Power Density Primer: Understanding the Spatial Dimension of the Unfolding Transition to Renewable Electricity Generation (Part V – Comparing the Power Densities of Electricity Generation)
By Vaclav Smil -- May 14, 2010Editor’s note: This is the conclusion of the series that provides an essential basis for the understanding of energy transitions and use. The previous posts in this series can be seen at:
Part I – Definitions
Part II – Coal- and Wood-Fired Electricity Generation
Part III – Natural Gas-Fired Electricity Generation
Part IV – New Renewables Electricity Generation
America’s dominant mode of electricity generation is via combustion of bituminous and sub-bituminous coal in large thermal stations. All such plants have boilers and steam turbogenerators and electrostatic precipitators to capture fly ash, but they burn different qualities of coal that may come from surface as well as underground mines, have different arrangements for cooling (once-through using river water or various cooling towers) and many have flue gas desulfurization to reduce SO2 emissions. Consequently, these conversions of chemical energy in coal to electricity feature widely differing power densities: for the power plants alone they are commonly in excess of 2 kW/m2 and can be as high as 5 kW/m2. When all other requirements (coal mining, storage, environmental controls, settling ponds) are included, the densities inevitably decline and range over an order of magnitude: from as low as 100 W/m2 to as much as 1,000 W (1 kW)/m2.
In contrast, compact gas turbines plants (the smallest ones on trailers and larger facilities that can be rapidly assembled from prefabricated units), which can be connected to existing gas supply, can generate electricity with power density as high as 15 kW/m2. Larger stations (>100 MW) using the most efficient combined-cycle arrangements (with a gas turbine’s exhaust used to generate steam for an attached steam turbine) will operate with lower power densities, and if new natural gas extraction capacities have to be developed for their operation then the overall power density of gas and electricity production would decline to a range similar to that of coal-fired thermal generation or slightly higher, that is in most cases to a range of 200-2000 W/m2.
Photosynthesis is an inherently inefficient energy conversion process, and production of biomass has large space requirements. Even with an intensively cultivated plantation of fast-growing trees, a wood-burning electricity generation plant would not have power densities higher than 0.6 W/m2, and for most operations the rate would be below 0.5 W/m2. Space demand for such facilities, then, would be two to three orders of magnitude (100 to 1,000 times) greater than for coal- or gas-fired electricity generation.
Photovoltaic plants can generate electricity with much higher power densities than wood-burning station — converting solar radiation to new biomass has overall efficiency no better than 1% while even relatively inefficient PV cells have efficiencies around 5% and today’s best commercial facilities go above 10%. Taking only the PV cell area into consideration, this translates to power densities of mostly between 10-20 W/m2. But when all ancillary space requirements are included, the typical density range declines to 4-9 W/m2, an order of magnitude higher than for wood-powered generation but one to three orders of magnitude lower (that is demanding 10 to 1,000 times more space) than the common modes of fossil fueled electricity production.
Power densities for central solar power are slightly higher, with rates as high as 45-55 W/m2, when only the area of heliostats is considered, but with overall power densities (including spacing, access roads and the tower facilities) on the order of 10 W/m2. Finally, wind-driven electricity generation has power densities similar to, or slightly higher than, wood-burning stations, with most new installations using powerful (1-6 MW) turbines fitting into a range between 0.5-1.5 W/m2.
| Power Source | Power Density (W/m2) | |
|
Low |
High |
|
| Natural Gas |
200 |
2000 |
| Coal |
100 |
1000 |
| Solar (PV) |
4 |
9 |
| Solar (CSP) |
4 |
10 |
| Wind |
0.5 |
1.5 |
| Biomass |
0.5 |
0.6 |
Implications of these differences are manifold. Changing the power density-determined infrastructure of energy systems that were created over more than a century for electricity generation from fossil fuel combustion will not be easy. A fossil-fuelled civilization has been securing the supply of its most flexible form of energy by “shifting downward,” that is by generating electricity with power densities 1-3 orders of magnitude higher than the common power densities with which electricity is used in buildings, factories and cities. In a civilization that would rely only on renewable energy flows, but that would inherit today’s urban and industrial systems, we would produce electricity at best with the same power densities with which they would be used –- but more often we would have to concentrate diffuse flows of solar radiation, wind, and biomass in order to bridge power density gaps of 2-3 orders of magnitude.
This new energy infrastructure would increase fixed land requirements and preempt any other form of land use in areas devoted to PV cells, heliostats or fast-growing wood plantations. Most of the area occupied by large wind farms could be used for crops or grazing but other land uses would be excluded, and large areas dotted with wind turbines would require construction and maintenance of access roads as well as the creation of buffer zones not suitable for permanent human habitation. And in all cases of renewable energy conversion, much more land would be needed for more extensive transmission rights-of-way in order to export electricity from sunny and windy regions, or from areas suited for mass-scale biomass production, to major urban and industrial areas.
