SPECIAL REPORT: US ethanol forecast identifies refiner, marketer opportunities

March 19, 2007
This spring, US ethanol production capacity and imports are expected to exceed blenders’ current ability to blend ethanol into gasoline.

This spring, US ethanol production capacity and imports are expected to exceed blenders’ current ability to blend ethanol into gasoline. During the next 2 years, new plants and expansions of existing plants will come on stream, creating by the end of 2008 an 8-10 billion gal/year (1.17 million b/d) ethanol production capacity. Only in 2009 will production capability finally match demand.

In the meantime, refiners and gasoline blenders will have an opportunity to develop ethanol supplies for both current and additional blending requirements on a very attractive economic basis.

This forecast by Houston BioFuels Consultants LLC (HBC) paints a picture of the US ethanol industry and market during the next 2 years. Figures are given both in gallons, which the ethanol industry currently is using, and in barrels, employed by the oil and refining industry.

Since May 2006, when ethanol replaced methyl tertiary butyl ether in the US gasoline pool, ethanol demand in the US has been steady, averaging 5.9 billion gal/year (385,000 b/d) through November 2006. This is considerably more than the amount mandated by the 2005 Energy Policy Act (EPACT) for 2006 (4 billion gal/year) and 2007 (4.7 billion gal/year.)

Then the 2006-07 winter oxygenate program and unseasonably strong gasoline demand increased ethanol demand to 6.2 billion gal/year or 404,000 b/d, which is 2.2 billion gal/year more than EPACT’s 2006 requirement.

Weekly variations have been within a narrow band of plus or minus 400 million gal/year (26,000 b/d) and are mainly the result of deviations in demand for Environmental Protection Agency-specified reformulated gasoline (RFG)-which comprises about one third of US gasoline demand but contains two thirds of ethanol demand (Fig. 1).

Click here to enlarge image

Weeks around major holiday weekends-Memorial Day, Fourth of July, Labor Day, and Thanksgiving-that always show volatile demand because of short reporting weeks and holiday demand represent a major component of the variation seen in the weekly ethanol demand estimated by HBC.

Viewing ethanol demand on a monthly basis eliminates much of this variation and provides a convenient method for comparing HBC’s ethanol demand estimates with the ethanol production, imports, and inventory reported in the Energy Information Administration’s EIA 819 Monthly Oxygenate Report, published 2 months in arrears. Agreement between the two is quite close, with known exports being one reason the EIA report tended to be higher in 2005 when there wasn’t such a call on ethanol production in the US (Fig. 2).

Click here to enlarge image

Supplies of ethanol from imports and inventory draws have provided the volume to fill the gap between US ethanol production, estimated at nearly 5.6 billion gal/year (365,000 b/d) at yearend 2006, and demand. By comparison, at the end of 2005 and 2004, ethanol production capacity in the US was 4.8 billion gal/year (313,000 b/d) and 4.3 billion gal/year (280,000 b/d), respectively, already significantly above the mandated amount of 4 billion gal/year for 2006.

By the end of first-quarter 2007, an additional 600 million gal/year (39,000 b/d) of ethanol production capacity is expected to be on stream from a dozen or more new ethanol plants. This is nearly the same capacity that was added in all of 2006 by 13 new plants; capacity expansions-about 200 million gal/year (13,000 b/d)-provided the balance of the 2006 increase.

HBC forecasts imports of 500 million gal/year (33,000 b/d) during the next 2 years, which is somewhat less than the 700 million gal/year (46,000 b/d) imported in 2006. Imports likely will continue because Brazil currently has no other market comparable to the US and blenders want to maintain an alternate supply chain.

Throughout the remainder of the 2-year forecast period, new plants and expansions of existing plants already well under way-and for all intents and purposes “locked-in”-will come on stream so that by the end of 2008 there will be at least 8 billion gal/year (522,000 b/d) of ethanol production capacity in the US and perhaps as much as 10 billion gal/year (652,000 b/d) (Fig. 3).

Click here to enlarge image

The low end of this range will occur if sponsors of the legitimate ethanol projects that are ready to proceed recognize the oversupply and tight margin situation (high corn prices but low ethanol prices), defer projects, and bide time until prospects for margins improve.

The high end of this range will result if, instead, the many independent ethanol project developers and their financiers believe their projects can withstand the market conditions during an indefinite period following initial operation.

At the moment, ethanol project developers speak confidently of their projects’ ability to weather the storm. Are they engaged in a game of waiting for other developers to blink, or are they going to proceed regardless? That is the uncertainty differentiating HBC’s high and low forecast production capacities.

