Bradley Olsen
[email protected] (713) 333-7693
Differentials and Differentiation TPH Midstream Sector Initiation uy , **IMPORTANT DISCLOSURES BEGIN ON PAGE 71 OF THIS REPORT**
What’s in This Report
State of the Midstream Industry
Performance Retrospective
Crude/NGL Growth Forecast
Pipeline Opportunities Gas Pipelines Crude Pipelines out of Cushing, OK NGL Pipelines out of Conway, KS
Storage Opportunities Natural Gas Storage Crude and Products Storage
NGL Value Chain Opportunities
Implications of North America as low-cost supplier Eagle Eagle Ford, Marcellus, Marcellus, Bakken Bakken NGL Discussion Discussion Downstream implications of massive NGL growth Many American NGLs won’t be consumed by Americans 2
Key Midstream Takeaways
We have seen a wave of unconventional “dry gas” activity; the drill bit has now turned to crude and wet gas.
Gas infrastructure received multibillion dollar investments throughout the 2000s; oil and – ’ .
What’s old is new again: the US is generally short liquids infrastructure in areas with existing assets; this increases barriers to entry, and gives existing operators a natural edge in bidding for future projects at attractive IRRs. Examples: □ Marc Marcel ellu luss take takeaw awa a i elin elines es – exis existi tin n infr infras astr truc uctu ture re is ex andi andin n to take take adva advant nta a e of new Marcellus volumes; expansion economics make these projects much more attractive than the the pipe newbuilds of the late 2000s. □ Cushing, OK – expect returns on contracted contracted storage and (non-Keystone) (non-Keystone) pipelines pipelines to the Gulf to be very attractive over extended (5-10 year) terms. , – , are high. NGL fractionation projects are currently being subscribed with 7-10 year takeor-pay contracts with 15%+ unlevered economics. □ Eagle Ford gathering gathering and processing – kudos to midstream midstream operators for seeing this thing thing coming years ago… well, maybe they didn’t, but there’s a lot of empty field-level infr infras astr truc uctu ture re that that is bein bein ram ram ed u for for attr attrac acti tive ve ofte often n feefee-ba base sed d IRRs IRRs.. Dust off that export export dock – you don’t don’t have to wait for LNG export terminals terminals – we can □ export gas BTUs in NGL form. America is seeing a burgeoning energy export renaissance in NGLs, and we are displacing crude and natural gas imports.
3
Suddenly, buried steel is a lot more interesting…
4
Quite Simply: It’s About Differentials
DIFFERENTIAL - in the energy industry, this word refers to a price difference between fungible units of energy.
’ When a barrel of oil is produced in western Canada or a cubic foot of gas is produced in the Permian basin, it is not worth very much. Between the wellhead and the final consumer lies the midstream sector. – sites. Midstream services include the pipeline to bring hydrocarbons to market. □
Seasonal differentials – you might produce oil when there is no way to take it off the lease, or deliver gas to a city on a 70 summer day. Midstream provides , production and consumption.
□
Quality differentials – you wouldn’t fill your car with crude; similarly, natural gas straight from the well is often dangerous to consume in a home or business. .
The midstream industry exists in order to reduce differentials, but it also profits from them. As a result of this fact, the midstream sector is dynamic, even when production, prices, supply, and demand are stagnant or even declining (as long as t ey on’t a stagnate toget er). 5
This Ain’t Your Parents’ Midstream Industry
The midstream industry is in the middle of an unprecedented transformation, a result of the revolution in North American unconventional resource development .
□
□
□
□
As a result of declining oil and stagnant gas domestic production in the 1970s-2000s, midstream assets were often underutilized, and not seen as value-drivers. Oil and gas production from unconventional resource development is increasing rap y an n ras ruc ure access s no onger a g ven or pro ucers Midstream has been largely occupied by small-caps with limited institutional ownership, mainly due to ~50% of sector market cap being partnerships (MLPs) This has begun to change as investor demand has reduced public valuation penalty for midstream assets housed in a C-corp structure Lower correlations with commodity prices than E&P and OFS Midstream investments provide leverage to trends that are arguably more durable , migration to unconventional basins, increased US import/export of oil/gas/NGLs) Assets historically found inside larger energy companies Asset rationalization by large-cap energy companies has provided a steady stream of m s ream oppor un es over e pas + years
6
The Midstream Value Chain
Generally riskier approaching the wellhead
1. 2.
Generally less risky approaching point of consumption
Proximity to supply increases risk – generally speaking, the closer to the wellhead, the higher the dependence on a single basin, single field or even a single well. Demand centers, on the other hand, are relatively stable and immobile over time. Field assets are more likely to have commodity exposure – historically controlled by producers, atherin and rocessin was once effectivel an extension of the E&P business. Crude gathering services are exposed to the front end of the futures curve, while gas processing is often exposed to gas and NGL pricing. 7
Midstream in a Portfolio
Correlation of midstream performance to commodities is substantially lower than for E&P/OFS
last decade as a result of two factors: □
□
disproportionate share of commodity-levered (gathering and processing, E&P) midstream IPOs in the late 2000s ncrease part c pat on y macro- r ven nst tut ona energy nvestors
Correlation of Daily Returns to Oil
Correlations of Daily Returns to Gas
70.0%
50.0%
60.0%
40.0%
50.0%
30.0%
40.0%
20.0%
30.0%
10.0%
20.0%
.
10.0%
(10.0%)
0.0% E&P
OFS
(20.0%)
MLP
E&P
OFS
MLP
2000
2001
2002
2003
2004
2005
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2006
2007
2008
2009
2010
2011
Source: Bloomberg
8
Midstream Businesses are NOT “Toll Roads”
Contrary to the misleading impression conveyed by prominent TV talking heads and some research analysts, midstream businesses and MLPs in general cannot accurately be described as “toll roads,” (i.e., volumetric businesses with minimal commodity exposure)
Almost every publicly-traded midstream company has margins that are impacted, directly or indirectly, by commodity prices. For example: □
Gathering/processing companies often go long the hydrocarbon at the wellhead
□
Pipelines often sell gas, allocated to them as fuel to operate their pipes
□
□
□
□
Refined product storage operators can benefit from blending cheaper NGL feedstocks, in effect capturing the spread between blendstocks and gasoline Gas storage operators charge a fee per mcf per month. Historically, higher gas prices have meant higher seasonal spreads in the gas futures market ranspor a on ac v y s corre a e o gas eman . ven a e-or-pay p pe nes can earn 10%-15% of total revenues by utilizing the pipe’s full capacity NGL fractionators are indirectly levered to liquids prices, as liquids prices determine the amount of drillin ri s tar etin dr vs. NGL-rich as
9
…But Midstream Businesses are Anchored by Stable Trends Refined Products Transportation and Storage MLPs vs. Consumption and Product Prices
Although we find the idea of “midstream = toll road” to be overly simplistic, we do not want to understate the stability of many midstream businesses.
Long-haul transportation
300 250 200 150 100 50
centers are the most stable midstream businesses, due to proximity to consumers
0 2006
2007
2008
Refined Products Performance
a ura
2009
RBOB/Diesel 2:1 Price Composite
2010
2011
Refined Products Consumption
as ranspor a on s vs. onsump on an Natural Gas Prices
250
200
.
At left, the equalweighted performance of long-haul natural gas and
150 100 50 0 2006
2007
2008
Gas Transportation Performance
2009 NG Price
2010
2011
illustrates that pipelines have been more stable than the commodities they transport
NG Consumption
Source: Bloomberg & EIA Notes: Refined products performance is equal-weight index of BPL, HEP, MMP, NS, SXL, TLLP, TLP; Natural gas performance is equal-weight index of BWP, EPB, SEP, TCLP
10
It’s Been a Heady 3 Years…
13
Midstream Performance Overview
Consistent with the broader market, we have seen a strong recovery in the midstream sector since the end of 2008
As midstream valuations recovered from the financial crisis, investors spent late 2009 and 2010 moving further out the risk curve, favoring companies with long commodity exposures and cyclical business models. , benefitted as NGL and crude prices have performed strongly vs. dry gas.
As the cycle has continued, we have seen consolidations and restructurings □
□
The majority of all publicly-traded MLP general partners have been consolidated into their respective MLPs at healthy premia. STR WMB and EP’s restructurin s have driven erformance.
Going forward, we prefer exposure to liquids volume growth, rather than prices (fee-based processing/fractionation). We believe the rapid increase in NGL volumes will cause choppiness in NGL prices, but we believe the volume growth trend is durable. 14
2009-Present: Taking Stock of a Wild 30 Months
Performance since 2008 has been broadly outstanding, with every segment generating >100% total returns after 18 dismal months.
400 ) % ( n r u t e R l a t o T
300 200 100 0 Gathering and rocess ng
Other
MLP GPs
C-Corp s
Refined Product s
Shippers
Midtream - orps
Regulated Gas pe nes
Diversified MLPs
Propane and ea ng
Alerian Total Return
) % ( n r u t e R l a t o
GPs have notably outperformed. Investors will need to find alpha elsewhere as the investible universe of GPs has been limited by buyouts/consolidations (buyouts marked with *).
