Modular and scalable low-head pico-hydro generation for off-grid networks

Over 1.3 billion people across the world do not have access to electricity, with most living outside urban environments. In these rural, remote locations with low population density, the cost of extending the grid is very high. If a location has a river nearby, pico-hydropower is a cost-effective wa...

Full description

Bibliographic Details
Main Author: Williamson, Samuel
Published: University of Bristol 2013
Subjects:
Online Access:http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.616869
Description
Summary:Over 1.3 billion people across the world do not have access to electricity, with most living outside urban environments. In these rural, remote locations with low population density, the cost of extending the grid is very high. If a location has a river nearby, pico-hydropower is a cost-effective way of providing off-grid electricity. Typically pico-hydro units are stand-alone and can only provide enough power for basic domestic services such as lighting and mobile phone charging. However, if these pico-hydro units were connected together in a network, the available power could scale up and provide a more reliable source to support health posts, schools and income generation activities such as grain processing, as well as domestic demand. This thesis investigates the interconnectivity of low-head pico-hydro modules to form a scalable off-grid network. A specification for a system is derived from user requirements and the current state of technology resulting in a concept for the network. Each unit is to be identical, operating at low head and connecting onto an AC grid using a power-electronic front-end . Using this specification, a new methodology is developed using weighted quantitative and qualitative criteria to select a turbine to match the specification. A single-jet Turgo turbine is chosen; a turbine design not commonly used at low heads. Using a low-speed generator and power-electronic front-end it is shown to satisfy many of the quantitative and qualitative criteria in this case. A quasi-steady-state 2D model for the Turgo turbine is developed with a jet interception efficiency model that calculates the percentage of jet that impinges the turbine cup. An experimental set-up for the turbine is designed and scaled performance tests are carried out and compared with the model. Using a Design of Experiments method and further optimisation tests, a Turgo system arrangement was found with a maximum jet-to-mechanical efficiency of 91 % at 3.5 m head. The generator unit, made up of turbine, generator, rectifier, DC-DC converter and inverter, is simulated in Simulink. The generator units do not communicate between themselves, so each inverter uses a modified version of droop control for low voltage grids where the output power is proportional to the available turbine power. This allows each unit on the network to support voltage and frequency regulation on the AC grid. A simulation of the system is developed and different control parameters are varied to understand their effect on the grid output and unit synchronisation. The control system is then validated experimentally using hardware-in-the-loop techniques, showing how the proportional droop control is able to successfully share load between generator units of different power rating, whilst mitigating distance between the source and load. Finally, a full-scale generator unit is designed and mocked up in the installation environment. The unit is modularised to aid fault identification and part replacement by unskilled labour, with an estimated cost of £1300 per unit, which can generate 1 kW.