Modelling the upgrade of an urban waste disposal system

Abstract

The waste intermodal station of Clyde, in the city of Sydney, Australia, is in the heart of a complex network of terminals connected by road and rail to transport urban waste from its first collection to its final disposal. The amount of waste the network is projected to handle in 2015 will increase from about 340,000 tonnes/year in 2006 up to about 1.5 million tonnes/year, following population and consumption raise. The paper proposes a discrete-event model to represent Clyde Transfer Station (TS) and its relations with the other terminals. Such a model allows one to evaluate the effects of different expansion plans of the station structures as well as different policies of the collection service. The results show that, as often happens, a careful consideration of management alternatives can decrease the necessity of structural enlargement, which is normally much more expensive, if at all possible.

Introduction

The worldwide expansion of the urban population and the parallel improvement of lifestyles make the disposal of urban wastes a critical problem in both developed and developing countries. Furthermore, most urban wastes are still disposed into landfills which have a finite capacity and were originally located relatively close to the urban areas. They are now approaching their planned capacity and their expansion is almost forbidden by the enlargement of residential areas and by the augmented concern of citizens with respect to environmental issues.

The situation can be partially compensated by the improvement of production processes to reduce the amount of wastes “from cradle to grave”, i.e. during all the life cycle of products from their design, to production, to packaging, to use and disposal, and by an enhancement of recycling and reuse practices. All over the world, legislation is pushing in this direction and the situation in Australia is no exception.

Projections for the whole Australia foresee for the next decades an increase in the total amount of generated urban wastes (disposed and recycled) from 31.6 million tons per year in 2002/03 up to 57 million tons per year in 2022/23 (ABS, 2006a, DEH, 2006). In particular in New South Wales (NSW) urban solid waste to dispose is projected to rise every year reaching 8.1 million tons per year in 2014 (NSW DECC, 2006), while the related waste legislation asked at least to hold its level for 5 years from the release of Waste Strategy 2003, when it was around 6 million tons, and it set other thresholds for the remaining six years.

Similar guidelines dealing with waste generation, transport and disposal have been suggested by the European Directive 2006/12/EC as well (codified version of the previous Directive 75/442/EEC).

The above figures explain why the region of NSW has been taking an increasing interest to the waste management issue. It is understood that a more sustainable transport, recovery and disposal of waste have to be pursued in order to reduce the considerable pressure on the environment (Beavis et al., 2006). Waste management policy needs to suggest a new interpretation of products handling, seeking to optimize outcomes against a range of economic, environmental and social objectives across all stages of a product's life, rather than being concerned only with the final disposal. NSW has recently updated the current guidelines, introducing the Waste Avoidance and Resource Recovery Strategy in 2006 developed under the Waste Avoidance and Resource Recovery Act 2001 (NSW DECC, 2006). Action needs to be taken in order to fulfil the objectives defined by this new strategy.

To tackle this situation, the scientific literature has addressed on the one side the classical problem of designing the expansion and management of disposal facilities (e.g., Cheremisinoff, 2003), and, on the other side, the optimal location of facilities on a given territory which also entails an optimal planning of waste collection (see, for instance, Abou Najm and El-Fadel, 2004 and the literature review presented therein). While the first point of view concentrates on a single plant and mainly requires a continuous time modelling to evaluate, for instance, the impact of a landfill on underground water quality or the combustion dynamics in an incinerator (e.g. Astrup et al., 2004, Lobo Garcìa de Cortàz and Tejero Monzòn, 2007, Huai et al., 2008), the second approach is normally non-dynamic, utilizes classical operation research tools as integer programming and network optimization, and is presently evolving towards an integrated assessment of the complete waste production and disposal cycle (e.g. Tchobanoglous et al., 1993, Everett and Applegate, 1995, Björklund et al., 1999, Tchobanoglous and Kreith, 2002, Costi et al., 2004, Salhofer et al., 2007).

A third, intermediate approach is however possible, though rarely used (C̆erić and Hlupić, 1993, Lu et al., 2006, Johansson, 2006). It looks at the waste disposal system as a classical industrial activity where wastes reach the plants in discrete lots at certain (almost random) time instants and plants perform a number of “transformations” on incoming wastes (actions required for the final disposal) using given resources and requiring a certain time. The modelling paradigm normally used in these cases is based on discrete events and is particularly suited to explore the trade-offs between an increase in the capacity of the waste disposal system (an action that, as already mentioned, may not be easy for technical and social reasons) and other non-structural changes (Humphries, 1986, Ballis and Golias, 2001), such as a modification of the collection schedule, that may be undertaken more rapidly and economically, as pointed out also by Wilson et al. (2002).

This type of modelling is used in this paper to analyze future scenarios of the Great Sydney Urban Waste disposal system and determine the required retrofit as well as possible management actions to improve its performance.

The present and future conditions of the system are described in the next section, while the third presents in more details the Clyde intermodal transfer station. Section 4 shows how a model for the systems was developed and verified, recalling also some basic ideas about discrete event systems and queuing theory. The following section defines the scenarios that have been simulated and the optimization performed and, finally, Section 6 shows the results obtained and highlights the trade-offs determined between facility expansion and management improvements.