Fault management

This paper sets out the fault management processes carried out by DNO’s on the 11 & 33 kV system. Included in the paper are the regulatory framework which Distribution network operators (DNOs) operate within including both Quality of Supply and Customer Service standards. The paper discusses the components that make up an Electricity Network and the influencing factors that can result in faults on the Network. The lifecycle of fault management from response, restoration, location and through to repair are discussed including the future technological advances enhancing the Fault Management process now and in the future.

Go to the profile of Grant McBeath
Aug 31, 2017
1
0
Upvote 1 Comment

Author(s): Grant McBeath

Distribution network operator fault management

What they do

DNOs are regulated businesses that distribute and connect electricity to homes and businesses over their network.

DNOs take electricity generated from distribution-embedded generation and from the transmission network and supply customers at the appropriate voltage. DNOs in Scotland manage the networks operating at 33 kV and below, whereas in England and Wales they operate at 132 kV and below. Connections are competitive and the DNOs, along with licenced independent Distribution Network Operators (IDNOs), compete to provide new connections for customers.

A typical DNO Distribution Network operating at voltages of 33 kV and below has 43,000 substations, 21,000 km of overhead lines and 41,000 km of underground cables and covers a substantial geographic area. This paper will discuss the fault management of the DNO network operating at 33 and 11 kV including both the external Office of Gas and Electricity Markets (OFGEM) regulatory requirements and the internal fault management processes.

The GSs associated with unplanned fault incidents are as follows:

Regulatory standard

Criteria

Time period

Compensation

Regulation 5

supply restoration during normal weather

interruption > 12 h

£75 domestic  £150 business  £35 each additional 12 h

Regulation 6

supply restoration during normal weather – incidents affecting 5000 customers or more

interruption > 24 h

£75 domestic  £150 business  £35 each additional 12 h

Regulation 7

supply restoration during severe weather – category 1

> 8× normal HV weather-related faults > 24 h

£70 domestic  £70 business  £70 each additional 12 h

supply restoration during severe weather – category 2

> 13× normal HV weather-related faults > 48 h

£70 domestic  £70 business  £70 each additional 12 h

supply restoration during severe weather – category 3

large event and formula-based approach

£70 domestic  £70 business  £70 each additional 12 h

Regulation 10

multiple interruptions

customer > interruptions > 3 h over 12 month period

£75

Regulation 11

distributor's fuse

attendance within 3 h weekdays 07:00–19:00  4 h weekends 09:00–17:00

£30

What they do not do

The DNOs do not generate or sell electricity. They transport it from the time it enters the DNO network to individual DNO customers. A proportion of any electricity bill accounts for the distribution of electricity. On the basis of an average domestic usage of 3300 kWh, every consumer will pay ∼£100 pa or 8% of their bill for the services of the DNO, which is collected by their energy supplier and passed onto the DNO. This amount will vary by DNO and is an agreed revenue return, based on charges approved by the regulator OFGEM.

Regulatory standards

Guaranteed standards

DNOs have a licence obligation to deliver certain standards and quality of service. In accordance with the electricity (standards of performance) Regulations 2015, DNOs are required to comply with specific guaranteed standards (GSs) associated with unplanned or fault interruptions. When DNOs fail to meet the standards set out by the regulator, consumers are entitled to financial compensation.

Broader measure of customer service

OFGEM has also created a mechanism that ensures companies deliver good customer service whilst carrying out their licence obligations.

The broader measure of customer service has three components:

  1. Customer satisfaction survey.
  2. Complaints metric.
  3. Stakeholder engagement incentive.

With regard to fault events, the customer satisfaction survey is the key component that determines the quality of customer service delivered by the DNO against a specific fault incident. Within the customer satisfaction survey there is a separate ‘Interruption’ category which looks at both planned and unplanned interruptions. The ‘Interruption’ category that measures the customer experience for unplanned faults is sub-divided into ‘Unplanned Message’ and ‘Unplanned Agent’:

  • ‘Unplanned Message’ is where the customer has contacted the DNO and heard a recorded interactive voice response message.
  • ‘Unplanned Agent is where the customer contacted the DNO and spoke with a call agent in regard to their interruption.

