A Robust Non-Gas Heating / Water System for Rentals

[Ferdinand]

Venturing gently into the Boffins' Corner, I am after some help to try and come up with a heating / water / ventilation system for rentals which will:

 

1 - Not involve gas (trying to get rid of Annual Checks, and the plain worry of gas appliances going wrong).

2 - Will not slug the EPC figure. I am regulated to be D or better from 2025, and C from 2030, and not to comply will be a criminal offence - though there are exceptions.

3 - Will not be overtly expensive. 

4 - WIll be essentially maintenance free, which ideally means minimal self-done maintenance once per year, and professional assistance once every 5 years.

 

This will be a slow-burn, but as a starting case study, I want to work on the real world renovation of the Little Brown Bungalow. This has been easy to renovate so is a good example for a first try - and there another 49 identical properties in the street to do next time. I may get a copy of the  Stoma data from my EPC an, but the project characteristics are:

 

a - Floor area is 64sqm. 3 beds, reception, kitchen, bathroom. Suspended wodoen floor over a concrete raft foundation.

b - EPC will be 76 or 77 in the C band. Previously was 47E.

c - I have addressed all the low hanging fruit and a bit more. Underfloor insulation 100mm rockwool + 25mm PIR.  27mm loft inculation. Good double glazing. Modern boiler (Ideal Vogue) + rads. 50mm cavity wall insulation. Background ventilation is a PIV fan at the far end of the hall, one end and an HR Trickle Fan in the kitchen. 

d - Attention has been paid to airtightness and detailing.

e - The obvious weakness is the 0.55-ish U-value of the walls.

f - Realistically energy costs should be around £500 per year give or take, and demand could feasibly be under 10kWh per year - the previous EPC number was 13.4kWh predicted from the EPC.

g - As a marker, the all new heating/water system and the ventilation kit added up to £4k installed.

 

EWI would be prohibitive unless substantial grants are available - looking at 25-30 year payback on the £8-9k I was quoted. If I need to I can do more modelling; I think I probably need to do some proper thermal modelling eg the @JSHarris spreadsheet. If there was a way of using them effectively without complications, PV would be a good thought as it would also benefit the EPC (2.5hWp would put it up to nearly the A band). 

 

My initial questions:

 

How close is this to being practical as a low cost electric only house? 

Does the thermal demand need to be mitigated further? How can that be done? Pump LECA or EPS beads into the subfloor, perhaps? 

What sort of technology mix would be suitable in this rental application? ASHP? Storage Heaters? Sunamp? Economy Seven? MHVR combined with warm air heating?

What to do with any solar panel output?

Are there any creative possibilities?

 

If we can come up with a good ideas, I would be willing to try it out on another identical bungalow in future if one comes up for sale.

 

Cheers

 

Ferdinand

 

Edited December 7, 2017 by Ferdinand

[Jeremy Harris]

All electric is the obvious way to go in terms of reduction of ongoing maintenance and inspection cost, but will hit the EPC quite hard.  One option may be to mimic what the big boys do to get around this problem, and fit small PV systems.  A few panels can earn you a fair number of EPC points, and may well be cheaper than more expensive insulation work.

 

It goes against the grain to suggest this, as clearly better insulation and airtightness would be a more responsible option, but you'r fighting costs here, and I suspect a small PV system may well be cheaper than something like EWI, and earn as many, or more, EPC points.  This would be particularly true if installing systems on several houses, as the chances are that you would get a better deal.

 

I'd just keep any revenue yourself, and let the tenants benefit from the slight reduction in electricity use, as the paperwork to do anything else would be too much hassle, I'm sure.

Edited December 4, 2017 by JSHarris

[A_L]

1 hour ago, Ferdinand said:

 

a - Floor area is 64sqm. 3 beds, reception, kitchen, bathroom. Suspended wodoen floor over a concrete raft foundation.

c - I have addressed all the low hanging fruit and a bit more. Underfloor insulation 100mm rockwool + 25mm PIR.  27mm loft inculation. Good double glazing. Modern boiler (Ideal Vogue) + rads. 50mm cavity wall insulation. Background ventilation is a PIV fan at the far end of the hall, one end and an HR Trickle Fan in the kitchen. 

f - Realistically energy costs should be around £500 per year give or take, and demand could feasibly be under 10kWh per year - the previous EPC number was 13.4kWh predicted from the EPC.

