| dc.description.abstract | The intent of this study was to provide an empirical analysis of project performance between the traditional design-bid-build (DBB) method and alternative design-build (DB) method of project delivery for highways. The study examined five major performance metrics: cost growth, schedule growth by notice to proceed, schedule growth by construction start date, award growth and construction engineering inspection cost factor. The data were collected from six selected state departments of transportation (DOT): Florida, Indiana, Oregon, North Carolina, Ohio, and Utah. These six DOTs have significant experience on using the DB project delivery method. Totally, more than 15,000 projects were mined and analyzed. To create a comparable pair between DBB and DB projects, six contract size bins were defined: projects under $2M, $2M - $5M, $5M - $10M, $10M - $20M, $20M - $50M, and over $50M. Performance data were collected from the six states by direct DOT official interviews and were mapped to the exploratory metrics. Three rounds of mining were conducted to accomplish the comparable sets of projects: missing data point removal, outlier removal, and project pairing by the bin-sampling method. Projects were matched one-to-one between DBB and DB at +/- 15% of contract award amount and +/- 1 year of respective project construction start date. The results of analysis showed that on average, the cost growth for DB projects is higher than the cost growth for DBB projects for smaller project ranges and lower than DBB projects for larger project ranges. The schedule growth based on notice to proceed dates and construction start dates on DB was found less than that for DBB projects across all sampled states and for all project sizes. DBB produced more negative award growth than DB for all contract size bins. For CEI factor, DBB showed higher mean values than DB for five out of the six contract size ranges. | |
