TY - JOUR
T1 - Pipe Networks Risk Assessment Based on Survival Analysis
AU - Tsitsifli, Stavroula
AU - Kanakoudis, Vasilis
AU - Bakouros, Ioannis
PY - 2011/11
Y1 - 2011/11
N2 - Integrated management of pipe networks should include methods for monitoring, repairing and replacing deteriorating components (usually pipes), but also methods and everyday operating practices towards a proactive risk assessment approach in order to give a solid answer to the unavoidable "repair or replace" dilemma. The present paper attempts to check whether the Discriminant Analysis and Classification (DAC) method can be used to achieve the above mentioned goals and predict the future behaviour of network pipes. Three pipe networks carrying different types of fluids (oil; gas; and water) are used as case studies. For each case study network, the DAC method is used to classify the pipes into two groups (failures/successes), based on simple variables (pipe/network characteristics) and dimensionless joint ones. Several scenarios are being analysed for each case. The results for the two cases of oil and gas networks are very satisfying. The implementation of the DAC method to water pipe networks needs to overcome serious problems related to the quality, reliability and compatibility of the data records provided by the Water Utilities. In this paper, these shortcomings are faced combining field data with theoretical one. Also the distinction between what "failure" and "success" actually mean in a water pipe network has to be determined. The present study uses the total water volume being lost as a definition criterion.
AB - Integrated management of pipe networks should include methods for monitoring, repairing and replacing deteriorating components (usually pipes), but also methods and everyday operating practices towards a proactive risk assessment approach in order to give a solid answer to the unavoidable "repair or replace" dilemma. The present paper attempts to check whether the Discriminant Analysis and Classification (DAC) method can be used to achieve the above mentioned goals and predict the future behaviour of network pipes. Three pipe networks carrying different types of fluids (oil; gas; and water) are used as case studies. For each case study network, the DAC method is used to classify the pipes into two groups (failures/successes), based on simple variables (pipe/network characteristics) and dimensionless joint ones. Several scenarios are being analysed for each case. The results for the two cases of oil and gas networks are very satisfying. The implementation of the DAC method to water pipe networks needs to overcome serious problems related to the quality, reliability and compatibility of the data records provided by the Water Utilities. In this paper, these shortcomings are faced combining field data with theoretical one. Also the distinction between what "failure" and "success" actually mean in a water pipe network has to be determined. The present study uses the total water volume being lost as a definition criterion.
KW - Discriminant analysis and classification
KW - Pipe networks
KW - Survival analysis
UR - http://www.scopus.com/inward/record.url?scp=80255141902&partnerID=8YFLogxK
U2 - 10.1007/s11269-011-9881-3
DO - 10.1007/s11269-011-9881-3
M3 - Article
AN - SCOPUS:80255141902
SN - 0920-4741
VL - 25
SP - 3729
EP - 3746
JO - Water Resources Management
JF - Water Resources Management
IS - 14
ER -