convergent energy disruptions...power price range wind ppas utility-scale solar ppas levelized 2014...

Amory B. LovinsCofounder and Chief Scientist

© 2016 Rocky Mountain Institute

Aspen Institute Energy Policy ForumAspen, 5 July 2016

RO

CKY MOUNTAIN

INSTIT UTE

WAR R O O M

CARBON

Convergent energy disruptions

[14 minutes] [Presenter’s Notes in square brackets will be unspoken; asterisk * denotes a click or build]

Thank you for the honor of helping you celebrate our 40th birthday by highlighting some of the disruptions that will make electricity’s next 40 years very different. Other disruptions will take the world off oil, and these two big stories overlap in electric vehicles, which Diarmuid will discuss, so I won’t for now. *

Inde

x of

U.S

. Prim

ary

Ener

gy

Per D

olla

r of R

eal G

DP

Heresy HappensU.S. energy intensity

0

0.25

0.5

0.75

1

1.25

1975 1990 2005 2020 2035 2050

Government and Industry Forecasts, ~1975

Reinventing Fire, 2011

Lovins, Foreign Affairs, Fall 1976

Actual

In 1975, our government and industry all insisted the * energy needed to make a dollar of GDP couldn’t go down. * A year later I heretically suggested it could drop 67% in 50 years. * So far it’s dropped 56% in 40 years—⅔ through technical efficiency, ⅓ through structural change. Those savings have been 31x bigger than the increase from doubling renewable supply. Yet our Reinventing Fire synthesis showed how innovations just through 2010 * can now save another threefold, twice what I’d thought, at a third the real cost, and six years later, that looks conservative. /

That’s partly because our little black-swan hatchery has developed “integrative design.” Optimizing vehicles, buildings, and factories as whole systems, not isolated components, can often make very big energy savings cost less than small ones, turning diminishing returns into expanding returns.*

U.S. buildings: 3–4× energy productivity worth 4× its cost (site energy intensities in kWh/m2-y; U.S. office median ~293)

284➝85 (–70%)2013 retrofit

~277➝173 (–38%) 2010 retrofit

...➝108 (–63%) 2010–11 new

...➝≤50 (–83% to –85%) 2015 new

Yet all the technologies in the 2015 example existed well before 2005!

The * 38% energy saving in our Empire State Building retrofit seemed pretty good until our * cost-effective Denver retrofit three years later saved 70%, making this half-century-old Federal complex more efficient than the * best new U.S. office. That NREL office in turn is * only half as efficient as RMI’s cost-effective new office some of you visited yesterday. So with nearly unchanged technologies, better integrative design has cut potential office energy intensity by more than twofold (and fivefold from the median) in five years—and * buildings, dominated by offices, use three-fourths of U.S. electricity. *In buildings and industry, just optimizing pipes and duct friction could save about a fifth of the world’s electricity, with paybacks under a year for retrofit and under zero for new-build. Yet integrative design isn’t in any industry forecast or official study, because this huge scaling vector is not a new technology but a design method. *

50

100

150

Lum

inou

s ef

ficac

y (lm

/W)

Incandescent lamp1879

200

250

300

1900 1950 20000

Years

1996

LED and PV

50

100

150

Lu

min

ous e

ffic

acy (

lm/W

)

Fluorescent lamp

Incandescent lamp

Halogen lamp

Sodium-vapor lamp

1965

1938

1959

1879

200

250

300

1900 1950 20000

Years

1996

50

100

150

Lu

min

ous e

ffic

acy (

lm/W

)

Fluorescent lamp

Incandescent lamp

Halogen lamp

Sodium-vapor lamp

White LED

1965

1938

1959

1879

200

250

300

1900 1950 20000

Years

1996

Sources: L: courtesy of Dr. Yukio Narukawa (Nichia Corp., Tokushima, Japan) from J. Physics. D: Appl. Phys. 43(2010) 354002, doi:10.1088/0022-3727/43/35/354002, updated by RMI with CREE lm/W data, 2015, www.cree.com/News-and-Events/Cree-News/Press-Releases/2014/March/300LPW-LED-barrier;. R: RMI analysis, at average 2013 USEIA fossil-fueled generation efficiencies and each year’s real fuel costs (no O&M); utility-scale PV: LBNL, Utility-Scale Solar 2013 (Sep 2014), Fig. 18; onshore wind: USDOE, 2013 Wind Technologies Market Report (Aug 2014), “Windbelt” (Interior zone) windfarms’ average PPA; German feed-in tariff (falls with cost to yield ~6%/y real return): Fraunhofer ISE, Cost Perspective, Grid and Market Integration of Renewable Energies, p 6 (Jan 2014); all sources net of subsidies; graph inspired by 2014 “Terrordome” slide, Michael Parker, Bernstein Alliance

