Heat loss and running cost

[Jeremy Harris]

Sometimes questions are raised as to whether it's worth increasing insulation levels and often there seems to be confusion as to what the "ideal" level of insulation is, or even what a "good" or "reasonable" level of insulation might be.  I'm not sure whether or not the non-linear impact of improving insulation, in terms of the effect on the heating requirement, and hence running cost during cold weather, is widely understood.

I've heard comments like "it's not worth improving the insulation from 0.16 W/m².K to 0.12 W/m².K because it would be 30% more expensive and only reduce the heat loss by 25%".  Most of the time this is incorrect, because homes have heat sources all year around, from the occupants, incidental heating from appliances, solar gain and even pets (a medium sized dog is probably a four-legged 40 - 50W heater).

So, I thought a really simple example might help some gain a better understanding of this non-linearity, and illustrate better why some are so evangelical about trying to improve insulation levels (and reduce ventilation heat loss, too, but I'll get to that another time). 

Let's build a pretend house, that for simplicity has no doors or windows and is a rectangular single storey box with a flat roof.  For simplicity we'll assume it's on raised piles, with an air space underneath, just so we can use the same insulation level on all six sides and to make the sums simple.  All I'm doing here is making a comparison, so this is a valid way of illustrating this effect.

In our rectangular box house we have an average of 100W of incidental heating, coming from things like internet kit, a PC, a cordless phone base station, a TV, a phone charger, a few lights and a handful of intermittently used kitchen appliances.  This is a pretty low figure - I struggle to keep our house background load below about 200W, without any lights on.  The box houses two adults, giving out around 80 - 100W each and a dog, so lets say there is 220 W of heating coming from the occupants.  The box also has a heating system that can deliver whatever power is needed to maintain a temperature of 20 deg C inside, and its night time, so there's no solar heating of the walls.  Outside it's 5 deg C, a chilly winters night.  This rectangular box is 10m long x 10m wide x 2.5m high inside, so has a total wall, floor and roof area of 300m² and an internal floor area of 100m², so fairly average in size (a bit bigger than our current 3 bed bungalow).

So, we have a temperature difference between the inside and outside of 15 deg C (20 deg C - 5 deg C), an internal surface area of 300m² and a constant incidental heating level of 320 W (220 W from two adults and dog, 100 W from electrical appliances and lights).

First, lets see how much heat we need to put into this box from the heating system, if we have U values for the walls, floor and roof of 0.2 W/m².K (K is degrees Kelvin, the same units as degrees Centigrade when only temperature difference is being compared):

The total heat loss power, in Watts, can be calculated from the U value, the area and the temperature difference, so for this first example we get 300m² area x 15 deg C temperature difference x 0.2 W/m².K U value = 900 W.  There is 320 W of heat coming from the occupants etc, so the heating system would need to deliver 900 - 320 = 580 W in order to keep the house at 20 deg C under these conditions.  If this were by direct electric heating, then the heating cost would be about £2.09 per 24 hours.

Next, let's see how much heat we need to put into this box from the heating system, if we have U values for the walls, floor and roof of 0.1 W/m².K , in other words, we've made the insulation twice as "good", so might think we've halved the heating cost:

The total heat loss power is now 300m² x 15 deg C temperature difference x 0.1 W/m².K U value = 450 W.  This is what we'd expect, double the insulation effectiveness and halve the heat loss.  However, when we now take away the incidental heat gain from the occupants, etc, of 320 W, the heating system needs to deliver 450 - 320 = 130 W in order to keep the house at 20 deg C under these conditions.  If this were by direct electric heating, then the cost would be about £0.47 per 24 hours.

So, by doubling the insulation level we've decreased the heating cost by about 78%, not the 50% that might have been expected.

This is a very simplistic example, but it does illustrate why doubling up in insulation can give a far greater benefit than might be expected.  It also shows why, when you improve the level of insulation you can reduce the heating requirement down to such a low level that for a lot of the time you don't need any heating.  In that last example, turning on a few extra lights could heat this imaginary box home to a comfortable temperature on a cold night, whereas with only half the insulation it needs something that delivers 446% more heat.

 

 

 

 

[Guest Alphonsox]

A useful analysis. Another view that could be useful is to look at the external temperature at which the 320W generated by the occupants is sufficient to allow the heating system to be turned off. i.e. the point where the incidental energy input alone is greater than that needed to maintain the 20C internal temperature. In the house with the 0.2U value insulation this occurs when the outside temperature rises above ~15C. For the 0.1U house this occurs at ~10C. Average temperature for London only rises above 15C for 4 months of the year so the heating will be on to some extent for the other 8 months in the 0.2U house. The numbers reverse themselves for the 0.1U house, the average temperature only drops below 10C for 4 months of the year leaving the heating system turned off for 8 months of the year. So not only are you paying less per 24hr in your 0.1U house your heating system is only on for half as long.

