A number that comes up frequently is the “what is the percentage of US land needed to make all our electricity with PV?”
The assumptions needed to calculate this are:
1. The efficiency of the PV system (because that defines the area needed)
2. The fraction of land the modules cover within the deployed system (packing factor)
3. The output in kWh per installed W, which depends on the sunlight, tracking or not, and other losses within the system.
These numbers can vary quite a bit. For example, system efficiencies are between 5% and perhaps 15% at this time, and as they evolve, will almost certainly rise. Traditionally one is safe to choose 10% as the system efficiency, knowing full well it is likely to be better.
Packing factor varies between tracking and nontracking systems, and seems to be about three for those where land is a constraint. I assume simple, nontracking systems.
Output (kWh/W installed) depends strongly on sunlight and tracking. Assuming nontracking systems but mostly in sunny places allows one to estimate about 1.8 kWh per installed W output per year.
Calculations Showing Land Area
| Assumption | Amount |
| m2/MW modules (10% System Efficiency) | 10,000 |
| Packing factor | 3 |
| m2/MW system | 30,000 |
| km2/TW | 30,000 |
| TWh/TW | 1,800 |
| km2/TWh | 16.7 |
| Area of US (km2) | 9372610 |
| US electricity use TWh/yr | 4000 |
| % US land for All US Electricity from PV | 0.7% |
So, with these assumptions it would take less than 1% of US land area to make all our electricity with PV. If we put more in the sunniest places, it would take less; if we put more all around the country, more. These numbers should be about right as we evolve, but probably conservative – that is, real land area will be a little lower.
By the way, hydro produces 7% of our electricity and uses 100,000 km2 for lakes behind the dams. This is about 1% of US land area. So in comparison, hydro is about 20 times more land intensive per kWh than PV ((1%/7%) / (0.7%/100%)). This is actually a more important answer than the raw number, since it puts it in perspective. PV land area is 20 times less than hydro land area, yet we value hydro a great deal. And PV use doesn’t make any permanent alteration on the land like making artificial lakes.
Ken Zweibel
Utilities don’t buy solar energy to provide energy to their customers. It’s too expensive and too unreliable. They don’t even buy wind for their customers, even though in most cases it is very close to being cost-effective. Neither of these intermittent sources suits their needs, either in price or dependability.
But they do buy solar and wind. They are trying to reduce their carbon dioxide emissions, and to a lesser degree, diversify their sources of electricity in case of fuel price shocks.
They are required by state and Federal regulations to buy wind and solar.
It’s really a simple equation. If -
Federal and state requirements – Added cost of electricity – Intermittency and transmission penalty > 0,
then they buy solar and wind.
Ok, for wind the added cost of electricity is small, even nonexistent, and can go positive with the Federal tax credit. But then there is the intermittency penalty and odd-ball things like wind forecasting and added spinning reserves. Utility executives wouldn’t buy added complexity if they didn’t have to.
Societal preferences for lower carbon dioxide, and insurance against energy price fluctuations is why we deploy solar and wind.
The competition for solar is the other non-CO2 sources of energy, not coal, and certainly not natural gas (which helps by being flexible, filling in when solar or wind aren’t available). It is also energy efficiency, and the metric is the cost of reducing a MT of CO2. Dollars per MT CO2 (evaluated over the operating life of the project), that is what counts. (Of course, other things count, too, like water use and time of day pricing. They are all in the mix.)
Yes, the price of solar will keep coming down. Of course that helps – simply look again at the equation! The less that added cost, the more solar will be bought. But it is not because it will be cheaper than conventional sources (even though, someday it may be), but because it is cheap enough based on societal needs.
Ken Zweibel
We might save money if we harnessed solar and wind to displace all our coal and all the gasoline used for light duty vehicles (cars, SUVs, pick-ups). Let’s see how this works.
The US uses about 4000 TWh of electricity, and about half of that comes from coal (about 2000 TWh/yr from about 23 Quads of primary energy).
Figure 1. Coal burning is about 23 Q in the US, expected to rise to about 27 Q in 2035
Our light duty vehicles require 17 Quads of oil, but only 3.4 Quads of it actually gets to the vehicles and moves them (20% efficiency from oil to movement). At ~300 TWh per Quad, this is about 1000 TWh of energy. If we did it with electricity and assumed 25% losses (electricity to batteries to motors to movement), we would need about 1333 TWh to move our light duty vehicles without oil. The total to displace both oil for cars and coal would be 3333 TWh. Let say this takes 25 years, so let’s assume 33% more demand by then (perhaps not warranted, since we may be saving energy, but just to be conservative) – that would be 4444 TWh in 2035. (This should be rounded, but it’s such a charming number, we’ll use it as is.)
So how much would this cost? Let’s do half with solar, half with wind.
Solar output varies by location. So we make the obvious choice to use above-average solar locations instead of Seattle. We might assume a 20% annual capacity factor for solar, or 1750 kWh/kW installed. This means putting most solar in the Southwest and a few other sunny places.
For wind, we can assume about a 30% capacity factor, so annually 2600 kWh/kW.
