Your Home Is Not the Average Home

Distribution of modelled annual heating hours across 107,336 Nottingham homes: a two-humped spread from 557 to 5,868 hours, with the 3,592-hour average marked on the shoulder of the first cluster, the most common homes clustered near 4,900 hours, and the old 1,800-hour assumption far to the left.

Almost everything you are told about your home's energy comes as a single number. The certificate gives it a band. The heat pump quote gives it a running cost. The retrofit estimate gives it a saving. Each of those turns your home's details, plus a set of standard assumptions, into one figure. They are useful for comparing homes quickly, and they are not the same thing as what your own house will do. Your bill, when it lands, is the gap between the single figure and the particular house, and that gap has a shape worth seeing.

One number, or a distribution

Here is the idea made concrete. When we built a small pipework calculator, we assumed a home's heating runs about 1,800 hours a year, the sort of single figure that sits inside a calculation and yields a tidy answer. Then we simulated 107,336 homes through the government's national calculation method, hour by hour, and asked each one how long its heating was modelled to run. The median is not 1,800. It is close to 3,600. One home in ten is modelled to run past 5,175 hours, one in ten under 2,474, and across the whole set the curve stretches from a few hundred hours to nearly six thousand. The chart above is that spread, and the 1,800 mark sits far to the left of almost all of it.

The point is not only that the round number is low, though for the median home it is. It is that a single number cannot stand in for the curve. There is a wide distribution, and every home is one point on it. An average drawn across it lands in the middle, where fewer homes actually sit than you might think, and the further your home is from that middle the less the average has to say about your bill.

The single average of 3,592 hours shown as one point, joined by a dropline to its thin slice of the full modelled distribution of 107,336 homes below. Most homes sit far from the average.

Why the same house is not the same house

Two homes that look the same on paper, same age, same size, same certificate band, can sit a long way apart on this curve. A band is a wide bucket, and inside it the details still differ: the loft that was topped up and the one that was not, a single draughty bay, a corner plot against a sheltered mid terrace. In this simulation those recorded differences are what move a home along the curve. Real life adds more on top, because how a home is lived in and how warm it is kept vary from household to household in ways no model here observed. Either way, none of it shows in a band, and all of it lands on the bill.

What to do with that

The averages are not useless. They are how a whole country is sorted quickly, and they are fine for that. They are simply the wrong tool for a decision about one home, because the thing they set aside is the thing you are deciding about: where your home sits on the curve. Before spending real money on a heat pump, a refit or insulation, the question worth answering is not what the average home saves. It is what yours does, and that can be a very different number.

Where this leaves you

For a single home, the useful first step is a reading of your own, built from what is already recorded about your house rather than an average laid over it. For anyone looking after many homes, the same truth scales into a harder and more valuable question: which of your homes are the outliers, and where. That is a reading our record can produce for any stock in England, home by home, from public records, before anyone visits.

About the numbers

107,336 homes across Nottingham, each simulated through the national calculation method, full year, hourly, from public energy certificates. Active heating hours are the hours a home is modelled to be calling for space heat.
Distribution of modelled active heating hours: minimum 557, 10th percentile 2,474, median 3,592, 90th percentile 5,175, maximum 5,868. The 1,800 hour round number sits below all but about one home in forty.
The spread illustrates the principle; it is not a claim about any single home. Weather and occupancy schedules are the method's stated conventions, so the spread shown comes from recorded property characteristics; your own year and how you run your home would move your point further.
Aggregates only. No figure for any single home leaves the analysis.