How do users of ageing electrical plant know that they’ve got a problem? It stopped working.It still works but something is obviously not right.‘There aren’t many of that type around elsewhere.’
Installation and commissioning of replacement equipment can be a nightmare, especially in COVID times. But in normal times it is bad enough. Aside from quality control of the actual installation, possible problem areas include access, production rescheduling, staff training, spares holdings, floor space footprint mismatches, delivery delays and documentation.
It is small wonder that, when capital costs are also factored in, it often seems attractive to just stick with whatever is on site and keep it running longer. But for how long? What are the limits for particular types of equipment? What are the failure mechanisms?What could make it last longer? And most importantly, is it possible to be warned of impending failure?
The plant engineer faced with these issues will no doubt have been diligently following the manufacturer’s instructions for periodic maintenance and checking, to ensure that the equipment remains safe and fit for purpose. But the maintenance instructions may go no further than that. A different set of issues arises if the instructions have been lost and the manufacturer is no longer in business:then the plant engineer and site staff will have to know everything that there is to know about maintaining that piece of equipment, or find a contractor who does, possibly at short notice.
It’s not ideal. There must be many engineers who are faced with these problems and say: “I think I could deal with this, if only I knew where to start and what to look for”.
What is not clear from the outline above is that a very significant part of the management of ageing electrical plant is the need for accurate, consistent and complete recordkeeping. This is because in practically every case the environment and circumstances of use will have a profound effect upon the useable life of the asset. One-off events such as lightning strikes or fault currents may result in a sudden reduction in residual life. Continual immersion, high temperatures, salt spray corrosion and prolonged operation close to maximum rating will result in progressive and possibly more predictable degradation.
Consider a transformer as an example. Transformers are generally one of three types:mineral oil-cooled, biodegradable fluid-cooled or cast resin air-cooled. The first type is the most common and will be used for this example (a damaged unit is shown on p42). The transformer is likely to have a tap changer to adjust the voltage, which may be done on- or off-load. The windings are insulated with paper, which is impregnated with the cooling oil. The external connections are via bushings which are generally porcelain. There may be a pump to circulate the cooling oil. A number of failure scenarios spring to mind: Slow degradation of the cooling oil and the paper insulation due to the operating temperatureDistortion of the windings and possibly the core due to fault currents or lightning strikes. This may result in immediate internal damage due to reduced clearances causing flash-over or shorted turns, or a slower degradation due to partial dischargeFailure of the tap changer mechanism due to frequent (or infrequent) useOverheating due to cooling pump failurePartial discharge failure of the bushings, perhaps due to manufacturing defect, poor installation or mechanical damage, surface contamination or lightning strikeWater ingress affecting any part of the transformer.
It must now be clear that there is value in keeping a detailed inventory of assets with a full history including incidents, environmental conditions, any changes of location or configuration, maintenance activities, tests and test results.
Testing may invoke a range of activities depending on the transformer’s condition and suspected degradation mechanisms. Perhaps the best known is analysis of the cooling oil to detect the products of degradation of the paper insulation and internal arcing. Useful results can be obtained by appropriately trained site staff during routine maintenance, but a more detailed analysis can be provided by a laboratory and is advisable from time to time for high voltage transformers or, for other transformers, to establish a baseline and confirm trends. Other tests include acoustic analysis and thermography while on line, and winding resistance, insulation, swept frequency analysis, capacitance and tan delta tests while off line. The value of a critical visual inspection should not be overlooked!
Where does this lead? The desired outcome is to establish a trend and show rate of progression of degradation. Studies indicate that there is a wide variation in typical transformer lifetimes depending on size and construction, as well as use and environmental factors, such that large transformers may last 30 to 50 years, whereas small transformers may last as little as 10 years. Whether to refurbish or replace is then an economic argument tempered with considerations of convenience.
