by Andrew D. Weissman, Haynes and Boone
The U.S. electricity market is in the midst of a sweeping transformation with implications for power producers, investors and end-use customers.
The electricity, natural gas and coal markets have converged into a tightly integrated market in which prices for the three commodities are determined by intense competition between natural gas-fired generation and coal.
Power producers realize that in a post-shale market, margins are likely to be thin.
Many plant owners are often reluctant, however, to develop a rigorously supported position regarding future price levels.
This issue cannot be avoided. Every decision to buy, sell or hold generation is a bet on future prices.
Change in Market Dynamics
The past decade was a halcyon period for independent power producers. It was not unusual for on-peak prices to average $60 to $70 per megawatt-hour for the year, reaching $100 per megawatt-hour or more on hot summer days. Cash flow and operating margins were strong.
This era of high prices was triggered by construction of a massive armada of new natural gas-fired generating units.
Between 1999 and 2005, the power industry built more than 200,000 MW of new, natural gas-fired generating capacity.
This construction program locked the U.S. into dependence upon increased use of natural gas to meet incremental demand for nearly a decade.
Between 2003 and 2007, annual power sector consumption of natural gas increased by 1.7 trillion cubic feet (Tcf)–an average of nearly 1 billion cubic feet (Bcf) per day.
In an unfortunate coincidence of timing, this increased dependence upon natural gas occurred as natural gas producers hit a wall in expanding production from conventional sources of supply.
A train wreck was inevitable. To keep the lights on, power suppliers had no alternative other than to bid natural gas prices as high as necessary to drive price-sensitive users out of the market.
The price availability curve for demand destruction, rather than the short-term marginal cost supply, determined prices for natural gas.
Demand destruction, however, was not solely responsible for price increases during the past decade. Instead, two other factors were at work.
First, for decades, the ability of energy users to switch from natural gas to fuel oil played a major role in tempering price increases, which enabled natural gas users to reduce demand quickly if prices for natural gas started to rise.
During 2001 and 2002, however, in compliance with Clean Air Act requirements, many oil-fired industrial boilers and electric generating units were retired.
This eliminated a critical safety valve that had been a factor in keeping natural gas prices in a narrow range.
Second, during winter 2002-2003 with less fuel switching available, total U.S. storage capacity was insufficient to meet needs in a slightly colder than average winter.
When the withdrawal season began, storage was nearly full. By late February, the amount of gas in storage was drawn down to perilous levels and triggered a steep run in prices. At the height of the crisis, prices at Henry Hub reached $32 per million British thermal units (MMBtu) intraday.
This price spike left an indelible imprint on the market. In response, a huge risk premium was incorporated into the price of winter-month contracts.
At its peak, this premium reached $1 to $2 per MMBtu and rippled through the entire 12-month strip, which raised prices for natural gas across the board.
This price increase flowed into the power market and raised prices in natural gas-dependent regions by an additional $10 to $15 per megawatt-hour.
Permanent Change in the Market
During the past four years, the market drivers that triggered price increases during the past decade have been turned on their heads.
The supply side of the story has changed in ways that would have been unfathomable a few years ago with unprecedented increases in total U.S. production and the addition of vast new reserves.
Now the issue is too much supply; not too little.
Abundance, however, is only part of the story. The speed at which production can be increased is as important.
During the past five years, U.S. onshore production has increased by a mind-boggling 16.5 Bcf per day–an average of 3.3 Bcf per day every year.
The natural gas crisis of the past decade was sparked by a total increase in power sector demand of 5.4 Bcf per day.
Post-shale production twice has increased by more than 5.6 Bcf per day in a single year (i.e., in 2010 and again in 2011).
If the growth in power sector demand during the past decade had occurred just a few years later, it easily could have been accommodated.
As production has been increasing, fuel switching has returned to the market with a vengeance, albeit this time with coal.
In addition, more than 700 Bcf of new underground storage capacity has been added, which provides a huge storage cushion if market conditions tighten.
These simultaneous changes–namely abundant supply, the ability to increase production at a breathtaking rate, the re-emergence of fuel switching and the development of a storage reserve–alter how the natural gas market functions.
Since 2008, the total supplies that flow into the U.S. market have increased by 3 Tcf per year, far outpacing growth in core demand.
