We study the stationary state of a simple exclusion process on a ring which
was introduced by Arndt et al. In a previous study by Rajewsky et al it wa
s claimed that the spatial particle condensation the model exhibits is not
associated with a phase transition in the framework of a grand canonical en
semble. The discussions were, however, based on an assumption about the mon
otonicity of a certain function appearing in the analysis. In this paper we
prove the assertion. The proof combined with the previous discussions show
s convincingly that there is no phase transition in the framework of a gran
d canonical ensemble.