Sentimental Journey: the long, winding road to a renewable energy future June 2016

The journey to a renewable energy future is taking longer than many analysts and agencies
1
expected. How so? The solid line in the first chart shows the percentage of US primary energy derived
2
from renewable sources, alongside some over-optimistic prior forecasts . Similarly, most forecasts for
electric vehicles (EVs and plug-in hybrid vehicles) have been too high as well, as shown in the 2nd chart.

The share of US primary energy coming from renewable Another generation of electric car projections out of
sources, and some notable forecasts sync with reality, EV+PHEV sales as % of total car sales
50% 12%
Physicist Bent Sorensen Deutsche Bank
Google 2030 Clean ● US
Energy Plan 10% PwC
40% Frost & Sullivan
● Global
Bloomberg NEF
Amory Lovins (RMI) 8%
30% IEA
Carter Admin Roland Berger
(solar only) Nat'l 6%
20% Renew. BCG
Energy 4%
10% Lab Deloitte
2%
Actual renewable share of primary energy Actual US EV+PHEV sales
0% 0%
1960 1970 1980 1990 2000 2010 2020 2030 '09 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 '20
Source: EIA, listed authors, Vaclav Smil, JPMAM. 2015. Renewables include Source: DOE, BEA and listed organizations. 2015. Note: global
wind, solar, hydropower, geothermal, biomass, wood and waste. EV+PHEV sales in 2015 were also around 0.6%.

Now that learning curves have contributed to declines in solar and wind capital costs 3, renewable
energy is finally showing up on the global electricity radar screen. In 2015, global investment in
renewable energy reached a new all-time high at $286 billion, despite sharp falls in oil, coal and gas prices.
Almost all of the recent growth is wind and solar, with the largest regional increases in China/Asia.
Renewables ex-hydro made up 54% of all capacity installed around the world in 2015 (the first time they
represented a majority), and accounted for 7% of global electricity generation in 2015. Wind has been
growing steadily, while solar remains at just 1% of total generation.
Annual global investment in renewable energy Global renewable electricity generation: still mostly a
USD billions hydro-electric story, but wind rising, % of total generation
$300 25%
Middle East & Africa
$250
20%
Americas ex. US US Hydro
$200
15%
$150
Europe Asia ex. 10%
$100 China
Solar
5%
$50 Geothermal, biomass, & other
China
Wind
$0 0%
'04 '05 '06 '07 '08 '09 '10 '11 '12 '13 '14 '15 '85 '87 '89 '91 '93 '95 '97 '99 '01 '03 '05 '07 '09 '11 '13 '15
Source: UNEP, Bloomberg New Energy Finance. 2015. Source: BP Statistical Review of World Energy. 2015

1
US primary energy use by sector: electricity 39%, transportation 28%, industrial 22%, residential/commercial 11%.
2
Rather than focusing just on the US, Jacobson and Delucchi (2009) proposed a global plan for 2030 in which
100% of world energy would come from renewables. Parts list: 3.8 million 5 MW wind turbines; 40,000 300 MW PV
plants; 1.7 billion 3 kW rooftop solar installations; 5,350 100 MW geothermal plants; 270 1.3 GW hydropower
stations; 720,000 wave devices; 490,000 1 MW tidal turbines; and unprecedented extensions of high-voltage
transmission. Assembly required.
3
Since 1976, solar module costs have declined steadily from $100 per watt to $1 per watt as production volumes
increased. Wind costs also declined from 1984 to 2004, after which prices stopped falling due to rising component
and labor costs. For more information, see our learning curve exhibit from last year’s energy paper.
1
Will renewable energy continue its recent rapid expansion? If so, wind and solar intermittency costs
will have to be absorbed or eliminated. In last year’s energy paper, we analyzed Germany’s Energiewende
plan to meet 80% of electricity demand from renewable sources by 2050. One notable finding about
Energiewende: Germany’s thermal (coal and natural gas) capacity needs would be almost unchanged from
current levels, given the need to meet electricity demand in winter months when German wind and solar
irradiance levels are low. As an indication that our estimates may be on the right track, consider the next
chart. Despite a lot of renewable capacity added since 2001, Germany’s coal and natural gas capacity has
not declined at all. In other words, grids with high wind and solar penetration can still be heavily reliant on
backup thermal capacity. This dynamic is not incorporated into levelized costs per kWh for renewable
energy, and may also explain why Germany’s electricity costs are the highest in Europe. We get into more
details in the New York section on pages 14-15.
Despite a large renewable energy build-out in Germany,
almost no reduction in natural gas and coal capacity
100 625

80
Thermal capacity 600
(coal & gas), GW
60
Electricity
575
consumption, TWh
40

550
20 Solar and wind
capacity, GW
0 525
'91 '93 '95 '97 '99 '01 '03 '05 '07 '09 '11 '13
Source: German Federal Ministry for Economic Affairs and Energy. 2014.

For investors, the journey to a more renewable and cleaner energy future has translated into
substantial risk. Since 2003 and 2009, two notable renewable and alternative energy indices have trailed
4
both the S&P energy sector (comprised mostly of oil & gas stocks), and the broad equity market as well .

Renewable/clean energy stocks, traditional energy Renewable/clean energy stocks, traditional energy
stocks and the broad market: 2003-2016, Cumulative return stocks and the broad market: 2009-2016, Cumulative return
350% 250%

300%
200%
S&P Large Cap Energy
Wilderhill Clean Energy

S&P Small Cap Energy

and Alternative Energy
FTSE EO Renewable
Wilderhill Clean Energy

S&P Large Cap Energy

S&P Small Cap Energy

250%
S&P Mid Cap Energy
S&P Mid Cap Energy

150%
200%
S&P Small Cap
S&P Small Cap

S&P Mid Cap

150% 100%
S&P Mid Cap

S&P 500
100%
S&P 500

50%
50%
0%
0%

-50% -50%
Source: Bloomberg. Date range: December 31, 2002 to June 14, 2016. Source: Bloomberg. Date range: December 31, 2008 to June 14, 2016.

4
The Wilderhill Clean Energy Index includes US-listed companies focused on renewable energy production, energy
conversion (DC-AC, fuel cells, LED screen manufacturers), pollution prevention, conservation and power delivery. The
FTSE Environmental Opportunities Renewable and Alternative Energy Index, launched in 2008, includes global
companies involved in renewable & alternative energy, energy efficiency and waste and pollution control.
2
Renewable and clean tech investments have been made by venture capital and private equity investors
as well. Private investment is by definition harder to track, but there are sources to look at. Cambridge
5
Associates compiles data on venture capital and private equity investments in clean tech companies . As
with publicly tradable equities, the Cambridge private investment renewable energy and clean
tech universe has generated substantial risk and low returns to-date:
• Through September 2015, the gross internal rate of return (before fees) on the Cambridge
renewable/clean tech universe of private companies is 3.4%
• US companies which comprise around ¾ of the universe generated a gross internal rate of
return before fees of 1.7%
• Cambridge notes that when measured across their entire dataset, such company-level returns
are typically 4.4% higher than net fund-level returns earned by limited partner investors

Bankruptcies and
A closer look at70%+ stock price
renewable energydeclines in thetech
and clean renewable, alternative
bankruptcies and clean
and large stockenergy sectors
price declines
The following companies are reminders of the risks and exaggerated hype often associated with renewable
energy and clean tech stocks. Each stock shown went bankrupt, or suffered a price decline of 70%+ from its
peak. The list is primarily drawn from our analysis of concentrated stock risk in the history of the Russell 3000
Index (“The Agony & The Ecstasy”). The list includes public companies only, and excludes failed private
companies such as Solyndra (solar panels); Better Place (electric car charging stations); Range Fuels, Qteros,
Coskata and Terrabon (biofuels); etc.

Biofuels Fuel Cells, Engines Geothermal Solar Wind

Amyris and Batteries Alternative Earth Resources Akeena Solar Broadwind Energy
Aventine A123 Systems Geothermal Resources Intl American Solar King China Ming Yang Wind
Bioamber Advanced Battery Technologies Applied Solar Energy Cleantech Solutions
Gevo Aura Systems Pollution control Ascent Solar Technologies
KiOR Azure Dynamics Earthshell Corp AstroPower
TerraVia Ballard Power Systems Environmental Elements Energy Conversion Devices Misc. services
VeraSun ECOtality Fuel Tech Enphase Energy Intermolecular
Ener1 Wahlco Systems Equinox Lime Energy
Energy conservation FuelCell Energy Evergreen Solar
Blue Earth H Power Power generation, conversion First Solar
Comverge Impco/Fuel Systems Sol. and storage Hoku Scientific
Magnetek Medis Technologies Aixtron Real Goods Solar
Orion Energy Systems Millennium Cell American Superconductor Sky Solar Group
SemiLEDs Plug Power Amtech STR Holdings
Power Solutions Int. Beacon Power SunEdison
Rare earth minerals Quantum Fuel Systems Capstone Turbine SunPower
5N Plus Sonex Research Proton Energy Suntech Power
Daqo New Energy Unique Mobility/UQM Satcon Technology Trendsetter Solar Products
Molycorp Valence Technology TerraForm Yingli Green Energy
Rare Element Resources Westport Innovations

Source: JPMAM, Bloomberg. May 31, 2016. The companies above are shown for illustrative purposes only. Their inclusion should not be interpreted as a
recommendation to buy or sell. The use of the above company logos is in no way an endorsement for JPMIM investment management services.

