Jump to content

Dew Point - What is it and why does it matter


Triassic

Recommended Posts

I’ve been looking at the wall construction of my timber frame and the dew point analysis. I’ve realised I just don’t  understand what’s going on and where the dew point should be and why?

 

Could someone explain why it’s important and how to interpret the results?

Link to comment
Share on other sites

Hopefully you do not have a dew point in your timber frame! Dewpoint is 100%relative humidity and water condenses from the gas phase to a liquid on any solid material whose temperature is less than the dewpoint temperature. It can be acceptable for a dewpoint to occur in the exterior layer of a wall buildup if that layer is unaffected by liquid water and it cannot spread to water sensitive layers (e.g. masonry façade of a cavity wall)

  • Like 2
Link to comment
Share on other sites

9 hours ago, A_L said:

Hopefully you do not have a dew point in your timber frame! Dewpoint is 100%relative humidity and water condenses from the gas phase to a liquid on any solid material whose temperature is less than the dewpoint temperature. It can be acceptable for a dewpoint to occur in the exterior layer of a wall buildup if that layer is unaffected by liquid water and it cannot spread to water sensitive layers (e.g. masonry façade of a cavity wall)

The analysis suggests it’s in the middle of the insulation .?

Link to comment
Share on other sites

2 hours ago, Triassic said:

So if the video is correct the condensing surface is on the back side of the OSB sheathing. If so, the OSB and the breathable Membrane should allow the passage of any moisture to the outside.

 

Capillary action will allow water to transfer inwards and affect the timber studs and depending on its type the insulation. Although OSB3 is designed for humid conditions this does not mean it is unaffected by liquid water, it will be.

 

As Ed Davies please post the analysis or at least a detailed description of the wall build up.

Edited by A_L
minor typo
Link to comment
Share on other sites

@A_L, think of the heat equation: the inner skin is above the dew point, the outer skin is often below the due point, so the due point will be somewhere in the wall profile.  So what?  Our frame has its airtightness membrane on the inside and a breathable membrane on the outside.  There is no flow of (moist) air so no moisture condensing out.  Any moisture within the profile will slowly evaporate off during the summer.

 

What you don't want to occur is having the dew point or below at any surface where there is air flow, as this will create a condensing surface.

  • Like 1
Link to comment
Share on other sites

@TerryE 'the heat equation'?,

 

5 hours ago, TerryE said:

the inner skin is above the dew point, the outer skin is often below the due point, so the due point will be somewhere in the wall profile

my instinct would be to design a wall profile where this did not occur or only occurred at an insensitive part.

 

you said

5 hours ago, TerryE said:

Our frame has its airtightness membrane on the inside and a breathable membrane on the outside

is your air tightness membrane also to any degree a VCL?

 

5 hours ago, TerryE said:

There is no flow of (moist) air so no moisture condensing out

Bulk air movement is not a prerequisite for water molecule movement through a wall profile. Vapour diffusion unless inhibited by a VCL, can transport H2O molecules without bulk air movement.   Interstitial condensation will potentially occur anywhere the vapour pressure exceeds the saturated vapour pressure.

Link to comment
Share on other sites

20 hours ago, A_L said:

my instinct would be to design a wall profile where this did not occur or only occurred at an insensitive part.

 

Physics trumps instinct, and I can't say that I am sorry to say it either: the sun seems like it revolves around the earth but we (or nearly all humanity) accept that it is is v.v.  The dew point has to occur somewhere within the profile when the outside is below the dew point and the inside is well about it.  Given that you want to live at the latitude that you do and you want the inside of your house to be a comfortable living temperature, then you don't really have any alternative.

 

20 hours ago, A_L said:

is your air tightness membrane also to any degree a VCL?

 

Yes

 

20 hours ago, A_L said:

Bulk air movement is not a prerequisite for water molecule movement through a wall profile. Vapour diffusion unless inhibited by a VCL, can transport H2O molecules without bulk air movement.

 

That's what happens to dry any residual moisture out if the TF internals.  The gas laws still apply.  What you don't want is a condensation pump. 

 

Edited by TerryE
Link to comment
Share on other sites

8 hours ago, TerryE said:

The dew point has to occur somewhere within the profile when the outside is below the dew point and the inside is well about it. 

