The Changing Landscape of US Power Generation

June 1, 2011
Brandon Blossman bblossman@tudorpickering.com (713) 333-2994 George OLeary goleary@tudorpickering.com (713) 333-2973

Crack a WindowU.S. Getting Gassier

*Important Research Disclosures on page 30 of this document

Contents
Takeaways Includes overall conclusion (p.3) FAQs Includes predictions, switching stats, etc.(p.4) Background The build out of the U.S. generation fleeteach technology had its day (p.5) Three drivers have fallen into place pushing coal and gas-fired generation into direct competition (p.6-12) An overabundance of efficient gas-fired generation resulting from an over-enthusiastic build out of capacity in the 2000s (p.6-7) Cheap gas driven by shale development (p.9) Expensive coal driven by an increasingly expensive Eastern coal cost structure (p.10) U.S. Generation supply stack circa 2006 vs. today Supply stack = generation arranged in order of variable cost 2006 Relationship between least efficient coal and efficient gas/combined cycle gas turbines (CCGTs)coals variable cost less than gass coal runs before CCGTs Too much generation capacity and lower demand vs. expectations meant CCGTs ran less than expectedway less than expected 2011 Coal and CCGT fleet intermixed; direct competition between Eastern (BIT) coal and CCGT fleets as variable costs move to parity.

The mechanics of switching- coal vs. gas price competition (p.13-20)

Regional case study Southeast (p.21-24) Price driven changes in switching behavior (p.25) Generation additions forecast (p.26) Break-even coal/gas prices (p. 27)

Glossary refer to this for unfamiliar terms/acronyms (p. 28) Supply stack assumptions (p.29)
2

Takeaways
Gas-fired generation has steadily gained market share since the 1990s, driven first by a gas generation overbuild then more recently by the collapse of the coal-gas price spread. Going forward, we expect incremental power demand to accrue to gas generation (ex- renewable additions)implies little to no change in power-generation coal demand. While power-gen gas demand has increased significantly, doubling from 10bcf/d in 1990 to more than 20 bcf/d last year, there is still more where that came from via an underutilized combined cycle gas turbine (CCGT) fleet. The ~230 GW CCGT fleet ran only 40% of the time last yearmoving from 40% to a 70% capacity factor would increase gas demand by 12bcf/dlikely possible over time but requires some combination of increasing power demand, enhanced transmission infrastructure and a continued tight coal/gas price spread. ORjust silly-low natty prices ~$2/mcf, a situation that we dont see as sustainable for any length of time. THE CALL/CONCLUSION: We saw coal to gas switching ranging from 2.5 -3.0bfc/d from 2009 to date: A 3bcf/d annual average in 2009, A decrease to 2.5bcf/d last year with a slightly looser coal-gas price spread and YTD 2011, a touch more at 2.8bcf/d average through March on a coal-gas price spread similar to 2009 averages. Were expecting coal and gas prices to remain closely tied at least through 2012 maintaining switching levels at ~3bcf/d. Longer-term were expecting some unwinding of price driven switching. Though we expect the decrease in switching driven demand will be more than offset by incremental power demand. Given a 1% annual power demand growth assumption, we forecast gas demand from power generation to be 27bcf/d by 2017, a 30% increase from 2010 levels.
3

FAQ
Best to get this out of the way firstanswers to Coal to gas switching questions that are easy to ask but often challenging to answer. How much switching has taken place? - Our regression model normalized to 2006-08 data shows: 2009: 3.0 bcf/d 2010: 2.5 bcf/d 2011: 2.8 bcf/d (forecasted)

How much more switching will there be tomorrow? 2.5-3.0bcf/d near-term. We expect price-driven switching to decrease over longer term as variable fuel input costs diverge given our long term price deck for Eastern coal at $75/ton and gas at $6/mmbtu in 2013. What is the maximum amount that can switch? Theoretically ~12bcf/d. This would require low, low natural gas prices allowing the CCGT fleet to overcome transmission costs and constraints. At what gas price does switching occur? At a $4.3/mmbtu much of the eastern coal-fired fleet has a similar cost structure to efficient gas-fired generationassuming $75/ton eastern coal price. How low can gas prices go? Again assuming a $75/ton coal price, materially below $4.3/mmbtu on a sustained basis should be self correcting over time. What happens if switching unwinds? How do gas prices react? It depends on why it happens. Higher power demand? More pricing support for both gas and coal. Higher gas prices? Self correcting (to a degree)higher prices should unwind switching Lower coal prices? Not likely, high cost of supply put a floor on coal prices.

Generation Build Out Each Technology Has Its Day

Additions to U.S. Electricity Capacity by Fuel Type (Pre-1951 to 2009)
180
Merchant overbuild Coals King

First, a bit of history. The chart to the left puts some context around the aging U.S. coal fleet and the much newer, abundant, efficient gasfired combined cycle gas turbine fleet (CCGTs). Over the course of the build-out of the current U.S. power generation fleet, each technology had its day. 1950s-early 1970s New coal generation satisfied the robust growth in electricity demand supported by a plentiful, inexpensive Eastern coal reserve base. This was long enough ago that the 70s are cool again; makes the weighted average age of U.S. coal fleet ~40 years old. 1970s-early 1980s Ushered in the Nuclear Age with the promise of cheap long-term base load power Mid 1980s to mid 1990s Regulatory uncertainty (the promise/fear of deregulation) resulted in an under-build of all generation types Late 1990s-early 2000 pendulum swings to big over-build with an explosion of new build capacity by nonutility entities. Technology of choice by a wide margin: CCGTs

140

Gigawatts (GW)

100

Nuclear era Under build

60

20
Pre-1951 1951-1960 1961-1970 1971-1980 1981-1990 1991-2000 2001-2007 2007-2009

-20

Coal

Hydroelectric

Natural Gas

Nuclear

Source: EIA/DOE, Tudor, Pickering, Holt & Co.