5
The Cambridge Clean Tech universe includes renewable power manufacturing and development, energy
optimization and conservation, energy storage, recycling and pollution control. As of Q3 2015, the universe was
comprised of 810 companies in which 480 venture capital and private equity funds invested. Over 90% of the
investments were made after 2005. Cambridge estimates that their sample represents 20%-40% of all clean tech
private investment. I consider the returns to be informative since 45% of the investments have been realized.
3
This year’s paper continues our research into the promise, potential and constraints of renewable energy,
after a brief section on the commodity super-cycle and oil. Click on the page links for each section.
[I] How far along in the commodity super-cycle, and is the oil bounce sustainable? Pages 5-6
Prior super-cycles have taken 15-30 years to play out. However, given price declines that
have already taken place, we are probably much closer to the end of the current one. And
on oil, by mid-2017, the world may be looking at a deficit of supply relative to demand.
[II] The tortoise, the hare and the electric car: the quiet progress of hybrids over EVs Pages 7-9
Electric cars generate excitement, but EV projections have been too high for generations.
The more impactful energy-reducing trends are taking place in the hybrid market.
[III] A New York State of Mind: Thoughts on an ambitious new energy plan Pages 10-15
By 2030, New York aims to generate 50% of electricity from renewable energy, and
decrease energy consumption in buildings by 23%. Parts of the plan are achievable while
others are aspirational, particularly since New York is not a very sunny or windy place.
[IV] The Utility Empire Strikes Back: Distributed solar power and billing changes Pages 16-18
US utilities are changing their approach to customer billing, which could slow the future
growth of distributed residential and commercial solar power.

[V] Dam it: A look at the modest potential for more hydropower in the US Pages 19-20
Despite its large stock of streams and rivers, the US gets a lot less electricity from
hydropower than Norway, Switzerland, Canada, Austria and Brazil where its share of
generation is over 50%. We take a closer look at US hydropower potential.
Sources and acronyms Page 21

As always, our energy piece is overseen by Vaclav Smil, Distinguished Professor Emeritus in the Faculty of
Environment at the University of Manitoba and a Fellow of the Royal Society of Canada. His inter-
disciplinary research includes studies of energy systems (resources, conversions, and impacts), environ-
mental change (particularly global biogeochemical cycles), and the history of technical advances and
interactions among energy, environment, food, economy, and population. He is the author of 39 books
and more than 400 papers on these subjects and has lectured widely in North America, Europe, and Asia.
In 2010, Foreign Policy magazine listed him among the 100 most influential global thinkers. In 2015, he
received the biennial OPEC award for research, and is described by Bill Gates as his favorite author.
Michael Cembalest
J.P. Morgan Asset Management

This year’s cover. The picture of Jinshui

Road in Taiwan at night shows light trails and
light bursts which seem to be happening in
real time, but which are actually the result of
changes to the camera’s shutter speed and
aperture settings. With its mixture of the real
and the unreal, it captures some of the
themes of this year’s energy piece. On a
sentimental journey, some of the images you
recall seeing were real, while others existed
only in the realm of your imagination.
Credit: ©katoh123/dreamstime.com

4
[I] How far along in the commodity super-cycle, and is the oil bounce sustainable?
Mary Erdoes (head of our Asset Management business) showed me data from Paul Tudor Jones dating back
to the 1700s. Tudor’s data shows how commodity super-cycles generally took from 15 to 30 years to
bottom after the peak. The implication: there’s a long way to go before the current commodity super-cycle
ends, since we are only 4-5 years into the process. But as we first wrote in February, commodity prices
generally declined by 50%-70% in prior super-cycles. In that regard, a lot of damage has already been
done: by February 2016, commodity prices had already declined by 55% from their peak. For investors, I
think “price” is more important than “time” when thinking about where we are in this cycle.
The collapse in commodity super-cycles since 1779 EM commodity producer equities and commodity spot
0 5 10 15 20 25 30 35 prices, Index level (Dec. 1999 = 100)
1779
600
1814 EM commodity producer
Cycle decling beginning in:

1864 500 equity price index

1920
1951
1980 Time in years to 400
2011 bottom of cycle
300
1779 % decline in
1814 commodity prices
1864
200 Commodity spot price index
1920
1951 100
1980
2011 Through Feb 2016 0
0% 10% 20% 30% 40% 50% 60% 70% 80% 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Source: Tudor Investment Corporation, Bloomberg. February 29, 2016. Source: S&P, MSCI, Bloomberg. June 14, 2016.

To be clear, this commodity super-cycle was a huge one. The 3rd chart shows the $3.6 trillion spent on oil
exploration and mining since 2004 when the super-cycle began. The latest reports suggest that the copper
market may be oversupplied for another 5 years. But as usual, producers of copper, nickel, zinc and
aluminum have responded with 35%-65% cuts to capital spending. Given long lead times, the impact of
these capex cuts will likely be felt in supply terms in a few years. Typically, price responses happen sooner.
As a sign that investment consequences of the commodity unwind are in its later stages, consider the YTD
performance of global equity markets. Emerging markets commodity exporters are leading, followed by
commodity-rich Canada and Australia. All things considered, the period from 2011-2014 was the time to
position for the commodity super-cycle collapse; now, we believe it’s more about sifting through rubble for
long-term opportunities.
Global crude oil exploration & production and mining: YTD total returns in USD on major equity markets/regions
capital expenditures, USD billions (both axes) Percent
Canada &

$500 $120 20%

Australia
Large Cap

Crude oil exploration & 15%

Small Cap

$400 $100
production spending
US

US

10%
$80
$300 5%
$60 0%
EM ex-China manuf.
EM commodity
All Country

$200
World ex EM

-5%
Nasdaq

Metal companies' $40

exporters
World

capital expenditures -10%

Japan

Europe

Eurozone

offshore

$100
China

$20 -15%

$0 $0 -20%
onshore

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012
China

-25%
Source: World Bank. June 2015. Source: Bloomberg. June 14, 2016.

5
Is the oil bounce sustainable? Oil prices have risen sharply in 2016, particularly at the short end of the
curve. Is the bounce for real? We think it is. While the world is still awash in supply (2nd chart), there has
been an abrupt 35%-40% decline in global oil capital spending (3rd chart) which will likely affect future
supply. The 75%+ drop in the US rig count is another sign of this phenomenon. Capex declines are
primarily taking place outside OPEC, where average field decline rates are expected to rise back to 5% per
year 6. Furthermore, oil investors were positioned very bearishly in early 2016, setting the stage for rumors
about OPEC production freezes, even when unsubstantiated, to impact oil prices (4th chart).
The wild ride in oil prices in 2016 World oil supply-demand balance
WTI crude oil futures curve as of date shown, USD per barrel Million barrels per day
$58 4
Excess supply
$56 Jun 14
3
$54
$52 Jan 4 2
$50
1
$48 Mar 31
$46 0
$44
-1
$42
$40 -2
$38 -3
$36 Excess demand
0 12 24 36 48 60 -4
Months until delivery 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015
Source: Bloomberg. June 14, 2016. Source: IEA, Bloomberg. Q1 2016.

World oil capex Oil prices, production freeze rumors & positioning
Upstream capital spending as a share of global GDP, % USD per barrel Oil short contracts, thousands
0.9% $110 350
First production freeze
$100 rumors in the press
0.8% 300
$90
0.7% $80 Spot oil price 250
2015e
$70
0.6% 200
2016e $60
0.5% $50 Oil short contracts 150
$40
0.4% 100
$30
0.3% $20 50
1985 1990 1995 2000 2005 2010 2015 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16
Source: Barclays, IMF. December 2015. Estimates for 2015 & 2016. Source: Bloomberg, CFTC, JPMAM. June 14, 2016.