Installed Solar and Wind (GW) To Make 4444 TWh/yr in 2035 (displacing all coal and the gasoline for light duty vehicles)
| Formula | Cumulative Installed 2035 | |
| Solar | 2222 TWh/(1750 kWh/kW-yr) | 1270 GW |
| Wind | 2222 TWh/(2600 kWh/kW-yr) | 854 GW |
Today solar is about $3/W for very large fields, and First Solar has signed a contract to supply China with 2 GW at about $2.5/W. Meanwhile, wind has risen to about $2/W. Given the history of cost reduction in PV, we can assume that PV will drop to $2/W within the next decade. Let’s assume that all this new power will cost $2/W, so $2540B for the PV and $1708B for the wind, or a total of about $4250B over 25 years.
How much do we save in avoided fuel? After all, we are turning off all the coal we now burn, and all the oil for light duty vehicles. Assuming $3/MMBtu for the coal (including a dollar a MBtu for rail costs), and oil at 100 $/bbl for the oil (including refining and distribution costs), we get:
Avoided Energy Costs per Year (2035)
| Avoided Source | Energy Avoided (2035) | Source Fossil Energy | Annual Avoided Fuel Cost |
Avoided CO2 (annual) |
| Electricity from Coal | 2666 TWh (33% increase) | 27 Quads Coal | $110B @ $4/MMBtu or 4.5 c/kWh (w/transport cost) | 2.6 billion MT/yr@95 kg/MBtu |
| Gasoline | 22 Quads (17 now + 33% increase) | 3.8B bbl oil/yr (@5.9 million Btu/bbl, or 170 million bbl/Q) | $380B @$100/bbl | 1.7 billion MT CO2(@3.15 MT CO2/MT oil & 0.141 MT/bbl) |
| Total Avoided (2035) | $490B/yr Avoided (2035) | 4.3 billion MT CO2 |
Note that almost four times more dollar savings comes from eliminating oil/gasoline than from saving coal. Oil is far more expensive than coal per Btu (about $16/MBtu at $100/bbl). But somewhat more CO2 saving comes from eliminating coal.
Figure 2. Coal and oil are the biggest contributors to CO2; and using solar, wind, and EVs reduces them significantly – over 4 GT/year in 2035.
Now how do we compare the costs with the savings? We have invested $4250B over a period of 25 years; and in the 25th year (and from then on) we are saving about $500B of fuel and avoiding over 4 GT of CO2.
The following table uses the simplest possible approach. On the left side (column 1) we have the annual outlay in billions to buy the solar and wind. On the far right (last column), we have the savings in avoided fuel costs. The annual fuel savings starts out too small to pay for the annual solar and wind, but catches up in 2024 as the accumulated installations eventually save more fuel than the new additions add to cost.
Outlays and Avoided Costs ($/B) Assuming a
Linear Increase in Deployment
|
|
Annual $B |
Annual TWh Added |
Cumulative |
Fuel Cost |
|
|
|
|
TWh/yr |
Avoided ($B) |
|
2010 |
12.1 |
12.6 |
12.6 |
1.5 |
|
2011 |
24.2 |
25.2 |
37.8 |
4.5 |
|
2012 |
36.3 |
37.8 |
75.6 |
9.1 |
|
2013 |
48.4 |
50.4 |
126 |
15.1 |
|
2014 |
60.5 |
63 |
189 |
22.6 |
|
2015 |
72.6 |
75.6 |
264.6 |
31.7 |
|
2016 |
84.7 |
88.2 |
352.8 |
42.3 |
|
2017 |
96.8 |
100.8 |
453.6 |
54.4 |
|
2018 |
108.9 |
113.4 |
567 |
67.9 |
|
2019 |
121 |
126 |
693 |
83.0 |
|
2020 |
133.1 |
138.6 |
831.6 |
99.7 |
|
2021 |
145.2 |
151.2 |
982.8 |
118 |
|
2022 |
157.3 |
163.8 |
1146.6 |
137 |
|
2023 |
169.4 |
176.4 |
1323 |
159 |
|
2024 |
181.5 |
189 |
1512 |
181 |
|
2025 |
193.6 |
201.6 |
1713.6 |
205 |
|
2026 |
205.7 |
214.2 |
1927.8 |
231 |
|
2027 |
217.8 |
226.8 |
2154.6 |
258 |
|
2028 |
229.9 |
239.4 |
2394 |
287 |
|
2029 |
242 |
252 |
2646 |
317 |
|
2030 |
254.1 |
264.6 |
2910.6 |
349 |
|
2031 |
266.2 |
277.2 |
3187.8 |
382 |
|
2032 |
278.3 |
289.8 |
3477.6 |
417 |
|
2033 |
290.4 |
302.4 |
3780 |
453 |
|
2034 |
302.5 |
315 |
4095 |
491 |
|
2035 |
314.6 |
327.6 |
4422.6 |
530 |
|
|
4247.1 |
4422.6 |
|
|
|
|
$B |
TWh/yr |
TWh/yr |
$B |
This can be seen more clearly in the next figure.
What this shows is that after a period of annual losses ending in 2024, we start saving more and more money from avoided fossil fuels. Our savings far outweigh our investment. In fact, by 2030, the avoided fuel costs outweigh the investment by $700B. If we invested no more after this, the savings would increase by about $500B per year, since the systems would still be producing electricity at almost no cost. (There is a small O&M cost, and the wind turbines would have to be replaced every 20 years. The solar PV would last indefinitely, even > 60 years, perhaps degrading a bit (under 0.5%) every year.)
