‹header›
‹date/time›
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
AM-04-75
Page ‹#›
SIn this panel we are going to talk about a wide range of refining issues from near term to long term. 
S
SI will focus on refining capacity and factors affecting decisions to alter that capacity. 
S
SCapacity decisions, by the nature of the assets involved, are influenced both by short- term considerations as well as long-term.  
S
S
SThe next few slides pose the capacity issue that the rest of the presentation will address.  SThe excess capacity that existed in the early 1980’s is gone.  But will U.S. capacity necessarily be the most economic supply source to meet growing demand in the short or the long term? SAnd although demand is currently still growing – how long will it continue to grow and at what rate?
S.
S
SAs we look ahead long term, there is much discussion about moving away from our petroleum dependence, which likely will require tremendous infrastructure changes.
S
SYet, if we look back in time, we recognize that the United States went through tremendous infrastructure changes to get from the horse and buggy era to today.  Tremendous changes can occur, and likely will occur.
S
SThe need for petroleum fuel is not going to diminish quickly.  But as we struggle with the short-term challenges, we need to keep an eye on the long term, looking for signs of significant changes. 
S
S
S
S In looking at the need for new capacity, we recognize that many factors influence each company’s decision.  Two significant factors are future demand and the marginal or most economic sources of supply to meet demand. 
S
SFor today’s discussions I will focus on
Historical demand for petroleum and what insights history can provide on the potential for future demand changes;
What seems to be happening to our sources of product imports – in particular gasoline;
What historical changes to U.S. capacity can tell us about potential future changes.
SClearly future demand growth is an important variable in the decision to increase capacity.
S
SU.S. demand more than tripled in the 50 years from 1950 to 2000, fueled by transportation, which went from just over half of petroleum demand in 1950 to over 2/3 of demand by the year 2000.
S
SBut will such growth continue in the future, and what might change that pattern?
S
SMost “reference case” or “business as usual” forecasts today show petroleum demand in the United States continuing to grow over the next 25 years or so, being driven by transportation fuels.
S
SEIA’s reference case shows petroleum demand for transportation fuels reaching a level in 2025 that is more than 50% greater than petroleum transportation demand today.  An additional 7 million barrels per day will need to be supplied, which argues for more capacity.
S
SWe know factors can change this outlook.  I would venture to say that a significant slowing in demand growth or a significant shift in the mix between diesel and gasoline could have tremendous impacts on the shape of the refining industry in the future. 
S
SBut do such scenarios merit consideration?  History can provide some insights.  For example, in the early 1980’s, high prices, a recession, and improved vehicle fuel efficiencies acted to reduce demand.
S
SThe next slides consider the impact fuel efficiencies alone can have.
S
SCorporate Average Fuel Economy (CAFE) standards and high fuel prices resulted in large improvements in vehicle efficiencies for almost 10 years in the late seventies and early eighties.
S 
SUltimately other factors such as the interest in minivans, SUV’s and heavier and higher performance automobiles countered those improvements, and light duty vehicle efficiency actually declined for the entire fleet.  Will this change?
S
SAs we look into the future, one factor that could change the current efficiency trend is a stronger movement towards limiting CO2 emissions.  Technology exists to improve efficiencies, but at a cost. 
S
SAssuming the United States decided to incur the cost, how quickly could changes in efficiency make an appreciable impact?
SWhile the situation historically was different than now, it does provide an illustration of how quickly efficiency changes can affect demand.  This slide focuses on the period from 1978 through 1987, which covers part of the time when fleet efficiency was improving rapidly. 
SThe graph illustrates that efficiency improvements can affect demand in less than 10 years.  During the early period from 1978 to 1982 shown to the left of the dotted line, vehicle miles traveled was basically flat, and gasoline consumption declined as car efficiencies improved.  Even after 1982, gasoline consumption growth was small as car efficiencies continued to improve, countering the increasing share of light-duty trucks in the fleet. 
During the entire 1978-1987 period, the total VMT (vehicle miles traveled) for light duty cars and trucks rose at 2.4%, which is about the same as it was between 1995 and 2001. 