7
On oil demand , we expect consumption to grow at its recent pace of ~1.3% per year. While consumption
has declined by 10% in the OECD since 2005, it has risen by 37% outside the non-OECD8, primarily due to
demand related to transportation. In barrel terms, the non-OECD finally overtook the OECD in 2013.
Bottom line: extrapolating oil demand growth and future supply in an environment of sharply reduced
non-OPEC capital spending, the current oil glut could turn into an oil supply deficit sometime in 2017-
2018, in which case the oil price rally makes sense to us.

6
As per Wood MacKenzie, non-OPEC field decline rates averaged 5%-6% per year from 2006 to 2011. Non-OPEC
declines rates dropped to 3%-4% in 2013-2014 as higher oil prices made some expensive recovery techniques more
economically viable. However, Wood Mackenzie now expects field decline rates to rise back to 5% by 2017.
7
Is renewable energy growth affecting oil demand? We doubt it. A research firm we subscribe to wrote in
2015 that reasons for lower oil prices include “technological advances that are rapidly improving the economics of
solar, wind and battery power.” Really? If so, there would need to be a lot of oil used for electricity generation that
could be displaced by renewable energy, or rapid electric car growth. However, oil used for electricity is just 5%-6%
of global oil production (having declined sharply in the 1980s and 1990s), and EV growth is very slow (see page 7).
8
While Chinese demand for coal and other commodities declined in 2015, Chinese oil demand grew at 5.7% in
2015 and rose 3% y/y in Q1 2016. While heavy industry and construction activity have fallen, car sales are rising.
6
[II] The tortoise, the hare and the electric car: the quiet progress of hybrids over EVs
There’s a lot of excitement about Tesla´s Model 3, Another generation of electric car projections out of
the Chevy Bolt and other electric cars coming to sync with reality, EV+PHEV sales as % of total car sales
market. However, auto analysts have generally 12% Deutsche Bank
been way too optimistic about adoption of electric ● US PwC
10% Frost & Sullivan
vehicles (EVs), as shown in the chart on 2020 ● Global
Bloomberg NEF
projections vs. reality. This isn’t the first time; over- 8%
IEA
optimism on electric cars began in the 1960s with Roland Berger
6%
Ford Motor’s announcement of an imminent BCG
4%
breakthrough on electric car production. Deloitte
2%
With respect to innovation in the car industry, the Actual US EV+PHEV sales
0%
more impactful developments involve the use '09 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 '20
of electricity in hybrid vehicles (the tortoise), Source: DOE, BEA and listed organizations. 2015. Note: global
rather than in EVs (the hare). EV+PHEV sales in 2015 were also around 0.6%.

What’s a hybrid vehicle? Well, it depends…

• Micro hybrid: internal combustion engine (ICE) cars with stop-start systems drawing power from
regenerative braking; the engine shuts off when drivers idle and restarts when the gas is re-engaged
• Mild hybrid: ICE cars with more powerful batteries (e.g., 48V or 110V instead of 12V), also powered by
regenerative braking, which provide engine stop-start functionality while stopping, coasting and
decelerating; and which can boost the engine during acceleration
• Full hybrid: ICE cars with larger electric motors that can power the car for short periods; battery packs
are charged both by regenerative braking and by the ICE itself
• Plug-in hybrid (PHEV): primary engine is usually the electric motor, recharged with an external electricity
supply; ICE mostly used to recharge the electric battery rather than to power the car

The next chart shows sales for US full hybrids, plug-in hybrids and EVs. Despite the excitement,
PHEVs + EVs are still just 0.7% of US light vehicle sales. Full hybrids, on the other hand, have reached
3% of sales. According to the International Council on Clean Transportation (ICCT), full hybrids are an
important part of the climate picture since they can reduce fuel consumption and CO2 emissions by up to
35% 9. In its analysis, ICCT cites automotive research firm Vincentric which compared 31 full hybrids to the
closest non-hybrid vehicle from the same maker. The 2nd chart shows the fuel consumption reduction for
some of these vehicles alongside their incremental cost. The tradeoff can be an expensive one: in only
~60% of full hybrids analyzed, the car’s extra cost is recouped through lower fuel expense over its life. This
suggests that full hybrid costs need to come down further.
US full hybrid, plug-in hybrid & electric vehicle sales Full hybrids: fuel savings and cost vs. non-hybrid
Thousand units sold, with examples of each Fuel consumption reduction Price premium
500 35% $7,000
30% $6,000
400
Full hybrids: Toyota Prius, 25% $5,000
Toyota Camry, Ford Fusion,
300 Hyundai Sonata, Lexus CT 200h 20% $4,000

Electric vehicles: 15% $3,000

200 Tesla Model S, Nissan 10% $2,000
Plug-in hybrids: LEAF, BMW i3, VW
Chevy Volt, Ford Fusion eGolf, Fiat 500e 5% $1,000
100 Energi, Ford C-MAX
Energi, Toyota Prius Plug- 0% $0
in, BMW i8 Toyota Ford Hyund, Honda Nissan VW, BMW, All
0 Kia Audi, Merce, hybrids
'99 '00 '01 '02 '03 '04 '05 '06 '07 '08 '09 '10 '11 '12 '13 '14 '15 Porsc Subaru
Source: US Department of Energy. December 2015. Source: ICCT, Vincentric Hybrid Analysis. July 2015.

9
The 2014 EIA fuel economy report reached a similar conclusion in terms of potential hybrid fuel and GHG savings.
7
The potential for lower hybrid costs and greater market share
“High-power electronics” is a relatively new field, and costs are coming down rapidly. ICCT expects hybrid
system component costs to fall in half by 2025. Manufacturers and suppliers are sorting out advantages
and costs of the different configurations, such as voltage level (from 12V all the way up to 330V), power
density, energy storage (lead-acid, NiMH, Li-ion) and drive type. New 48V mild hybrid systems may turn
out to be the most promising, achieving 1/2 to 2/3 of the benefits of a full hybrid at less than half the cost.
Barclays projects greater acceptance and penetration of mild and full hybrids; the table below shows actual
2015 data and their projections for 2020 and 2025.
The big picture:
• Port fuel injection technology is expected to give way to gasoline direct injection in many ICE cars.
• 48V mild hybrid and full hybrid cars are projected to increase market share. The EPA has a similar
outlook, expecting mild and full hybrids to capture 31% market share in the US by 2025 (mostly
through mild hybrids).
• Combined plug-in hybrid and EV shares are projected to hit just 5% in 2025, way below most of the
2020 forecasts shown on the prior page.
• What about Tesla? In 2015, Tesla delivered 50k units and is projected to deliver 75k in 2016. The
company mentioned 500k as annual capacity for its new US gigafactory by 2020, and mentioned one
million units as its global target by 2020. This seems ambitious, but even if these levels are reached,
Tesla’s EVs would still represent just 1.2% of global vehicle production.
Bottom line: compared to EVs, hybrids are making more impactful but less glamorous use of
electricity, and can be similar to EVs in terms of GHG emissions as well (see next page)

A 2015 projection of future global light vehicle Where does hybrid car energy conservation come
production by powertrain type from? As per the ICCT:
Powertrain 2015 2020 2025 • Using energy normally lost during braking
ICE port fuel injection 50.9% 36.8% 20.8% • Maintaining performance while using a smaller,
ICE gasoline direct injection 24.0% 31.2% 29.4% more efficient engine
ICE diesel 19.7% 19.9% 17.0% • Shutting the engine off at idle and at low load
Compressed natural gas 1.9% 1.9% 1.8% conditions
48 Volt (mild) hybrid 0.0% 3.6% 15.5% • Enabling the engine to be run at lower speeds,
Full hybrid 2.8% 4.4% 10.4% where it is more efficient
Plug-in hybrid (PHEV) 0.4% 1.5% 2.0% • Replacing the alternator as a means of generating
Electric vehicle (EV) 0.2% 0.7% 3.0% power with more efficient motor/generator systems
Source: Barclays, IHS. August 2015. • Replacing less-efficient mechanical water and oil
pumps with electrical pumps that only operate when
needed
• Supplying power required by safety features, heated
seats, dynamic chassis control and other power-
hungry components

8
Sparse penetration aside, how “green” are EVs compared to hybrids?
Energy use and GHG comparisons of hybrids with EVs are very complex. A comprehensive “well-to-
wheels” analysis has to factor in the current electricity mix and how it’s changing over time, energy used
and lost in transmission of electricity and gasoline, energy used on oil extraction and refining, energy used
in the manufacturing of EVs vs. hybrids, battery disposal and other subtleties such as driver habits.
Based on data from the ICCT and a November 2015 paper from the Union of Concerned Scientists, our
sense is that the GHG footprint of a full hybrid like the Prius is roughly the same as an EV in a state with the
average US electricity mix. The chart below shows the electricity mix by country and by US state. Where
coal usage is high, hybrid cars can produce better GHG results than an EV. As penetration of renewable
energy increases, the EV becomes the greener choice.

