You put it so elegantly

So what evidence do you have for that statement, djh? What meaning do you attach to 'importance' ? Heat loss? Comfort? Building deterioration? Moisture penetration and air infiltration should be eliminated by good design detailing, and are unrelated to thermal mass. The thread topic raised was about thermal mass effects. That inter-relates with wind chill (bigger mass, lower cooling rate).

Steamy,

sunny Hampshire.

We are taking part in a DECC funded trial of control systems for heat pumps. The company conducting the trial for our system is Passiv Systems Ltd.

I have their agreement to post their graph showing the temperature changes in our house whilst we were away for six days last week.

In order to understand this graph, some background information:

Our house is a hybrid house, part timber frame built about 1984 and part modern block built 2006/7. Passiv Systems have installed temperature sensors in the rooms:

Zone 1 (A): Lounge (blue line) orientation NW, windows on NE, NW and SW, modern block
Sensor 1 :Old hall (blue line) interior, timber frame
Sensor 2 : New Hall (lime line)interior, modern block
Sensor 3 : Master bedroom (pink line) small window on NE, main window SW timber frame
Sensor 4: Guest bedroom (light blue line) window on NW timber frame

Apologies for the two different blue lines, which make it a bit harder to separate.

It does look like the modern block part loses heat more slowly and heats up more slowly and the timber frame part responds more rapidly to heating.


We are just over two months into this trial and have only been running the new control system since the end of February. So it is far too early to draw any definitive conclusions. However, I have found it easy dealing with Passiv Systems and their staff are very approachable and responsive. They will be looking for further triallists next year, so I will be happy to provide feedback to anyone thinking of volunteering.

I'll have the solar data from the PV, it is quite a good proxy. Sunday Mar 23rd PV peaked at over 3.3 kW and generated over 19 kWh for our 3.84 kW array. Best day for the year so far for PV.

Yes, the grey line is the external temp taken from the local Met Office (I think), there isn't an outside sensor. The bottom yellow blips are the heat pump cycling and the top two lines are the measured flow and return temps for the heat pump. The temp sensors record to two decimal places, what they are accurate to is a different question.

My view would be in a traditional blockwork house there is too much thermal mass. I've just finished a build and went with a 300mm blockwork construction with full fill insulation, but I also insulated the outside of the building with a batten and sheeted exterior finish. Inside I dry lined all the external walls. The theory being that I would still have the interior blockwork walls for the thermal mass.

The living spaces are all open plan with large quantities of South and west facing triple glazing and the passive solar works great. During the winter with the lower sun these spaces do not require any heating during the day, during the evening the spaces would rarely drop lower than 19 degrees. During the daytime with a good sunny day the space would heat up to 25 degrees, with the rooms 23-24 during the evening

Posted By: ArchmocoMy view would be in a traditional blockwork house there is too much thermal mass.

What makes you take that view?

If you only want to heat the house for a few hours in the evening then you could have too much thermal mass but is there any other reason?

Just spent the last week logging some temperatures in my thermal mass experiment.
It is two identical polystyrene boxes with two identical double glazed lids (fish boxes and small fridge doors). One has two bricks in it, the other does not. I covered the bricks in tin foil to reduce the amount of direct sunlight hitting them (test 3 may be without tin foil)
Have stuck them in the garden were they can catch the sun until 1PM. There is a bit of a difference as the sun goes into shadow on one box first.
I have since re-setup to correct some inconsistencies (added to fake bricks to the non thermal mass box so that the volumes are the same) and moved the temperature sensors to the same position in each box. Not sure how much difference it will make but last Saturday I dislodged a probe and got very high readings (have taken that day out of the data).
The boxes are .55m x .32m x .115 m, the window aperture .55m by .32m. The bricks are .21m x .08m x .1m and .2m x .08m .1m and 3.23kg and 3.83kg respectively.
I also put a temp sensor in my shed, and one on the shed wall (have since put these in a better position too).

It is worth pointing out the the 1Wire sensors are really only accurate to 0.5°C and often give misreadings (which I filter out).
I have averaged the data into half hour blocks, and charted this over the last few days, I have also added some solar and temperature data from a local weather station. This makes for a cluttered chart but shows that my data correlated well with someone else's.

Below are the basic average results:

Temp,Box,Box with ,Ambient,Shed
Mean Temp /°C,12.2,12.3,10.0,12.2
Standard Deviation /°C,6.4,4.2,2.2,2.9
Count,2478,2472,2465,2478
Standard Error/ °C,0.1,0.1,0.0,0.1
T Test (1st 6 days),0.36

And as a couple of charts, two as time series and the other a correlation.
There is more analysis to do, but seems with this set up that it limits the maximum temp, but makes little difference to the overall temperature over a day.

I wouldn't, as the variations is not so different, especially when the temperatures are within normal house ranges i.e not 40°C plus or below 15°C (this is a problem with small scale models).
I will try and filter out the extremes and see what we are left with. But if you look at the correlation chart there is little difference at the 'normal' temps.

As I said, I do need to do some more analysis (night times, daytimes, shaded, non shaded, radiation levels...).
This is just a simple analysis and this not really looking at typical data (it was sunny in Cornwall!!)

I did that so I get an initial baseline. There are more tests to be done, one will be with only one brick (the thermal mass to air ration is huge at the moment), another will be with a smaller window aperture, another will be with uncovered bricks, another with gravel instead of bricks.
It is way too easy to jump to a quick conclusion without having data. I am trying to provide data and a test methodology that gives meaningful results (within the limits of what I can do).

Posted By: Ed DaviesMaybe just put it on the outside of the ISS with 50 minute days and 45 minute nights would work and be less hassle.

Or get people with high and low mass houses to supply temperature data. Get enough places and the means will be, umm, meaningful.

Posted By: bot de pailleOne could also look through the 60+ years of academic research that has been carried out on the usefulness of thermal mass in building design

Show me the UK research about the UK climate, I really want to see it.

http://etheses.nottingham.ac.uk/1872/2/PhD_Thesis_-_Lucelia_Taranto_Rodrigues_Single_Side.pdf

Posted By: tonyOK then can I conclude that living in a high thermal mass box will be more comfortable and not get as cold or as hot as a box with no mass would?

To a simple biologist it does look from the second and third plots as if you have demostrated rather nicely the difference between light-weight (1) and heavy-weight (2) buildings ie that 1 heats up rather faster and to a higher level that 2. And given your experiement why wouldn't you.

It may not be quite right to assume the relation between the temperatures in the two boxes is linear. Could you modify the third plot slightly by giving the x and y axes equal scales? The eye is good at detecting non-linear relationships but has trouble adjusting for scaling differences.

Posted By: SteamyTeaDo you mean make the y-axis go to 55°C like the x-axis.

Not quite, something simpler - I was suggesting that your third graph be modified by plotting those temperatures using the same intervals (spatially) on each axis so that the eye does not need to correct for the wider intervals along the x axis. I assumed that each point represented a pair of measurements taken at the same time - correct? Does the graph show that the high mass box stays as lower temperatures for the same solar input?

One would expect the two sets of data to be closely correlated (both boxes exposed to the same powerful stimuli under similar conditions - overnight low temperatures followed by day time sun and high temperatures). Is the question whether from a baseline (early morning lowest temperatures) the rates of temperature change deviate in a useful way? Graph 2 suggests they do, the high mass box moderating the rate of rise and fall. So a high mass building with a low cost heat source (sun on south facing windows or a porch or some other solar collector (or Economy 7 electricity come to that) doesn't need to "dispose" of heat but can retain it for the time when only a higher cost heat source is available? This is sounding obvious - have I missed something?