But in the 1978-1987 period, there was virtually no growth in gasoline consumption, compared to an annual average growth rate for gasoline of 1.7% for the 1995-2001 period.
S
SBecause vehicles are further along the efficiency curve now, we might not see the dramatic percentage improvements in the future that occurred in the past.  During the late seventies and 1980’s, the fuel efficiency of the new vehicles was between 45% and 55% of the average fleet on the road.  Still, the technology exists to improve fuel economy by over 30%. 
S
SIf the United States targeted fuel economy improvement, we could see a dramatic reduction in gasoline demand growth. For example, assume 3 years to design efficiencies and then gradually introduce fuel economy improvements into new vehicles.  If new vehicle efficiency reached a 28% improvement in 7 years, by the end of 10 years, we could see demand growth more than cut in half.  That is, a “base case” 2% growth rate would fall to less than 1% within a decade.
SNow consider U.S. capacity decisions in the context of competing with capacity from abroad.
S
SHistorically, capacity abroad has been very competitive in the U.S. gasoline market, as demonstrated by gasoline imports having more than doubled in volume since 1990.  As a result, imports have become an essential source of gasoline supply to the U.S. East Coast, where they met 24% of East Coast (PADD 1) demand in 2003.
S
SBut will these competitive imports continue to be available and grow to help meet growing U.S. demand?
S
SOne of the short-term issues is how might changing gasoline specifications in the United States affect the availability of imports from other regions.  As EIA has discussed in other reports, the ban on MTBE requires very low RVP gasoline components to blend with ethanol in order to produce RFG. The low RVP reformulated gasoline blendstock for oxygenate blending (RBOB) is unique to the United States, and we see no signs of other countries moving this direction.  This reduces the number of import suppliers that can provide RFG components from the number that were able to provide MTBE-blended RFG. 
SU.S. sulfur constraints are also expected to reduce the number of suppliers, but more for timing reasons than because the United States is taking a unique specification path.  The United States is moving to low sulfur product before many other regions do so.
Europe is relatively close to U.S. standards, and its exports to the United States may not be limited by sulfur.  In fact, with its tax incentive programs, gasoline being sold in some European countries today has lower sulfur content than U.S. requirements.
South America generally is producing gasoline with higher sulfur levels than in the United States, and that is not expected to change soon.
Asia is moving to lower sulfur levels in some countries, but also has higher sulfur specifications in many regions still.
S
SThese regional specifications along with knowledge of where our imports are coming from can give us a clue as to how the U.S. specifications may impact import availability.
S
SThis chart shows imports stacked roughly by their ability to produce low sulfur gasoline.  Those with the highest ability are on the bottom. SOne could assume that refineries that depend on the United States for much of their output would invest to comply with U.S. regulatory changes.  Facilities dependent on U.S markets include Eastern Canadian refineries, the Hovensa refinery in the U.S. Virgin Islands, and refineries in Venezuela.  In 2003, these three regions supplied 42% of total gasoline imports.  SIn addition, Western Europe has become an important supplier to the United States, and with European Union (EU) specifications being close to U.S. specifications, we might expect a large share of those volumes to continue to be available.  SHowever, much of the historical supply from Latin America outside of Venezuela may not be able to meet our specifications in the near future, and Venezuela’s ability to supply future specifications on time is uncertain as a result of the changes in PDVSA that occurred since the strike in December 2002.  SWe may also lose some Eastern European volumes – particularly when we move to 30 ppm requirements next year.  Furthermore, some of our Asian sources of imports (e.g., India and China) may not be able to supply low-sulfur product for some time, although Asia is a small source of imports, supplying 4% of total imports in 2003. SThis implies that in the short run, we may need higher prices to attract needed volumes both because fewer sources can produce U.S. product and because the product we need is higher valued.  SOne of our biggest import source areas, Europe, may need to supply more product in the future if other areas drop out.  But will Europe have extra supply for export?