 

a - I currently live in a 68m² version, admittedly with a 20m² extension, so I hope I can say I have some experience.

c - with your existing floor insulation nothing additional here would be economic

f - If that is 10000kWh/yr  for space heating and DHW then £500 energy costs implies around 5p/kWh for energy. I think given current off-peak and daytime electricity rates only an ASHP running at a COP of 3.5 or better could achieve this, with spare P.V. heating DHW to relieve the ASHP of the high temp/low COP end of the DHW.

[Ferdinand]

On 04/12/2017 at 13:21, JSHarris said:

All electric is the obvious way to go in terms of reduction of ongoing maintenance and inspection cost, but will hit the EPC quite hard.

 

Cheers, Jeremy. An ASHP (playing with Stroma) seems to give an increase in the EPC number of a few points - tested by simply replacing the Combi Boiler in the Model with an ASHP. Which is, however, a slightly crude test. 

 

On 04/12/2017 at 16:04, A_L said:

a - I currently live in a 68m² version, admittedly with a 20m² extension, so I hope I can say I have some experience.

c - with your existing floor insulation nothing additional here would be economic

f - If that is 10000kWh/yr  for space heating and DHW then £500 energy costs implies around 5p/kWh for energy. I think given current off-peak and daytime electricity rates only an ASHP running at a COP of 3.5 or better could achieve this, with spare P.V. heating DHW to relieve the ASHP of the high temp/low COP end of the DHW.

 

@A_L

 

a -  thanks for the comment. Keen to hear more of your experience.

 

c - I am not sure that I agree entirely with your "c" reply.

 

My current floor renovations give me a floor U-value of just under 0.25, which I think is the basic requirement for renovated 'flooral' thermal elements (not that I had any formal inspections just a "what if" conversation with the department while considering, and they said 'some improvement is better than nothing'). The wall u-value is 0.55 (50mm filled cavity brick). The roof u-value is about 0.14 (270mm rockwool).

 

The cost-effectiveness or not comes from 2 things ... energy saved and cost of installation , and can be judged-ish on a simple payback. The cost of my whole floor buildup (ignoring the floor finishes and labour as most are self-builders here) was well under £10 per sqm. I will post the detailed buildup next. Even if I include labour, it is probably a 5-7 year payback in energy terms (depending on tariff). (*)

 

I think that perhaps leaves room for a bit more, which would mean either more insulation in the layer-cake or EPS beads pumped into the underfloor cavity. I am already quite close to my doors' 65mm trimming margins at 60mm incl. ventilation gap, so to put more insulation in would need some redesign of the build-up. I could fit 38-40mm of celotex plus Engineered Floor attached directly to battens, or click-laminate/vinyl on 6mm ply on celotex. Just. Putting 40mm in as I planned would have given a u-value of about 0.21, which is still imo not really good enough.

 

I estimate that 300-350mm of EPS beads would give me a u-value of 0.10-0.12, which may be worth it if it put me in a position to use an ASHP not a Gas Boiler.

 

f - Than 10000kWh/yr is all energy. It is a conservative guesstimate as I do not have my EPC back yet with their modelled Guesstimate. The previous EPC (will post) says 13.5k with a potential 50% saving, but includes solar panels as well. 

 

Ferdinand

 

(*) There are issues around transferring part of that saving from the Ts energy bills onto a slightly increased rent - say sharing the benefit 50:50, but at least I am now in a position to see if a dramatically better EPC number can give say £595 /month not £575.

Edited December 6, 2017 by Ferdinand

[Ferdinand]

Floor Buildup for Little Brown Bungalow.

 

U-value = 0.23 (ignoring joists). Thickness above old floor = 53mm.

 

Service cavities run alongside some internal walls.

 

Calculations look slightly off but it is under £10 per sqm.

