Power outages can be frustrating, inconvenient, and even dangerous in certain situations. That’s why Northern Powergrid has been working hard to develop and test low-cost pre-fault detection techniques that can locate developing faults in LV circuits before a power outage occurs. Working with VisNet’s parent company, EA Technology, the project has successfully tested this technology on various types of LV networks and cable types. With the valuable data that has been collected, the end goal is to create a process that can predict when faults will occur and prioritise how to deal with them proactively. This exciting development has the potential to improve the reliability and safety of electrical networks, ultimately benefiting customers and communities across the board.
Foresight – LV Network Visibility and Fault Recognition
The Foresight project outlines a new approach to LV fault management using innovative, ground-breaking low-cost technology that can not only detect gestating faults but can successfully locate them before a loss of supply occurs. It will also provide an indication as to whether the fault is imminent or not so that operational resources can be targeted in the most effective way. This type of approach will require a whole new paradigm shift in operational culture within electricity distribution companies to be effective and avoiding these faults would deliver significant quality of supply benefits.
Electricity Regulators continue to focus on increased availability and reliability of networks. Many Electricity Network Operators have improved the performance of High Voltage networks and consequently a significant proportion of customer interruptions and customer minutes lost now result from Low Voltage (LV) faults. In addition, Low Carbon Technologies, such as electric vehicles, heat pumps and distributed generation are increasingly being connected to the LV network, changing power flows, voltage profiles and increasing stresses on the network.
Due to an ageing network, much of which is over 50 years old, LV fault management is becoming increasingly difficult. Restoration times can be lengthy, as most of the LV network is neither comprehensively monitored nor controlled automatically. Responses to faults tend to be reactive as the condition of LV cable systems at any point in time is unknown and there is no capability of predicting the timing and location of faults before the event occurs.
The Foresight project involves using a range of LV network monitoring equipment in combination with automated switching devices. The aim of the project is to recognise LV pre-fault signals and then use low-cost monitoring devices to identify and locate developing faults before supply outages occur. This enables pro-active deployment of reclosing devices and pro-active cable repair before the fault. Planned interventions rather than post-fault repair on cables produces both financial and customer service benefits.
Traditionally, LV cable replacement has been driven by the reliability history of cable sections, however, the Foresight project has provided the opportunity to radically change this strategy to incorporate condition-based information to guide replacement decisions. This is especially important as the age of the LV cable asset base advances towards the end of its design life where there is a need for cost-effective asset replacement.
Northern Powergrid has worked with EA Technology to develop and test novel, low-cost pre-fault detection and location techniques. These techniques have been proven to detect and locate developing faults in LV circuits before a loss of supply event. The project has utilised this technology on different types of LV network and cable types and the data collected is being used to develop a process that aims to predict when faults will occur and provide a level of prioritisation in terms of proactively dealing with faults before a loss of supply occurs.
Back in 1974 when EA Technology was the Electricity Distribution Research Centre, some work was started on the recognition of “transitory faults” on the Low Voltage network. One of the collaborating electricity companies was Yorkshire Electricity. These events were observed to occur as dips or notches in the voltage waveforms measured by, at the time, expensive equipment at substations. They were found to be due to degradation of the cable insulation at a particular location which, following a short passage of current, self-healed before fuses blew and therefore before customers lost supplies. In 2017 the same companies, now called EA Technology and Northern Powergrid, collaborated again to further research these phenomena.
At the start of the Foresight Project in 2017, the project team wondered whether these events (now named pre-fault events) could be used to enable a brand new way of managing LV underground cable faults. At the end of the project in 2021, we discovered that they could. The Foresight Final Report charts the journey between those two dates and presents the equipment, systems and methods used to move away from reactive fault management to management based on proactive action.
This comes at a critical time in the UK and elsewhere in the world. While electricity networks are being looked at as a critical ally in the transition to zero carbon and move away from fossil fuels, they are also ageing, most being over 50 years old. Increasing the probability of faults at the same time as increased reliance upon it for the connected customers is an unwelcome combination and calls for a step change in the way network performance is managed.
The Foresight Project has demonstrated through trial that the equipment, system and methods proposed will enable a proportion of developing faults to be detected, located and proactively mitigated in the early of development, drastically reducing unplanned off-supplies in comparison with present methods. The presented proactive method of managing LV faults uses new low-cost/ quick-deploy monitoring equipment (Guards) designed for widespread deployment. Signatures from the data from that equipment create a measured cable condition data set. This data set has been used for two purposes. In the operational timeframe (months, weeks) it is used to recognise circuits that are accelerating toward failure. These are prioritised for proactive intervention to mitigate the impact of the developing fault.The intervention trialled involved targeted deployment of fault management equipment (Reclosers) and new techniques to pinpoint the location of the developing fault with the ultimate goal of being able to find and fix these faults whilst they are still in the early stages of development and before they fail. In network investment timescales (years), the same cable condition data set has also been used to provide previously unavailable circuit condition data. This has been integrated into a LV Cable Asset Management Model (EA Technology’s CNAIM model) capable of providing strategic insight into movements on the health indices of this significant asset group.