700 600 500 400 300 200 100 0 -100 -200
* X * * C L O P M S I T * C P P P R P P E P U P B Y * D P * H L P B S L A B * E P P U S P P K P H * P P H * * E P L S S Z M O P G E P P S P M H N E N L M G O G L E X K X A N D S G W E E V E L K L E P G Z P L S E T R P G L T N P S P P P E E P N M Q P L L M S O L U T W T M G T P T V R T H K P E X O G G A G T W E G E C O N B M G O S P E N F B A K S T S G P E T G M M S S E R M C D P W G T X L M W L E P U R X E B R C C N M C B T N A C P E M N W H H N
Alerian Total Return
s
- orp
s
Gathering & Processing
Shippers
Diversified MLPs
Other
Refined Product MLPs
Propane & Heating Oil
s ream - orps Regulated Gas Pipelines
15
2009-Present (Continued)
Although much of 2009 performance was “mean reversion,” subsequent outperformance has been driven by secular growth from leverage to NGL economics (red), dropdowns (yellow), takeouts and corporate actions (green), recoveries from distress (sky blue)
250 ) % ( n r u t e R l a t T
250 ) % ( n r u t e R l a t T
200 150 100 50 0
200 150 100 50 0
EEP
EPD
OKS
PAA
KMP
ETP
TLP
Gathering & Processing
MMP
SXL
NS
200 ) % ( n r u t e R l a t o T
600 400 200 0
150 100 50 0
MWE
APL NGLS WPZ DPM XTEX RGNC CPNO CMLP WES EROC PVR HLND*
Propane & Heating Oil ) % ( n r u e R l a t o T
BPL
Regulated Gas Pipelines
800 ) % ( n r u t e R l a t o T
HEP
EPB
WMZ*
TCLP
BWP
SEP
Group Average
200 150
50 0 SGU
NRGY
FGP
APU
SPH
16
2009-Present (Continued)
Although much of 2009 performance was “mean reversion,” subsequent outperformance has been driven by by secular growth from leverage to NGL economics (red), dropdowns (yellow), takeouts and corporate actions (green), recoveries from distress (sky blue) Shi
) % ( n r u t e R l a t o T
Other ) % ( n r u t e R l a t o T
NMM
TOO
TGP
OSP*
CPLP
CQP
GEL
) % ( n r t e R l a t o T
200
100 50
MMLP
CLMT
BKEP
EXLP
GLP
TNH
C-Corp GPs
250
0
200
100 50 0
SUG
EP
WMB
ENB
SE
TRP
STR
MLP GPs
XTXI
OKE
TK
Group Average
600 ) % ( n r u t e R l a o T
700 600 500 400 300 200 100 0
USSPQ*
Midstream C-Corps
250 ) % ( n r t e R l a t o T
ers
300 250 200 150 100 50 0 -50 -100
500 400
Alerian Total Return
300 200 100 0 ATLS
NRGP*
EPE*
BGH*
ETE
PVG*
NSH
MGG*
HPGP*
17
2010-Present: “Risk On” Trade has Driven Performance
) % ( n r u t e R l a t o T
Whether financial leverage (GPs) or commodity leverage (gathering & processing), the last 18 months have been good for risk. While 2009 involved a mean reversion play, emboldened investors were rewarded for moving out the risk curve in 2010.
100 80 60 40 20 0 C-CorpGPs/
Gathering and
Midstream -
Shippers
Refined Product
Diversified MLPs
Other
Regulated Gas
Propane and
Alerian Total Return
Interestingly, the 5 top performers over this period were dividend/distribution cutters during 2008-2009
250 ) % ( n r u t e R l a t o T
200 150 100 50 0 -50
S Z E G E O O S P L E P M P S B D P B H B K E M H P U P P L R C P P A T P L P S U P P H P P P P Y P P L S C X P I E P W P L O E E X L X V N L R A M M P L N P E L P G T L E G W E Q E T G M K M P E N S P T S M N L G L U K O N L E S E P S B K A T R T T G C G E T P E H E N E S T C S P G M T P L K B X D M O W W N T M M S O T P A E X X W E A G F C R T C C N B C N R M
Midstream C-Corps
Refined Product MLPs
C-Corp GPs/MLP GPs
Other
Diversified MLPs
Gathering & Processing
Shippers
Regulated Gas Pipelines
Alerian Total Return
Propane & Heating Oil
18
2010-Present (Continued)
Commodity-levered names (red), dropdown MLPs (yellow), post-distress (sky blue) and corporate action stories (green) have generally outperformed Diversified
Refined Product 60 50
50 ) % ( n r u t e R l t o T
) % ( n r u t e R l a t T
40 30 20
40 30
10
10
0
0 OKS
EPD
PAA
KMP
EEP
MMP
ETP
Gathering & Processing 250
t e R l a t o T
) % ( n r
200
t e R l a t o T
150 100 50 0
Group Average
TLP
SXL
BPL
NS
Regulated Gas Pipelines
300
) % ( n r
HEP
Alerian Total Return
50 45 40 35 30 25 20 15 10 5 0
19
2010-Present (Continued)
Commodity-levered names (red), dropdown MLPs (yellow), post-distress (sky blue) and corporate actions (green) have generally outperformed Propane & Heating Oil
) % ( n r u t e R l a t o T
50
) % ( n r u t e R l a t o T
40 30 20 10 0 SGU
APU
SPH
FGP
Shippers
80 60 40 20 0
NRGY
TOO
TGP
Other ) % ( n r u t e R l a t o T
CPLP
Midstream C-Corps
70 60 50 40 30 20 10 0 -10 -20
120 ) % ( n r u t e R l a t o T
100 80 60 40 20 0
CQP
GEL
TNH
MMLP
CLMT
EXLP
GLP
BKEP
C-Corp GPs/MLP GPs
250 ) % ( n r u t e R l a t o T
NMM
EP
SUG
WMB
ENB
SE
STR
TRP
Group Average
200
Alerian Total Return
150
50 0 ATLS
XTXI
OKE
ETE
NSH
TK
20
2011 YTD… Reaching Cruising Altitude
) % ( n r u t e R l a t o T
Total returns shown below for 2011 YTD performance. GPs and C-corps have outperformed, the former due to performance by liquids-levered GPs, the latter due to corporate actions (WMB, EP)
Gas storage, propane and shippers have been held back by weak fundamentals and weak results
30 20 10 0 -10 Midstream C-Corps
C-Corp GPs/ MLP GPs
Gathering and Diversified MLPs Refined Product Processing MLPs
Other
Shippers
Regulated Gas Pipelines
Propane and Heating Oil
Gas Storage
Alerian Total Return
) % ( n r u t e R l a t o T
80 70 60 50 40 30 20 10 -10 -20
S E L I P B E E E M S O P P S L B D O P L A T U P H M P P P L P P P M P C P R P H K S P U P Y Y P G P G A P H X C B P Z S G P S P K E M L E P P N M X A M G L S K G L E P E E L M L N L V T P T N L P W E K P X P E R N T W G G L N G K Q U E L A T N T O M G T E E M D O O G E E P K S P L S M N H T M K B E S T N X G P P E S G A B R R C P F M S T R W W E N C T H M G A O X T X R E R C N E W T C C C M B S C N N T
Alerian Total Returns
Gathering & Processing
Refined Product MLP’s
Propane & H eating Oil
GP Holding C-Corps/MLP GP’s
Other
Shippers
Gas Storage
Midstream C-Corps
Diversified MLP’s
Regulated Gas Pipelines
21
2011: Life as a Fairly Valued Sector
After perhaps once-per-lifetime returns in 2009-2010, 2011 has seen continued outperformance from liquidslevered names (red), post-distress stories (sky blue), corporate actions (green), and dropdowns (yellow). Diversified
Refined Product
25 ) % ( n r u t e R l a t o T
) % ( n r u t e R l a t o T
15 10 5 0 WPZ
OKS
EPD
KMP
PAA
EEP
12 10 8 6 4 2 0 -2 -4 -6 HEP
ETP
Gathering & Processing ) % ( n r u t e R l a t o T
) % ( n r u t e R l a t o T
XTEX EROC WES
MWE
SXL
BPL
TLP
NS
Regulated Gas Pipelines
40 35 30 25 20 15 10 5 0 -5 APL
MMP
DPM NGLS CPNO CMLP CHKM RGNC PVR
8 6 4 2 0 -2 -4 -6 -8
Gas Storage ) % ( n r u t e l a t o T
8 6 4 2 0 -2 -6 -8 -10 -12
Group Average Alerian Total Return NRGY
PNG
NKA
22
2011: Life as a Fairly Valued Sector-Continued
After perhaps once-per-lifetime returns in 2009-2010, 2011 has seen continued outperformance from liquidslevered names (red), post-distress stories (sky blue), corporate actions (green), and dropdowns (yellow). Propane & Heating Oil
) % ( n r u t e R l a t o T
) % ( n r u t e R l a t o T
6 4 2 0 -2 -4 -6 -8 SGU
SPH
APU
NRGY
TOO
TGP
) % ( n r u t e R l a t o T
30 20 0 -10
NMM
CPLP
Midstream C-Corps
80
-20
60 40 20 0 -20
TNH
GEL
CLMT
MMLP
BKEP
EXLP
GLP
CQP
SUG
EP
TRP
ENB
SE
STR
SEMG
GP Holding C-Corps/MLP GP'S
50 ) % ( n r u t e R l a t o T
10 8 6 4 2 0 -2 -4
FGP
Other
40 ) % ( n r u t e R l a t o T
Shippers
40
Group Average
30 20
Alerian Total Return
0 -10 ATLS
OKE
XTXI
WMB
TRGP
ETE
NSH
TK
23
Midstream Valuations Imply Growth
MLPs are not compellingly cheap, compared to historical yield spreads
MLPs are Reasonably Cheap vs. 10-Year Treasury
MLP are Fairly Valued vs. 10-Year Baa Industrial 4.00%
8.00% 7.00% 6.00% 5.00% 4.00% 3.00% 2.00% . 0.00%
. 2.00%
Cheap
Cheap
1.00% 0.00% -1.00% - .
