Thoughts on Thermal Mass

Posted on April 23, 2017
Tags: building, green, tomar

It is worth considering what thermal mass is, and what it does.

Basics

All materials have a specific heat capacity (or shc) which is the amount of energy in Joules required to raise 1Kg by 1K. Also, all materials have a density which is the mass of 1m^3 of material. It ollows that for any given volume of material we can identiify a heat capacity (or thermal capacitance) which is the product of the shc and the mass, where we get the mass from the volume and density.

Effects of Thermal Mass

There are three ways that the heat capacity of a building element can be influential in the behaviour of a building: - interaction with its own thermal resistance and that of adjacent elements to create a decrement effect on the amplitude and phase delay of thermal oscillations - reducing thermal oscillations inside the building envelope - storing direct-gain solar or biomass energy for later release

In effect, the second and third are the same effect, but viewed from a different perspective: - in the second case we have temporary mismatch of demand and supply that we wish to smooth. - in the third case we deliberately create such a mismatch by inviting solar gain (or perhaps direct radiation from a log fire) that is hard to control, but that we would like to capture for release over an extended period.

We can also argue that there is a single effect, analogous to an RC network, where the massive element's thermal capacity is acting like a capacitor, and the thermal resistance of the element itself and the border and adjacent element is the resitence.

ISO13786 and Academic Texts

In ISO13786 we see these effects as:
  • decrement delay and decrement factor, where the element is providing the RC effect
  • aerial heat capacity, where we consider the ability of the exposed inner surface to absorb and release energy to provide storage

The former of these is discussed in a number of academic texts, including Design criteria for improving insulation effectiveness of multilayer walls.

Note that in this text - as in ISO13786 - we consider the case where the outside temperature is varying and the inside temperature is stable. Somewhat confusingly, the outside is 'the source'.

Practical Observations

In the case where the massive element is in a wall with insulation, we look for behaviour where the 'R' component is quite high, and the desired effect is one of a filter stage. In the other cases we wish the resistance to be low so that energy can flow into and out of the massive element.

If a building element has a very high heat capacity then we might expect there to be a 'high thermal mass' but this is not necessarily a practical way to consider things - if we want to smooth interior fluctuations or store and release energy on a daily cycly, then having a large thermal mass that is protected behind a layer of insulation will be ineffective.

Some observations are in order:
  • the amount of thermal resistance that is necessary to reduce the aerial heat capacity is surprisingly small: we rely on exposed surfaces with high conductivity and high heat capacity. Modest thermal resistance (such as a cavity behind plasterboard) will significantly reduce the effect between daytime and night.
  • significant thermal capacity within the envelope can still have an effect if it is insulated but the effect will be subtle and over an extended period.
  • high levels of exposed thermal capacity can smooth unwanted oscillation, but it can also make it harder for a control system to operate effectively

The impact on control systems is significant, because thermostats have a designed hysteresis which is intended to stop continuous on/off operation with a short period. Also, they are not always very sensitive even when they report temperature to significant precision (not accuracy!) - and they need to smooth out random short term fluctuations of air temperature. As a result, by the time that the control system identifies that a thermostat in a very high capacity environment is reading low, the absolute size of the energy deficit may be quite large. If the control system does not have a way to deliver energy into the environment quickly - and without overheating the air in the environment - then the under-temperature state will last for an extended period.

If the energy is delivered to the core of a massive element which then heats the environment then there is also a significant chance of 'overrun' as the temperature in the massive element and the air in the environment equalises: this can be a problem with high mass underfloor heating systems with screed where the spaces to be heated have low thermal mass.

In practice the number of days on which we would have both heating and cooling demand is rather small in a temperate European climate like the one in Central Portugal. The daily swings are higher than we are normally used to in the UK, but the amount of thermal mass needed to smooth them out does not seem to be very large.

Decrement Delay, Phase and Decrement Factor

The decrement factor resulting from mass and resistence do not substitute for thermal resistance unless the outside temperature is oscillating around the desired internal temperature: if the mean temperature is different then it allows us to disregard the peak temperature discrepancy when sizing a control system.

