This paper presents observations of buffer implementation and usage during the construction of a five-story, full-scale, reinforced concrete, earthquake test structure. Over 40 private industry partners performed a majority of the construction work inkind, each covering a different scope of work. Activity durations were often longer than expected, particularly because research interests often resulted in changes in scopes of work for partners, sometimes requiring design work. Due to the fluctuation between the initial baseline construction lookahead schedule developed early in the project, and the pace of actual construction, inventory often sat in the laydown area or in the workface area for lengthy periods of time. This was true even though a pull approach was used to coordinate activities. A simulation experiment was used to determine the impact on inventory levels of activity durations exceeding planned durations. Because material lead times were on the order of the planning horizon, orders for upcoming activities were often made without knowledge of delays in intermediate activities, resulting in additional inventory time on the site. The simulation experiment exhibited a similar behavior, and showed that excess inventory levels increase rapidly with the degree to which actual durations extend beyond planned durations.
Buffers, production control, inventory control.
Espino, E. , Aranda, C. , Walsh, K. , Hutchinson, T. , Restrepo, J. , Hoehler, M. & Bachman, R. 2012, 'Relationship of Time Lag Buffer to Material Stockpile Buffer Levels' In:, Tommelein, I. D. & Pasquire, C. L., 20th Annual Conference of the International Group for Lean Construction. San Diego, California, USA, 18-20 Jul 2012.