denote by W the set of waste commodities, where each commodity
wAW represents a specific type of waste. Concerning the data
required to model a strategic SWM system, the following sets and
constants are given. Let qt
jw denote the capacity of facility jAVS [ VP [ VL for commodity wAW in period t (t ¼ 1;…; T), and aj
denote the overall estimated capacity of landfill jAVL that cannot
be exceeded during the planning period. All the information
related to the total amount of waste generated at each source
iAVO in a given period t (t ¼ 1;…; T) for each commodity wAW is
represented by Gt
iw. Observe that uncertainty may affect the value
of Gt
iw. Finally, let bjww′ denote the reduction coefficient per unit
weight (or volume) of the waste commodity wAW into the waste
commodity w′AW at facility jAVS [ VP . This coefficient is used to
describe how the waste flows are converted when they reach
a processing facility. The costs involved in these processes are:
a non-negative transportation cost per unit of waste cij that is
associated with each arc ði; jÞAA, an initial cost f t
0j for opening a
new facility jAVS [ VP [ VL in period t (t ¼ 1;…; T), a fixed cost ftj
for operating facility jAVS [ VP [ VL in period t (t ¼ 1;…; T), and a
cost pt
jw when processing a unit of waste of type wAW at facility
jAVS [ VP [ VL in period t (t ¼ 1;…; T). These costs are timedependent
to account for possible variations over time. Decision
variables ztj
are binary variables taking the value 1 if and only if a
facility jAVS [ VP [ VL is open in period t (t ¼ 1;…; T), 0 otherwise.
Variables ytj
are also binary variables taking the value 1 if
facility jAVS [ VP [ VL is operating in period t (t ¼ 1;…; T), and
0 otherwise. Finally, variables xt
ijw represent the flow of waste
commodity wAW traversing arc ði; jÞAA in period t (t ¼ 1;…; T).
A mathematical formulation of the problem is