0

100

200

300

400

500

600

700

800

1990

1994

1998

2002

2006

2010

2014

Coal-fired steam turbine, fuel cost onlyOil-fired condensing, fuel cost onlyNatural gas CCGT, fuel cost onlyUtility-scale solar PV, total costOnshore windpower, total costGerman PV residential feed-in tariff

Real

bus

bar p

rice

or fu

el c

ost,

2011

US$

/MW

h

(Seattle-like climate)

Efficiency technologies used to advance slowly: * Edison’s incandescent lamps and their * successors took decades. But * then white LEDs each decade got 30x more efficient, 20x brighter, and 10x cheaper. In the next five years, they’ll save an eighth of the world’s electricity, prying open an old crack in electric utilities’ business model. / Thomas Edison didn’t sell electricity; he sold light. He charged one cent to light a lamp for an hour, so more-efficient lamps would cut his costs and raise his profits. But in 1892, New York Edison Company switched to selling kWh, so ever since, greater customer efficiency has cut utilities’ revenues, not their costs. They sell a commodity when customers want an infrastructure or a service, like hot showers and cold beer. / What else changes this fast? LEDs inside out, also known as * PVs. Their meteorite strike makes solar power cost less than utilities’ fossil fuels (the dashed lines). Windpower, the aqua curve, costs still less, often making old thermal plants shut down as uneconomic to run. Happily, closing distressed nuclear plants and reinvesting their saved opex in far cheaper efficiency or renewables can save even more carbon than keeping them running. *

http://www.cree.com/News-and-Events/Cree-News/Press-Releases/2014/March/300LPW-LED-barrier

Utility revenues

Efficiency Distributed renewables

Storage (including EVs)

Flexible demand

New financial and business models

Regulatory shifts

Customer preferences

Integrative design

$

Disruptors are * converging on the * electricity industry from all * sides. But these * “eight PacMen of the * Apocalypse” multiply quickly, they hunt in packs, and they don’t add; they multiply and exponentiate—faster than most utilities’ cultures and stakeholders can cope. *

Two years ago, all central power plants were called “dinosaurs”—“too big, too inflexible, not even relevant for backup power in the long run.” By whom? Not Greenpeace but Union Bank of Switzerland.

2002 2004 2006 2008 2010 2012 2014 2016

Renewable Energy’s Costs Continue to PlummetWind and photovoltaics: U.S. generation-weighted-average Power Purchase Agreement prices, by year of signing

250

200

150

100

50

U.S. wholesale power price range

wind PPAs

utility-scale solar PPAs

leve

lized

201

4 U

S$/M

Wh

x

**

lowest unsubsidized world bids

* The range of prices for US wholesale electricity is now widely undercut by the average PPA prices of new * windpower and * utility-scale solar power. These renewables are temporarily subsidized, but generally less than the permanent subsidies to nonrenewables. That’s why renewables are winning in * unsubsidized auctions worldwide, like these prices for Moroccan wind and for Mexican PVs near Texas. *

0 10 10 91 0Years

“Cathedral” Photovoltaics

0 10 20 30 40 50 60 70 8

0 GW-y1 GW-y3 GW-y6 GW-y10 GW-y15 GW-y21 GW-y28 GW-y36 GW-y45 GW-y0 GW-y3 GW-y1 GW-y2 GW-y

[animation plays for 17s] Modern renewables also scale up in a fundamentally different way. Traditionally, we built giant cathedral-like power plants, each costing billions of dollars and taking many years to license and build. But now each year, with roughly comparable capital, you can build a factory that produces each year thereafter enough solar cells to generate each year thereafter as much electricity as your “cathedral” ultimately will. So solar output worldwide is scaling faster than cellphones. *

International Energy Agency global wind and solar forecastsCumulative GW installed

0

500

1,000

1,500

2,000

2,500

3,000

2000 2005 2010 2015 2020 2025 2030 2035 2040

WEO 2002WEO 2004WEO 2006WEO 2008WEO 2010WEO 2012WEO 2014WEO 2015actualBNEF forecast

