JamesPa

If you have an EV then the EON Next Drive (or Next Drive Smart) is hard to beat. 6hrs (was 7) of cheap electricity (7.5p) and not much of an uplift daytime. If you dont have an EV and have a Vaillant heatpump then EDF heat pump add on at 15p for the consumption of the heat pump only is worth considering.

Love that statement, I think the discussions on here and elsewhere bear it out. Of course there are many exceptions, people who take the trouble to learn/re-learn the necessary physics, think about the job, question what they are being told, seek to understand, and continually to improve their understanding. The rest are just painting by numbers guys (some of whom cant even read the numbers) bluffing their way through.

I bet your consumption will be going down today, and even more tomorrow. See my post above which was actually deadly serious despite the light tone I used to lift the discussion, and merely reiterates what others have said.

Yes, but currently you have air at 20C in contact with the slab not pipework at 35C, makes a big difference. Your choice of course!

Obviously there is quite a bit dependent on construction detail (something builders are notoriously sloppy about!), but if the pipes are in good contact with 'grout' ('grouted in') then that wouldn't be much different to being buried in concrete. Only if there is a void underneath (ie insulation!) can I see it making much difference. On the strength of this I wouldn't do it personally, nor would I dig up the floor because, as you say, it costs a lot. Id fit radiators!

Its nothing to do with 'your' fridge, its to do with the fridge outside which was delivered by the almighty (other explanations for the natural world are available) around the same time. Its warming up at the weekend and consumption will drop back down again. However you heat your house its going to consume a lot more when its -1 than it will most of the time, typically twice as much! With an ASHP the relationship is a bit exaggerated, but it evens out over the season. Did you watch your consumption daily in your previous house/with your previous heating, I'm guessing not.

Duck.ai came up with ~50:50 room/soil with a slightly different methodology. For that I has to specify a room 7m square (ie a house), but otherwise the same assumptions. I also gave it 10m square and it came up with a similar ratio. I am not sure which underlying model Duck.ai uses so this may not be 'independent'. Im now definitely not buying the handwaving argument that most of the heat goes up though. I would say the jury is out, do this at your own risk!