SThe United States has benefited from Europe’s dieselization program historically.  As Europe’s dieselization program evolved, demand for diesel fuel grew faster than gasoline, and Europe’s refineries produced more gasoline than the region could use.  This provided an economic source of supply for the United States
S
SEuropean diesel demand overtook gasoline demand after 1995, and is expected by many forecasters to continue to outstrip gasoline – even to the extent that gasoline demand might decline.
S
SThis forecast shows the ratio of gasoline to diesel halving between 1990 and 2015, which has significant impacts for refiners.
S
SThis forecast is based on a continuation of the change in vehicles in Europe that has occurred historically.
SAs in the United States, European diesel fuel consumption has grown because of growth in road freight movements.  But in Europe, most of the growth in diesel fuel has been due to the growth in diesel-fueled light-duty passenger vehicles.
New vehicle registrations showed new diesel vehicles exceeding 40% of new light-duty vehicles. 
This compares to 1-2% in the United States.
SIn Europe, tax incentives favoring diesel fuel over gasoline have existed for a number of years in many countries.  Recently, concern over greenhouse gases and carbon dioxide emissions have resulted in additional diesel fuel tax incentives.
S
SFurther penetration of diesel-fueled vehicles is predicted, and as shown in the previous slide, the result could be a drop in gasoline demand.
S
SThis shift from gasoline to diesel has been occurring over many years, which raises the question of whether refiners are adjusting their capacity to better match this changing fuel mix, thereby reducing the amount of excess gasoline available to the United States.
SIn response to the market shift from gasoline to diesel fuel, European refiners have increased hydrocracking capacity to increase production of diesel compared to gasoline.  Since 1990, hydrocracking as a percent of distillation capacity has increased from 2% to almost 7%, while FCC capacity as a percent of distillation has leveled at about 16%.
S
SDespite the growth in hydrodcracking capacity, the deficit in diesel production compared to demand and the excess in gasoline production has continued to grow.  This is expected to continue for some time, even with continued hydrocracking capacity growth. 
S
SThis implies Western Europe may be able to provide the United States with higher imports in the near future, to some limited extent. 
S
SAlso, as both European and U.S. gasoline specifications become more stringent, more Eastern European refiners may have the incentive to upgrade as well.
SIn the short term, as U.S. product specifications become more stringent than in many other areas of the world, the number of available import sources is likely to diminish, which in turn, all else being equal, would increase the cost of imports.
S
SFurthermore, growing demand in the rest of the world where less stringent product specifications exist, could provide markets for refiners that previously supplied the United States.
S
SBut if margins increase in the Northeast (the major product import area), then an opportunity arises for the large number of importers who have supplied that market historically to figure out a way to take advantage of the new situation.
S
SThat is, a shortfall in import availability may not be a lasting problem. If more imports are needed, especially at an attractive price, they will be found.  And if import prices remain high enough for a long enough period of time, then we should see increased capacity growth in the United States, or new, more price-competitive import sources will emerge.
S
SAlthough factors in the short term may improve the competitive position of U.S. capacity, will that situation be sustained over the long term?
SNow I turn to historical capacity changes to see what insights might be gained from this information as to what to expect in the future.
S
SThis section of the presentation will cover both closures and operating refinery capacity changes, and end with noting how yield changes resulting at least partially from conversion capacity increases have helped to meet growing demand historically.
SFrom 1981, when refinery capacity utilization was 69%, until the mid 1990’s, refinery industry concerns were focused on surplus capacity. 
S
SBut since 1996, refinery capacity has increased. 
S
SThe net growth hides some interesting details that may help us to understand the future.  The next few slides will explore the separate effects of shutdowns and of  refineries that have stayed in operation.
SRefinery closures have occurred every year over the past two decades.  Since 1988, we have lost over 1.6 million barrels per day of capacity, which is about 10 percent of today’s capacity.  Several factors are driving this situation:
The refineries that have closed are smaller and have less favorable economics than other refineries in their market area.
Even though refinery utilization has improved since the 1980’s, refinery margin improvements have been modest. 
