WETS’D15 WORKSHOP

World Energy Transmission System

"Renewal of cables distribution networks"

Was held within the scope of JICABLE'15 in Paris on June 25 rd 2015

Session 1 - General data of the network: Network topology

Length of overhead and underground lines (kilometres)
MV and LV networks, in urban zones and in rural areas

Network Topology
(WETS D'15 1.1 Tambrun - ERDF)

Presentation of ERDF networks (description of MV topologies, main figures about length, type of cable and accessories and their progression over the years). Since accessories are the weak point, we rather use 3 drums with 1-phases cable each that we assemble on site than drum of 3-phases to reduce the number of accessories. Concerning faults/km: 20% for cables mostly cables, and 80% for accessories, mostly joints.

Session 2 - Technologies

2.1 – Cables

What kind of cables for a long lasting life
Under-ground and submarines cables construction

Technology: Cables
(WETS D'15 2.1 Brument &Tourcher - EDF R&D)

Presentation of ERDF's cables and accessories. It also presents the faults statistics of the different MV accessories. The main parts are due to transition joints and normal joints (40% each). Transition joints are focused on very few cables whereas normal joints concern much more cable length. So failure/km is much worse with transition joints.

Aging of Cables and Diagnostics
(WETS D'15 2.2 Kim - Kepco)

Kepco has made a study based on a sample of XLPE cables (no accessories). They studied the evolution of Tan delta (TD) (trend and shape of the evolution: increase, decrease, linear,… ). They introduced a new parameter as diagnostic feature: the "skirt" (corresponding to the statistical departure of the TD measurements to a virtual line joining the min and max values of TD). This criteria revealed to be statistically relevant to estimate the remaining life of the cable. However no physical explanation has been found yet.

The chairman strongly invited Kepco to consider the possibility of attending one of the next IEEE ICC meetings (Insulated Conductors Committee) to present their research.

Evaluation of XLPE MV Lifetime
(WETS D'15 2.3 Dardel - Nexans)

Following bad results in cable reliability for the first generations of synthetic cables (PE based insulation) and technology improvements in cable design and manufacturing processes, Nexans has developed systematic tests (15kV/mm during 30 min, the voltage being adapted depending on the insulation thickness) on cables in order to estimate cable aging and to detect all cables with rapid breakdown (under factory tests) and to implement corrective actions to improve quality. Tests at higher voltages gave estimation on longer duration. They can estimate the lifetime for cables working at a service stress of 3kV/mm, after these corrective actions are implemented. For example according to their statistical model, they can predict that their failure rate at 100 years should be < 0,001 (For a DSO, this is the term that is relevant for the network).

They can also estimate the number of failure to be expected in the future based on the evolution of the results on existing former cables.
Such methods has not been applied to LV cables and their problems are more due to mechanical problems during installation or exploitation than electro-mechanical and such problem are not always critical for the network.
They use this method on standard XLPE and thus haven't made any comparison between NFC 223 and NFC 226 which is a French standard where other cables are used.
Several criteria are important for the robustness of the cable: cable in water, dry cross linking, compatibility between material... but none prevails over the other.

2.2 - Accessories, experience feedback

Banded technologies,
Heat-shrinkable technologies,
Cold retractable technologies,
Hybrid technologies, banded and shrinkable

Effect of Performing VLF Withstand Tests on Field Aged Degraded Joints
(WETS D'15 2.4 Drapeau – Hydro-Quebec)

In Hydro-Quebec's underground distribution system, joints are in manhole and thus accessible. Since the early 2000's, as part of their maintenance, they are performing measurement campaigns through thermal examination of the joints in search of heating points that reveal the weak points of the joints. About five years ago, they performed a study in order to investigate the effect of performing VLF withstand tests (according to the protocol proposed in IEEE 400.2) on field aged joints, according to their level of degradation. Results have shown that the most degraded joints experienced further degradation through the VLF withstands procedure and that this procedure appeared not well suited to be used in the Hydro-Quebec underground distribution system.
How much time between 2 condition assessment (before and after VLF withstand procedure) ? In the same week so that condition and parameters doesn't change too much in-between.

