Heat Loss - Pen & Paper Check

It’s amazing how often radiators seem to come in a bazillion varieties but never exactly what you need. Stelrad tripples seem to come in 300, 500 and 600 high, but not 400.

Does anyone else make them?

Edit: how would 500 work out for you?
 
Edit: how would 500 work out for you?
I think 450(h) is about my max -
that gives me 25mm (to cill) wiggle room to mount rad on brackets
and 125mm underneath for air convection (whereas Stelrad recommend 150mm)

D
 
New boiler plumbed with two zones:
- Z1 17sqm UFH installed and running now.
- Z2 x9 rads; flow/return not yet connected - radiators being replaced.

By modern standards; house insulation isn’t great (parts are >200yrs).
Mix of solid walls and cavity (of sorts) walls, single glazing, sketchy loft insulation, gaps, drafts……….
I can make some improvements over time but fundamentally - I can’t see me ever looking at heat pumps.
Which is kinda why I was happy to work on T50 specs.
I wouldn’t write off running lower flow temps/ashp in the future, and overcompensating now with respect to radiator sizing is definitely a good strategy.

If it helps our house has very poor insulation in places. I remember experimenting with our old boiler where we would be running setbacks of 15/16 overnight and when we were out, and I couldn't get away with outlet flow temps less than mid sixties. The house just wouldn't heat back up to 19/20 when it was very cold outside.

Ashp is now constantly running mid 30s and I've had to dial my weather compensation curve down to run at outlet temps of 46 at -3, as the house was getting too warm when the outdoor temp was dropping. The difference is all in the weather comp approach and running 24/7 with no [or very limited] set backs.
 
Thanks reply and pointer; I have been looking at these.

Eg….,window cill at 600mm.
The two outgoing rads were 300h
k3 at 300h x 1000 and 2000 are 1349W and 2698W respectively
I’d need one of each (2x 2000mm would block off the 13A mains socket)
That gives me 4,047W - I’m looking for 5,500W so I’d need a third rad.
Stelrad don’t do a K3 in a 450mm.
I’m also concerned about wall fixing K3s.
Walls are very old - I’m unable to add feet.
The practicalities in an old house are very real.

Set in context of ‘designed to operate at the lowest possible temperature while still meeting heating needs of the dwelling’.
Albeit not ideal, the DT40 sized rads are looking more viable.

D
5.5kW for one room is crazy, my whole 4 bed house is only a bit more than that! You could go with the DT40 radiators and run a suitable flow temp as required. Then fix the building insulation issues, especially the loft insulation as it’s a quick and cheap fix, then gradually lower the flow temp.
 
5.5kW for one room is crazy,
I may have confused things there - I was meaning my DT50 equivalent figure is 5.5kW.
Ie I’ve taken my room heat loss of 2.8kW and used a DT30 correction factor (equiv 5.5kW in the ‘standard tables’).

That said, I’ve attached my workings out for that room - please let me know if you spot any clangers.

D
IMG_6108.jpeg
 
You've basically got the method right: room by room fabric plus ventilation loss, plus your 10% for heat-up, then pick the rad that covers that loss at the mean water temp you actually plan to run, not the catalogue headline.

One thing on the 1957 wall: bridging from debris, ties and mortar makes a cavity wall lose more heat, not less, so it pushes U up. An unfilled cavity of that age is usually nearer 1.5 than 1.0 (RdSAP uses about 1.5), and since yours sounds a bit sketchy I'd take 1.5 to 1.6 to stay on the safe side. On the chimney breast endecotp is right, with similar setpoints either side the ΔT is small so its U-value hardly moves the total.

The DT bit is where the online tools disagree. Most catalogues quote at ΔT50 (75/65 flow/return, 20 room). Output scales roughly with ΔT to the power 1.3, so at ΔT40 (say 70/50, mean 60) you get about 75% of the ΔT50 figure, and at ΔT30 about half. If wall space only allows the ΔT40 size, that's a reasonable compromise on a condensing combi. The return sits around 50 °C so it should still condense, more so if you drop the flow temp when it's milder. Part L asks for emitters sized for 55 °C flow or lower when a wet system is installed or fully replaced, and where that isn't feasible it allows the lowest practicable temperature, which in an old bungalow with limited wall space is often the honest answer.
 
It was a year ago, so one would hope sorted by now, but to me the big problem is, if I read the temperature in this room it varies, with a fan assisted radiator it may be that the room is at 22°C for the whole room, but using non fan assisted it will vary.

Hand on radiator behind me, and its cold, so Kasa TRV 18.7°C the Energenie TRV 19°C, the wall thermostat 20°C and the thermometer on my dinning table is 19.3°C the curtains are closed on the windows, and the heating has been running most of the day, so furniture has warmed up.