As a result, these new energy infrastructures would have to be spread over areas ten to a thousand times larger than today’s infrastructure of fossil fuel extraction, combustion and electricity generation: this is not an impossible feat, but one posing many regulatory (environmental assessments of affected areas, rights-of-way permission and inevitable lawsuits), technical and logistic challenges. Higher reliance on renewable energies may be desirable (mainly because of perceived environmental and strategic reasons) and technical advances would also make it an increasingly appealing economic choice –- but inherently low power densities of these conversions will require a new system of fuel and electricity supply that will be able to substitute for today’s dominant practices only after decades of gradual development.
Power densities of fossil fuel extraction, thermal electricity generation and renewable modes of electricity production. Reproduced from: V.Smil Energy Transitions: History, Requirements, Prospects (Praeger: 2010).
I was really hoping that this series would include nuclear.
Including nuclear here would have very dramatically invoked issues of scale that would have made comparison seem silly, given that the power density of nuclear is millions of times more than natural gas. Even more interesting would have been a comparison with hydro, by far the most effective “renewable”–indeed the renewable of choice after so much forest was clearcut a hundreds years ago, and it continues to remain so today.
Dr. Smil documents quantitative land use; he does not discuss here qualitative land impacts produced by the various power sources, which is a subject for another examination. And qualitative assessments are much more subjective and volatile. Still, readers here should be encouraged to think about them.
We already know what a civilization looks like that relies for its electricity production only on renewables like wind, biomass, and solar. It’s called the pre-modern past.
I, for one, fail to see how the standards of performance of modernity can be maintained principally by wind and solar, even with the Smart Grid and buckets of PR. Neither produce capacity, which is the cornerstone of the reliability, affordability, and security of our electricity supply. And wind’s variability, given the prerequisite of continuously matching supply with demand, would require it to be entangled with some kind of high capacity, flexibly responsive generation, even if it’s well trained hamsters at the wheel.
There is an enormous difference between four 9s reliability and “most of the time when the sun is out or the wind is blowing”.
That difference is admittedly far more obvious to utility executives than to solar and wind advocates, who appear to believe that the utilities should operate for their convenience, rather than the convenience of the utilities’ customers. So far, the state utility commissions and consumer advocates have stayed focused on reliability and low rates.
The state regulators in Maryland, Pennsylvania, New York, and Massachusetts, Ed, have hardly stayed focused on reliability and low rates. Quite the contrary, despite all the posturing. FERC continues its pretentious ways, playing footsie with wind’s tail wagging the dog technology, which actively subverts mission on behalf of reliability, affordability, and security of the electricity supply.
[…] diesem Zusammenhang ebenfalls sehr empfehlenswert ist die Lektüre einer Beitragsserie über das Problem der Energiedichte von Vaclav Smil (Power Density Primer) auf dem Weblog […]
coal,nuclear,wood, solar, thermal.hydroelectric?
Why try to sell a dead horse. Geoffery Ballard finished in 1991. The entire universe is 70% hydrogen. 1 billion cars need to be recharged=1 trillion tons of battery acid. What’s to understand?
[…] Vaclav Smil is one of the leading energy scholars of our day. He has, time and again, tried to inject energy reality into energy fantasy. Some of his previous posts at MasterResource (see here) include ‘The Limits of Energy Innovation’: Timeless Insight from Vaclav Smil and the five-part Power Density Primer. […]
The main deficiency in this analysis is the neglect of the space required to “deposit” the effluents of fossil fuel plants. Common wisdom is that (approximately) a doubling of the natural carbon content of the amosphere is just still acceptable as an environmental effect (just as wind turbines are considered a “somehow acceptable” change to the landscape. Similarly, hilltop clearing is considred an acceptable result of coal mining by its proponents, as is the open pit mining in your examples. Covering an area with solar panels or with fast-growing trees also will not destroy it competely, but alter its natural value). Neglect the differences in effect for each square meter of affected area for now.
For the coal plant, every year an additional area will thus be needed, where the natural carbon content (in tons) of the part of the atmosphere above it is (or was) equivalent to the annual effluent of the coal plant. Maybe less, if the oceans take up some part of the carbon, maybe less, if the oceans will release additional carbon after heating up.
If you add this dumping area for fossil fuel effluents into the calculation, you will easily see that wind and solar need mch less space per energ produced.
Estimate of aggregate power density of coal plants: Factor around 500 to 1000 for solar PV
I am referring to Prof. Smil’s comparisons of areas required for different kind of energy sources as also in
http://www.vaclavsmil.com/wp-content/uploads/docs/smil-article-power-density-primer.pdf
The power density for wind and solar power is significantly higher than for fossil fuel, if the space was added that is required for a damaging, but not devastating dumping of the main end product.
