Can Renewable
Electricity
Generation Scale?
Choices and Challenges for the
Current Century
Three Main Challenges



Electricity Production:  per capita
consumption is increasing faster than
energy efficiency
Developing Nations: 2 Billion humans
are still without electricity
Electricity Distribution:  Aging grid
already at capacity; can’t easily
accommodate new sources of
generation
Production and Consumption on the
Century Timescale
A Century of Change
(1900 (=1) vs 2000)
•
•
•
•
•
•
Industrial Output: 40
Marine Fish Catch: 35
CO2 Emissions: 17
Total Energy Use: 16
Coal Production: 7
World Population: 4
No More Fish by
2100 at this rate
of Consumption
Electricity Demand Scaling
•
•
•
•
•
•
1890 – 1955: D a P1.6
1955 – 2005: D a P3.5 (!)
US Nameplate Electricity Gen: 1.4 TW
World 4 +/- 0.5 TW (China Uncertainty)
2050 Estimate by Pop Scaling: 12-16 TW
2050 Estimate including improvements in
electricity generating efficiency: 8- 12 TW
Why China is Such A Concern
Factor 16 Growth (4
doubling times) in 35 years!
International Energy Outlook 2010
4 TW
But this is Net Generation: Divide
by 0.4 (40%) to Get to Nameplate
World Energy Use 2050
Energy = electricity + transport + heat
 Electricity Production: 30% of total Energy
 So we need approximately 35 TW of
Power in 2050
 Assume that as the annual average over
2000-2100
 We need energy source with 3500 TWyrs
 Make an Inventory of Energy sources

Ultimately Recoverable Resource







Conventional Oil/Gas
Unconventional Oil
Coal
Methane Clathrates
Oil Shale
Uranium Ore
Geothermal Steam conventional







1000 TWy (1/4 need)
2000
5000
20,000
30,000
2,000
4,000
Non Fossil Fuels





Hot Dry Rock
Sunlight/OTEC
Wind Energy
Gulf Stream
Global Biomass





1.5 MW
1,000,000 TWyrs
harv
9,000,000 est
devic
es
100,000 - 200,000*
140,000
Radia
10,000
tive
transf
er
In Principle, Incident Energy is Sufficient
 but
theor
y
how to recover and distribute it in the most
cost
effective manner?
Not Considering Nuclear As
Scalable
Current timescale from design through
approval, construction and turn on for a
typical 1200 MW reactor is 16 years !
 Insufficient Uranium Ore
 Thorium reactors are certainly better but
special handling conditions are required
for Thorium and we have zero
infrastructure for this

Scaling now becomes critical


Which renewable energy technologies scale
the best in terms of unit yield, production
timescales and material costs?
This need to properly scale is now coming at
a time when civilization has essentially
exhausted about 70% of available earth
resources
More Earth Limitations




Total fuel cell production limited by amount of
accessible platinum on the planet; 500 million
vehicles  lithospheric exhaustion in 15 years
Higher efficiency PVs limited by accessible
amounts of rare materials (e.g. Tellurium,
Sellenium, even Cooper)
Conventional Transmission media limited by
available new Copper
Clear need for Carbon based materials (fiber,
nanotubes) to overcome this.
This Decade (2010-2020) is Crucial: We must
do much better than the Last Decade
1998
1% PER DECADE SUCKS
Evaluation Rubric For All forms of
Renewables
MW output per surface area (MW/KM2)
 2. MW output per material use (MW/Ton)
 3. MW output per job created (Jobs/MW)
 4. MW output versus production time scale
to bring on line (months/MW)
 5. Capital cost per MW ($/Watt)
 6. Realistic Levelized Cost (cents per
KWH)
 1.
Dollars Per Megawatt per unit Land
use per unit Material Use






20 KW power buoy
5 MW Wind Turbine
LNG closed cycle
Wind Farm (3Dx6D)
Solar Trough
Pelamis Farm






850 Tons per MW
100 Tons per MW
1500 MW sq km
50-60 MW sq km
20-25 MW sq km
30-35 MW sq km
To Evaluate Competing Electricity
Generating Technologies
 Develop an internally consistent indexing
system for the 6 attributes listed previously
(the dow jones is an index)
 Use real world data and real world physics to
best determine the values
 Weight the indexes appropriately (real world
cares about $/Watt and Jobs Created)
 Choose Baseline – we will use Solar in the
following exercise
Indexing – Solar Troughs
 1. Land ~20 MW/km (over 24 hour day) = 1
 2. Materials ~3 tons per kw = 1
 3. Jobs ~3 jobs per MW
 4. Time ~10 MW per month
 5. Capital ~3$ per watt real facility cost
 6. Levelized 10 cents per KWH
Index
Solar Wind
Waves Biomass
Land
1
2.5
1
.2
Material 1
3
.2
1
Jobs
1
1
1
5
Time
1
3
.5
.5
Capital
1
2.5
.5
.5
Level
1
3
.75
1
Cumulative Index = 1+2+(1.5)3+4+1.25(5)+1.25(6)
Highest Index is Best
Relative Ranking
Solar = 7
 Waves =4.75
 Biomass =11 (because of jobs created)
 Wind = 17 (lower material intensity and
low Levelized costs)
 In general, wind is more scalable than
Solar and wind always beats Solar PV