 

 

Edited December 6, 2017 by Ferdinand

[pdf27]

I'm not sure how much going the extra mile on floor insulation would really help you. Shifting from 0.23 to 0.12 saves:

 

(0.23-0.12) x (64m2) x (30+3 C) = 230W. at roughly the design low temperature condition (-3C for me) and assuming underfloor heating running at 30C for a heat pump. It adds up over a year if you're going for a particular standard like Passivhaus, but if you aren't then the value of additional insulation is really pretty small. I would also be wary of what is available with a ventilated underfloor cavity - the advice we got from our architects was that the ventilation itself makes improving the U-value up to around the 0.1 level almost impossible, but replacing the whole thing with an insulated concrete slab  is horribly expensive, so I think you've done about all that is feasible on a retrofit. You've already discussed external wall insulation and airtightness - realistically EWI should help with airtightness a bit , but I think if you're trying to reduce heat demand then this is the place to go.

 

One thing that might be worthwhile considering if you're chasing SAP points is a shower heat exchanger. That's harder but not impossible in a bungalow (you have to use the drain type rather than the vertical pipe type, which is more expensive and less effective), with something like 50% recovery being feasible. For my current bungalow (3 bed), SAP would probably interpret that as saving 1,000 kWh/year. Fitting it to the only bathroom which would presumably be being ripped out and replaced anyway is probably not too painfully expensive, particularly if it's an alternative to other means of compliance.

[Trw144]

@Ferdinand I have a similar issue on the horizon, taking a recent purchase - a ground floor flat, part of an old building with about 8 other flats. I ve put in new windows, and replaced the old night storage rads with straight electric radiators with time and temp control (not sure they get me any points but much better for a tennant). I ve moved it from 34 to 50, but where do I go from here in the future - it’s a solid floor, and solid walls, and in a conservation area, plus the complexity of all the other owners to contend with.

Edited December 7, 2017 by Trw144

[Ferdinand]

32 minutes ago, Trw144 said:

@Ferdinand I have a similar issue on the horizon, taking a recent purchase - a ground floor flat, part of an old building with about 8 other flats. I ve put in new windows, and replaced the old night storage rads with straight electric radiators with time and temp control (not sure they get me any points but much better for a tennant). I ve moved it from 34 to 50, but where do I go from here in the future - it’s a solid floor, and solid walls, and in a conservation area, plus the complexity of all the other owners to contend with.

 

Over the w/e. I'll try and draw up a list of everything I did, though there are various threads talking about it.

 

It might be worth looking up all the other flats on the EPC Register (https://www.epcregister.com/reportSearchAddressByPostcode.html) for a start for ideas and hints, since that information is published and free. And remember that there will be exceptions to the rules where improvement does not give a 15 year simple payback etc.

 

Lining your outside wall looks like one option that should give benefit, or perhaps an overfloor. That would not disturb neighbours. Or EWI the entire building :-). Then you are into improving internal systems. That shower heat exchanger suggestion sounds good, especially with an electric shower - no moving parts,

 

I would also check the very basics ... eg does your ventilation fan in the bathroom have a backdraft shutter, are you OK for condensation, do you have background ventilation that is adequate now you have well-sealed new windows? I now fit a Lo Carbon Tempra trickle/boost fan in all properties, plus a PIV loft fan (which doesn't apply to you).  None of these will make much difference to the EPC, but they may make for a happy tenant and a lack of niggling problems.

 

One important point is photographic evidence of everything for the EPC man.

 

There is a reason I have never bought a flat :-), though I wish I had found a way to afford the short lease one bed I was offered by my LL in EC2 in 2000 for 70k.

 

Ferdinand

 

Edited December 7, 2017 by Ferdinand

[Ferdinand]

@pdf27

 

Thanks - useful comments.

 

If I take an engineering approximation and say that your 230W peak saving saves an equivalent of 100w over the entire year for ufh (assuming I find a way of doing ufh :-), that would be say £0.1 per unit for 9000 hours = £90 per year. If we call it £30-50 per year as a buffer for my approximations, then that suggests that it might be worth spending £100-£200 for a marginal improvement to a u-value of 0.18 and £300-400 for an improvement to 0.12 (-ish), if extra material can be installed without consequences and extra labour when the floor is up.

 

That would probably leave as realistic options:

 

1 - a little more insulation in the over-floor for 0.18.