For carbon dioxide, a doubling of its natural atmospheric content is considered as damaging, but still acceptable as a rule of thumb; thus on the basis of the carbon content per square meter of the air above us, the required dumping space can be determined.
The natural carbon dioxide content is about 4 kg/square meter. Assuming one kg of carbon dioxide effluent per kWh for an existing coal power plant, 8766 hours/annum, 80% capacity factor, the space requirement is 8766 h/a * 80% * 1 kg/kWh / (4 kg/m2) = 8.766 *0.8 / 4 m2/(W*a) = 1.75 m2/ (W*a). This is the area where a doubling of atmospheric carbon content would occur, if no mixing would take place. Given that we will probably “achieve” doubling, the maths of extrapolation fits well.
The space efficiency is the inverted value, i.e. 0.57 W /(m2/a).
This looks like about the same as the efficiency for Wood-Fired Electricity Generation that you determined at about 0.6 W/m2, but 0.57 m2 is actually the figure of deposit space required for one year of 0.8 Watt power production only. Taking the heroic assumption that ½ of the carbon would have dissipated into the ocean after 50 years (I estimate that value in lack of more precise figures at hand), it would be reasonable to divide the figure by 100 for a power efficiency of coal of 0.0057 W/m2 (but of course, comparing an annual space requirement with a permanent use of space is always problematic, higher ratios of dissipation to the ocean take much longer).
As a result, planting forests for power production is about 100 times more space-efficient than using coal without CCR.
The additional space required for mining will not change the prior value significantly; I would note your result for the 15 m seam as 4,814 W/(m2/a) , as this area is used for mining every year. For a 50 year lifespan of power plant and mine, this would go down to around 100 W/m2 (however, the assumption of full rehabilitation of the land area after 50 years would be quite optimistic). In summary, coal mining is around 12,500 times more space-consuming if the area for not-yet-devastating disposal of effluents is taken into account.
Considering that 15*1.4 = 21 tons of coal are present in each square meter of your model mines, i.e. 5,000 times more than the carbon dioxide content of the air, and each carbon molecule reacts with two oxigene molecules of roughly similar mass, the result of 12,500 is plausible.
The space for the power plant is used permanently during operation, on the other hand, so that the corresponding figure of 1000 W/m2 is of minor relevance anyway.
Solar PV efficiencies also have improved a little since your paper, as you forecasted, 15 W/m2 is reasonable today. Given the lower full load hours, that should be translated to up to 5 W/m2 for comparability. For low-irradiation areas like Germany or Alaska, half of that. Solar PV thus allows for a 500 to 1000 times more efficient use of space compared to coal power plants without CCR and with conventional efficiencies.
For natural gas, it would be interesting to know how large areas of underground have to be fracked into pieces in relation to power output.
[…] compared to government-enabled wind power, (on-grid) solar power, and ethanol. Fossil fuels are energy dense and have built-in storage; they are thus affordable and reliable (non-intermittent) compared to […]
[…] compared to government-enabled wind power, (on-grid) solar power, and ethanol. Fossil fuels are energy dense and have built-in storage; they are thus affordable and reliable (non-intermittent) compared to […]
[…] scholar, has prolifically educated about energy reality. (See, for example, his five-part post on energy density at MasterResource.) While stopping short of advocating an end to renewable energy subsidies and […]
[…] and oil reserves, all three can be found in most parts of the world (ref. here and here and here). Even better, America possesses the best technology to extract all three economically, […]
[…] gas, and oil reserves, all three can be found in most parts of the world (ref. here and here and here). Even better, America possesses the best technology to extract all three economically, […]
Gentlemen,
I have read through all the five parts of this presentation and the comments. There is very little discussion of Hydro Electric generation of power. The only illustration is the chart in this part:
“Power densities of fossil fuel extraction, thermal electricity generation and renewable modes of electricity production. Reproduced from: V.Smil Energy Transitions: History, Requirements, Prospects (Praeger: 2010).”
I realize the main parameter you are stressing is Land Area Vs power output but for Hydro that is not a realistic comparison. Especially IF you are talking about an existing impound (dam) facility. The example I am thinking about is the Jackson Lake Dam, part of the Minidoka irrigation dam system. A dam at this location and size would be classified as between a small and large facility, the dividing point being 30MW.
Some of the brief estimates for this Hydro generation system could supply from 1MW to 30MW of output that follows the flow of the river and closely matches the demand in the region. The power would be available nearly 99% of the time, would suffer no diurnal or seasonal loss of power, would not require any new dam construction – and more. The flow of the river would dictate the power available to be fed to the adjacent grid, not vice-versa.
Sincerely,
Bob Culver – BSEE – PE (Ret) Jackson Wyoming