Wind Solar

0

500

1,000

1,500

2,000

2,500

3,000

2000 2005 2010 2015 2020 2025 2030 2035 2040

5x upward revision since

2000

14x upward revision since

2000

Source: IEA WEO, BNEF (forecast from June 2015), slide inspired by Michael Liebreich’s 2016 BNEF Summit keynote

When renewables get cheaper, we buy more, so they get cheaper, so we buy more, leaving forecasters in the dust. Thus last year alone, modern renewables, excluding big hydro dams, added 135 GW, 22% above Bloomberg’s forecast and over half the world’s new capacity, and they got $266 billion of asset investment, ⅔ of it private. *

Yet we’re often told that only coal, gas, and nuclear stations can keep the lights on, because they’re “24/7,” while windpower and photovoltaics are “variable” and hence unreliable. Neither statement is true. *

French windpower output, December 2011: forecasted one day ahead vs. actual

Variable Renewables Can Be Forecasted At Least as Accurately as Electricity Demand

Source: Bernard Chabot, 10 April 2013, Fig. 7, www.renewablesinternational.net/wind-power-statistics-by-the-hour/150/505/61845/, data from French TSO RTE

GW

00.5

11.5

22.5

33.5

44.5

5

First, “variable” doesn’t mean “unpredictable.” Here’s how accurately the French grid operator in one stormy winter month * forecast a day ahead the * output of the country’s windfarms. I’ll bet they wish they could forecast demand that accurately! *

http://www.renewablesinternational.net/wind-power-statistics-by-the-hour/150/505/61845/

!10% Downtime

!12% Downtime

Second, we built the grid because no generator is 24/7. They all fail. When * giant coal and nuclear plants fail, a billion watts vanishes in milliseconds, often for weeks or months and often without warning. * Grids routinely handle this intermittence by backing up failed plants with working ones, * and in exactly the same way, but often at lower cost, grids can handle the forecastable variations of PV and windpower....

[automated transition continues...]

...by backing up variable renewables with other renewables, of other kinds or in other places. Thus highly reliable power can come from a portfolio of largely or wholly renewable sources when they’re forecasted, integrated, and diversified by type and location. Here’s a difficult case, for the isolated grid of Texas… *

[adjust length of animation if needed—can’t continue to slide 51 until animation ends]

0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7

Original loadLoad after efficiency

Geothermal etc.

Choreographing Variable Renewable GenerationERCOT power pool, Texas summer week, 2050 (RMI hourly simulation, 2004 renewables data

HVAC ice/EV storageBiomass/biogas

Storage recoveryDemand response

Solar (25 GW)Wind (37 GW)

Spilled power (~5%)

* Its expected loads in a 2050 summer week can get much * smaller and less peaky with efficient use. Then we can install enough * wind and * PV power to make 86% of the annual electricity—and get the other 14% from dispatchable renewables like * geothermal, small hydro, solar-thermal-electric, * feedlot biogas burned in existing gas turbines, and burning wastes and obsolete energy studies. This 100% renewable supply can then be matched to the load by putting excess electricity into * two kinds of distributed storage worthy buying anyway—ice-storage air-conditioning and smart charging of electrified autos, both fully deployed—then * recovering that energy when needed, and * filling the last gaps with unobtrusively flexible demand. * Only ~5% of the annual renewable output is left over, implying good economics. *


0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7

Choreographing Variable Renewable GenerationERCOT power pool, Texas summer week, 2050 (RMI hourly simulation)


0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7


Wind (37 GW)


0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7




0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7



Geothermal etc.


0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7



Geothermal etc.Biomass/biogas


0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7

Geothermal etc.





0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7

Geothermal etc.



Storage recovery



0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7

Geothermal etc.






0

10

20

30

40

50

60

GW

Day

1 2 3 4 5 6 7

Geothermal etc.





