Civil Engineering April 2022 | Vol 30 No 3

Civil Engineering April 2022 17 A n inadequate understanding of ground conditions is often considered a leading cause of cost overruns on road projects. To investigate whether this was the case for South Africa, 25 major road upgrade and rehabilitation projects with at least five years of service were examined. Projects covered different geological conditions at both coastal and inland sites. Twenty of these projects performed satisfactorily, with no geotechnical or pavement failures within five years of completion, but for five performance was unsatisfactory, resulting in cost overruns or delays. METHODOLOGY Project and site investigation costs were collected from project records and pay- ment certificates from provincial and national roads agencies. In cases were remedial work fell outside the original project scope, costs were obtained from area-based consultants. In addition to remedial costs, details of suspected technical causes of problems were obtained from area-based consult- ants. While permission was granted by relevant agencies to collect data, no identifiable specifics are provided. FINDINGS Figure 1 illustrates that site investigation costs for unsatisfactory projects were signifi- cantly lower (t test, p < 0.001) than for satis- factory projects. However, cost discrepancies could not be linked to differences in work conducted (Figure 1b and c) or companies; as companies that performed site inves- tigations for unsatisfactory projects also performed work on satisfactory projects. Nevertheless, when considering the actual failures, insufficient data was avail- able and means to obtain the data were not utilised: Q Q Unsatisfactory case 1 : A single test pit provided to design a road cutting was subsequently supplemented by borings when the slope began to unravel. This caused a three-month delay and the original project cost increased by 0.4%. Q Q Unsatisfactory case 2 : Interpreting soil conditions on experience without laboratory tests failed to identify ex- pansive clays, resulting in a two-and- a-half-month delay and cost increase of 23%. Q Q Unsatisfactory case 3 : Insufficient labo- ratory tests to identify expansive clays resulted in pavement distress three years later and cost an additional 2% to repair. Q Q Unsatisfactory case 4 : Test pits 1 m deep failed to identify weak soils below a fill that resulted in slope failure, a delay of six months and a cost increase of 20%. Q Q Unsatisfactory case 5 : With no geotechnical issues in the past, no specific investigation was done for a culvert, which settled differentially on a clay foundation delaying the project by nine month and increasing costs by 2% for repairs. While it may seem prudent to ques- tion existing minimum site investigation costs or the number of tests per kilometre (or structure or project cost), the cases reviewed show that these can only ever be a guideline. Site investigations must be planned to investigate anticipated prob- lems based on a review of local geological conditions.  Stellenbosch Uni vers i ty geotechni cal research abstracts Site investigations and road project cost overruns Sbekezelo Brian Nkwanyana Candidate Engineer KwaZulu-Natal Department of Transport sbekezelobrian@gmail.com Dr Charles John MacRobert Pr Eng Senior Lecturer | Department of Civil Engineering Stellenbosch University macrobert@sun.ac.za Stellenbosch University has compiled a series of geotechnical research abstracts outlining research undertaken by the Department of Civil Engineering. The first abstract examines cost overruns on road projects as a result of geological conditions. Figure 1 Site investigations for road projects: a) Associated costs, b) Test pits conducted, c) DCPs conducted