One way to imagine this is that we taxpayers pay for the systems each year through taxes. Meanwhile, we pay less for our electricity and gasoline, without the charge for burning fuel. By 2024, our outlay is smaller than our savings, and we then start making money and have energy self-sufficiency, guaranteed steady energy prices, no energy blackmail from the Middle East, and have avoided the greater part of the CO2 we would otherwise produce.
Why does this look so good versus prior analyses? Are we missing something?
- It looks better because the solar costs are lower – $2/W instead of $4-$8, as in the past, before the recent major progress and drop in prices. It also assumes using solar in sunnier places.
- It looks better because it assumes that we have no finance charges. Interest payments would increase the annual costs. But if we pay as we go, we avoid them.
- How about electric storage? Do we need it? The electric vehicles themselves can partially smooth solar and wind variability. We will need the smart grid for dispatching, and hydro as back up. Probably some of the coal plants can be converted to natural gas to add back up flexibility. Compressed air storage can add further responsiveness.
- We need some transmission, but transmission is not that expensive (perhaps 1-2 c/kWh) – it is the access to right-of-way that is hard. With a national mandate, this will not be a problem.
- The crucial missing piece is electric vehicles. How are we going to have these savings, if we cannot avoid oil?
So this analysis is different because it includes the savings from avoided gasoline. It requires that we have electric transport. The solar and wind are ready.
Ken Zweibel
<!–[if !mso]> <! st1\:*{behavior:url(#ieooui) } –>
The US uses about 4000 TWh of electricity, and about half of that comes from coal (about 2000 TWh/yr from about 23 Quads of primary energy).
Figure <!–[if supportFields]> SEQ Figure \* ARABIC <![endif]–>1<!–[if supportFields]><![endif]–>. Coal burning is about 23 Q in the US, expected to rise to about 27 Q in 2035
Our light duty vehicles (cars, SUVs, pickups) require 17 Quads of oil, but only 3.4 Quads of it actually gets to the vehicles and moves them (20% efficiency from oil to movement). At ~300 TWh per Quad, this is about 1000 TWh of energy. If we did it with electricity and assumed 25% losses (electricity to batteries to motors to movement), we would need about 1333 TWh to move our light duty vehicles without oil. The total to displace both oil for cars and coal would be 3333 TWh. Let say this takes 25 years, so let’s assume 33% more demand by then (perhaps not warranted, since we may be saving energy, but just to be conservative) – that would be 4444 TWh in 2035.
So how much would this cost? Let’s do half with solar, half with wind.
Solar output varies by location. So we make the obvious choice to use above-average solar locations instead of Seattle. We might assume a 20% annual capacity factor for solar, or 1750 kWh/kW installed. This means most solar in the Southwest and a few other sunny places.
For wind, we can assume about a 30% capacity factor, so annually 2600 kWh/kW.
Installed Solar and Wind (GW) To Make 4444 TWh/yr in 2035 (displacing all coal and the gasoline for light duty vehicles)
|
|
Formula |
Cumulative Installed 2035 |
|
Solar |
2222 TWh/(1750 kWh/kW-yr) |
1270 GW |
|
Wind |
2222 TWh/(2600 kWh/kW-yr) |
854 GW |
Today solar is about $3/W for very large fields, and First Solar has signed a contract to supply China with 2 GW at about $2.5/W. Meanwhile, wind has risen to about $2/W. Given the history of cost reduction in PV, we can assume that PV will drop to $2/W within the next decade. Let’s assume that all this new power will cost $2/W, so $2540B for the PV and $1708B for the wind, or a total of about $4250B over 25 years.
How much do we save in avoided fuel? After all, we are turning off all the coal we now burn, and all the oil for light duty vehicles. Assuming $3/MMBtu for the coal (including a dollar a MBtu for rail costs), and oil at 100 $/bbl for the oil (including refining and distribution costs), we get:
Avoided Energy Costs per Year (2035)
|
Avoided Source |
Energy Avoided (2035) |
Source Fossil Energy |
Annual Avoided Fuel |
Avoided CO2 (annual) |
|
Electricity from Coal |
2666 TWh (33% increase) |
27 Quads Coal |
$110B @ $4/MMBtu or 4.5 c/kWh (w/transport cost) |
2.6 billion MT/yr @95 kg/MBtu |
|
Gasoline |
22 Quads (17 now + 33% increase) |
3.8B bbl oil/yr (@5.9 million Btu/bbl, or 170 million bbl/Q) |
$380B @$100/bbl |
1.7 billion MT CO2 (@3.15 MT CO2/MT oil & 0.141 MT/bbl) |
|
Total Avoided (2035) |
|
|
$490B/yr Avoided (2035) |
4.3 billion MT CO2 |
Note that almost four times more dollar savings comes from eliminating oil/gasoline than from saving coal. Oil is far more expensive than coal per Btu (about $16/MBtu at $100/bbl). But somewhat more CO2 saving comes from eliminating coal.