Last Build Cycle Natty-fueled CCGTs Won Hands Down

Net Capacity Additions Coal vs. Gas (1990-2009)
250

No contest really, gas generation new-builds out numbered coal by more than 25 to one by the end of 2010 (net of retirements). Relative to coal generation, Combined cycle gasturbines (CCGTs) are: Cheaper less than 50% of the capital cost of a new coal facility. More efficient use 30% less fuel (mmbtu) to produce a MWh of electricity. Easier to site and faster to build permit to operation in as little as two years vs. four+ years for coal. Cleaner 50% less CO2 and in some cases several times lower SO2, NOX and Mercury emissions. These are the same reasons that CCGTs will dominate generation capacity adds in the near to midterm.

200

Gigawatts (GW)

150

Coal

Natural Gas

100

50

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Cheap and efficient more than offsets higher nat gas prices. Despite natural gas being 2-3xs coal prices during the planning stage for much of the CCGT fleet (1995- 2000), the lower construction costs and higher efficiency of the CCGTs more than offset higher input fuel prices making CCGT the easy choice to satisfy new generation needs.

*Sources: EIA, SNL, TPH Research

Nat Gas CCGT Overbuild Still With Us

Additions to U.S. Electricity Capacity (1990-2010)
50 45 40

Gigawatts (GW)

35 30 25 20 15 10 5 0

In 1990, the U.S. had 800 GW of generating capacity. So, with peak demand growing by 1.7%/year (20002007), the industry needed to add about 14 GW/year to ensure adequate supply. As the industry grappled with the uncertainty of deregulation during the 1990s, utilities all but stopped building new generation (additions only about ~4 GW/year) and merchant generators didnt really exist yet.

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

(5)

2010

U.S. Electricity Capacity v. Peak Demand (1990-2007)

1,100 1,000

Gigawatts (GW)

900 800 700 600 500 400

But, boy-o-boy did the merchants make up for it in the 2000s! Peak year for new builds was 2002 two years after the CA energy crises and a year after the collapse of Enron. The average 2000-2003 capacity add was ~37 GW/year. Over-build resulted in excess power generation capacity. 2010s generation capacity exceeded peak demand by 31% well above the ~15% or so insurance margin grid reliability requires.

Trendline of peak demand growth

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Peak Demand

Capacity

2010

Source: EIA/DOE, Tudor, Pickering, Holt & Co.

Gas vs. Coal Prices Changing Relationship

80 70 60 50
$/MWh

CCGT planning period (Late 1990s)

Coal vs. Nat Gas Prices 1990-2011YTD

In annual average cost of power production CCGT vs Eastern (Bituminous) Coal $/MWh

Gas

Industry looks at historic ~$2 natural gas prices and gets ready for a big merchant (nonutility) generation build out Gas at a premium to coal but adjusting for CCGT construction cost and efficiency advantages, CCGTs are obvious choice for new builds.

40 30 20 10 -

Coal

Build out period (2000-07)

Build out Period CCGT/Coal Full cycle economic parity Ouch! Variable cost Competition

Full cycle economic parity Average coal/gas price relationship during this period implies indifference between building a CCGT or new coal generation

Variable cost competition (2009+)

Planning Period Advantage CCGT

Structural changes in coal and gas production move coal prices up and gas prices down. Cost of power production coal vs. gas on a $/MWh basis moves to parity.

*Sources: EIA, SNL, TPH Research

Gas Price Down WHY?

US Wellhead Supply Forecast [Shale Only]
25 2,000 1,800 20
Updated Forecast

GAS Gas prices have fallen relative to coal prices on a $/mmbtu basis over the last few years due one main factorSHALE! Shale plays = TooMuchGas
Supply growth at current rig countsupply grows 2.5bcf/d per our estimates. Demand growth only ~1bcf/dunless low coal/gas spread yields lost market share from coal.

1,600 1,400

Natural Gas Production (bcf/d)

1,000 10 800 600 5 400 200 0

Historical Production Historical Rigcount Old Production Forecast Old Rigcount Forecast Current Forecast Current Rigcount

12/99 12/00 12/01 12/02 12/03 12/04 12/05 12/06 12/07 12/08 12/09 12/10 12/11 12/12 12/13

Rig Count

15

1,200

US Wellhead Supply Forecast [Onshore]

70
Updated Forecast

5,000 4,500 4,000 3,500 3,000

65 60 55 50

Gas supply has increased meaningfully due to the emergence of shale gas production thanks horizontal drilling and fracturing! See supply study for addl detail - TPH Research Report: Natural Gas Supply Study Update 10-15-10

Natural Gas Production (bcf/d)

2,500 45 40 35 30
Historical Production Historical Rigcount Old Production Forecast Old Rigcount Forecast Current Forecast Current Rigcount

2,000 1,500 1,000 500

12/99 12/00 12/01 12/02 12/03 12/04 12/05 12/06 12/07 12/08 12/09 12/10 12/11 12/12 12/13

Rig Count

*Sources: EIA, SNL, TPH Research

Coal Prices Up WHY?

COAL
$70 $60 $50
2002-2010 2007-2010

Production Cost Creep Over Time ($/ton)

10% CAGR 14% CAGR
$40.95 $41.20 $46.65 $50.48

$60.31

$/ton

$40 $30 $20 $10 $0

$28.64 $28.23 $30.50

14% y/y cost creep

$34.00

Eastern Coal prices have risen largely due to continued operating cost pressure in Appalachia. Marginal cost of production sits at $70+/ton (includes capex) and environmental /regulatory pressures add to costs. Material cost creep has become a Q/Q phenomenon.

