Learnings

Outcomes

The project considered the potential challenges of switching aged low voltage AC network apparatus to DC operation. A test programme was executed to assess the behaviour of the assets in a range of DC scenarios. This considered the application of DC load cycles, a DC accelerated aging regime and DC faults. Regular health assessments were conducted to monitor changes in the insulation under DC conditions. All network apparatus for test was carefully selected by the team at SP Energy Networks. The test programme assessed the impact of DC voltages on the most relevant cables, joints and linkboxes. Load cycles were applied to investigate performance under realistic load regimes. DC accelerated aging was performed on all of the samples to ascertain longer term performance. One group of samples experienced 40 years of accelerated aging and the remaining sample groups had 10 years of accelerated aging applied. Extrapolations of the sample health were made for the samples which experienced 10 years of aging. The remaining useful life of the conventional network apparatus, failure modes and the effect of failure were considered. All samples under test completed the programme but one partial failure was observed. No adverse behaviour was detected during the regular health assessments. Based on the results reported in this test programme the conversion of existing LVAC feeders to LVDC merits further investigation through follow up work streams. 

All samples were supplied by the team at SP Energy Networks. This ranged from service cables to mains cables with PILC and XLPE insulation. Linkboxes and cable joints were also considered. The PILC cable sample age ranged from 1952 to 1958 and the XLPE samples were mainly manufactured in 2019. Mechanical damage, caused during sample removal, was noted on the 2 way linkbox with 185 mm2 PILC mains cable tails. In general, the retrieval process for the PILC samples was a success as the majority of samples were in a good condition on arrival. This enabled an effective test campaign to be applied to service aged PILC samples. The lab testing of service aged apparatus has been demonstrated in this project and there may be other novel/emerging applications that benefit from a similar de-risking approach. 

The health assessment considered the measurement of; insulation resistance, conductor resistance, capacitance and dissipation factor. Polarization Index (PI) and Dielectric Absorption Ratio (DAR) were calculated using the time based insulation resistance measurements. The measurement of Partial Discharge under AC and DC conditions was also considered but significant background noise over the IEC 60270 [8] frequency range was apparent preventing a meaningful measurement campaign. A sample health assessment was completed at all key intervals (on arrival, post load cycle, following 10 years of accelerated aging and post fault) to determine whether the applied DC regime had a detrimental effect on the insulation. The most meaningful measurement was found to be the insulation resistance with threshold values could be defined from available literature. Further value was added by performing measurements over a 10 minute window to allow the calculation of PI and DAR. A recommendation for further health assessment tests would be to add visibility and automation by employing a modern Megger. Measurement systems such as the Megger MIT1025 have a wider measurement range (1 MΩ to 3 TΩ) and can provide preconfigured diagnostic tests (PI, DAR etc.). The adoption of such techniques for measurements on cable circuits would add a period of downtime during the switch over to DC but it is critical to understand the insulation condition for health assessment purposes and fault finding. Ultimately, the availability of such data would enable more effective run to failure trends to be developed for the different cable/joints/linkboxes. This would enhance asset management practices where faults are dealt with more proactively.

Analysis considered the expected operational lifetime when converting LVAC circuits to LVDC. Projections were made for the sample groups which experienced 10 years of accelerated aging. The health assessment data from the group which completed 40 years of aging was used as run to failure data. The 40 year degradation profile was used for projection purposes in the groups that experienced 10 years of accelerated aging. It may have been beneficial to have multiple groups that completed the 40 years of accelerated aging to enhance the robustness of the projection. During the 40 years of accelerated aging no prescribed health assessment thresholds were breached in the projections. A ratings matrix was compiled whereby the projected sample health response was ranked. Group 2 was the best performing group (300 mm2 CNE XLPE Waveform, 25 mm2 Single Phase Split Concentric XLPE SNE and the Glasgow city council 16 mm2 armoured cable) followed by group 9+ (25 mm2 Single Phase 2 Core PILC and 185 mm2 4 Core LV Mains Branch Joint with PILC Service) and group 8 (25 mm2 Split Concentric XLPE SNE and the Glasgow city council 16 mm2 armoured cable). Based on the projections no threshold values were breached and when no threshold was available no sustained adverse behaviour was observed. It is anticipated that the sample groupings would be able to complete the prescribed 40 years of service. The projections highlight the opportunity of operating these aged assets under 40 years of LVDC conditions.

The potential fit for purpose tests which could be performed before converting an existing LVAC feeder to LVDC were discussed. Tests considered included; baseline health measurements, a feeder layout assessment, insulation resistance, VLF/Tan delta, Conductor arrangement or integrity in cable runs and how to deal with dormant cable runs. Tests of this manner will enable initial assessment of potential circuits of interest and will avoid engineering time on unsuitable circuits. The fit for purpose testing will elongate the pre assessment time but will ultimately confirm suitability for switchover. 

Some suggested operational practices were discussed. It would be interesting to consider alternative operational regimes for the DC converter as ultimately, SP Energy Networks have full control of the DC link. These regimes could benefit the health assessment, maximise lifetime of assets or deal with short term extreme loading conditions. Sympathetic operation of the converter via soft start functionality could avoid rapid changes in the DC voltage. Establishing the earthing arrangement and defining which portions of the feeder are CNE/SNE will be critical to the overall operation of the LVDC feeder. Significant power transfer benefits exist for SNE cables over CNE and harnessing the opportunities will drive the ultimate benefits of transitioning the LVDC. The main failure in this programme was caused by mechanical damage during sample removal from the field. Therefore minimising the ground disturbance on service aged network apparatus must be a key objective. The study highlighted the potential pitfalls of disturbing assets which are over 60 years old but still have operational life left. 

The network apparatus revisions centre around the changes required at substations and customer premises to accept equipment for DC energisation. Standard approaches could be developed for installing the DC converter in the substation or where conventional equipment may be replaced. There may be a need to revise upstream portions in the case where a SNE cable is configured CNE, increased power transfer may be required as part of the justification in switching to LVDC. There will be tangible benefits in certain portions of the feeder to have 4 active conductors in the DC bipolar arrangement. Requirements around warning field staff or contractors that the feeder is operating LVDC. These operational changes need to be clearly highlighted and differentiated in a clear way to avoid safety issues. Based on the experiences in the test programme the options and requirements for potential health monitoring equipment was presented.

Lessons Learnt

From the laboratory testing discussed in the progress section, the following conclusions were made:

·         All samples completed the programme and only one partial failure was observed, due to mechanical damage rather than the DC test regime;

·         The testing confirmed that no adverse behaviour was evident whilst the LVAC cables were energised under representative DC load cycles;

·         No samples failed during the accelerated aging programme and there was no significant degradation in insulation performance on the range of samples;

·         All samples withstood the applied fault current and were deemed safe in the post fault health assessment. One sample group completed the prescribed 40 years of service and projections suggest that groups which experienced 10 years of aging would also complete the 40 years of service; and

·         The availability of cable/accessory health assessment data would enable more effective run to failure trends to be developed for the different cables/joints/linkboxes. 

Therefore DC energisation of existing DNO AC LV cable assets is viable and justifies further work within the area of LVDC to progress to BaU reinforcement option,