Spilled power (~5%)

Europe, 2014 renewable % of total electricity consumed

Choreographing Variable Renewable Generation

Such grid integration is a reality today. * In 2014, Germany met 27% of its annual electricity needs from renewables, Italy 33%[, Ireland 20%, France and Britain 19%]. * But four other * European countries with * modest or no hydropower met about * half their electricity needs from renewables, adding no bulk storage and with superior reliability—for Denmark and Germany, about 10x ours. The ultrareliable former East German utility got 49% of its electricity last year from PVs and wind. These grids are run the way a conductor leads a symphony orchestra: no instrument plays all the time, but the ensemble continuously makes beautiful music. *

27%Germany (2015 peak 78%)

59%Denmark (33% wind; 2013 windpower peak 136%—55% for all December)

50%Scotland

46%Spain (including 21% wind, 14% hydro, 5% solar)

64%Portugal (peak 100% in 2011; 70% for the whole first half of 2013, incl, 26% wind & 34% hydro; 17% in 2005)

Such grid integration is a reality today. * In 2014, Germany met 27% of its annual electricity needs from renewables, Italy 33%[, Ireland 20%, France and Britain 19%]. * But four other * European countries with * modest or no hydropower met about * half their electricity needs from renewables, adding no bulk storage and with superior reliability—for Denmark and Germany, about 10x ours. The ultrareliable former East German utility got 49% of its electricity last year from PVs and wind. These grids are run the way a conductor leads a symphony orchestra: no instrument plays all the time, but the ensemble continuously makes beautiful music. *

Grid flexibility supply curve cost

efficient use

demand response

(all values shown are conceptual and illustrative)

accurate forecasting

of wind + PV

diversify renewables by

type and location

dispatchable renewables and

CHP

bulk storage

fossil-fueled

backup

distributed electricity storage

thermal storage

ability to accommodatereliably a large share ofvariable renewable power

So we have not just one way (bulk storage) but about nine ways to make the grid more flexible so it can gracefully and reliably accommodate variable PVs and windpower. This schematic sketch arranges them in order of increasing cost. Your actual costs will vary, but clearly you’d do bulk storage last, not first, and the eight cheaper methods are so big that we needn’t wait for a storage miracle. Those who believe renewables need one will be gone first. *

Load control + PVs = grid optional

0"

2"

4"

6"

8"

10"

12"

kW#

Uncontrolled: ~50% of solar PV production is sent to the grid, but if the utility doesn’t pay for that energy, how could customers respond?

EV-charging

!"!!!!

!2.00!!

!4.00!!

!6.00!!

!8.00!!

!10.00!!

!12.00!!

kW#

Unc!Load! Smart!AC! Smart!DHW! Smart!Dryer!

0"

2"

4"

6"

8"

10"

12"

kW#

Controlled: flexible load enables customers to consume >80% of solar PV production onsite. The utility loses nearly all its windfall and most of its ordinary revenue.

AC

DHW

Dryer

Other

Solar PVAC

DHW

Dryer

Other

Solar PVEV-charging

Source: RMI analysis “The Economics of Load Flexibility,” 2015

Efforts to block or tax rooftop PVs only annoy and educate customers so they leave the grid faster, or buy less grid power by combining PVs with timeshifting storage. But customers hold a third PacMan too. * Half a typical Hawai‘ian household’s daily loads occur within solar hours (the yellow curve), but HECO, having failed to block solar hookups, now proposes to confiscate all surplus PV power without paying for it. Soon * customers can riposte with smart appliances that shift 80–90% of their loads into the solar hours. HECO will then lose nearly all its desired windfall and most of its ordinary revenue. We just studied six utilities’ anti-solar tariffs, like APS’s. We found they’re all well-aimed boomerangs that tend to expand and accelerate solar adoption. *

重能：中源塑国⾯面向2050年能源消费和⽣生产⾰革命路线图研究

Five years ago, our business book Reinventing Fire rigorously showed how a 2.6x bigger 2050 US economy could need no oil, coal, or nuclear energy and ⅓ less gas by tripling efficiency and quintupling renewables, $5t cheaper, needing no new inventions nor acts of Congress, the transition led by business for profit. Now in press is our * Chinese-American consortium’s Reinventing Fire: China, a 54-FTE 2.5-y effort by the Chinese government’s best energy modellers, aided by RMI, Berkeley Lab, and Energy Foundation China, together creating a roadmap for China’s energy revolution. It shows how to raise China’s energy productivity 7x by 2050, carbon productivity 12x, $3.4 trillion cheaper if carbon is priced at zero. This roadmap is now the foundation of China’s energy strategy, embedded in the 13th Five Year Plan approved in March. *

Price > CostValue >

In such turbulence, short-term worries can obscure strategic danger. Utility managers understandably focus on the need for * price to exceed cost. But the other part of the business condition says * value must exceed price. If competitors offer a superior value proposition and grab your revenues, it doesn’t matter if you can profitably sell what customers are no longer buying. *

1900: where’s the first car?

