Highlighting the importance of assessing all relevant impact pathways of renewable energy sources, in order to identify the main environmental impacts and identify trade-offs between different energy production options and places of operation, has been the focus of recent research.
The impacts of hydropower electricity production can be felt through land use and land use change (LULUC). Storage and pumped storage hydropower plants cause LULUC. Reservoir filling causes LULUC by raising water levels and inundating land. Besides reservoir filling, further LULUC is caused by the construction of infrastructure, including power lines and access roads. However, according to the authors of a new research paper, no operational model exists that covers any of these cause-effect pathways within life cycle assessment (LCA).
In an attempt to contribute to the assessment of LULUC impacts of Norwegian hydropower production in LCA, Martin Dorber, Roel May and Francesca Verones from the Norwegian University of Science and Technology and the Norwegian Institute for Nature Research, quantified the inundated land area associated with 107 hydropower reservoirs with remote sensing data and related it to yearly electricity production.
LCA is a commonly used methodology for analysing the complete environmental impacts of a product or process throughout its life cycle, and is a particularly suitable method for identifying potential trade-offs between impact pathways. However, LCA is still developing and cannot currently assess all relevant biodiversity impacts from hydropower production on a global scale. Dorber et al say that in their paper they address this research gap from a LCA perspective with focus on LULUC.
The authors explain that remote sensing data provides an opportunity for assessing net land occupation in a spatially explicit manner. The data is useful for monitoring actual surface area, as well as wetland identification in general. In addition, case studies on land-use transitions from lakes and lake desiccation have shown that remote sensing data can be used to calculate natural lake surface area prior to inundation. To identify land cover types, like water, from satellite images, these studies use the different spectral responses of different land cover types, assessed by the satellite sensor. Therefore, the first aim of this study was to utilise these case study-based approaches, in combination with remote sensing data providing global coverage, to quantify spatially explicit inundated land area values due to the installation of storage hydropower plants in a globally systematically applicable approach.
Inundated land area
Dorber et al say that they were able to quantify the inundated land area for 184 of the 265 hydropower reservoirs in Norway that have a commissioning year of 1972 or after. The main reason for not quantifying all of the hydropower reservoirs was due to the fact that many Global Land Survey-1975 images were acquired in early May or October when lakes are frozen in Norway, thus making identification of some water surface areas impossible. This was also the main reason for the low number of hydropower reservoirs with quantified ILA in the north.
In total, the authors assessed land occupation for 75% (808km2) of the reservoir surface area (RSA) from hydropower reservoirs with a commissioning year of 1972 or after in Norway. This represents 13.4% of the total RSA of all Norwegian hydropower reservoirs. The 107 hydropower reservoirs have an average annual electricity production of 27.059GWh, representing 19.6% of the total average annual hydropower electricity produced in Norway between 1981 and 2010.
Implementation and use
The unit of the modelled land occupation is m2·yr/kWh. This is in accordance with the unit of m2·years for land occupation in the land use inventory and therefore the net land occupation values calculated for storage hydropower reservoirs are directly implementable in lifecycle inventory databases.
The authors assumed that the average annual electricity production is the “true” electricity production in a normal year, and the average annual electricity production can be used to calculate the land occupation per kWh produced. This means that their values are designed to calculate the average land occupation over the complete operational phase. They are, however, not applicable for individual years, as this can either lead to over- or underestimation of the average yearly land occupation, because the annual inflow to hydropower reservoirs in Norway has varied from 1990 to 2013 by about 60TWh. This, for example, caused the variation in the hydropower electricity production in the whole of Norway from 143TWh in 2000 to 106TWh in 2003, the latter being a very dry year.
However, if the efficiency of the hydropower plants change over time (e.g., due to changes in precipitation patterns), the inventory has to be updated as this will reduce the land occupation per kWh. The authors’ calculated land occupation is only representative for the period from 1972 to 1985 and should not be used to quantify the land occupation of newly built hydropower reservoirs. For newly built facilities, the inundated land area itself can be directly modelled with digital elevation models.
The authors also noted that their values do not account for land occupation and hydropower electricity production changes that may occur due to possible precipitation and related hydrological regime changes under different climate change scenarios. Furthermore, LCA does not account for potential positive effects on species like fish or water and shore birds.
REFERENCE
Modelling Net Land Occupation of Hydropower Reservoirs in Norway for Use in Life Cycle Assessment by Martin Dorber, Roel May, and Francesca Verones, Environmental Science & Technology 2018 52 (4), 2375-2384 DOI: 10.1021/acs.est.7b05125