@JohnMo building on your idea I asked chatgpt a more specific question. I have quickly read through and it seems plausible, I might ask another AI tool the same question. If its right then UFH on uninsulated slab is mad, just put radiators in! However I dont know if its right. It is however prima facie more convincing than the handwaving arguments made by UrbanPlumber, but it could of course be BS which I have certainly experienced with ChatGPT in the past. Q: Suppose I have underfloor heating with pipes at 150mm centres buried 15mm deep into concrete that is 100mm thick, sitting directly on damp soil assumed temp at sufficient depth 10C. Suppose that on top of the concrete is say 5mm of luxury vinyl tile and the room above is at 20C. How much heat goes down and how much goes into the room? You may assume, for simplicity, that the heating continues for several months, that the room is infinite is size and that the pipes are supplied with water at 35C A: Nice — I’ll do a simple heat-resistance (steady / long-run) calculation and give numbers, plus a short sensitivity check and practical notes. I’ll keep the math short and show the assumptions so you can change any number if you want. Model & assumptions Pipes run at 35°C (we take the water/pipe as an isothermal heat source at 35°C). Room air temperature above floor 20°C. Soil at “sufficient depth” fixed 10°C (an infinite sink). Geometry: slab total 100 mm thick; pipe level 15 mm below top surface → 15 mm to top, 85 mm to bottom. Pipe spacing 150 mm (0.15 m) centre-to-centre (so each metre of pipe serves 0.15 m² of floor area). Vinyl tile 5 mm thick on top. Material properties (typical values): concrete thermal conductivity k = 1.7 W·m⁻¹·K⁻¹, vinyl k ≈ 0.25 W·m⁻¹·K⁻¹. Convective (and radiative) heat-transfer coefficient at the floor surface h ≈ 8 W·m⁻²·K⁻¹ (natural convection + radiation typical value). Slab sits directly on damp soil; I add a small contact/soil surface resistance of 0.01 m²K/W (per m²) to account for imperfect contact and soil conduction. Steady / long duration (months) so temperatures and fluxes are taken as time-steady. This is intentionally simple (1-D vertical resistances from the pipe location to the top surface and to the bottom surface). The main limitation: real piping is a line source, soil is semi-infinite and will warm over time, and pipe internal resistance / pipe outer surface temp would change the numbers — see caveats below. Calculation (concept) Treat upward and downward paths as thermal resistances per metre of pipe (area associated with one metre of pipe = spacing × 1 m = 0.15 m²). Upward path (pipe → concrete 15 mm → vinyl 5 mm → surface → convective to room): [ R_{up} = \frac{t_{conc,top}}{k_{conc}A} + \frac{t_{vinyl}}{k_{vinyl}A} + \frac{1}{hA} ] Downward path (pipe → concrete 85 mm → soil sink): [ R_{down} = \frac{t_{conc,bottom}}{k_{conc}A} + R_{soil_contact_per_m^2}/A ] Heat flows per metre of pipe: [ Q_{up}=\frac{T_{pipe}-T_{room}}{R_{up}},\qquad Q_{down}=\frac{T_{pipe}-T_{soil}}{R_{down}}. ] Results (base assumptions above) Per metre of pipe (spacing 0.15 m): Upward heat to room: (Q_{up} \approx 14.6\ \text{W per m of pipe}) Downward heat to soil: (Q_{down} \approx 62.5\ \text{W per m of pipe}) Total supplied to ground+room: ( \approx 77.1\ \text{W per m of pipe}) Split by percentage: Upwards to the room ≈ 19% Downwards into the soil ≈ 81% Per square metre of floor (divide the “per metre of pipe” numbers by spacing 0.15 m): Upwards to room ≈ 97.5 W·m⁻² Downwards to soil ≈ 417 W·m⁻² Total ≈ 514 W·m⁻² (Those per-m² numbers look large because the model assumes the pipe is at 35°C and the slab/soil are cold — it’s a purely conductive picture with a warm source close to the bottom. The per-m values and percentages are the most useful result for comparing up/down split.) Sensitivity (short) A few changes strongly affect the split: Lower convective coefficient h at the top (poor surface heat transfer) → less heat goes up (down fraction increases). Higher soil contact resistance or insulation under slab → more heat goes up. Example: if soil contact resistance increases to 0.05 m²K/W (larger insulation effect), downward heat falls and the upward fraction increases substantially. Closer pipe spacing (more pipe per m²) → increases total heat available per m² (but the % split depends on resistances). Higher concrete conductivity → more symmetric flow; with high k more heat reaches soil faster (but details depend on thicknesses). I ran a few numeric variants (keeping other parameters same): Lower h = 5 W/m²K → upward fraction drops (less to room). Raise contact resistance (soil R = 0.05 m²K/W) → Q_down falls from 62.5 W/m to ~37.5 W/m and upward share increases (about 28% up instead of 19%). Tighter spacing (100 mm) reduces per-pipe Q but increases per-m²; percentage changes modest if material resistances unchanged. Practical interpretation & caveats Most of the heat is lost downward with the geometry you gave (pipe fairly close to slab bottom relative to top temperature difference). The key reason: although the temperature difference to the room is smaller (35→20 = 15 K) than to the soil (35→10 = 25 K), the conduction resistance downward (85 mm concrete) is lower per unit area than the whole upward path (thin concrete + tile + convective/radiative exchange), when evaluated per pipe area. The geometry and the fact the pipe is closer to the top or bottom can swing the split strongly. A real system normally includes insulation under the slab precisely to stop the large downward losses. If you add even modest insulation under the slab, the downward path resistance rises fast and most of the heat will go up into the room. In the sensitivity run, increasing bottom contact resistance from 0.01 to 0.05 m²K/W cut downward heat by ≈40% and boosted the upward share. Transient soil warming. Over months the soil immediately beneath the slab will warm and the downward heat flux will reduce compared with the initial value. I assumed an infinite 10°C sink (worst case continuous draw) so the numbers are conservative for downward loss. If the heating has been on for a long time, the steady downward flux will be smaller (because the soil warms). Pipe/pipe wall & internal convective resistance were ignored (we treated pipe water as 35°C source). In reality pipe outer temperature will be slightly below water temp; that would slightly reduce flux numbers but not qualitatively change the split unless the pipe/fluid resistance is large. Line source & lateral spreading. A rigorous solution would model the cylindrical pipe as line source in a slab and account for lateral spreading and the semi-infinite soil transient. The simple 1-D resistive model is fine for quick engineering estimates and to show how to reduce losses. Quick takeaways / recommendations With the numbers you gave roughly 80% of the energy goes downward and only about 20% into the room (under the simple assumptions above). If that is undesired, add insulation under the slab (or a thermal break) — even modest insulation under the slab will shift heat upwards dramatically. Reducing spacing (more pipe per m²) raises the heat available to the room for a given supply temperature; shifting pipe closer to the top surface also helps. If you want, I can: rerun the calculation with different soil contact resistance (insulation thickness/k), different h, or different spacing; compute the required underside insulation (thickness/k) to get e.g. 70% of heat upward; or do a more realistic line-source slab transient estimate (requires a bit more detail). Which follow-up would you like? (I can re-compute numbers right away with any changed assumptions.) .