2 - or complete removal of the ventilatd cavity via pumped insulation / sealing air bricks, which could also be costed against the labour of lifting the floor, and a chap with a staple gun and rolls of rockwool for a couple of days, and the insulation in the over-floor.

 

Both might be viable options, depending on the circs. I'm not competely heartless, so I am usually willing to do a bit extra if it is the right thing for the sake of a few ££.

 

Unfortunately, EWI will not be viable for me here without at least a 50-70% grant. I had quotes and they were coming in at £8.5-£11k, even with us doing some things. And a small bungalow is the easiest type of building to EWI. In the event I have settled for installing all my 2G with extended windowsills should the EWI be realistic in the future. It is still a payback of many decades, as it would be 130sqm of wall with a u-value reduction of 0.55 to 0.22 (ish).

 

I make the calculation to be money saved per annum by transmission if the heating was on 24*365 approx = 130 sqm * 0.33 u-value reduction * 12 C ave. temperature difference * 0.001 to give kWh * 0.05 £/Kw * 8760 hours / yr = £225, without the air leakage, and ignoring all kinds of subsidiary effects. EPC says the heating bill may be £300 a year, and removing the whole thing would be a 20-30 year payback. Ergo not viable.

 

But I need to improve my heat modelling skills.

 

Ferdinand

 

Edited December 7, 2017 by Ferdinand

[A_L]

On ‎06‎/‎12‎/‎2017 at 07:56, Ferdinand said:

 

c - I am not sure that I agree entirely with your "c" reply.

 

8 hours ago, Ferdinand said:

 

1 - a little more insulation in the over-floor for 0.18.

 

- a reduction of floor U-value from 0.23 to 0.18 will save about 4kWh/yr/m² , a reduction to 0.1 will save about 9kWh/yr/m² (Full SAP model)

 

On ‎06‎/‎12‎/‎2017 at 07:56, Ferdinand said:

I estimate that 300-350mm of EPS beads would give me a u-value of 0.10-0.12, which may be worth it if it put me in a position to use an ASHP not a Gas Boiler.

 

- I know from the winter of 2010 I could heat my house, on a continuous basis, with 6kW heat input.

- A quick estimate using your existing U-values suggests a 7KW boiler size (including hot water), so I think a 10kW ASHP with appropriate heat emitters would do already.

[A_L]

9 hours ago, Trw144 said:

@Ferdinand I have a similar issue on the horizon, taking a recent purchase - a ground floor flat, part of an old building with about 8 other flats. I ve put in new windows, and replaced the old night storage rads with straight electric radiators with time and temp control (not sure they get me any points but much better for a tennant). I ve moved it from 34 to 50, but where do I go from here in the future - it’s a solid floor, and solid walls, and in a conservation area, plus the complexity of all the other owners to contend with.

 

- if this means, as it suggests,  that you have moved from an off-peak tariff to a daytime tariff it will be seriously deleterious to your EPC Band. Probably a two band reduction (or more). Particularly since the DHW heating is now also likely to be daytime.

 

- If someone with storage rads and an off-pk tariff has a single immersion 110litre DHW cylinder one of the most effective steps is to install a 210litre twin immersion cylinder

 

[pdf27]

9 hours ago, Ferdinand said:

@pdf27

 

Thanks - useful comments.

 

If I take an engineering approximation and say that your 230W peak saving saves an equivalent of 100w over the entire year for ufh (assuming I find a way of doing ufh :-), that would be say £0.1 per unit for 9000 hours = £90 per year. If we call it £30-50 per year as a buffer for my approximations, then that suggests that it might be worth spending £100-£200 for a marginal improvement to a u-value of 0.18 and £300-400 for an improvement to 0.12 (-ish), if extra material can be installed without consequences and extra labour when the floor is up.

 

That would probably leave as realistic options:

 

1 - a little more insulation in the over-floor for 0.18.

2 - or complete removal of the ventilatd cavity via pumped insulation / sealing air bricks, which could also be costed against the labour of lifting the floor, and a chap with a staple gun and rolls of rockwool for a couple of days, and the insulation in the over-floor.

 

Both might be viable options, depending on the circs. I'm not competely heartless, so I am usually willing to do a bit extra if it is the right thing for the sake of a few ££.