Expensive 5 9 9 1
6 9 9 1
7 9 9 1
8 9 9 1
9 9 9 1
0 0 0 2
AMZ-10 Yr Spread
1 0 0 2
2 0 0 2
4 0 0 2
Avg Spread
s are ompe ng y
6.00% 5.00% 4.00% 3.00% 2.00% 1.00% 0.00% -1.00% -2.00%
3 0 0 2
5 0 0 2
6 0 0 2
7 0 0 2
8 0 0 2
+ 1 Std Dev
eap vs.
9 0 0 2
5 9 9 1
0 1 0 2
-1 Std Dev
6 9 9 1
7 9 9 1
8 9 9 1
9 9 9 1
AMZ-Baa Spread
are
s
0 0 0 2
1 0 0 2
2 0 0 2
3 0 0 2
Avg Spread
4 0 0 2
5 0 0 2
6 0 0 2
7 0 0 2
8 0 0 2
+ 1 Std Dev
o era e y xpens ve vs.
9 0 0 2
0 1 0 2
-1 Std Dev
- r
un s
8.00% 6.00%
Cheap
Cheap
. 2.00%
Expensive
Expensive 5 9 9 1
6 9 9 1
7 9 9 1
8 9 9 1
AMZ-REIT Spread
9 9 9 1
0 0 0 2
1 0 0 2
2 0 0 2
Avg Spread
3 0 0 2
4 0 0 2
5 0 0 2
6 0 0 2
+ 1 Std Dev
7 0 0 2
8 0 0 2
9 0 0 2
0 1 0 2
-1 Std Dev
0.00% 5 6 9 9 9 9 1 1
7 9 9 1
8 9 9 1
AMZ-Muni Spread
9 9 9 1
0 0 0 2
1 0 0 2
2 3 0 0 0 0 2 2
Avg Spread
4 5 0 0 0 0 2 2
6 0 0 2
+ 1 Std Dev
7 0 0 2
8 0 0 2
9 0 0 1 0 0 2 2
-1 Std Dev
Source: Bloomberg
24
Midstream C-Corp Valuations
Midstream C-Corps are cheaper than MLPs, compared to historical yield spreads
C-Corps are Reasonably Cheap vs. 10-Year Treasury C-Corps are Cheap vs. 10-Year Baa Industrial 1.00%
4.00% . 2.00% 1.00% 0.00% -1.00% -2.00% -3.00% -4.00%
0.00% -1.00%
Cheap
Cheap
-2.00% -3.00% -4.00% -5.00%
Expensive 9 9 9 1
0 0 0 2
1 0 0 2
xpens ve
-6.00% -7.00% 2 0 0 2
C-Corp-10 Yr Spread
3 0 0 2
4 0 0 2
5 0 0 2
Avg Spread
- orps are ompe ng y
6 0 0 2
7 0 0 2
8 0 0 2
9 0 0 2
+ 1 Std Dev
eap vs.
0 1 0 2
9 9 9 1
1 0 0 2
2 0 0 2
C-Corp-Baa Spread
-1 Std Dev
s
0 0 0 2
3 0 0 2
4 0 0 2
5 0 0 2
Avg Spread
6 0 0 2
7 0 0 2
8 0 0 2
9 0 0 2
+ 1 Std Dev
- orps are a r y a ue vs
- r
0 1 0 2
-1 Std Dev
un s
4.00%
1.00%
3.00%
0.00% Cheap
-1.00%
2.00%
-2.00%
1.00%
-3.00%
0.00%
-4.00%
Cheap
-1.00%
-5.00%
Expensive
-2.00%
-6.00% 9 9 9 1
0 0 0 2
1 0 0 2
C-Corp-REIT Spread
Source: Bloomberg
2 0 0 2
3 0 0 2
4 0 0 2
Avg Spread
5 0 0 2
6 0 0 2
7 0 0 2
+ 1 Std Dev
8 0 0 2
9 0 0 2
0 1 0 2
-1 Std Dev
Expensive
-3.00% 9 9 1
0 0 2
1 0 0 2
C-Corp-Muni Spread
0 0 2
0 0 2
0 0 2
Avg Spread
5 0 0 2
0 0 2
0 0 2
+ 1 Std Dev
0 0 2
0 0 2
1 0 2
-1 Std Dev
25
Crude and NGL Volumes are Driving Midstream Opportunity
26
Midstream Growth Will Come from Liquids Supply
Shift from dry gas BTUs to wet gas BTUs has been underway for the past 18 months US Rig Count by Target Target – NGL Targets Targets Included in Gas Rig Count, Understating Impact of Shift , 1,000 500 0 Oct-06
Apr-07
Oct-07
Apr-08
Oct-08
Apr-09 Gas
Oct-09
Apr-10
Oct-10
Apr-11
Oil
Source: BHI
, current levels. This includes production declines in Gulf of Mexico and mature onshore basins, requiring requiring construction of nearly 1 mmbpd of pipe capacity
This tectonic shift in the composition com position of US hydrocarbon production presents midstream opportunities throughout the value chain, from upstream (gathering, trucking, processing, storage terminals) to downstream (pipelines, fractionation, storage) □ □
Crude oil/condensate growth will come from Canada, Bakken, Eagle Ford, Permian growt w
come rom ag a g e or or , ran te
as , erm an
27
Liquids Growth is Going to be BIG Forecast Liquids Growth from Eagle Ford and Granite Wash 2,000
y a D / s 1,500 l e r r a 1,000 B d n 500 a s u o 0 h T
2011E
2012E
2013E
2014E
2015E
2016E
2017E
Source: TPH Estimates
Although emerging horizontal plays within the basin make estimates more difficult, Permian Basin NGL production production could add over 200 kbpd by 2015, by our estimates estimates
Midstream service providers are positioned to participate in that growth from the wellhead to the . (relatively brief) history as a substantial standalone sector.
This wave of growth will not fuel a 2007/08-style boom, when creeping commodity leverage drove rapid growth followed by a cash flow collapse for some midstream companies: , exposure) exposure) – much less volatile volatile but slightly lower returns returns through an economic economic cycle □
□
Increasingly Increasingly contracted contracted – term contracts for processing/fraction processing/fractionation ation assets, more life-of-lease life-of-lease acreage dedications, minimum volume gathering commitments 7/1/11 NYMEX gas price of $4.32/BTU is well below NYMEX NGL price of over $15.00/BTU, providing huge economic incentive for E&Ps to develop wet gas and crude
28
US Liquids Infrastructure is Especially Outdated
We believe the midstream industry has a historic opportunity as US unconventional production becomes wetter and reinvigorates long-stagnant US crude and NGLs production.
The Cushing price dislocation has illustrated that US liquids infrastructure is based on simple assumptions that are becoming antiquated by unconventional production 70
) d / f c b ( n i t c u d o r P s a G
5,000 4,500
Bakken/Permian/Eagle Ford growth is here
60
4,000
50
3,500 , 2,500
30
2,000
20
1,500 1,000
10 0
C r u d e / N G L P r o d u c t i o n ( k b p d )
0 1999
2000
2001
2002
2003
Gas Production (ex. Federal GOM)
2004
2005
2006
2007
2008
Crude Production (Ex. Federal GOM)
2009
2010
2011
NGL Production
Source: EIA
29
Remediating Differentials Between Hubs
30
US is Long Gas Pipe and Short Oil/NGL Pipe
Natural gas – differentials have been drastically reduced over the past several years by three factors: □ >$60 bn of capital spending since 1996 on large-diameter pipelines □ Unconventional gas supply (Marcellus gas could reduce imports to NE by up to 6 bcf/d)
Crude oil – opportunities from Cushing/WTI dislocation: □ 1970s-2000s: US increasingly short crude, current imports 6mmbpd higher than 1970’s. Pipelines from Canada/ports typically terminate at Cushing, OK (NYMEX hub, largest in US) □ 2000s: pipes prepare for coming wave from W. Canada, but they don’t prepare or: Bakken (~400 kbpd, operators indicate could reach 1.2 mmbpd by 2015) Eagle Ford crude (~100 kbpd, expected to reach nearly 1 mmbpd by 2015) US government delaying “oil sands” pipe (TRP’s XL) □ Cushing now has ~500 kbpd more inbound than outbound capacity. EPD/ETP JV has proposed a “Double E” pipeline to take 450 kbpd to Houston □ Rail as swing capacity: shifting supply basins, Cushing bottleneck have led Kinder Morgan and WATCO, a private rail company, to announce rail projects out of Cushing 31
2000s: The First Wave of Midstream Spending
After a few sleepy decades, the midstream sector saw sharp increases in spending in the mid-2000s
The infrastructure boom in the 2000s was a distinct development from projects being developed today: □
□
2008: Peak of the First Wave of the Midstream Boom Projects Debottlenecking Southeast, Importing LNG
High regional gas price differentials indicated that US was short long-haul transportation Primary focus: getting emerging gas supplies – LNG, Barnett, Canada, – Northeast, Midwest, Florida via underutilized pipes (Transco, TETCO, Tennessee)
boom is over, but expansion projects on existing gas pipelines are expected to provide an average of $5bn per year of s endin throu h 2020
2000s Saw Revitalization of Midstream Investment $12
$4 Gas pipeline data shown
$10
$3
$8 $6
$2
4
$1
$2 $0
$0
, 2010-20 Est Capex/Yr ($bn, left axis) Cost per Mile ($mm, right axis) Source: EIA, El Paso
32
Gas Pipeline Opportunity Set Will Be Smaller
After the “first wave” of conventional gas-focused infrastructure investment in the 2000s, much of North America is long gas pipeline capacity eren a s ave a en across mos ma or u s, con ro ng or enry u pr ce decline.