How Green is My Valley: the CO2 footprint of your electric car is heavily dependent on where you live
% of total electricity generation by source
Renewables (hydro, wind, solar, etc) Nuclear Natural gas Coal and oil
100%

80%

60%

40%

20%

0%
Can Chi Fra Ger Jpn Nth Nor Swe UK US Bra Ind Indo Ita Kor Mex Pol Por Rus Spn Tha Tur AZ CA FL GA IL MA MO NY OH PA TX VA WA
Countries with largest # of EV cars Other select countries Select US states
Source: World Bank (2014) for countries, EIA (2015) for US states, JPMAM.

A few things to keep in mind when looking at this chart:

• Coal vs. natural gas. An EV charged by a grid powered 100% by natural gas has ~50% of the GHG
footprint of an EV charged by a grid powered 100% by coal. Of the countries shown, only the US
experienced a double digit decline in coal's share of generation from 2006 to 2014, mostly due to
expansion of natural gas. From 2006-2014, coal's share of global electricity generation was unchanged.
• Renewable energy mostly hydro. As per the chart on page 1, green renewable energy segments
are mostly made up of hydroelectric power, whose global share of electricity generation was 3.6x
higher than wind and solar power combined in 2015.
• How green is nuclear? Difficult question. The Clean Air Task Force, climatologist James Hansen of
Columbia University, David Mackay of Cambridge, Robert Hargraves from Dartmouth and 65 biologists
who signed an open pro-nuclear letter in 2015 support it. On the other hand, there are disturbing
reports of Fukushima’s aftermath: rain and other sources of water flowing through the site and
becoming radioactive, some of which seeps into the ground or the ocean; signs that cesium and
strontium were still leaking into the ocean as of 2013; and the lack of knowledge as to exactly where
the molten cores from the reactors are located. I don’t know what color to make the nuclear bars in
the chart: green when it works and vantablack (the darkest color in the universe) when it doesn’t?

9
[III] A New York State of Mind: Thoughts on an ambitious new energy plan
In December 2015, New York State announced an ambitious energy plan designed to reduce GHG
emissions by 32% by 2030 10. The plan has several components, but there are two primary ones:
• Cut energy use. By 2030, reduce energy consumption in commercial, residential and industrial
buildings by 23%.
• Decarbonize the grid. By 2030, generate 50% of electricity from renewable sources.
While the first objective is ambitious over a 15-year timetable, it relies on established methods of energy
conservation. But will the second objective be a sentimental journey, or a real one? New York assumes
that an eight- to ten-fold increase in wind and solar will propel it to 50% generation from renewables.
However, New York is not a particularly windy or sunny place, which explains its low wind and solar
penetration to-date. Furthermore, its projections may not adequately reflect wind and solar intermittency,
which may result in substantial back-up thermal capacity needs. This phenomenon is observed in Germany
after its 15-year renewable energy journey. Grid de-carbonization is critical 11 given adverse human health
externalities associated with fossil fuels, but the economic and political levers required to make this vision a
reality, and its associated costs, are still unclear.

New York State goals for wind and solar power by 2030 Despite a large renewable energy build-out in Germany,
Wind plus solar energy as % of total electricity generation almost no reduction in natural gas and coal capacity
30% 100 625
Baseline: assuming 1.1% annual
25% electricity demand growth
80
Thermal capacity 600
20% Conservation: assuming no growth in
(coal & gas), GW
electricity demand, with 2030 demand 60
15%
equal to 2014 levels Electricity
575
consumption, TWh
40
10%
Current wind + solar 550
5% 20 Solar and wind
capacity, GW
0% 0 525
'90 '92 '94 '96 '98 '00 '02 '04 '06 '08 '10 '12 '14 '16 '18 '20 '22 '24 '26 '28 '30 '91 '93 '95 '97 '99 '01 '03 '05 '07 '09 '11 '13
Source: EIA, NYSERDA, JPMAM. December 2015. Source: German Federal Ministry for Economic Affairs and Energy. 2014.

10
New York cites a 40% GHG reduction target, but this is vs. a higher 1990 baseline; the 2030 target is a
32% reduction when compared to 2014 emissions. In addition to renewable sources of electricity energy and
reduced energy consumption in buildings, New York also assumes GHG reductions from changes in transportation:
increased use of plug-in electric vehicles (see pages 7-9); upgrade of less energy-efficient mass transit infrastructure,
such as the NYC subway system; and reduced idling and enhanced traffic flow from sensors, improved traffic
avoidance guidance and synchronized traffic signals.
11
Vaclav, on electricity grid de-carbonization: “Underlying all of the recent moves toward renewable energy is the
conviction that such a transition should be accelerated in order to avoid some of the worst consequences of rapid
anthropogenic global warming. Combustion of fossil fuels is the single largest contributor to man-made emissions of
CO2 which, in turn, is the most important greenhouse gas released by human activities. While our computer models
are not good enough to offer reliable predictions of many possible environmental, health, economic and political
effects of global warming by 2050 (and even less so by 2100), we know that energy transitions are inherently
protracted affairs and hence, acting as risk minimizers, we should proceed with the de-carbonization of our
overwhelmingly carbon-based electricity supply – but we must also appraise the real costs of this shift.”
10
Objective #1: New York’s plan to reduce energy consumption in buildings by 23% by 2030
Conservation is a critical component of a more efficient energy future, and is a primary building block in
most CO2 emission plans. New York’s conservation objectives rely upon established methods (solid-
state/LED lightning, light fixture efficiency, heat pumps, adaptive and optimized thermostats, more efficient
HVAC and insulation systems, reduced stand-by loads and smart plugs, remote appliance control, etc).
Conservation has a good track record: since 1980, US energy use per sq ft has fallen by 21% and 11% in
commercial and residential buildings. In Denmark, a conservation leader, energy consumption in buildings
has fallen by 45% per sq meter since 1975. As shown in the first chart, energy intensity has declined in
many regions, with conservation gains in buildings and machinery playing a large role.

Examples of declining primary energy intensity NYSERDA electricity use scenarios

Index, 1997=100, BTUs of energy per unit of real GDP New York electricity sales, TWh, historical and projections to 2030
200
100
190 Forecast without conservation
India China
95 180
90 170
160 Historical data
85 through 2013 Conservation plan
Europe 150 incorporating market and
80 140 economic barriers
75 UK 130
US 120
70 All conservation projects
110 implemented
65 100
60 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030
1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Source: NYSERDA, JPMAM. April 2014. Pre-1990 electricity sales
Source: National statistics offices, BP Stat. Review of World Energy. 2015. extrapolated from published load data.

While New York’s energy conservation methods are well-established, the

Declines in TWh of
magnitude of achievable savings is not easy to quantify. The New York State
electricity consumption,
Energy Research and Development Authority (NYSERDA) is tasked with coming up 2015 vs. prior peak
with plans to reach the state’s goals. NYSERDA reports released in 2014 contain two
scenarios for energy conservation in buildings. In the first, all projects with positive Denmark -34%
Belgium -21%
economics are implemented, while in the second, market barriers to implementation
Finland -19%
and extra costs are incorporated. NYSERDA estimates that energy consumption could
UK -16%
decline by 29% in the first case, and by 11% in the second. Italy -12%
The first case is ambitious; let’s look at electricity, which accounts for 40% of total Netherlands -11%
projected energy savings in buildings 12. As shown above (right), in the first case, New Sweden -5%
York’s electricity usage would fall by 35% and decline below 1980 levels, while in the Germany -4%
second, 2030 usage would be roughly the same as in 2013. Double-digit electricity Australia -4%
France -3%
usage declines have occurred in some European countries (see table), so while New
Norway -2%
York’s plans are very ambitious, they’re not totally unprecedented.
Canada -2%
All things considered, reducing energy consumption in buildings by 10%-15% seems US -1%
quite achievable, while a 23% reduction would take a more aggressive action plan Source: BP Statistical Review of
World Energy. 2015.
with explicit incentives for businesses and households to make the necessary
investments in energy-saving technology.

12
Reduced natural gas use accounts for another 40% of building energy savings in NYSERDA’s plan. The remainder is
from reduced petroleum use.
11
Objective #2: a large expansion in New York’s wind and solar power
New York is not a particularly windy or sunny place. New York ranks 31st out of 35 states based on
its wind capacity factors (a measure of electricity generation efficiency described below 13), and with respect
to solar, New York ranks 48th out of 50 states based on capacity factors from the National Renewable
Energy Laboratory. Wind and solar only play a small role in New York electricity generation (around 3%),
so a plan relying heavily on their expansion is worth examining. New York is also an interesting bellwether:
its solar irradiance is similar to China, Japan, Russia, the UK, Germany and the Philippines, and its wind
speeds are similar to India, Brazil and large parts of Central Europe.