 

I am not so sure. If we take air at 20°C and 65%RH then the dew point as I understand it is at about 13.2°C. Yes a temperature of 13.2°C will occur in a profile with an internal temperature of 20°C and an external temperature of say 0°C. However by the time we reach the point in the profile where the structural temperature has fallen to 13.2°C vapour resistance has lowered the vapour pressure and thus the temperature at which dew will form.

 

Can we use this example? The first image gives the construction details, the second tabulates the structural temperature and dewpoint temperature through the profile, the third shows this graphically.

 

 

TerryE0.jpg

TerryE1.jpg

 

Edited by A_L
simplify 3rd image
Link to comment
Share on other sites

I think @TerryE is technically right: there has to be a region of the wall below the dew point under some conditions because we know dew forms. However, if the wall is properly designed (for 100% outdoor RH) then that region will just be a small layer near the outside and the condensation in it will evaporate away soon enough to not be a problem so for practical purposes I think @A_L's view is more useful.

 

If there's a dew point under reasonably average conditions (just about to be condensing outside) then there's a real problem.

Edited by Ed Davies
Link to comment
Share on other sites

My concern has always been how a structure will behave under rapidly changing conditions, as I'm not sure that the tools often used for construction modelling can account for this properly.

 

The situation that springs to mind is a fairly typical one for this time of the year, where a wet day may be followed by a very cold, clear night, which is then followed by a bright, sunny morning.  The wet day may result in there being moisture on, or in, the outer skin.  The cold night then lowers the temperature of the wall.  The warm sun the following morning then heats up the wet outer wall and, because the core of the wall may well still be pretty cold, there would seem to be a risk that water vapour (from the near-saturated layer of air in the cavity) may condense out at some point deeper in the wall.

 

Whether that then turns back into vapour and moves out of the wall structure depends on how quickly enough heat can travel through the wall so as to cause it to evaporate, as well as how vapour permeable the structure is outboard of where moisture may have condensed.

 

I have a suspicion that this sort of dynamic effect may act to "pump" moisture into a condensing region deep in the wall, that may remain too cool to allow the moisture to evaporate out again before the next cold overnight period.

 

I've not seen any interstitial condensation model that seems to allow for dynamic changes like this, as to do so requires more data than just the relevant material U values, the heat capacity of each layer within the wall also needs to be taken into account, as that will determine how long each element takes to heat up or cool down under any given set of changing conditions.

 

Having spent years modelling the dynamic conditions experienced when weapons hit the water at relatively high speed, I'd say that this is probably a pretty difficult thing to model accurately.  The big question that I've never seen answered is whether or not it's worth modelling rapid changes like this.  My gut feeling is that it may not have been for conventional construction methods, but may well be for new methods of construction.  My reasoning is partly driven by the apparent lack of thought given to the design of some standard details for things like SIP construction (and I know this has now been reconsidered by at least one SIPs supplier).

Link to comment
Share on other sites

@A_L your figure might to a computer generated graph, but it makes no sense to me.  Heat flow through a wall cross-section is approximated by a 1-D Fourier equation.  In the thermal gradient is proportional to 1/Rvalue so if your three infill materials are all similar mineral wool slabbing, then the gradients of all three are the same, but they aren't on the graph with the middle layer being about 30-35% steeper.  Why?

 

Ditto the dew point temperature.  This makes absolutely no sense to me.  The dew point is largely a function of RH, so no way would this again have a piecewise linear profile.  @JSHarris Jeremy, can you make sense of this?

Edited by TerryE
Link to comment
Share on other sites

1 hour ago, TerryE said:

but they aren't on the graph with the middle layer being about 30-35% steeper.  Why?

 

The middle one doesn't have 12% softwood.

 

2 hours ago, TerryE said:

The dew point is largely a function of RH,

 

Yes it is, but it's much more directly a function of the absolute humidity, the water vapour density.

 

2 hours ago, TerryE said:

Ditto the dew point temperature.  This makes absolutely no sense to me.  The dew point is largely a function of RH, so no way would this again have a piecewise linear profile. 

