The role of concentrating solar power in a South African large scale renewable energy scenario



Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
This thesis investigates a hypothetical large scale renewable energy system for South Africa over the long term future using 2050 as a nominal year of analysis. Simulation modeling was used to assess the integration of three renewable energy technologies, which are currently being installed at utility scales in South Africa, at much higher penetrations. These technologies, namely wind, photovoltaic and concentrating solar power, are studied here within future scenarios (2050) over a period of one year at hourly intervals using real weather and solar irradiance data from 2010-11. The electricity demand profile for these scenario years was projected by scaling up the 2011 South African demand profile to the total annual energy demand projected for 2050 by the Draft South African Integrated Energy Plan reference case. The dispatch of electrical energy from CSP plants with thermal energy storage has the potential to support other more intermittent renewable energy technologies. To evaluate this potential role, CSP production and available storage is assessed over one year relative to the production from wind and photovoltaics. Electrical energy is dispatched from thermal energy storage in a controlled fashion, thereby complementing the use and integration of more intermittent technologies. The effects of using thermal storage are monitored with regard to system cost and impact on overall system reliability. The results of the modelling exercise identified potential benefits gained when using thermal energy storage to backup intermittent renewables. These results were consistent with those reported in the literature. CSP plants which are equipped with thermal energy storage facilities are termed ‘dispatchable' due to the amount of electricity that can be dispatched when needed with reasonably good reliability. The additional costs incurred to build CSP plants, relative to wind and photovoltaic plants, can be offset by better system reliability. A more expensive dispatchable system using thermal energy storage displays better reliability. When utilising the more expensive fully dispatchable renewable energy system which is defined here by the use of thermal energy storage primarily being used to back-up other renewables, the fractional increase in cost can be argued on the grounds of increased fulfilment and decreased curtailment achieved by the renewable energy technologies installed on the grid. This allows for improvements to the economics of clean energy and lowers total carbon emissions produced by the supply of electricity. Scenarios of high penetration of the three renewables in different ratios, with unfulfilled energy being met by a single representative fossil fuel alternative (natural gas fired combined cycle gas turbine), were compared to each other using a range of indicators with cost of generation being determined with the Levelised Cost of Electricity (LCOE) method. The cost of renewable energy technologies has declined in recent years and is envisioned to continue to become more costcompetitive with conventional generation technologies. To account for this expected reduction in renewable energy costs, technology learning has been applied to the LCOE calculation. Additionally, a sensitivity analysis of one of these scenarios assesses the LCOE of a system with increasing RE penetration, spanning from low RE with 95% of energy supplied by gas-fired CCGT to high RE with 26% of energy supplied by gas-fired CCGT. Technology learning rates, as estimated by the Integrated Resource Plan for 2030, are used in the interests of being conservative on estimating actual costs in 2050. When technology learning rates are assumed for 2030 and a fuel cost of R860.38 per MWh electricity generated for gas is used, the three RE technologies within the modeled system can iv contribute 85% to total energy requirement over the year (TWh) and attain an LCOE of R0.75/kWh. In the event that the same system is moved towards a heavier CCGT reliance of 95%, the LCOE is raised to R0.96/kWh due to the impact of expensive variable operating and maintenance costs for CCGT. Capital costs for the different RE technologies assessed here are expected to be lowered by continued learning from 2030 to 2050.