2002 2003 2004 2005 2006 2007 2008 2009 *Based on MEE production data (MEE large public CAPP producer)

2010

Appalachia Production on the Decline

420 400 380

CAPP total production ~215mm tons in 2011. Average production costs based on public company data are ~$65/ton + ~$10/ton maintenance capexso marginal production cost is ~$75/ton. Supply cost rather than demand, is the main driver for Eastern coal prices in the US. In addition, Appalachian coal production has been on a steady decline for many yearsmore challenging geology both increases cost and reduces production.

mm tons

360 340 320 300 2005 2006 2007 2008 2009 2010

*Sources: Massey Energy, EIA, SNL, TPH Research

10

CCGTs More Run Time = Market Share Gains

CCGT Fleet Capacity Factors
45% 40% 35% 30% 25% 20% 2003 2004 2005 2006 2007 2008 2009 2010

Gas market share gains driven by 1) CCGT fleet build-out, 2) increasing utilization of that fleet and 3) reduced coal/gas commodity spread post 2008. In reaction to the under-build of the 1980-90s, the market built way more CCGTs than needednot surprisingly those CCGTs ran much less than forecasted. Once built, the CCGT fleet competed with coal generation on a variable cost basis and generally lostmade for low run times with capacity factors ~30%. CCGT capacity factors are edging up butstill a long ways to go. They can and do operate as baseload facilities (for example CCGT-based cogeneration facilities producing both steam and power often run 24/7 providing process steam). The underutilized CCGT fleet allows for significant (really significant) gas market share gains. The 230GW fleet of CCGTs averaged ~16bcf/d of consumption in 2010 running at a 40% capacity factor (CF). Increasing CF to 70% (similar to the 2010 subbituminous coal fleets capacity factor), would add ~12 bcf/d incremental gas demand.thats a lot of gas consumption. 12 bcf/d of natural gas demand is an increase of 18% over total US natural gas demand and a 60% increase in gas demand from power. At current power demand growth trajectory, which includes coal/gas switching, gas demand for power generation would get to 30Bcf/d by ~2020.

Gas Market Share as % of Gas + Coal Generation

40% 35%

Gas Market Share

30% 25% 20% 15% 10% 5%

2003 2004 2005 2006 2007 2008 2009 2010

0% .

*Sources: EIA, SNL, TPH Research

11

Post Buildout Price Spread Supports Continued Gas Gains

Coal-Gas Price Spread vs. Gas Market Share
$50
Gas coal power generation spread in $/MWh Peak of CCGT buildout Gas demand increasing @ 3% CAGR 1998-2011

40% 35% 30%

$40 $30 $20 $10 $1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Gas generation market share flat-ish prior to CCGT build out

25% 20%

Gas generation market share (gas/gas + coal)

15% 10% 5% 0%

1990 -1997 Gas generation market share was relatively stable at 20% of combined gas and coal generation output. Relationship between coal and gas prices also shows limited volatility. 1998-2008 New gas generation drives gas market share and demand increases while commodity spread widens 2009-Present Commodity Spread collapses supporting further gains in market share Future Power demand growth supplied by natural gas along with mandated renewables. Coal gen output is essentially flat with rising capacity factors offset by retirements of older coal generation.
12

*Sources: EIA, SNL, TPH Research

Gas Market Share

$/MWh

Coal to Gas Switching Price Driven Competition

An overbuilt and underutilized CCGT fleet allows for big increases in natural gas consumption. Both over time with increasing load demand, and in response to coal vs. gas price competition. Post 2008, lower natural gas and higher coal prices have resulted in variable cost parity for some gas and coal units. Some of the most efficient gas and least efficient coal units are competing for market share with similar cost structures on a $/MWh of power production basis. Coal to Gas Switching Its a confusing term. Once upon a time, fuel switching in the power generation world generally meant switching between natural gas and oil at a given facilityeasy to explain, easy to quantify. Coal/gas switching means something like: A gas generation facility running more than a nearby(ish) coal facility due to a change in the price differential between the two input fuel costs. No problem right? Well, in practice it ends up being hard to definerunning gas generation more than what baseline? What coal generation are you comparing to? And finally, what is its baseline runtime expectation? Plus, additional considerations like spot vs. contracted fuel costs make a precise answer to how much coal to gas switching is going on difficult to determine.

13

The Mechanics of Switching Generation Supply Stack Decomposition

$135.00

Reflects Henry Hub Gas @ $4.30/mmbtu

$115.00

$95.00

Gas-fired Steam Turbines All older, less efficient generationsame technology as coal-fired generation, replaced by combustion turbinebased technologies for gas-fired generation built in the last 2+ decades.

Gas-fired Combustion Turbines New simple cycle (vs. combined cycle) gasfired generation. Less efficient than CCGTs but quicker starting with lower fixed costs

Bituminous Coal Eastern coal fueled generation

$/MWh

$75.00

$55.00

CCGT Young, efficient gas-fired generation

$35.00

Other Coal (LIG/SUB) Lignite (generally mine mouth facilities) and low-cost PRB fueled facilities near (sub-bit 1) and further (sub-bit 2) from MT and WYs Powder River Basin
$15.00 0 50,000 100,000 150,000 200,000 250,000

CapacityMW

Break out of coal and gas fired generation plants at current market prices ordered left (lower cost) to right (higher) by variable cost of power generation in $/MWh. Other generations types with lower variable costs not shown include nuclear, hydro and renewables. Lignite and Subituminous plants generally dont compete with CCGTs All coal plants virtually always cheaper than gas steam & gas turbine generation CCGT and BIT g d generation variable costs are ~in-line, hence the overlapping supply stacks aboveCOMPETITION! ti i bl t i li h th l i g l t k b COMPETITION!

NOTE: SUB plants split into two categories by distance. SUB 2 plants located further from coal source higher transport costs.

