In open shops with renewable resources, an operation may require addit
ional resources, besides a machine, for its execution. All resources r
equired by the operation are allocated to it all the time during its e
xecution. At no time may total resource requirements exceed resource c
apacities. We consider the problem of minimizing makespan in a two-mac
hine open shop with a single renewable resource. We show that optimal
nonpreemptive schedules are not longer than optimal preemptive schedul
es, which we then translate into a polynomial time algorithm for the m
akespan minimization problem. This is an important generalization of a
well-known result obtained by Gonzalez and Sahni (1976) for the two-m
achine open shop without additional resources. We also study the probl
em of minimizing makespan in a two-machine open shop with at least two
different resources. In this case, optimal nonpreemptive schedules ma
y be longer than optimal preemptive ones. We show that this makes the
makespan minimization problem NP-hard.