The study attempted to analyse the impacts of the ‘fast track’ land reform on maize production in Zimbabwe. This purpose was tackled by constructing a partial equilibrium model that depicted what could have happened if no further policy shifts had taken place after 2001. Setting up a partial equilibrium model required a sound understanding of the functioning of the Zimbabwe’s maize market. The institutional structure of the Zimbabwean maize market was explored to inform the model development process that would allow for the development of the baseline model. Developing the model started off with the estimation of single equations which were collapsed into a simultaneous system of equations through the use of a combination of ordinary least squares and generalised least squares techniques. The development of the simulation model required that assumptions be made for exogenous variables, and crafted assumptions were based on the 2000 macro-economic and institutional environment as well as agricultural policies. The re-simulated baseline model that was constructed in this study was used to make projections based on the various trends of exogenous variables in 2000. This means that the model generated an artificial data set based on what the maize market would have looked like under a set of the pre-2000 existent policy conditions. As such, all the shifts in the political and economic environment that took place after 2000 were not introduced in the model. The ‘fast track’ land reform policy was thus assessed based on the performance of the baseline model using a range of “what if” assumptions. Therefore, the re-simulated baseline solutions discussed result not only from policy shifts that occurred before 2000, but also from the convergence of hypothetical political and economic stability within the period in question. The results of the re-simulated baseline indicated that the commercial area harvested was negatively affected by the expropriation of commercial farms. The arguments in literature that the ‘fast track’ land reform policy shift contributed the loss in area planted owing to the stalling of farming operations due to political unrest, economic instability and input shortages were supported by the model results which showed that total area harvested would have been higher under pre-2000 conditions. From the re-simulated baseline results, the difference between actual and would be outcomes revealed that the total maize production was 13.27% less than what could have been produced in 2001, the year that the ‘fast track’ land reform policy was formally implemented. In view of the 2002/03 drought, output was 57.44% less and 33.53% less than what could have actually been produced for the 2002 and 2003 seasons respectively. In the 2005 drought season, the total maize production was 41.8% less than what could have been produced without the ‘fast track’ land reform. This may imply that droughts would have been less severe if the ‘fast track’ land reform was not implemented. In 2007, the baseline showed that the nation could have produced almost 48.03% more than what was actually produced. Therefore, according to the model results, the assertion that the ‘fast track’ land reform contributed, to a fair extent, to the underperformance of the maize sector still holds. The model developed in this dissertation contributes to an understanding of not only the general structure of the maize market, but also of the impact of the ‘fast track’ land reform policy on the Zimbabwean maize market based on how the market itself could have performed under the absence of these land reforms. The baseline model revealed that the maize sector performed below potential within the period of the ‘fast track’ land reform. The maize market model could thus be used as a tool that may assist policymakers to design future strategies that will help enhance maize sector performance. / Dissertation (MSc(Agric))--University of Pretoria, 2010. / Agricultural Economics, Extension and Rural Development / unrestricted

The earliest design of a wind power system with multiple rotors on a single support structure dates back to the late 1800s. Such a system called a Multi-Rotor Wind Turbine (MRWT) was proposed by several researchers due to its perceived advantages over a single-rotor wind turbine. As turbine size increases, power produced by a rotor tends to scale up as the square of its diameter, as opposed to rotor weight which varies as its cube. So, several smaller rotors will weigh and cost less than one large rotor producing the same power. MRWTs offer several advantages such as better distribution of loads, better logistics of the components and scope for standardization. The MRWT system can also continue operation even if some of the rotors fail. However, MRWTs require a complex support structure to connect the rotors to the tower and an arrangement to yaw them into the wind. A recent study involving a scaling model for a three-rotor MRWT system estimates a cost saving of 13.1% as compared to the NREL 5 MW single-rotor model. A triangular truss type support structure for the MRWT model is designed and its preliminary static analysis is performed in that study. This thesis is a continuation of that study where the scaling model is extended to include MRWT systems having two to seven rotors. A systematic design method is developed for modeling any MRWT support structure for two to seven rotors for the given 5 MW configuration. The structure consists of frames and cables and the design constraints for the static analysis are stress, deflection and buckling. A dynamic analysis of the MRWT solution is also carried out to verify that the structure can withstand loads induced at varying wind conditions and design load cases – especially steady, turbulent and extreme wind conditions. Some special cases for the three-rotor MRWT system, such as use of two-bladed rotors, direct-drive machines, analysis for zero wind loads, load analysis for each of the assembly stages are also discussed. Finally, as the support structure design for the three and seven-rotor models is the main focus of the thesis, the scaling model is validated by comparing these models with similar turbines having rated power corresponding to the rotors used in the models.