For some reason the radiators are placed on outside walls, under windows, so the losses are higher.

The two radiators are at 90° to each other, so anyone's guess as to air movement. One wall has large billy book case units against it, so is better insulated, as very little air movement behind them, floor carpeted with a void under the floor boards, but then a Rockwell wool insulated area before flat below, above again a void, with two lightly heated rooms above it.

The important factor is recovery time, as room allowed to cool when we go to bed, so the most important thing is room temperature at 9 am and how long it will take for the room to recover.

But also we have a 20 kW boiler, so how long can it run before the return water temperature turns off the boiler?

So since I have heating already, I need to log the outside temperature, and how long the boiler runs for, so if total in 24 hours it runs for 2 hours, and outside temperature is 10 degrees, then I need 40 kWh to raise the home by 10 degrees, so from zero degrees I will need 80 kWh add a bit for very cold days and varying temperature so to maintain temperature likely need 120 kWh, so boiler needs to be 120/24 = 5 kW to maintain the temperature.

But that has not answered the question, if I want the room at 22°C by 9 am, and radiators are 1.5 kW each, when will they need turning on?
 
But that has not answered the question, if I want the room at 22°C by 9 am, and radiators are 1.5 kW each, when will they need turning on?
You are trying to be far too precise. I want the house comfortable, so I turn it on, say, when I wake up and it takes an hour or so. If it isn't then I turn it on a bit earlier the next day, if it heats up too quickly then I turn it on a bit later.

Comfort is the key, forget trying to be pin point accurate with temperatures.
 
That's a slightly different question from Diggory's heat loss check, but a useful one.

You can't work out the switch-on time from radiator kW alone. It depends how far the room drops overnight and how much heat the walls and furniture take back on the way up. Your two 1.5 kW radiators are probably rated at ΔT50, so they give about 3 kW when the water is that hot, and less at a lower flow temperature. If one is cold to the touch, check why first: the TRV may have shut it, or the wall stat at 20 °C may have turned the boiler off.

A simple way to time it: one morning, note the room temperature when the heating comes on, the outdoor temperature, and how many minutes it then takes to reach 22 °C. The next similar morning, start that many minutes before 9 am. If it is colder outside, allow a bit longer. After a few days you will be close enough. If your Kasa heads can run a schedule, set them to that earlier time. Some programmers also have an optimum start or learn setting that does the same thing automatically.

The boiler run-hours idea is good for a rough whole-house heat loss (daily kWh divided by the indoor-to-outdoor difference), but it does not tell you when a single room will recover. It also assumes the boiler ran at its full 20 kW for those hours, and a modulating boiler usually doesn't, so gas meter readings give you a better number.

If you want the room at 22 °C by 9 am, it is usually better to start a bit earlier at a lower flow temperature than to fire short bursts of hotter water. The house needs the same heat either way, but a condensing boiler runs more efficiently with a cooler return. Once the return goes above about 55 °C it mostly stops condensing and you lose several percent. So it pays to know the lowest flow you can still get away with, which means comparing each room's heat loss with what its radiators give at that water temperature. I ran that for my own house here:
https://heatlosspro.cloud/s/YAzppFumRwF3TzRGRBaV0g
It took under 15 minutes, and it showed the heat loss in each room plus the lowest flow temperature I could use to keep any given room at the temperature I wanted. I didn't have this app back then, and I undersized the heat pump. The calculation shows exactly that
 
One extra thought on the 9 am warm-up. If you currently let the room go quite cold overnight and then heat it hard with hotter water, weather compensation and the lowest flow temperature you can get away with might use less gas and give you steadier comfort. It is also simpler: the curve sets the flow from the outdoor temperature, so you are not juggling on/off timers and a hot morning blast. Set the wall stat and the TRVs a little above target so they act as limits rather than switching the boiler on and off.

When I first bought my place I still had a gas boiler, and I set it exactly like that. No night drop, just weather compensation and the lowest flow I could get away with. That took a lot of testing, dropping the temperature a bit at a time, because I didn't have a tool for it then. My consumption was cut in half, and it has never been as comfortable everywhere in the house!
 
In my mothers house, at 7 am would set to 24°C then at 8 am back to 22°C as the anti-hysteresis was far too good, and if I set to 22°C to start with it would be 11 am before at 22°C, but within the hour it had reached 22°C when set to 24°C this was with Energenie TRV heads. But she had a really over sized boiler for the house 35 kW and it modulated, this house twice the size, and the oil boiler does not modulate, so all learnt with mothers house, does not work here.