See appendix (page 30) for assumptions *Sources: EIA, SNL, TPH Research

14

The Mechanics of Switching U.S. Generation Supply Stack Circa 2006 vs 2011
$121

2006 Assumptions: Gas price - $7/Mcf Delivered Prices: BIT - $50/ton SUB 1 - $30/ton SUB 2 - $45/ton LIG - $18/ton

2006 Power Generation Supply Stack Coal & Gas Generation Only

$101

Gas

$81

Supply stack = generation arranged in order of variable cost (fuel + transport + variable operation and maintenance expense) 2006 - Note relationship between least efficient coal and most efficient gas (CCGT fleet)coal runs before CCGTs

$/MWh

$61

Coal
$41

$21

$1 400,000 500,000 600,000 700,000 800,000 Capacity Megawatts 900,000 1,000,000 $121 1,100,000 1,200,000

2010 Power Generation Supply Stack Coal & Gas Generation Only
2010 Assumptions: Gas price - $4.3/Mcf Delivered Prices: BIT - $80/ton SUB 1 - $35/ton SUB 2 - $50/ton LIG - $20/ton

Illustrative - variations in power demand and generation supply along with transmission constraints means the full U.S. supply stack wont actually dispatch as illustrated. However, conceptually this is big picture correct 2010 Coal and CCGT fleet intermixed direct competition between eastern (Bit) coal and CCGT fleets

$101

$81

Coal & Gas Generation Overlap Zone

h W $61 M / $
$41

$21

$1 400,000 500,000 600,000 700,000 800,000 Capacity Megwatts 900,000 1,000,000 1,100,000 1,200,000

15

CCGT fleet vs. Bituminous Coal by Nat Gas Price

$75.00

Gas Displacing Coal at Various Gas Prices*

$65.00

$55.00

@ $5.50/Mcf - Coal/gas generation relationship similar to historic normmajority of coal dispatched before any of the CCGT fleet. Little meaningful price competition

Bituminous coal generation @ $75/ton

@ $4.25/Mcf Much of the CCGT and Coal fleet is in direct competition with similar variable cost structures. Who wins and is dispatched first depends on a variety of local unit/local market level factors

$/MWh

$45.00

$35.00

$25.00

@ $3.00/Mcf The vast majority of the CCGT fleet is cheaper to run though transmission constraints limits complete displacement of coal fleet. If all CCGTs ran at a 70% CF rather than the current ~40%, Gas demand would increase by 12 bcf/d.

$15.00 0 50,000 100,000 150,000 200,000 250,000

Capacity Megawatts

*Gas/Coal prices shown as Henry Hub/NYMEX pricingsupply stack includes transport **See appendix (page 30) for assumptions 16

*Sources: EIA, SNL, TPH Research

The Competitors in this corner: The CCGT Fleet

250 233

200

Gas Fleet Name Plate Capacity (GW)

CCGT fleet is new, big and by far the largest consumer of gas of the gas-fired gen typesand currently well underutilized The fleets efficiency means that when it comes to coal/gas competition, the CCGT fleet is on the front lines
85

Capacity GW

150

139

51%
100

30%
50

19%

The balance of the gas fleet well behindsee next page. The next tranche of gas-fired generation is 30% less efficient, in-line w/ coal fleet on a heat rate basis.
1%
3

0 CCGT GT ST Generation Type IC

The 233GW of CCGTs averaged ~16bcf/d of consumption in 2010 running at a 39% capacity factor (CF). Increasing CF to 70%, would add ~12bcf/d incremental gas demand.
Heat Rate (Btu/KW) 7.44 8.92 8.35 11.28

Fuel Type Natural Gas CCGT GT IC ST

% of Total TWh bcf/d Gas + Coal Produced Power Gen Consumed 794 89 2 96 28% 3% 0% 3% 16.2 2.2 0.0 3.0

% of Total US Gas Demand 25% 3% 0% 5%

Capacity Factor 39% 4% 0% 5%

By the end of 1997, only 47GW (20%) of the current CCGT fleet existedbingo bango, most plants are relatively new.

*Sources: EIA, SNL, TPH Research

17

The Gas Generation Supply Stack

$140.00

Gas Generation Supply Curve

$120.00

$100.00

$80.00

$/Mwh

Gas turbine, Steam turbine, and Internal Combustion FLEET

$60.00

$40.00

CCGT FLEET

Doesnt Compete w/ Coal

$20.00

Competes w/ Coal
$0.00 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 Capacity Megawatts

CCGT generation = more efficient lower variable costs GT, ST, IC generation = less efficient higher variable costs
*See appendix (page 30) for assumptions
*Sources: EIA, SNL, TPH Research

18

The Competitors in this corner: The Coal Fleet

Age (capacity) X<30 30-45 45+ (X<250MW) 45+ (X>250MW) Total
200 180 160 140 136

Coal Fleet Capacity GW LIG BIT SUB 2.10 31.65 35.34 9.19 103.30 72.04 0.17 13.39 9.87 0.00 37.90 18.31 11.5 186.2 135.6
186

Total 69.10 184.53 23.43 56.21 333.3

Coal generation varies by efficiency (boiler type/size/age) and coal type. The least efficient, generally the oldest and smallest coal units, are the first to lose market share to the CCGTs. Market competitiveness of coal generation is driven by a combination of efficiency and fuel type/source. Coal is not a homogenous commodity: Coals with lower heat content are less expensive, but more costly to transport. In order of heat content: Lignite. Low energy content coal only used at mine mouthtoo expensive to transport (5,000 8,300 Btu/lb).

Coal Fleet Name Plate Capacity (GW)

Capacity GW

120 100 80 60 40 20 0 BIT SUB Generation Type LIG

56% 41% 30% 19% 3%

11

PRB = subbituminous coal: More energy content than lignite but at a transportation disadvantage realtive to higher energy content bituminous. (8,400-9,400 Btu/lb). Bituminous coal is 10,000+ Btu/lb coal. This encompasses Western Bit, Illinois Basin and Appalachian coal. There is approximately 15GW of old, inefficient, high variable cost (Bituminous) coal generation in the USthe most likely to be retired. Coal Supply Stack on following page

TWh Fuel Type Produced Coal LIG SUB BIT 81 825 931

% of Total Gas + Coal Power Gen 3% 29% 33%

~MM tons Coal Consumed* 68 490 414

Capacity Heat Rate Factor (Btu/KW) 62% 69% 57% 10,998 10,458 10,223

*Assume LIG 6,500 Btu/lb, SUB 8,800 Btu/lb, BIT 11,500 Btu/lb to estimate tons of coal consumed via conversion of actual mmbtu consumed. KEY: LIG = Lignite, SUB = subbituminous, BIT = bituminous *Sources: EIA, SNL, TPH Research