Wind and solar: just 3% of NY electricity generation New York ranks near the bottom in terms of electricity
NY state electricity mix, 1990 - 2015, % of total generation generation efficiency from wind & solar, capacity factor %
100% 45% 27%
90% Oil
40% Wind capacity 25%
80% Coal factor
70% Nuclear 35% 23%
60% New York,
30% #31 of 35 21%
50% Natural
40% gas 25% 19%
Solar capacity New York,
30% Biomass/other
20% factor #48 of 50 17%
20%
10% Hydroelectric Solar/wind 15% 15%
0% 1 5 9 13 17 21 25 29 33 37 41 45 49
'90 '92 '94 '96 '98 '00 '02 '04 '06 '08 '10 '12 '14 State rank based on capacity factor, 1= highest
Source: Energy Information Administration. December 2015. Source: Wind Action Group, NYSERDA, NREL. April 2014.

The next chart shows NYSERDA’s renewable energy projections for 2030 alongside New York’s current
renewable electricity mix. NYSERDA examines two different cases: a “Technically Feasible” case which
assumes all projects are built irrespective of their cost, and an “Economically Viable” case which assumes
projects are only built if they make economic sense. The latter benchmarks wind and solar against the
most expensive form of generation (typically natural gas peaker plants), and results in 8x-10x growth in
wind and solar power by 2030. If these generation goals were achieved, New York would make good
14
progress, particularly if energy conservation reduces demand .

Share of New York State electricity generation from

renewable sources, Current and 2030 NYSERDA projections NYSERDA renewable energy scenarios for 2030:
Biomass/other
70% • Technically Feasible case relies heavily on
60% Hydro offshore as well as onshore wind, and a
50%
combination of utility-scale, residential and
Offshore wind commercial solar power
40%
Biomass/other
• Economically Viable case mostly relies on
30% Onshore wind
Biomass/other onshore wind, and utility-scale and large
20% Hydro commercial solar power
10%
Hydro Solar
Onshore wind
• In both cases, share of generation from
Onshore wind Solar biomass, hydroelectric, etc remain at current
0%
Current (Dec. 2015) Technically Feasible Economically Viable
levels
Source: NYSERDA, EIA, JPMAM. April 2014.

13
Capacity factors measure actual production relative to theoretical production (generation at full-power over the
entire period). For example, assume a 1 MW wind turbine with potential output of 24 MWh in one day; if the turbine
generated 8 MWh of electricity, its capacity factor would be 33%.
14
In its 2014 renewable energy projections, NYSERDA assumed annual electricity demand growth of 1.1% to 2030,
rather than incorporating the impact of conservation goals stated elsewhere in the plan. If electricity loads in 2030
were flat vs. today, the Economically Viable case would imply 37% renewable generation rather than 30%.
12
Our take on NYSERDA’s assumptions. While there are over 500 pages of NYSERDA documents, wind
and solar assumptions matter most since NYSERDA’s 2014 analysis did not assume higher shares from
biomass, biogas, hydropower, solar thermal, wave energy, etc. For the most part, the Economically Viable
case assumptions regarding wind and solar seem reasonable to us:
• Onshore wind capacity factors. New York has 1.8 GW of onshore wind whose 2013 capacity factors
were 23%-25%. Most plants were built from 2006 to 2009, and are located in New York’s windiest
counties (map). NYSERDA assumes capacity factors of 33% to 40% on future installations, levels
typically observed in much windier states (see US map). However, two recent installations in New York
show capacity factors in the mid 30s. As new GW of wind are added, the state average should rise,
although 40% by 2030 seems like a stretch to us. Challenges include community receptivity 15 and
degradation in capacity factors over time 16, a trend which has adversely affected wind investor returns.
• Upfront capital costs for onshore wind and solar. NYSERDA estimates of upfront capital costs for
onshore wind ($2,380 per kW) and utility-scale solar ($1,720 per kW) look reasonable and are maybe
too high, particularly for onshore wind where we have seen lower estimates of $1,600 to $2,000 per
kW. As we discussed last year, learning curves show continued unit cost declines for wind, solar and
battery storage as production volumes increase.
• Capacity factors for utility-scale, commercial and residential solar PV. NYSERDA assumptions of 14%-
15% for all 3 types are reasonable, although commercial/residential capacity factors are often 2%-3%
lower. NYSERDA estimates are consistent with NREL regarding areas with NY’s solar irradiance.
• Offshore wind. Unsurprisingly, offshore wind’s projected share falls by 90% in the Economically Viable
case. While offshore wind speeds are generally higher than onshore, many estimates of their costs
(including grid interconnection) more than offset this benefit. A 2014 paper from the University of
Sussex determined that offshore wind is the most expensive commercially available renewable energy
source in the UK, and that rather than compelling economics, its development reflects the
alignment of political and financial interests of its advocates 17. Papers like these are useful since they
reflect the experience of countries with actual offshore wind installations; the US currently has none.
• Distributed solar PV. In the Economically Viable case, residential and small commercial solar PV
disappears due to higher costs. What remains is a projected 10 GW build-out of large commercial and
utility-scale solar PV. To-date, most utility-scale solar PV has been built in states that are much sunnier
than New York. However, we believe that NYSERDA solar capital cost, capacity factor and levelized
cost assumptions of 11 to 12 cents per kWh are reasonable (the latter 1-2 cents too low at most).

15
There’s local opposition in the air, but it may not have an impact given passage of anti-NIMBY legislation.
Three New York counties (Erie, Orleans and Niagara) passed resolutions opposing a proposed 200 MW project that
entails 500-600 foot wind turbines across 20,000 acres. However, in 2011 Governor Cuomo signed Article 10 into
law which allows the State to provide necessary approvals in a unified proceeding for clean energy projects, and which
does not appear to require approvals from local communities.
16
Examples of wind capacity factor shortfalls and degradation. A study by the Renewable Energy Foundation
on UK and Danish wind farms found that capacity factors fell as wind farms aged, from 24% to 15% in year 10 in the
UK, and from 22% to 18% at age 15 in Denmark. A similar study from the Imperial College of London found that
wind farm output declines by 1.6% per year as turbines age. In 2012, when S&P downgraded several single-asset
wind projects below investment-grade, S&P cited capacity factors that were below what industry experts had cited as
lower-bound estimates, as well as operating costs that were 30% to 40% higher than original forecasts.
17
What explains the development of offshore wind? “Our analysis demonstrates how the close alignment of
economic and political interests of key actors within the specific context of the UK has led to the rapid deployment of
offshore wind – by circumventing anti-onshore wind protest in the short term and meeting 2020 renewables targets
in the medium term but at potentially high economic and political costs when the further deployment of offshore
wind adds up to a significant impact on electricity bills”. Source: From laggard to leader: Explaining offshore wind
developments in the UK, Kern et al, University of Sussex, February 2014.
13
While NYSERDA’s individual assumptions seem reasonable, there’s another important question to answer:
what happens when there’s not enough wind or solar irradiance to meet electricity demand?
The issue of wind and solar intermittency: the need for back-up thermal generation capacity
Costs for wind and solar power cannot simply be compared to thermal sources without accounting for
wind and solar intermittency, since back-up thermal capacity is typically needed for periods when there’s
not enough wind or solar power to meet demand. NYSERDA is aware of this issue and is not assuming
that renewable capacity immediately displaces thermal capacity on a 1:1 basis. Based on our reading of
the documents, NYSERDA assumes that thermal capacity is replaced at a constant rate that is less than 1:1.
Even so, this may be optimistic, since existing research shows a diminishing (non-constant) ability to shutter
thermal capacity as more wind and solar power are added to the grid18.
However, our even bigger concern is that everyone’s estimates of how much thermal capacity
19
can be shuttered may be too high. Consider the next 2 charts from our 2015 energy paper . Working
with the Clean Air Task Force, we examined hourly generation and load patterns in both Germany and
California. Even though Germany’s planned build-out of wind and solar (Energiewende) would result in
80% of electricity coming from renewable sources over the course of the year, we estimated that its back-
up thermal capacity needs would be practically unchanged vs. current levels, given periods of low wind and
solar power during winter months. The same outcome occurred in our analysis of California. Results were
not very different when assuming energy storage (via batteries, pumped storage or hydrogen fuel cells),
“demand management” or geographical grid expansion. Unfortunately for New York, its winter solar
irradiance is similar to Germany, and its winter wind is like California’s (i.e., the worst of both).
Even if wind and solar capacity were expanded to generate 80% of electricity from renewable energy over the course
of the year, in winter months, there would be extended periods when extensive back-up thermal capacity is needed

Germany: January load vs. renewable generation California: January load vs. renewable generation
Hourly generation by source with load, gigawatts Hourly generation by source with load, gigawatts
90 35
Solar Solar
80 Periods of unmet demand, filled by 30
Load
natural gas and coal generation
70
Load 25
60
50 20

40 15
30
10
20
Wind 5 Wind
10
Hydro/biomass Hydro/biogas/biomass/geothermal
0 0
Jan 1 Jan 6 Jan 11 Jan 16 Jan 21 Jan 26 Jan 31 Jan 1 Jan 6 Jan 11 Jan 16 Jan 21 Jan 26 Jan 31
Source: Germany grid operators, JPMAM. 2015. Source: CAISO, JPMAM. 2015.

