A1: Conductor breakage by over-twisting during installation:
During an early ACFR installation, back-to-back pulling of the conductor for more than 10km without torsion releasing technique through poorly maintained running sheaves caused breakages of ACFR conductors by over-twisting at multiple locations. Causes were as follows: excessive pulling length, improper use of running blocks, lack of torsion releasing means and inconsistent back-tensioning operation caused excessive accumulation of torsional stress (over-twisting) within a relatively short length of the conductor. The length of the back-to-back pulling section was more than 10km which caused higher pulling tension to maintain clearance (and more torsional stress in the conductor as it tends to unwind itself when pulled with higher tension). Some of the running blocks used in the section were poorly maintained (some were not even in rotating condition with rusty bearings and the conductor was being dragged over the grooves during pulling). Also, the position of the running blocks at angles were not adjusted in a manner that the conductor runs in the center of the groove which caused even more torsional stress on the conductor as it was rotated by the friction with the wall of the groove when pulled through angles. Torsion releasing swivel between the drums and counterweight at the head of each drum were not used as the connection between drums were done by permanent conductor splices installed at the tension site. All these factors combined, torsional stresses kept accumulating on the conductor during the pulling operation which eventually caused excessive twisting of the ACFR conductor within a relatively short length, when the pulling tension was increased and dropped dramatically. All the breaks were in the same section installed by same installer.
Countermeasure:
TRI’s installation support personnel provided feedback and recommended corrective actions. All the tools and equipment were inspected and any portion of the conductor with aluminium distortion (evidence of over-twisting) were cut and replaced with new conductor. The installation specialist provided scope-specific on-the-job training and detailed recommendation for: selection of sheaves, torsion releasing technique as well as dividing the pulling sections into < 20 spans. All the poorly maintained sheaves were removed from the sites and replaced with properly working sheaves and position of sheaves at angles were readjusted, torsion releasing swivels and anti-twist counterweights were used in between drums, and conductor splices were installed after the conductors were pulled into place. TRI’s installation experts remained on site until it was assured that all members of the installation team became aware of the specific cautions and limitations of the conductor for each scope of work. The rest of the installation was resumed without any issue. The line has been energized and operating without any problem after installation to this date. Scope specific preventive cautions were added in our ACFR installation guideline.
A2: Conductor breakage by over-twisting after installation:
Conductor breakage occurred at one location 3 years after installation. After investigation of the broken portion of the conductor, it was concluded that the CFCC at the location of the breakage had already been sheared by over-twisting of the conductor during the installation so the core was not fully functioning as tension bearing member. The tension on the conductor (approximately 2 tons) was taken by the outer aluminum strands (almost 90% UTS of the aluminum strands combined) which eventually led to the breakage by creep rapture of aluminum strands after 3 years. The causes of damage were (similar to aforementioned case): excessive pulling length, improper use of running blocks, lack of torsion releasing means and inconsistent back-tensioning operation which caused excessive accumulation of torsional stress (over-twisting in loosening direction of outermost layer). Although the sign of distortion of aluminium was evident, the twisted portion of the conductor was repaired in a manner to restore the conductivity of the aluminium strands but not the tensile strength of the core (replacing the distorted portion with new conductor or full tension splice is recommended). The line was installed by the same crew that experienced conductor breakage during the installation mentioned above.
Countermeasure:
TRI provided additional training for the crew members to reinforce the best installation practices. The very same crew successfully completed another reconductoring project without any issues during and after the installation and no issues or problems have been reported on the line. Any portion of conductor with sign of overtwisting (birdcage and/or squeezing) were identified and replaced prior to energizing.
A3: Failure of incorrectly pressed dead-ends during and after installation:
Conductor slipped out from dead-end clamp during and after installation. Two different dead-end fittings (which require different die sizes) were used for the project (due to some delays in delivery of initially scheduled fittings). Slightly larger sized die was specified for the initial fittings (across-flat-width after compression was 18.90-19.20mm for the initial fittings and 18.75-19.05mm for the newly added fittings). After investigating the failed fitting, it was found that some of the inner steel sleeve of newly added fittings were compressed with wrong dies (which was prepared for initial fittings), resulting in insufficient compression (gripping of the composite core) of some of the steel sleeve (which in general provides 80 % of the RTS of the dead-end (the rest, 20% comes from the aluminium sleeve). It was then communicated to the installation crews that it must be verified that across-flat width of steel sleeves after compression must match with the manufacturer’s specification. However, some completed dead-ends were overlooked and not corrected prior to completion. Main causes of this issues were lack of communication and installation quality control failure.
Countermeasure:
Double-checking across-flat-width after compression was included in the quality control procedures/method statements and dies and press, as well as fittings, were marked with matching identifications to avoid unnecessary confusion. It was instructed that installation tools and equipment, procedures and QC parameters to be reviewed and verified at daily toolbox meetings. The rest of the dead-ends, where it could not be assured that the inner sleeve was compressed with correctly sized die, were replaced with properly installed dead-ends. No failure of conductor nor fitting have been reported on the line after the corrective actions have been taken.