 

I think you're right, the dew point temperature will curve within the spans of materials with the same properties. However, in the past I've convinced myself that it'll curve downwards (sag) so approximating it with a straight line is safe (conservative). If there isn't a dew point problem on either side (interface) of the material then there won't be one anywhere within the material. Therefore, for diagrammatic purposes approximating it with a straight line is understandable. But then, with a bit of computer power to hand it does seem a pity not to calculate it every 10 mm or whatever.

  • Thanks 1
Link to comment
Share on other sites

2 hours ago, TerryE said:

so if your three infill materials are all similar mineral wool slabbing, then the gradients of all three are the same, but they aren't on the graph with the middle layer being about 30-35% steeper.  Why?

 

@Ed Davies has just beat me to it, layers 3 & 6 in the first attachment have 12% timber and thus average to a lower value of thermal resistance per unit thickness than layer 5

 

similarly the vapour resistance per unit thickness is higher for the 12% timber layers 

 

 

Link to comment
Share on other sites

The condensation analysis normally ignores the effect of of the timber frame proportion - both thermal and vapour properties. I don't think the Glaser method is sophisticated enough to include these 2 items. I cannot find the reference section in the BS but the software I use certainly ignores it.

Link to comment
Share on other sites

16 hours ago, Ed Davies said:

The middle one doesn't have 12% softwood.

 

OK, I missed the previous table to the graph.  The higher R-value is consistent with this. Strike this one :)

 

16 hours ago, Ed Davies said:

 

18 hours ago, TerryE said:

The dew point is largely a function of RH,

 

Yes it is, but it's much more directly a function of the absolute humidity, the water vapour density.

 

No.  The dew point is occurs at the point where the RH is 100%, though the RH is a function of the AH and the temperature You can talk in terms of RH, or AH and T, but not AH alone.

 

16 hours ago, Ed Davies said:

I've convinced myself that it'll curve downwards (sag) so approximating it with a straight line is safe (conservative). If there isn't a dew point problem on either side (interface) of the material then there won't be one anywhere within the material. Therefore, for diagrammatic purposes approximating it with a straight line is understandable.

 

Re your first statement: EH???  Sorry Ed, you've lost me on this one.  (The reason for the robust response is that you are someone that I know that I can have a useful debate around this.)  Can you cite any physics or engineering references that support your conviction?   I've been trying to get my head around this on and insulated TFs will broadly suffer three types of water ingress:

  • Direct leakage of water in liquid phase as a result of failures in weather sealing.  The moisture gradients will typically follow some 2-D dispersion from the actual water path, but this is entirely avoidable with proper detailing in the design and construction.
  • Transit of moist air through the TF as result of failures in the airtightness barrier, e.g. an air pathway from a hole in the inner airtighness layer out through the TF.  Such systemic air flow can end up dumping lots of air into the frame if undetected -- which is why houses of this class should be properly air-tightness tested and any paths identified and remedied.  This time moisture gradients will typically follow primary a 1-D deposition alone the air path depending on the moving dew point with some 3-D dispersion from this path.
  • Transit of moist air through the TF as a result of internal convection cycles.  This can be a real issue with TFs using loosely fitted slab (e.g. PUR) insulation, but this rarely an issue with blown insulation or well packed wool.  Even if there is some internal convection, then you still need an external source of moisture to transport this along the convective path.
  • Internal absorption transport.

I think that we are really discussing this last mechanism, and I am trying to get my head around it.   With something like an animal fibre (wool) based model with no material transit (from the fist three points), this should be treat as a largely static air, IMO, in that there are vapour paths to allow diffusion, but the extremely high internal surface area (the crenelated keratin wool hair surfaces) to volume will prevent any bulk flow.  Can you think of or cite why Raoult's law, etc. would not apply here?  The overall gas mix and pressure across the air-infill will be constant.  I can't think off the top of my head why the AH of any water vapour in the interior would be anything other than uniform, so the RH will vary across the profile in lock with the equilibrium vapour pressure at the corresponding temperature.  What causal mechanism could there be to create any gradient here, and even if we found a boundary value differential, what is the causal mechanism for the straight-line fit -- other than the "accountant's rule": pick any two fixed points and the answer is a straight line between them.   We would need some physical mechanism for pumping, transporting, a gas (water vapour) in a convection free medium.  One boundary is an air-tight membrane and therefore unconstrained in terms of moisture, the other is a tenting fabric between air in an airgap and some frame covering such as panel vent.  What have I missed here?  I am pausing due to brain-freeze :) 

 

 

Edited by TerryE
Link to comment
Share on other sites

3 hours ago, TerryE said:

No.  The dew point is occurs at the point where the RH is 100%, though the RH is a function of the AH and the temperature You can talk in terms of RH, or AH and T, but not AH alone. 

