Contents lists available at ScienceDirect Electric Power Systems Research journal homepage: www.elsevier.com/locate/epsr Investigation of different methods to generate Power Transmission Line routes S. Ghandehari Shandiz a , G. Doluweera b , W.D. Rosehart c , L. Behjat c , J.A. Bergerson a, ⁎ a Department of Chemical and Petroleum Engineering, University of Calgary, Calgary, Alberta, Canada b Canadian Energy Research Institute, Calgary, Alberta, Canada c Department of Electrical and Computer Engineering, University of Calgary, Calgary, Alberta, Canada ARTICLE INFO Keywords: Power Transmission Lines Routing Siting methods Stakeholder values Multicriteria decision-making ABSTRACT In this study, we analyze and then evaluate four methods for siting more than one Power Transmission Line (PTL) simultaneously. Specifically, we look at (1) the least cost path (LCP) inside a macro-corridor, (2) the simultaneous definition of two routes (3) many routes inside a macro-corridor and (4) non-corridor routes generation. We apply the methods to a case study of siting two lines to deliver power for Northeastern Alberta and evaluated based on a set of metrics including overall impacts, computational complexity, and the spatial variability of the proposed alternatives. The results show that when the range of stakeholders’ values and concerns are incorporated into the siting model, the conventional LCP between the source of electricity and the destination is not necessarily the best solution. Rather, our findings show that among the methods examined in this study, non-corridor routes generation method tends to find the lowest impact alternatives. 1. Introduction There is an increasing worldwide need for new Power Transmission Lines (PTLs) as the demand for electricity grows. PTLs have a variety of impacts on the environment, ecosystem and society. Most of these im- pacts are spatial, as they depend on the location of the terrain where the PTLs are placed. ‘Power Transmission Line siting’ is the regulatory process for the identification of the corridor in the terrain where a PTL can be placed. Siting PTLs is usually a very long process because of the difficulty in coming to a universally agreed solution amongst all the stakeholders, which includes all parties potentially affected by the construction and operation of PTLs [1] such as property owners, mu- nicipalities and transmission facility owners. Traditionally, models that are developed to support the siting de- cisions are spatial models that optimize the techno-economic para- meters of transmission lines while determining the line corridor. However, this approach fails to address the concerns of a large number of stakeholders in particular, affected landowners. Thus, the decision- making process can lead to significant stakeholder oppositions and subsequent delays in the approval phase [2–4]. An attempt to integrate the economic and environmental criteria is presented where satellite images are used as the input map and different qualitative weights are applied to select the best route [5]. However, conflicting stakeholders’ values are not incorporated in this siting study type. Besides stakeholder satisfaction, reliability is also a factor that shapes the siting decision. Spatial reliability becomes important parti- cularly when system operators identify a need to build two (or more) PTLs [6]. These routes are normally separated by a pre-specified minimum distance to ensure the reliability of the power system [7] thus that the potential reasons for the failure of one line is unlikely to impact the parallel line simultaneously. The common practice in transmission line siting that most studies in this field focus on is the utilization of one least cost path (LCP), such as Dijkstra’s algorithm [8], to find a route with minimum cost across a set of specified criteria [9–15]. A limited number of studies discuss specific al- gorithms for generating alternative routes. The K-shortest loopless path (KSP) method and its variants were some of the earliest attempts to solve the problem of alternative routes generation (see online bibliography [16]). KSP uses a brute-force method of systematically listing all possible routes between a given origin and destination, then ranking them in order of length. Although the KSP method guarantees all possible paths are found within a cost threshold, it is not practical as it generates a massive number of possible paths that are spatially similar and share most of the attribute values [17]. Paths with similar attribute values do not aid in the decision-making process since they do not convey the full range of options that are available to decision makers. The KSP runtime and memory re- quirements increase factorially with solution space. Hence it is limited in all practicality to trivially small networks [17]. Subsequent algorithms https://doi.org/10.1016/j.epsr.2018.08.012 Received 11 December 2017; Received in revised form 13 July 2018; Accepted 19 August 2018 ⁎ Corresponding author. E-mail address: jbergers@ucalgary.ca (J.A. Bergerson). Electric Power Systems Research 165 (2018) 110–119 0378-7796/ © 2018 Elsevier B.V. All rights reserved. T