<< ^
San Bernadino


Copyright 2008-10,
V 1.06
site last updated: 6/26/2010

C a l i f o r n i a   P H O T O N

Can Rooftop Solar Satisfy Most of Humanity's Energy Needs?

At an average flux of 89 petawatts, the solar energy reaching the earth's surface dwarfs all other sources of energy on earth. At 15 terawatts, the average rate of energy use by civilization is 6,000 times smaller. 1   Solar energy could easily supply all of human energy needs, replacing all fossil fuels, given the requisite technological conversions and allocation of space in the sun.

A more interesting question is: Could solar energy replace fossil fuels without introducing a huge environmental impact of its own?

Two factors prominently figure into such an impact. One factor is the cost of manufacturing solar energy systems. Conventional silicon PV panels, for example, have considerable embedded energy costs, and thereby have energy payback times of, typically, two to four years. 2   Possible solutions to such impacts are discussed below.

A second factor in the environmental impact of a worldwide conversion to solar power is the space occupied by solar energy installations over time. The use of space that is not already occupied by man-made structures is an environmental impact, if for no other reason that it removes lands from their natural state. Hence the main motivation for the question posed in the title: if solar energy systems can be confined to the surfaces of buildings, they will have essentially zero land-use impact. Aside from the massive benefits of displacing fossil fuels, rooftop solar could improve urban climates by reducing the heat-island effect. Another major advantage of widespread rooftop solar is the distribution of power generation at points of use, reducing the need for power transmission infrastructure.

Recent studies have examined the potential of a full-scale deployment of rooftop solar in particular nations, if not the world. A 2005 study by Navigant Consulting concludes that covering about half the rooftop space in the US could replace the use of coal for electricity generation. 3   A less detailed analysis of England's rooftop resources concludes that the yearly production of rooftop PV panels could exceed the nation's electricity use. 4   5   Both of these studies posit PV conversion efficiencies of below 20 percent, and neither considers the use of thermal or hybrid PV/thermal panels. In the present analysis, I will explore a worldwide conversion scenario using solar panels which, although having much higher efficiencies than today's, are feasible without fundamental technological breakthroughs.

Estimating Energy Production From a Worldwide Conversion to Rooftop Solar

In the following, I will outline a set of assumptions supporting a back-of-the-envelope estimate for the quantity of energy that could be produced by pressing into service the rooftops of the world, using nearly half of that space for solar energy capture. After reviewing those assumptions in some detail, I will proceed to examine key developments needed to transform the vision of worldwide rooftop solar into a reality.

The following assumptions are made for the year 2015.

  1. A worldwide mean human energy use of 15 terawatts
  2. A world population of 7 billion people
  3. A mean rooftop area per person of 15m2
  4. A mean solar panel rooftop coverage of 40 percent
  5. A mean solar panel load factor of 20 percent
  6. A mean solar panel conversion efficiency of 50 percent
  7. A mean solar flux during operating hours of 1000 watts/m2

Multiplying the last six items, starting with the last, gives:

1,000*0.5*0.2*0.4*15*7,000,000,000 = 4,200,000,000,000

The result, 4.2 terawatts, is less than 1/3 of the 15 terawatts currently being consumed, but several facts are worth noting:

  • 4.2 terawatts is similar to the magnitude to the worldwide energy usage for electricity production, and similar to that of the energy generated from coal -- the most environmentally damaging fuel for a host of reasons, from greenhouse gas emissions to mercury emissions to strip mining.
  • Because of inefficiencies in the production and distribution of electricity, only about half of the approximately 4 terawatts used in its production is delivered to the end users. In contrast, the vast majority of solar energy produced by rooftop solar is used on location or nearby, avoiding those losses.
  • The estimate that human energy usage averages 15 terawats is based on the current energy economy in which there are huge inefficiencies due to the low prices for fossil fuels -- prices that don't capture their environmental impact. As an indicator, the average energy use per person is 12,000 watts in the USA but only 3,000 in the Switzerland, 6   a nation with a higher standard of living than the USA. 7   8   Switzerland has a goal of reducing that number to 2,000.
  • Trends of increasing efficiency in energy consumption are likely to accelerate, driven by increasing prices for fossil fuels and enabled by new technologies such as LED lighting, hybrid vehicle drive trains, and intelligent building climate control.
  • Continued growth in world population and per capita energy consumption will be paralleled by increasing roof area and therefore the resource of rooftop space for solar generation.