 

 

[Jeremy Harris]

I keep meaning to do some pictures to try and explain this better, as I know some find the numbers a challenge.

There are some interesting consequences resulting from improving insulation/reducing heat loss, by only a modest amount.  Having visitors around very quickly warms the whole house up, for example, as a couple of adults added to the occupancy can easily just reduce the need for any heating, even in cool weather, to zero.  Similarly, the house becomes more sensitive to incidental gains, so the added heat from a bit of sun shining though a window can easily give more heat than is wanted, even in winter.

 

[TerryE]

Three other factors that played into my decision to go for a twinwall (~0.12U) rather than a single wall frame were:

• Our twinwall is pumped fill rather than factor slab fill.  There are less risks of voids and convection circuits in a twinwall.  And the pumped filler significantly reduces the risk of air leaks, so the design is intrinsically more robust in its thermal performance.
• Decrement delay factor.  A twinwall pumped with cellulosic filler has a DDF measured in days, so you really don't need to worry about those winter nights when the temperature can go far below 0°C -- you only need to design for actual average temperatures.
• House zonal variations.  With a U value in the 0.12U zone the internal heat flows are significantly greater than external ones.  The temperature delta ground floor to first floor (even with all of the occupants and internal heating on the ground floor) is less than 2°C in winter extremes: you don't need to worry about designing in "backup" central heating for the upper floors -- it will be a total waste of money.

In simple terms you can be confident that you have the design margins with a twinwall construction to discount / ignore a whole bunch of issues that would add to the complexity and cost of the build.

[TheMitchells]

22 minutes ago, TerryE said:

House zonal variations.  With a U value in the 0.12U zone the internal heat flows are significantly greater than external ones.  The temperature delta ground floor to first floor (even with all of the occupants and internal heating on the ground floor) is less than 2°C in winter extremes: you don't need to worry about designing in "backup" central heating for the upper floors -- it will be a total waste of money.

In simple terms you can be confident that you have the design margins with a twinwall construction to discount / ignore a whole bunch of issues that would add to the complexity and cost of the build.

Sorry Terry, you've lost me - Temperature delta.... is that the difference between upstairs and downstairs.  its a new term for me so may be for others. 

Anyway - thanks for the simplified picture (words) - that makes it a lot clearer Jeremy.  will add that to my talk!

[TerryE]

In our current farmhouse it can be toasty downstairs with the log fire on etc. but is bloody freezing upstairs.  Delta is just science / maths-speak for difference.   If you've got no heating upstairs then any heat losses through the walls have to be balanced by air circulation and heat coming through the floor.  The current BRegs for acoustic insulation between floors means that you've also got reasonable thermal insulation between floors, so if you aren't careful then you might end up needing some low background heating in the bedrooms that you use.  

[TheMitchells]

thanks - will try to remember that.  I obviously wasnt listening properly in the science class. 

[Crofter]

A very useful discussion!

In my own project I have chosen to go down the low-energy route because it saves me having to fit a central heating system, and by opting for controlled ventilation I hope to improve air quality and avoid damp and mould problems.

 

One other factor that I have come to realise is the synergy between low energy and low build cost: most of the things that make a house 'high energy' actually increase the cost, e.g. lots of little windows, complex shapes, chimney breasts.

[AliG]

Hi Jeremy,

I hate disagreeing on line but I wanted to bring some other points to bear.

No one would disagree that more insulation will save money and the savings are not linear due to the incidental heating of electrical items and occupants as well as solar gain. There is a particularly large saving right at the more extreme point if you can avoid installing heating systems at all.

However, I think it is better to think in terms of cost effectiveness or the return on the investment in extra insulation , thus there is a point where the costs of more insulation outweigh the benefits. A few things effect this.

1. A quick look on Uswitch gave me a lowest cost for electricity of 8p/kwh and for gas of 2.4p per kWh. Most people use gas to heat their homes. Assuming even 4p per kWh to take into account VAT, boiler efficiency and longer term gas price volatility, in your example the heating cost would fall from 56p a day to 12p a day. Now the ratio is still the same, the more insulated house only has circa 22% of the heating cost of the less insulated house, but the heating cost per year of the less well insulated house would be circa £200 if the temperature was 5C all year round. The heating would not be needed for half the year so the heating cost would be somewhat lower. The other house may have a cost of only a few pounds  as it will have lower costs and less days of heating, but the £100-200 saving may not be great relative to the cost of the extra insulation. It should be noted, however, that if you can get the heating requirement down to the point where you need no gas at all and no heating system then you could save thousands of pounds in capital costs and gas standing charges which may run to around £100 per year.