Figure <!–[if supportFields]> SEQ Figure \* ARABIC <![endif]–>2<!–[if supportFields]><![endif]–>. Coal and oil are the biggest contributors to CO2 and using solar, wind, and EVs reduces them significantly – over 4 GT/year in 2035.
Now how do we compare the costs with the savings? We have invested $4250B over a period of 25 years; and in the 25th year (and from then on) we are saving about $500B of fuel and avoiding over 4 GT of CO2.
The following table uses the simplest possible approach. On the left side (column 1) we have the annual outlay in billions to buy the solar and wind. On the far right (last column), we have the savings in avoided fuel costs. The annual fuel savings starts out too small to pay for the annual solar and wind, but catches up in 2024 as the accumulated installations eventually save more fuel than the new additions add to cost.
Outlays and Avoided Costs ($/B) Assuming a Linear Increase in Deployment
|
|
Annual $B |
Annual TWh |
Cumulative |
Fuel Cost |
|
|
|
|
TWh/yr |
Avoided ($B) |
|
2010 |
12.1 |
12.6 |
12.6 |
1.509972 |
|
2011 |
24.2 |
25.2 |
37.8 |
4.529915 |
|
2012 |
36.3 |
37.8 |
75.6 |
9.059829 |
|
2013 |
48.4 |
50.4 |
126 |
15.09972 |
|
2014 |
60.5 |
63 |
189 |
22.64957 |
|
2015 |
72.6 |
75.6 |
264.6 |
31.7094 |
|
2016 |
84.7 |
88.2 |
352.8 |
42.2792 |
|
2017 |
96.8 |
100.8 |
453.6 |
54.35897 |
|
2018 |
108.9 |
113.4 |
567 |
67.94872 |
|
2019 |
121 |
126 |
693 |
83.04843 |
|
2020 |
133.1 |
138.6 |
831.6 |
99.65812 |
|
2021 |
145.2 |
151.2 |
982.8 |
117.7778 |
|
2022 |
157.3 |
163.8 |
1146.6 |
137.4074 |
|
2023 |
169.4 |
176.4 |
1323 |
158.547 |
|
2024 |
181.5 |
189 |
1512 |
181.1966 |
|
2025 |
193.6 |
201.6 |
1713.6 |
205.3561 |
|
2026 |
205.7 |
214.2 |
1927.8 |
231.0256 |
|
2027 |
217.8 |
226.8 |
2154.6 |
258.2051 |
|
2028 |
229.9 |
239.4 |
2394 |
286.8946 |
|
2029 |
242 |
252 |
2646 |
317.094 |
|
2030 |
254.1 |
264.6 |
2910.6 |
348.8034 |
|
2031 |
266.2 |
277.2 |
3187.8 |
382.0228 |
|
2032 |
278.3 |
289.8 |
3477.6 |
416.7521 |
|
2033 |
290.4 |
302.4 |
3780 |
452.9915 |
|
2034 |
302.5 |
315 |
4095 |
490.7407 |
|
2035 |
314.6 |
327.6 |
4422.6 |
530 |
|
|
4247.1 |
4422.6 |
|
|
|
|
$B |
TWh/yr |
TWh/yr |
$B |
This can be seen clearly in the next figure.
What this shows is that after a period of annual losses ending in 2024, we start saving more and more money from avoided fossil fuels. Our savings far outweigh our investment. In fact, by 2030, the avoided fuel costs outweigh the investment by $700B. If we invested no more after this, the savings would increase by about $500B per year, since the systems would still be producing electricity at almost no cost. (There is a small O&M cost, and the wind turbines would have to be replaced every 20 years. The solar PV would last indefinitely, even > 60 years, perhaps degrading a bit (under 0.5%) every year.)
One way to imagine this is that we taxpayers pay for the systems each year through taxes. Meanwhile, we pay less for our electricity, without the charge for burning fuel. By 2024, our outlay is smaller than our savings, and we then start making money and have energy self-sufficiency, guaranteed steady energy prices, no energy blackmail from the Middle East, and have avoided the greater part of the CO2 we would otherwise produce.
Why does this look so good versus prior analyses? Are we missing something?
- It looks better because the solar costs are lower – $2/W instead of $4-$8, as in the past, before the recent major progress and drop in prices. It also assumes using solar in sunnier places.
- It looks better because it assumes that we have no finance charges. Interest payments would increase the annual costs. But if we pay as we go, we avoid them.
- How about electric storage? Do we need it? At this level, probably only minimally, since the electric vehicles themselves can also smooth solar and wind variability. We will use the smart grid for dispatching, and hydro as back up.
- We need some transmission, but transmission is not that expensive – it is the access that is hard.
- The crucial missing piece is electric vehicles. How are we going to have these savings, if we cannot avoid oil?
So this analysis is different because it includes the savings from avoided gasoline. It requires that we have electric transport. The solar and wind are ready.
We know that for some amount of money we could change from an oil-dependent society spewing CO2 to one that is self-sufficient and not damaging our world. The technology for doing so exists, and improvements are occurring quickly.
Besides electric vehicles, wind and solar, there are other potential contributors to this transition, but they are each relatively small. You can save just so much energy without turning yourself into a primitive society. Biomass can provide back-up power and portable fuel. Geothermal, ground-source heat pumps, solar hot water heaters, and wave power can add to our diverse portfolio. But the bulk of new energy is most likely to come from the two most powerful sources, wind and solar. And biomass is never going to supply enough portable fuel to be of significance.
