Decrement delay greater than 12 hours

[davidc]

I know that the sweet spot is 12 hours, but is having a delay of 14 hours any less desirable than 12 ?

 

Thanks

[ProDave]

What do you perceive as a problem with a longer decrement delay?

[joe90]

No.

[Nickfromwales]

4 hours ago, davidc said:

I know that the sweet spot is 12 hours, but is having a delay of 14 hours any less desirable than 12 ?

 

Thanks

I'd be very interested in how you modelled the 2 hour divide?

[SteamyTea]

As 'decrement delay' is the time taken for a set temperature drop, knowing the starting points and the ending points are important.

And the drop in temperature is only half the story, there will be a rise in temperature sometime, which in average, will need to balance out (or the temperature will always rise or fall).

Then there are the measurement conditions i.e. energy forcing, are they positive or negative.  This will make a huge seasonal difference i.e. what happens in the summer and winter can be very different.  Even spring and autumn, which may have similar hours of daylight to your decrement time may well be very different. 

 

Taking my house, which I have long data sets for, I have a delay of 12 hours per -1K between midnight and 6 AM, 6 AM to 9 AM, it is 12 hours per +1K, 9 AM to 4 PM, it is stable, so infinite hours per 0K, a classic divide by zero, 4 PM to midnight it is 24 hours per -1K.

So taking simple mean averages, it is 18 hours to cool by 1 K, but 12 hours to get back to the same energy level.  These do not match, but one is in part darkness and the other full daylight, the 6 hours difference is due to solar forcing and my heating/ventilation system.

Looking at the last week (01/08/20 to 08/08/20) there is no noticeable difference in times, just higher means temperatures.

Worth noting that, so far this year, the internal temperature range is still only 1K on an average of 21.6°C internal (this will drop to about 20°C for the full year) on a mean external temperature of 11.6°C (this may creep up to about 12°C for the full year, autumn is warmer than spring).

The mean temperature difference, between internal and external, so far this year is 9.9°C.

So a better way to measure temperature stability may be to look at internal temperature range for different external temperature ranges i.e how much does my house vary when the external temperature is 0°C, 1°C, 2°C etc.

This would take a bit of analysis, but shall see if I can do it sometime.

 

[A_L]

I believe @SteamyTea that the OP is referring to the time taken for energy arriving at the external surface of an opaque building element to transfer (in a temperature wave) to the internal surface and the room.

 

@davidc, There is probably very little difference in 12 and 14 hours. I would be interested in your build-up and like Nick which software/method have you used?

[SteamyTea]

30 minutes ago, A_L said:

I believe @SteamyTea that the OP is referring to the time taken for energy arriving at the external surface of an opaque building element to transfer (in a temperature wave) to the internal surface and the room.

That is just the thermal inertia of the materials.  Easy to work out as it is the product of the conductivity, density and the heat capacity.

 

https://en.wikipedia.org/wiki/Thermal_effusivity

 

Edited August 12, 2020 by SteamyTea

[SteamyTea]

Have sorted and cleaned up this years temperature data, interesting result.

As a house cannot be placed into a lab, one can only work with the prevailing conditions.  This can be described as 'normal usage'.

What seems to be happening is that the decrement delay runs into the millions of hours, with the only times it is in the sub tens of hours is at the extremes.

During the normal external temperature ranges of between 6°C and 18°C (this is Cornwall, so no big extremes), there is very little change in internal temperature , it just tracks external temperature (y = 0.2792x + 16.869, where y in internal temperature and x is external temperature).  It increases by 0.28°C for every 1°C external temperature rise.

Over the whole temperature range this year so far, it is only 0.34°C for each 1°C rise in external temperature.

I suspect that most of this is down to the ventilation i.e. I close windows when it is cold, open them when it is hot, solar gain.

I may have a look at night time temperatures later, and at extremes.  This may give a fuller picture as it eliminates one element.

Deaggregating is not an easy task as the signal to noise ratio is quite low, but I shall see what comes out.