18
Electricity grid analysts refer to “capacity credits”: the amount of thermal capacity (natural gas and coal) no longer
needed once wind and solar are added to the grid. The Lawrence Berkeley National Laboratory compiled various
studies that all projected sharply falling capacity credits as wind and solar penetration increases. They are not alone; a
2015 paper from the Potsdam Institute for Climate Impact Research notes that integration costs in systems with high
levels of renewable energy can be up to 50% of generation costs, and that the largest factor is the cost of backup
thermal power. Hence the problem with analyzing renewable energy simply based on stand-alone levelized
costs per kWh. Lazard and Bloomberg New Energy Finance are well-known sources for levelized costs per kWh by
generation source; they are usually cited in media reports on renewable energy without caveats on backup power.
19
See Brave New World, Annual Energy Eye on the Market, October 19, 2015.
14
Here’s some real-world evidence: despite a large buildup of solar and wind capacity at a time of
stable electricity consumption, Germany’s thermal capacity is almost unchanged and is still very
20
actively used . Thermal generation and CO2 emissions have declined, but not thermal capacity. This may
partly explain why Germany and Denmark, countries with the highest renewable capacity per capita in
Europe, also have the highest residential electricity prices at around 0.30 Euro cents per kWh.
Despite a large renewable energy build-out in Germany, The benefits: a 15% reduction in German thermal
almost no reduction in natural gas and coal capacity generation, Share of total electricity generation
100 625 70%

60%
80
Thermal capacity 600
50% Thermal (coal & gas)
(coal & gas), GW
60
Electricity 40%
575
consumption, TWh
40 30%

550 20% Solar and wind

20 Solar and wind
capacity, GW 10%
0 525 0%
'91 '93 '95 '97 '99 '01 '03 '05 '07 '09 '11 '13 '91 '93 '95 '97 '99 '01 '03 '05 '07 '09 '11 '13 '15
Source: German Federal Ministry for Economic Affairs and Energy. 2014. Source: German Federal Ministry for Economic Affairs and Energy. 2015.

All things considered:

• New York’s 50% renewable generation target is ambitious, particularly in just 15 years
• If Germany is any guide, New York’s back-up thermal capacity needs may be higher than those
assumed by NYSERDA
• Even though New York is relying on market forces to bring this transition about, taxes/subsidies and
electricity prices will almost surely end up playing a role. Even before New York’s energy transition
th
begins, its electricity prices are the 8 highest in the US, and its state taxes are the highest in the US
• While renewable energy growth can drive electricity grid de-carbonization, its power density and
intermittency dynamics suggest that some hard choices involving taxes, spending and electricity prices
may eventually have to be made.

New York: 8th highest electricity price in the US New York: highest individual tax rates in the US
Electricity price, USD/kWh, all consumers Effective tax rate for the middle 20% of taxpayers
$0.18 13%
$0.17 12% New York: #1 of 50
$0.16 New York: #8 of 50 11%
$0.15
10%
$0.14
$0.13 9%
$0.12 8%
$0.11 7%
$0.10
6%
$0.09 Note: effective tax rate shown after benefit of federal
Note: highest state (Hawaii) 5% tax deduction; includes state and local income taxes,
$0.08
at $0.26 per kWh property taxes and sales taxes.
$0.07 4%
1 5 9 13 17 21 25 29 33 37 41 45 49 1 5 9 13 17 21 25 29 33 37 41 45 49
State rank based on electricity price, 1 = highest State rank based on effective tax rate, 1 = highest
Source: Energy Information Administration. 2015. Source: Institute on Tax and Economic Policy. 2015.

20
Coal still accounts for 44% of electricity generation in Germany, and as shown in the linked chart, coal plants are
still very actively used. In 2014, brown coal plant capacity factors were over 75%, close to their highest level in 25
years. Black coal plant capacity factors have fallen from their historical average of 50% to around 40% in 2014.
15
[IV] The Utility Empire Strikes Back: Distributed solar power and billing changes
Several factors have contributed to growth in US distributed residential and commercial solar capacity from
less than 1 GW in 2010 to over 10 GW in 2015 (1st chart):
• Declining costs of distributed solar power (2nd chart), and the availability of third party ownership/
leasing which allows customers to put less money down upfront
• The prevalence (until very recently) of most utilities compensating solar PV owners for electricity sold
back to the grid at the customer’s retail rate, a practice known as “net metering”
• The practice by most utilities to recover the majority of costs through volumetric (per kWh) charges
instead of fixed dollar charges, magnifying potential savings for PV customers and undercharging them
for transmission infrastructure
• Attractive Federal and state subsidies for solar PV purchasers
Total installed distributed solar PV Upfront capital costs for distributed PV
GW of residential and commercial capacity USD per kW-DC, residential and commercial systems
12 $7,000
Balance of System
10 $6,000
Inverter

Residential
$5,000 Module
8

Comm.
$4,000
6
$3,000
4
$2,000
2 $1,000

0 $0
2010 2011 2012 2013 2014 2015E 2009 2010 2011 2012 2013 2015
Source: Solar Energy Industries Association. 2015. Source: National Renewable Energy Laboratory. August 2015.

Growth in distributed solar does not come without complications: how will utilities deal with the portion of
transmission and infrastructure costs that PV customers no longer pay for? If utilities simply raise prices per
kWh on everyone, some believe it could make distributed solar more attractive, leading to greater solar
adoption and continually collapsing utility revenues (the so-called “death spiral”). There are a lot of
debates about how realistic a death spiral really is 21. Nevertheless, there are several options utilities are
considering to prevent such a spiral from ever happening:
• Reallocating cost recovery from volumetric charges to fixed dollar charges
• Time-varying electricity rates, a system in which customers buy and sell electricity based on its
instantaneous price at that point in the day. The implication: solar power sold in the middle of the day
could be much less valuable since it would be more abundant
• Partial net metering: customers use what they produce, and sell their surplus electricity back to the
grid at a price that’s lower than the retail rate at which they buy electricity from the grid
• Lower feed-in tariffs: this system effectively assumes that PV users are in the distribution business.
All the power they produce is sold to the grid (not just their surplus generation) at a lower rate. Then,
they have to purchase all power they consume at a higher rate. As a result, lower feed-in tariffs are less
rewarding for PV customers than partial net metering

21
Death spiral? LBNL notes that there have been no studies to-date conclusively showing death-spiral dynamics in
motion. They believe a utility death spiral is unlikely, since feedback effects operate in opposing directions: higher
costs per kWh improve solar PV economics for residential customers, but worsen them for commercial PV customers
on time-varying rates since higher PV adoption by residential customers increases electricity supply. The former is
found to increase PV adoption by 2050 by 8%, while the latter reduces deployment by 5%. These are national
estimates; effects for individual utilities could vary greatly, depending on their customer mix.
16
This not an abstract discussion. In California, Arizona, Nevada and Hawaii where 54% of US
distributed solar capacity has been installed to-date, these changes are already happening:
• Arizona. In November 2013, the Arizona Corporation Commission (the regulatory body overseeing
privately owned utilities in Arizona), voted to install a $5 monthly fixed charge on new residential solar
PV users. Utilities governed by the ACC serve 2 million customers in Arizona (~1/3 of the population).
Other smaller privately owned utilities in Arizona have sought regulatory approval to lower net
metering rates from retail to wholesale rates, and to increase fixed charges.
• Arizona. The net metering debate is not confined to privately owned utilities. The Salt River Project, a
public utility serving 1 million customers in Arizona, changed pricing for new solar PV users to be based
on time-varying rates, and began applying monthly fixed charges and demand surcharges (extra fees
based on periods of maximum customer demand).
• California. In January 2016, the California Public Utilities Commission (CPUC) voted to apply a one-
time interconnection fee to all solar customers, to apply time-varying rates to all customers, and to
accelerate them for solar PV customers. The three largest utilities in California who serve ~24 million
customers had been arguing for a lower feed-in tariff instead, which would have been more negative
for solar users. Our sense is that none of the newly adopted approaches are written in stone, and are
subject to change; California has agreed to revisit the issue in 2019.
• Nevada. In February 2016, the Nevada Public Utilities Commission finalized new rules on all solar PV
customers (not just new adopters) that increase fixed monthly charges on net-metering customers by 2x
- 3x over the next 12 years, decrease net metering credits for surplus generation by 60% - 70%,
decrease volumetric charges by 10% - 30%, and give net-metering customers the option to move to
time-varying rates. This is a very complex billing system; from our perspective, this new approach is the
most economically negative for solar PV customers since they include both fixed charges and lower
net metering rates. The NPUC regulates NV Energy which serves 1.3 million customers, or around half
of the state’s population.
• Hawaii. Hawaii has the greatest amount of distributed solar per capita. In October 2015, the Hawaii
Public Utility Commission lowered pricing on surplus PV generation to 15 - 28 cents per kWh vs. an
average retail rate of 38 cents per kWh.

