The idea of comparing power sources on the basis of their spatial footprint was popularized by University of Manitoba professor Vaclav Smil. We found his book Energy at the Crossroads to be an enlightening read. It presents a fresh perspective and a few new ways tools for looking at power sources. Smil extensively uses the metric (in SI units no less) of how much power can be generated from a square meter of land (W/m^2).
While this metric allows the scientist to bring their depth of experience using these units in different contexts to bear, it has some distinct disadvantages. Power densities leave out large portions of an energy story, such as the reliability of the energy bring produced or the externalities that will go on to effect other areas, such as when pollution spreads.
Robin Whitlock recently wrote a renewable energy article
in which he looks at wind power through the lens of Smil’s land use figures. Reading his article led us at Vision of Earth to an intense discussion of the pros and cons of simplifying our descriptions of power systems down to even as useful of a metric as areal power density. In this piece we demonstrate some problems with using this concept as the sole metric of comparison for power systems.
What is areal power density?
While Smil uses the term ‘power density’, we find this may be confusing because it has other technical definitions already.1 Instead we will be using the term areal power density to distinguish it from the other definitions. Areal power density refers to the power produced per unit area of land. In SI units this quantity is watts per meter squared (W/m2). Higher power densities are more desirable, because it means that more power can be produced from less land area.
What does this mean? Let us consider some examples for clarification.
If solar panels (along with their requisite maintenance roads and appropriate spacing) cover an area of 100m2 of land, and produce 1000 watts of power on average (over a year), we would say that this solar installation has a areal power density of 1000 watts per 100m2 or equivalently: 10 W/m2.
In practice, the sun shines with varying intensity every day, and over the day; depending on factors such as the weather, latitude, and season. In considering intermittent power sources, Smil and VOE will be considering the yearly averaged power output. When the full potential power output (or nameplate capacity) is discussed, it differs from the actual average power output by a factor known as the capacity factor.
Things get a bit more complicated when we begin to consider other infrastructure and other land required for power generation. In the case of solar power, this is relatively straightforward since we only need to consider maintenance systems and power distribution equipment. These systems take up some area, and would contribute to the total area required to use solar power.
Looking a bit deeper, we could also add the land use for acquiring the materials that go into solar panels.
An additional consideration for solar power is transmission infrastructure. The areas with the most yearly sunlight (such as deserts) may be very far from the places where power is needed most, such as cities. High-voltage transmission lines can carry the power over great distances, but this does require additional land use.
However, for power sources such as coal, we are faced with a more complicated calculation. We need to consider not only the area required for the power plant and transmission lines. We must also consider systems like the cooling water loop, mining facilities, transportation systems, fly ash settling ponds, and fly ash storage.
Coal mining, fly ash, and coal pollution all contribute to a incrementally growing areal footprint. That is, as more coal is mined, more ground is damaged through the effects of mining. Remediated coal mine areas are generally not as productive or healthy as they were prior to the coal mining commenced. Substantial remediation requires a large (additional) investment of time and money.
Fly ash, one of the products of coal combustion, also requires extensive storage. The more coal that is burnt, the more fly ash storage is required unless the fly ash is used for other purposes or is buried in a previously mined coal seem. Currently only about 40% of fly ash is used in other products, meaning that 60% of it is stored.2 This creates a large, and growing, waste disposal issue. For more detail on the broader effects of coal, see our article on coal power’s products and impact.
In short, the power density of coal power shrinks over time, because it impacts a larger area the longer it is used.
Why is this metric useful?
Areal power density is very useful for evaluating the relative merits of various power sources. Land use is a very crucial factor in determining the feasibility of large-scale power systems.
In this perspective, power production can be regarded as analogous to food production from agricultural land. This is an apt analogy since agriculture is also a form of power production. Humans get their energy from food, which is ‘produced’ using tracts of land.
If producing our electricity requires exclusive use of too much area, it will crowd out other land uses such as food production, human habitat, and wild ecosystems. Since all of these land uses are of prime importance for the continued existence of our civilization, areal power density becomes a factor of some importance.
Some issues with Smil’s technique
The area required to generate electricity does not tell the whole story; and we’re not suggesting that Smil meant it as such. However, we wish to clarify the fact that something as complicated as power production cannot be captured in a single quantity.
For his scenarios, Smil assumes a coal-fired power plant has an efficiency of 38% for his high-efficiency scenario, and 33% for a low-efficiency scenario. According to the World Coal Institute, average coal power plant efficiencies are actually 28%. This is a deviation from Smil’s usual technique of using real-world numbers for his analyses. For coal, he chose relatively high efficiencies that have been attained by some power plants, but do not reflect the real world averages.
Long term land damage
Land will still be damaged even after the coal plant stops producing power. Coal power plant sites are not remediated back to their former quality. This is partially due to soil contaminants and heavy metal buildup. Also, the burning of coal produces particulate and gaseous emissions that can severely affect health of ecosystems, cropland, and humans on a broader scale. These effects are also not limited to the time frame of the plant operation.
Land that has been used for coal mining is generally seriously degraded, even after the mining is completed. Mining coal creates liquid and solid wastes that are then spread into the surrounding environment. This degradation is long-term, since the biology, chemistry, topology, and ecology of the area has often been irrevocably changed. Wildlife in the area will be slow to recover, and may suffer from chronic setbacks due to contaminants left over from the coal mining process.
