation of Section 6-4-101, et. seq., Colorado Revised Statutes. Authority for Regulation No. 13 can be found in the Colorado Air Quality Control Act, Section 25-7-101 et seq., C.R.S. 1982 (1986 Supp.). Specifically, Section 25-7-106(1 )fe) authorizes the Colorado Air Quality Control Commission to develop "a control or prohibition respecting the use of a fuel or fuel additives in a motor vehicle to the extent authorized by Section 211(c) of the federal act". Specific authority can also be found in Section 25-7-109(3)(d). TH£ cooe OF COLORADO REGULATIONS 3 CCR 1001-16 ------- Pigifi V. FISCAL IMPACT STATEMENT Proposed Regulation No. 13 would require the use of oxygenated fuels In the AIR Prograa area starting January 1 1988 to March 1 of 1988, and for each period of November 1 to March 1, thereafter. This regulation would require the use of an oxygenate such as an alcohol or an ether to be blended Into gasoline. The Increase In the retail price of gasoline due to this regulation Is being estimated at .051 to 3.51 per gallon. The affected AIR Prograa (Front Range) area consumes an estimated 1.2 billion gallons of gasoline a year. For the four month winter period oxygenates would be required each year, this would affect an estimated 380 million gallons of gasoline annually. Using the estimated high range of 3.51 per gallon increase, this would result in an annual cost of $13.3 million. Based on Division testing, a 2% fuel econoay penalty was seen for closed loop vehicles operating on oxygenated fuels. This Is estimated to be 250,000 vehicles 1n the affected area. Assuming each vehicle Is driven 4,000 miles from November 1 to March 1, each vehicle averages 25 mpg, 160 total gallons of gasoline would be consumed per vehicle. A 21 Increased fuel consumption would Increase consumption an additional 3.2 gallons. This results In a total Increase of 800,000 gallons. At one dollar per gallon, this 1s an additional .$800,000. This brings the total program cost to $14.1 million annually. With both ethanol and MTBE in use 1n the marketplace, this regulation 1s estimated to result 1n a 300 ton per day, or 141 reduction In ambient carbon monoxide levels. This results In a cost of $128 per ton of carbon monoxide reduced. These figures estimate increases 1n the retail price of gasoline, and encompass all associated costs to provide oxygenated fuels. Gasoline refiners, blenders, distributors and marketers will all have costs as a result of this regulation. Refiners may have a reduction In crude oil through-put from their refinery operations, as oxygenates will be displacing gasoline. Refiners or fuel blenders will need to purchase and blend an oxygenate Into gasoline. This may require additional tankage and blending facilities In some cases. Retail fuel marketers, and private fleets with fueling facilities will need to ensure underground tanks are clean and water free before handling any alcohol blended gasoline. Final filters at the fuel pump nozzles may need to be added when dispensing alcohol blends. These costs are included 1n the estimates made above regarding cost per gallon Increase. It should be noted that these estimates are using the high end of retail gasoline cost estimates, and are the costs applicable to the fully implemented program, and would be less during the phase-in years. There was conflicting testimony at the public hearing regarding costs. Representatives of the petroleum industry estimated that costs of implementing the program could range up to 8.3 cents per gallon and total $49 million. These cost estimates vary so widely because an oxygenated fuel program In the AIR Program area may encourage the use of such fuels in other areas and states with carbon monoxide problems. Some owners of motor vehicles may also Incur repair costs, 5 CCA 1001*1Q THE COOE OF COLORADO REGULATIONS ------- |