But basically my feeling is the adding additional mass to a building makes almost no difference to the thermal response times.  This is more affected by ventilation, heating/cooling and solar gain.

But to initially answer the question, does 2 hours difference actually make a difference, then no.

 

 

 

[davidc]

Thanks to all who responded. I couldn't see how the extra two hours would cause an issue - but didn't want to mistakenly assume that as a given.

[Nickfromwales]

4 hours ago, davidc said:

Thanks to all who responded. I couldn't see how the extra two hours would cause an issue - but didn't want to mistakenly assume that as a given.

"The longer the better" is the bottom line  Will add sustain to the thermal inertia of the internal fabric and core structure, ( that which is within the heated envelope ), which is a good thing for heating via load shifting eg ASHP > UFH in slab via 3x E10 'chunks' which should afford you zero input in between those off-peak times.

Quite a 'big' question / subject, so specifics would be appreciated if we're to expand  

[SteamyTea]

24 minutes ago, Nickfromwales said:

"The longer the better" is the bottom line

I am not so sure.

Looking at my data, when there is a very small, or very large difference between internal and external temperatures, the instability is greatest.  Exactly what one is trying to avoid.

Some of this will be down to the low number of samples, 322 for the bottom two temperatures, 43 for the top two, but 52,922 for the ones in the middle.

The rest will be down solar gain and heating/ventilation.

I am of the opinion that it is almost impossible to design a normal house that can take advantage of time shifting the thermal loads.  This is mainly a UK problem as we are above 50° North, which means we have large change in solar gain throughout the year. We are also surrounded by a warm ocean on one side, and a cold sea on the other side and are windy and cloudy.

It is also worth remembering that at low temperature difference, there is little capacity to store excess energy, the energy difference is the driver.  Then at high differences, losses are greater as there is more energy to disperse.  That is why cooling and heating does not follow a straight line, or even a simple sinusoidal curve.  Not even solar power follows a sinusoidal curve (it is the sum of at least 4 curves, and then a random element for clouds, Fourier showed us this).

 

I just added a trend line to my chart, the equation is y = -4E+08x - 2E+10.  So basically flat.  I expect the second part the, -2E+10, to get closer to zero as the year progress, though this is weather dependant, and weather is fickle.

 

Edited August 12, 2020 by SteamyTea

[Nickfromwales]

11 minutes ago, SteamyTea said:

I am of the opinion that it is almost impossible to design a normal house that can take advantage of time shifting the thermal loads.  This is mainly a UK problem as we are above 50° North, which means we have large change in solar gain throughout the year. We are also surrounded by a warm ocean on one side, and a cold sea on the other side and are windy and cloudy.

Exactly my point above, to create a space largely unaffected by external influences. Wind and cloud are outside, and only thing affected by wind really is the MVHR. Yes the wind adds to the heat / cool 'wrap' around the exterior of the dwelling, but investing in staving off any big influences from external conditions is a plus afaic. 

[SteamyTea]

3 minutes ago, Nickfromwales said:

but investing in staving off any big influences from external conditions is a plus afaic

Not sure what you mean here.

Shielding a house from the elements is a plus for stability, but may be a negative for overall energy usage.

Stick a house on our western seaboard probably has the greatest benefits overall.  It is why I get away with low energy usage, even though I am 200 metres ASL, when I get strong winds, they are warm.  And when the temperatures are high, with little wind, I get cloud (last two days have been misty and wet here).

[Nickfromwales]

1 minute ago, SteamyTea said:

Not sure what you mean here.

Shielding a house from the elements is a plus for stability, but may be a negative for overall energy usage.

Stick a house on our western seaboard probably has the greatest benefits overall.  It is why I get away with low energy usage, even though I am 200 metres ASL, when I get strong winds, they are warm.  And when the temperatures are high, with little wind, I get cloud (last two days have been misty and wet here).

