Weighing up small hydro24 June 2004
A new approach for assessing environmental impact may help to change perceptions of hydro power, says Björn Svensson
FROM the arguments put forth by non-governmental organisations (NGOs), hydro power constitutes a larger threat to biodiversity than any other type of renewable energy. Life Cycle Analyses (LCA) provide one method of comparing the environmental performance of different kinds of electricity generation. Biodiversity is not easily included in LCAs but most interest centres on the protection of species, so analyses of different energy systems could at least include the fate of species as a variable when characterising the environmental impacts.
The LCA method is particularly well-suited to characterise complex systems, where the amount of service varies considerably between different options and where the use of natural resources or the emission of pollutants is strongly coupled to the geographical setting. Until now, it has been customary to stop the LCA when the inventory phase has been completed, thereby making a Life Cycle Inventory (LCI).
Thus, results of such LCIs are usually given as lists, where the total appropriation of a particular resource or emission of a particular pollutant is summed up over the entire lifetime of the system and divided by the total output of electricity during the same period. Consequently, the contribution to global warming is, for example, given as amount of carbon-dioxide equivalents per unit electricity generated (gCO2-equivalent/kWh.)
A compilation of LCIs made on hydro power reveals that large projects tend to be more efficient in terms of resource use compared to small projects, at least as far as hydro schemes with reservoirs are concerned (international-energy-agency, 2000).
Debate over the last few decades clearly shows that few environmentalists accept this conclusion, and in countries like Sweden, new hydro power is not accepted at all as an option for creating a green energy supply system.
NGOs usually relegate hydro power as an option without highlighting suitable alternatives. And when alternatives are suggested, it is usually wind and biomass energy that are advocated as environmentally more benign sources of electricity. But are they really?
There is a widespread contention that hydro power associated with reservoirs constitutes a larger threat to biodiversity than any other type of renewable energy (McAllister et al. 2001). Yet statements or quantifications of the impacts of hydro power plants as a decision making tool are not meaningful unless comparisons are made between different options on an equal base.
Biodiversity is not easily included in LCAs and, besides, there is still a lack of consensus about the practical meaning of the concept (Schenck 2001). However, most interest centres on the protection of species (May 1994). So analysis of different energy systems should at least include the fate of species as an additional variable when characterising the environmental impacts. Very few studies provide data that are sufficient for such analyses to be carried out.
The author has used information from the literature to obtain realistic estimates of the abundance and richness of species in different habitats. By combining this information with the amount of land required to supply a certain amount of energy, the author has attempted to estimate the loss of individuals and/or species per unit electricity produced with hydro power or when biomass is used as fuel.
Impact of hydro power
The ecological consequences of river regulation; damming, flow alterations, and perhaps diversion of water are complex and comprehensive. They include direct impacts related to habitat alteration and obstruction to dispersal, as well as indirect impacts that are coupled to physico-chemical and ecological changes.
General compilations of these effects have been carried out by Ward & Stanford (1983 & 1995) and the World Commission on Dams (2000). LCAs on hydro power have, so far, only treated these kind of impacts qualitatively and descriptively, while detailed quantitative estimates of the flow of material and matter, including emissions have been provided (Vattenfall 1999; Vold et al. 1996; Frischknecht & Müller-Lemans 1996; Beals & Hutchinson 1993; International Energy Agency 2000).
In addition, most studies on the impacts of hydro power relate to power plants with reservoirs. Run-of-river schemes are dealt with only exceptionally (Hildebrand et al. 1980; Turbak et al. 1981; Loar & Sale 1981; Olson et al. 1985; Hildebrand et al. 1980; Loar et al. 1980).
While opponents to large-scale hydro power quote lists of potential negative impacts to support their arguments, only aesthetics and the potential effects on biodiversity and fishery seem to be valid arguments against small hydro power development.
There are numerous success stories that describe how stocks of fish have been maintained following the construction of low-head dams (Swales 1989; Olson et al. 1985; Trussart et al. 2002). Measures that maintain or restore the habitats for plants and invertebrates, i.e. which enhance biodiversity in general, usually require less efforts (Gore 1985).