19

The Coal Generation Supply Stack

$50.00 $45.00

Bituminous Plants
$40.00

$35.00

$/Mwh

$30.00

$25.00

Subittuminous Plants Group 1

Subittuminous Plants Group 2

The most expensive, least efficient coal falls in this group. This competes for market share with the CCGT fleet. This segment shown on page 14

$20.00

$15.00

Lignite Plants

$10.00 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000

Capacity Megawatts

Lignite generation = lowest variable costs (mine mouth generation + cheap commodity) SUB generation = higher variable costs (cheap commodity, high delivery cost) BIT generation = highest variable costs (expensive commodity, high cost production)
*SUB plants split into two categories by distance. SUB 2 plants located further from coal source higher transport costs. **See appendix (page 30) for assumptions *Sources: EIA, SNL, TPH Research

20

Hotbed of Switching Southeast and PJM

Y/Y Delta in Gas Consumption by Region (as % of Total Consumption)

GAS
2003 2004 2005 2006 2007 2008 2009 2010 YTD 2011

MISO -1.6% 0.5% 1.5% -1.0% 1.0% -1.4% -0.3% 1.1% 0.0%

PJM -1.1% 1.1% 0.5% 0.3% 1.7% -0.4% 1.4% 2.0% 2.8%

ERCOT -1.9% -1.1% 1.2% 0.0% 0.1% -0.5% -0.8% -0.6% -1.6%

West 0.9% 3.3% -1.6% 1.8% 1.8% 1.0% -0.7% -1.8% -3.9%

Northeast Southeast -3.3% -2.9% 0.4% 2.0% 1.3% 2.2% 1.7% 1.8% 1.1% 1.8% -0.8% -0.2% -0.2% 3.8% 2.2% 3.7% 1.4% 2.7%

Other -0.6% -0.1% 1.9% 1.1% 1.5% -0.4% -0.2% 0.5% -3.0%

Annual Delta MMcf/d (1,469) 893 1,117 969 1,693 (476) 556 1,386 (48)

Annual Consumption MMcf/d 14,069 14,969 16,080 17,047 18,744 18,270 18,828 20,213 2,882

Y/Y Delta in Coal Consumption by Region (as % of Total Consumption)

COAL

2003 2004 2005 2006 2007 2008 2009 2010 YTD 2011

MISO 0.5% 0.5% 0.0% -0.2% 0.6% 0.3% -2.0% 1.0% -1.9%

PJM 0.2% -0.5% 1.0% 0.0% 0.2% -0.5% -2.9% 0.8% -1.1%

ERCOT 0.5% 0.1% -0.1% -0.2% 0.1% -0.1% -0.7% 0.8% 0.7%

West 0.1% 0.0% 0.0% -0.8% 0.3% 0.2% -0.4% 0.2% -0.5%

Northeast 0.1% 0.0% 0.0% 0.0% 0.0% -0.1% -0.5% -0.1% -0.2%

Southeast 0.4% 0.5% 0.5% 0.3% 0.5% -0.5% -3.9% 1.1% -1.3%

Other 0.8% -0.1% 0.5% -0.1% -0.2% 0.3% -1.1% 0.3% 0.1%

Annual Total Annual Consumption Delta (000s) Tons (000s) 26,475 1,014,058 6,465 1,020,523 20,925 1,041,448 (10,891) 1,030,556 16,239 1,046,795 (4,461) 1,042,335 (107,651) 934,683 41,369 976,052 (6,976) 163,793

Gas consumption has been steadily rising over the majority of last decade, both nominally and relative to coal. Coal consumption was relatively flat from 2003-2010; uncertain environmental regulations played a big role in capping the growth of coal generation. In 2009, the fundamentals changed as the coal-gas spread thinned. This coupled with the environmental regulation backdrop drove gas demand up 3% while coal demand fell 11% As seen above, the Southeast, PJM, and the Northeast have all seen gas demand tick up more than other regions, again both nominally and relative to coal over the last few years.
*Sources: EIA, SNL, TPH Research

21

Whats the Southeast fleet look like?

2010 Generation Capacity % of SE Capacity % of total US TWh GW Fossil Gen Factor Fossil Gen GAS FLEET CCGT GT COAL FLEET LIG SUB BIT 250 24 8 110 298 67.5 41.1 1.2 17.1 62.0 36% 22% 1% 9% 33% 42% 7% 74% 73% 55% 9% 1%
GWh Generated
65% CF

400,000
354,939

SOUTHEAST - CCGT vs. BIT Generation/Capacity Factor

350,637 297,680 269,518 250,090 212,541

350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 2007

*CF = Capacity Factor 29% CF 65% CF 173,105

0% 4% 11%

65% CF 176,695

50% CF

55% CF 36% CF 42% CF

The Southeast fleet represents about 25% of the total U.S. generation fleet and has a heavy concentration of CCGT and BIT generation ~70%. The Southeast CCGT fleet includes ~30% of the overall US CCGT fleet

30% CF

2008 BIT Generation GWh

2009 CCGT Generation GWh

2010

The CGGT fleet in the region is in direct competition on a variable cost basis with the BIT coal fleet. As you can see above, CCGT & BIT power generation quantities have begun to convergegas generation has stolen market share from coal as the relative pricing of the two input fuels has converged. Given the abundance of CCGT/BIT, generation in the region, the magnitude of switching is greater than in other areas of the country. The following slides illustrate this point with month to month detail

*Sources: EIA, SNL, TPH Research

22

Regional Case Study - Southeast

45% 40%

2008-2011 Gas/Total Fossil Fuel Consumption Ratio Southeast CCGT /(CCGT + BIT) Consumption Demand responds to price over the
long term and short term