 

Unfortunately, EWI will not be viable for me here without at least a 50-70% grant. I had quotes and they were coming in at £8.5-£11k, even with us doing some things. And a small bungalow is the easiest type of building to EWI. In the event I have settled for installing all my 2G with extended windowsills should the EWI be realistic in the future. It is still a payback of many decades, as it would be 130sqm of wall with a u-value reduction of 0.55 to 0.22 (ish).

 

I make the calculation to be money saved per annum by transmission if the heating was on 24*365 approx = 130 sqm * 0.33 u-value reduction * 12 C ave. temperature difference * 0.001 to give kWh * 0.05 £/Kw * 8760 hours / yr = £225, without the air leakage, and ignoring all kinds of subsidiary effects. EPC says the heating bill may be £300 a year, and removing the whole thing would be a 20-30 year payback. Ergo not viable.

 

But I need to improve my heat modelling skills.

 

Ferdinand

 

Umm... one thing to be aware of if you stick with radiators is that the temperature difference between the floor and the outside world will drop a lot - with underfloor I've assumed 30C because the floor will be heated to keep the rest of the house warm. If you're using radiators then 18C is more realistic since the floor will be the surface the radiators find hardest to heat. That moves the peak heat loss down to about 150W or so, and more importantly for most of the heating season it's about 50W or so since the difference between floor and outside temperature is so much lower without the floor being used as a heat emitter.

In other words, the £2-300 you've calculated is only valid for underfloor heating - shift to radiators and the value drops to something in the region of £100 - £1-2 per m². That's why I don't think anything extra is worthwhile unless you're chasing certificates - the cash benefit of doing so is pretty small already. That's why I was looking at other things - chasing the last few Watts is always a painful and expensive job.

[Jeremy Harris]

26 minutes ago, pdf27 said:

............... with underfloor I've assumed 30C because the floor will be heated to keep the rest of the house warm.

 

 

 

30 deg C is VERY hot for an under floor heating system, and the heat losses downwards to the cool ground beneath, even with very good insulation, will be high.  You can easily calculate the heat output from a heated floor surface using this formulae: 8.92 * (floor surface temperature - room temperature)^1.1, where the floor surface temperature and the room temperature are in K or deg C and the heat output is in watts. 

 

For example, our house has a heated floor area of 75m² and needs around 1600 W of heat to balance the total heat losses (fabric plus ventilation heat loss), when the room temperature is 21 deg C, the outside air temperature is -10 deg C and the ground temperature is 8 deg C (our ground floor sits on the ground, it's not suspended above it, for a suspended and ventilated floor the losses would be higher, from the colder air allowed underneath).

 

To deliver 1600 W to the whole house, each m² of floor area needs to deliver 1600 / 75 = 21.33 W.  From the above formulae, with a room temperature of 21 deg C, the floor surface needs to be at a temperature of about 23.21 deg C.  That's very much a worse case, in extremely cold weather, with no incidental heat gain from occupants or appliances, etc.  Taking a more normal winter heating requirement, of around 500 W, means that the floor surface only needs to be at 21.91 deg C.

 

 

[pdf27]

This isn't for anything like a Passivhaus though - it's a retrofit done on a tight budget. If you look online for underfloor heating temperatures, you get links like this which suggest 45C flow and 35-40C return temperatures for a screeded floor, higher for a floating floor. Those will be based on their experience of supplying systems for typical UK housing stock, and seems to be supported by other sources as well - Baxi for instance suggest that a 7kW boiler would be needed (100W/m²). That gives you a floor surface temperature roughly 10C warmer than the air temperature to provide the required heating, which depending on floor coverings will probably give you a slightly higher temperature on the downwards insulation surface.

[Jeremy Harris]

2 hours ago, pdf27 said:

This isn't for anything like a Passivhaus though - it's a retrofit done on a tight budget. If you look online for underfloor heating temperatures, you get links like this which suggest 45C flow and 35-40C return temperatures for a screeded floor, higher for a floating floor. Those will be based on their experience of supplying systems for typical UK housing stock, and seems to be supported by other sources as well - Baxi for instance suggest that a 7kW boiler would be needed (100W/m²). That gives you a floor surface temperature roughly 10C warmer than the air temperature to provide the required heating, which depending on floor coverings will probably give you a slightly higher temperature on the downwards insulation surface.

