This article is the third in a series on peak demand in electrified heating systems. The previous piece – The Design Sequence: What Happens Before Plant is Sized – described the decisions that determine peak heat demand before a heat pump is selected. This one examines what the process produces when those decisions are skipped.
Every credible route to net zero, including those modelled by the Climate Change Committee (CCC) and the National Energy System Operator (NESO), treats demand reduction as the first intervention, and a prerequisite to decarbonisation. In heat electrification, it is too often skipped, and the cost is embedded in energy centres across the country.
Demand Not Reduced
The energy efficiency hierarchy is straightforward: reduce demand before sizing plant to serve it.
Fabric performance, ventilation strategy, DHW treatment, and distribution temperature all determine how much peak demand reaches the energy centre. When these decisions are made without reference to their cumulative effect on the plant, or not made deliberately at all, the input to sizing is already inflated before a heat pump has been considered. Plant is then sized for a worst-case condition that good upstream design could have reduced substantially. The plantroom that results is sized for a condition that may occur once in a decade, with the capital cost to match.
Plant Sized Directly to Peak
The second condition compounds the first. Without thermal storage to absorb peak demand, generation must meet it in real time. Plant capacity and electrical connection are directly coupled, but the relationship is not simply one of installed capacity. The electrical input required to serve a given heat demand depends on the efficiency at which plant operates. A heat pump with a higher coefficient of performance draws less electrical input for the same heat output, reducing the connection the plant requires. First generation units, operating at lower COP under peak conditions, offered little relief on this front.
Without storage to decouple from peak, and without the efficiency to moderate the electrical draw at that peak, connection charges are significant capital items, fixed at the point of infrastructure agreement and not easily revised. They do not reduce when the building operates at typical load.
Like-for-Like Replacement
The problem is most acute in refurbishment. When gas boilers are replaced with heat pumps on a like-for-like basis, plant is sized to match the installed capacity of its predecessor – which was almost certainly oversized to begin with, specified to a design standard of its era and inflated by successive contingency allowances over its operational life. Replacing it like-for-like does not electrify a building’s heat demand; it electrifies its historical oversizing. The electrical connection must then be capable of serving that inflated figure. The building has never operated at that demand. It never will. But the infrastructure cost is real, and it will remain for the life of the building. Avoiding it requires metering data and energy bills, not installed capacity figures.
The Cumulative Effect
Each condition makes the next worse. Unmanaged demand inflates the sizing input; peak-led sizing locks in the connection cost; like-for-like replacement adds historical oversizing on top. The result runs through capital expenditure systematically: more plant, a larger electrical connection, a larger plantroom, and a more complex hydraulic circuit to manage the mismatch between installed capacity and typical operating conditions. The cost of the larger energy centre appears in the budget. The design sequence that produced it does not – it is simply absorbed as if the outcome were inevitable.
Reinstating the Hierarchy
The CCC and NESO assume demand reduction when they model affordable decarbonisation. Fabric-first design has always intended it. What the current pace of electrification has bypassed is the sequence that makes it real. When the hierarchy is reinstated – demand reduced, storage used to decouple from peak, plant sized to what remains – the energy centre is smaller, the connection is cheaper, and the plant operates at a load factor that reflects how the building actually behaves.
The technology that makes the correct sequence deliverable, and why first generation plant could not, is the subject of the next article.