40.00 35.00 30.00

Gas/Total Fossil Fuel Consumed (%)

20.00 30% 15.00 10.00 5.00 20% 0.00 -5.00 15% -10.00 10% -15.00

25%

Gas/Total Fossil Fuel Consumption

Spread

Confirmation! The southeast is full of both coal and gas generation capacityspecifically CCGT and BIT generation fleets that we believe compete on a variable cost basis. The above indicates a tight correlation between the relative cost of power production from gas vs. coal and gass market-share of total power-gen fossil fuel usage. As coal prices have risen and gas prices have fallen, weve seen gas demand for power generation in the Southeast both in absolute terms and in market share tick higher. The chart above demonstrates thisthe green line indicates gas prices relative to coal prices; demand for gas relative to total fossil fuel demand responds ~concurrently to movements in the relative price of the two commodities. The following page puts some numbers around the above.
*Sources: EIA, SNL, TPH Research

23

Coal vs. Gas Price Spread $/MWh

35%

25.00

Regional Case Study Southeast (contd)

Delta in Gas Demand for Power Generation - Southeast
BIT Coal Consumption CCGT Gas Consumption $/MWh Price Spread Gas Demand bcf/d Delta vs. 2007 bcf/d Delta vs. 2007 % Sep-07 Sep-08 Sep-09 Sep-10 308,950,757 294,051,864 215,695,311 250,408,800 126,669,337 113,323,158 153,865,243 171,819,641 $20 $4 -$5 -$6 4.2 3.8 5.1 5.7 0.9 1.5 21% 36%

Lets put some numbers around this. September has been a big month for switching in recent history. September 2010 vs. 2007 - incremental +1.5 bcf/d or 36% in CCGT gas demand in the Southeast. We performed a similar analysis for the PJM region and reached the same conclusionthe relative coal/gas price does affect power generation behavior and leads to switching. In PJM, CCGT gas demand was up nearly 50% or 0.55bcf/d in 2010 vs. 2007.

*Consumption figures in mmbtu

Delta in Gas Demand for Power Generation - PJM

PJM BIT Coal Consumption PJM CCGT Gas Consumption $/MWh Price Spread Gas Demand bcf/d Delta vs. bcf/d Delta vs 2007 %
*Consumption figures in mmbtu

Sep-07 Sep-08 Sep-09 Sep-10 277,525,456 272,402,927 209,816,668 244,719,574 35,944,424 39,263,723 53,401,543 52,375,773 $20 $4 -$5 -$6 1.2 1.3 1.8 1.7 0.6 0.5 49% 46%

*Sources: EIA, SNL, TPH Research

24

Coal/Gas Price Spread Impacts Switching

40.00 35.00 30.00 25.00 20.00

Generation Gas Demand Actual vs Predicted - Bcf/d

Actual less Predicted Gas Demand Estimates Coal Gas Price Spread

7.0

5.0

Estimates
Bcf/Day
3.0

$/MWh

15.00 10.00 5.00 0.00 (5.00) (10.00)

1.0

(1.0)

Jan-08

Jan-09

Jan-10

Nov-08

Nov-09

Nov-10

Sep-08

Sep-09

Sep-10

Jan-11

May-08

May-09

May-10

May-11

Mar-08

Mar-09

Mar-10

(15.00)

Mar-11

Sep-11

Jul-08

Jul-09

Jul-10

Jul-11

(3.0)

Above is a historical presentation of coal/gas switching and our April/September 2011 estimates based on our regression model. Short term switching is primarily driven by the relative prices of coal/gas. We saw switching range from an annual average of 2.5 -3.0 bfc/d (2009 to date). Were expecting coal and gas prices to remain closely tied at least through 2012 maintaining switching levels at ~3bcf/d.

*Sources: EIA, SNL, TPH Research

25

Generation Additions Forecast

Cumulative Incremental Generation (2010 Base) 2011 Expected Cumulative Change in Power Demand (TWh) Incremental Change in Generation - Non-Gas Renewables Capacity Adds (GWs) Change in Renewable Generation (TWh) Coal Capacity Adds (GWs) Expected Generation (TWh) 70% CF Capacity Retirements (GWs) Expected Generation (TWh) 50% CF Change in Coal Generation (TWh) Total Non-Gas Natural Gas Total from Gas (TWh) Total from Gas BCF/day
Incrementa l Cha nge i n Ga s Dema nd (BCF/da y)

2012 145

2013 188

2014 231

2015 274

2016 318

2017 363

103

We expect the vast majority of incremental power demand over time to accrue to gas-fired generation. Gas demand from power generation is a function of power demand growth less incremental new-build generation. Our forecasted supply through 2017 assumes: no incremental nuke capacity, flattish coal generation output and a muted renewable generation build-out. Renewable additions average ~6GWs/year 15% 2010s renewable capacity . A good amount but small potatoes vs. a 1,000GW total U.S. generation portfolio.

14 36 3 17 (5) (22) (5) 31 72 1.4

1.4

22 57 8 46 (10) (45) 2 59 86 1.7

0.3

32 84 8 46 (15) (67) (21) 63 124 2.4

0.7

37 98 8 46 (20) (89) (43) 55 176 3.4

1.0

43 113 8 46 (26) (112) (65) 48 227 4.3

1.0

48 125 10 59 (31) (134) (75) 50 269 5.2

0.8

52 137 12 71 (36) (156) (85) 51 311 6.0

0.8

4,600 4,400 4,200

Electricity Demand
1997-2017E

TWh

4,000 3,800 3,600 3,400 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017

Power demand forecast predicated on 2% GDP growth with an electricity demand ratio of 0.5 or a 1% power demand growth post a 2011 2.5% rebound.