 

The dew point is the temperature that a parcel of air needs to be cooled to so that condensation starts; strictly, the temperature at which condensation starts on a flat surface of pure water. In other words, the temperature at which the RH reaches 100%. That's the dew point shown (with a linear approximation/interpolation) in the graphs @Patrick and @A_L posted.

 

Of course you can talk about absolute humidity alone. It's just the amount of water vapour present, typically measured in g/m³. If you have a jar of moist air and warm or cool it without evaporation or condensation happening then the absolute humidity will stay the same.

 

It's the RH which is dependent on the temperature. If you take a parcel of air and heat or cool it without condensation or evaporation then the RH will change.

 

3 hours ago, TerryE said:

I think that we are really discussing this last mechanism,

 

Agreed. That's the major limitation of this sort of condensation risk analysis, it only really takes the flow of moisture from the higher-vapour-pressure side (usually the warm, indoor side) into account. However, looking at that flow does give you some idea of the robustness of the structure to dry itself out from the other types of water ingress so long as they are not too sustained.

 

3 hours ago, TerryE said:

With something like an animal fibre (wool) based model with no material transit (from the fist three points), this should be treat as a largely static air, IMO, in that there are vapour paths to allow diffusion, but the extremely high internal surface area (the crenelated keratin wool hair surfaces) to volume will prevent any bulk flow.  Can you think of or cite why Raoult's law, etc. would not apply here? 

 

I'm not sure where you're going with this but I think you're making it all way too complicated.

 

3 hours ago, TerryE said:

can't think off the top of my head why the AH of any water vapour in the interior would be anything other than uniform, so the RH will vary across the profile in lock with the equilibrium vapour pressure at the corresponding temperature.  What causal mechanism could there be to create any gradient here,

 

The mechanism driving water vapour through the wall is the difference in partial pressures. In a typical British house (not, say, an air-conditioned one in Washington DC) most of the year the partial pressure of water vapour indoors will be higher than that outdoors despite the fact that the relative humidity indoors is usually a bit or a lot lower than outdoors. The partial pressure will therefore drop from the inside to the outside depending on the vapour resistivity of the materials following “Ohm's law” in the same way that the temperature drops depending on the thermal resistivity of the materials. We have a flow of water vapour (probably µg/m²/s) and a flow of heat (J/m²/s or W/m²) each dependent on the wall's overall vapour and thermal resistance and the relevant “forces” pushing them through: pressure and temperature difference. The gradients in partial pressure and temperature across the various layers of the walls are then proportional to those flows and the relevant resistivities of the materials.

Edited by Ed Davies
Link to comment
Share on other sites

  • 2 weeks later...

@Ed Davies Ed, I've spent this last week pondering this one on and off.  First, you say "the temperature at which condensation starts on a flat surface of pure water", and I talk about the "equilibrium vapour pressure": there is a shared understanding here, but we are simply using different terminology to describe the same mechanism.

 

Returning to @Triassic's original Q: "what is dew point and why does it matter", we can agree on a definition of the due point, but in my mind the issue is really "when does it matter?" and in my mind what this comes down to is when there is sufficient mass-flow of moist air to create a systemic condensing out of water vapour into its liquid phase on or within the fabric of the build.   Direct leakage of water is always going to be a problem and dew point is irrelevant to this.  My thesis is that dew point is only relevant when you have mass flow of air, as this is an essential precursor to the delivery of moisture for transfer to the liquid phase at the dew point.

 

So this mass flow scenario and dew point is relevant in the case of MVHR design which is why units include condensate collection and waste removal.  It might also become an issue for leaky houses where there are material exhaust flow paths from the interior though the insulation fabric, but for airtight walls the only possible transport is gas diffusion. 

