Transformer life extension |
A new approach to estimating the remaining useful life of LV transformers
Combining its non-invasive continuous monitoring technology with probability-of-failure modelling and long-term load scenarios, Oktogrid has developed a data driven approach to estimating remaining useful life of distribution transformers. A recent pilot project in Germany has demonstrated how substantial hidden capacity can be unlocked and the life of ageing infrastructure significantly extended
Jonas Kutteruf Oktogrid, Denmark
Before the energy transition, planning in the low-voltage network was comparatively straightforward. Demand profiles were stable and predictable, and transformers were installed with substantial capacity buffers. According to demand forecasts at the time, those capacity limits were unlikely ever to be approached. The situation today is markedly different. The ageing of assets coincides with rising demand from heat pumps, electric vehicle charging and other electrification trends. At the same time, consumers have increasingly become prosumers, feeding power into the network through residential PV systems. Together, these developments are driving low- voltage transformers closer to their capacity limits. Operational challenges are intensified by workforce constraints and extended procurement lead times for new transformers, making asset replacement more complex and less flexible. As a result, grid operators require far more precise assessments of the remaining useful life (RUL) of their assets. Yet many low- voltage transformers – originally designed to operate for 50–80 years – still run without any meaningful condition data. With limited visibility of thermal behaviour or asset condition, planners must rely on conservative loading assumptions and age-based replacement limits, often leading to premature retirements or unexpected failures. Figure 1 outlines the typical mechanism behind conventional replacement decisions. A recent pilot project undertaken jointly by a distribution system operator (DSO) in Germany and Danish company Oktogrid explored a new approach to estimating RUL. By combining continuous measurements with CNAIM probability of failure modelling and long-term load scenarios, Oktogrid was able to present a data driven approach to estimating RUL and facilitating replacement planning for the operator. CNAIM (Common Network Asset Indices Methodology) was jointly developed by all six UK distribution network operators and recognised by UK energy regulator Ofgem. It provides a standardised method for evaluating asset condition and probability of failure, based on inspection and maintenance data from more than 10 000 distribution transformers.
Non-invasive monitoring, condition and performance data, and probability of failure To better understand the condition and utilisation of its low-voltage transformer fleet, the German DSO equipped ten 10–20 kV/0.4 kV transformers with Oktogrid’s non-invasive monitoring system. Over an eight-month period, around 2.7 million measurements were collected for the entire fleet. The monitored parameters included: transformer load;
dynamic thermal rating (according to IEC 60076-7);
top-oil and hot-spot temperatures; total harmonic current distortion; vibration response; acoustic sound level;
partial discharge activity (ultrasonic emissions); and
ambient temperature and humidity. Using those parameters it was possible to estimate the performance and condition of the transformers.
This condition information was then incorporated into the CNAIM model to calculate the future probability of failure (PoF) for each transformer. Condition data was combined with static factors such as age, installation environment and inspection history. CNAIM
22 | November/December 2025|
www.modernpowersystems.com
provided a structured framework to quantify health indices and establish RUL. The inclusion of condition data allowed a refined PoF calculation beyond conventional age-based assessments.
Taking account of future loading Because condition alone does not determine replacement timing, Oktogrid also evaluated future load development using Scenarios A (light rise in energy demand) and C (substantial rise in energy demand) from the German grid development plan (Netzentwicklungsplan) 2037/2045. These scenarios were extrapolated to 2075 to reflect long-term trends in energy demand. Historically, many DSOs define end-of- life at approximately 100% rated load. However, thermal modelling based on real-time top-oil and hot-spot temperature indicated that actual dynamic thermal rating could be significantly higher. The ten monitored transformers showed: minimum additional capacity: +8%; maximum additional capacity: +26%. This difference reflects the gap between nameplate-based planning assumptions and the true thermal state recorded in operation for the lowest observed additional capacity per transformer. Figure 2 displays the utilisation prognosis for a selected transformer with rated and dynamic thermal rating limit.
Oktogrid data collectors. Photo: Oktogrid
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53
