Cluster tools are integrated machines for wafer processing in semiconductor wafer fabrication facilities (wafer fabs). Since different kinds of wafers can circulate in a cluster tool simultaneously, it is a fully automated machine environment. The development of scheduling approaches for these extremely expensive machines is essential for improving their operations. We study a scheduling problem on the work center level, i.e. for parallel cluster tools. The performance measure is the total weighted tardiness. All jobs are available at scheduling time. Two different biased random-key genetic algorithms (BRKGAs) are designed to assign jobs to cluster tools and to compute sequences for each single cluster tool. We establish algorithms to compute the time a job spent in a cluster tool, called processing time, given the sequence in which the jobs are processed there. They are embedded into the BRKGAs. Computational experiments based on randomly generated problem instances are carried out. The results demonstrate that the proposed metaheuristics perform well with respect to solution quality and computing time.