A large and under-developed source of renewable energy exists, due to the properties and associated processes attributable to the substantial and ubiquitous presence of water, globally. Hydrokinetic energy conversion refers to the conversion of kinetic energy in moving water to electricity. It offers an alternative to conventional hydropower, with benefits of modularity and scalability, in addition to being environmentally and socially less impactful. Hydrokinetic energy conversion can benefit isolated communities currently without access to electricity.

This study aims to determine the global theoretical riverine hydrokinetic resource. A 35 year modelled daily discharge data set and vectorised representation of rivers, with near-global coverage and suitable spatiotemporal resolution, is used to determine the mean annual energy yield of 2.94 million river reaches. Two methodologies are applied, considering the hydrokinetic resource as: a conversion from gravitational potential energy to kinetic energy; and, directly, as the transfer of kinetic energy.

Using the former method, and aligning to the convention of previous large-scale hydrokinetic resource assessments, the mean global resource (excluding Greenland and Antarctica) is estimated to be 58000 TWh/yr (6.7 TW). Consideration of global spatial distribution, by river reach, illustrates regional variation and shows a tendency for potential to be concentrated along major rivers and in areas of significant elevation change. China, Russia and Brazil are found to be the countries with the greatest potential. After normalisation by total river length, Bhutan, Nepal and Tajikistan also show great potential.

With the latter method, a novel approach is proposed, and an argument is presented to suggest that there are pragmatic advantages to the perspective this provides. In this case, the mean hydrokinetic energy of global rivers (excluding Greenland and Antarctica) is estimated to be 1.642 TWh (5.911 PJ). In contrast, this time the overall resource is quantified as a measurement of hydrokinetic energy, rather than power. The two approaches differ in their perspective, with the former giving a Lagrangian description and the latter a Eulerian description. With this latter approach, the resource is shown to have great potential on the continents of South America, Asia and Africa. Hydrokinetic power within individual reaches is considered, as before, and comparable with the earlier method. Yet, due to the Eularian view, in this case, global power cannot be quantified, reflecting an alternative perspective of the global riverine hydrokinetic resource. This novel perspective concludes that major rivers, and particularly the lower-courses of major-rivers, are found to offer the most potential for hydrokinetic energy conversion. Areas with large elevation change are not found to be as significant, using this method. Recognising the importance of major rivers, basin-scale resource assessment of prominent rivers identifies the Amazon, Ganges and River Plate as those offering the greatest potential for hydrokinetic energy conversion.

Comparison with a field-scale study in Lithuania, that used a hydraulic model at 3 locations, demonstrates the extent to which the conventional method results in significant over-estimation, with a percentage bias of +11000%. In contrast, the novel method proposed is found to have a percentage bias of +40%. Large uncertainties in the parameters used in the estimation of channel form and flow velocities in this method certainly affect precision, but potentially offer a more accurate perspective. The limitations of this imprecision are improved by application at large-scale and the use of a Monte Carlo approach. For this reason, this methodology is more appropriately applied in this context, rather than at reach-scale. Improvements in large-scale determination of hydraulic geometry parameters offer a promising way to improve this methodology, potentially providing a more precise and accurate perspective of the global distribution of riverine hydrokinetic resource.