But few ageing electrical assets are as uncomplicated as transformers. Variable speed electrical drives are a case in point. They combine electrical, electronic, software and mechanical failure points in a highly proprietary and integrated package. Moreover, the technology used has changed rapidly over the last 30 years, both in terms of drive electronics and control hardware and software. That is not to imply that knowledge of their failure mechanisms and possibilities of life extension are beyond the grasp of a site electrical engineer; only that for some assets obsolescence of the component parts may play a greater part in assessments of residual life.
For example, where nowadays would you go to source a replacement for an Intel 8008 microprocessor or some of the older memory chips? Perhaps eBay? But what would be their provenance? That said, some electronic components such as electrolytic capacitors have seen general improvements – for example being routinely available to higher rated temperatures. Routine replacement of electrolytic capacitors in control electronics and power supplies could bring real benefits for life extension because overheating is a major cause of failure, especially where drives are used in dusty conditions, as they often are.
In contrast, obsolescence tends not to be a very significant factor for the life expectancy of cables. There are many recorded instances of cables which are over 100 years old and still providing useful service. This is of course dependent on factors such as the quality of manufacture and installation, as well as use within specification in a benign environment. Environmental factors can result in service life being a lot less – perhaps 20 years. Ultraviolet exposure, minor manufacturing defects and inconsistencies, partial discharge and water treeing may all be significant degradation factors.
Cable insulation and construction tend to be optimised for particular types of use, for example fire resistance or high temperature. But optimising for one characteristic is generally to the detriment of others, such as flexibility or permeability to water, which may be particularly so for some earlier generations of cables. There may also be batch differences between nominally identically specified cables. Again, it comes down to knowing what you’ve got, observing changes, testing as appropriate and keeping detailed and accurate records.
These examples have been chosen because in some respects they represent extreme cases of a type of ageing management. They are certainly not the only assets with age-related problems. One might mention switchgear and protection as an example of the need for regular maintenance activity and where aftermarket upgrades of ageing plant may bring real benefits, or UPS as an example of where accurate understanding of the original requirement may have already determined the achievable asset life, given good maintenance.
Mindful of many of the issues outlined above, EEMUA has tried to give a helping hand by producing new guidance developed by its electrical engineering committee, which brings together engineers, technical authorities, policy managers and others from across industry. EEMUA 227, ‘Management of ageing electrical assets’, has sections dealing with UPS, electrical drives, electrical protection devices, power cables, switchgear, transformers, and generators and motors.
Each section has a similar structure: technical introduction, typical failure modes, condition monitoring techniques, expected and acceptable values, assessment of failure risk, life extension methods. The guiding principle has been not to duplicate the content of some of the existing excellent guidance available in standards, from regulators, or on the internet generally, but to give a broadly applicable overview with pointers to more detailed information.
Probably there aren’t too many engineers who would deliberately set out to be experts on ageing electrical plant. Hopefully they won’t need to be if they can be better equipped to understand the issues and find the right information.
BOX: Other advice on ageing equipment
Over the last five years, a new methodology of hazard identification has emerged that aims to pick up on hazards caused by the accumulation of many small changes – so-called ‘creeping change’. A guide was published in 2017 by the Petroleum Institute (see www.is.gd/cojebo). It says: “The CCHAZID methodology covers both engineering (including process safety; mechanical engineering, and electrical, control and instrumentation (EC&I)) and human/organisational changes.” It argues that the techniques, which were developed for high-hazard industries, are applicable to any industry, and should be used to regularly review plant risks in terms of safety, environment and commercial aspects. The subject continues to be taught by HSL, including in a Buxton course in early February 2022 (see www.is.gd/iricol).
Ageing assets are a particular problem in the nuclear power industry, as many commercial stations have had their initial operational term extended. International nuclear agency IAEA has expanded its guidelines for long-term operation of nuclear reactors through its peer-review process called SALTO (safety aspects of long term operation; www.is.gd/ekiguh). The first such mission was in 2005, and more than 40 have been carried out.-Will Dalrymple