In either case, production capacity would exceed HBC’s demand forecast.

If developers restrain production capacity as we expect, an oversupply will still occur, but ethanol prices will fall just low enough to continually induce more blenders to use ethanol. In the unrestrained production capacity case, ethanol prices are likely to fall below what is needed as incentive for blenders. In this latter scenario, a battle could result among producers to lower prices enough to reduce capacity utilization so that supply and demand balance-as ultimately they must.

Demand increase forecast

The HBC-forecast increase in ethanol demand from today’s 5.9 billion gal/year (385,000 b/d) level is predicated on discretionary blending taking place-that is, additional blending beyond the amounts mandated by EPACT. Ethanol producers will offer economic inducements sufficient to overcome the blenders’ capital and switching costs, so they will blend more ethanol than required by EPACT.

The forecast includes an additional 300 million gal/year (20,000 b/d) of demand in California, starting in the second half of 2008. HBC believes changes in the state’s predictive model regulating California gasoline quality and possibly a state mandate for greater use of renewable fuel will bring about this increase.

Demand will increase in lumps because of varying lead times to install needed facilities at terminals to enable ethanol blending.

With current usage far exceeding the federal renewable fuels standard requirements and with future mandates unlikely to significantly contribute to ethanol demand until 2009 and onward-when the mandate will be 6.1 billion gal/year-the main driver for increased ethanol usage during the forecast period will be favorable blending economics giving gasoline blenders incentive to blend more ethanol.

Most current discretionary ethanol blending is in the Midwest, and this is where most of the additional ethanol blending is expected to occur during the forecast period because the proximity to ethanol production will result in the lowest ethanol cost in the country relative to gasoline. Hence Midwest blenders will have the most economic incentive to blend more ethanol than is mandated.

HBC estimates that about 30% of the terminals in the Midwest can begin blending ethanol at relatively low cost within 4 months of deciding to do so. Other Midwest terminals will require more time and capital to install ethanol blending facilities, 18 months in some cases, but by the end of 2008, an additional 1.7 billion gal/year (111,000 b/d) of discretionary ethanol blending is expected to take place in the Midwest, most of which will be 10% ethanol.

Discretionary blending will increase in other areas of the US, as well, but to a lesser extent and on a less systematic basis. One example of this would be the Atlanta area, which has had facilities to blend ethanol at its terminals since 2005. More examples would be Florida and other main population centers in the Southeast. However, should these coastal areas convert to ethanol blends, we expect that imports from Brazil or the Caribbean Basin Initiative (CBI) countries would increase and provide much of the supply. Overall US ethanol demand would increase but would not significantly impact the forecast increase in discretionary blending in the Midwest or help provide demand to meet the US ethanol production capacity.

To a certain extent, it is in the ethanol producers’ best interest to encourage additional discretionary blending in the Midwest states to reduce the threat of foreign ethanol taking market share. This is but one area where ethanol producers and marketers interests may begin to diverge.

The HBC forecast shows production capacity exceeding demand by about 240 million gal/year (16,000 b/d) -4% of capacity-starting in second-quarter 2007 and persisting until at least yearend 2008. This excess capacity will be the main driver pushing ethanol prices down to the point that discretionary blenders will take the steps needed to increase ethanol blending.

Events in the ethanol market, Washington, DC, state legislatures, and the oil markets will have a major impact on what happens in 2009 and beyond, but for the next 2 years, the large increase in ethanol production capacity will be the key driver in the US ethanol market, and the laws of supply and demand will prevail.


The ethanol industry has evolved such that ethanol marketing companies currently are the intermediaries between ethanol producers and blenders, supplying rail cars and other logistics and scheduling as well as coordinating plant output with blenders.

Ethanol producers are predominantly from agribusiness rather than the oil industry. This arrangement has worked well until now, with companies and terminals interested in using ethanol relatively easy to locate, as most of the country’s ethanol demand has been generated by federal RFG requirements or state governmental mandates such as California’s.

The marketing companies typically represent dozens or more ethanol plants whose individual production is relatively small compared with the needs of any one blender. They locate the supply and arrange logistics for reliable delivery of the ethanol to the blending or regional transloading terminals.

Starting in the spring and continuing through at least yearend 2008, the situation will be reversed: Ethanol supply will be well-defined and easy to locate, but the additional demand is going to be more difficult to pin down because it will be in the form of discretionary blending. Refiners and gasoline blenders have an opportunity to develop the ethanol supplies for both current and additional blending requirements on a very attractive economic basis.