Main problem was the hall cools too slowly, so the Nest Gen 3 in the hall was not turning the boiler on when required, had to put a Wiser hub in parallel to it, so now three thermostats, two wall and one TRV head type which can turn on boiler.

I tried measuring the oil in the tank, expansion can change height as much as a days use, so that does not work, would need to connect a clock to boiler to work out run time, and that seems too much faffing around to work out run time.

Wiser and Nest have some learning features, but these rely on doors always being closed or open, so when the doors vary as to closed or open, the learning features get it wrong. So all turned off.

Gone to great lengths to only heat as an when required, with 10 programmable TRV heads, but the boiler was designed well before the heads, so the two don't match as they should, the pump is pumping water around the system, but that water temperature varies depending on when boiler last fired.

And of course on how many TRV heads are open, today I checked my bedroom to find batteries discharged, which means that TRV head would remain wide open. Need another 14 AA cells to do whole house, so heating using more than it should.
 
Last edited:
That makes sense. The 24 °C then 22 °C trick worked there because the boiler could modulate. Here the oil boiler is either on or off, so the same boost just runs it at full output and does not transfer.

The hall sounds like the bigger issue. If it holds its heat, a Nest there will not call when the rooms you actually want warm are still cold. Letting a Wiser wall stat or TRV fire the boiler as well is the right workaround. Whichever room you most want warm in the morning should be one of the ones that can call for heat.

Weather compensation and no night drop can still give steadier comfort, but I would not expect the same consumption drop I saw on gas. A non-modulating oil boiler cannot turn down, so if it is quite a bit bigger than the house needs it will cycle. If you can set a lower flow temperature, that is still worth doing for more even rooms, more than for a big fuel saving.
 
Whichever room you most want warm in the morning should be one of the ones that can call for heat.
Why not all rooms being able to fire the boiler? In my case cost is the reason, the radiators are on outside walls, yes I know that's an old fashion idea, but that's the way it is, so each room would need both a smart TRV head and a wall thermostat linked, so a cold wall does not activate the boiler, it has to be a cold room.

But then we have a secondary problem, a boiler firing up just to heat one room.

I can accept a room at 18 - 22°C, for intermittent use, so a non linked programmable TRV head set to 16°C overnight and 22°C during the day for example the dinning room, is acceptable, for the living room want it 20 - 22°C and the dinning room radiator is far bigger in relation to room size to living room, so the TRV head does not need to be linked in the dinning room.

This can be repeated throughout the home, my bedroom only has one window, wife's has two, her room North facing mine South, so both rooms have programmable TRV heads, but only her room needs to be linked, so if not linked it can be set to 18°C at night, 16°C in the day, but do that with a linked TRV head, and it would fire the boiler because the wall is cold, not the room, so it needs a wall thermostat on an internal wall or free standing.

But the times set are important, if the dining room set 10 minutes before living room, then living room will open before the boiler has shut down, but set to 20 minutes, boiler will shut down and likely at least ½ hour delay before it cools enough to re-fire. So setting a sequence so rooms are heated in a set order can both reduce and increase re-heating times, depending on delay set.

On moving here, what became apparent, is there is no one system that suits all, what had worked in mothers house, did not work in this house, so had to work it out from scratch again.

Re-heat times do seem to be slow, radiators are hot, so only way to speed up would be fan assisted radiators, but then can't use TRVs, so there is no magic wand fix.
 
Why not all rooms being able to fire the boiler? In my case cost is the reason, the radiators are on outside walls, yes I know that's an old fashion idea, but that's the way it is, so each room would need both a smart TRV head and a wall thermostat linked, so a cold wall does not activate the boiler, it has to be a cold room.

I would try the opposite approach, leave the TRV heads fully open and properly balance the radiators, rather than having each room opening and closing on a schedule. Aim for the lowest flow temperature that keeps the rooms comfortable, within what your oil boiler allows. That way the rooms aren't constantly switching in and out, although the boiler itself will still cycle.

That's what I did alongside the weather compensation and no night drop I mentioned earlier. My usage literally halved. I probably inherited a really badly tuned system from the previous owners, but I honestly didn't expect anything like that much difference.

That was with my gas boiler, so I'm not saying yours would halve, especially when you're deliberately leaving some rooms cooler. But I'd try balancing with the heads open before spending more on controls. You might get steadier comfort and lower running costs too.
 

If you need to find a tradesperson to get your job done, please try our local search below, or if you are doing it yourself you can find suppliers local to you.

Select the supplier or trade you require, enter your location to begin your search.


Are you a trade or supplier? You can create your listing free at DIYnot Local

 
Back
Top