The weighted energy demand, part II

Yesterday's post was about the calculation of the energy demand of a house and the weighting factors assigned to different energy sources. Another factor that must be taken into account is the efficiency of the particular device, e.g. a furnace, used to produce the heat energy. A note on terminology: In the case of heat pumps, the word efficiency is not accurate and the phrase coefficient of performance[1] is used instead.

It is important to note that these are the "standard" values accepted in the Minergie system. There are heat pumps which beat these numbers and our architects have recommended one of the best performing models on the market. The COPs for the air-source heat pumps made by this company[2] are claimed to range from 3.56 to 4.2, depending on the particular model. I plan to write a post about these systems before too long. Update 23 August 2009: Instead of a heat pump we will probably use district heating. Please see post 038.

Table 3. Efficiencies or coefficient of performance (COP) for a selection of different heat generating systems. Larger is better.

	Heating	Warm (hot) water
Source	ηH	ηWW
Oil or gas furnace	0.85	0.85
Oil, condensing furnace	0.91	0.88
Wood-fired furnace	0.75	0.75
Wood pellet furnace	0.85	0.85
Waste (district) heat	1.0	1.0
Electricity	1.0	0.9
Heat pump, outside air	2.3	2.3
Heat pump, ground source	3.1	2.7
Heat pump, ground water	3.2	2.9
Photovoltaic	0	0
Solar collector	0	0

The effect of this weighting system can be seen in figure 8 below. The standard values for the weighting factors for the different energy sources have been used here to compare the different results for the same house. I've left off the numbers on the graph because I just want to give an idea of the relative differences. A process of optimization leads to the the best solution for a particular house.

Figure 8. The weighted energy demand calculated using different energy sources.

---

[1] Read about COP at this Wikipedia page → Coefficient of performance

[2] The company is Heliotherm → Heliotherm air-source heat pump

The weighted energy demand

All this talk of insulation, but I haven't yet gone over the fundamental point of Minergie which is to reduce the energy demand of buildings, in this case our house. I've mentioned several times that the limiting value for houses is 38 kWh/(m²·a). Then the question is how this energy demand is defined. Different systems of certification (e.g. the German KfW-40) have different ways of calculating this and in the Minergie system under consideration here it is a weighted sum of the energy required to
(i) maintain a comfortable indoor temperature (usually taken to be 20°C), call this Q_H,eff
(ii) to heat water Q_WW and
(iii) to run the ventilation system[1], Q_V.

We have the following relationship (click on the equation to see a larger version):

In Equation 1 the g terms are the weighting factors for the particular type of energy source chosen and the η's (eta) are (or are analogous to) the efficiencies of the devices used. You can see from the inequality that small g's and large η's are good.

More about the η's in the next post, here I'll just say a few words about the weighting factors g. This is where the differences in the different certification systems become apparent[2]. They are basically an attempt to compare the losses associated with the conversion of the energy from different sources to heat (see Table 2 below for the list). Burning fossil fuels to generate heat is taken to have a g of unity. Using the sun directly, as in absorbing the radiation and storing it as heat 'costs' nothing so it is given a weighting of 0. Using electricity, say to run a heat pump or (horror!) a resistance heater is considered least desirable (I imagine because the electricity itself is generated from other sources and there are losses in that chain of production and in the transmission). However, in the case of heat pumps this is mitigated by the ability of these devices to extract energy from the surroundings and this should be clear in tomorrow's post. The energy source to run the ventilation system is almost certainly electricity. A typical value for the Q_V is between 3 kWh/(m²·a) and 4 kWh/(m²·a).

Table 2. Weighting factors for different energy sources. Smaller is better.

	Weighting factor
Source	g
Solar, geothermal, ambient	0
Biomass (Wood, biogas)	0.5
Waste heat	0.6
Fossil fuels	1.0
Electricity	2.0

---

[1] A little aside about the ventilation system: The aspect of the building envelope that I've talked about so far, namely the insulation, deals with the loss of heat via conduction through the shell. Another very important mechanism of heat loss is the movement of warm air from (i.e. leakage) and cold air into (i.e. infiltration) the house through gaps in the shell. In a well-insulated house, this can account for upto 50% of the total heat loss. It turns out that it is possible to neutralize this effect and still have good air quality by building a very tight shell and by relying on a high-efficiency mechanical ventilation system with a heat exchanger to capture back more than 80% of the heat of the exhaust air. A topic for other posts.