Following an unplanned fault, the DNOs submit data to OFGEM's independent customer survey organisation. This company contacts the relevant customers and ask a series of questions associated with the customer's experience during the interruption. The customer scores the DNO out of ten on ease of contact, politeness, accuracy of information and usefulness of information. The customer survey company then asks the killer question, ‘Overall, how satisfied are you with the customer service?’, and it is this rating that determines the score and financial reward/penalty position. All DNOs are benchmarked against each other in an effort to ensure high standards across the industry through comparison and competition.

Interruption incentive system

The regulator agrees with each DNO an interruption incentive scheme (IIS) annually over the price review period which rewards or penalises DNOs depending on their interruption performance against their targets. The DNO is then penalised whenever a customer is off supply for more than 3 min, which is referred to as a customer interruption (CI). The financial penalty varies for each DNO; however, typically this can be £8 for every customer interrupted for longer than 3 min. The ‘customer minutes lost’ element of the IIS arrangement aggregates the number of customers each minute that are off supply. Typically, the financial penalty can be 22p/customer/minute during the interruption period.

The financial penalties during fault outages can be significant and understandably this has, where possible, driven DNOs to avoid interruptions and restore customers quicker when a fault occurs. Initiatives to avoid interruptions can be targeted asset investment where apparatus has reached the end of its life; therefore, removing plant before failure occurs. Reducing interruption times via the introduction of new technology that improves fault discrimination, fault sectioning and fault re-closure have been instrumental to improving the DNO's interruption performance.

Influencing factors on fault incidence

Asset performance

The Distribution Network has multiple assets such as cables, overhead lines, transformers, switchgear etc. Every asset has a different level of reliability and this may vary with manufacture, age, service history etc. DNOs have to understand the condition of their network and assets to allow the correct allocation of investment to replace and undertake routine maintenance. Under the regulatory price review, the DNO has to justify what investment in their assets is required and what benefit this will have to the consumer.

Underground cables are inherently very reliable if installed correctly and not subjected to overloads. Generally, the investment strategy for cables is to replace when there is a fault, unless there is a known reliability issue or safety concern associated with the asset. Underground cables can be subject to third-party damage and it is essential that parties working in the vicinity of cables deploy safe digging practises including detailed cable records. The main cause of cable faults are joints or wear and tear of an ageing asset. There is very little opportunity for non-invasive inspection or maintenance of these assets.

Overhead lines are a significant asset type for most DNOs and generally perform well; however, they can be affected by severe weather and other environmental factors. Weather is the main cause of network faults, though trees and third-party interference are also common factors. Overhead lines can be easily inspected and maintained and are replaced following a condition-based assessment.

Weather

Weather is the single most influential factor that causes faults on the DNO network. Different weather parameters such as wind, temperature, snow and rainfall all have the potential to cause faults to different types of assets. Regular detailed weather forecasts, example in table 1 below, are provided to DNOs along with weather warnings from the Met Office and other agencies.

In managing networks and preparing resources in advance of major weather incidents, accurate local forecasts are central to ensuring a fast response when the adverse weather affects the network. Manpower will be put on standby, vehicles prepared and cooperative agreements for obtaining manpower from other DNOs less affected will be triggered.

Flooding is an ever-increasing issue for DNOs, as the UK's weather patterns continue to deliver wetter, milder conditions, especially in winter. DNOs have recently invested in additional flood-resilience schemes to protect key strategic sites that supply significant customer groups. DNOs have the ability to deploy temporary barriers or high-volume pumps; however, if this is insufficient then de-energisation of the site maybe the only option. De-energisation of sites will likely interrupt supplies and can hinder other attempts to alleviate flood damage by pumping. Much of society's infrastructure is electricity-dependent, ranging from house digital phones through to cash machines and retail outlets. If flood defences fail, inspection of affected properties maybe required to ensure that re-energisation can be carried out safely, and this may prolong an outage. The Royal Society of Engineering has produced a report setting out the impact following extensive loss of supply due to flooding, and the impact on that locale and its consequences are set out: http://www.raeng.org.uk/publications/reports/living-without-electricity.

High winds are the most common environmental factor that can affect the performance of the overhead line network. The DNO network is generally resilient up to wind speeds of around 60 mph in the prevailing direction, which is South Westerly in the UK. Wind speeds >60 mph can result in conductor clashing and also carry the risk of broken conductors and poles. Damage to overhead lines due to debris or falling trees is common and reinforces the need for regular tree cutting to minimise these occurrences. The direction of the wind and time of year is also a factor, with the network more resilient to the prevailing wind direction and the trees in full leaf at greater risk during high winds.

