Previously:
The design works.
After a full heating season, the system delivered a seasonal COP of 4.21, used 68% less energy than the LPG boiler it replaced, ran at 46% lower cost, and held the house at setpoint through every observed outdoor condition. The 12 kW Vaillant aroTHERM Plus, sized on paper at 11.94 kW, met its design demand on the coldest day of the year. Indoor temperature stayed comfortably above 19°C throughout. By every measurable functional standard, the design is doing what it was designed to do.
That part of the story is closed.
But the design isn’t perfect. A year of operation has shown me where the gap is. And the gap is worth talking about — not as a confession, but as the next iteration.
What experience would have caught at design time. There is a real difference between an experienced professional and a first time amateur, and it is not a question of effort or care.
An experienced professional designs many systems. An amateur designs one. The professional brings calibration the amateur can’t yet have — a sense for which conservatisms cost something and which ones don’t. The first iteration of an amateur’s design is, almost by definition, more conservative than a professional’s would have been, because the amateur doesn’t yet know where conservatism is free and where it has a price tag.
In my case, the conservatism or the amateurism that had a price tag was the design flow temperature.
The system was sized for 50°C flow at the −2.7°C design point. That number was set by the existing radiators — most of which are K2 panels, kept in place because replacing every radiator in the house was off the table at design time. The original installers, the ones who wanted to quote a 37 kW heat pump, had also wanted to replace many radiators in the house. Rip out the lot, fit triples, start again. I had pushed back on that by instinct not knowledge — the radiators were fine, I said, the system would work with what was there — and broadly I was right. The radiators do work for the designed flow temperature. The house is warm.
But broadly right is a specific kind of trap. It’s the trap of being correct on the headline question and wrong on the subtler one underneath it. The headline question was do the radiators need replacing? The answer was no. The subtler question was do all of them carry their share at the same flow temperature? And the answer to that one was no, but I didn’t know to ask it.
A professional would have known that 50°C design flow was higher than it needed to be, and that leaving the radiators alone wasn’t a free choice. it would cost efficiency in the regime that carries most of the season’s energy — getting the system to a lower flow temperature on day one.
I didn’t. Not for lack of attention. The subtlety wasn’t in my vocabulary at the time.
What the data showed. A year of operation made the gap visible.
Across December, January and February, the system ran at an average outdoor temperature of 4.9°C, an average flow temperature of 39.3°C, and a heating sCOP of 4.03. That’s the meat of the heating season. That’s where the kWh actually go.
On the coldest day, the outdoor temperature dropped to −1.6°C through the daytime hours and the flow temperature peaked at 44°C. Indoor temperature held at 19.5°C. Then, briefly, in the small hours of that morning, the outdoor temperature reached −2.7°C — the design point itself, the temperature for which the whole system had been sized. The flow rose to 46°C. Indoor sat at 19.8°C.
Six degrees below the 50°C design ceiling. Even at the worst moment of the worst day of the year. The radiators were not, in fact, working at their thermal limit.
That headroom is where the inefficiency sits.
Not on the cold days — those are rare. The inefficiency sits in the long shoulder weeks of December to February, the stretches at outdoor temperatures between 0°C and 8°C, where the heat curve is forced higher than it needs to be because the design ceiling was set high. January is the highest-energy month and the most expensive. It is also the month in which the gap between operational reality and design ceiling does the most damage.
What the change is. The fix is small.
Two large K2 radiators in the main living space are going up to a K3 format. The rest stays. Every radiator upstairs stays. They have margin already at the current flow regime; replacing them would buy nothing.
The target is to reduce the design flow temperature from 50°C to roughly 40°C at the −2.7°C outdoor design point. The radiator change is the lever. The heat curve will then re-tune itself, both in the coarse layer and the fine trim, around the new operating envelope.
This is not a system overhaul. It is two radiators in two rooms, picked because the data identified them.
What it will buy us, in numbers. I don’t know.
That is the honest answer, and it is the answer this chapter is built around.
A professional, designing from experience, would commit to a predicted gain. An amateur with monitoring will measure. The exact figure is the kind of thing a heating season of post-change data will settle better than any prediction.
What I can say is the shape of it. The lever is flow temperature. The regime where the lever matters most is the 0–8°C outdoor band. That regime carries the bulk of January’s energy. Lowering flow temperature in that band lifts COP. The COP improvement, applied across the kWh that actually flow through the band, produces a reduction in electricity used and money spent.
The shape of the gain is clear. The magnitude is what the data will tell us.
I would rather not predict a number that turns out to be wrong. I would rather measure a number that turns out to be right.
The position. The design met the requirements. It didn’t meet them perfectly.
A professional, with experience, would have caught at the start what a year of data caught at the end. The amateur route is slower. It is also, given good monitoring, more precise — because the second iteration is informed by observation that the professional would have had to estimate.
Two radiators are going to change. The flow temperature ceiling will come down. The COP will improve in the regime where most of the season’s energy is consumed. The data will settle the question of by how much.
The design works. The design has room to be better. The data shows where. The discipline is patience.
you might enjoy 
.
.