CHAPTER 4 ANALYSIS OF DECISION VARIABLES
4.2 FAULT DATA ANALYSIS
108
109
3rd party damage Mechanical ferrule Pulled off main/seal Corrosion
Poor workmanship Faulty Product 3rd party damage Hole (in ferrule body) Corrosion
Faulty Product 3rd party damage
Bolt Failure Corrosion
Poor workmanship Faulty Product 3rd party damage Electro-fusion fittings
(incl. top-tees/saddles) Leak/ blown off /pull out Faulty Product Poor workmanship Overpressure 3rd party damage
Mechanical Fitting Seals Ground movement
Loose/too few bolts Corrosion
Faulty Product Poor workmanship Overpressure/ surge 3rd party damage Hole (in fitting body) Corrosion
Faulty Product Overpressure/surge 3rd party damage Service pipes As for pipe
Valve/meter As for mechanical fittings
EVIDES BV Faults Database
Meijer, (2012) reported that the fault data detailed in Table 4.6 is that collected by the Dutch Water Utility Evides BV. The management of assets at this Utility is fairly advanced and analysis of this fault data enables Evides to expand the traditional reporting framework and consider the triple bottom line of social, environmental and financial performance (people, planet and profit). Their capital expenditure programmes are designed to minimise risk and achieve optimal management of their physical assets. The likelihood and consequence of failure of all system components is examined thus embracing the concepts of the Publicly Available Specification (PAS55) published by the British Standards Institution. The balance of the database structure is shown in the Appendix.
Of interest, the NRW in The Netherlands is the lowest in Europe and is approximately five percent.
110
Table 4.6 Fault Data used by Evides BV
Table 4.7 Fault Codes used by Evides BV FAULT DESCRIPTOR
Awareness time Location time Repair time
Duration to repair the fault Plumber assigned
Fittings used for fixing the fault Cost of repair
Estimated rate of leakage
Last fixer of the fault on the section
Date of last fault for the section of the pipeline Frequency of bursting
Nature of fault Mode of failure Size of the fault
Possible cause of the fault Point of fault
Function of the pipeline Depth of pipeline Material
Diameter
Construction year of the pipeline Construction date of the pipeline General condition of the pipe Average zone night pressure Average pressure
Recommendation General Remark
POINT OF FAULT TYPE OF BREAK MODE OF FAILURE
On the joint Burst pipe Longitudinal
Along the pipeline Broken pipe Circular (Circumferencial)
Valve point Leaking Valve Hole/ Pit
Underground leak Joint Leaking Meter Broken pipe Multi broken pipe Unclassified No water
Leaking pipe Low pressure Heavily leaking pipe
111
One cannot fault the exceptionally low levels of losses achieved by Evides BV. Comparing the data in Table 4.6 and Table 4.7 with that collected in the UK national faults database, there is some duplication and more importantly, a number of important options have been omitted. As the pipes age and the levels of failure rise in the Dutch system, it will be more difficult to analyse the meagre data collected to pin point causes of failures. This omission will cause their replacement programmes to be sub-optimal.
EWS Fault Database
A fault recording and job scheduling system called “Faultman” is used at EWS. It was written in-house in the early 1990’s and was designed to record and track the progress of reported faults. Faultman fulfils this primary task adequately, but the fault data in this system is extremely difficult to interrogate as much of the data is recorded in simple text fields with no control or consistency applied to the input.
Apart from the material and pipe diameter, the following data is currently recorded for all faults that are repaired by EWS.
Table 4.8 EWS Faultman Codes
The data collected by EWS is more comprehensive than the data collected by Evides BV but not as comprehensive as the in UK National Faults Database. It was found in this investigation that the database in which this data is stored is extremely difficult to access as several fields of free
ITEM TYPE TYPE OF BREAK CAUSE OF BREAK
None None None
Pipe Broken Back Soft AC
Coupling / Collar Split Damaged
Valve Hole Roots
Ferrule Leak Pressure
Meter Other Corrosion
Gasket Earth Movement
Tee Shallow
Bend Bedding
Meter Assembly Workmanship
Saddle Packing
Hydrant Unknown
Air Valve PRV Other
112
flowing text exist that contain critical data. This data is also not currently spatially linked so the engineer is unable to graphically visualise where the faults are taking place. This would be an invaluable enhancement. It is also imperative that standard drop down lists with programmatic logic checks on the input. Ideally, this information will be captured electronically by the artisans on site with a low cost ruggedized device such as the “Juno” manufactured by Trimble (Pty) Ltd, which has the added advantage of being able to capture photos and the coordinate of the repair. As this device is fitted with GSM this data can be transmitted back instantaneously to the server at head office.
It would be of further advantage to collect and analyse samples from the field to conclusively determine the cause of failure. These tests can also for instance predict the remaining useful life in the case of asbestos cement pipes.
Of greater concern however, is that no training or quality control is currently taking place to ensure that this data is reliably and accurately captured. As a consequence, the data recorded in this system is not only extremely difficult to interrogate but the credibility of much of the data is also questionable. This data is the most important data that is required in the mains replacement decision making process. It is essential that attention is given to these aspects so that in the future, optimised decisions can be made with confidence.