Abstract:
Given the importance of improving municipal solid waste management, the Optimisation
Group from the University of Pretoria completed preliminary studies on data obtained
from waste collection vehicle global position system traces from the City of Cape Town
and found that some waste vehicles spend more time travelling between service locations
and dump-sites as opposed to collecting waste. In order to improve these operations it was
proposed that intermediate facilities be used to reduce the travel distance. The purpose
of this project is to identify potential locations for intermediate waste facilities (transfer
stations) by using a location model and then evaluating the impact of the intermediate
waste facilities on waste vehicle collection operations.
This was achieved by splitting the project into three phases. Since some of the data that
was required was not readily available, it had to be generated rst. In phase one that data
sets were generated. In phase two a variant of the location modelling formulation known as
the multi-facility location problem was solved using the generalised Weiszfeld method for
the multi-facility location problem. Up to 25 potential intermediate facility locations were
identi ed and used for the remaining phase. In the next phase two models were built in
order to quantitatively evaluate the impact of said facilities on waste collection operations.
The rst model was concerned with the travel distance and travel time for each collection
vehicle associated to a single collection beat (service location). The second model was
concerned with the waste allocation from the collection beats to either an intermediate
facility or a land ll site and from the intermediate facilities to a land ll. The model was
also used to conduct a cost bene t analysis.
It was concluded that as the number of intermediate facilities increased, the travel
distance, travel time as well as the total transportation cost of the system decreased.
Additionally, the location of the two existing intermediate facilities in the City of Cape
Town were ine cient when compared to the same number of intermediate facilities located
with the Weiszfeld algorithm. Furthermore, it was identi ed that the optimal number of
facilities to locate, in the City of Cape Town, in order to reap the cost savings from
improving the existing intermediate facility locations with respect to the capital cost of
building new facilities, was three.