Lightning is another environmental factor that can cause faults on the DNO overhead network. The overhead network is designed to withstand lightning strikes through the installation of surge arrestors and, since most lightning-induced faults are transient, auto-reclosers re-energise the circuit following the strike. However, damage to pole transformers and insulators is common during a lighting event. The DNOs have lightning monitors that detect cloud to ground strikes. This, as well as enhancing fault-location, allows field teams to be notified of the lighting risk for safety purposes.

Temperature, both high and low, can be a factor that causes faults on the DNO network. Low temperatures can cause gas pressures to fall on circuit breakers, rendering them inoperative. Similarly, low temperatures can cause transformer oil to contract and activate Bucholz protection on 33/11 kV transformers. High temperatures can also cause issues with increased sag on overhead lines and the bitumen used in older generation equipment can melt; therefore, degrading the insulation quality.

More recently, winters tend to be wet and mild; however, snow can be damaging to the overhead line network if combined with cold winds, increasing the risk of ice build-up, particularly if the temperature is close to freezing, which allows water to turn to ice on the lines. The build-up of ice on overhead lines can be catastrophic, with broken conductors and poles commonly resulting in extensive repair times. Gaining access to locate and repair faults can be challenging with access road being impaired. Helicopters are used to patrol lines and identify fault locations and can be used to deliver materials to site where access is an issue.

Third-party interference

DNOs experience third-party interference both accidently and maliciously. Accidental interference is dangerous to the parties involved and affects the surrounding customers. Accidental damage to cables is not uncommon and reinforces the need for DNOs to keep accurate cable records to enable them to minimise the risk for third parties working in their vicinity by preventing physical interference during excavation.

Accidental contact with overhead lines can be more prevalent when the agricultural community is more active in the countryside such as during crop harvesting. Carbon fishing rods are a hazard to life. Wildlife can also be a factor, especially birds, and often overhead lines near ponds and breeding areas have bird deflectors fitted to reduce the risk of inadvertent contact. Extensive use of ‘Danger’ notes helps to minimise this risk, and public awareness campaigns both protect the public and reduce the risk of network faults affecting customers.

Malicious interference or metal theft is experienced by all DNOs and legislation associated with the reclamation of scrap metal in England and Wales has acted as a deterrent. Instances of stolen copper earthing, cables and overhead lines have all been experienced, incurring CIs and significant fault repair costs on DNOs.

Response, location, repair and restoration

Response

As all faults are unplanned DNOs cannot predict when they will occur, and therefore are reactive in nature! DNOs can become aware of a fault on their network in two ways: first via a SCADA event or plant status change and second via a telephone call from a customer or third party. The Distribution Network is designed with protective devices ranging from fuses to complex electronic devices that identify the presence of a fault and discriminate and/or grade with other protection devices to minimise the fault zone and hence customers affected. Many faults on the overhead network are transient in nature, allowing the line to be re-energised after the fault arc has extinguished. Auto-reclosing devices are fitted, resulting in only short interruptions and no manual intervention.

DNO's have invested significantly over the past 20 years to increase the volume of telemetry at primary and secondary Sub-Station (S/S), allowing the central DNO Supervisory Control and Data Aquisition (SCADA) system to monitor and control devices on the network. Fault-breaking devices in primary S/S, secondary S/S and overhead line auto re-closures were initially targeted, as this would increase the DNO's awareness of a fault and allow remote fault switching. Further telemetry has been installed at normally open points and midway on a circuit to improve the DNO's performance under fault sectioning scenarios.

DNO awareness of a network fault could be from the customer who is off supply or from third parties where the faulted circuit is protected by a non-telemetered device or no plant has operated. DNOs have centralised call centres to manage faults and emergencies and, depending on the number and location of calls, can use these to determine the probable fault category. If a fault only affects one property, then it is likely single premise or service-type issue. If adjacent properties are affected, then it is likely Low Voltage (LV) in nature. If more than one S/S is impacted and calls are received over a wide area, the SCADA system predicts the likely non-telemetered device that has operated. If a high voltage (HV) telemetered device operates, then the SCADA system will immediately record this event and inform the control engineer via an alarm, plant status change and altering the network diagram to indicate a dead section.