Such approach is not being used to assess condition of LV cables. They made a system to assess this problem with high pressure tests (they observe the evolution of the pressure in the tube), but the system turned out to be useful also as a commissioning tool because early experience with the system has shown, in some instances, insulation weaknesses which were revealed right at the installation stage.

Session 3 - Diagnosis of ageing and estimation of the residual life.

What diagnosis methods, on and off line mainly?
What feedback on these methods?
What progress can we expect in a near future?
Based on the principle of applying an electrical stress to an equipment, and then investigate the specific response.

Diagnostic methods can be separated into two families :

  • dielectric loss (no failure expected, but impedance measurements)
  • PD measurements (partial breakdowns of the global insulation)

For the PD measurements, there are 2 types : Off-line and On-line ( relates to the voltage source):

    • On-Line: PD measurements are performed while the circuit is energized under normal service conditions. The concept is to install any type of sensors and follow partial discharge activity over the time, so it is possible to detect trends.
    • Off-Line: take the circuit out and use an independent source that you can select through the following options: power frequency / VLF sinusoidal or cosine rectangular (typically 0.1 Hz) / Damped AC sinusoidal. In this case, you get rather a snapshot of the cable, but you could afford to solicit it with stress levels higher than service stress.

The question is then what is the best method to detect a given problem.

Diagnostics for Aging of Cable Systems and Residual Life Assessment
(WETS D'15 3.1 Espilit - EDF R&D)

Method based of PD and Tan δ measurements. PD is used to categorize the risks. First Tan δ and then PD (most of the time). Most of the time you can't rely on only one method, especially if you have no idea of the type of default you are looking for. Eventually, you can localize the problem using breakdown. (Explanation to be found in JICABLE 2007 about the limit for Tan δ method).

How to estimate the voltage to use do detect a PD for long distance lines? The signal is lower but the energy is mainly conserved. Another solution is to place several sensors and synchronize them.

ERDF don't accept PD in polymeric cable.

You can use standards for threshold, but they apply to specific systems. It is better to make your own criteria. ERDF has established its own criteria based on the characteristics of its network and their experience.

PD Diagnostics on MV cables
(WETS D'15 3.2 Hummel – Omicron)

Present several PD testing methods available at different frequencies, the type of signal available and the places you can use them. VLF is for Off-line test, mostly on site and other closer to power frequency for factory are on site test.

Aged PD XLPE reliability program
(WETS D'15 3.3 Ben Lanz –Imcorp)

The contribution shows the summary of off-line PD measurements @ power frequency performed by Imcorp over time, cumulating > 16000 cable assessments (covering a cable population size of 1600 km). The outcome is: if you can identify the limited weak points of your network, you can dramatically improve the quality of the network at minimal cost.

Diagnostic Selection for Condition Assessment of Long MV Underground Feeder Circuits
(WETS D'15 3.4 Jean Francois Drapeau – Hydro-Quebec)

This document presents a description of the ongoing R&D project carried-out at IREQ (Hydro-Quebec Research Center) regarding the various cable diagnostic techniques. The project consists in addressing 2 technical challenges: propose the most relevant and accurate selection of diagnostic techniques and protocols to:
a) Detect and pinpoint "suspicious components" to be replaced.
b) Confirm the "back to normal" condition of the "refurbished" circuits, once the remediation actions (targeted component replacements) have been done.

The document shows a preliminary version of a portfolio of diagnostics devised for addressing these tasks. One of their points is to use multiple diagnostic methods in order to make sure to detect all the various insulation defects present on the cable sections targeted for refurbishment.

Session 4 - Renewal of the distribution networks.

Methodology to optimize technically and economically the network renewal
Experience feedback and now what best strategy to replace accessories and cables
Which progress and how to obtain them, Easy assembly, precise instructions, simplicity and optimized design, How to improve the qualification of the fitters?

Renewal of the Distribution Network
(WETS D'15 4.1 Resmond - EDF R&D)

Rather than replacing "randomly" cables, ERDF targets the line to replace. First using statistical criteria and aging simulation to determine when cables have to be replaced. It then uses diagnostics to refine this plan, for example using diagnostic trucks on long risky connection. Research has been done to improve the estimation of aging of the network depending on the type of cable. For each type of cable : what is its weak point, aging model… They did so with sample from the field.