Now let's examine the assumptions underlying the above calculation, starting with the third.

Mean rooftop area per person: 15m2

Published studies of rooftop area are available for some regions, but apparently not for the entire world. However, rough estimates of global rooftop area per person can be inferred from other data. Many publications provide estimates of floor area per person, because of its value as a measure of economic development. One estimate of average global rooftop area per person puts the median value at 14.1 m2 in 1970. 9   The mean value is likely significantly higher because the distribution is asymmetric, with a 'long tail' at the upper end. The mean value in the US, for example, approaches 80. 10   Also, that value is increasing over time, both in the developed and the developing regions of the world such as China. 11   12   Given that, a value of 20 m2 per person in 2015 is plausible.

Inferring rooftop area from floor area requires making additional assumptions, such as a mean ratio of floor area to roof area in residential buildings. Two seems like a reasonable guess for that ratio, given that the vast majority of residential structures are single-story. According to the Navigant study the ratio in the US is 1.375. Using a ratio of two, my estimate of worldwide mean per capita residential rooftop area is half of 20 m2, or 10 m2.

Next, the rooftop space of non-residential buildings is considered. According to the Navigant study, total rooftop area in the US divides almost equally between residential and commercial, where commercial includes government buildings. It seems likely that about the same ratio holds throughout the world. For simplicity I assume that the two sectors have equal rooftop area.

Mean solar panel rooftop coverage: 40 percent

The suitability of rooftop space for solar panels varies according to a number of factors, including roof pitch orientation, shading by trees and other buildings, rooftop obstructions such as HVAC equipment, and local climate. Due primarily to these factors, the Navigant study estimates that 65 percent of the total commercial and 22 percent of the total residential rooftop space is suitable for coverage by solar panels. Given the assumption that residential and commercial rooftop space are equal, I adopt the estimate that 40 percent of total rooftop space is suitable for coverage.

Mean solar panel load factor: 20 percent

Because solar panels only operate in daylight, and because they operate near their peak efficiency only when the sun is high in the sky, I assume a mean load factor of twenty percent. That corresponds to a solar panel operating for four hours and fourty-eight minutes per day and at its peak power. This assumption is meant to account for large differences due to season, local climate, and requirements of particular types of solar energy systems.

Mean solar panel conversion efficiency: 50 percent

This may seem like very optimistic assumption, given that current PV panels have conversion efficiencies of at most 20 percent. However, a concentrating photovoltaic (CPV) panel could potentially harvest nearly 40 percent of the sunlight's energy, and much greater aggregate efficiencies are possible by combining PV and thermal in a concentrating panel. Although the thermal energy could only be used in close proximity to the panels, It could supply heat for climate control and industrial processes, offsetting one of the largest consumers of fossil fuels, and be used to drive heat engines to generate additional electrical energy.

Mean solar flux during operating hours: 1000 watts/m2

This is the reference illumination used to test solar panels, and corresponds, approximately, to the energy reaching a square-meter-sized aperture perpendicular to the sun's rays at sea-level.

Implementing a Worldwide Conversion to Rooftop Solar

The transformation of the world's roofscapes required to meet the goal of supplying energy at an average rate 4.2 terawats will be a vast undertaking. However, the project is less far-fetched than it may appear, for a number of reasons described below.