In extreme cases, decrement factor means that we do not have days in which we run cooling during the day and heating during the night - so in a sense it can also reduce the overall energy consumption.

If we can organise for the delay to be 12 hours then we have an 'out of phase' system where we a driving external oscillation and a resulting oscillation at the inside surface of the wall. This is useful because it means that the inside wall is at its highest temperature when the heat loss through windows and air replacement is highest, and the inside wall is coolest when the outside temperature is highest and we have solar gain and the incoming air is over the target.

This is a useful effect but we should be careful not to overstate the benefits. Walls of 'reasonable' thickness that exhibit 12 hour delay tend to be very well insulated with a commensurate decrement delay and the total energy flow is limited. Even in the case where the inner wall surface has good aerial heat capacity, the absolute temperature fluctuation on the inner surface is very small, so the temperature differential between the wall and the room is small and small amounts of energy are exchanged.

My opinion is that the decrement related effects are secondary in a building that has good static thermal resistance an is properly ventilated. Achieving a good static thermal resistance limits the heat transfer flows to such an extent that the absolute benefit that would be seen from the additional decrement factor is small.

Consider the case of a light steel frame structure that has very limited decrement effect. Such a structure will tend to have a lower U value than a conventional wall structure since - even with some thermal bridging in the inner part of the wall where the frame is - there is insulation outside the frame also within the frame elements.

A 300mm deep LSF structure with 150mm frame filled with mineral wool and 100mm neopor ETICS will can have a static U value of about 0.16W/sqm.K.

A 200sqm house with 2 floors and an 8x12.5 footprint might have 41m of perimeter and 200-250sqm of wall. For a 10C fluctuation (ie +-5) then the peak energy flow is 200W, and all the internal furnishings and surfaces are providing thermal capacity even if the building envelope provides little.

Note that the average energy input rate over the heating half-cycle is about 127W, and the average loss over the cooling cycle is also about 127W.

In all likelihood, the resulting fluctuation would not result in active heating or cooling demand on the control system.

If there is demand, it is likely to arise because of localised energy inputs inside the building, such as devices, people, andactions like cooking and washing. These local inputs are primarily addressed by aerial heat capacity anyway.

Aerial Heat Capacity

If we consider the massive components of a building structure then we tend to have quite a significant heat capacity. However, it is not entirely useful to think in such terms, because much of the capacity may be unable to exchange energy with the living space.

Over an extended period the whoe of the building fabric will tend to cool or warm based on the average internal temperature. However, the effect can be very gradual and does not directly help with reducing the control system sizing unless the mass can exchange energy with the living space.

It is not easy to plan the 'right amount' of aerial heat capacity.

The 'gold standard' might be considered high density concrete.

Internal structure Aerial Heat Capacity
dense concrete 86
dense concrete, 25mm cavity, dry lining 43
insulated lsf, drylining 23
insulated lsf, drylining, 40mm clay plaster 62
insulated lsf, drylining, 40mm dense plaster 54
ETICS and plastered thermal block 51

We see that it is possible to add effective aerial capacity to a lightweight structure using a suitable dense plaster - clay is the most effective but it does need a deep application and it can be expensive.

The key factor is that the surface that is exposed to the living space is the critical one for heat capacity and we need it to have good specific heat capacity and be as thick as is practical.

The temptation is to think in terms of the inside face of the external walls. Doing so does have some benefits in terms of decrememt behaviour, but we can also think in terms of inside walls, floors, and ceilings.

In the case where we are trying to capture solar energy that is directly incident through windows then the floor and interior wall surfaces are the most important since those surfaces will be warmed directly. In the case where we have a log burner or wood fire then the floor, walls and ceilings that are closest to the burner are most important.

In the case where we have neither - and do not have major heat sources such as cooking or washing or landry, then there seems limited benefit to adding large aerial heat capacity or in treating surfaces to add it.

How much aerial capacity is 'enough'? I don't know!

I have no great desire to have a log burner, so it is really driven by solar gains and the kitchen and laundry areas.