Economies of density for on-site waste water treatment

Abstract

Decentralised wastewater treatment is increasingly gaining interest as a means of responding to sustainability challenges. Cost comparisons are a crucial element of any sustainability assessment. While the cost characteristics of centralised waste water management systems (WMS) have been studied extensively, the economics of decentralised WMS are less understood. A key motivation for studying the costs of decentralised WMS is to compare the cost of centralised and decentralised WMS in order to decide on cost-efficient sanitation solutions. This paper outlines a model designed to assess those costs which depend on the spatial density of decentralised wastewater treatment plants in a region. Density-related costs are mostly linked to operation and maintenance activities which depend on transportation, like sludge removal or the visits of professionals to the plants for control, servicing or repairs. We first specify a modelled cost-density relationship for a region in a geometric two-dimensional space by means of heuristic routing algorithms that consider time and load-capacity restrictions. The generic model is then applied to a Swiss case study for which we specify a broad range of modelling parameters. As a result, we identify a 'hockey-stick'-shaped cost curve that is characterised by strong cost reductions at high density values which level out at around 1 to 1.5 plants per km2. Variations in the cost curves are mostly due to differences in management approaches (scheduled or unscheduled emptying). In addition to the well-known diseconomies of scale in the case of centralised sanitation, we find a similar generic cost behaviour for decentralised sanitation due to economies of density. Low densities in sparsely populated regions thus result in higher costs for both centralised and decentralised system. Policy implications are that efforts to introduce decentralised options in a region should consider the low-density/high-cost problem when comparing centralised and decentralised options.