 

But the heat losses though the floor will be horrendous, and such a system would cost  a lot more to run than radiators.

 

There isn't scope here for decent underfloor insulation; we have 300mm of EPS and still waste around 5 to 8% of our UFH input energy directly to the ground.  UFH makes no sense at all unless the heating demand is down below 50 W/m² and even then the losses will be higher than with radiators.

[pdf27]

Err... not convinced by that. Very crude spreadsheet model. X-axis is the heat provided to the whole house in W/m2. That is then used to work out floor surface temperatures and hence losses through the floor for three insulation levels, given as a percentage of the total heat loss for the house. For a well-insulated house, floor heat losses form a significant proportion of the total. As heat losses elsewhere become more important, however, the quality of the floor insulation becomes much less important as a fraction of the whole, and the heat loss impact of using underfloor heating becomes less significant. For a gas boiler you're probably better off using radiators still (just - the impact is small enough that you **might** get it back from improved condensing), but with a heat pump I think it's pretty clear that the lower flow temperatures and hence improved COP from the bigger emitter area would make underfloor heating helpful - although in a retrofit case I suspect it's unlikely to be economic.

[Jeremy Harris]

The problem is that UFH is always less efficient than something like radiators, because UFH increases the floor temperature differential, and hence heat loss rate.

 

Take a typical renovation with a solid ground floor, where you might be able to get 75mm of PIR with a lambda of 0.023 W.m/K under the floor before laying UFH in a new 100m slab.  The ground underneath in the UK will be at around 8 deg C all year around, +/- 1 deg or so.  Lets say the house needs the massive 100 W/m² of heat from the UFH, quoted earlier, to stay at 21 deg C.

 

The U value of 100mm of concrete (lambda ~ 1.2 W.m/K) plus 75mm of PIR (lambda around 0.023 W.m/K) is about 0.3 W/m².K.  To achieve 100 W/m², with a room temperature of 21 deg C, the floor surface needs to be at 30 deg C.  The floor heat loss rate to the underlying ground is therefore (30 - 8) * 0.3 = 6.6 W/m², ignoring the significant additional heat losses from the UFH pipes running at close to UFH flow temperature (higher than floor surface temperature) directly, via the downward heat flux from the water in the pipes (which adds a lot to the losses, but complicates this simple model).

 

If the same house had radiators mounted on internal walls (the sensible place to put them to reduce heat losses) then the floor temperature would be just below room temperature (probably at least 1 deg C cooler because of the thermal gradient).  The heat losses through the floor would significantly decrease, because of the decrease in Δ t, to (20 - 8) * 0.3 = 3.6 W/m².

 

So, in this example, just as in any example of a house with UFH on a solid floor, there are higher ground heat losses with UFH than there are with radiators.  The higher the heating requirement, the greater the ground heat loss.  UFH is definitely nice to have, but there is no getting away from the fact that it is always less efficient at heating the house when fitted to the ground floor.  Radiators fitted to internal walls are virtually loss free, as there is no heat loss path to the outside as there is with UFH, so will always be cheaper to run when compared to them, whatever the heat source.

 

Edited December 8, 2017 by JSHarris

[Ferdinand]

Will respond to the excellent comments from all views later on.

 

There is an interesting analogy between ufh losses due to proximity to the ground, and radiators on external walls, perhaps?

[Jeremy Harris]

1 minute ago, Ferdinand said:

Will respond to the excellent comments from all views later on.

 

There is an interesting analogy between ufh losses due to proximity to the ground, and radiators on external walls, perhaps?

 

 

There is, and really putting radiators on outside walls isn't rally needed any more.  Like all things in house building, old ideas, that were sound in their day, survive for decades beyond their use by date.

