Analyses of radiation fields resulting from a deuterium-tritium (DT) plasma in fusion devices is a critical input to the design and validation of many aspects of the reactor design, such as shielding, material lifetime and remote maintenance requirements/scheduling. Neutronics studies are performed using radiation transport codes such as MCNP, TRIPOLI, SERPENT, FLUKA and OpenMC. The Serpent 2 Monte-Carlo code, developed by VTT in Finland, offers a promising alternative to Monte-Carlo N-Particle (MCNP), which at present is the most commonly used code among the fusion neutronics community. There are several fundamental requirements for such a code, namely, the ability to capture the system in all its complexity, employ acceleration techniques, and have parallelisation capability for deployment on high performance computer architectures. This work focuses on the benchmarking the Serpent 2 code through performing a selection of the benchmarks in the established radiation shielding and dosimetry experiments database (SINBAD). In total, 31 fusion relevant benchmarks are present in the database, characterised by DT neutron irradiation sources. A subset of the benchmarks from SINBAD are selected to perform some initial validation on the consistency of both the particle transport models and interpretation of the nuclear data. The benchmarks are distributed with an MCNP model of the experimental set-up, and translation to a Serpent input file is automated using a python script. Experimentally determined quantities such as reaction rates in activation foils, the tritium breeding rates and neutron/photon flux spectrum are calculated, and a comparison is made between the two transport codes and the experimentally derived results. We extend this validation by utilization of several different evaluated nuclear data libraries relevant to the defined energy range, including FENDL and JEFF. Dosimetry cross sections using IRDFF relevant to activation foil analysis are applied where the data is available. To extend the benchmarking of Serpent 2 to next step fusion reactor analysis, a Serpent model of an EU DEMO concept is generated. We extract several nuclear quantities and quantify the differences between MCNP and Serpent. To extend this analysis beyond in-vessel regions, we discuss recent developments in variance reduction capabilities in Serpent, as well as further work which is being conducted collaboratively in parallel with the Serpent development team.