Returns generally protected by take-or-pay contracts, low utilization can still marg na y re uce revenue rom var a e c arges
Tens of billions of dollars of newbuild pipe capacity now, on average, “out of the money” (i.e., netbacks are worse when shipped vs. sold locally). □
Rockies Express (Opal Hub, WY)
□
Barnett area buildout (W.TX)
□
Midcontinent Ex ress E.OK
□
Florida Gas Transmission VIII (FL)
□
Ruby (Opal Hub, WY)
b u H y r n e H f o % s a l a i t n e r e f f i D
30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% (5.0%)
Average, 2005-2009
Average, 2010-Current
33
Northeast Gas Differentials Persist, Presenting Opportunity
New York City Gate differential has persisted at an average >$0.80, increasing as % of HH price from 12% to 19% □
□
□
O ortunities to connect nearb Marcellus su l to eastern cit ates: >$4bn of announced Marcellus opportunities for TGP (EP, $1.2 bn), Transco (WPZ/WMB, $0.6bn), TETCO (SE, $1.75bn) Expansions are typically offer higher returns than newbuilds, as expansions can leverage existing property rights-of-way and compression facilities An effective “FERC put” can offer return upside on expansions while providing downside protection through rate cases
Operators’ strategy for profiting from the NY differential is in its first phase: □
□
Phase 1 (current-2012): plumb Marcellus supplies (TGP’s 300 Line, TETCO’s TEAM, Transco’s NE Supply Link), move oversupplied Midwest market eas war o s pro ec Phase 2: move Marcellus gas to markets (TETCO’s NJ-NY Expansion, TGP’s NE Upgrade, Transco’s NE Connector and Rockaway Lateral) and displace ’ , ’ Algonquin backhaul) 34
“Phase 1” Marcellus Projects Access Supply… TEAM/TEMAX in same corridor
TGP Projects Underway
NJ-NY
TETCO Projects Underway
Operator EP EP EP SE SE SE WPZ
Project 300 Line Expansion NE Supply Diversification NE Upgrade TEMAX/TIME III TETCO Appalachia to Mkt NJ-NY Expansion NE Supply Link
In Service 4Q 2011 4Q 2012 4Q 2013 4Q 2012 4Q 2012 4Q 2013 2013
WPZ
Rockaway Delivery Lateral
1H 2014 Total:
Cost ($mm) $660 $70 $400 $700 $200 $850 $341 $182 $3,442
Capacity (mm cf/d) Supply/Dem and Driven? 350 Supply 250 Supply 636 Demand 455 Supply 190 Demand 800 Demand 250 Supply 647 3,678
Demand
Source: public filings, company presentations
35
…“Phase 2” Marcellus Projects Will Displace Imports
Proposed but not-yet filed with FERC are projects to move gas up Algonquin and Iroquois and displace Canadian gas from New England markets
El Paso’s proposed Iroquois interconnects could displace 500 mmcf/d by Q4 2014; small backhaul volumes possible w/out compression
NJ-NY
Spectra’s TETCO has proposed a project to move Marcellus gas up Algonquin Gas to New England markets
Source: public filings, company presentations
36
Coal-to-Gas Switching Should Drive Southeast Growth
Coal-to-gas switching (as detailed in TPH’s recent report) has driven large gains in Southeastern gas demand since 2007. This trend is set to continue: □
□
8.5 GW of announced coal retirements throu h 2020 e ual to ~1.1 bcf/d of demand) Efficient CCGT power gen running at a 42% capacity factor, outlook for gas demand remains positive
Southeast demographics remain strong:
Population Growth, Census 2010 vs. 2000
Source: US Census Bureau
20% 10% 5% 0% LA
MS
AL
TN
SC
FL
GA
NC
Southeast
US
Despite current softness in Florida gas differentials, Florida’s long-term gas market growth should be strong, driven by retirements □
Softness due to the recent FGT Phase VIII expansion, a project to feed power demand which was unable to contract final 26% of 820 mmcf/d of capacity 37
Escape from Cushing and the Mid-Continent
38
Well, This is a Conundrum...
Volumes chase volumes: as Cushing has grown larger (18.3 mmbbls of shell capacity to be added between Q1 2011 and 1H 2012, bringing total to 76 mmbbls) and more liquid (see NYMEX volumes below), pipeline operators have treated Cushing as a deep demand market, despite constraints: □ limited local refinery capacity - 1.1 mmbpd of refining capacity is proximate to Cushing, less than one ay o us ng s n oun p pe ne capac y □ lack of access to Gulf Coast – largest Houston-Cushing pipeline, Seaway Crude, sits empty. It is owned by COP (50% owner, along with EPD), whose refineries enjoy the benefit of cheap crude feedstocks at its Midcon refineries (Ponca City, OK; Borger, TX; Wood River, IL)
,
Cushing Seaway Ke stone Keystone XL Ponca City – 170 kbpd Borger – 146 kbpd oo ver – p Source: CAAP
Fallacy of composition: industry has increasingly built towards Cushing, assuming that everyone was a “ ” “ ”
39
Crude Pipelines: “Cushing-or-Bust” Backfires
Despite record-high storage levels, inflows from oil sands, Bakken, Permian, and even the Rockies have not sufficiently abated, and have continued to push down WTI pricing
Due to limited Cushing storage, WTI pricing has not been discounted to levels comparable with local Bakken pricing; for western Canadian producers and Permian Basin producers, comparative WTI pricing is still attractive
Cushing Storage vs. Historic Range on Left Axis; ) s l b b ( g n i h s u C t a e g a r o t S e d u r C
$60.00 40,000
$50.00
, 30,000 25,000
$40.00
Storage well above historical norms
$30.00 $20.00
20,000 15,000
$10.00
10,000
.
D i f f e r e n t i a l ( $ / b l )
5,000
($10.00)
0
($20.00)
6/25/10
9/24/10 2006-2010 Storage Range
12/24/10 2006-2010 Storage Average
LLS-WTI Spread
Canadian Sour-WTI Spread
3/25/11 6/24/11 Trailing 12 Mos. Cushing Storage Permian-WTI Spread
Source: EIA
40
Potential Outlets to the Gulf Coast
Transcanada’s Keystone projects added 156 kbpd into Cushing, with a planned 700 kbpd extension (“XL”) to move heavy Canadian crudes from Cushing to Houston markets
Enterprise and Energy Transfer’s “Double E” Pipeline would provide the earliest solution, utilizing existing underutilized natural gas pipeline in Texas to enter service before the end of 2012 – ’ 350 kbpd of light crude to the Gulf Coast. It was re-announced (after being shelved) in September 2010 has been on hold, awaiting a US gov. decision on XL and evaluating the long-term demand for Cushing-Houston transportation
Reversing Available PADD2-PADD3 Transportation is not Enough to Stabilize the Imbalance 180,000
2,500,000
y150,000 a D r120,000 e , s l e 60,000 r r a B 30,000
2,000,000 1,500,000 1,000,000 500,000 0
0 To PADD3 Pipeline
To PADD3 Barge
From PADD3 Pipeline
From PADD3 Barge
Source: EIA
41
Well, Where the Heck are those Pipes?
XL is technically a separate pipeline for purposes of government approval, has been protested by environmental groups due to the fact that it will increase US imports of oil sands crude
XL has been delayed by at least a year due to a heavily politicized review process by the US State Dept., which, unless accompanied by a comprehensive national energy policy that uses non-market forces to radically alter US crude slates, will do nothin to chan e the fact that US crude su l is ettin heavier and more Canadian □
□
Keystone was approved in 2008 to import 590 kbpd of “oil sands crude” to Illinois, and is flowing volumes now ’ their project rather than create a bottleneck and distort the price of crude in the largest oil hub in the world? One more: won’t delaying the approval just encourage competitor i elines who avoid “im ort” i eline status but still shi Canadian crude to the Gulf Coast? The answers to the former is almost certainly “yes,” the answer to the second is objectively “yes,” as “non-import” competitor pipelines have been announced see followin a e)
42
Handicapping Pipelines out of Cushing
We believe that XL will almost certainly be built (eventually), as Transcanada can most likely modify their pipeline design to make XL technically a “non-import” pipeline. Given the political climate, it is impossible to predict whether this is a 2013 or 2014 event. ’ . in January 2011 that the project is close to receiving the anchor shipper commitment needed to make it economically viable. We still think that Monarch is unlikely, after XL and Double E are built.
