Legacy constraints¶

Legacy constraints explain the architecture better than any design document would.

Equipment reaching end of life¶

Openly reported. The 2022 station programme in Amersfoort, described in the 2022 report, removes or replaces outdated components and renews a 50/10 kV transformer station to secure quality and add capacity, three of the four stations finished that year.

This sits alongside two end-of-life populations: brittle gas mains (grey cast iron and asbestos cement) and the ageing cable-and-joint stock whose internal-defect failures dominate the fault statistics. So end-of-life is running across gas mains, cables and station primary plant at once.

Spare part shortages¶

Framed as structural material scarcity rather than incidental. In 2022 part of the planned network-driven investment slipped partly on material scarcity at contractors, with prices rising from that scarcity, and the network-driven scope came in at 93 per cent for electricity and 96 per cent for gas rather than complete (2022 report).

Stedin’s investment plan for 2024 names the scarcity explicitly as sitting in technicians, in space, and in materials such as cables, lines and transformers. Transformers in particular are the item where scarcity and lead time compound.

Long procurement lead times¶

Quantified in a study the operators commissioned jointly. A high-to-medium-voltage station runs three to seven years from proposal to commissioning, of which materials ordering alone is around a year and construction one and a half to two, with environmental and permitting the largest and most variable slice; a medium-to-medium-voltage transformer station runs two and a half to three years, per a McKinsey 2020 study in a VNG guidance document.

Stedin’s own framing is that a transport-network adaptation has a lead time of many years, and that permitting trajectories are long, which is why connection lead times have lengthened and why it now says it would build more than it can, from its new-connection page and its 2050 study release. A one-year transformer order against a three-to-seven-year station is the constraint that sets planning horizons.

Multi-decade replacement programmes¶

Two horizons run in parallel. The near, hard deadline is brittle gas removal before 2028. The far horizon is the Masterplan 2050: Stedin split its area into 18 zones and built a masterplan per zone with a scope to 2050, precisely because a transport-network change takes many years to realise and has to be started early against uncertain demand ( 2022 report).

The strategic investment plan reaches to 2037, and the alternatives are compared on a minimal-regret principle to avoid stranded spend under deep uncertainty. So the replacement estate is governed on a multi-decade planning cadence, not a rolling annual one.

Mixed analogue and digital environments¶

Acknowledged in the operator’s own words, that it is putting ever more IT-like components into a traditional technical landscape, which helps steer more precisely but raises complexity (from Stedin’s public security account).

Concretely, smart sensors and Smart Grid Terminals are retrofitted into existing medium-voltage stations rather than deployed on a clean build, and on metering roughly 87 per cent of households had a smart meter by end-2023, which is also to say a non-trivial analogue remainder persisted alongside successive DSMR digital generations. The estate is genuinely hybrid: decades-old primary plant instrumented with a thin, newer digital layer.

Brownfield integration¶

This is where the constraints become visible as design choices. The GIS and asset-register modernisation is explicitly aimed at a loosely-coupled, API-first architecture to reduce the management burden on the legacy application estate ( Smallworld and Lovion), rather than a rip-and-replace (from a Stedin GIS-developer recruitment advert).

The station work is retrofit into live brownfield sites, and the security posture leans on island mode so critical, older operational technology can keep running detached from external connections.

The enterprise side shows the same brownfield pattern in migration form: an SAP move to S/4HANA and the shift of most business applications into Azure, both legacy-carrying transitions rather than greenfield builds.

Legacy constraints as architecture drivers¶

Taken in sequence, the legacy constraints account for the architecture more cleanly than the technology choices do on their own.

The loosely-coupled, API-first target is a rational response to a fixed legacy application core (Smallworld, Lovion, SAP) that cannot be replaced wholesale, so integration is designed to wrap it rather than remove it. Island mode is a rational response to a mixed analogue-digital OT estate where much of the plant predates network security, so isolation is the compensating control where modern authentication cannot be assumed. The standardised and clustered station design and the wijk-by-wijk “production caravan” delivery model are a response to the lead-time and materials constraints, since uniformity is what lets scarce components and scarce crews be scheduled and reused. And the Masterplan 2050 with its minimal-regret rule is a response to procurement lead times measured in years against demand no one can forecast, so early, low-regret commitments substitute for accurate prediction.

The through-line is that Stedin’s architecture is shaped less by what is technically ideal than by what a brownfield estate under materials scarcity and multi-year lead times permits. The scarce, slow, end-of-life physical layer is the independent variable; the API-first integration, the island-mode segmentation, the standardised stations and the 2050 masterplans are the dependent ones. A technical document would present those as design decisions; the annual reports show them as constraints made visible.

Last updated: 11 July 2026