The only way to absorb this huge increase has been to drive down natural gas prices enough to induce massive amounts of substitution of natural gas-fired generation for coal (see Figure 3).
Coal Displacement vs. $5.50 per MMBtu Baseline
In a coal displacement-driven market, the equilibrium price for natural gas is driven by the amount of price-induced substitution of natural gas fired-generation for coal required to balance the market.
The more gas has to be absorbed, the further prices must fall.
End of Demand Destruction-driven Market
In the post-shale market, the market-clearing price for natural gas is likely to be permanently set by the amount of excess natural gas and the cost availability curve for coal displacement.
The era of demand destruction is over; for the foreseeable future, prices likely will be set by the need to induce demand.
During this same period, the market’s vulnerability to price spikes has been reduced greatly. Three factors have changed.
First, competition between natural gas and coal has driven down prices for both commodities and created a far more robust fuel-switching option.
While many coal-fired generating units are scheduled to be retired, 75 percent or more of the U.S. coal-fired fleet–240,000 to 260,000 MW of coal-fired capacity–likely will continue to operate for many years.
If natural gas prices start to rise, use of these units will increase immediately. This will cut demand for natural gas and relieve upward pressure on the market.
Second, as a result of the massive underground storage increase, for the first time in years–even during cold winters–the amount of available storage capacity is more than adequate to meet market needs.
The fear of winter supply shortages, which previously drove up prices, has become a relic.
Further, a storage reserve provides time to adjust if market conditions tighten.
In the post-shale market in which production can be increased by several billion cubic feet per day in a few months, storage can be replenished quickly.
This eliminates the justification for including a significant risk premium in prices for natural gas.
What does this mean for 2013 prices? And long term? Several conclusions stand out:
1. In a post-shale market, during any given 12-month cycle, prices for natural gas are likely to be driven primarily by the cost availability curve for coal displacement. Moving the demand needle significantly takes many years. In its most recent long-term forecast, for example, the Energy Information Association (EIA) estimates that even at low price levels, demand for natural gas is likely to increase less than 1.5 to 2 Bcf per day from 2012 levels–an average increase of just 0.3 Bcf per day per year. Decreases in U.S. production also occur slowly. Short-term, price-induced coal displacement is likely to be the only effective means to balance supply and demand.
2. Natural gas prices are extremely sensitive to swings in weather and relatively small shifts in production. The magnitude of this impact was demonstrated in December. During November and December, weather was expected to be slightly colder than normal. This forecast, however, proved wrong, with nearly a 200 heating degree-day (HDD) drop in gas-weighted HDD. This loss of demand reduced the likely market clearing price this winter by 35 to 50 cents per MMBtu as a result on loss of weather-dependent demand over 31 days. This potentially reduces wholesale power prices this winter by $3 to $5 per megawatt-hour. During the remaining three winter months in plausible scenarios, HDDs could be 150-250 HDDs lower than expected or up to 100 HDDs higher. Natural gas prices could range anywhere between $2.50 and $3.75 per MMBtu with a commensurate impact on wholesale prices for electricity.
3. Upside potential for natural gas prices is limited, constraining spark and dark spreads except when scarcity conditions apply. Power producers recognize that natural gas prices are not likely to fully recover quickly until exceptionally warm weather arrived in December, but many believed prices in the $4 to $4.50 per MMBtu level were possible in 2013. Further, hope is still widespread that prices will return to the $5 to $6 per MMBtu range by the middle of the decade. This possibility cannot be ruled out definitively partly because of the impact of weather. Hopes for a quick return to $4 per MMBtu prices, however, underestimate the extent to which the U.S. is oversupplied and the momentum that is keeping U.S. production near all-time highs. Even if weather during the remainder of winter matches historical norms and temperatures this summer are hotter than normal, prices at Henry Hub are unlikely to average more than $3.50 per MMBtu, which equates to a dispatch cost for an efficient combined-cycle unit of $23 to $28 per megawatt-hour.
Many factors could drive prices significantly lower, including the potential for continued mild winter weather, a return to normal temperatures this summer or modest increases in production.
It remains possible, therefore, that by the third quarter of 2013, prices could be driven down to as low as $2.75 per MMBtu, shaving an additional $5 to $6 per megawatt-hour off on-peak power prices during nonscarcity conditions in many regions.
