Classical models for wear accumulative indicators are univariate Gamma and compound Poisson processes, which both are univariate subordinators (increasing Lévy processes). Bivariate subordinators are here proposed to model correlated couples of univariate wear indicators, namely bivariate wear indicators. A few properties of these bivariate subordinators are pointed out, and a special simple construction is provided, for application purpose. The use of these bivariate wear subordinors is illustrated through the study of a classical block replacement policy. The influence of the dependence between the marginal indicators on the optimal block replacement policy is pointed out.