The DOE study varied the head, nozzle diameter, jet inclination angle and aim position, all parameters identified as critical by the model developed in Section 2. The DOE results in Fig. 14 show that the efficiency of the turbine is sensitive to all of these parameters, most especially to the nozzle diameter and aim position. The smal- ler nozzle provides an efficiency improvement of 15% on average, and aim position provides over 20% change in efficiency over the range of positions. Section 6.2 discusses the improvement due to reducing the nozzle diameter. As the aim position moves down the cup not only does the volume efficiency change, as shown in Section 2.2, the jet impacts earlier in the rotation cycle, reducing the moment arm of the jet on the turbine and therefore reducing the output torque. The maximum efficiency obtained from the DOE and optimisation was 91%, with the aim at 6 mm from the top of the cup, inclination angle of 10, a nozzle diameter of 20 mm and head at 3.5 m. It was found in the DOE study that the optimum jet inclination angle was 10, but implementing this at a larger scale, this may not be the case. With a higher jet inclination angle, the nozzle exit can be placed closer to the impact location without the nozzle struc- ture hitting the turbine disc. With the nozzle closer to the turbine disc the jet will expand less, droop less due to gravity, experience lower losses due to windage and have a reduced chance of interfer- ence between splashing water and the incoming jet [12]. A similar set of tests were carried out to those in Fig. 16, but this time at a 20 jet inclination angle. It was found that the maximum efficiency was approximately 90%, when the aim point was 11 mm from the top of the cup. In all the tests apart from the final cup design study, the cup de- sign was held constant. Further efficiency improvements could be found by redesigning the Turgo cup, especially the trailing edges. An improved design on the trailing edge of the cup can reduce the amount of kinetic energy leaving the wheel, as is achieved with larger scale turbines [13]. As the jet spreads out on the trailing edge of the cup, Fig. 12, the flow deviates from the modelled flow and the absolute velocity of the flow is increased. However, as the turbine is required to operate over a head variation, the rotational speed will be dependent on the site and so the design cannot be optimised for all possible locations. The turbine tested was a scale model which can be designed to full scale using non-dimensional analysis [14]. Using these laws and assuming similarity between the experimental turbine and the full scale version, a turbine that generates 1.3 kW at 3.5 m head