Building systems

Heat pumps in existing buildings: when replacement actually pays

Swapping a boiler for a heat pump does not always pay. Three conditions of the building decide the outcome, long before the choice of equipment.

Reza Salimi · Co-founder and CEO, CarbonysCo-founder and CEO of Carbonys, a Turin-based company building energy analysis software for Italian SMEs. He studied economics and management, and works on how estimated energy savings become numbers a company — or its bank — can act on.

The question almost always arrives in the same form: is it worth switching to a heat pump? It is the right question, asked about the wrong object.

Efficiency is not a property of the machine. It is a property of the coupling between the machine and the building it has to heat.

Three conditions, in order of weight

In an existing building the outcome depends almost entirely on three conditions. They are listed in order of impact, and the order is surprising: choosing the equipment comes after all three.

1. The flow temperature required

A heat pump performs better the lower the temperature of the water it has to produce. A radiator circuit sized for 70 °C and an underfloor system at 35 °C are not two variants of the same measure: they are two measures with different economics. Before evaluating any unit, it is worth knowing what temperature the building can actually run at on the coldest days — almost always lower than the design value.

2. The load set by the envelope

The demand the system has to meet is decided by the building envelope. In a building with widespread thermal bridges and high U-values, replacing the generator shifts the cost of energy but not the quantity: you pay less per unit and keep consuming too many. Here the order of measures matters more than their sum, and the best sequence rarely starts with the system.

3. The ratio between electricity and gas prices

This is the condition that changes fastest and gets considered least. The saving on the bill depends on three quantities, and none of them is a constant:

• the electricity price actually paid per kWh, levies included
• the gas price per kWh over the same period
• the real seasonal performance of the system, not the nameplate figure

A calculation built on one month's price describes that month, not the investment. The historical series published by Eurostat are a more honest starting point than a recent bill.

Reading the return without fooling yourself

Simple payback is the most quoted number and the least useful: it ignores the cost of capital and the useful life of the measure. A decision that commits ten or fifteen years needs NPV and IRR — the same indicators a bank or an accountant will use to assess it. EN 17463 (ValERI) sets out exactly how to appraise energy-related investments on those terms.

Where an analysis like ours sits

Carbonys is an analysis instrument, not an energy audit: an audit under UNI CEI EN 16247-2 requires a named professional, a measurement plan and a site visit, and remains the reference when a formal document is needed. What a physics-based simulation of the building adds before that step is the ability to compare several scenarios against the same model — lower flow temperature, improved envelope, different generator — and to see which combination holds up.

The choice of heat pump is, at that point, the last variable to fix. And it is the one the market spends almost all of its conversation on.