The rapid development of microfluidic biochips requires matching automated synthesis methods. In particular, high-level synthesis methods for microfluidic chips need to consider various bio-constraints regarding time and device occupancy. Recent biochemical applications show that the duration of some bio-operations cannot be predicted in advance, which increases the likelihood that current synthesis methods would waste on-chip resources or even violate given bio-constraints. In this work, we present a relative-scheduling-based high-level synthesis method to optimize the bio-assay schedules and the usage of on-chip devices considering bio-operations with indeterminate durations. Experimental results show that our method significantly reduces the total execution time of bioassays without violating bio-constraints when using the same on-chip resources.