One reason the project appears so formidable is that, in the present energy economy, the infrastructure and costs associated with energy production and distribution are largely hidden from the populace. Drilling for petroleum and mining for coal and uranium happens far away from most of us (though the scarring of the land is increasingly apparent as one scans the landscape from a trans-continental flight at 35,000 feet). And the bulk of the costs of fossil fuels are being deferred to the future, which may see flooded coastal cities worldwide by the end of the century, among other catastrophes.

In contrast to the status quo, the solar energy economy will be visible as an architectural element of most buildings. Energy production will no longer be out-of-sight and out-of-mind, nor will energy consumers be at the mercy of faceless corporations. Of course that doesn't mean that the rooftops of places like Florence need to be covered with solar panels -- the vast majority of rooftops worldwide are nondescript surfaces that are aesthetically compatible with solar panels. Furthermore, solar energy devices can be designed to blend in with architectural elements.

But what of the economic barriers to the conversion to the solar rooftop energy economy? I consider three key objectives in turn:

Unprecedented Economies of Scale

Before looking at these objectives, I make a general observation that that has bearing on all of them. Because the scale of the project will be so vast, it will afford huge economies of scale. Imagine the drivers of technological innovation and the optimizations of production methods in a project involving the production of 100 billion solar panels!

The economies and efficiencies of scale may be much larger than indicated by the mere number of solar panels needed to produce an average output of 4.2 terawatts. Solar panels, unlike automobiles, are relatively simple devices with few parts replicated in large numbers. A single CPV panel might contain a total of 1000 parts but only 20 unique parts. Also, because the panels are so simple in construction, small factories for producing all but the most advanced components of the panels might be mass-produced and geographically distributed to minimize transportation costs.

Reducing the energy payback time of solar panels

A key metric of solar panels is their energy payback time (EPT) -- how long they have to operate to recoup the energy that went into their manufacture. Energy of manufacture is, perhaps, the dominant factor of solar panel cost. Crystalline solar panels typically have EPTs, depending on their siting, of two to five years. Various types of thin-film or amorphous panels have EPTs that are about half that of crystalline silicon panels, but also have much lower efficiencies, and some use elements that are toxic and have larger mining footprints than silicon. Although most solar panels have a net energy that is several times the energy of their production, access to capital to finance the acquisition of equipment remains a principle barrier to rapid conversion.

One solution is the use of CPV systems that use tiny quantities of photovoltaic materials, rendering EPT primarily a function of the manufacture of the optics and tracking systems -- components whose energy footprint has a great deal of downward elasticity.

Also, Solar thermal panels, or the thermal component of hybrid panels, tend to have much lower EPTs than comparable PV panels.

A hybrid CPV panel could easily have a fraction of the energy of production and several times the aggregate efficiency of today's panel's, reducing the EPT by an order of magnitude, and slashing the capital costs of equipment acquisition.

Achieving the required conversion efficiencies

As noted above, there are a number of approaches that may yield solar panels with dramatically greater efficiencies than today's. A hybrid CPV/thermal panel could provide an aggregate efficiency of greater than 70 percent -- assuming thirty-five percent efficiency of each of the PV and thermal subsystems -- using *today's* triple-junction PV cells. Future advances in high-efficiency PV cells will allow a greater fraction of the harvested energy to be in the form of electricity, with its far greater versatility and mobility than thermal energy.

The development of efficient electrical micro-storage methods, such as compact electrolysis and fuel cell systems, will decrease the dependence on the grid for load balancing. Again, such methods do not require fundamental technological innovation.