 

In the case of radiator positioning, back in the mists of time, when central heating was first introduced, the highest heat loss areas in any room were the windows, by a massive margin.  The heat loss was so great that cold drafts would often descend from the window and then run along the floor.  When central heating came along, it seemed eminently sensible to place radiators under windows, to that these cold drafts would be eliminated.  The snag is it increased the heat losses, a lot, as now the air next to the windows was warmer than the room by a fair bit, the windows were still the poorest thermal element in the fabric, so a lot of heat went straight out through the window.

 

We now have pretty good windows available, and the days when the inside of a window might be covered with frost and ice after a cold night are no more, so there is rarely any need to fit radiators under windows.  Moving them to internal walls makes sense, as they then have no practically heat losses at all; almost all of the heat that goes into the radiator is delivered to the room.

 

 

[pdf27]

1 hour ago, JSHarris said:

So, in this example, just as in any example of a house with UFH on a solid floor, there are higher ground heat losses with UFH than there are with radiators.  The higher the heating requirement, the greater the ground heat loss.  UFH is definitely nice to have, but there is no getting away from the fact that it is always less efficient at heating the house when fitted to the ground floor.  Radiators fitted to internal walls are virtually loss free, as there is no heat loss path to the outside as there is with UFH, so will always be cheaper to run when compared to them, whatever the heat source.

I agree with the loss calculation, what I'm not so convinced by is the significance of it. What you're showing is that underfloor heating is a bit less efficient than radiators, when the radiators use water at the same temperature as the underfloor heating. That seems entirely reasonable to me, but it also doesn't seem like a very real-world situation - unless you're using fan coil type radiators then using the entire surface area of the floor as a heating surface is going to be pretty hard to equal. That isn't really a problem in a very well insulated house since the heating requirement is very low and "normal" sized radiators are now drastically oversized and will work just fine.

 

It's a bit different in a less well insulated house, say at the 50W/m² point. A smallish 10m2 room would need 500W - that puts the floor temperature 5C above ambient and the water temperature no more than 30C. A heat pump is operating at peak efficiency and a boiler will be fully condensing at this point - perfect. The correction factor seems to be about  (air-water temperature difference/50)^1.344 - so to match the underfloor the radiator would have to be 8.7 times larger than the book value - i.e. sized for 4.3kW. That means specifying a triple panel 600 x 1800 radiator for a room 3m x 3m!

 

Now realistically you don't need to match the low temperature of the underfloor heating to get all of the efficiency gains from the rest of the heating system, and that gets you away from ridiculously big radiators (at 40C you're down to a 1.7kW radiator for instance) - but there is clearly a risk that by optimising to reduce total heat losses from the building you're increasing rather than decreasing total energy demand, depending on how your heating system of choice reacts to varying output temperatures.

[billt]

I think the reason for placing radiators on outside walls under windows is a bit more subtle than that.

 

With a normal house and radiators on an internal wall there will be a big temperature gradient between the inner wall and the outer wall, which makes the temperature distribution uneven leading to an uncomfortable room and a tendency to increase the room temperature. With a radiator on the outside wall the temperature distribution is much more even leading to a comfortable room at a lower temperature. There is also the issue of the warm heat source of the radiator compensating for the cold sink of the window. This may be less of an issue in a Passivhaus, but it's still going to be more effective to place radiators in the traditional place in most radiator heated buildings; windows are still a cold spot.

[Jeremy Harris]

22 minutes ago, pdf27 said:

I agree with the loss calculation, what I'm not so convinced by is the significance of it. What you're showing is that underfloor heating is a bit less efficient than radiators, when the radiators use water at the same temperature as the underfloor heating. That seems entirely reasonable to me, but it also doesn't seem like a very real-world situation - unless you're using fan coil type radiators then using the entire surface area of the floor as a heating surface is going to be pretty hard to equal. That isn't really a problem in a very well insulated house since the heating requirement is very low and "normal" sized radiators are now drastically oversized and will work just fine.

 

It's a bit different in a less well insulated house, say at the 50W/m² point. A smallish 10m2 room would need 500W - that puts the floor temperature 5C above ambient and the water temperature no more than 30C. A heat pump is operating at peak efficiency and a boiler will be fully condensing at this point - perfect. The correction factor seems to be about  (air-water temperature difference/50)^1.344 - so to match the underfloor the radiator would have to be 8.7 times larger than the book value - i.e. sized for 4.3kW. That means specifying a triple panel 600 x 1800 radiator for a room 3m x 3m!