DNOs have fault reclosing policies that ensure necessary safety risk assessments are considered prior to any re-energisation. The risk assessment will consider any critical information received, location, network type, time of day, alarms etc. To locate the faulty section, it will often require fault re-closures to positively confirm the location of the fault for underground networks, as the fault location likely will not be visible. On overhead networks fault sectioning will also occur with the aid of temporary fault indicators that detect the passage of fault current to assist in identifying the faulty equipment. To minimise the effect on the surrounding network and sensitive customers, protection settings can be altered to speedup the fault clearance time or the network altered to reduce the fault level during re-closure.

Location

On high-voltage DNs, cable faults are the greatest risk to security of supply. Most public DNs are structured in a ‘ring’ configuration where the network can remain live with a single cable fault but with reduced security. Therefore, the process of repairing the faults usually has to be a swift one. The most critical part of the process is locating the faulted piece of cable. Distribution cables (11 and 33 kV) can range from a few metres to several kilometres, and therefore the fault-location process has to be very accurate to avoid unnecessary excavation costs.

To achieve an efficient fault location, DNOs have over the past 50 years used various methods and technology, ranging from loop techniques (Murray/Varley) with Bridge Meggers to voltage and current impulse techniques with time-domain reflectometers (TDRs). Both aforementioned methods are ‘pre-location’ practises that locate the fault to within a few metres and the final location (pin-pointing) is done with an acoustic process of putting a high-voltage surge down the cable that discharges at the point of fault which can be heard. The loop method and TDR are shown below.

The current techniques all employ the TDR, which in simple terms is cable radar: a small pulse is sent down the cable every few milliseconds and any reflection from an impedance change is shown on a screen. The difficulty with this process is that very few HV cable faults are a low enough impedance to show up on a TDR. To assist with this process, we have to change the properties of the fault either permanently or just long enough for the TDR to capture a picture.

The most successful technique currently used is arc reflection method (ARM), which sends a significant charge (c. 1750–3500 J) via a surge-wave generator into the cable at a voltage high enough to cause a breakdown. When the breakdown occurs, the resultant arc is stabilised for around 20–30 ms with a further surge generator; the stabilised arc is usually zero impedance and the TDR can capture several pictures of this breakdown. On the basis of the information seen on the TDR screen and available cable records, we can calculate a rough estimate of the fault location. An example of this fault is represented and can be seen in the traces below. The cursor on the TDR shows the estimated location.

Despite ARM being the most successful method, we occasionally need to use others, the majority of which are known as transient methods utilising voltage or current. The most utilised of these is impulse current equipment. Two differing methods are encompassed here: a single-phase option or a three-phase one; the single phase is purely a high charge discharged into the cable with the TDR measuring the ‘echo’ of the breakdown on the cable. The three-phase option is used very successfully to locate very high resistance faults, particularly on 33 kV cables where the protection systems are extremely quick and the cable damage from the system energy is low. This method uses the TDR to identify the breakdown and then uses a healthy core on the cable as a reference that allows a measurement to be taken between the two to calculate an accurate distance to the fault.

Once a location has been obtained, this distance is mapped on the cable records, which gives the engineer an area to visit and listen for an acoustic signal to pinpoint the exact location of the fault. To achieve this, we again employ the ‘surge-wave generator’ to apply a suitable charge to the cable, again typically 1750–3500 J that will discharge at the point of fault; this discharge can be heard or even felt as a shock wave on the ground. To assist on this process, the engineer utilises a listening device that can detect very small acoustic and magnetic pulses – this aids in the search for the exact point of fault. When this location is obtained, the ground above is usually marked with paint and coordinates taken to be passed onto the excavation team to open the ground around the cable in preparation for the repair process to begin.

Repair and restoration

The repair phase can vary depending on the type of fault and the asset to be repaired. Commonly, overhead line faults are repaired immediately, as often customers cannot be restored until the repair has been completed. Cable networks that operate in a ring configuration can be repaired over a number of days, given that the customers have been restored via alternate feeding arrangements. The timescales to repair the fault and secure the network will depend on the network risk and sensitivity. Typically for 11 kV faults they would be located, repaired and returned to service within 3 days. For 33 kV faults, the timescales are 5 days, given the larger excavations and more complex jointing techniques.

