Consistent cost curves for identification of optimal energy savings across industry and residential sectors

Henrik Klinge Jacobsen, Mattia Baldini

    Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

    138 Downloads (Pure)

    Abstract

    Energy savings are a key element in reaching ambitious climate targets and may contribute to increased productivity as well. For identification of the most attractive saving options cost curves for savings are constructed illustrating potentials of savings with associated costs. In optimisation modelling these cost options are then compared with the cost of producing energy and all savings with negative costs and cost below the cost of producing the energy including the associated externality costs are expected to be implemented. There are however several methodological issues associated with constructing and applying the cost curves in modelling: • Cost curves do not have the same cost interpretation across economic subsectors and end-use technologies (investment cost for equipment varies – including/excluding installation – adaptation costs – indirect production costs) • The time issue of when the costs are incurred and savings (difference in discount rates both private and social) • The issue of marginal investment in a case of replacement anyway or a full investment in the energy saving technology • Implementation costs (and probability of investment) differs across sectors • Cost saving options are not additive - meaning that marginal energy savings from one option depends on what other options implemented We address the importance of these issues and illustrate with Danish cases how large the difference in savings cost curves can be if different methodologies are used. For example, the difference between marginal investment costs in residential heating of a more efficient building element (windows) in a larger renovation project compared to the costs of just replacing the windows. This is done based on some of the results from Zvingilaite & Klinge Jacobsen 2016. We compare to the results found for residential savings in Giraudet et. al. 2012 and Amstalden et. al. 2007. For our case the resulting savings potential below a given level of costs can be up to a factor of 5 times larger if only the marginal cost measure is used. For national energy plan strategies this results in much more emphasis on energy savings, than renewable energy expansion as a way to achieve fossil fuel reductions if it is possible to implement all heating savings with their marginal costs. As saving potentials are not additive for savings in a specific end-use entity it is difficult to compare savings in one sector comprising many options together and single options in another sector. We illustrate that a saving option in one sector (eg a more efficient pump) would be difficult to compare with the savings from replacing an entire production line in a factory with a more efficient one. If the average cost of the two are compared then probably the efficient pump would be preferred due to low costs compared to the full production line. This would leave out the elements of the production line where independent savings investments might have cost that are just as low as for the pump. We argue that comparing across sectors should be carried out with similar sets of savings options (small individual replacements in each sector, and comparable larger technology switches in each sector).
    Original languageEnglish
    Publication date2017
    Publication statusPublished - 2017
    Event40th Annual IAEE International Conference - Singapore, Singapore
    Duration: 18 Jun 201721 Jun 2017
    http://iaee2017.sg/

    Conference

    Conference40th Annual IAEE International Conference
    CountrySingapore
    CitySingapore
    Period18/06/201721/06/2017
    Internet address

    Fingerprint Dive into the research topics of 'Consistent cost curves for identification of optimal energy savings across industry and residential sectors'. Together they form a unique fingerprint.

    Cite this