Article Hub

Article Hub

Article Hub

Technical Articles

Written Interviews

In Focus

Media Hub

Media Hub

Media Hub

Podcasts

Webinars

Tech Talks

Power Panels

CEO Forums

Technology Showcases

Event Interviews

Other Brands

Trustworthy coverage of the transformer and transformer-related industries.

Visit Website

Trustworthy coverage of the transformer and transformer-related industries.

Visit Website

Copper to Code: Protecting Substation Transformers in the E-Mobility Era through Time-Aware Compliance 

The shift from centralized fossil generation to highly decentralized, weather-dependent renewables has changed how European distribution networks are operated. The most exposed asset in this transition is the oil-immersed distribution substation transformer. Its thermal loading profile is now influenced as much by human charging behavior as by traditional demand growth. 

Historically, utilities planned around transformer lifespans of 30–40 years, but real service life has always depended on design margins, cumulative loading history, cooling conditions, and temperatures reached in the winding hotspot region. Modern loading guides such as IEC 60076-7 explicitly model insulation aging as a function of hotspot temperature to estimate how quickly a transformer is consuming its thermal life, rather than a simple “years in service” rule. 

The rapid buildout of localized, high-capacity electric vehicle charging introduces highly variable and synchronized load patterns at low voltage nodes. Evening commuter charging in dense residential districts can create sharp, simultaneous load increases on individual transformers, pushing top oil and winding hotspot temperatures toward their permissible limits. Over time, repeated exposure to elevated temperatures accelerates the depolymerization of cellulose paper insulation and contributes to dissolved gas signatures associated with thermal aging in mineral oils. If distribution system operators (DSOs) want to protect these critical assets without slowing the e-mobility rollout, they need more than conservative current limits — they need to treat transformer and network capacity as a problem that can be examined in software, bridging copper and code to reveal how much real headroom is hidden behind static planning assumptions. 

Traditional connection assessments boil down to a single threshold: if the projected current is expected to exceed the rated current under worst-case conditions, the application is rejected or heavily constrained. In practice, this “above or below the nameplate” approach leaves significant usable capacity unused, as transformers have thermal inertia and can ride through short-duration overloads safely when ambient temperatures and preceding loading are favorable. 

Loading guides such as IEC 60076-7 and IEEE C57.91 share the same core principle: insulation aging accelerates rapidly as the hotspot temperature climbs above a defined reference level for the insulation system. For conventional oil-paper designs, that reference is placed around a specific hotspot temperature used to represent “normal” relative aging, while higher temperatures are only allowed under short-time or emergency operating conditions with explicit limits on duration and frequency. In other words, transformers are not binary devices. What matters is how often and how long an asset operates above that reference hotspot temperature, and under what ambient conditions. A brief overload on a cool night after a period of light loading has a very different aging impact than the same current on a hot afternoon following a congested morning peak. Dynamic thermal assessment captures this reality; static current checks do not. 

Planners can, in principle, track how ambient temperature, oil temperature rise, and local hotspot rise combine at each time step to produce the actual hotspot temperature, by embedding thermal relationships from IEC 60076-7 directly into network models. Instead of asking, “Does the current exceed 100 percent of nameplate?”, the question becomes, “Does this loading pattern keep the hotspot within our agreed aging budget over the horizon we care about?” That shift in framing is the conceptual foundation for everything that follows — including a narrower, related question this article demonstrates directly: how much of a transformer’s apparent capacity deficit disappears once time and coincidence are properly accounted for, even before full thermal modeling is applied. 

Apart from thermal aging, local grid strength also determines how aggressively a transformer must work to keep voltages within acceptable limits. When large controllable loads connect to electrically weak points — nodes with low short-circuit power — the same change in current causes a much larger voltage dip than it would at a strong node. 

To compensate for these voltage dips, the On-Load Tap Changer (OLTC) has to move more often. Over years of operation, high tap-change frequency translates into mechanical wear, contact erosion, and localized heating in the tap-changer mechanism and surrounding oil. Utility experience and dissolved-gas-analysis literature show that elevated thermal stress, even at temperatures below classical “fault” levels, can drive so-called stray gassing in some hydro-treated mineral oils, producing gases such as methane, ethane, or hydrogen in the absence of a discrete electrical fault. 

For planners, short-circuit power at the transformer bus remains a practical proxy for local grid strength, as it combines the nominal voltage with the available fault current into a single indicator. Lower short-circuit power means that each additional kilowatt of flexible load causes a larger voltage disturbance, more OLTC activity, and, in turn, more mechanical and thermal stress on the transformer. When DSOs map calculated transformer loading together with these grid-strength indicators, a clear pattern emerges: the riskiest nodes are not just the thermally loaded ones, but those where weak system strength, heavy flexible loads, and frequent tap-changer operations overlap. 

Across Europe, DSOs are facing a wave of connection requests for EV charging, heat pumps, and other flexible loads. Each request requires multi-scenario checks for thermal limits, voltage quality, and protection constraints. Processing these studies manually, case by case, was manageable when volumes were low, but is becoming a structural bottleneck as applications scale. 