*Sources: EIA, SNL, TPH Research

26

Nat Gas Breakeven Varies on a Plant by Plant Basis

Midwest Scrubbed PRB Break even with Natural Gas Northeast Unscrubbed NAPP Break even with Natural Gas

No single coal vs gas breakeven price

7.0 12.0 $65 $12 $5.36 $4.96

CCGT Heat Rate Coal Generation Heat Rate Coal Price ($/ton) Coal Transport ($/ton) Breakeven Gas Price @ Del Point Breakeven @ Henry Hub
Southeast Scrubbed NYMEX Break even with Natural Gas

7.0 9.5 $15 $24 $3.18 $3.03

CCGT Heat Rate Coal Generation Heat Rate Coal Price ($/ton) Coal Transport ($/ton) Breakeven Gas Price @ Del Point Breakeven @ Henry Hub
Northeast Scrubbed NYMEX Break even with Natural Gas

Varies primarily by coal transport distance/cost and competing coal plant efficiency At current coal prices competition starts at ~$5.25 and increases through $3.25/mmbtu Coal generation bidding behavior not necessarily tied to spot pricingalso impacted by contract terms

CCGT Heat Rate Coal Generation Heat Rate Coal Price ($/ton) Coal Transport ($/ton) Breakeven Gas Price @ Del Point Breakeven @ Henry Hub

7.0 9.5 $70 $18 $5.03 $4.98

CCGT Heat Rate Coal Generation Heat Rate Coal Price ($/ton) Coal Transport ($/ton) Breakeven Gas Price @ Del Point Breakeven @ Henry Hub

7.0 9.5 $70 $12 $4.69 $4.29

27

Glossary: Power Terms Defined

CCGT: Combined cycle gas turbine power plant ST/GT: Steam/Gas turbine power plant. For purposes of this report, ST/GT only refer to gas plants but coal fired generation are also ST plants. IC: Internal Combustion power plant BIT: Bituminous, Eastern coal fired power plant SUB: Subituminous, PRB coal fired power plant LIG: Lignite coal fired power plant GW/MW: Measure of generation capacity. GW = 1,000 MW Supply stack: Power generation arranged in order of variable cost. Demonstrates which plants are more economic to run given a certain power demand level. Capacity Factor/CF: The measure of actual output of a power generation unit vs. its maximum possible output over a period of time. Bcf/d: Billion cubic feet per day MMcf/d: Million cubic feet per day Btu: British thermal unit. Measure of heat content/energy. GWh/MWH/KWh: Amount of power generated in an hour (Giga, Mega, Kilo) HH: Henry Hub

28

Base Case Supply Stack Pricing Assumptions

Pricing Assumptions: Henry Hub Natural Gas - $4.30 plus 50c transport charge. Coal (delivered prices): Bituminous coal - $80/ton. Average price, includes off-spec lower cost coals plus $12/ton transport costs. Subbituminous 1 coal - $35/ton. For facilities relatively close to the Powder River Basinincludes a $20/ton transport charge. Subbituminous 2 coal - $50/ton. For facilities further away from the Powder River Basinincludes a $35/ton transport charge. Lignite coal - $20/ton. Generally mine mouth plantsno transport charges included.

29

Analyst Certification: We, Brandon Blossman and George OLeary, do hereby certify that, to the best of our knowledge, the views and opinions in this research report accurately reflect our personal views about the company and its securities. We have not nor will not receive direct or indirect compensation in return for expressing specific recommendations or viewpoints in this report. Important Disclosure: The analysts above (or members of their household) do not own any securities mentioned in this report. Analysts compensation is not based on investment banking revenue and the analysts are not compensated by the subject companies. Tudor, Pickering, Holt & Co. has not received in the past 12 months compensation for investment banking or other services from the companies mentioned in this report. We intend to seek compensation for investment banking services from the companies we follow in the next 3 months. Ratings: B = buy, A = accumulate, H = hold, T = trim, S = sell, NR = not rated For detailed rating information, distribution of ratings, price charts and other important disclosures, please visit our website at www.tudorpickering.com. To request a written copy of the disclosures please call 800-507-2400 or write to Tudor, Pickering, Holt & Co. Securities, Inc. 1111 Bagby, Suite 5000, Houston, TX 77002.

OTHER DISCLOSURES Trade Name Tudor, Pickering, Holt & Co. is the global brand name for Tudor, Pickering, Holt & Co. Securities, Inc. (TPHCSI) and its non-US affiliates worldwide including Tudor, Pickering, Holt & Co. International, LLP. Legal Entities Disclosures U.S.: TPHCSI is a member of FINRA and SIPC. U.K.: Tudor, Pickering, Holt & Co. International, LLP is authorised and regulated by the Financial Services Authority. Registered in England & Wales No. OC349535. Registered Office Pellipar House, 1st Floor, 9 Cloak Lane, London EC4R 2RU. Canada The information contained herein is not, and under no circumstances is to be construed as, a prospectus, an advertisement, a public offering, an offer to sell securities described herein, or solicitation of an offer to buy securities described herein, in Canada or any province or territory thereof. Any offer or sale of the securities described herein in Canada will be made only under an exemption from the requirements to file a prospectus with the relevant Canadian securities regulators and only in the relevant province or territory of Canada in which such offer or sale is made. The information contained herein is under no circumstances to be construed as investment advice in any province or territory of Canada and is not tailored to the needs of the recipient. To the extent that the information contained herein references securities of an issuer incorporated, formed or created under the laws of Canada or a province or territory of Canada, any trades in such securities must be conducted through a dealer registered in Canada. No securities commission or similar regulatory authority has reviewed or in any way passed judgment upon these materials, the information contained herein or the merits of the securities described herein and any representation to the contrary is an offense.