Enterprise and Energy Transfer have announced a JV (“Double E” pipeline) to convert existing exas n ras a e gas p pe nes an ay new p pe o u a p pe ne rom us ng o e u Coast. Inbound Pipes
Capacity (kbpd) Outbound Pipes
Capacity (kbpd)
Basin Seaway Spear ea Sout Centurion North Keystone Pipeline EOG Rail to Stroud Other
450.0 BP No. 1 350.0 Ozark 193.3 Osage 175.0 Ponca City 156.0 PAA to CVR 60.0 Borger 133.0 Other , . TPH Estimated Max. Structural Imbalance: Proposed Outbound Lines
Likelihood
"Double E" Pipeline ETP/EPD High Keystone XL TRP High Monarc ENB <50% Pro Forma Structural Outbound Capacity
175.0 170.0 135.0 130.0 80.0 59.0 280.0 , . (488.3) In‐service Capacity (kbpd)
Q4 2012 2H 2013 2014
450.0 700.0 350.0 661.7
Source: CAPP, press releases, filings
43
Can Trucks or Rail Provide a Release Valve for Cushing?
Over the past year, we have seen Cushing inventories at levels approximately 5.8 mmbbls higher than the previous year (average 16 kbpd of inventory growth)
Rail will contribute to reducing the bottleneck in 1H 2012. Rail involves limited capital and typically less than a year of lead time. It also provides destination flexibility for □ shippers. Rail is more expensive than pipelines - $4-$6 bbls from Cushing to the Gulf Coast. Currently, the nearest rail depot to Cushing is Stroud (~20 miles), where EOG ships up to 60 kbpd of inbound Bakken □ volumes on their WATCO-operated rail line. In March 2011, KMP and WATCO announced a project to provide rail takeaway to the Midcon. We expect that this project could provide up to 70 or 80 kbpd. We doubt that many rail operators will be lured by the Cushing opportunity, when >1mmbpd of outbound pipelines are scheduled to come online in 2012/13.
Trucking cannot meaningfully reduce the imbalance. Trucks can be employed with minimal lead time and capital, but they are expensive - $9-$11/bbl. □ >1,400 trucks would need to be employed full-time to take 100 kbpd out of Cushing. In reality, the local infrastructure (roads, truck racks, etc.) limits trucks to a minor role in offtake. Trucking industry sources have indicated that hiring truckers on a long-haul route dependent on a volatile + Despite the current spread, prospect of being long $11 transport in a trade that has historically seen ~$0 differentials is not attractive for producers/marketers.
Bottom Line: Until pipes arrive in late 2012-2013, spread is likely to remain wide
Cushing-Gulf Coast Spread vs. Costs of Transportation l . e r r a $10.00 B / $ n $5.00 i d $0.00 a e r S
.
2000
2001 2002 Rail Cost/Barrel
2003 2004 2005 Truck Cost/Barrel
2006 2007 WTI-LLS
2008
2009
2010
2011
Source: EIA, TPH estimates
44
Crude Transport Opportunities in the Eagle Ford
Although the lack of a local hub makes the differential difficult to observe, the current lack of crude off-take capacity in the Eagle Ford has forced producers to rely on high-cost trucking takeaway, reducing crude/condensate netbacks. □
Large-diameter oil pipelines have recently been announced out of the oil window of the Eagle Ford, with an aggregate capacity of at least 1,130 kbpd by 2014 perator EPD PAA/possibly CHK and Koch KMP Koch MMP/M3 Private TexStar Velocity/NS Koch/Harvest Pipeline Koch/Harvest Pipeline/NS Total Proposed Capaci ty g - e oo apac y
pe ne r g n- erm nus Wilson/LaSalle-Houston LaSalle Co.-Corpus DeWitt Co.-Houston Frio/Atascosa-Corpus LaSalle/Live Oak-Cor us S.TX-Corpus San Antonio-Corpus Frio/Atascosa-Corpus Frio/LaSalle-Corpus
apac ty 360.0 300.0 300.0 200.0 180.0 120.0 100.0 90.0 80.0 1,730.0 , .
nc or CHK CHK HK None named None named None named None named APC None named
e oo Under construction High High Moderate <50% Moderate Moderate High In service
Source: public filings, company presentations □
US Development is nearing completion on a 40,000 crude rail hub out of the Eagle Ford to service local refinery demand
45
Conway to Belvieu – The Quiet Cash Machine
Mont Belvieu, Texas is home to roughly 36% of total US nameplate fractionation (NGL separation) capacity and is America’s largest NGL hub. MB fracs typically command the highest fees of any US frac facilities, due to proximity of major petrochemical plants (55 bn lbs/ r eth lene ca acit as well as the lar e NGL stora e caverns nearb .
Conway, Kansas, is the second-largest NGL pricing hub, with 21% of fractionation capacity, but its lack of local downstream consumers (2.3 bn lbs/yr ethylene capacity) and relative lack of storage capacity makes it a much less liquid hub than Mont Belvieu.
Conwa -Mont Belvieu Ethane Price S read
200
80%
) l a g150 / s t n e100 c ( e i r P
60% 40% 20% 0%
0
C o n w a y D i s c u n t
(20%)
6/1/01 6/1/02 6/1/03 6/1/04 6/1/05 Belvieu 80% Ethane Mix Conway 80% Ethane Mix
6/1/06 6/1/07 Conway Discount
6/1/08 6/1/09 Avg. Disc. 2001-2008
6/1/10 6/1/11 Avg. Disc. since 2009
Source: Bloomberg
ue o e ane s app ca on as a pe roc em ca ee s oc , e ane pr c ng ou s e o s prone to substantial volatility, reflecting transportation costs to Belvieu as well as local supply/demand factors.
The differential between Conway and MB has steadily grown as the Conway market has . , Pass Pipeline was placed into service, bringing 140,000 bpd of eastern Rockies capacity online (expandable to 255,000 bpd). 46
Conway to Belvieu – The End of a Differential?
Differentials are valuable things, and companies will fight to control them – sometimes at the expense of the differential. , failed to keep pace with the demand for transportation to high-value Gulf Coast markets.
’ – historically provided all takeaway capacity from Conway to Belvieu, allowed the spread to widen, thereby forcing Midcon and Rockies NGL producers without additional transport capacity to MB to take a discount on NGL sales.
In June 2011, DCP Midstream (private SE/COP JV) announced that they would be purchasing COP’s Seaway Products pipeline from Cushing to Houston, extending it to Conway and converting it to 150 kbpd of NGL service. This is timed to be in-service b 2 2013 before ONEOK com letes a 193-kb d expansion of their Conway-Belvieu Sterling system (see details later in report).
It is likely that ONEOK and DCP fill capacity on both pipelines, but the golden goose – the spread from Conway to Belvieu, will be diminished by a corridor with multiple pipeline operators competing for volumes. 47
Intermediating the Seasonality of Supply and Demand
48
Gas Storage: A Quick Primer
The need for gas storage is a result of the simple fact that demand is seasonal and supply is not
As shown in the illustration below, in the Northeast, winter demand often exceeds the amount of daily available gas (by as much as 10 bcf/d), necessitating that withdrawals from storage compensate for the shortfall
Holders of storage make money by injecting gas into storage during the summer season – roughly April through October – and then agreeing to sell that gas in the winter. The futures market places a premium on winter gas based on perceived concerns about a supply shortfall (premiums have fallen – see next pg) Peak Demand
23.0 21.0
) d / f 19.0 c b ( 17.0 d n 15.0 a
Withdrawal
6.5 bcf/d shown in illustration based on monthly averages; actual daily peak several bcf/d higher
e 13.0 D s 11.0 a G E 9.0 N
Shoulder Season
Production near storage locations effectively lifts this red bar, reducing need for storage
7.0 .
0%
10%
Source: EIA, TPH estimates
20%
30%
50%
60%
70%
80%
90%
100%
Dispersion of Monthly Average NE Gas Demand, 2000-10
Historical Monthly Demand (bcf/d)
40%
Max. Pipe Capacity
Min. Storage Deliverability Needed
The development of unconventional shale gas resources drastically reduces exploration risk, to an extent . “ reservoir now serves as storage,” due to the low geologic risk associated with incremental production
49
Gas Storage: A Casualty of “Conventional Wisdom”
During the 2000s, gas storage construction was overbuilt in the Southeast and Northeast, driven by the same (ultimately incorrect) macro forecast as gas pipelines: □ Dependence on imported LNG expected to reduce the consistency of gas supplies, increase storage demand □ Declining domestic supply, increasing power gen and industrial demand would require greater gas deliverability from storage □ Increasing dependence on offshore gas would make supply vulnerable to weather-related disruptions
Projects have Been Focused on NE/SE…
The Golden Goose May Rise Again, but He’s Dead for Now …then storage buildout, shale, recession reduced seasonal spread
$4 $3
400 300
f c m $2 /
200
$1
100
$0
0
-$1 1/1/00
-100 1/1/02
1/1/04
6 Mo. Fwd. Sale Margin
1/1/06
1/1/08
1/1/10
G u l f C o a s t / N E S t o r a g e C a p a c i t y
Avg. 6 Mo Fwd Sale Margin
Source: EIA, Bloomberg
50
Gas Storage – Declining Fees, Increasing Capacity
We’re Not Seeing a Bottom: new gas storage facilities are predominantly highdeliverability salt-dome caverns, which can deliver gas 7x-10x faster than depleted reservoir storage.