@JohnMo I didnt think of that and its definitely interesting. I guess the bit that purports to justify the the key statement it makes ie 'Short answer up front — without insulation under the pipes most of the heat goes into the ground rather than up into the room.' is the section 'Example conductive heat-flux numbers (simple resistance estimate)' and the key variable here is L, the depth into the ground before you reach the 10C assumed ground temp. The bit about seasonal variation is also interesting, suggesting it takes months to get to a depth of metres. Another thought - when we bury flow/return pipes we worry a lot about insulation, surely somewhat the same logic applies to UFH!

I have never understood how Cosy is a good fit to a heat pump unless you also have a battery. Sure you get cheap leccy for a few hours per day, but at relatively poor COP because you are trying to get the same amount of energy into the UFH in one third of the time, so have to jack up the flow temperature (49C for UFH apparently, I'm running at only 43 on radiators). Obviously you also need a heat pump which is a factor of 3 oversized. However, that notwithstanding (almost) everything above, is also true. Its very cold at present so consumptio9n is at its max, (as it would be with any heating but who monitors gas or oil daily?) If you have only just switched it on you have got a lot of making up to do. It might be best to just monitor it for a few weeks and see how its averaging. Also consider whether cosy plus jacked up flow temp is more cost effective than a different tariff plus running the UFH at a more typical 30-35C flow temp. What COP is your heat pump reporting?

Thanks everyone, great discussion. Lets leave aside the financial for a moment (I accept its a sound argument but its another dimension which different people will view differently). Lets just consider the physics and what the urban plumber said. He said that (in round numbers), for a room at 20C, its DT 10 to the floor (because the soil is at say 10C), and DT20 to the room (because the outside air is at 0C). But for the walls and ceiling its the internal air that is in contact with the cold elements whereas in a floor with UFH its heated pipework that is in contact with the cold elements. So it may be DT from room to soil but its DT20-25 from pipework to soil, and that's what matters because that is where the heat is lost from. He also said he has no evidence to support his assertions and 'gets good scop'. Well yes, you can get good scop with a house that consumes 10kW or with a house that consumes 3kW, the SCOP is not dependent on the loss, its dependent on the flow temperature. Im not saying he is wrong, but he admits he has no evidence to prove it and some of his arguments do appear to be wrong. There has to be a thermal gradient from the pipe temperature of 30-35 to the soil temperature of 10, and heat will continue to fall down that gradient. So I accept its a thermal buffer but it is a lossy buffer, same as a river, the river level stays constant but water still pours out of the estuary and has to be replenished at the source. Not convinced, I wonder if anyone has actually modelled this?

This is just mad right? Or is there some part of the physics I don't yet understand which means that its somehow sensible? Just checking, I don't have it nor do I plan to, but I have been asked and want to check whether my immediate reaction is right or not.

Its very cold at present, this is the time it will be at its most expensive (roughly double what you might normally see most of the season), Dont forget to factor that in! With a cost of £6 presumably about 24kWh? - difficult to tell as you are apparently on agile (do you have a battery? If not agile might not be the best tarrif) For comparison Im currently consuming about 28kWh/day in a house with a loss of 7kW @ -2 and on the few days when its consistently sub zero I can consume as much as 50kW. The heating is still cheaper and more comfortable than the gas boiler it replaced was, so Im not complaining! You seem to have your WC set to 49@-2 and 39@20. Does it need to be that high (answer no, certainly at the high OAT end)? Are you operating 24x7 with WC adjusted and zones balanced, thermostats/trvs set to max so they have no effect or alternatively 2C above desired so they act as limiters? Tell us a bit more and someone may be able to give some specific suggestions.

Now you are expecting people to take responsibility for their own actions. How dare you, don't you realise that everything is the fault of the Government/the Council, the EU/immigrants/whatever other target for hate somebody chooses to invent. If course there definitely are people who can afford neither home improvements nor 'expensive holidays and weddings'. Assuming that they are trying to improve things for themselves then they deserve help with improvements to make their houses more efficient IMHO. Here are some stats from ABTA. Nearly 60% of people in the UK (according to ABTA) take one or more foreign holidays in any given year (allowing for Covid recovery).

I run the numbers for batteries at least every 3 months and can never quite get it to work out after accounting for cost of capital (even if I allow for tax on savings). I want them to and its close, but they don't for me. A decent nighttime tarrif, doing washing and charging ev overnight, and running the heat pump 24x7 without setback still works out the best bet. Drop the price by 20% and it changes, but the installers don't drop the prices. Even factoring in arbitrage doesn't tip the balance.