Wind and solar’s competitive characteristics are:
- Wind:
- Advantages
- Cheap in windy places – about 4-8 c/kWh (without transmission costs)
- Located within a thousand miles, and usually half that, of most demand
- Large enough to supply almost all world energy demand and second only to solar in size
- Under-exploited offshore wind more dependable than on-shore wind, but more expensive (50%?)
- Disadvantages
- Blows 50% or more at night, sometimes much more (when electricity is less needed)
- Blows more in winter than summer, missing some peak demand
- More variable than solar annually, seasonally, and for most time intervals
- Very sensitive to local wind speed, so locations almost always require some transmission
- Unsightly towers can be seen from a distance; often located in visually appealing places due to correlation of terrain to wind speed
- Carries tangible operating costs and turbines must be replaced about every 20 years
- Advantages
- Solar PV
- Advantages
- Largest potential source of renewable electricity on Earth; dwarfs second largest (wind) and all others > 100
- Can be harnessed most places, but is cheapest where there is the most sun (linear relation to sunlight)
- Rapidly dropping in cost, so that the cheapest systems are about 15 c/kWh or less; and future systems will approximate 8-10 c/kWh in the sunniest locations
- Value of daytime electricity, and electricity supplied for summer peaks is much higher (even double) value of nighttime electricity from wind
- Lasts over 30 years with minimal losses, and may last 100 if built to last that long;
- Near-zero operating costs
- Can be built in months rather than years
- Disadvantages
- Cost is about double wind and may only reach levels that are similar to wind in the sunniest locations
- Output varies by season, and during the day as well due to clouds
- In some circumstances for systems under 30 MW in size, variation from cloud movement can be quite abrupt (variation for large systems over 100 MW in size is much smoother and smaller in magnitude)
- Use of the best solar locations (e.g., US Southwest) usually means that transmission will be needed
- Advantages
- Solar Thermal Electric
- Advantages
- May be built with relatively economical, on-site thermal storage, smoothing and extending output
- Simpler technology than PV (boilers, mirrors, no semiconductors)
- Disadvantages
- Uses water for cooling, or else costs 10% or so more if uses air cooling
- Can only be economical in near-perfect cloudless sunlight like deserts
- Must be built in relatively large systems
- Takes several years to build
- Cost reduction potential unclear
- More O&M than PV (more like wind)
- Advantages
There are parallel solutions to the variability of both wind and solar. They are:
- Using existing natural gas turbines to back them up when their resources are absent;
- Creating smarter and more responsive “balancing regions” that include solar and wind and their compensating gas turbines to smooth output to a reliable level;
- Developing some storage to further compensate as solar and wind deployment becomes large enough to need it.
Many studies suggest that the near-term cost of solar or wind variability is about 1 c/kWh or less – not a major factor. This will be true for about the next twenty years or until they are deployed at greater than about 30% of our electricity. This is a lot of electricity, especially if we start using it for electric vehicles – enough to power all our light duty vehicles instead of gasoline.
We can avoid most CO2 and remove our dependence on foreign oil with solar, wind, smarter balancing regions with responsive natural gas turbines, and electrification of our transport – with what we have now and improvements already in the pipeline.
In the next blog, I’ll examine how much it might cost, and what an affordable but still effective strategy to do it might be.
And we can address the “deniers,” who really are saying, “Sure, we’d love to do this, but it would cost too much.” Because who wouldn’t want to do it, if it didn’t cost too much?
Ken Zweibel
Senator Feinstein has expressed what a lot of people feel by making an effort to introduce a bill to protect a large area of the US Southwest from solar development (http://s.nyt.com/u/vqR and http://www.renewablesbiz.com/article/09/12/worlds-largest-solar-project-prompts-environmental-debate). There is also some indication that she is being sensitive to the need for solar development by indicating support for solar installations elsewhere, including in pre-arranged solar energy zones. It is quite possible that this represents a balanced, even productive approach to solar in the US Southwest. After all, we are talking about a major paradigm shift in how energy is made, so it’s not surprising that new laws are needed.
It’s important to be clear about why solar needs the desert lands. After all, one viewpoint is that there are plenty of roofs and otherwise disturbed land, so why bother with the desert?
For one kind of solar, based on using concentrated sunlight to boil water to make electricity, the desert is the only place cloudless enough for it to work. This is the kind of technology represented by BrightSource and the like (http://www.renewablesbiz.com/article/09/12/california-labor-unions-and-bechtel-construction-company-reach-agreement-build-solar-thermal-facility-0).
For photovoltaics (PV), which can work almost anywhere, the answer is lower cost in the desert:
1. Solar electric prices are inversely proportional to sunlight. The more the sunlight, the more the output, and the lower the solar electric price, plain and simple. The US Southwest has about 50% more sunlight than most of the rest of the country, and that means that prices of solar electricity are 50% higher outside the Southwest. If solar electricity is 12 c/kWh in Arizona, it will be 18 c/kWh in Washington, DC, or New York City.