Consequently, there is justification to assume that run-of-river hydro power plants, where properly managed, have insignificant impacts on biodiversity. Yet, in order to measure them up to other means of electricity production, quantifications of impact are needed. In the absence of actual measurements, and as a starting point for such comparisons, it is perhaps reasonable to assume a total loss of animals and plants within the area when the power plant was constructed.
From an LCA perspective, this means that the measure of impact on biodiversity can be reduced to a function of specific land use, i.e. area reclaimed per unit electricity produced, and some expression of the content of biodiversity, e.g. average abundance or species richness, within the same area before it became exploited.
With this simplification, one major problem still exists, namely the quantification of biodiversity. Very few ecosystems have been completely surveyed with respect for their organism content. Birds, mammals, groups of insects and plants are usually well known, but inconspicuous groups of minute and taxonomically difficult invertebrates are not.
In the absence of complete sets of data, ecologists therefore commonly use indicators (easily sampled and easily identified organisms) such as birds, butterflies and vascular plants to characterise biodiversity. In this context, indicators must also reflect the overall biodiversity or richness of species in an area.
Correlations between different groups of potential indicators show that this latter requirement is sometimes, but not always, met (Malmqvist & Hoffsten 2000; Heino et al. 2003; Lawler et al. 2003; Paavola et al. 2003). For the time being, the author has chosen to use insects as indicators of biodiversity because they are the most numerous multi-cellular organisms in terrestrial (Stork 1988) as well as freshwater environments (Illies 1978). In addition, there are many field studies that provide data on the abundance of insects in different environments.
Hence, it is possible to obtain information about the richness of insects with enough confidence to calculate the magnitude of maximal losses due to different land use practices, including hydro power development. A compilation of data from numerous published surveys reveals that insect abundance in aquatic as well as riparian environments is on average in the order of 5000/m2.
The appropriation of land and water for the operation of two Swedish hydroelectric power systems has been estimated (Table 1). One system is Hornsö, a run-of-river plant located on the Alsterån, a south Swedish river (Klercq in press).
The other system is Luleälven in northernmost Sweden, a river with several reservoirs and 15 power stations arranged in a cascade (Svensson & Ericson 1993).
The difference between the two options in terms of specific loss of insects is probably larger than indicated because the simplistic approach used here exaggerates the negative impact on biodiversity caused by run-of-river hydro power vs hydro power with reservoirs. Although the riparian and aquatic habitat in the immediate surrounding of the run-of-river plant is altered, there is no reason to think that these alterations are transplanted downstream as long as the natural discharge is maintained.
When most water that runs the turbine is intermittently abstracted from a natural river channel, the impact on the aquatic biodiversity depends on the frequency and magnitude of the flow changes (Fig. 1).
Ecological mix up
The impact on biodiversity of hydro power associated with reservoirs as quantified by the method used above is too large. Even if the draw-down zone loses most of its set-up of flora and fauna, it is invaded by species that can cope with the harsh conditions and make use of this new environment (Fisher & LaVoy 1972; Friden 1984; Shafigullina 2000; Greenwood et al. 1995).
The net effect is, however, still comprehensive and always negative as regards the number of species.
The construction of dams is often said to cause fragmentation of affected rivers meaning that the river system is cut into more or less isolated pieces. According to ecological theory, habitat fragmentation will always bring about reductions in the richness of species.
The use of fragmentation to describe the ecological effect of dam construction is, however, erroneous. When the flow is maintained, be it for various periods of time, the aquatic habitat is basically kept uninterrupted.
A dam that prevents migrating species such as salmon and sea trout to reach their spawning grounds is not causing ‘fragmentation’ but ‘loss of habitat’, which is something completely different. It is unfortunate that ecologists frequently tend to mix up these two phenomena because the interpretation in terms of potential ecological impacts then also tends to be biased (Fahrig 2003).