Pollution from these sources diffuses, impacting an area massively larger than that used to generate the power. The true tally of the land impact of coal power should attempt to include these factors. If areal power density is to be used to compare power sources, this must somehow fairly be accounted for.
In Smil’s estimation of coal, he takes a whole coal mine, power plant, and waste disposal area into account so that he doesn’t have a growing land use figure. We think he fails to mention an important fact when he does not account for land being only marginally useful even after the coal resource is exhausted.
When we consider coal power in the long term, we see that it will continue to use more land over time. More mining and more fly ash ponds will contribute directly to this fact. However, the additional degrading factor of the spreading pollutants from the mining and burning of coal will also be applied to vast areas of land. This degradation has been scaled back by the western nations to some extent, but lobbying battles continue as coal producers look to keep their waste management costs low. Smil does not mention the fact that generations of coal plants will continue to do incremental long-term damage to land, and will also continue to require new sites.
No land use synergies
No land uses that we are aware of can be conducted in harmony with an open pit or ‘mountaintop removal’ coal mine. An underground mine may also cause land subsistence or an underground coal seam fire, causing long-term problems. This is an important fact when we compare with other forms of power production (such as wind and solar), which can be conducted on land in synergy with other land uses.
Ideal spots used already
Smil also glosses over the fact that generally only old coal power plants have the ideal spots that he talks about in his optimistic estimate. We have used the ideal spots, and are quickly moving on to less ideal ones.
When considered for only one turbine, wind power appears to have the potential to be one of the most energetically dense forms of renewable power. However it does require sufficient spacing of turbines, so its areal power density drops substantially in real-world application.
Land use synergy
Here are some questions that we will use to examine the important ways in which the land use of wind turbines is dramatically different than it is for coal power:
- How was that land used before wind was built on it?
- What other land uses are possible synchronously with wind power?
- How might that land be used after the turbines exhaust their usefulness?
Wind turbines are often build on hilltops, ridges, and shallow ocean. These are generally areas that are not intensively used by humans except for farming, herding, fishing, and recreation. On land, the area around the bases of the turbines is still largely useful for cultivation and herding.3
There are some limits of course on the land uses that synergize with wind. Some people prefer not to live near wind turbines for various reasons, so this currently precludes intensive human dwelling development near large-scale wind installations.
Healthy land afterwards
Turbines may be removed after their useful lifetime, leaving behind few traces of their previous residency. This allows the land to return to its former land use at full productivity.
Land is cheaper than turbines!
Smil analyzes current wind farms for his analysis. While this makes sense, we would like to point out an unwritten implication from his analysis which has dubious basis. Smil’s estimates assume that power production is conducted on minimal land use. This agrees with common sense because power companies won’t bother to pay for more land than they need.
However, a very important facet of wind power is that there is a bit of a trade off between cost-effectiveness and land area. That is, if we pack in the turbines into too small of an area, we see reduced cost effectiveness. That is why turbines are spaced 5-10 blade lengths away from each other, depending on the situation. Sometimes they are placed in lines when the prevailing wind is from a very predictable direction.
Analyzing our current wind power installations gives us a snapshot of the turbine density per unit area that maximizes the cost-effectiveness of the wind installation. For most installations currently, the turbines themselves cost a lot more than the land that they are on. This means that in order to be cost-effective, they are spread out very substantially. Using more land currently doesn’t change the cost of wind power very much, but having inefficiently spaced turbines matters a lot.
Since wind synergizes with other land uses such as agriculture, there are vast tracts of land available for wind power development. To maximize cost-effectiveness, these turbines are placed quite far from one another. With very sparse wind farms, transmission infrastructure tends to be become a notable cost that will limit their geographic spread in a definitive manner.
We argue that Smil’s numbers are not measuring the potential of wind as an energy resource at all. They are measuring the areal energy density of the currently cost-effective wind power solutions. The areal energy density of wind may change dramatically in the near future with innovations like larger, slower spinning turbines, or greater market penetration of wind. It may soon make economic sense to place more turbines in closer proximity in areas that have very strong winds and high capacity factors. This may be more cost-effective than spread-out wind turbines that are built in less favourable locations.
On a last note, wind power and solar photovoltaics are intermittent power sources. This means that they will have to be matched with either energy storage or dispatchable generation in order to be baseload like coal or solar thermal power. We must keep this fact in mind when comparing their areal energy density with a baseload sources, or a dispatchable sources.
For more information on dispatchable power, and how we can meet it using renewable energy, see our article: How can renewables deliver dispatchable power on demand?
Call for submissions
This concludes the fourth installment of the Renewable Energy Review Blog Carnival. For a complete list of all publications in this series, see our post regarding the launch of the carnival. If you are interested in submitting an blog post or article to this carnival, see our submission page on the Blog Carnival website. This carnival is published weekly, and we are always interested in seeing new material.
The intent of this publication is an ongoing investigation of the progress and potential of renewable energy in our world. Our goal is to collect the best writing and news on the subject of renewable energy projects and policies. We have observed that humanity is innovating rapidly as the energy security of the future becomes a global priority.