Easter Parades on Fifth Avenue, New York, 13 years apart

1913: where’s the last horse?

Images: L, National Archive, www.archives.gov/research/american-cities/images/american-cities-101.jpg; R, shorpy.com/node/204. Inspiration: Tona Seba’s keynote lecture at AltCar, Santa Monica CA, 28 Oct 2014, http://tonyseba.com/keynote-at-altcar-expo-100-electric-transportation-100-solar-by-2030/

?

Disruption can be brutally quick. Tony Seba notes that in this Manhattan scene, in 1900 you have to look hard for * the first car, then * just 13 years later you have to look harder for the * last horse (if there is one). The horse-and-buggy industry thought it had many decades to adapt, but Henry Ford made the Model T 62% cheaper in 13 years. Today, PV modules just got 80% cheaper in five years. Car-owning households went from 8 to 80% in ten years, ¾ financed by an innovation called car loans. The same fraction of our rooftop solar is now innovatively financed. /The pace of transformation is set not by incumbents but by insurgents. They’re not inhibited by incumbents’ business models, cultures, or legacy assets. As Jack Welch said, “If the rate of change on the outside is greater than the rate of change on the inside, the end is near.” /Incumbents actually have even less time than insurgents grant, because investors flee before customers do. Moving at the pace of Twitter, capital markets keenly sniff out disruption. Once they think you’re in or headed for the toaster, they don’t wait for the toast to get done before they decapitalize you and invest in your successors. That’s now happening. *

0 1 2 3 4 5 6 7 8

May 2015

SolarCity

Exelon

A new and old utility

Inde

xed

stoc

k m

arke

t pric

e(1

3 De

cem

ber 2

012

= 1)

12 December 2012(SolarCity’s IPO)

$6b market cap

$34b market cap

Here’s how the stock price of Solar City—which recently became a utility—has moved compared to Exelon. And the market hasn’t yet realized that such utilities could lose half their revenues in the 2020s. *

29 June 2010(Tesla’s IPO)

May 2015

Tesla

0

2

4

6

8

10

12

14 A new and old automaker

General MotorsInde

xed

stoc

k m

arke

t pric

e(3

0 Ju

ne 2

010

= 1)

$30b market cap

$57b market cap

50 thousand cars per year

8 million cars per year

Or here’s how a 2003 automotive startup we’ll soon hear from [Diarmuid] gained half the market cap of General Motors while selling 0.6% as many cars. *

www.rmi.org | [email protected] | +1 970 927 7202

RO

CKY MOUNTAIN

INSTIT UTE

WAR R O O M

CARBON

So we live in interesting times, but at root the issue for utilities is simple. Customers are figuring out that they can use electrons far more productively and timely, make more of their own, and even trade them with each other. It’s smart to sell customers what they want before someone else does. All the rest is detail. *

http://www.rmi.org

mailto:[email protected]

Rapid Growth of Electrified Cars!

Source: Tom Randall (Bloomberg), “Here’s How Electric Cars Will Cause the Next Oil Crisis,” 25 Feb 2016, http://www.bloomberg.com/features/2016-ev-oil-crisis/; see also RMI, “Electric Vehicle Charging as a Distributed Energy Resource,” in press, spring 2016

U.S. EV sales flattened—but global sales are growing ~60%/y

Global EV sales are growing 60%/y. China last year sold more EVs than the world did in 2012. Bloomberg expects EVs to save 2 Mb/d in the next 7–9 y and 12 by 2040 (9x ExxonMobil’s top forecast). That was before India targeted and Germany considered 100% EVs by 2030. And it doesn’t count accelerating EVs by feebates, which have made ⅓ of Norway’s new cars electric—50x the U.S. share; or ultralighting, which saves up to ⅔ of the costly batteries (I drove here in a mass-producted Hypercar whose carbon fiber is paid for by needing fewer batteries); or monetizing EVs’ value as a grid resource, which can repay about half their cost; or shareable, autonomous, and mobility-as-a-service auto business models that strongly advantage electric traction.EVs should reach sticker-price parity in the 2020s as their * annual battery production nears 1 TWh. That means abundant cheap batteries, faster distributed-solar growth, weaker gas prices, and many GW/y of new distributed EV storage and demand-response resources whose grid benefits will attract many competitors. *

http://www.bloomberg.com/features/2016-ev-oil-crisis/

convergent energy disruptions...power price range wind ppas utility-scale solar ppas levelized 2014...

Documents