□
□
by 14%, and deliverability by 44% Lease fees for high-deliverability capacity have declined from approximately $0.25 to as low as $0.12 per month per mcf , 3 high deliverability salt-dome assets on the Gulf Coast sold in 2010, at valuations that are difficult to square in the current low and declining storage lease rate environment Acq ui siti on Acq ui ro r Bobcat Spectra Southern Pines Plains Ex pansion Bobcat Expansion Tres Palacios Ph. 3 & 4 Southern Pines Expansion Bobcat Tres Palacios Southern Pines
□
Date Annc. 7/15/10 12/29/10 In-service 2015 2013 2012
Cost ($mm) $540.0 . $750.0 Cost ($mm) $425.0 $85.5 $50.0
Capacity (bcf) 18.0 . 18.0 Capacity (bcf) 28.0 20.9 23.0
$965.0 $810.5 $800.0
46.0 47.9 41.0
$/bc f TPHe EV/EBITDA $30.00 12.50x . . $41.67 17.36x $/bcf TPHe EV/EBITDA $15.18 8.43x $8.69 2.63x $2.17 1.17x $20.98 $16.92 $19.51
11.65x 10.85x 10.49x
In each case, cheap expansion capacity was cited in each case as major motivation behind the deal… when incremental high-deliverability capacity is so cheap, it makes it ar o see e o om o e mar e
Source: public filings, company presentations, TPH Estimates
51
Crude/Refined Products Storage
Regulatory mandates and unconventional resource development have contributed to changes in supply/demand dynamics that are creating opportunities in crude and refined product terminals
Crude: □ Canadian oil sands and Bakken light crude production have created an oversupplied situation at Cushing , crude. Typically, contango is caused by anticipation of future demand; this contango market is caused by extraordinary pressure on the front of the futures curve long-term throughput at Cushing
Refined products: □ Refiners’ increasing ability to process “bottom of the barrel” crudes into g er pro uc s ave re uce supp y o r y pro uc s, w e regu a ons an low natural gas prices have decreased domestic demand for dirty products. □ Despite the diminishing market size, increasing exports out of the US and decreased availability of dirty products have given providers of storage, . 52
Cushing Storage and the Contango Market
Contango describes a state of the futures market when near-term prices are lower than vs. longer-term prices
Contan o can exist due to seasonal demand atterns but it can can exist also when there is insufficient capacity to handle deliveries □ An admittedly simplistic example to help think about Cushing: Grills in January at Home Depot are in contango. HD is paying you to take
You would pay full price if you bought in June. Depending on your family’s discount rate and the amount of junk in the garage, this may or may not be attractive. This is a normal contango market. At Cushing, it’s like the Home Depot is getting truckloads of grills every day, and they are forced to drop the price just to keep inventories from overwhelming the store. HD has started building new square footage so that they can inventory more grills. ven gran ng a cus omers now n o s oca on as a gr superstore;” it’s still obvious that this is not a sustainable solution.
Bottom line: More storage at Cushing is not the solution; the market needs
pipeline offtake to structurally narrow the arbitrage
53
Cushing Storage Set to Grow Rapidly in 2011 Cushing Contango is a Result of Insufficient Pipeline Takeaway; Market Incentivizes Storage, but May Not e ec ong- erm ee or orage
Storage Buildout Serves as a Stop-gap Until Pipelines Arrive; Once Cushing Takeaway Expands, Storage will ro a y e n urp us
10%
70.0
8%
y t i a p 50.0 a C e g a r 40.0 o t S g n i . k r o W f o 20.0 s l e r r a B 10.0
60.0
6%
4%
2%
0%
-2% 2005
8.5 5.4 2.6 4.1 5.6
5.4 2.6 4.1 5.6
2.6 4.1 5.6
9.3
9.3
9.3
2.6 4.1
. 5.8
. 5.8
. 5.8
5.6
5.6
5.6
5.6
5.6
10.0
10.0
10.0
10.0
10.0
15.8
15.8
15.8
15.8
15.8
8.5
8.5
2.6 4.1
.
6.5
7.6
12.2
12.6
14.2
12.3
12.3
12.3
.
.
.
.
.
Q1 2011
Q2 2011
Q3 2011
Q4 2011
Q1 2012
Q2 2012
Q3 2012
Q4 2012
0.0 2006
2007 2009 CL3-CL1 Average Contango 2005-2008 Average Contango 2009-2011
2010
o
emgroup
n erpr se
ers
54
Crude Terminal Opportunities Outside of Cushing
Rapidly increasing Eagle Ford light crude production has created another dynamic that has upended industry expectations □
, Gulf Coast refiners spent billions through the 1990s and 2000s becoming, on average, the US’s highest complexity refineries, capable of handling the world’s heaviest and sourest crudes
TPH estimates that Eagle Ford crude production will increase to nearly 900 kbpd by 2016 □ Mostly high gravity crude (~API 50 ), not optimal for the heavy sour slates of Gulf Coast refineries □ Although Flint Hills (Koch), Valero and NuStar have announced plans to refine some of this new crude supply, we expect that much of this highvalue crude will be exported or blended into heavier crudes in Houston and Louisiana as the US imports cheaper, heavier crudes for refining
Midstream operators with marine terminals along the Gulf Coast from Corpus Christi (MMP – 1-2 mmbbls, NS – 1.6 mmbbls, PAA – 1.5 mmbbls planned 4Q 2012) are well-positioned to capitalize on this emerging waterborne crude trade
55
Refined Products: Exporting the Bottom of the Barrel
Large refineries have reduced “bottom of the barrel” output by building cokers to produce higher-value light refined products from heavy crude. Refiners are selling associated heavy product storage facilities, providing M&A opportunities for smaller midstream operators. eavy term na transact ons comp ete on t e u
oast n t e past 8 mont s:
□
1.8 mmbbls of storage in Mobile, Alabama purchased by NS in May 2010 for $25/bbl
□
544 mbbls of storage in Channelview, Texas bought by NGLS in March 2011 for $50/bbl . -
,
The withdrawal of refiners from the storage market presents opportunities for independent storage providers and marketers – NS and GEL are active marketers with terminal operations. NGLS is evaluating heavy product terminalling as a potential business segment. US Residual Production by Grade vs. Net US Imports 1000
y a D r 500 e p s l r r a B (500) d n a s (1000) u o h 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 T
<0.31% Sulfur
0.31%-1% Sulfur
>1% Sulfur
Net Imports 6 Mo. Trailing Avg
Source: EIA
56
Profiting from the Dry Gas Price Dislocation
57
North America Moves Down the Energy Dispatch Curve
The unconventional resource revolution has transformed the North American gas market. Even if the US has not yet figured out how to fully capitalize on our novel position as a low-cost energy supplier (as evidenced by overflowing gas storage, shale moratoria, and other head-scratchers , the chan es it has brou ht and will brin to our economy are profound.
The good news is that North American midstream operators are uniquely positioned to benefit from a wide array of these changes.
We believe that the true game-changing opportunity for the midstream industry is the price dislocation between wet and dry BTUs of energy Nat Gas and NGL Prices per BTU
Most Pressure on Light NGLs: NonBlendstocks with Limited Trade
140% T 120% B r e 100% p e c 80% i r P I 60% T
os re uc on o lighter NGLs driven by gas supply
n o i t c u d o r P S U f o % s A
f 40% o %20% s A
0% Natural
Ethane
Propane
2000-2007
Butane 2010-2011
Isobutane Natural
90% 80% 70% 60% 50% 40% 30% 20% 10% 0% -10%
64%
61%
0% 0% 0%
11% 1% 0%
Ethane
Propane
Imports, 2000-2009 Used as Refining Blends
12%
-3% Butane
9% 5%
6%
-4% Natural Gasoline Imports, 2010-2011 YTD Isobutane
Source: Bloomberg, EIA
58
NGL Primer: How Do NGL Recoveries Affect Netbacks?
Basins are like a box of chocolates, so there’s no rule of thumb for netbacks. US average is 2 gallons per mcf (GPM). Gas associated with oil production can be 10 GPM, increasing netbacks by >100% vs. selling the gas at its “dry gas value”. verage mer can c o as From the wellhead 1.123 mmbtu of gas Processed Components 1.000 mmbtu of dry gas 0.123 mmbtus of NGLs 1.410 . allons of NGLs 0.601 gallon of ethane 0.399 gallon of propane 0.103 gallon of butane 0.127 gallon of isobutane 0.180 gallon of natural gasoline Total Netback for Gas and NGLs Increased Netback Under Current Pricing
r ce
Eagle Ford " NGL Window " Gas From the wellhead 1.386 mmbtu of gas Processed Components 1.000 mmbtu of dry gas . mm tus o s 4.500 gallons of NGLs 2.295 gallon of ethane 0.945 gallon of propane 0.315 gallon of butane 0.315 gallon of isobutane . Total Netback for Gas and NGLs Increased Netback Under Current Pricing
Price
e ac
$4.32
$4.85
$4.32
$4.32
$0.75 $1.52 $1.73 $1.97 $2.45
$0.45 $0.61 $0.18 $0.25 $0.44 $6.24 28.7%
% of NGL Bbl 42.7% 28.3% 7.3% 9.0% 12.7% 100.0%
er mmbtu $10.20 $16.60 $16.75 $20.82 $22.30
% of NGL Bbl 51.0% 21.0% 7.0% 7.0% . 100.0%
$ per mmbtu $10.20 $16.60 $16.75 $20.82 .