 

Now realistically you don't need to match the low temperature of the underfloor heating to get all of the efficiency gains from the rest of the heating system, and that gets you away from ridiculously big radiators (at 40C you're down to a 1.7kW radiator for instance) - but there is clearly a risk that by optimising to reduce total heat losses from the building you're increasing rather than decreasing total energy demand, depending on how your heating system of choice reacts to varying output temperatures.

 

The temperature of the water in the radiators doesn't change their efficiency one iota, only their heat output.  Thire heat output relative to surface temperature is fairly similar to that for UFH, although the equation is slightly different because of the changed relationship between the convected and radiated components.

 

The point I'm making is that any heating system that wastes some of the heat output directly into the ground is always going to be less efficient than one that doesn't.

[Jeremy Harris]

3 minutes ago, billt said:

I think the reason for placing radiators on outside walls under windows is a bit more subtle than that.

 

With a normal house and radiators on an internal wall there will be a big temperature gradient between the inner wall and the outer wall, which makes the temperature distribution uneven leading to an uncomfortable room and a tendency to increase the room temperature. With a radiator on the outside wall the temperature distribution is much more even leading to a comfortable room at a lower temperature. There is also the issue of the warm heat source of the radiator compensating for the cold sink of the window. This may be less of an issue in a Passivhaus, but it's still going to be more effective to place radiators in the traditional place in most radiator heated buildings; windows are still a cold spot.

 

 

In practice, the windows are no longer a practical cold spot with modern glazing.  One of the first things I noticed with our new build was that you feel no cooler standing by a window in winter than you do standing by a wall, and this was backed up with thermal imagery, that showed that the inner pane of glass was only very slightly below room temperature not enough to detect with your hand) and even the coldest part of the window (on a cold day), the edge of the frame, was only about 1 deg C cooler than room temperature.

[oldkettle]

Whether the radiator is under the window or somewhere else on the external wall, it still feels way more comfortable than when the radiator is on the internal one. My home office is on the ground floor, I sit facing the outside wall and the radiator is way behind me because it was likely simpler to fix there. Unless I put woolly socks on it is ridiculously cold  as my feet are right next to the wall. It may be less heat loss but it is certainly more discomfort. Although I guess if the radiator was next to me on the wall perpendicular to the external it might have been OK. 

[Jeremy Harris]

21 minutes ago, oldkettle said:

Whether the radiator is under the window or somewhere else on the external wall, it still feels way more comfortable than when the radiator is on the internal one. My home office is on the ground floor, I sit facing the outside wall and the radiator is way behind me because it was likely simpler to fix there. Unless I put woolly socks on it is ridiculously cold  as my feet are right next to the wall. It may be less heat loss but it is certainly more discomfort. Although I guess if the radiator was next to me on the wall perpendicular to the external it might have been OK. 

 

 

Depends on how even the temperatures are around you.  One of the most noticeable things with our new house is that all the surfaces, walls, ceilings, windows, floor etc are at near enough the same temperature.  The variation is less than 1 deg C, even on a cold day.

 

The effect of this is to make you feel more comfortable, and I have a theory that this is to do with the way the body senses thermal comfort.  Air temperature seems only a part of the overall feeling of comfort, it seems that we are quite sensitive to directional heat loss, mainly by radiation, but also by conduction.

 

I can't prove this, but empirically, standing inside in front of a window on a cold day feels a great deal more comfortable at the new house than at the old house, even though the old house probably has a warmer room temperature.  I think this is down to our bodies being pretty good at detecting directional radiative heat loss, and with 3G windows, with two panes with a low emissivity inward facing coating, the radiative heat loss through the glazing is very much lower than from the 12 year old double glazing at the old house. 

 

The same goes for sensing conductive heat loss.  The solid concrete floor of our old house has no insulation, and if you stand around for a while, or sit with your feet on the floor, it doesn't take long for your feet to feel cold.  I tested this out with a 50mm thick offcut of PIR foam, that I put on the floor under my desk, to rest my feet on.  It makes a very significant difference to how my feet feel!

Edited December 8, 2017 by JSHarris