Uncertainty and error in locational allocation modeling and route networking

Keywords: Surface, Adjusted, Dijkstra, Uncertainty, Route, Networking, Distance, Error, Metrics, Anaerobic, Digestion
Session Type: Virtual Paper Abstract
Day: Friday
Session Start / End Time: 2/25/2022 03:40 PM (Eastern Time (US & Canada)) - 2/25/2022 05:00 PM (Eastern Time (US & Canada))
Room: Virtual 59

Abstract

An Anaerobic Digester project (AD) is comprised of different layers of analysis that are tightly knit to the geographical domain. Lack of data and uncertainty require modifying current methodological approaches to location modeling and siting analysis. A good example is the use of Dijkstra’s algorithm (1959) that conventionally operates by accumulating Euclidean distances, but does not account for distance underestimation in rough terrain. The paper reports a refinement of the conventional algorithm to address potential errors in distance metrics (Buttenfield et al 2019; Qiang et al 2020). A quantified relationship between terrain roughness and distance errors will shed light on how the economic feasibility of an AD project can be affected, depending on the terrain (Poser and Awad 2006). A second problem is that a lack of data on source point locations can negatively affect the AD project,. The problem arises when the input organic waste is situated at an unknown location, as is the case of farms in many US counties. Farms are important as they are considered the main source of the AD input (organic waste). Location allocation modelling cannot be carried out under those circumstances. The paper implements Monte Carlo simulation to probabilistically allocate farm seeds around research area inclusive zones that were created under a multi-criteria evaluation. The result showcases a range of possible farm locations, that is used to quantify a minimum distance error so that the modified Dijkstra algorithm can capture the uncertainty in the most precise way possible.