However, Concentrating Solar Power plants are often designed to use water for cooling at the back-end of the thermal cycle, typically in a wet cooling tower. These water requirements can result in difficulties in arid areas, particularly in the MENA region, being the region in the world experiencing the hardest water stress.
Large-scale implementation of CSP in Europe and the MENA region requires that additional water needs can be effectively met, or technologies with lower water use must be implemented.
A typical 50 MW parabolic trough plant uses 0.4–0.5 million m3 of water per year for cooling: roughly the same as agricultural irrigation of an area corresponding to that occupied by the CSP plant in a semi-arid climate (and less than half that used for irrigating food crops in Andalucia in Spain).
In the MENA region, withdrawal of renewable water resources is already above 70%, i.e. close to exhaustion. Water could possibly be diverted from its massive, in some cases inefficient, use in irrigation. The water withdrawal for agriculture in the MENA region was 188.3 billion m3 in 2002, while the corresponding figure for the entire MENA region’s industrial sector was only 7.9 billion m3 in the same year.
But the prospect of withdrawing large amounts of fresh water for CSP cooling is not appealing, particularly when the MENA region water demand is conservatively expected to almost double in the period 2000–2050. Water is also used for cleaning the mirrors to maintain their high reflectivity, although water use for cleaning is typically a factor of a hundred lower than that used for water cooling.
It may be more significant in desert areas where dust storms may require more frequent cleaning, and the associated water consumption is relatively higher when compared with precipitation.
Experience with CSP plants in Spain is that soiling rates and hence washing requirements are a little higher than initially expected. Water use can be decreased by cooling with air instead, but this lowers the effi ciency of the system. A study conducted by the US National Renewable Energy Laboratory indicates that the switch from wet to dry cooling in a 100 MW parabolic trough CSP plant can decrease the water requirement from 3.6 l/kWh to 0.25 l/kWh.
Using dry instead of wet cooling increases investment costs and lowers efficiency, adding 3–7.5% to the LEC. For areas with high irradiation and available land close to the sea, such as the Egyptian north coast, using salt water for cooling could be an attractive option. It also opens up the possibility of integrating desalination with the CSP plants.
Finally, there are some CSP plant designs that have inherently low fresh water requirements, such as gas turbine towers and parabolic dishes with Stirling engines.
Land use and visual impact
To compare CSP land use to that associated with other energy conversion technologies, a basic estimate of land use has been made. Land use refers to the area directly occupied by a power plant structure (in a CSP plant the collector/heliostat fi elds dominate), by extraction of fuel, or by plantations for biomass. It is presented in relation to the energy generated annually by each plant, and hence is expressed in units of m2/(MWh/y).
The ‘visual impact’ gives the area over which a power plant disturbs the view, divided by the energy generated annually by the plant (and hence is also expressed in units of m2/(MWh/y)). For comparison, for wind power. Visual effects are most noticeable in tower CSP plants where very bright points appear in the rural landscape. However, due to contemporary social attitudes the signal has been interpreted by the population as a technical novelty and a sign of progress, not causing rejection (so far).
One advantage of CSP plants is that they are often located in areas with limited amenity or aesthetic value. Using desert land for solar energy plants could in many ways be seen as better than, for instance, agricultural land for biomass energy. The placement of power plants or fuel extraction (such as lignite) close to highly populated areas can be almost completely avoided. The areas available globally for CSP development far exceed present needs.
Nevertheless, arid regions do have environmental value, and contain some biotopes or species that are threatened. The harshness of the desert climate also makes it take longer for an arid biotope community to recover from the effects of disturbance. Massive establishment of solar plants in an area may affect regional animal or plant populations by cutting dispersion routes and partially isolating populations from each other. This is hardly unique for CSP plants, but calls for some caution.
Energy and materials use
In evaluating the sustainability of CSP plants it is useful to compare their energy balance and material use over their life cycle to other power generation technologies. A life cycle assessment of CSP power shows that the cumulative (non-renewable) primary energy invested in construction and operation of a plant over its lifetime is gained back as renewable power in less than one year of the assumed 30-year life. This gives an energy return on investment (EROI) of about 30.
The cumulative (non-renewable) primary energy needed to produce 1 kWh of electricity is comparable to that of wind power and orders of magnitude lower than for fossil-fired power plants.
CSP plants are more material intensive than conventional fossil-fired plants. The main materials used are commonplace commodities such as steel, glass and concrete whose recycling rates are high: typically over 95% is achievable for glass, steel and other metals. Materials that cannot be recycled are mostly inert and can be used as filling materials (e.g. in road building) or can be land-filled safely.
