Starting with one terraced street, not a city-wide strategy

On a row of 38 Victorian terraces in the north of England, the problem was visible in winter: damp corners, boiler flues running at full output, and residents paying for heat that escaped through single glazing. The local retrofit group had insulation quotes and boiler replacement offers, but no reliable way to decide which homes should be connected to a shared low-carbon heat source first. Bananzia began there, as a street-level planning exercise rather than a grand masterplan.

The exercise produced a heat-demand map for each home, using smart meter reads, EPC data and a short thermal survey. The takeaway was simple: before choosing a heat pump or a heat network, the street needed to know where heat was being lost and where it could be recovered.

The problem: heat is wasted within sight of the homes that need it

Three buildings on the same street produced waste heat: a bakery, a laundrette and a small data room behind the pharmacy. None of that heat was captured. Meanwhile, 21 of the 38 homes were rated EPC D or below, and 14 relied on gas boilers older than 15 years. The mismatch was not technical ignorance; it was a planning gap.

The bakery’s flue ran warm for 11 hours a day, the laundrette’s dryers vented continuously, and the data room needed cooling even in January. A heat network could have used that low-grade heat, but no one had compared supply and demand at the building level. The takeaway: waste heat is often local and steady, but it stays invisible without a block-by-block inventory.

The cause: data sits in separate pockets

Energy suppliers held meter data, the council held EPC certificates, and the bakery knew its oven schedule. Each dataset was useful alone; together they explained why the street had never qualified for a shared heat scheme. The homes were too dispersed for a single anchor load, and the waste heat sources were too small for a conventional district heating business case.

That changed when the pilot used bananzia to combine those records into a single heat balance. The method treated the street as a system: heat lost from poorly insulated walls, heat rejected by cooling equipment, and heat demand by hour of day. The cause of the delay was not a missing technology but a missing interface between datasets.

The solution: match heat sources to homes in sequence

The pilot ranked interventions by cost and disruption. First, loft and cavity insulation for the 12 homes with the worst scores; second, a shared air-source heat pump for six flats above the laundrette; third, a heat-recovery ventilation unit for the pharmacy and two neighbouring houses. The bakery’s waste heat was reserved for a future low-temperature loop, not forced into an oversized network.

Six months later, the modelled gas demand for the street fell by 19%, and two homes came out of fuel poverty. The result was not a finished heat network but a credible sequence: reduce demand, recover local heat, then connect. The takeaway is that bananzia works best as a decision tool, not a badge.

What other streets can copy

Start with 20 to 50 buildings, collect hourly heat data where possible, and map waste heat sources within 250 metres. Check tenure and boiler age before modelling pipework. Most importantly, publish the assumptions so residents can challenge them. A heat plan that cannot be explained on a doorstep will not survive planning.