More than 50 senior figures from the electricity industry gathered at London's Royal Automobile Club on 5 February 2009, for NetWork 2009 - the first ever international DNO strategy conference. Top of their agenda was how the ability to measure the condition of live assets is making the management of network assets more efficient, at lower cost. Neil Davies from EA Technology Instruments investigates
The inaugural NetWork 2009 event in February was an extremely valuable opportunity for UK DNOs to share knowledge on the key strategic management issues facing network operators and learn from the examples of two of the world's most reliable and efficient networks - SP Powergrid of Singapore and China Light and Power (CLP) of Hong Kong.
The pressures are common to every operator across the world: how can they manage an ageing asset base so that it will deliver greater network reliability, power quality and safety, while reducing costs to consumers? At the same time, how can they make a watertight business case for investment in maintaining, upgrading and replacing assets to stakeholders, including industry regulators?
The answer to these questions is being found in two developments which are inextricably linked: new techniques for accurately measuring the condition of live assets, plus new methodologies for managing assets more effectively, based on their actual condition.
Let's look at what has been achieved in Hong Kong and Singapore, where condition based asset management has become the driver for remarkable improvements in both reliability and cost efficiency:
SP Powergrid, Singapore
SP Powergrid's network includes nearly 10,000 substations, 40,000 switchgear sets, 14,000 transformers and 30,000km of cable. Since incorporating condition monitoring into its systems, it has dramatically improved an already excellent performance. The System Average Interruption Duration Index (SAIDI) has averaged less than 1 minute per year over the last three years.
NB: The blip in 2004/5 was caused by a third party supply issue outside SP Powergrid's control.
SP Powergrid estimates over the last eight financial years, condition monitoring has enabled it to avert 450 network failure incidents, with a net financial saving of US$29m. In addition to improving customer service, it has been able to pass cost savings on to them.
CLP Hong Kong
The China Light and Power network in Hong Kong includes nearly 13,000 substations and 22,000km of overhead lines and underground cables, serving 2.26 million customers.
As a result of focusing over the last 10 years on condition based maintenance, to predict faults and improve reliability , it has reduced its SAIDI figures from more than 40 to 2.68 minutes lost per year
Demand from customers has continue to grow, but in the last two years, greater operating efficiencies have enabled CLP to reduce tariffs.
The UK Business Case
Taken as a whole, the UK electricity network is relatively efficient. But an in-depth analysis by EA Technology Consulting of preventable, condition-related failures, shows there is considerable scope for improvement:
Using condition monitoring as a failure prevention tool is a valuable technique, but is only part of a much wider move towards condition based asset management techniques.
Using Condition Data
The ability to collect data on the condition of live assets is transforming the industry's approach to asset management itself: from one based on time-scheduled maintenance and replacement, to one based on a detailed understanding of the condition of the asset base. It also provides accurate intelligence for investment programmes.
Maximising the value of data is essentially carried out at two levels:
Asset condition registers
Expert analysis and interpretation of PD activity readings gives a clear indication of the condition of assets, including accurate predictions of when they are likely to fail. In EA Technology's case, this is based on a unique database, built up over more than 30 years, which shows how tens of thousands of asset types have deteriorated over time.
This approach enables operators to develop registers of assets, in which each asset is accorded a ‘health index' showing its present condition, its predicted date of failure and/or its remaining service life.
Condition Based Risk Management (CBRM)
CBRM is a comprehensive new methodology, which takes condition based asset management to a higher level, enabling managers to take more intelligent decisions on revenue and capital spending. It also reduces the cost of network operation, while improving their efficiency and reliability.
The effectiveness of CBRM derives from factoring together probability (derived from the asset condition) and consequences of asset failure, to determine risk in terms of financial cost.
In addition to managing the health of assets, CBRM provides the answers to the key questions:
- If an asset costing £XX fails, what will be the consequential loss to the business?
- If an asset is refurbished or replaced at a cost of £YY, what will be the benefit to the business?
- Therefore, where should we prioritise our spending?
EA Technology's experience shows that partial discharge (PD) activity is a factor in around 85% of disruptive substation failures. It has thus become increasingly clear the ability to detect and measure PD is key to assessing the health of assets. PD activity provides clear evidence that an asset is deteriorating in a way that is likely to lead to failure. The process of deterioration can propagate and develop, until the insulation is unable to withstand the electrical stress, leading to flashover.
Partial discharges emit energy, in the form of effects which can be detected, located, measured and monitored:
- Electromagnetic emissions, in the form of radio waves, light and heat.
- Acoustic emissions, in the audible and ultrasonic ranges.
- Ozone and nitrous oxide gases.
The most effective techniques for detecting and measuring PD activity in live assets are based on quantifying:
Transient earth voltages (TEVs)
The importance of TEV effects (discharges of radio energy associated with PD activity) was first identified by EA Technology in the 1970s. Measuring TEV emissions is the most effective way to assess internal PD activity in metalclad MV switchgear.
PD activity creates emissions in both the audible and ultrasonic ranges. The latter is by far the most valuable for early detection and measurement. Measuring ultrasonic emissions is the most effective way to assess PD activity where there is an air passage e.g. vents or door in the casing of an asset.
PD activity can also be measured in the UHF range, and is particularly useful in monitoring EHV assets.
The latest PD instruments typically use a combination of ultrasonic and TEV sensor technologies, characterised by the EA Technology UltraTEV range. These include:
- Handheld dual sensor instruments which provide an instant indication of critical levels of PD activity, ideal for ‘first pass' PD surveys and safety checks. Traffic light warning levels are precisely calibrated using a database of known patterns of asset deterioration.
- More sophisticated handhelds, which provide audible and numerical readings of ultrasonic and TEV activity.
- PD location instruments which pinpoint and quantify the source of PD activity.
- PD monitoring instruments, which measure, record and analyse PD activity over time.
- PD alarm systems, which give immediate warning of critical PD activity in groups of assets or whole networks.
- Specialist PD monitoring systems for strategically important assets, including Gas Insulated Switchgear (GIS).
Other Asset Classes
Condition based management is by no means confined to assets which present faults in the form of PD activity.
The same principle is equally effective, using a range of condition measurement techniques, to all types of electricity network assets including substations and cables. It can apply to the complete asset, such as an overhead line, as well as to the component parts, such as the overhead conductors, poles, towers and footings.
The ability to assess the condition of live assets is changing the way assets are managed on many levels: as a technique for preventing faults from developing into failures, as a means of moving from time-based to condition-based maintenance, as a way of quantifying risk and as the basis for justifying and prioritising investment.
But the ultimate rationale for condition measurement is that it pays for itself, many times over.
This article includes material from presentations made at NetWork 2009, the first international distribution network strategy conference, held in London in February 2009. The full presentations are available from www.networkconference.co.uk, where readers can also register their interest in NetWork 2010.