 

In the case of closed cell insulation mediums such as PUR/PIR, the very closed cell isolation of insulating gasses is core to their insulation performance: they typically embody a hydrocarbon pentane as the filler gas that has roughly double the R-value of air which is why these materials can claim roughly have the U-value of the same depth of an open cell material such as wool which relies on still air as its main insulating material.  However there could be clear flow and circulation paths around slab insulators if poorly fitted and not sealed, but whilst packed materials such as wool and blown cellulosic filler rely on air as the insulator, they are also intrinsically void filling and prevent mass flow: the only transport mechanism is diffusion.

 

In our case we have a Larson strut TF with metal wall ties to an outer stone leaf.  In cases of driving rain I could image scenarios where the gap-side surface of the stone skin becomes wet by driving rain through the odd pointing gap, so the air-gap could set to saturated water vapour pressures relative to the temperature of the stone skin, but there are only pathological circumstances when the inner panelling will be at a lower temperature than the stone skin to cause a condensing transport of water across the air-gap.  

 

Any surface water on the tenting surface of the frame will cause liquid phase transport of water into the frame structure, but can you think of any circumstances dew point mechanisms (vapour -> liquid transition) will cause material transport of water into the frame?  I can't. 

Edited by TerryE
Link to comment
Share on other sites

On 12/03/2019 at 14:35, TerryE said:
On 11/03/2019 at 22:02, Ed Davies said:

I've convinced myself that it'll curve downwards (sag) so approximating it with a straight line is safe (conservative). If there isn't a dew point problem on either side (interface) of the material then there won't be one anywhere within the material. Therefore, for diagrammatic purposes approximating it with a straight line is understandable.

 

Re your first statement: EH???  Sorry Ed, you've lost me on this one.  (The reason for the robust response is that you are someone that I know that I can have a useful debate around this.)  Can you cite any physics or engineering references that support your conviction?

 

I've been pondering, too, and I'm now a lot less convinced of that point. I now think that actually the dew point line should curve upwards and it is possible to have condensation in the middle of a layer without there being a condensation risk at the interfaces to the adjacent layers. I've been trying to think of a succinct way of describing that.

 

3 hours ago, TerryE said:

First, you say "the temperature at which condensation starts on a flat surface of pure water", and I talk about the "equilibrium vapour pressure": there is a shared understanding here, but we are simply using different terminology to describe the same mechanism.

 

Yes, they're closely related but not quite the same. The equilibrium vapour pressure (EVP) is the pressure at which condensation starts if you increase the vapour pressure at constant temperature whereas the dew point is the temperature at which condensation starts if you decrease the temperature at constant pressure.

 

img_1862-small.jpg.5be8bdd111db96d908b3b3fbecaa5814.jpg

 

The rest of what you write seems quite reasonable to me but I'm not sure where you're going with it. Is it that you think that diffusion alone cannot transport sufficient water into a structure to cause problems?

Link to comment
Share on other sites

@Ed Davies you seem to be implying some form of hysteresis in this.  Can you cite a basis for this?  I can see possible mechanisms for supercriticality, etc, but I wouldn't think that they are critical in the real world.  As to EVP vs Dew Point: clearly they aren't the same as one is a pressure and one a temperature, and the only occasion that they are locked is at the triple point, but for a given material such as H2O, they are directly related.  But the EVP for a given temperature is the pressure at which the rate of evaporation from a liquid/gas interface is the same as the condensation rate: there is no net evaporation.

 

Coming back to @Triassic's OP and Q, my assertion is that the dew point is largely irrelevant as far as the interior of a well designed insulated wall: so long that there are no construction flaws which lead to internal circulation or mass air flow through the wall profile.  For a given temperature profile, the dew point can only be well defined if the AH of the water vapour content can be well defined.  For an open celled medium such as wool or cellulosic filler with an interior air tightness membrane, I can't think of any mechanism why the AH should have a systematic gradient as the gas laws would tend to self-level this subject to diffusion rates, and you'd need some form of forcing or pumping mechanism to sustain a gradient.  At worst the cavity will be at dew point and since there almost always is a positive thermal gradient through the wall profile, the entire profile should be above the dew point IMO.  The only possible interface where this could occur in practice is where there is discontinuity -- e.g. of AH between the room environment and the wall interior at the airtight membrane.  And this could lead to surface condensation on such surfaces.  But this isn't an interior issue.

Link to comment
Share on other sites

  • 1 month later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...