2. Large systems are cheaper than small systems. Right now, residential systems are about 40% more expensive than larger, commercial rooftop systems; and those are another 25% more expensive than big (multi-MW) systems on the ground.
If you combine these two factors, you can get a factor of at least three between the price of solar electricity for a large system in the desert and a small system in the mid-Atlantic, and worse in really bad solar locations like off Lake Erie or near Seattle. So if PV electricity were to cost 12 c/kWh for a large system in Arizona or California, it might be 36 c/kWh on a house in DC, and 50 c/kWh on a house in Seattle.
In the future, we might expect the range between small and large systems to narrow as the markets mature, but we will likely always see at least a 10% difference between systems in the 1-10 kW range and those above 100 kW; and another similar gap from there to multi-MW systems. This means at least a factor of two between the desert and the majority of systems outside the desert, where sunlight is lower and system sizes more limited.
There are other factors that make this a bit more complex, but do not change the basic picture. Solar from the desert has to be transmitted elsewhere, adding cost. But it is also more dependable than solar in cloudier climates. The use of a smart grid will facilitate integrating either kind of solar, but each with a different set of components and costs.
It is important to note that once strip mining is included, solar uses no more land than coal mining. Solar is about 30 times less land intensive than hydroelectricity (where about 1% of US land is behind dams as artificial lakes), and about 50-100 times less than biomass production. It is our perception and our local preferences that drive these solar land conflicts, not national priorities about land use. Only wind uses land less intensively than solar per kWh.
It is clear that the resistance to large solar installations will be ubiquitous. It will be as bad or worse outside the sunny regions. No one is going to accept cutting down trees to build solar. Everyone will want it put only on roofs and parking lots and similar, and some will resist even that. We cannot solve global commons problems with local preferences.
Solar cannot afford the extra costs. What some may perceive as a small added permitting cost may in fact be fatal to large-scale solar development in the US, and this may already be showing up in cancellations (http://www.nytimes.com/external/venturebeat/2009/12/22/22venturebeat-new-report-is-first-solar-all-hat-no-cattle-55429.html). Solar may use no fuel, but it is only on during the day. Baseload plants are on 90% of the time. Thus, solar cannot defray added one-of-a-kind costs (like permitting and NIMBY delays) as easily as coal or nuclear can. It does not have as many annual kilowatt-hours to offset them with. Solar’s costs per kWh rise about four times faster for each dollar of initial cost.
We must not forget that what we don’t do in our back yard must be done in someone else’s. If we don’t have solar because we drive up its costs by segregating it from sunny regions, we will have coal, carbon dioxide, mountain top removal, and numerous health hazards. As a nation, we must decide. We have learned enough to know that if we are to have economic strength and avoid foreign energy dependence, we have to have something productive in our own backyards.
The combination of our preferences on cost, economic sustainability, and environmental protection will lead to the amount and mix of solar in the Southwest and elsewhere. But let us hope it is done at a national level and not haphazardly, based on local interests alone. For private lands, providing expedited environmental clearance, helping locals understand the value of their sacrifice for accepting big systems nearby, minimizing the impact of such big systems via prequalified acceptances (e.g., the BrightSource use of air instead of water cooling); and for public lands, similar rapid clearance and designated solar zones would go a long way towards balancing national (global) priorities with local ones.
Ken Zweibel
We Are Replacing Current Infrastructure and Incurring Added Costs, Because That Is the Only Way We Can Rapidly Turn Down Fossil Fuels
What has mostly been understated or otherwise ignored is that we are not going to solve climate change simply by adding low/no carbon sources as power plants and cars get old – we are forcibly turning down existing power plants and removing gasoline-powered cars while we add a much larger amount of low/no carbon sources. We are not simply evolving towards a new equilibrium by slowly transitioning the energy fleet as it ages. Instead, we are replacing it while it is still running.
This is so different from how we usually think that we are often at odds over simple issues because we are starting from different sides of this question. For example, we are not “adding new electricity capacity with wind and solar,” as most engineers familiar with the power system might think. We are subtracting carbon electricity, and replacing it with low/no carbon electricity.
We want to reduce the burning of fossil fuels, preferably coal. We can replace coal with intermittent sources like wind and solar and still retain reliability for when the wind and solar are not available by keeping the coal plants for when they are needed. In utility terminology, we are retaining capacity (through coal plants) while reducing carbon (by turning them down and replacing the lost energy with solar and wind).
Until we reach a point where we have so much wind and solar that we cannot operate reliably with existing fossil fuel back up, we have no major issue with solar and wind intermittency. After that, we have to add sufficient storage or new flexible fossil fuel backup to add more wind and solar and continue to avoid more carbon emissions.
Meanwhile, we will face a choice about coal. Coal is a big problem in terms of carbon emissions. Critics argue that must continue using it, or use it with carbon sequestration because we have few alternatives. But the recent growth in availability of natural gas may obviate this argument.
Switching baseload electricity from coal to natural gas would cut emissions in half. But it also has a salutary effect on the use of wind and solar. Natural gas is more easily turned on and off, which allows greater flexibility for adding intermittent wind and solar. It is very compelling to imagine us switchimg to natural gas baseload electricity generation while we are scaling up wind and solar, because it will allow us to include more wind and solar while also reducing carbon intensity from our baseload plants.
