One of the most important challenges in planning and main-
taining a digital repository is to predict the needed resources
on a long term basis, especially storage size and processing
power. The main problem emerges from the need to mi-
grate the data at certain times to newer le types, which
takes time and alters the needed storage space, potentially
branching into several migration paths for individual ob-
jects. Understanding the e ect of di erent policy decisions,
such as when to migrate or whether to stay within a format
family or branching into several format families turns into
a complex task, speci cally when considering non-trivial in-
gest structures and assumptions on format evaluations. In
this paper we present ReproSim, a framework that simulates
the evolution of a digital repository and helps predicting
these factors. We demonstrate the complexity and power of
simulation to assist in preservation decisions in a set of sce-
narios involving di erent ingest and preservation planning
pro les.