30

United Kingdom Tudor, Pickering, Holt & Co International LLP does not provide accounting, tax or legal advice. In addition, we mutually agree that, subject to applicable law, you (and your employees, representatives and other agents) may disclose any aspects of any potential transaction or structure described herein that are necessary to support any UK income tax benefits, and all materials of any kind (including tax opinions and other tax analyses) related to those benefits, with no limitations imposed by Tudor, Pickering, Holt & Co International LLP or its affiliates. The information contained herein is confidential (except for information relating to tax issues) and may not be reproduced in whole or in part. Tudor, Pickering, Holt & Co International LLP assumes no responsibility for independent verification of third-party information and has relied on such information being complete and accurate in all material respects. To the extent such information includes estimates and forecasts of future financial performance (including estimates of potential cost savings and synergies) prepared by, reviewed or discussed with the managements of your company and/ or other potential transaction participants or obtained from public sources, we have assumed that such estimates and forecasts have been reasonably prepared on bases reflecting the best currently available estimates and judgments of such managements (or, with respect to estimates and forecasts obtained from public sources, represent reasonable estimates). These materials were designed for use by specific persons familiar with the business and the affairs of your company and Tudor, Pickering, Holt & Co International LLP materials. This information is intended only for the use of professional clients and eligible counterparties or persons who would fall into these categories if they were clients of Tudor, Pickering, Holt & Co International, LLP, or any of its affiliates. Retail clients must not rely on this document and should note that the services of Tudor, Pickering, Holt & Co International, LLP, are not available to them. Under no circumstances is this presentation to be used or considered as an offer to sell or a solicitation of any offer to buy, any security. Prior to making any trade, you should discuss with your professional tax, accounting, or regulatory advisers how such particular trade(s) affect you. This brief statement does not disclose all of the risks and other significant aspects of entering into any particular transaction. Tudor, Pickering, Holt & Co. International, LLP is a limited liability partnership registered in England and Wales (registered number OC349535). Its registered office is Pellipar House, 1st Floor, 9 Cloak Lane, London EC4R 2RU. Tudor, Pickering, Holt & Co. International, LLP (TPH International) is authorised and regulated by the Financial Services Authority, and is a separate but affiliated entity of Tudor, Pickering, Holt & Co. Securities, Inc. (TPH Securities). TPH Securities is a member of FINRA and SIPC. Unless governing law permits otherwise, you must contact the Tudor, Pickering, Holt & Co. entity in your home jurisdiction if you want to use our services in effecting a transaction. See http://www.tudorpickering.com/Disclosure/ for further information on regulatory disclosures including disclosures relating to potential conflicts of interest.

Copyright 2011, Tudor, Pickering, Holt & Co. This information is confidential and is intended only for the individual named. This information may not be disclosed, copied or disseminated, in whole or in part, without the prior written permission of Tudor, Pickering, Holt & Co. This communication is based on information which Tudor, Pickering, Holt & Co. believes is reliable. However, Tudor, Pickering, Holt & Co. does not represent or warrant its accuracy. The viewpoints and opinions expressed in this communication represent the views of TPH as of the date of this report. These viewpoints and opinions may be subject to change without notice and TPH will not be responsible for any consequences associated with reliance on any statement or opinion contained in this communication. The viewpoints and opinions herein do not take into consideration individual client circumstances, objectives, or needs and are not intended as recommendations of particular securities, financial instruments or strategies to particular clients. Past performance is not indicative of future results. This message should not be considered as an offer or solicitation to buy or sell any securities.

31

RESEARCH
Oil Service Jeff Tillery 713.333.2964 jtillery@tudorpickering.com Joe Hill 713.333.2963 jhill@tudorpickering.com John Lawrence 713-333-7682 jlawrence@tudorpickering.com Rhett Carter 713-333-2983 rcarter@tudorpickering.com Macro Dave Pursell 713.333.2962 dpursell@tudorpickering.com Coal & Power Brandon Blossman 713.333.2994 bblossman@tudorpickering.com George OLeary 713.333.2973 goleary@tudorpickering.com Integrateds/ Downstream Robert Kessler 713.333.7696 rkessler@tudorpickering.com Brandon Mei 713.333.7689 bmei@tudorpickering.com Midstream Brad Olsen 713.333.7693 bolsen@tudorpickering.com E&P David Heikkinen 713.333.2975 dheikkinen@tudorpickering.com Brian Lively 713.333.2970 blively@tudorpickering.com Brad Pattarozzi 713-333-2993 bpattarozzi@tudorpickering.com Jessica Chipman 713.333.2992 jchipman@tudorpickering.com Matt Portillo 713-333-2995 mportillo@tudorpickering.com Oliver Doolin 713-333-2989 odoolin@tudorpickering.com Hubert van der Heijden 713-333-3983 hvanderheijden@tudorpickering.com *London- E&P Anish Kapadia +44 20 3008 6433 akapadia@tudorpickering.com

SALES
Houston Clay Coneley 713-333-2979 cconeley@tudorpickering.com Tom Ward 713.333.7182 tward@tudorpickering.com Mike Bradley 713.333.2968 mbradley@tudorpickering.com Mike Davis 713.333.2971 mdavis@tudorpickering.com Josh Martin 713.333.2982 jmartin@tudorpickering.com Paige Penchas 713.333.2969 ppenchas@tudorpickering.com Denver Chuck Howell 303.300.1902 chowell@tudorpickering.com Jason Foxen 303.300.1960 jfoxen@tudorpickering.com New York Ken Johnson 212-610-1650 kjohnson@tudorpickering.com *London Jon Mellberg +44 20 3008 6430 jmellberg@tudorpickering.com Win Oberlin +44 20 3008 6431 woberlin@tudorpickering.com

TRADING 800.507.2400
Michael du Vigneaud mduvigneaud@tudorpickering.com Scott McGarvey smcgarvey@tudorpickering.com Seth Williams swilliams@tudorpickering.com

DEBT SALES & TRADING

Clay Border 713.333.2974 cborder@ tudorpickering.com

*Office of Tudor, Pickering, Holt & Co. International, LLP. Anish Kapadia is employed by Tudor, Pickering, Holt & Co. International, LLP in the United Kingdom and is not registered/qualified as a research analyst with FINRA. Mr. Kapadia is not an associated person of Tudor, Pickering, Holt & Co. Securities, Inc. and as such is not subject to NASD Rule 2711 restrictions on communications with subject companies, public appearances and trading securities held by a research analyst account.

32