The potential for prices to return to $5 to $6 per MMBtu in another few years also is not high.
Two issues are key: potential growth in demand and the supply curve for natural gas.
Increasing demand rapidly, however, is difficult. Further, unless prices sink back to $2.50 per MMBtu–the average for the first nine months of last year–power sector demand for natural gas is likely to be below 2012 levels for several years.
The prospects for higher prices, therefore, depend primarily on the supply side of the production.
As recently as two months ago, many analysts asserted that prices would return to $4 to $4.50 per MMBtu soon and then slowly rise to $6 per MMBtu with further increases in subsequent years.
This prediction was based upon supply curves’ showing that few dry gas-only plays could be developed for prices below $4.25 per MMBtu, with breakeven costs for several plays of $5.50 per MMBtu or higher.
These analysts asserted that as a result of drilling cutbacks during the past two years, production will decline soon.
Once this decline begins, these analysts contended, prices inevitably must rise to induce sufficient production to meet market needs. This market assessment is seriously flawed for three reasons.
First, the data used to develop these supply curves is based upon a snapshot taken at a moment that is already outdated.
To develop these supply curves, analysts frequently used published reserve estimates developed using data that was already one or two years old and has since been superseded.
Further, they relied upon well-by-well production data that looked backward in time and did not fully reflect improvements in techniques for shale development that the industry has implemented.
In the pre-shale world in which estimates changed slowly, use of this data might have made sense.
In a post-shale market, however, in which reserves are continuing to grow rapidly and techniques for shale development are continuing to improve rapidly, it is a recipe for faulty decision-making.
By the end of the year, some analysts already revised their estimates of breakeven costs downward by 10 percent or more. The EIA’s most recent long-term forecast suggests even steeper cuts are appropriate.
Second, published supply curves reflect the expected average breakeven cost for a typical producer in each play.
Production costs for many producers are lower. Encana, for example, has indicated it can earn acceptable returns in Haynesville, the deepest of the dry-only plays, at $3 per MMBtu.
Further, nearly every producer has at least some targets that can be developed at lower costs using current techniques.
Finally, during the next few years, development costs are nearly certain to continue to decline. Yield per well could improve even more dramatically.
Shale is a technology play. Development of most plays began just three or four years ago. Production from some important plays (e.g., Utica Shale Point Pleasant) is just starting.
As Jack Welch, the former CEO of General Electric Co., said recently, shale development is still in its first inning. Like every technology play, further improvements are inevitable.
As a practical matter, given the huge resources available and the speed with which production can be expanded, it is unlikely that prices higher than $4 per MMBtu will be needed to balance the market. Further, breakeven costs for natural gas are likely to continue to decline.
Periodic price declines are inevitable.
Finally, although the breakeven cost for natural gas establishes a ceiling, it does not create a floor.
Instead, from time to time producers overshoot and produce more natural gas than the market requires.
Weather-driven demand also often is likely to fall short of expected levels as occurred in December.
When this occurs, prices drop to induce sufficient price-induced coal displacement to keep the amount of natural gas in storage at manageable levels. In some instances, the plunge in prices could be steep, as occurred last spring.
Managing Downside Risks in an Asymmetrical Market
The shale revolution is creating an asymmetrical market for electricity and natural gas.
The ready availability of low-cost supply is establishing a ceiling on prices for natural gas, which is likely to decline and is creating continued downward pressure on wholesale prices for electricity. At the same time, downside price risks will remain and create an imbalanced risk-to-reward ratio.
To understand potential risk exposure, producers rigorously must assess a range of scenarios and examine a range of assumptions regarding weather, natural gas production and other key market drivers.
This requires new forecasting tools, which make it possible to analyze the electricity, natural gas and coal markets on a fully integrated basis.
In the new asymmetrical, low operating margin paradigm, power producers are likely to become increasingly dependent upon monetization of capacity rights to earn acceptable returns.
Long-term power supply agreements at acceptable price levels affect long-term hedging strategies, and favorable market rules regarding capacity rights are likely to become even more important than they have been.
Sweeping Changes in Grid Operation
The shale revolution also has sweeping implications for the operation of the grid and the physical infrastructure used to generate and transmit power.