A pragmatic response is to automate the repeatable part of the process. In this context, we refer to such workflows as Indicative Capacity Analysis (ICA): scripted studies that estimate local hosting capacity against agreed thermal and voltage criteria before an engineer performs the final review. Python is a natural glue language here, because it can interface directly with power-system simulation tools — in the case study below, via Pandapower — to run realistic, time-resolved scenarios rather than single worst-case snapshots. 

In a typical setup, the Python layer performs three main tasks. First, it prepares realistic load and generation profiles across a full operating year. Second, it drives a time-series simulation, stepping through those profiles while continuously reading out transformer loading and voltage levels at key buses. Third, it applies a set of business rules: if all relevant constraints remain within defined margins for the vast majority of the year, the connection is flagged for expedited approval; if violations occur, the workflow can propose mitigation options such as power limits, time-of-use windows, or minimal curtailment. 

Importantly, this is not an “AI black box.” Automation here acts as an advanced decision-support system that takes over the repetitive simulation and data-filtering work, keeping engineers in the loop for interpretation and final decisions. The responsibility for asset risk and customer impact remains with the engineering and asset-management teams. 

Traditional connection assessments often rely on a worst-case snapshot: does peak generation exceed the transformer’s rated capacity, assuming both occur simultaneously? For a representative 30 MVA distribution transformer modeled against a peak generation scenario of  35.11 MW, this static check shows a 5.11 MW deficit — under conventional planning rules, the connection would be rejected outright. 

But peak generation and peak load rarely coincide exactly. To test this, we ran a time-series simulation in Python using Pandapower, across 35,040 fifteen-minute intervals — a full year at high temporal resolution — modeling actual coincidental load alongside generation, rather than assuming worst-case simultaneity. At the moment of peak generation (35.11 MW), coincidental local load absorbed 12.4 MW, leaving a net transformer flow of 22.71 MW — 75.7% of rated capacity, well within safe limits. Across the full year, the transformer exceeded its rating for only 14 hours (0.16% of the time), manageable with minimal curtailment (0.4% of annual energy). 

The static view rejected the connection outright. The time-aware view reveals that most of the apparent deficit was never real: local load was already absorbing a significant share of the peak generation the static check assumed would flow entirely through the transformer. That gap — 

12.4 MW of capacity the static check couldn’t see — is hosting capacity that exists today, without any physical upgrade, once planners account for real operational timing rather than worst-case coincidence. 

This time-series result addresses capacity and loading directly; it does not, on its own, quantify winding hotspot temperature or insulation aging in the way described in Sections 1 and 2. The natural next layer for this work — and the direction we are taking it — is coupling this kind of time-resolved capacity analysis with transformer-level thermal modeling, embedding IEC 60076-7 hotspot calculations directly into the same time-series framework, so that “how much power can flow” and “what that flow costs the transformer’s insulation life” become one integrated assessment rather than two separate questions. We see that integration as a concrete, buildable extension of the approach demonstrated here, and one other DSOs facing similar e-mobility rollouts could adapt directly. 

The energy transition is pushing substation transformers into a new operating regime. Evening charging peaks, evolving grid codes, and longer procurement lead times converge on a single asset: the copper windings and paper insulation quietly heating and cooling in thousands of distribution substations. 

Engineers who can translate loading and thermal models into robust, transparent software workflows are in a unique position to protect these assets. By replacing worst-case, static assumptions with time-resolved, software-driven assessments — and, over time, embedding IEC-and IEEE-aligned thermal thinking into those same workflows — utilities can move from blanket conservatism toward informed risk management: unlocking capacity that already exists, targeting reinforcements where they deliver the most value, and extending the useful life of the existing transformer fleet under clearly defined conditions. 

Ultimately, compliance by design is not about surrendering control to algorithms. It is about giving asset managers and planners a shared, data-driven view of how each new kilowatt of flexible demand will interact with the copper, steel, and insulation that anchor the grid. When copper and code are treated as one system, the e-mobility transition stops being a thermal threat to transformers and starts becoming a solvable engineering problem. 


Clifford Omari Ondieki, power systems engineer 

Clifford Omari Ondieki is a power systems engineer based in Berlin, specializing in grid modelling and power-flow studies (DIgSILENT PowerFactory, Pandapower, Python). His research on equitable EV-charging deployment received the Best Paper Award at IEEE ETECOM 2025. He serves as Secretary of CIGRE Study Committee C6 and NGN OneGrid Representative, and is a member of IEEE, VDI, VDE. 

This article was originally published in the November 2026 issue of the Critical Transformer Components & Next-Gen Design magazine.

View Magazine

Like this article? Share on

Lightning bolts strike across a dark blue night sky.

Subscribe to our Newsletter

Subscribe to our newsletter and stay ahead with the latest innovations, industry trends, and expert insights in power systems technology. Get updates on cutting-edge solutions, renewable energy advancements, and essential best practices delivered straight to your inbox.