2. The 100 square metre floor area you have used is roughly the UK average house size, but this includes flats etc. The average detached house is roughly 150 square metres and the average self build is probably larger, say 200 square metres. Thus incidental heating is likely to account for a lower percentage of the heating costs than you have allowed.

3. When the discussion comes up on the site it usually centres around changing one element of the build such as the walls or roof. If I look at the calculation for my house, the walls only account for 20% of the heat loss, this could vary considerably house by house. I have to note that I have used Jeremy's heat loss calculator to figure this out, so thanks for that Jeremy. The triple glazed windows, which are not excessively large in area by self build standards and have around a 0.8 U value account for around 40% of the heat loss, the roof another 20%, then floor and ventilation around 10% each. I would say it is more appropriate to look at individual elements and the cost benefit of improving each. Moving from double glazed to triple glazed windows may save more than moving from 0.2 U value to 0.1 U value walls and may cost less. A lot depends on the design of a house, the materials it is constructed from etc. In the example of a gas heated house of 100 square metres floor area, the saving of changing the wall U-Value from 0.2 to 0.1 would probably be less than £50 a year.

The point I wanted to make is yes, more insulation saves money and yes it is non linear. But, if people are simply looking at cost effectiveness, and many people are budget constrained, then the costs may still outweigh the savings. There may of course be other benefits in terms of the comfort of the house and some people may be more worried about the environmental impact of their house than costs.

Edited June 15, 2016 by AliG

[ToughButterCup]

46 minutes ago, AliG said:

I hate disagreeing on line but I wanted to bring some other points to bear.

[...]

However, I think it is better to think in terms of cost effectiveness or the return on the investment in extra insulation , thus there is a point where the costs of more insulation outweigh the benefits.

[...]

 But, if people are simply looking at cost effectiveness, and many people are budget constrained, then the costs may still outweigh the savings. There may of course be other benefits in terms of the comfort of the house and some people may be more worried about the environmental impact of their house than costs.

Productivity in relation to cost (cost effective by another name) will differ by [self?] builder won't it?

In our build,  can't afford a foundation made to the highest spec, but many others could. So cost effectiveness is a subjective measure. What's OK for me, isn't for others.

And I'm almost sure J was focusing on just the numbers - so that that logic can inform the decision about how much money to invest. J's contribution is just  a framework on which to hang a very personal decision.

[Jeremy Harris]

I've no problem with any disagreement, this was just a very simplistic model.  Everything you say is right, and it's unlikely that anyone would use a full price direct electric tariff for heating, but the saving proportion is the same whatever fuel is used - I just picked a number (15p per kWh including standing charges) that illustrated a price difference.  One odd thing is that when I've looked at the cheaper tariffs for the sort of electricity consumption our 130m² home would have (without the solar panels) they have (so far) all turned out to cost more than this per unit when the standing charge is taken into account.

All I was trying to do here was illustrate one point, very simplistically, as an illustration, not a practical home.  There are dozens of compromises that will impact on a real house, but it was this single principle of the non-linearity of heating cost as you reduce the heating requirement that I was trying to illustrate, as that seems to be something that some don't always take account of, as I mentioned in the first two paragraphs.

Not everyone here is building a new house, either, some here are renovating, converting or restoring, and I'm not sure that many here are building houses of 150m² or more, even if that is the average for new self-builds.  I may be wrong, but I get the impression that there are a lot of us are building in the 100 to 150m² range and some smaller than that, with only a few that are bigger.  I'd also guess that the more cost-constrained are going to be building at the lower end of the size range and may well reap greater benefits (in proportion to their income) from having significant lower energy bills, or no energy bills at all, which really isn't either costly or difficult to achieve, we've found. 

Our total build costs, including the solar panels and MVHR, oak joinery throughout, fairly high to mid range kitchen and bathrooms and passive house levels of insulation and airtightness have come to about £1380/m².  If we'd cut out the oak, travertine flooring, gone for budget kitchen and bathroom components, then that could easily have come down to £1200/m².  Using an architect/project manager and budget to mid range kitchens and bathrooms would probably have ended up at around £1350/m².  The average self-build cost for our part of the UK for this size of house is probably around £1400/m², for someone, like me, who uses a main contractor for all the ground works and another main contractor for the wind and weather tight shell, with sub-contractors for windows, roofing, electrics, plastering etc.  Bigger houses may well be a bit cheaper, but then bigger houses don't need as high an insulation level, either, because of the better volume to surface area ratio.

