Static and dynamic reservoir modelling are common to predict resource potentials and flow behaviours in geological reservoirs. However, existing reservoir modelling workflows often create barriers between geoscience and engineering. So called "flow diagnostics" (FD) can overcome this problem. FD approximate and quantify the reservoir dynamics using intuitive measures such as the time-of-flight (Fig. 1). The seminal work at SINTEF demonstrated how FD can be computed in the matter of seconds, providing real-time feedback on reservoir dynamics while the geological model is built and changed. This allows geoscientists and engineers to jointly explore how uncertainties in geology impact flow behaviours before commencing further reservoir studies. FD also enable rapid screening of a large ensemble of reservoir models to select a smaller subset of models for additional studies without compromising on the geological uncertainties that have been captured in the original ensemble (Fig. 2). FD are hence suited to analyse reservoir behaviours in challenging situations where subsurface data is limited and geological uncertainty is high, such as (fractured) geothermal reservoirs. FD therefore enable value-based decisions when characterising and developing geothermal reservoir and help to mitigate risks. Examples from developing geothermal reservoirs in Germany have shown that acquiring the right additional geological data (e.g. 3D seismics) is very valuable to constrain uncertainties and reduce risks (e.g. dry wells) that have adverse impact on public acceptance and project economics. This PhD project aims to adapt FD that were originally developed for (fractured) hydrocarbon reservoirs to geothermal reservoirs. Year 1 of the project will focus on developing and implementing a new FD methodology, using existing mathematical theories developed by the supervisors to model the movement of thermal fronts, geomechanical effects, and fracture-matrix interactions. Year 2 will focus on a proof-of-concept study that demonstrates how the new FD technology can be used to screen an ensemble representing a semi-synthetic geothermal reservoir to select an appropriate subset of reservoir models without compromising on the ability to forecast and optimise reservoir performance under geological uncertainty. Year 3 will focus on a real-field application, showcasing how the reservoir modelling and simulation studies that are carried out under the NERC-funded GWatt project at the United Downs Deep Geothermal Project in Cornwall can be accelerated and improved using FD.