As conventional oil reserves are depleting, the rising energy demand as well as advancements in well drilling and stimulation technologies have attracted increasingly interests in exploiting the unconventional oil resources (e.g., tight oil and heavy oil). Tight oil resources in Canada are mainly located in Cardium and Viking pools in Alberta, Bakken formation in Saskatchewan, and Western Canada Shale plays, respectively. Also, tremendous heavy oil resources accounting for almost half of the total world heavy oil reserves are mainly found in Western Canada, though they are contained in thin and depleted reservoirs. Such unconventional oil resources either have been deposited in extremely unfavourable environments or possess high carbon intensity, requiring more efforts and energy to recover than conventional oils. In practice, three-dimensional (3D) displacement models show unique advantages for designing appropriate well configurations, though their physical constraints make it impossible to duplicate the real reservoirs under certain conditions. Therefore, it is of fundamental and practical importance to develop scaling criteria for describing fluid flow behaviour in unconventional reservoirs. Scaling criteria have been developed and validated to evaluate the performance of both waterflooding and immiscible CO2 flooding in tight formations. Experimentally, saturation pressures of the CO2-light oil systems are determined with a versatile PVT system and then used to tune the binary interaction parameter (BIP) correlations between CO2 and the lumped pseudo-components. Subsequently, waterflooding and immiscible CO2 flooding have been respectively conducted with light oil in the 3D physical model. Theoretically, mathematical formulae have been derived to reveal fluid flow of immiscible displacements in tight oil reservoir by performing dimensional and inspectional analyses. Since not all of the scaling groups can be satisfied, relaxation of the scaling groups will be made to neglect capillary pressure force. Scaling criteria have been validated by history matching the experimental measurements, and then extended for field applications. Geometric factor is demonstrated to be negligible while gravitational and viscous forces have been considered for scaling up waterflooding and immiscible CO2 flooding. Since capillary pressure and relative permeability have a negligible effect on general movement of the displacing fluid, sacrificing these two parameters has been proved permissible. Scaling criteria have then been modified and validated to evaluate performance of waterflooding and immiscible CO2 flooding in heavy oil reservoirs by using the 3D sandpacked displacement model. Experimentally, waterflooding and immiscible CO2 flooding of heavy oil have been conducted with 3D sandpacked models. Theoretically, scaling criteria have been relaxed by neglecting the ratio of gravitational force to viscous force while other scaling groups including geometric factor, diffusion groups, and ratio of capillary forces to viscous forces can be satisfied. The relaxed scaling criteria are validated by comparing the simulation results of synthetic reservoirs with experimental measurements, and then extended for field applications. There exists a reasonably good agreement between the laboratory measurements and the simulation results of the synthetic models. This is consistent with the findings from literature that, as for thin heavy oil reservoirs (i.e., ratio of well spacing to thickness and mobility ratio larger than 50) gravity effect is negligible.