In parallel, we should develop electric storage via demonstrations of existing technology and through enhanced R&D. There are two clear economical paths to the bulk electric storage that would be suitable to the challenges of wind and solar. One is the thermal storage relatively easily included in solar thermal electric facilities. The other is to have wind and solar electricity drive a compressor to store high pressure air in large caverns or aquifers. On demand, this air can then be allowed to expand through turbines, with (but eventually without) the help of burning a small amount of natural gas.
Both thermal and compressed air storage are substantial and efficient enough to go a long way towards satisfying the needs of intermittent solar and wind for bulk energy storage as they accumulate past some level, perhaps 30%, on our electricity grid. Other storage is being examined and will play a part: e.g., pumped hydro, flywheels, batteries. Even transmission from daylight or windy areas further and further away can assist grid stabilization as long as transmission losses are contained.
The electrification of transportation will evolve in parallel and will be significant for both climate change and reduced oil dependence. Electricity is already cheaper than oil to move cars (10 c/kWh is about 1 dollar per gallon equivalent per mile). What we do not have is the batteries and then the cars to use this electricity. But as gasoline cars are replaced with those that can use electricity, we will use somewhat more electricity (much of it from wind and solar), but save both oil and carbon emissions (and perhaps driving costs). But it will be another case of replacing “before its time” infrastructure that works (oil and cars) and is mostly paid for, with a new infrastructure that has more sustainable characteristics.
This replacement of infrastructure “before its time” is the added cost that is being talked about when climate change is being discussed. It is not the cost most people think – the cost of the new technologies versus the old ones, on a steady-state basis. That cost will be fairly small, even negative, and is largely unpredictable due to technology progress uncertainties and fuel price uncertainties. Instead, the cost that stands out is the one of making the fast change from today’s equipment to the new equipment.
We are replacing current infrastructure and incurring added costs to do so, because that is the only way we can rapidly turn down fossil fuels.
Ken Zweibel
How can we be so profoundly behind in our awareness of solar PV? China signs an agreement with the world’s largest PV company (which just happens to be an American company) for the world’s largest PV system (equivalent to Hoover Dam in output) using the most advanced, lowest-cost technology, and we haven’t even heard about it? The company, the technology, the concept of big PV. All that is new. Our press and our government are in the dark. Why?
We hear about self-promoting Silicon Valley PV start-ups manipulating the press for coverage while they raise money (First Solar is from the Rustbelt). We hear about Chinese silicon PV companies using low-cost labor to take the market away from everyone, because that is a cliché of our psyche – the foreign threat.
And we hear about how hapless our government-run applied research is, yet First Solar came from just such a program, started during the Carter Administration out of the National Renewable Energy Laboratory.
We are surprised by progress in solar because we always look for averages in the energy field. We don’t look for disruptive progress because our energy analysts are conditioned by fossil fuels and utilities to use averages. Nothing ever changes, so why look for change?
We think America is helpless versus the Chinese megalith, but here it is, buying from us. We don’t learn lessons from what we have done better, we ignore it or never hear of it because we are looking the other way, far across the sea toward our fears.
Maybe it is because we don’t really need the lessons of how to compete, to innovate, to be committed and visionary – we already know how.
PV Tech, perhaps the best web newsletter out there, and Tom Cheyney (ditto for kudos for him, individually), had a valuable article on Solibro’s CIGS status and plans. For the first time, it seems that CIGS could be becoming mainstream and predictable. Solibro is at 45 MW of production and is finishing 90 more, for 135 MW total in 2010. Their modules are at about 11% (like First Solar’s CdTe) but aim to surpass 12%, and may even reach 15% with steady progress in a five or so years. (These are very similar to First Solar’s recent projections, too.) In the past, such projections from the CIGS start-ups were purely speculative. Now at least we are getting a firmer set of actual manufacturers.
The other steady producer is Wurth, also evaporation of the elements onto glass. This is the traditional method developed over the last 25 years in the US and Europe, with the National Renewable Energy Lab in Golden, CO, still the world record holder at almost 20% cell efficiency. Lars Stolt works at Solibro, and he is one of the long-term contributors to the CIGS technology.
Global Solar of Solon makes CIGS on flexible foil; Ascent had recent success making integrated CIGS modules on flexible plastic; and a few others such as Showa are working on glass.
Interestingly, we have not seen the new breed of VC-backed companies emerge from the tunnel yet. Similarly, the ones with tangible production and expansion are working on glass. Ten years later, all the famous talk in the late 1990s about the new breed of CIGS start-ups was either the ignorance of new entrants or the spin of those raising money.
Solibro didn’t tell us what cost they are making CIGS modules for now, but they told us their goals of about 0.8 euros/W. I will guess this means about $1.1/W based on a wildly fluctuating dollar-euro price (of course, it is closer to $1.2/W now). It wasn’t clear what their projected efficiencies assumed were, but let’s take a round number 12% for comparison sake, and $1.3/W for right now.
Using our own spreadsheet estimates (and uniform 30% module margins and near-10% integrator margins), we can stack up Solibro’s CIGS modules against the field and see how they rate. These are engineering estimates, and not purchasable prices – they do not account for the margins and middle men between us and our hardware. But at least they are on a common basis, which allows for comparisons.
