If you’ve spent time on both sides of the electricity world – inside the grid and outside it you eventually realise something that isn’t obvious from either vantage point alone: perception is not reality. From the outside, the grid can appear slow, bureaucratic, awkward, inflexible. Thirty years after privatisation, people still call it “public sector in disguise.” But from the inside, another picture emerges, a system built with one purpose so clear it could be printed on a mug: keep the lights on. Through storms, budgets, elections, outages and ageing infrastructure, it has largely done exactly that.
Tonight, perhaps a thousand people across the UK will lose supply out of sixty-four million. A 99.99% reliability rate, delivered by a network that stretches over valleys, beneath rivers, through cities and across coasts. No modern software platform would dare promise the same. But the grid didn’t achieve this performance through magic or monotony. It achieved it through structure, structure that grew slowly, sometimes painfully, over generations.
That structure wasn’t created to exclude contractors. It wasn’t even created with decentralised energy in mind. It was built to manage a national system at national scale, in an era where assets were large, teams were stable and careers formed over decades. Understanding that context is crucial, not because we need to replicate it, but because the decentralised world now expects contractors to deliver reliability without ever having been given the same foundations.
In the centralised world, competence deepened because the environment stayed still. People entered through apprenticeships, learned from crews who’d been on the job for decades and absorbed the unspoken knowledge that no textbook tries to capture. Engineers recognised the quirks of circuits, remembered which substations flooded and understood the rhythm of faults across the seasons. Much of that knowledge was passed on through conversations that, from the outside, might sound like small talk “that circuit again,” “remember dragging that switchgear up the hill,” “has anyone checked that retrofit radiator?”, but in reality, formed the backbone of a living, collective intelligence.
Contractors today don’t lack competence. They lack the conditions that allow competence to compound.
Decentralised systems are agile by design. Teams rotate. Companies evolve. Workloads rise and fall with policy, incentives and market cycles. The rooftop world doesn’t provide decades of static geography or a closed loop of repeated experience. It provides variety – huge, valuable, entrepreneurial variety but variety without continuity comes at a cost. Not a competence gap, but a context gap.
Hierarchy inside the old system is often misunderstood. From outside it looked rigid; from inside it provided clarity. When a director once banged the table and said, “This is not a democracy,” it wasn’t arrogance it was recognition that, in a high-risk environment, somebody must hold authority clearly and openly. Everyone knew who could isolate, who could approve, who carried responsibility if something went wrong. The outcome wasn’t oppression; it was coherence.
Contractors rarely get the luxury of that coherence. They inherit sites designed by different hands, installed to different standards, documented to different levels of completeness, maintained by different firms operating under different commercial pressures. Responsibility blurs not because anyone lacks professionalism, but because decentralisation without shared frameworks creates fog.
One of the grid’s greatest strengths was the socialisation of failure. When something went wrong in one region, the entire system learned. If a switchgear type began failing disruptively, every depot knew within days. In one case, the failures initially looked national in scale with a pattern of five failures in six months. It looked catastrophic until a few conversations revealed it wasn’t a national issue but a regional oversight: a radiator retrofit designed for a hotter climate needed checking every two years, separate from the four-year service. Nobody had been doing the two-year check. The fix was simple; finding the pattern required a hive that spoke to itself.
Now imagine that pattern occurring in rooftop solar. Does the learning travel? Does the cause get recorded? Does a contractor two counties away benefit from the discovery? Most of the time, no. Not because contractors don’t care, they deeply do but because the system does not yet give them a way to see what each other sees.
Standards in the centralised world weren’t PDFs; they were behaviour. Switching protocols, isolation procedures, PPE requirements, clearances all internalised through repetition. “I do it my way” wasn’t celebrated because in a tightly coupled network it could be dangerous. Variation wasn’t creativity; variation was where accidents began. Rooftop solar, by contrast, invites variation. Different teams. Different methods. Different documentation. In some cases, companies guard their forms as if transparency threatens their competitive advantage. But a system cannot learn if its basic data is inconsistent or inaccessible.
Centralised systems also had time, decades of it. Thirty-year investment cycles shaped decisions. There was no rush to beat a subsidy deadline or capitalise on a six-month political window. The challenge wasn’t finding funding; it was applying it responsibly. That long view allowed skills to mature, standards to stabilise and institutional memory to deepen without interruption. Rooftop solar, incredibly, has grown in the opposite conditions: fast, urgent, policy-driven, customer-led, often reactive. Contractors are expected to deliver long-term reliability while operating in short-term environments. That tension is systemic, not personal.
So when people describe centralised systems as rigid, slow or old-fashioned, they are describing the exterior not the performance. The reality is simpler: centralised systems were powerful because they were aligned. Training, standards, competencies, planning and operations all pulled in the same direction. Everyone worked within the same hive. Decentralisation removes that alignment unless we rebuild it deliberately.
Rebuilding it is exactly what the next decade of solar depends on. Because once you decentralise, you lose the very thing that made the grid so dependable: the ability to rely on the same process, the same authorisation, the same learning and the same behaviour across every asset in the ecosystem. Without that, a decentralised system is not a distributed hive at all – it is a scattering of solitary bees, each working hard, each doing their best, but without a shared choreography to hold the whole together.
For all its limitations, the centralised energy system succeeded because it treated reliability as a discipline. It recognised that complexity must be engineered, not improvised; that variation must be managed, not indulged; and that serving millions of people, every hour of every day, requires a culture where responsibility is clear, competence is predictable and learning is shared. That discipline is why the system endures.
Decentralisation does not remove the need for that discipline. If anything, it amplifies it. When thousands of micro-assets behave like miniature power stations, the coherence that once held the hive together begins to dissolve. The challenge of the next chapter is not merely embracing decentralisation it is rebuilding the invisible framework that made centralisation work: methods, standards, shared memory, and the quiet choreography of competence.
Because now, the weakest link is no longer buried in a substation diagram. It could be sitting on any rooftop in the country. Our future depends on whether we recognise that and act before the failures begin to rhyme.
