Comparison of bacterial regrowth in distribution systems using free chlorine and chloramine: a statistical study of causative factors.

Bacterial regrowth was investigated over a 15-month period in distribution systems (DSs) of Durham and Raleigh in North Carolina. These two water utilities were chosen because they are adjacent to one another, have similar service area characteristics, and treat surface waters of similar characteristics with conventional processes (coagulation-sedimentation and dual-media filtration). The finished waters have similar chemical quality and regrowth potential as measured by assimilable organic carbon (AOC). The major difference in treatment is the choice of final disinfectants (chlorine in Durham and chloramine in Raleigh). Ten sampling sites (monthly sampling) were chosen in each system to give wide geographic coverage and correspondingly, a wide range of water residence times. Significant losses were observed in both chlorine and chloramine residual in the DSs that produced bacterial regrowth as measured by heterotrophic plate count (HPC). The frequency distributions for log HPC (133 observations from Durham and 135 observations from Raleigh) were statistically the same in the chlorinated and chloraminated DSs. A correlation analysis indicated that disinfectant residual is the most important factor determining HPC level. However, the resulting R2 value for a non-linear regression model that also included AOC, temperature, and pH as independent variables was less than 0.7. Bacterial regrowth as measured by HPC, is dependent upon a complex interaction of chemical, physical, and operational parameters that may not be captured by such a simple statistical relationship.