Before looking at their goals, let’s look at Solibro using their assumed efficiency of 11% and $1.27/W total, today. This may be too kind to them (perhaps in terms of yield and current equipment costs).
Figure 1. Comparing technologies at the large-ground mounted system level using today’s assumed Solibro numbers: 11% modules and $1.27/W direct cost.
Note that they are only competitive against amorphous silicon on glass using these numbers. Now let’s look at their goal of 12% and $1.1/W.
Figure 2. Solibro’s goals: 12% modules and $1.1/W. Still not competitive with either today’s CdTe or better silicon producers, but closer.
In the past, lack of cost competitiveness was not as much an issue, as every module made was sold, and those who were cheaper had larger margins. Today it is different. Still, these are good numbers for an emerging technology, and it would be healthy for consumers if markets were robust enough to support the continued progress of CIGS to keep up the competitive pressure on the incumbents.
Dr. Stolt made the remark that 1% of module improvement was worth as much as halving the capital cost. This actually allows one to back out their capital costs. 1% module improvement from 11% to 12% is a 1/11th , or 9% cost reduction. 9% of their area cost is about $12/m2. This means their capital cost is about $24/m2, which at 11% and 7 years depreciation is about $1.5/W. This seems like a reasonable assumption for their factories, now that they are being simplified. It used to be $3/W, so this is big progress.
Is CIGS turning the corner. Alright, maybe it is still a little too soon, but it is good to see them reach a point where it makes sense to actually calculate their place in the competitive landscape.
It does not have to be mysterious anymore what the US, and by implication, the world can do about climate change and peak oil. It is to deploy the appropriate amount of wind, solar, and electric transportation. With this strategy, we have the knobs for all the results we want: less and less carbon dioxide, and reduced need for oil. What more do we want?
Naturally, we must answer two key questions:
- How much would it cost?
- How do we deal with wind and solar intermittency?
It used to be that wind cost too much and solar cost way too much. Those days are gone. Now wind costs about the same as new coal plants (which is to say, as little as anything to make electricity), and solar costs (depending on local sunlight) only about half again more. (Of course, in less sunny places, solar prices go up significantly. This is why you hear so many different economic numbers quoted for solar. Small systems are also significantly more expensive than large ones, although most of this is the cost of middle-men and not hardware.)
There should be little resistance to implementing a great deal of solar, starting in sunny places, because:
- It will sustain continued cost reductions in solar (which have been documented over about 30 years as about 20% per doubling in production); and
- Solar electricity, by being produced during the day, is worth more than wind-electricity, which is produced usually over 50% at night (this means the economic disparity between the two is actually marginal in terms of customer value).
We have few solar and wind intermittency challenges in America and other developed countries, because we have a power grid capable of filling in for solar and wind when they are not available. That will work to about 20% of our electricity (which is a lot, even enough to power all our light duty vehicles, if they were electric).
Electric vehicles and mass electric transport are key to moving away from fossil fuels. Electricity is cheaper than gasoline as a “fuel” (one dollar a gallon is about the same as 10 c/kWh, the retail rate in most places), so we would get an immediate economic benefit, if we had the batteries. Everyone knows that batteries are still a challenge, and we ought to move forward with their development expeditiously.
In parallel, we ought to also expedite the development of alternative electric storage methods for the period after we can use existing backup. That may be about twenty years from now, so we have time if we work hard.
What will this cost? Those with other preferences often try to portray it as breaking the bank. This is far from the truth. The economics of wind and solar are similar to existing options, now, and will get better with further development. Conventional fuel costs will rise if we wait for further resource depletion. How can we quantify that avoided cost? Or the avoided climatic costs of GHG reduction?
In fact, a major cost of the transition away from carbon fuels is simply the rapid rate of building a replacement set of power plants for the existing ones. It’s not that the electricity from the new plants costs that much more. It’s that we will be building a huge number of solar and wind plants in a short time. That’s the unusual cost of this transition. But how can we avoid it, unless we imagine that conservation by itself is going to get us there? That isn’t going to happen.
Wind, solar, and electric transport are the core technologies for solving climate change and peak oil. They will work in the US, and would work worldwide. The rest (efficiency, geothermal, tides, waves, biomass, others) are icing on the cake. They will be there, but they will not be the kernel.
Why does it matter to have this discussion? Because until we crystallize at least an outline of a plan, as a society we won’t be able to move forward.
![[Image: Smoke Stacks]](http://thesolarreview.files.wordpress.com/2009/11/appollo-beach-power-plant.jpg?w=600&h=400)
Individuals interested in solar energy and climate policy are likely aware that the U.S. Environmental Protection Agency (EPA) published its final Mandatory Greenhouse Gas Reporting Rule in the Federal Register on October 30, 2009. (74 Fed. Reg. 56260) This regulation represents the first U.S. effort to require public reporting of certain greenhouse gas (GHG). However, few of these observers may be aware that the final rule will not require the tracking of progress by electricity consumers in reducing greenhouse gas (GHG) emissions by substituting on-site solar energy for purchased fossil fuel-fired electricity.
Nonetheless, a close reading of the final EPA rule indicates that solar energy supporters should not pack their bags and go home. read more…