I'd be the first to say that any self build is a stack of compromises, driven by planning, the site, personal wishes, cost, etc, so there will be very wide variations.  However, one consistent factor is that it is damned hard and costly to improve insulation levels once a house is built.  After the house is built and been in use for a few years, and you, or the next owner, is in a financially better place, you can always change the windows and doors for better ones, fit a better heat recovery system, upgrade the kitchen and bathrooms, install solar panels, lay better flooring.  So, if you're looking at making compromises it seems far better to not compromise on wall, floor and roof insulation, just because these can prove to be hard to improve later, and because, in the overall cost of even a budget build, insulation is only a small part of the cost.

 

[DeeJunFan]

I can only dream of the day I start on my 100-150m2 self build.

We took some advice upfront about our plans and rolled with it and it only when you see the beast of a thing in 1:1 scale in 3d that you think "why are we building such a big house?"

Stone guy came round last week and said "That's not a house, thats a hotel"  

I didn't think 330m2 was that big until I saw it.

On the OP I think it's a really clear representation of the insulation benefit/ratio.  I also used Jeremy's spreadsheet quite extensively (Also PHPP but found JSH easier to work with and pretty much as accurate)

The general level of knowledge around about such things is pretty grim.  I was only directed towards lower energy when my sister got a "consultant" to do up a plan for a self build that was supposed to provide a well insulated/airtight house.  (since joining the forums a few years ago I found out this guy is/was a copy/paste merchant who was giving some pretty poor advice but at least he was actively encouraging the thought process.

I was very gung ho when I started out the design, but over time I have had different levels of advice (some good, some awful) but I have used the knowledge gained here and the tools to make more informed decisions and overall I'm happy with where things are.  Insulation is one thing I'm coming back to over time as I was more drawn to making bang for buck decisions as mentioned by AliG above.  Going from 2g to 3g was far more impactful that going from 0.17 on the walls to 0.14 but it's because of topics like this that I was in a position to make those choices and understand the impact.

So a big shout out and thanks to everyone who cares enough to share knowledge/experience/mistakes etc with the rest of us!

 

[Jeremy Harris]

Yes, with a big house the impact of improving the insulation level decreases quite markedly.  By the time you get to the likes of a DIY shed supermarket it's barely worth having more than a thin layer of insulation, because of the change in surface area to volume ratio.  For big houses it's ventilation heat loss that is very dominant, so good airtightness and a high performance MVHR will reap a greater benefit than improving the wall, floor and roof insulation.

Our idea was to downsize with the new house, but at  2 bedrooms and 130m² habitable floor space it's 44% bigger than the three bed bungalow we're in at the moment!

[Stones]

I'm actively considering the spec of the insulation to the vaulted ceiling in my build.  The current spec is for U value of 0.1 but changing insulation type gives me a U value of 0.14, is cheaper in terms of materials and installation time.  I'm waiting on a revised price from the builder to see what the price difference is so I can work out whether it is worth spending the extra to achieve the £17 a year saving in heating costs that I gain with the 0.1 U value roof.

[Jeremy Harris]

Your conditions are probably atypical for many other parts of the UK (exposed, near-constant wind etc) but you do get very long days in summer, so I think I'd look long and hard at the decrement delay.  It may well be worth making a compromise on U value to get a longer decrement delay, in terms of comfort level during any long sunny days you might get up there. 

[SteamyTea]

When it comes to cost effectiveness there are a number of different ways to calculate this.

You can take 'today's' figures and assume that all proportions will be the same in the future.  This is my personal favourite as it is easy.

You could try and look at long term earnings and energy price inflation.  This is harder to do, but fairly easy still.  Just work on an average house spending 5% of income on heating, you won't be far out.

Those are the simple ones.

The harder one is working out what each square metre of floor area is worth, then compare U-Values to loss of value because of thicker walls, or walls of the same thickness but better insulation i.e. aerogel backed plasterboard.

The really hard one is to start combining all the options, so inflation, property value, more heating technology control, alternative sources of energy, building design (a sphere is pretty good), better foundations.

 

All the calculations are pretty simple, though tedious to do, but with a spreadsheet they soon become easy.

Or just aim for an overall U-Value of 0.1 W.m^-2.K^-1.  This will almost certainly mean you are spending more on windows and doors than the walls.