Netback
$4.32
$5.99
$4.32
$4.32
$0.75 $1.52 $1.73 $1.97 .
$1.72 $1.43 $0.54 $0.62 . $10.18 70.0%
Source: EIA, company press releases, TPH estimates 59
Wet Shale Gas: Game-Changing for the American Economy
The unconventional resource revolution has transformed the North American gas market. Even if the US has not yet figured out how to fully capitalize on the novel position of being a low-cost energy supplier (as evidenced by overflowing , , , wrought in our economy are profound
Oil and gas are not directly fungible (i.e., a gasoline-powered car cannot run on natural gas), which is the main reason for the price disparity between the two. ere are ways o su s u e ower-cos gas ypro uc s or g er-cos o byproducts, effectively increasing the demand for wet gas: □
□
□
Ethylene, the most widely produced plastic feedstock, can be produced from ethane or propane, or from crude byproducts like naphtha and gas oil, a ow ng petroc em ca compan es to exp o t t e ow pr ce o gas Knock-on economic effects include the effective export of BTUs of gas in the form of plastics (polyethylene and polyethylene derivatives) Propane has transportation and heating applications. The US, once an importer of propane, now exports propane, taking advantage of demand in developing markets that are transitioning away from open flame cooking and heating. Cheap natural gas is driving the obsolescence of residual fuel oil power generation in certain US markets (primarily East Coast). 60
Eagle Ford: The Tidal Wave of NGLs Needs Processing
The Eagle Ford is an area with emerging gathering/processing “bottlenecks” – current processing capacity handles >4bcf/d of gas, and produces >200 kbpd of NGLs (much drier average than Eagle Ford gas)
Eagle Ford Processing
DCP Midstream
3Q 2012
48.2
Bottlenecks have developed due to the drilling pace outstripping the pipe-laying pace, this is a temporary phenomenon that should be sorted out by late 2011.
ETP
1Q 2013
64.3
□
□
□
□
Eagle Ford, Granite Wash and the Permian have substantial processing facilities, which are wellpositioned to handle unconventional volumes. Eagle Ford, located in S. TX with substantial legacy dry gas production, has major gas processing capacity needs for 380 kbpd of Eagle Ford NGLs by 2016 (TPHe). In addition to existing underutilized capacity from DCP and EPD, processors have announced another 2.5 bcf/d of processing capacity, and 600 mmcf/d of expansions . We do not believe that major processing bottlenecks will arise, as announced additions can handle the vast majority of estimated peak NGL production.
CPNO/KMP - Formosa
TPH est. NGL Prod. Capacity (kbpd) Currently
21.4
CPNO/KMP - WPZ
3Q 2011
10.7
Southcross
2Q 2012
25.2
EPD
2
64.3
CPNO
2012
1Q 2013
42.9
Announced NGL Production Capacity
277.0
Announced Potential Expansions CPNO/KMP
2014+
10.7
ETP
2014+
21.4
EPD
2014+
32.1
Annc. Potential NGL Prod. Capacity Total Potential NGL Prod. Capacity
64.3 341.3
Source: Company press releases, Oil & Gas Journal, SEC filings, TPH estimates
Processing (extraction from dry gas stream) capacity is manageable; it’s fractionation believe will drive more Eagle Ford spending
We do not believe that significant new processing providers will enter the Eagle Ford, as currently firm, announced ca acit will be able to handle ~70% of TPHe maximum NGL production 61
Marcellus Field Gathering - Still A Competitive Landscape
We believe the Marcellus competitive environment is more fluid than that in the Eagle Ford, Granite Wash and Permian Basin, with less dominant midstream players and more opportunities for M&A □
□
□
WPZ’s Marcellus exposure has expanded from a JV with a small-cap midstream company in mid-2010 to now emerging as the largest gas gatherer in the Marcellus with two systems, serving COG and CVX/RIL, among others ma a erers ave s a e ou oo o s n arce us. r va e opera ors e a man a er n process n an Momentum (gathering) have entered the Marcellus, possibly offering opportunities for larger players to “buy in” DCP, one of the largest processors in the US, has indicated an interest in becoming involved in the Marcellus
Major producers – APC/UPL, XCO/BG - are providing their own midstream services. Smaller producers - REXX, MHR providing their own services. This could provide M&A opportunities as bottlenecks ease and E&Ps monetize non-E&P assets
Release Date Producer
9/15/2009
Geography
Gatherer
Processor
CHK/STL
Marshall/Wetzel Co, WV
NI - Columbia Gas
MWE JV - Majorsville, WV Plant
CVX/Reliance
SE PA; Dry Gas
Laurel Mtn (CVX/WPZ JV)
na
3/10/2010
RRC
Lycoming/Tioga Co, NE PA
PVR
na
6 1 2010
HK
Wet Marcellus
na
MWE JV - Houston, PA Plant
2009
11/18/2010
TL RR
COG
NE PA
WPZ
na
1/11/2011
CHK/STL
N. WV
na
MWE JV - Majorsville, WV Plant
1/26/2011
Chief/CHK/GST/SGY/Others Marshall/Wetzel Co, WV
Caiman
Caiman's Cameron, WV Plant
Announced Processing Capacity
Facility
.
mmcf/d
an
.
Announced Marcellus Frac. Capacity Facility
.
kbpd
ous on rac.
ropane+
.
2011
MWE JV
SW. PA and N. WV
335.0
4Q 2011
Caiman
Ft. Beeler Frac.
2011
Caiman
Ft. Beeler, Cameron, WV
320.0
2012
MWE
Houston Frac. (Propane+)
18.0
2012
D
Wetzel, WV Plant
300.0
2013E
MWE
Houston Frac. (Ethane+)
75.0
2012
Caiman
Ft. Beeler, Cameron, WV
200.0
4Q 2013E
Caiman
Ft. Beeler Frac. Exp.
25.0
. 2013+
MWE
.
SW. PA and N. WV
Potential Processi ng Total: Note: all Houston frac trains can be upgraded to recover ethane as needed Source: company press releases, filings, TPH estimates
. 455.0
12.5
, Total Potential Fracti onati on Capacity:
. 204.5
2,220.0
62
DJ Basin Processing and NGL Takeaway
There is plenty of gas takeaway capacity in the DJ basin; however, the bottleneck is processing capacity. Due to richness of gas in the basin (~4 GPM) production is constrained by processing. Due to recent
buildouts natural gas takeaway is not an issue, however, rich gas is hard to flare and processing is tight.
In our view, processing infrastructure will eventually be necessary and is at least one year out. Processing
Near-term, raw NGL export capacity should be adequate relative to processing capacity. Current 77mbpd
capac y s om na e y na ar o e ro eum an s ream. n arc , Wattenberg processing plant, consolidating 195mmcf/d of processing capacity.
purc ase
’s
total NGL takeaway capacity vs. ~750mmcf/d of processing implies a capacity ratio of 3.9 GPM. We believe that incremental takeaway will be accommodated by the planned expansion of Overland Pass pipeline (additional 115kb d with 60mb d earmarked for Bakken and the remainin 55mb d available for Rockies NGLs . Processing Capacity
Facility
mmcf/d
Online
APC/WES
Wattenberg (fmr. BP), Ft. Lupton, Platte Valley (fmr. ECA)
385.0
Online
DCP Mids tream (DPM )
Eaton, Greeley, Mewbourn, Plattev ille, Roggen, Spindle, Lucerne
400.0
Online
Other
Various <20mmcf/d plants
urren
48.0
rocess ng o a :
Potential Expansions
. Facility
mmcf/d
3Q 2011 APC/WES
Ft. Lupton (expansion underway to existing plant)
2Q 2013 DCP Midstream LLC
La Salle (announced new plant and gathering system)
2013
APC/WES
Ft. Lupton (potential expansion to existing plant)
Potentia Processing Tota : Pipeline Capacity
Pipeline
kbpd
DCP Midstream Partners
Wattenberg
22.0
Online
Overland Pass Pipeline (OKS-WMB JV)
DJ Basin Lateral
55.0
. 2013
Overland Pass Pipeline (OKS-WMB JV)
Potential NGL Takeway Capacity
110.0 40.0
998.0
Online
Pipeline Capacity
15.0
Pipeline
kbpd
DJ Basin Lateral
55.0
132.0
Marcellus/Bakken NGLs: Further Afield
The end market for Bakken NGLs remains to be seen: □
We believe that Bakken ethane will receive the best netback in Alberta, Canada
OKS will provide a 60 kbpd NGL line to Conway, which we believe will ship heavier NGLs to larger refining/NGL consumption markets.