Overproduction effects

With the large increase in new ethanol production capacity, refiners will recognize the oversupply situation in the ethanol market. In an oversupply situation, refiners are forced by market conditions to sell products on the basis of something over the variable cost of production rather than on the basis of value-added supplied to the customer.

Oversupply will tie ethanol prices to the variable cost of production. Ethanol producers will operate so long as their net-back prices cover variable costs. At prices any lower, they will be better off shutting down. Some ethanol plants during this overcapacity period might be shuttered, at least temporarily.

Cooperatives that own ethanol plants and supply corn might maintain ethanol production and forgo some of the corn value-added even if ethanol prices are below variable cost.

Refiners can relate to this, remembering when high-cost refineries (e.g. in Europe and Asia) continued to operate during periods of global overcapacity because they were government-owned or because government policy dictated that they stay online.

As in any other manufacturing industry, there is a cost curve for all the ethanol plants, with some in better cost positions than others. The large wet mills are some of the lowest-cost producers because netting out byproduct sales provides a greater credit than distiller grain solubles do for dry mills.

Cost considerations

Refiners and gasoline blenders in the Midwest should be in position to gain the most advantage from the oversupply. Blenders in the coastal markets may also have the opportunity to source economically advantaged gasoline blendstocks from the oversupplied ethanol industry-even with corn at $4/bushel. At $4/bushel of corn and $8.50/Mcf of natural gas, the variable cost for a dry mill, with no freight costs, is about $1.60/gal.

If the total cost of logistics from the ethanol plant to the blending terminal is 25¢/gal or less, the delivered cost to the blender-less the excise tax credit of 51¢/gal-would be $1.34/gal or less. Incentives grow to the extent that corn prices, natural gas, or other costs are lower. Readers can compare this with their gasoline price forecasts.

The capital cost of modifying a facility and the average ethanol throughput per terminal can give an indication of the price incentive blenders will need before investing. The average terminal ethanol throughput in the Midwest, assuming all gasoline is E10, is about 10 million gal/year (650 b/d). For a simple, 3-year payback, a rough proxy for a 25% discounted cash-flow after-tax return on capital-about 3.3¢/gal-would be needed for a $1 million investment. Double that-or 6.7¢/gal-would be needed for a $2 million investment.

The foregoing should only be used as a rough indication of the incentives that will be required, as each blender will have its own cost for facilities modifications, potential ethanol throughput, view on forecast crude oil and gasoline prices, risk-adjusted return hurdle rate, and other metrics for gauging the potential attractiveness of ethanol blending.

All potential additional discretionary blenders must be convinced that ethanol prices will stay low enough long enough for them to recover their capital and switching costs and still profit by switching to ethanol blending. Because of the prevalence of exchanges for areas such as Florida and the Southeast, where no ethanol is currently being blended, more than one blender may need to be independently convinced of the economics before discretionary blending becomes a viable option.

Another implication of this forecast is that refiners contemplating investments in ethanol production may wish to consider several options. Oversupply will invariably mean low return on investments for ethanol plants until demand catches up to production capacity. It will mean that some current ethanol plant owners will decide to exit.

Ethanol production assets should be available at prices better than today’s, reflecting lower valuations because of the low-margin environment (lower free cash flows).

Another opportunity and alternative to investing directly is toll processing. Refiners should find some ethanol producers receptive to toll processing arrangements that will provide producers with reliable outlet channels for their production and enable them to increase capacity utilization rates. The producer will have to provide economic incentive for the refiner to assume the offtake commitment and commodity price exposure, but this may seem like a bargain to the producer whose alternative is to shut down or sell out.

Relationships between blenders and producers will strengthen from cooperation in tolling and other long-term arrangements. In addition, blenders will gain understanding of the ethanol industry, including the corn market, and the stage will be set to forge more permanent arrangements with producers, including joint ventures.

Refiners, marketers, and ethanol producers all have the opportunity to profit by recognizing the coming changes and opportunities and acting on them in a timely fashion.

The author

Click here to enlarge image

Logan Caldwell ([email protected]) is president of Houston BioFuels Consultants LLC, where he provides strategic advice and insights concerning the biofuels markets and industry. He has experience in business development, procurement, and sales for biofuels, traditional fuels, petrochemicals, and specialty products. During more than 30 years in the downstream oil industry prior to founding HBC, he worked for Chevron, Exxon, Hess Oil, and Air Products on US and international assignments. He has an MBA in finance from Columbia University and a BS in chemical engineering from the University of Kentucky.