In a demonstration of the proactive management methodology described in the report, a developing fault was located, customers informed and was repaired as planned work. Not only was the (obviously) damaged component removed from the system whilst it was (surprisingly) still capable of providing supplies (see figure below), a seemingly healthy adjacent joint of similar type and age also removed and sent for forensic examination. Although visually healthy, the forensic investigation showed indications of accelerated aging judged to be consistent with an expectation of failure within 5 years. LV Cable networks, to which the majority of customers are connected and are becoming increasingly reliant, are not getting any younger.
Frequently Asked Questions
In the method used to locate pre-fault events, do you need to have a sensor at the remote end of the network (i.e. sensors either side of the fault position)?
No you don’t need sensors/measurements either side. All the measurements are taken from the substation. In any case, most UK LV underground networks are branched (i.e., they have several end points) and when location equipment is being deployed, it wouldn’t normally be known which branch the developing fault was on. The substation measurements we make seek to address this somewhat. We capture information that leads to us establishing an impedance to fault and a distance to fault. On branched networks with different cabling types in various sections, these can act as ‘co-ordinates’ that can often prioritise one branch over another as the candidate location
How many substations did the project monitor?
In order to get a better understanding of the activity levels per circuit, we made the decision to deploy on a much larger scale. With over 1800 measuring devices and nearly 1000 different substations, we had the opportunity to gain valuable experience in equipment installation across various network types. Not only did this approach give us the insight we needed, but it also allowed us to develop our skills and knowledge in a practical setting. It’s important to explore different avenues when faced with a challenge, and this widescale deployment certainly proved to be an effective solution for our project.
Did you manage to measure the efficiency of using the method, hours to repair etc?
No, this wasn’t an explicit part of the Foresight Project. In the Project, the proactive fault management method and supporting equipment was developed, trialled and refined. One of the next necessary steps will be to observe the efficiency savings that we expect will accrue in ‘business as usual’ use of the method. Under existing fault management practices, fault location and repair is left until the fault goes permanent at which point traditional location methods can be used. A permanent fault is one in which the circuit can no longer be re-energised by replacing a fuse or resetting/reclosing a different circuit protection device. Faults can go permanent at any time of the day or night on any day of the year, irrespective of whether there are enough staff available to reactively respond. Using the proactive method, even if the fault is left to go permanent, there is an advantage in having already identified the likely location. Efficiencies in this area would be measured in terms of hours per fault, not days. However, there are thousands of LV cable faults to which these efficiencies could be applied per year.
Is it planned to deploy this method more widely now it has been shown to work?
Yes it is. Northern Powergrid has been instrumental in developing an innovative fault management methodology that is set to revolutionise the way we approach this critical aspect of our operations. With their first-hand experience of working with this approach, they are as excited about its potential as we are. They have identified considerable benefits for their customers, as well as for improved operational efficiencies. This approach has already piqued the interest of other UK network operators, who are keen to explore how they can incorporate it into their own operations. With so much positive feedback, it’s clear that this new approach is set to make a real impact on the industry.
Is there a special requirement for the substations you want to use to deploy this method?
No, not really, though some substation designs make it simpler than others. The pre-fault detector (the Guard) is a very small piece of equipment that can be deployed within distribution cabinets.
We spent a lot of time on the Project determining the minimum sensor set required to give the information that is required to determine the presence of developing faults on the LV circuits fed from Ground Mounted Substations and their probability of failure. Some issues were experienced with designs of totally enclosed distribution cabinets which restrict access to the voltage connections and some clever thinking was also required on some cable terminations to get the required measurement of neutral current.
The biggest issues come when additional deployment of ‘Locate and Manage’ Equipment is signalled. Not all distribution cabinets are able to accommodate the reclosers that are the ‘go-to’ devices for proving impedance to fault whilst managing the customer supplies should transient faults occur during the location phase of the methodology. For this reason, we developed a separate piece of equipment (the Pre-fault Locator) that provides the same measurements of distance to fault and impedance to fault in situations where Reclosers are not the preferred option.
Do you have an idea of the precision of the location method?
The answer isn’t a straightforward percentage. The method has been shown to be highly accurate but is dependent upon three factors not all of which are under the control of EA Technology. The accuracy of the substation measurements, the assumptions made as part of the method/ algorithms used on those measurements and, importantly, the accuracy of the cable plans both geographically and in terms of cable asset information. Accurate location is therefore a team effort. Higher energy pre-faults (sometimes manifesting as transient off-supply events) give more accurate results than low-energy pre-fault events captured during the early stages of fault development. The substation measurements, marked-up plans and notification identify the area of the fault rather than an ‘x marks the spot’ location. Confirmation of the actual location using normal site techniques (e.g. fault gas sniffing etc) is highly beneficial before proactive repair. These methods are also more successful for higher energy events.
Recognising this, the Foresight Project undertook the development of a pavement-level fault passage indicator that could be used to determine the passage of pre-fault current as it flows from substation to the developing fault StreetSense). This is a very exciting development which we are aiming to trial further and, if it is successful, are sure it will be a game changer in the Foresight Find and Fix before Fail methodology.