End market for Marcellus ethane is also a subject of debate: □
We believe that Shell’s recent announcement of plans to build a Marcellus-area cracker will effectively prevent Marcellus ethane from reaching the Gulf Coast or being blended into pipelines over the long term. Cracker will likely require up to 70 kbpd of ethane capacity (no timing , - . storage/contracted pricing to persuade Marcellus producers that local ethane prices will not be wildly volatile without a Gulf Coast outlet. – The local geology does not support salt dome storage, so we expect that high-cost, above-ground “bullet tanks” may be necessary. – Alternatively, it is possible that southern NY gas storage may be converted to hold ethane and ethylene, as gas storage demand has been and will be diminished by Marcellus production. Remaining 30-50 kbpd of 2015E Marcellus ethane will be sent to Sarnia, Ontario pe roc em ca p an s a manannounce or p .
64
NGL Supply Driving Billions in Petchem Investment
In December 2010 and subsequently in April 2011, Dow Chemical, the 2nd largest ethylene producer in the US (7.7 bn lbs/yr, 13.4% of US), announced plans to increase ethane consumption in North America □
Restarting ethylene cracker at St. Charles, LA by YE 2012
□
Improving ethane capacity at Plaquemine, LA in 2014, Freeport, TX in 2016
□
Building new ethylene cracker on Gulf Coast by 2017
Potential newbuild crackers (worldscale facilities produce >2.2 bn lbs/year, and can consume up to 70 kbpd of ethane): □
ChevronPhillips - feasibility study announced March 2011 for potential 2014-15 construction
□
LyondellBasell – announced interest in potential ethylene JV
□
Shell – plans to build worldscale Marcellus cracker, location and timeframe undisclosed
Smaller expansions announced: □
Formosa – expanding ethylene production in S. TX
□
Westlake – expanding ethane consumption in Louisiana and Kentucky
□
INEOS – studying debottleneck of existing cracker in Texas
65
US Ethane Demand Set to Increase by Up to 40% by 2017 Ti mi ng Dow Announ cements - Decemb er 2010 and April 2011 Increasing Total Ethane by 20%-30% (Dec 2010 annc.) . < * Plaquemine E/P (50-90% ethane currently) Freeport 2 Flexi (<50% ethane currently) Potential Newbuild GC Ethylene Cracker Total from Dow Announcements Potential Expansions Westlake - phase 1 ethane increase at Lake Charles Westlake - phase 2 ethane increase at Lake Charles Westlake - evaluating ethane feedstocks in Kentucky Formosa - 1bn lb/yr expansion under consideration INEOS - 250 mm lb/yr debottlnecking under study
Cu rr en t Eth yl en e TPHe Ad d' l Eth an e Prod (m ml bs/yr) Consum pti on (kbpd)
2011-13 2014 2017 2017
7,749 , 2,700 2,200 0
35,000 , 8,000 12,000 60,000 115,000
2012 2014E 2013-14E 2015E 2014E
, 2,400 2,400 430 3,300 1,825
, 10,000 7,000 6,000 23,000 10,000 ,
0 0 0
60,000 60,000 60,000 180,000 213, 000 153,000 356,000
Potential New build Crackers Chevron Phillips - Gulf Coast feasibility study by YE 2011 2014-15E Shell Marcellus Cracker - announced, no details yet 2014-16E LyondellBasell - JV under consideration 2015-17E Total from Potential New builds Estimated Consum tion from Hi h Likelihood Pro ects Estimated Consumption from High Likelihood Projects on Gulf Coast Total Estimated Consumption, Including Low er Likelihood Projects
TPH Est. of Li kel ihood High High High Moderate
High Moderate Moderate High Moderate
High High 50/50
*Included in 20%-30% expansion of GC ethane crack ing shown on top line
Sources: Company press releases and presentations, TPH estimates
66
The Gulf Coast is Currently Short Ethane
Recent US demand for ethane has averaged 975 kbpd vs. production of roughly 925 kbpd, leaving the US approximately ~50 kbpd short
Conway Hub 520 kbpd
pe nes n o u oas Line EZ W. TX LPG W. TX NGL Arbuckle Sterling I and II Seminole Chapparal Total Inbound NGL Capacity
p 101.0 230.0 144.0 180.0 300.0 . 250.0 135.0 1,367.0
Gu lf Co ast Fra cti on ati on Ca pa ci ty
LA Gulf Mont Belvieu 880 kbpd
SE TX
480 kbpd
NGLS Cedar Bayou Fractionator OKS MB-1 Fractionator COP Gulf Coast Fractionator Mont Belvieu Copano HCPP EPD Armstron and Shou Formosa Other South Tex as Lou isia na Gulf Coast Total Gulf Coast Capacity . *
aw ur y Y-Grade Y-Grade Y-Grade Y-Grade Purity Y-Grade Y-Grade
kb pd . 293.0 160.0 102.0 860.0 22.0 97. 0 40.0 210.0 369.0 477.0 1,706.0 .
*Includes frac capacity + purity prods from Sterling
67
Paving the NGL Highway to the Gulf Coast
NGL pipeline and fractionation projects have been announced at a breakneck pace in 2011. Fractionation currently lags pipeline capacity. NGLs from the Eagle Ford could add another Future Pipes to GC Timi ng kbpd Raw /Purity 390 kbpd requiring fractionation by 2016.
onway
u
520 kbpd 6 0 kbpd
Sterling I 2Q 2011 MAPL 3Q 2012 Arbuckle 2Q 2012 Lone Star 1Q 2013 Sandhills 2Q 2013 Southern Hills 2Q 2013 Mariner East 2013-14 MEPS 2013-14 Sterling III 4Q 2013 MAPL/Seminole 3Q 2014 Total Inbound NGL Capaci ty Hi gh Likelihood NGL Capaci ty Fractionati on Adds
Mont Belvieu SE TX 390 kb d
LA Gulf 480 kbp d
880 kbpd 4 7 0 kb d
132 kbpd
15.0 15.0 75.0 130.0 130.0 150.0 50.0 60.0 193.0 50.0 868.0 758.0 Ti mi ng
GCF Expansion 2Q 2012 Lone Star JV 1Q 2013 MB2 1Q 2013 Cedar Bayou 2Q 2013 MB2 Expansion 4Q 2013 EPD Phase VI 2013 Mont Bel vieu Copano HCPP 4Q 2011 Formosa - Pt. Comfort 1Q 2013 Oxy - Ingleside 2013 South Tex as Total Gulf Coast Expansions High Likelihood Expansions Add 'l Ethane Prod uct ion *
Li kel ihood Purity High Y-Grade Hi h Y-Grade High Y-Grade High Y-Grade High Y-Grade High Purity Low Purity Low Both High Y-Grade High 63.5% increase 55.4% increase
kbpd . 43.0 100.0 75.0 100.0 50.0 75.0 518.0 22.0 35.0 75.0 132.0 650.0 575.0 266.7
Likelihood High High High High High High 60.2% i ncrease High High Moderate 35.8% i ncrease 38.1% increase 33.7% increase 28.8% i ncre ase
Note(*): Max Gulf Coast ethane supply increased by 60 kbpd Bushton frac expansion 68
Announcements Indicate Possible Softness in Ethane in 2013
Announcements have been issued at a pace that makes forecasting supply/demand dynamics difficult; however, announcements of incremental ethane supplies have clearly outpaced announced demand from petrochems The graph below does not consider any potential demand from unannounced petrochemical cracker turnarounds; industry sources indicate that these could add well over 100 kbpd of incremental ethane demand urren
orecas e
ane upp y- eman
a ance, ase on nnounce
ro ec s
400 d p 300 b k , e 200 n a h t 100 E B 0 M f o y -100 l p p u S -200
-300 -400 2Q 2011 2Q 2012 2Q 2013 2Q 2014 2Q 2015 MB Fractionation Announced Purity Pipeline into MB Potential Newbuild TX Crackers DOW Cracker Ex ansions WLK Cracker Expansions ChevronPhillips Cracker Expansions Current Projected Ethane Supply/(Deficit)
2Q 2016 2Q 2017 Ethane from Raw NGL Pipelines Formosa Cracker Ex ansions INEOS Potential Cracker Expansion
Source: company press releases, presentations, TPH estimates
69
More American LPGs Will Be Consumed Overseas
Potential fractionation overbuild will likely put some pressure on ethane. Longer-term, we believe ethane is going to have a long-term advantage as a petrochemical feedstock due to North American shale gas supplies.
Other products should weather ethane oversupply due to the industry’s growing ability to export LPGs (propane and butane). □
□
EPD announced plans in March 2011 to double their LPG export capacity (already the largest in North America) to roughly 150 kbpd to accommodate the excess NGLs that will resu rom e com n wave o s. We believe that NGLS, the owner of the second-largest LPG export facility, will upgrade dock facilities (the 2 nd-largest in North America) to export higher-quality (international grade) propane from Houston from approximately 60 kbpd to up to 125 kbpd. xpor s an
e
xpor s
150 y a 100 D / s 50 l e r 0 r a d n -100 a s u -150 o h -200 T -250 1981
1983
1985
1987
Propane Exports
1989
1991
1993
Butane Exports
1995
1997
1999
Net Propane Exports
2001
2003
2005
2007
2009
Net Butane Exports
Source: EIA
70