Creating mechanisms for siting rooftop solar installations given the varied needs and means of the rooftop owners

A key requirement for covering half of the world's rooftops with solar panels will be matching capitial and rooftop resources, given that much of the world's rooftop area is owned by people without the means to invest in the requisite equipment. Although this will require the creation of new economic tools, it is not difficult to visualize such. One example involves the development of leasing arrangements wherein a rooftop property owner permits the solar development of their asset in return for receiving a portion of the energy produced by the equipment. 13  

Overcoming entrenched interests wedded to the centralized energy economy

Much as the worldwide conversion to rooftop solar as the world's primary energy source seems daunting at first glance because of its scale, so too, the transformation of the energy economy from a centralized to a distributed model may appear to entail insurmountable political obstacles. These perceptions may be reinforced by the failure of the world's governments to forge meaningful agreements to limit carbon emissions.

However, although proactive measures by governments and large corporations can be very effective in fostering such transformations, they may not be necessary to a revolutionary change in the energy economy given the right technical solutions combined with a sufficient breadth of enlightened self interest. Policies favoring distributed renewable energy production -- such as feed-in tariffs, investment tax credits and subsidies, and production credits -- are often implemented as temporary measures to foster the development of renewable technologies and accelerate improvements in their economy and efficiency. As solar panels approach "grid parity", the driving incentives for conversion will shift from government policies to underlying economics.

Recent decades have seen numerous disruptive technologies succeed spectacularly without major subsidies and despite displacing powerful entrenched interests.

Will the Exxons and Shells of the world attempt to thwart the conversion to distributed renewables? Almost certainly. The petroleum companies are making large investments in the emerging biofuels sector, where they can leverage their existing infrastructure for refining and distributing liquid hydrocarbon fuels. Anyone who watches PBS has seen their public relations efforts to paint themselves as the saviors of humanity. One reading of such efforts is that they indicate a fear on the part of the energy giants that they stand to lose their stranglehold of the energy economy.


Solar energy is often dismissed as being "too diffuse" to be a practical replacement for fossil fuels as humanity's primary energy source. When such a role for solar energy is considered in the mainstream media, it is usually within the context of the centralized energy paradigm, where vast solar farms in the desert harvest the sun's energy and pipe to to distant consumers via a greatly upgraded transmission infrastructure. For example, a 2009 New York Times article purporting to illuminate the future of energy with many statistics about the land-use footprint of renewables reads as if rooftop solar panels don't exist. 14   Similarly, the January 2008 issue of Scientific Amercian outlining "A Solar Grand Plan" ignores a role for rooftop solar. 15  

As the above analysis shows, relatively straightforward estimates suggest that rooftop solar could supply about one third of humanity's energy needs, mostly replacing other forms of electricity production such as the mining and combustion of coal. With a re-toooling of the energy economy to remove large sinks of wasted energy, it is conceivable that rooftop solar could become the primary energy source for civilization.

The short shrift given to distributed rooftop solar by influential media organizations may reflect a loyalty to entrenched power structures or just a failure of imagination, but it seems irresponsible given the magnitude of the environmental impacts at stake. The quest for renewable energy doesn't need to be at the expense of wild lands and habitat, given the vast untilized rooftop resources ideally situated for supplying power to the very buildings they cover. Rooftop solar stands out as easily the most environmentally benign solution to humanity's energy needs, and one that is despirately needed to avert a catastrophic future promised by continued reliance on fossil fuels.


1. Photovoltaics: The Future Energy Source,,
2. Solar PV Environmental Impact,,
3. PV Grid Connected Market Potential under a Cost Breakthrough Scenario,,
4. A second look at solar power on roofspace,, 2009.03.08
5. How much solar power will fit on Britain’s roofs?,, 2009.02.25
6. Switzerland and the 2000-Watt Society,,
7. List of countries by GDP (nominal) per capita,,
8. World's Richest Countries,,
9. The mega-urban regions of Southeast Asia, , page 124
10. Giant American houses: another symptom of overconsumption,,
11. International Energy Outlook 2009,
12. Climate resilient urban infrastructure in China – Insights into the buildings sector,
13. Leasing America's Rooftops for Solar Energy,, 2009.01.27
14. Illuminating the Future of Energy,, [cached]
15. A Solar Grand Plan,,