Edited June 15, 2016 by SteamyTea

[CH1]

Could any knowledgeable person out there advise whether it is worth installing treble glazing over double glazing where the improvement in u-value would go from 1.3/1.4 to 1.1/1.2 for two large areas of glazing (one curtain wall system east elevation approx 10m sq) and other sliding door plus 2 fixed panes approx 12m sq south elevation, albeit slightly recessed glazing.  The cost to upgrade is approx £2.5k.  Is an improvement of 0.1/0.2 a decent improvement for the cost?  thanks in advance

[SteamyTea]

7 minutes ago, CH1 said:

Could any knowledgeable perso

Just knock up a quick model of the house and see what the differences are.

Solar gain may also be an issue, so worth taking a stab at those calculations as well.

[Jeremy Harris]

Welcome.

 

1.1 to 1.2 W/m².K seems a really poor figure for 3G, typically 3G will give a U value of around 0.7 to 0.8 W/m².K, perhaps slightly lower for large glazed areas. I would question the glazing supplier, as something doesn't seem right, as 1.1 to 1.2 W/m².K is about the same performance as a very good 2G glazing system, rather than 3G. 

 

The U value isn't the whole factor, either, as 3G allows two low e coated panes to be used, which just about halves the radiated heat loss, and subjectively makes the glazing feel warmer, by reflecting more IR back into the house.  This is quite noticeable if stood or sat near glazing, where you can feel the reduced radiated loss on a cold day.

 

Of more concern should be the potentially high solar gain from 12m² of South facing glass.  I would very seriously look at ways to mitigate this.  We have about 2/3rds that area of South facing glass and have had to install solar reflective film to try and cut down the overheating it creates, as well as add air conditioning.  Others here have done much the same, and fitted external shutters, or installed Sage glass, as alternative ways to reduce solar gain.

[ProDave]

I suspect over heating is going to be your issue. Look at and choose carefully to get the right coatings on the glass to minimise that risk.

[MikeSharp01]

3 minutes ago, JSHarris said:

as well as add air conditioning

Sorry Jeremy did I miss something you now have air con, or does this refer to your reversible ASHP / slab and the possibilities of cooled air via the MVHR?

[Jeremy Harris]

32 minutes ago, MikeSharp01 said:

Sorry Jeremy did I miss something you now have air con, or does this refer to your reversible ASHP / slab and the possibilities of cooled air via the MVHR?

 

Yes, I installed a Toshiba split unit recently, just before the really hot spell (thank goodness!).  There's some details here:

 

[CH1]

The glass manufacturer gave me u values for glass and frame combined.  The G value for glass alone will no doubt be better, but taking into account the frames it then changes the u-values.  So it sounds like a jump of 0.1/0.2 isn't that much of a saving?  The south elevation will be recessed so hopefully that will help the potential overheating 

[Jeremy Harris]

1 minute ago, CH1 said:

The glass manufacturer gave me u values for glass and frame combined.  The G value for glass alone will no doubt be better, but taking into account the frames it then changes the u-values.  So it sounds like a jump of 0.1/0.2 isn't that much of a saving?  The south elevation will be recessed so hopefully that will help the potential overheating 

 

 

I'd be inclined to look around at alternative glazing, as typically a 3G glazing unit will have a Ug of around 0.5 to 0.6 W/m².K, so unless you have small panes of glass and lots of frame area it's hard to see how you could end up with a Uw as high as 1.2 W/m².K for 3G.  We have cheap 3G and the Uw works out at between 0.7 W/m².K for the large glazed areas down to 0.8 W/m².K for the smaller windows.

 

Our main glazing sits under a 0.5m overhang, but that has zero effect on overheating, as the heat comes in when the sun is still low in the sky.  We've ended up changing the ASHP around to cool the ground floor (via the UFH pipes), installed an air con unit upstairs and installed (expensive) solar reflective film to the glass to reduce the over heating.  I regret not having known about Sage glass originally, as despite the cost, it seems a very good solution to over heating.  Planning conditions meant we couldn't fit external shutters or blinds, which is a shame, as others here have found them to work well.

 

If you have a read around here you will find that several members have encountered overheating problems from large areas of glazing.  Very well worth considering the impact and trying to design in mitigation measure, IMHO, as it has been a PITA trying to change things in order to cool our house down in warm weather.  It's not just a summer problem; this year our house overheated in April, such that I had to turn the cooling system on, from the low angle of the sun at that time of the year.

[MikeSharp01]

18 minutes ago, JSHarris said:

Yes, I installed a Toshiba split unit recently

Completely missed that - interesting read. Got a question and will pop that on the end of the thread above.