States with high distributed solar PV and net-metering

changes, Distributed solar MW per 100,000 people
28
States that have enacted net metering changes
26
10 ~~
8

0
HI AZ NJ CA VT MA CO NV DE MD CT NM
Source: EIA, US Census Bureau. January 2016. States shown are those
with more than 3 MW per 100,000 people.

17
How might such utility responses affect the rate of solar PV adoption? To answer that question, we
turn to a 2015 analysis from Lawrence Berkeley National Laboratory on distributed solar power growth 22.
The chart below shows their results, which we interpret as follows:
• If “death spiral” utility responses are confined to small monthly fixed charges or a shift to time-varying
rates, the projected impact on overall PV adoption is not radically different from a status-quo base case.
Note that the time-varying adoption curve flattens out, since the greater the abundance of solar PV, the
lower mid-day electricity rates would be for PV customers selling back to the grid.
• However, if more fundamental changes to customer billing were to take place (larger fixed monthly
charges and lower feed-in tariffs), distributed solar adoption rates could be materially affected.
• What are the early signs? Mixed. In California, while time-varying rates adopted by the CPUC
should only have a modest impact, ongoing utility counter-proposals would dampen solar PV incentives
further. In Arizona, the outcome is modest so far with a $5 charge, while in Nevada, the outcome was
much more negative for solar PV economics, leading to large solar industry layoffs in the state.
National distributed solar PV deployment by scenario
GW
160
Base case: continuation of the status quo
140 $10 increase in fixed monthly charge for residential customers only
120 All residential and commercial customers on time-varying rates
Partial net metering (see description on page 16)
100
80
60 $50 increase in fixed monthly charge for residential customers only
40
Low feed-in tariff at $0.07/kWh
20
0
2015 2020 2025 2030 2035 2040 2045 2050
Source: Lawrence Berkeley National Laboratory. July 2015.

Below, we show each scenario in 2040: distributed solar capacity, its electricity generation assuming a 15%
23
capacity factor and its share of all electricity generation. Even in the base case, distributed solar’s
2040 share would be less than what wind already contributes today.
Implied 2040 distributed solar PV under various scenarios
Implied Implied Share of 2040
2040 2040 projected
Capacity Generation electricity
(GW) (GWh) generation
Base case 139 182,560 3.7%
$10 fixed charge 118 155,208 3.1%
Time-varying rates 115 150,925 3.0%
Partial net-metering 93 122,585 2.5%
$50 fixed charge 52 68,937 1.4%
Low feed-in tariff 32 41,850 0.8%
Source: Law rence Berkeley National Laboratory, EIA, JPMAM. July 2015.
Note: assumes 15% capacity factor.

22
LBNL expanded on an NREL model that simulates customer adoption of distributed PV. The model uses a bottoms-
up approach where customer adoption depends on a comparison of PV system costs with reductions in customer
electricity bills, using data from 216 solar resource regions and more than 2,000 electric utilities.
23
Current national weighted average rooftop solar capacity factors are 15%, using NREL data. While there
have been large declines in the price of solar modules, capacity factor improvements have been more limited.
18
[V] Dam it: A look at the modest potential for more hydropower in the US
Hydropower is a reliable, cheap form of energy with little to no CO2 impact from ongoing operations. In
some countries, hydropower represents 50%+ of total electricity generation: Iceland, Norway, Switzerland,
Canada, Brazil, New Zealand, Colombia and Austria. In the US, however, hydroelectric power accounts for
just 6% of generation, a lower share than in 1990. The US has plenty of rivers, but hydroelectric power is
heavily reliant on topography: the mass of flowing water and the gradient of the river. Mountain rivers
have good hydropower potential, while rivers like the Mississippi do not. In this section, we look at a
realistic case: US hydropower could rise from 6% to 9% of total US electricity generation.

Continental US electricity generation by source How a reservoir impoundment dam works

% of total generation
100% Oil
Headgate Electricity
90%
80%
70% Coal Water
intake Transformer
60%
Dam
50%
40%
Nuclear Generator
30% Biomass/other Penstock
20% Natural gas Turbine
10%
Hydroelectric Wind/solar
0%
'90 '92 '94 '96 '98 '00 '02 '04 '06 '08 '10 '12 '14
Source: Energy Information Administration. December 2015. Source: Xeneca Power Development Inc.

There are two primary kinds of hydropower, one based on storage and the other on diversion:
• Reservoir impoundment dams create a reservoir of stored water. The water flows through a channel
(called a penstock) and spins a turbine, which in turn powers a generator. The penstock inlet is
normally as high as the lowest likely water level in order to maximize the height of the water flow
(“generating head”). The amount of energy a dam can create is the product of the mass of water
flowing through the turbine, the generating head and a gravitational constant.
• Run-of-river facilities divert a river from its natural course into a channel. The water in the channel
flows past a turbine, powers a generator and is then returned back into the river.

Could the US increase its hydropower footprint? The National Hydropower Asset Assessment Program at
Oak Ridge National Laboratory (ORNL) examined the potential for US hydroelectric power. They looked
separately at existing non-powered dams and at new stream development.
The potential from existing non-powered dams. ORNL looked at over 50,000 non-powered dams and
assessed their potential for electricity generation based on head-heights, average annual flow, seasonal
weather patterns and run-off. ORNL assumed 85% efficiency (the % of potential energy converted into
electricity), and excluded dams that were too low and streams with minimal flow. It’s a work in progress
with a few assessments for Ohio, West Virginia and Kentucky still pending, but most of the work is done.
ORNL estimates that the US could build another 5.6 GW of hydropower on existing non-powered dams,
generating 32 TWh of electricity annually. This would help the US renewable energy transition, but only on
the margin, representing just 1% of total US electricity generation in 2015 (4,071 TWh).
The linked map shows where these sites are located, superimposed on maps of US wind speeds and solar
irradiance. For the most part, non-powered dam sites are located in areas without the best wind and solar
resources, which could help such states meet higher renewable energy targets.

19
The potential from new stream development. ORNL scanned for suitable streams and rivers on which
dams have not yet been constructed. This is a more complex process, involving physical, environmental and
political obstacles that existing dam projects have already dealt with.
ORNL used the flood high water mark from the last 100 years as a measure of the potential head height (by
definition, a best-of-all-possible-worlds assumption) when estimating reservoir height and volume. ORNL
excluded national parks, scenic rivers and wilderness areas; rivers with insufficient water flow; and areas
with project incompatibility issues, environmental concerns or existing hydropower facilities. Their results:
61 GW of new potential hydropower, capable of generating 315 TWh of electricity annually. The linked
map shows where ORNL sees the greatest potential for new stream development; sites are generally
concentrated in mountainous regions.
The chart below combines ORNL estimates for existing non-powered dams and new streams. In ORNL’s
theoretical case, hydropower could increase from 6% of US electricity generation to 15%. However, ORNL
conducted their assessment of more than 3 million undeveloped streams at a “reconnaissance level” which
does not incorporate all issues arising from environmental impacts, local politics or cost. They see their
paper as a rough road map for future study. As a result, we added a second scenario in which new stream
development is 30% of ORNL’s estimate, and in which non-powered existing dam development is 75%. If
so, hydropower would rise from 6% to 9% of US electricity generation 24. This would be a welcome
addition, but wind and solar would still represent the bulk of the US renewable effort.