In a nutshell, stopping the worst of the external temp swings from affecting the dwelling interior. Once you are close to equilibrium the space heating ( and cooling ) requirements will be quite negligible, just a question of how big a hill you have to climb ( and how much it costs ) to get there vs what you will save on energy.

[SteamyTea]

1 minute ago, Nickfromwales said:

In a nutshell, stopping the worst of the external temp swings from affecting the dwelling interior. Once you are close to equilibrium the space heating ( and cooling ) requirements will be quite negligible, just a question of how big a hill you have to climb ( and how much it costs ) to get there vs what you will save on energy.

Right, yes.

Why the outside of a building needs to be wrapped in a windproof barrier.

It is also why adding excess mass does not aid temperature stability.  There is enough mass in plasterboard and OSB for that.

[Nickfromwales]

14 hours ago, SteamyTea said:

Right, yes.

Why the outside of a building needs to be wrapped in a windproof barrier.

It is also why adding excess mass does not aid temperature stability.  There is enough mass in plasterboard and OSB for that.

I'd always promote using a constructional slab with UFH though, as that's the most sympathetic method IMO and also lends itself to ASHP 'background' cooling. It does bode well for thermal inertia too, but the biggest benefit is deffo smoothing out the heating delivery. Have had feedback from folk who have built an insulated cube ( to PH standards, eg no slab / screed / UFH ) and the results are less than comfortable when in the heating season as most have had to use the MVHR for circulating warm air ( much higher airflow rates required / increased system audibility etc etc ) and have also had to employ auxiliary heating to fortify this via 'feature' radiators downstairs and towel radiators in every bathroom which are left on 24/7. Conversely cool feet vs very warm head was one comment.

[A_L]

16 hours ago, Nickfromwales said:

"The longer the better" is the bottom line

 

I would have to partly disagree, while longer decrement delay will reduce the percentage of the energy arriving at the outside getting through to the occupied space (the decrement factor) it would be better if the heat wave did not arrive at a time when overheating caused by other sources could occur.

 

15 hours ago, SteamyTea said:

There is enough mass in plasterboard and OSB for that.

 

I disagree, I have been in too many overheating lightweight sloping ceilinged warm roof constructions. 11m OSB and 12mm Plasterboard both have Thermal Capacities close to 0.0025kWh/°C per m² .

[ProDave]

3 minutes ago, A_L said:

 

I disagree, I have been in too many overheating lightweight sloping ceilinged warm roof constructions. 11m OSB and 12mm Plasterboard both have Thermal Capacities close to 0.0025kWh/°C per m² .

That will be too little insulation, and insulation of a type (probably PIR) that has a very short decrement delay so by the afternoon the heat will be streaming through the coombed ceilings.

 

Our new build with plenty of insulation and a type (wood fibre and earthwool) that have a longer decrement delay has none of these issues.

[SteamyTea]

As Dave says, bad design and execution are not a substitute for reality.

 

[Nickfromwales]

1 hour ago, A_L said:

I disagree, I have been in too many overheating lightweight sloping ceiling-ed warm roof constructions.

Best to compare apples with apples as the OP is in their situation, not the situation of others?

1 hour ago, A_L said:

I would have to partly disagree, while longer decrement delay will reduce the percentage of the energy arriving at the outside getting through to the occupied space (the decrement factor) it would be better if the heat wave did not arrive at a time when overheating caused by other sources could occur

Solar gain through glazing is, a little behind, linear with it's presence, but heating by other means, eg auxiliary systems, is by design and deemed ( known ) from feedback of the occupants being in habitation. Those events colliding to cause overheating ( from them being so far out of phase that they overlap or meet more significantly ) would be an error that you would have to strive to create as we're only talking about a model suitable for a 24hr period at the end of the 'day'.

Seasonal adjustments may be required, but the fabric first rule resides over the auxiliary systems so the collide can be anticipated and mitigated by a wee dram of human intervention.

Edited August 13, 2020 by Nickfromwales