Impact of biomass extraction on biodiversity
Biomass used for energy generation is derived from many sources. In Sweden, forest residues (FR) produced from the logging process has attracted much interest. The use of such fuel currently amounts to 10TWh, but the future potential is estimated to be 30-40TWh (www.stem.se).
Although the use of FR is encouraged in Sweden, it is still not clear how the impact on biodiversity compares to other means of energy generation. Studies have mainly concentrated on comparisons between areas that have been cleared from such residues and those that have not (Bengtsson et al. 1997; Åström & Nilsson 2003); the results of these studies are dubious.
There is also an allocation problem when attempting to quantify the specific burden of FR removal. Since the logging of saw timber and pulpwood – activities that cause the major impact on biodiversity – has priority, one generally views the possible extra impact of fuel extraction as insignificant.
However, if it is accepted that the change of habitat following logging is entirely attributed to those prioritised activities, we still need to know if and how much the FR would support biodiversity if left to decompose naturally. Unfortunately, few studies deal with this question. So far, The author has not been able to find any data in the scientific literature that integrate the content of insects and other animals in woody debris over its entire cycle.
Since the life span of woody debris extends over years and decades, depending on size and origin (Lundborg 1994), studies of life support potential are not easily carried out.
The author has anticipated that the number of insects that would make use of 1kg of woody debris from its deposition until fully decomposed amounts to at least 500, but in reality this number is probably considerably higher. The information used to arrive at this figure derives from many sources, e.g. Abbott & Crossley 1982; Schiegg 2000; Irmler et al. 1996; Schiegg 2001; Dajoz 2000; Marshall et al. 1998; Elton 1966; Hövemeyer & Schauermann 2003.
The heat content of 1kg dry weight of FR reaches 5kWh at most (AB Svensk Energiförsörjning 1998). The average yield of biomass in Swedish forests is estimated at 0.5kg/m2 per year (Johansson 2001; Parikka 1997).
About 20% of the annual growth is available as FR, which means that the extraction of 1kWh of heat brings about a net loss of at least 1000 insects, i.e. an order of magnitude higher than the two Swedish hydro power systems described above. The difference would be even higher with adjustment for the energy conversion efficiency of biomass burning.
Clearly, there are many uncertainties that need to be considered and evaluated when the impact of renewable energy generation on biodiversity is estimated. Moreover, the described attempt to do so provides only one of many possibilities.
Opponents to the approach used here could for example claim that insects are not the most important biodiversity indicators and that the threat to red-listed or endangered species would be more relevant.
However, the intention with the above reasoning is not to advocate an immediate shift in the political attitude to different energy sources but merely to introduce a necessary complement to the description of different energy supply options and consequently a better decision support system.
Nevertheless, the result may seem surprising given the massive information often quoted about the damaging impact of hydro power on river ecology. This is partly a result of reluctance to consider the degree of service provided by a particular power plant. Hydroelectric power makes more efficient use of the reclaimed land compared to the extraction of energy from biomass. Therefore, small-scale hydroelectric power in particular will always be a better option compared to energy production based on forest residues.
When energy crops are used instead, considering that agriculture is the prioritised land-use, the picture is less clear. Using willow planted on arable land as the source of biomass energy will probably bring about a net gain in biodiversity compared to traditional production of cereals (Göransson 1994; Berg 2002), not least because of the need to use pesticides to keep the number of weeds and insects low in modern agriculture.
It is, however, not only necessary to divide the total environmental burden by the amount of electricity produced. One must also consider when the capacity to provide electricity is available. In countries with a high share of hydro power, this energy source not only constitutes a stand alone option, but also provides the necessary back-up that makes the entire energy mix match the demand. Hydro power with reservoirs is a firm energy source as opposed to most other sources of electricity.
A fair comparison between different options in terms of their environmental characteristics should also look at the possible need to use complementary power from sources that are partly or fully operated for peaking-power production. Unfortunately, the allocation of impacts that takes this aspect into account is a complicated matter and good studies are still awaited.
TablesTable 1 Table 1