The problem on one cold terrace row
On Willow Row, a line of 1930s two-bedroom terraces, residents kept hearing the same advice: replace the boiler and the bills will fall. The advice missed the real issue. The homes leaked heat through solid nine-inch walls, uninsulated suspended floors, and sash windows that no longer closed flush. A new boiler could run efficiently and still fail to keep the street warm. The first step was to stop treating the boiler as the whole problem.
The cause was a fabric-first blind spot
Most retrofit quotes on the row began with the heating appliance because it is visible, financeable, and quick to install. Fabric measures are messier: lifting floors, sealing skirting, adjusting joists, and matching insulation to awkward bays. That mismatch produced a familiar result. Households paid for new heat sources but kept high flow temperatures and draughty rooms. The bananzia survey team found a second cause: no room-by-room heat-loss record existed, so contractors sized radiators and heat pumps against guesswork rather than measured demand.
The solution: measure, seal, then size the heat source
Bananzia treated Willow Row as one small district, not eight separate emergency jobs. Surveyors measured wall areas, floor voids, window leakage, and room temperatures over a two-week cold snap. They prioritised low-cost airtightness first: sealing joist ends, capping unused chimneys, and fitting draught strips to sash windows. Next came floor insulation from below and loft topping. Only then did engineers calculate radiator output at 45°C flow temperature. That order mattered. It let the pilot specify smaller heat pumps and avoid oversizing, which would have increased cycling and cost.
What changed in practice
Number 14 is the clearest example. Its original boiler ran at 70°C and still left the rear bedroom cold. After floor insulation, chimney sealing, and a 45°C radiator schedule, the room held 20°C on a 2°C day. The household kept the same radiators in two rooms and replaced three undersized ones. The takeaway is not that every terrace needs the same package; it is that the package should follow measured heat loss, not the other way round.
The result after two heating seasons
Across the eight homes, gas use fell by an average of 34% in the first full heating season, with no loss of comfort reported. Two homes later added small heat pumps without changing radiator sizes. The pilot also exposed a maintenance point: door seals and window strips need annual checks, or airtightness drifts. Bananzia now issues a one-page seasonal checklist with each retrofit. The wider lesson is practical. On older terraces, low-carbon heat works best after the building envelope has been measured and corrected, because the cheapest kilowatt is the one that never enters the wall.