[2] An entry on the German language Wikipedia compares three systems: Primärenergiebedarf

Limits on the overall heat transfer coefficient through the building envelope

I've mentioned before that the first step in reducing energy consumption is to invest in a well-insulated building shell. Just like choosing a thick and warm winter jacket. I've also stated before that in the end what is important for the Minergie specification is that the weighted (I'll talk about this weighting very soon) energy consumption per unit area be less than 0.38 kWh/m² per year. It is possible to achieve this by adjusting the different factors that play into this equation and to simplify the matter the Minergie group has come up with a set of "standard solutions". The limits given in the table below are those for the building elements. It is possible to build a Minergie house with components that exceed these values but make up the difference in other ways. In fact, not only it is possible, it seems to be often the case. This clarifies some of the confusion I had in earlier posts.

Table 1. Limits on the total heat transfer coefficient for building components.

	Separation from exterior	Separation from unheated interior	Separation from ground
	W/(m2·K)	W/(m2·K)	W/(m2·K)
Roof	0.15	0.20	-
Wall	0.15	0.20	0.20
Floor	0.15	0.20	0.20
Window	1.00	1.60	-
Door	1.20	2.00	-

Minergie vs the average building

I find it useful to remind myself from time to time what it is that we're doing that's different from conventional construction. If you haven't already, download the following flyer (in English) from the Minergie website. It seems to be a little outdated[1], but there's still useful information. Clicking on the link should start the download of the PDF file: The Minergie Standard for Buildings

The standard requires that general energy consumption must not be higher than 75% of that of average buildings and that fossil-fuel consumption must not exceed 50% of the consumption of such buildings.

So what's an average building? Rather than try to describe that, I find it easier to consider the limiting value from my third post, figure 1: The weighted energy performance value of a Minergie house must not exceed 38 kWh per m² per year. See page 4 of the flyer for more information.

---

[1] E.g. In this flyer which is edition January 2008, the U-value (the overall heat transfer coefficient) limit for the walls and roof is given to be 0.20 kWh/(m²·K), while as far as I can tell the 2009 standards set it to 0.15 kWh/(m²·K).

The wheels are in motion

Figure 7a. A rough representation of how the proposed house will sit on our land.

It was almost a year ago that we found a piece of land we liked and we initiated the process of acquiring it. We also started conversations with a selected group of Minergie-specialist builders in parallel. Back in those early days we were pretty sure that we would have one of these companies design and build our home. Their publicity materials assured us that they specialize in individual solutions and the combination of that with (i) the fixed-price guarantee they can offer and (ii) the good reputation for quality they enjoy is what interested us. As our discussions continued and we visited a number of houses they had built we realized that our individual solution would cost no less than what we could expect if we worked with an architect, with possibly a rather constrained design palette. We simply saw no examples that assured us that they would really build us something special. There was no doubt that they would build us a good, solid house, it was just that we were not convinced of their willingness to push the design envelope for us.

Figure 7b. The east elevation of our proposed house. Note how the land slopes off and the rear of the lower level is underground. The rectangular opening on the side is to the carport which is integrated into the building.

In the meantime I had compiled a short list of local architects based on examples of their work that I found either in magazines or through web searches[1] or actual houses I saw around our area. It was during this time that I decided that the websites of many architects left much to be desired[2]. In many cases the navigation is too cumbersome, the images miniscule or embedded in some fancy but slow-as-molasses display presentation, the information hardly useful. But there was one that stuck out for me. Both in the accessibility of the website and the aesthetics of the houses, especially the interiors.

Figure 7c. The north elevation. This is the side at street level. It looks a little menacing with the slits for windows. I'll say it: almost bunker-like. From the energy-loss point of view though, the north side (at these latitudes) is the worst and it makes sense to minimize glazing here. It's all made up for on the south side.