Scalability of Wind
Thinking Real Big – Gulf Current
Gulf Current Total Yield
 Ocean currents have approximately 830 times the energy
density of moving air (i.e. wind). For a 2 m/sec current, the
power density is 4 KW per square meter.
 Assume underwater array of 3 blade turbines with blade
length of 2 meters. That means 50KW unit capacity.
 Build a 1 km leg with turbines spaced by 4 meters. Each leg
then has 250 units x 50 kw = 12.5 MW per km.
 Build parallel legs over a 10 km wide section which each leg
separated by 100 meters  1.25 GW in this 1x10 x-section
 Now repeat this over 100 km of Gulf stream and you have 125 GW.
Repeat this over 1000 km and you have 1.26 TW (or equivalent of US
Total nameplate capacity) Gee, Isn't this a solution?
Scalability in the Real World
 Energy Payback Times (EPT)
 Material supply chain limitations
 Good current example – Lithium for Li-
Ion batteries.
1 Kg per laptop (200 million per year)
250 KG per EV
Supply Chain Limitations
All Supply Chains can reach
physical saturation limits
Historical Growth Patterns
Sustaining high growth is hard
Where is the wind going? US has
biggest 1 year gain
Current World Build Out
Annual Growth Rate is 22%
Looks Good but .6 TW by 2020 will only be
about 10% of total nameplate production
Wind Reality




144,000 MW installed capacity is about 44,000
MW real produced (capacity factor of 0.31) 
about 1% of global now
A 10% goal can reached in 2018 but would
require production to be 5 times higher than
now
Its unclear at the moment, which aspect of wind
turbine construction can’t scale to meet this goal
Transmission line limitations, may be the limiting
factor unless we start using HYDROGEN as the
energy carrier
Real World Wind Example:





3600 turbine blades per year
Requires 2500 workers
Requires 2 x 400,000 sf facilities.
1200 2.5 MW wind Turbines per year
3000 MW per year  115 yrs to replace
340,000 MW of Coal (NP)
 Need to ramp this up by a factor of 6 to
replace all Coal (NP) by 2030
Solar: Harvesting a Vastly
distributed Resource
A Large Scalability Challenge
Wind vs Solar
 Wind
has a materials only supply chain
 Solar PV has a chemical processing
component
Solar Trough Supply Chain
 Is mostly in materials, similar to wind
supply chain process:
Solar Growth
29% annual growth rate
Reality: Solar




18000 annual MW of Peak is closer to
3000 MW of average annual
Current Global is ~4.8 TW or 1600 times
larger
Supposed Global electricity growth is 2.5%
By 2027 10% electricity would come from
PVs which would now be produced at the
rate 60 times greater than now  in just
17 years!  not possible
Real World




Hard to predict how scalable an industry really is
Averaged over next 25 years solar is likely to be around
15% due to various material limitations
Wind is likely to be only 10% due to material limitations
and production physical space and
transportation/distribution constraints
With these rates by 2035 and a 9 TWe planet
 Solar = 8%
 Wind = 50%
 So 2050 Renewable Energy Planet is possible if one
includes the gulf current projet.
Energy Pay Back Times





Energy Payback Time  the timescale over which the total
supply chain energy of manufacturing is paid back by
harvested energy from the device.
EPBT for Solar PV at best is 2.5 years: This depends
primarily on PV efficiency and PV insolation
EPBT for Wind Turbine is now at about 3-4 months: This
depends mostly on turbine capacity. The energy required to
make a 3 MW turbine is a lot less than 2 times the energy it
takes to make a 1.5 MW wind turbine.
The bottom line here is that, in every way measureable, Wind
is more scalable than Solar PV.
But, the energy payback times of concentrating solar power
systems (e.g. solar troughs) is comparable to wind energy
(5-8 months) because that technology has a similar material
supply chain it just doesn't have the land use scalability of
wind.
The Need for the Hydrogen
Nano Battery
Molecular Storage Containers
Current best molecule is Ca32C60
 Molecular weight is high so addition of H
does not increase weight significantly

Major Physics
Problem:
You can put the
Hydrogen in But
you can’t get it
out Fast enough to
Power a Fuel Cell
Time to Think of Big Projects
Thinking Big -Solar

Sonoran Desert Project:
300,000 square km @
2% coverage yields
100,000 MW
10% coverage yields
500,000 MW
Thinking Big - Wind
Lake Michigan Wind
project down North South
Axis: Populate 400 x 30 km
box with 30 legs each
containing 1200 5 MW
turbines: 180,000 MW
Thinking Real Big - Wind
Great Prairie Wind Farm with 100
MW vertical Wind Turbines:
Construct 10,000 of these (Space
Needle Size) @1 per 125 square
km. This produces 1TW of
electricity and effectively replaces
all other forms of electricity
generation in the US.
Thinking Real Big – Aleutian Wind Wall
TW wind
power scale
incident on
the
Archipelago
(nearly
constantly)
(wind power
density
averages 600700 watts per
The Necessary Smart Grid




Route electricity like IP
Each node on network can buy, sell,
or store electricity
100 million network nodes – each
home is a power plant
SMS messaging on system state; poll
once every 2 minutes  4 Terrabits
per second
Summary
• Yes this is a big infrastructure challenge. Need to install
50,000 MW per year for 20-30 years. This is not
physically impossible. Big infrastructure can be built if
you put enough resources into it:
The Interstate Highway System
The Depression Era Federal Hydroelectric Projects (45% green)
Going to the Moon in 10 years
• Think seriously about using Hydrogen as a proxy for
transmission of electricity within the new smart grid
• No one technological solution (e.g. fusion) yet exists 
need Network of regionally based alternative energy
facilities
2010 Realities
• The Obama Unit: (780 Billion dollars)  insufficient
renewable energy infrastructure investment
• New (post 2006) domestic Natural gas discoveries are
moving the market back towards fossils and away from
renewables
• BAU depletes most planetary resources at the 90% level
by 2030  Chinese Sledgehammer
• We are in a collective stupor on all of these issues; if we
don’t prioritize the long term, then we will become stone
aged in the short term.