Hydroelectric power: current vs. potential generation Current and potential electricity share from hydropower
Share of continental US electricity generation
Country Current Potential Country Current Potential Country Current Potential
16% Non-powered
dams China 19% 50% Mexico 10% 21% Ukraine 4% 14%
14%
U.S. 6% 15% Italy 16% 33% Nether. 0% 0%
12%
New stream Non-powered India 10% 43% Spain 10% 23% Venez. 60% >100%
10%
development dams Russia 16% 96% Australia 5% 17% Vietnam 39% >100%
8% New stream
development Canada 61% >100% Turkey 26% 80% Belgium 0% 0%
6%
Germany 3% 6% Thailand 2% 11% Kazakh. 9% 41%
4%
Existing hydro Existing hydro Existing hydro Brazil 62% >100% S. Africa 0% 2% Sweden 44% 97%
2% France 9% 22% Poland 1% 5% Uzbek. 19% 45%
0% Indo. 7% 24% Argent. 29% 81% Norway 95% >100%
Current (Dec. 2015) Theoretical 75% NPD, 30% NSD
Source: World Energy Council, International Hydropow er Association, BP. Countries selected
& Current
represent largest overall energy consuming countries in 2015 for w hom the WEC projected
Source: EIA, ORNL, JPMAM. April 2014. hydropow er potential, sorted by 2015 consumption. Potential includes current generation.

The last table shows a cross-country comparison of current and potential hydroelectric generation from the
World Energy Council and International Hydropower Association. Like the ORNL analysis, the potential
amounts should be considered upper bounds given environmental, cost and local issues involved. The US
is at the low end of the range; the comparison sends a similar signal to ORNL with respect to the
modest incremental potential for US hydropower.

24
Separately, Idaho National Laboratory generated “theoretical” and “realistic” US hydroelectric power scenarios. The
two scenarios cite a doubling in generation and an increase of 50%+, respectively, which are similar to the two
projections shown in the chart above.
20
Bibliography
2014 Wind Technologies Market Report, Lawrence Berkeley National Laboratory, August 2015.
2015 Key World Energy Statistics, International Energy Agency, 2015.
A Brave New World: Deep De-Carbonization of Electricity Grids, Eye on the Market Special Edition, J.P. Morgan Asset
Management, October 2015.
An Assessment of Energy Potential at Non-Powered Dams, Oak Ridge National Laboratory, April 2012.
Calculating Electric Drive Vehicle Greenhouse Gas Emissions, International Council on Clean Transportation, Aug 2012.
Cleaner Cars from Cradle to Grave – How Electric Cars Beat Gasoline Cars on Lifetime Global Warming Emissions,
Union of Concerned Scientists, November 2015.
An Evaluation of Solar Valuation Methods Used in Utility Planning and Procurement Processes, Andrew Mills and Ryan
Wiser, Lawrence Berkeley National Laboratory, December 2012.
The Danish Energy Model – Innovative, Efficient and Sustainable, Danish Energy Agency, 2014.
Energy efficiency: A compelling global resource, McKinsey, March 2010.
Energy Efficiency and Renewable Energy Potential Study of New York State, Optimal Energy, ACEEE, Vermont Energy
Investment Corporation, NYSERDA, Volumes 1-5, April 2014.
The Energy to Lead: 2015 New York State Energy Plan, NYSERDA, June 2015.
Energy Transitions, Vaclav Smil, (Prager, in print).
From laggard to leader: Explaining offshore wind developments in the UK, Kern et al, University of Sussex, Science
and Technology Policy Research, February 2014.
Future Powertrain: From sail to steam, Barclays Equity Research, Global Autos & Auto Parts, August 2015.
Global Trends in Renewable Energy Investment 2016, Frankfurt School-UNEP Centre/BNEF, 2016.
Hybrid Vehicles: Technology Development and Cost Reduction, John German, International Council on Clean
Transportation, July 2015.
Integration Costs Revisited – An Economic Framework for Wind and Solar Variability, Hirth, Ueckerdt and Edenhofer,
Potsdam Institute for Climate Impact Research, Renewable Energy 74, 2015.
Net metering and market feedback loops: exploring the impact of retail rate design on distributed PV deployment,
Naϊm Dargouth, Ryan Wiser, Galen Barbose & Andrew Mills, Lawrence Berkeley National Laboratory, July 2015.
New Stream-reach Development: A Comprehensive Assessment of Hydropower Energy Potential in the United States,
Oak Ridge National Laboratory, April 2014.
New York State Greenhouse Gas Inventory and Forecast: Inventory 1990-2011 and Forecast 2012-2014, New York
State Energy Research and Development Authority, June 2015.
Perils of Long-Range Energy Forecasting: Reflections on Looking Far Ahead, Vaclav Smil, Technological Forecasting
and Social Change, Volume 65, 2000.
Survey of Energy Resources, World Energy Council, 2010.
Thoughts on the Oil Investment Cycle, Bridgewater Daily Observations, March 31, 2016.
Utility-Scale Solar 2014: An Empirical Analysis of Project Cost, Performance, and Pricing Trends in the United States,
Lawrence Berkeley National Laboratory, September 2015.

Acronyms
BNEF: Bloomberg New Energy Finance; EIA: Energy Information Administration; EV: Electric vehicle; GHG: Greenhouse
gases; GW: Gigawatt; GWh: Gigawatt-hour; HVAC: Heating, ventilation and air conditioning; ICCT: International
Council on Clean Transportation; ICE: Internal combustion engine; IEA: International Energy Agency; kWh: Kilowatt-
hour; kW-DC: Kilowatt of direct current; LBNL: Lawrence Berkeley National Laboratory; LED: Light emitting diode; Li-
ion: Lithium-ion; MW: Megawatt; NIMBY: Not in my backyard; NiMH: Nickel-metal hydride; NPD: Non-powered dams;
NREL: National Renewable Energy Laboratory; NSD: New stream development; NYSERDA: New York State Energy
Research & Development Authority; OECD: Organization for Economic Cooperation & Development; OPEC: Organization
of the Oil Exporting Countries; ORNL: Oak Ridge National Laboratory; PHEV: Plug-in hybrid electric vehicle; PV:
Photovoltaic; SRI: Stanford Research Institute; TWh: Terawatt-hour; WTI: West Texas Intermediate

21
NOT FOR RETAIL DISTRIBUTION: This communication has been prepared exclusively for
institutional/wholesale/professional clients and qualified investors only as defined by local laws and
regulations.

The views contained herein are not to be taken as an advice or a recommendation to buy or sell any investment in any jurisdiction,
nor is it a commitment from J.P. Morgan Asset Management or any of its subsidiaries to participate in any of the transactions
mentioned herein. Any forecasts, figures, opinions or investment techniques and strategies set out are for information purposes only,
based on certain assumptions and current market conditions and are subject to change without prior notice. All information
presented herein is considered to be accurate at the time of writing, but no warranty of accuracy is given and no liability in respect of
any error or omission is accepted. This material does not contain sufficient information to support an investment decision and it
should not be relied upon by you in evaluating the merits of investing in any securities or products. In addition, users should make an
independent assessment of the legal, regulatory, tax, credit, and accounting implications and determine, together with their own
professional advisers, if any investment mentioned herein is believed to be suitable to their personal goals. Investors should ensure
that they obtain all available relevant information before making any investment. It should be noted that investment involves risks,
the value of investments and the income from them may fluctuate in accordance with market conditions and taxation agreements
and investors may not get back the full amount invested. Both past performance and yield may not be a reliable guide to future
performance.

J.P. Morgan Asset Management is the brand for the asset management business of JPMorgan Chase & Co. and its affiliates
worldwide. This communication is issued by the following entities: in the United Kingdom by JPMorgan Asset Management (UK)
Limited, which is authorized and regulated by the Financial Conduct Authority; in other EU jurisdictions by JPMorgan Asset
Management (Europe) S.à r.l.; in Hong Kong by JF Asset Management Limited, or JPMorgan Funds (Asia) Limited, or JPMorgan Asset
Management Real Assets (Asia) Limited; in India by JPMorgan Asset Management India Private Limited; in Singapore by JPMorgan
Asset Management (Singapore) Limited, or JPMorgan Asset Management Real Assets (Singapore) Pte Ltd; in Taiwan by JPMorgan
Asset Management (Taiwan) Limited; in Japan by JPMorgan Asset Management (Japan) Limited which is a member of the Investment
Trusts Association, Japan, the Japan Investment Advisers Association, Type II Financial Instruments Firms Association and the Japan
Securities Dealers Association and is regulated by the Financial Services Agency (registration number “Kanto Local Finance Bureau
(Financial Instruments Firm) No. 330”);; in Australia to wholesale clients only as defined in section 761A and 761G of the
Corporations Act 2001 (Cth) by JPMorgan Asset Management (Australia) Limited (ABN 55143832080) (AFSL 376919); in Brazil by
Banco J.P. Morgan S.A.; in Canada by JPMorgan Asset Management (Canada) Inc., and in the United States by JPMorgan Distribution
Services Inc. and J.P. Morgan Institutional Investments, Inc., both members of FINRA/SIPC.; and J.P. Morgan Investment Management
Inc.

Copyright 2016 JPMorgan Chase & Co. All rights reserved.

22