So one day at the beginning of last summer I plucked up my courage and popped into the office to make an appointment with said architect. Initially we had thought that we would want to talk with at least two architectural firms but we soon decided that we felt confident that this group would build us something lovely. So here we are now.

Figure 7d. West elevation. The door opens out from the kitchen. This side borders a public path connecting the two streets that can be seen on the map view so we'll have to have some sort of structure to provide privacy.

One of the wishes we conveyed to the architect was that we didn't want a "box". Boxes are very popular here right now, and it is true that they are more energetically favorable because the surface-area-to-volume ratio is lower than say, something enlongated. But, that's not the way we wanted to go. Actually, after our initial meeting where we set down our wants and desires and budget we were actually presented with not one but TWO very different plans. This was a very pleasant surprise. The version we decided against was also very interesting. It was more compact – it would probably have had better energy performance – and we felt it didn't make the most of the southern exposure of the land that we had. It also had a more complicated sunken-courtyard thing going on which was rather cool but we decided the simpler design would work better for us.

Figure 7e. The very important south elevation. This is where most of the solar gain will be made hence the plethora of windows. Here you can see the terraced garden on the left. The rest of the land we envision as a wildflower meadow, much as it has been until now.

It seems that issues of accessibility (as in this definition on Wikipedia) are on the minds of many people right now. It wasn't so different for us. We wanted a house where it would be possible to live on one level if necessary. So the upper floor at street level had to include a bedroom and a full bathroom.

Figure 7f. The upper-level floor plan.

A few words about the upper-level. The entrance hallway (A) can be closed off in the winter to reduce the infiltration of cold air. There's a built-in coat closet (B) with a glass inset on the stairwell side to let in light. The stairwell (C) is illuminated by a skylight. I haven't done the calculations yet (it's on my long to-do list) but my gut feeling is that a skylight provides more light per unit of heat lost through it than a similarly sized north-facing window[3]. The labels BF (Bodenfläche) and FF (Fensterfläche) refer to the areas of the floor surfaces and the window surfaces, respectively, in each room. There's a minimum ratio of 8:1 that is mandated for rooms in which people "live". As we have no real basement nor attic nor garage, we don't have many areas in which to accumulate stuff. There's a generous storage room (D) on this level which in combination with all the built-in wardrobes we have planned should cover all our storage needs and then some.

The division of the interior space was partly driven by my insistence on good natural light in all rooms, including bathrooms. Light shafts are frequently used to bring light into basement-type areas but I'm not a fan. This meant that on the lower level all the rooms had to be arranged on the south wall because all other sides lie underground, so to speak.

Figure 7g. The lower level.

The "technical" room is where the heating system, water boiler, air handling device etc will be located. It will be unheated and will lie outside the insulated hull of the house. Actually, the same is (obviously) true of the carport just above it.

---

[1] Many of the architects are listed through this website: www.swiss-architects.com

[2] I'm not alone in this. See this Guardian blog post from January of 2008: Why are architects' websites so badly designed?

[3] If you know I'm wrong, please explain why.

Warm fuzzy insulation

Figure 6. The thermal conductivities of a selection of materials typically used as insulation in construction.

Generally speaking, the effective insulating value of a wall is really determined by that of the insulation used. This was the case of the wall I discussed in my last post a few days ago. The thermal conductivities, λ, of some of the commonly used insulating materials can be seen in figure 6. You can see that they range roughly from 0.025 W/(m·K) to 0.060 W/(m·K). For any given type of insulating material there's a host of products made in different ways and they can have different λ values. In some instances, the different forms are necessary for particular applications, for example, the physical requirements for flooring components are quite different from those for walls. In the graph, the areas represented by the solid colors depict the best (lowest) and worst (highest) measured products in these groups. The lone values at the top of the columns are to be used for those particular products that remain untested by an approved facility.