In recent years, some smaller, less-economic refineries that faced additional investments for environmental reasons in order to stay in business found closing preferable because they predicted that they could not stay competitive in the long term.
S
SWhile the rate of closures as measured by average capacity shut down per year has slowed, EIA expects to see closures continue in the future. 
SWhile some refineries have been closing, many remaining in operation have been expanding.
S
SSince 1993, refineries that were in continuous operation grew on average 1.5 percent per year.  Refineries in the mid-size range (140-210) exhibited the largest percentage growth pattern.
S
SThe term “capacity creep” has come into use to describe the relatively slow aggregate capacity increases.  However, the year-to-year data show additions to individual refineries are larger than the term “creep” would imply.
Much of the capacity growth in individual refineries occurred in fairly large increments over 1, 2, or 3 years during the 15-year period, rather than in small increments over many years.  
There were 10 instances in which individual refineries increased distillation capacity in one year between 30-50 MB/D, and there were 93 instances were capacity increased between 10-30 MB/D over the prior year.
SWhile these increases are modest in size and not being done by all refineries, the expansions are more than incidental efficiency improvements.
SIn addition to distillation capacity, downstream unit capacity has also grown.  This graph illustrates that distillation capacity has grown to maintain balanced refinery operations in that FCC and distillation capacity growth are very similar. 
S
SCoking and hydrocracking can be changed to accommodate different types of crude feeds and alter product mix, but may need little associated distillation capacity growth.  However, some distillation capacity increase is fairly common in conjunction with coker projects. 
S
SIn recent years, refiners have been increasing treating capacity to lower product fuel suflur and improve other product qualities.  During this time, we have seen little change in distillation capacity.
S
SWhat this tells us is that historically there has been a much stronger link between conversion capacity growth and primary distillation capacity growth than with increased treating capacity increases.
S
SThis implies that with refiners focusing on treating capacity right now, the United States may see little distillation capacity increase for another year or two.
SWith increases in conversion capacity, light product yields increased, which helped to meet growing transportation demand.
S
SVirtually all of the increase in yield from crude and unfinished oils was in distillate and jet fuel.  The yield increase in gasoline came from the addition of other components, mainly MTBE. 
S
SLooking ahead, the past yield trends cannot be projected into the future.
SFactors that are working to decrease yields outnumber those that tend to increase yields.
S
SThis places the burden back on capacity growth to meet demand increases.
SAs we think back through the factors that have affected capacity, we are left with much uncertainty about whether or not capacity investments will realize a reasonable return in the future.  
S
SFor much of the 1990’s returns on refining investments were not attractive.
S
SThose returns improved since 2000, but not smoothly or predictably.  Price spikes during the spring and late summer in gasoline, and winter distillate price spikes in the past several years contributed to improved returns, but that is not a predictable environment.  
S
SThis presentation touched on a number of factors that could affect refiners’ capacity decisions,  and attempted to illustrate the challenge the industry faces as we look ahead.  In the short-term, many factors point to the potential for improved potential for return on capacity investment.
Most forecasts show increasing gasoline and diesel fuel demand growth.  Even if vehicle fuel economy again improves significantly, the impacts will not occur immediately.
S
Our primary product import is gasoline, and with tightening U.S. specifications relative to much of the rest of the world, we might need to pay higher prices to attract increasing import volumes – at least for the next few years.
S
As refiners direct investment dollars to treatment capacity in order to reduce gasoline and diesel fuel sulfur content, capacity expansion may slow.  In addition, some of the fuel specification change requirements will reduce refinery capability to produce gasoline through yield loss. 
S
Refinery closures are likely to continue. 
S
All of this adds up to increased margins in the short term, which should improve the environment for capacity investment.  However, margin improvement may be the result of short-term price swings rather than a smooth steady increase.
S
SLong-term changes could dramatically affect the shape of the refining business, but such changes are not likely to occur quickly, and should produce signs of change in advance of the impacts, which is an incentive to keep an eye on the long term.
S