A dissertation presented to the Higher Degrees Committee of the Cape Peninsula University of Technology in fulfilment of the requirements for the degree of Master of Technology: Construction Management, 2012 / This study investigated the underlying causes of rework in construction projects and the impact on overall project performance so that effective containment and reduction strategies can be developed. The objectives of the study were as follows: (i) to determine the influence different project types have on the causes of rework in construction projects; (ii) to determine the impact of rework on organisational and project performance; (iii) to determine the influence various project types have on rework costs (direct and indirect) in construction projects; (iv) to determine the influence various procurement methods have on total rework costs in construction projects; (v) to design and develop rework reduction and containment strategies. The research was motivated by several international and local studies demonstrating a lack of concern for the root causes of rework and the potential impact on cost, overall project performance, and the ‘value-addedness’ to the completed project. The research approach adopted included an exploratory and main study targeting purposively selected construction professionals and stakeholders in the Cape Peninsula metropolitan area of the Western Cape Province. The exploratory case study was carried out at the initial stage of the study to gain more insight into the causes and impact of rework on overall project performance. Specifically, data was collected by means of observation of physical works, semi-structured interviews with relevant parties directly involved in site operation and the analysis of site instruction record documents. The main study obtained data from 78 construction professionals and stakeholders via questionnaire survey, a survey conducted among design consultants and contractors in the general building category ranging from grade 3 to 9 who are registered with Construction Industry Development Board (CIDB). Descriptive, inferential statistics and probability distribution functions were used to analyse the data. The findings revealed that changes initiated by the client, changes initiated by the design team due to errors and omissions, poor coordination, and finally, integration among the design team were the major contributing factors to rework. Moreover, non-compliance with specifications, setting-out errors, low labour skills, and emphasis on time and cost aggravated the occurrence of rework on site. The study revealed that while there is no significant difference between the causes of rework and various project types, rework can and often does make a significant contribution to any project’s cost overrun. The total mean cost of rework as a percentage of the original contract value for new build project and refurbishment/renovation projects was 4.89% and 6.28% respectively. However, rework costs do not differ relative to project type or procurement method. Furthermore, the study revealed that cost overruns, time overruns and design team dissatisfaction all impacted on project performance. The findings indicate that designrelated rework can be minimised by implementing the following strategies: team building, involvement of subcontractors and suppliers, and design for construction. Moreover, involvement of subcontractors during construction, and the implementation of quality control and site quality management systems could also lead to reduction in rework during the construction phase. Furthermore, the probabilistic analysis of rework occurrence was determined in the projects selected; this analysis predicts the occurrence of rework so that a quantitative risk assessment could be undertaken prior to the commencement of construction. The research concludes by recommending that design and construction firms must develop organisational measurement systems for recording rework occurrence and its associated costs. It is by determining the frequency and costs of rework that effective strategies for its containment and reduction can be identified.

In general, dreams are conceived, ideas are conceptualised, and initiatives are embarked upon in order to alter the state of realities. Dreams change realities; when mechanisms are put in place to realise them, dreams succeed. However, anecdotal evidence as well as empirical findings has continued to reiterate the difficulties associated with realising dreams related to construction projects. Extending the „dream‟ analogy to the South African construction industry context therefore paints an uninspiring picture. Dreams associated with construction do not have a 100% chance of becoming realities as evident in reported poor project performance in the industry. Shattered dreams in the form of poor performing projects, poorly implemented construction processes, or worst, projects delivered at the expense of unexpected cost to the client as a direct result of poor H&S or time overruns, negate the intent of dreams. This thesis is primarily concerned with project performance related bottlenecks in South African construction. After an extensive review of related literature that entails the analysis of publications related to non-value adding activities (NVAAs), supply chain management (SCM), and system dynamics (SD) in the construction project management realm, an exhaustive mixed-mode quantitative survey was conducted among key participants in the South African infrastructure sector. Public sector clients, consulting engineers and contractors that were involved in civil engineering projects were surveyed repeatedly with approximately five survey instruments at convenient intervals. Results arising from the study, inter-alia, indicate that: an appreciable amount of NVAAs occur in South African construction; these NVAAs become further compounded when propagated into other value adding activities (VAAs) in the construction process; the identified NVAAs equally perpetrate the menace associated with poor performance to the detriment of the achievement of cost, H&S, quality, and time project targets, and the root cause of these NVAAs that often contribute to poor performance is not far from the much reported „shortage of skills‟ in South Africa. Notable contributions to the body of knowledge include SD models are extendable regardless of the source of their empirical data as evident in the qualitative models proposed in this study; within the SD domain, it is advisable to consider the „competence‟ of individuals assigned to tasks especially in a developing country as this study revealed that human resources issues predominate among the sources of NVAAs that eventuate in a range of poor project performance; the NVAAs that occur, and their causes on projects are perceived to be due to lapses and / or inadequacies that involved the entire construction supply chain; there is no single construction process / task that is immune from being affected by NVAAs; and within the South African, and by implication construction context generally iii in developing countries, the adequacy of required knowledge among project stakeholders is the most crucial determinant of project performance. As opposed to what is obtainable in developed countries, the construction industry in developing countries, particularly in South Africa, should take advantage of knowledge management (KM) techniques such as brainstorming, communities of practices, and face-to-face interactions. These techniques can be driven through appropriate mentorship programmes, industry focused built environment education, and other human resources driven avenues that do not necessarily require substantial investment in technologies, so that to a large extent organisations in the industry can prioritise KM, and thereafter, continually engage in it for future performance improvement. Using inferential statistical methods for hypotheses testing, and SD concepts for creating qualitative models led to a range of recommendations which, inter-alia, propose that halting the tide of NVAAs and poor performance requires the management of both tacit and explicit knowledge gained in construction; and most importantly, it requires the assurance that „competence‟ is the overriding criteria for selecting project partners, and also, for assigning either design or construction related activities to responsible parties involved in project realisation in South Africa. In effect, in order to engender a culture of continuous improvement in South African construction, other considerations should be subservient to „competence‟ in the construction supply chain. Competence must be located among everyone involved in project realisation, that is, enhancing the competence of all involved in project realisation is as good as ensuring performance improvement, which in turn, equates to the acceleration of project delivery in South Africa.