The most highly publicized effect is the expected retirement of many coal-fired units.
During the next three to four years, up to 70,000-90,000 MW of coal-fired capacity could be shut down–25 percent of the total U.S. fleet.
The impact of market convergence, however, runs deeper. During the past 12-18 months, far-reaching changes have been occurring in the operation of coal-fired and natural gas-fired units.
In the past, except for planned maintenance, coal-fired units were seldom taken offline for less than five to seven days.
To avoid excessive wear and tear and reduce the risk of forced outages, many older units were operated at relatively high capacity factors even if it would have been economic to reduce output further.
In a post-shale market, however, many coal-fired units are being shut down for months at a time, especially in spring and summer. During the first quarter of 2012, the number of coal-fired units retired or shut down for at least one week climbed by more than 80 units year over year.
Further, even during periods of higher demand, coal-fired units are being cycled far more aggressively than in the past.
Many coal-fired units are being routinely shut down on weekends. Units also are being operated frequently at minimum load.
As a result, operating units that were considered bedrock capacity to be operated at flat out whenever they were available to be dispatched are being operated as peaking units only when demand spikes (e.g., most recently, First Energy’s Sammis units in Ohio).
At some units, plant operators also have started to switch repeatedly between coals, using expensive, high-quality BTU coal when power prices are high to maximize output but switching to less expensive Powder River Basin coal to reduce dispatch costs when prices are low.
Post-shale in low-margin markets, skill in operating coal-fired units and decisions about when to bid capacity into markets for ancillary markets are likely to become increasingly important.
Natural gas-fired generating units are operating at significantly higher capacity factors than in the past.
Partly because of operating limitations for coal-fired units, however, most combined-cycle units still are being ratcheted down during evenings and weekends.
Combined-cycle units also are still being used heavily to provide ancillary services and as a contingency resource to adjust to unexpected swings in production from intermittent resources. Combustion turbines also are seldom used for coal displacement.
Even in regions with high levels of coal displacement, the average capacity for combined-cycle units is frequently no more than 60 to 70 percent.
The capacity factor for combustion turbines is only slightly higher than historical norms.
The scenario envisioned by many analysts in which natural gas units might operate at 85 to 90 percent capacity factors on a sustained basis shows no signs of materializing.
Instead, in an increasing number of regions, the potential to use existing natural gas-fired generating units is close to being maxed out.
Significant increases in gas use are not likely to occur until new combined-cycle units are added to replace coal-fired units that are operating as baseload units.
Shift for LMP Pricing, Construction of New Transmission
The retirement of huge amounts of coal-fired capacity coupled with the other changes noted are changing the resource base available to serve load in many markets, with far-reaching consequences for locational marginal pricing (LMP), designation of must-run units and price caps imposed by grid operators in structured markets.
The location of congested nodes and the frequency with which transmission is constrained a few years from now may bear little relationship to what it has been historically.
LMP prices at some nodes might be considerably higher, triggering more frequent intervention by grid operators.
In other situations after new natural gas-fired capacity is added, historically high-priced nodes could disappear.
The use of the grid to transmit power between geographic markets likely will change significantly.
With less coal-fired generation available and minimal differences in dispatch costs between natural gas and coal-fired units, the power flows that have occurred historically between low-cost and high-cost markets have started to fall sharply.
Further, in many instances as coal-fired units with high dispatch costs are taken off the grid and additional natural gas-fired capacity is added, there might be little economic justification for transmitting power between regions.
An increasing number of transmission projects designed for when market conditions were significantly different might never be built, and the role of regional transmission organization might need to be re-examined.
Adjusting to a Different World
The U.S. electricity market has seldom changed as rapidly or radically as it is changing now.
Most power producers and investors recognize the need to thoroughly review strategic plans that seemed valid one or two years ago.
In reassessing asset values and reformulating strategic plans, though, it is important to remember that the past is no longer prologue.
Far-reaching changes are occurring in grid operation, with profound consequences for the industry.
Further, the hard questions regarding likely future price trends no longer can be shrugged off because they go to the heart of sound decision-making.
Andrew D. Weissman is senior energy adviser to Haynes and Boone LLP, a 550-attorney international law firm with offices in 11 cities. He is also editor-in-chief and publisher of “Energy Business Watch,” an energy market analysis service.