Many real-world problems are dynamic with mul-tiple, often conflicting, objectives, referred to as dynamic multi-objective optimisation problems (DMOPs). Furthermore, many problems' changes occur at varying intervals, with varying frequency and severity. Benchmark functions are used to eval-uate the performance of dynamic multi-objective optimisation algorithms (DMOAs) by representing various characteristics that are representative of typical real-world problems. This paper presents an approach to change existing benchmarks into DMOPs incorporating changes at varying intervals with varying frequency and severity. To demonstrate the approach, three variations of the dynamic non-dominated sorting genetic algorithm II (DNSGA-II) are evaluated on both benchmarks with changes occurring at the same interval and with the same frequency and severity throughout the run, as well as benchmarks with changes occurring at varying intervals with varying frequencies and severities of change.