There are few toxic substances used in CSP plants: the synthetic organic heat transfer fl uids used in parabolic troughs, a mix of biphenyl and biphenyl-ether, are the most significant. They can potentially catch fi re, can contaminate soils and create other environmental problems, and have to be treated as hazardous waste.
One aim of current research activities is to replace the toxic heat transfer fluid with water or molten salts. These also have the benefit of being able to be used at higher temperatures, giving better efficiencies and hence decreased specific emissions.
The emissions of greenhouse gases are strongly linked to the cumulative (non-renewable) primary energy demand. Greenhouse gas emissions for CSP plants are estimated to be in the range 15–20 grams CO2-equivalent/kWh, much lower than CO2 emissions from fossil-fired plants which are 400–1000 g/kWh. Greenhouse gas emissions of about 9–55 g CO2-eq/kWh for large-scale CSP technologies.
Using nitrous salts as heat transfer fluid and/or storage medium creates life cycle emissions of nitrous oxide (N2O). Although the amounts are roughly 500–1000 times smaller than the carbon dioxide emissions associated with a coal plant, they are not negligible as N2O is about 300 times stronger than CO2 as a greenhouse gas. Again, coal-fired plants have the highest emissions, but in this case, natural gas-fired plants have values not much higher than the renewable technologies.
Impacts on flora and fauna
Local impacts of CSP plants on the environment may be associated with traffic, building works, ecosystem disturbance, and loss of ecosystem functions. Traffic, plant construction and surface treatment of parking plots cause indirect mortalities to local fauna at a level depending on the surface area of the facility and the land use type before plant construction.
Mortalities caused to vertebrates are the main concern in respect of the local environmental impact of CSP plants. Direct mortalities take place under two main circumstances: collision with top mirrors and buildings (the tower in particular), and heat shock or burning damage in the concentrated light beams. Birds rarely collide with CSP plants when visibility is good, but when vision is impaired casualties have been documented.
A poorly illuminated solar tower can be hit by birds at night, but this is rare. Birds may mistake the reflecting surfaces for air or water and collide with them, for instance when taking fl ight from the ground. Insects can also mistake the glass surfaces for water and be killed, or lose eggs they are carrying, in attempts to enter the surfaces.
If a plant is built on former agricultural land, available nutrients in the soil may facilitate growth of vegetation up to 1 m in height below and between solar collectors.
Under Mediterranean climates the vegetation can dry up and contribute to fire risk. Herbicides can be used to prevent plant growth, but they typically have toxic effects at some scale, persist in the soil profile, and may be exported with runoff. Alternative treatments of soil surface that impair seedling establishment include compacting the ground, enabling the development of a surface crust, or adding gravel.
The water used in mirror cleaning drips onto a narrow ‘wet band’ at the base of the collectors whose area is around 15–25% that of the collectors’ surface. This cleaning water supply to the wet band may range from 10 to 20 mm/year, which can be a significant amount during dry summer months (particularly in desert areas), stimulating and/or maintaining plant growth.
As mentioned earlier, CSP plants may indirectly harm local animal or plant populations by cutting off migration routes. Another impact related to plant construction and operation is the introduction of species previously alien to the area. Gardening, goods and equipment, and public works machinery all contribute to introductions. Some other species actively follow contractors and colonise their area of activity, gaining from the removal of local species from the disturbed land.
Although CSP plants can have several effects on the local environment, compared with other technologies, particularly fossil-fired plants, they are relatively benign.
The direct damage from solar plants is low: a monitored CSP tower plant operating since 2007 in Spain has so far only recorded two bird deaths. Even with a much larger implementation, environmental impacts will not be on the same scale of direct and indirect effects from fossil fuels, like the Deepwater Horizon oil disaster in the Mexican Gulf in 2010.
All power generation has some effects on the environment but it is evident that CSP plants on the whole have much better environmental performance than today’s fossil-fi red technologies. Not using extractable fuels means that CSP is free of the impacts from coal mining, spills from oil rigs, leakage of methane from gas extraction, etc.
On the other hand, use of commodities such as steel, glass and concrete is relatively high, although most of these materials are readily available and have high recycling-potential. Issues that need to be addressed are water requirements in arid areas, use of toxic synthetic oils as heat transfer fluids, and use of pesticides to restrict vegetation growth in heliostat fields.
For all of these issues, technical solutions are available or under development. Environmental impacts vary between technologies and over time. Although some CSP technologies today are proven and commercialised, they are less mature than conventional fossil-fired power stations. This means that they can be expected to progress faster with innovation and improvements of efficiency, and hence the environmental impact of CSP technologies, relative to fossil-fired power, is likely to get (even) better over time.