During the cable fault repair phase safety, documents will be issued to positively identify, spike and obtain phase colours. On completion of this phase, a further safety document will be issued to complete the jointing. On completion of the cable joint, the cable will be subject to continuity, insulation and phase checks to ensure the integrity of the cable and newly installed joint.

Technology

The regulator has driven all DNOs to innovate and modernise their network to improve the security of supply and reduce the impact of faults on the network. Initiatives such as deploying telemetry at all primary S/S and strategic secondary S/S locations have significantly improved the DNOs’ fault response capabilities.

Fault recorders

New technology is used in fault recording devices located in primary S/S which can measure the fault impedance, and therefore map potential locations for the DNO to focus on their investigations. Impedance mapping is approximate enough to identify the faulty section, thus reducing fault re-closures and allowing customers’ supply to be restored more quickly. The fault recorders can also measure other electrical parameters such as voltage, current and fault clearance times of circuit breakers to ensure all are within operational limits.

SCADA automation schemes

Modern SCADA systems also have automation applications that can create a logic-driven decision-making tool based on plant indications and alarms received under fault situations. Exploiting the increased telemetry deployed in the Distribution Network and bespoke SCADA applications results in customers being restored safely and quickly. These applications deliver tangible benefits to both customer service and the DNO IIS targets during fault outages.

SCADA mobile solutions

Modern SCADA systems have mobile applications which allow the centralised control centres to communicate with field staff electronically via hand-held devices. This delivers efficiencies in reduced call volumes and allows the field staff to have accurate network and customer information where previously they relied on paper diagrams and verbal updates from control centres. Field staff also can update fault information in real time, allowing this to be shared with customers regarding the cause of the fault, fault progress and estimated restoration times.

Smart metering

Smart meters will be rolled out across the UK over the coming years will provide the DNOs with a number of opportunities to improve their fault response. Smart meters will be capable of communicating with the relevant DNO through a data aggregator to provide basic electrical parameters. Additionally, smart meters will have functionality that sends a ‘last gasp’ when supply is lost and ‘first breath’ when supply is restored. This will be an advance as for single premise and LV incidents, as the DNO will be aware of exactly which customers are on and off supply. This will allow the DNOs to despatch the correct resource type with the necessary skills and in certain circumstances give a better location of the fault. For HV faults in rural areas where protective devices operating in fault situations are non-telemetered, then again this will alert the DNO of a fault situation without the need for a customer contact.

Fig 1: Fault location loop method and TDR

Fig 2: TDR fault trace

Fig 3: Fault distance trace

Fig 4: Acoustic listening device and test van equipment

Fig 5: Example of faulty underground cables and overhead lines

Fig 6: Typical distribution automation architecture

Fig 7: Basic SCADA mobile functionality

Risk of disruption all hazards

GREEN

GREEN

GREEN

GREEN

GREEN

GREEN

GREEN

overall critical period all hazards

N/A

N/A

N/A

21-06

N/A

N/A

18-06

Temperature max/min

12/4

12/4

12/4

10/3

10/3

9/3

9/2

Wind chill temp Av. max/min

10/0

10/0

11/0

7/0

8/−1

7/−1

7/−2

Wind gusts over 60 mph (timing)

10–20%

10–20%

10–20%

10–20%

10–20%

10–20%

30%

Extremes + direction, mph

wind

West 25–30

West 25–30

West 25–30

West 25–30

West 25–30

West 25–30

West 30–35

gust

gust: 40–5

gust: 40–45

gust: 40–45

gust: 40–45

gust: 40–45

gust: 40–45

gust: 45–50

Snow

potential

<10%

<10%

<10%

<10%

<10%

<10%

<10%

amount/intensity/height

Line icing

potential

<10%

<10%

<10%

<10%

<10%

<10%

<10%

moderate/severe

Flood risk > 30 mm in 24 h

<10%

<10%

<10%

50%

20%

50%

50%

Lightning risk

0600–1200

3

3

3

3

3

3

3

1200–1800

3

3

3

3

3

3

3

1800–0000

3

3

3

2C

3

3

2C

0000–0600

3

3

3

2C

3

3

2C

 Table 1: Typical DNO weather report

Go to the profile of Grant McBeath

Grant McBeath

Control room manager, Scottish Power

No comments yet.