At the right edge of the figure I've included some other materials of interest. Things such as porous concrete have better insulating properties compared to regular concrete but they're still a long way away from what is necessary for a Minergie-type house. Walls made of them would be too thick. In the area of bricks, there are perlite-filled ones which have very good insulating properties. Poroton-T7 is one that is actually appropriate for Minergie homes. A wall thickness of 42.5 cm satisfied the required U-value limit of 0.16[1] W/(m²·K) for the Minergie-P house BE-028-P. A blog (in German) about the construction of this house can be found here.

---

[1] If you're wondering why it's 0.16 and not 0.15 as I've been saying, I have to say that I don't know yet. I'm working on finding the answer.

A closer look at a wood wall

Figure 5. An example of a Minergie wall module. This one is by Isover.

I was not completely satisfied with the wall discussion from three days ago so I spent a little more time looking through the PDF of the Bauteilekatalog I had linked to. The big challenge of calculating the U-value for the type of construction I've shown here is that the layers making up the sandwich don't extend uninterrupted through the entire length. There is a lattice made of wood that runs through the assembly holding things in place and providing stability. Wood is a better conductor of heat (λ in W/(m·K) is between 0.13 and 0.18) than insulating materials such as glass wool (λ in W/(m·K) runs from 0.031 to 0.048) and these areas where there's wood instead of insulation constitute a leak for heat flow (a thermal bridge) from the warm side to the cold side of the wall. In poorly insulated walls they don't make a big difference but in well insulated ones they do. Unless these regions are taken into account, the U-value that is calculated is lower than the true U-value (remember that a lower U-value is better).

The walls for our house are currently being designed. In the meantime let's consider a Minergie-certified wall module. There are a number of these units from different manufacturers that are guaranteed to satisfy the requirements. This one is made by Isover[1], a glass wool manufacturer. It has a total thickness of 35.75 cm and a thermal-bridge corrected U-value of 0.15 W/(m²·K). (According to the product description the non-corrected value is 0.12 W/(m²·K) though my own calculations give 0.13 W/(m²·K) – I must be using wrong λ values for some of the components). Anyway, the breakdown of the components is as follows. I've looked up the λ values when not provided by Isover (the numbers in red) :

1. 12.5 mm at 0.25 W/(m·K) | Gipsbauplatte (gypsum board of some sort).
2. 40.0 mm at 0.032 W/(m·K) | Support lattice; installation space embedded in Isotwin.
3. 0.05 mm – too small to matter | Vapor retarder/air sealer Vario KM Duplex.
4. 15.0 mm at 0.12 W/(m·K) | OSB (oriented strand board is 95% wood and 5% binder).
5. 200 mm at 0.035 W/(m·K) | Isofix.
6. 30 mm at 0.06 W/(m·K) | Fiberboard.
7. 40 mm – not counted | Ventilation gap.
8. 20 mm at 0.14 W/(m·K) but not counted | Wood siding.

Isotwin has a λ = 0.032 W/(m·K) and Isofix has a λ = 0.035 W/(m·K).

---

[1] Here is the page where I found the information: Isover.ch 24-201

Walls of wood and concrete compared

Figure 4. Two different above-ground wall constructions. Numbered elements are: 1 ‐ Interior wall surface. 2 ‐ Support structure. 3 ‐ Vapor retarder. 4 ‐ Insulation. 5 ‐ Post. 6 ‐ Wood fiberboard. 7 ‐ Ventilation gap. 8 ‐ Exterior cladding. 9 ‐ Exposed concrete.

In my last post I claimed that we were expecting the exterior walls of our house to be a minimum of 35 cm (14") thick to achieve the required heat transfer coefficient (U-value). I had also remarked that massive construction – masonry or concrete – would result in thicker walls, a few things being equal. Well, I ran some numbers through for a couple of wall types[1] I found in a compendium[2] of such things.

First I picked out the wall types seen in the figure. Then I chose the same material (Steinwolle[3] of the fiberglass persuasion) for insulation in both cases and I increased the thickness of this until the U-value dropped to 0.15 W/(m² K). The total thickness of the wall in the case of the wood construction turned out to be 35.5 cm and 40.0 cm in the case of exposed concrete.

---

[1] There must be a couple of hundred wall element combinations possible! If you think I'm exaggerating, you can download a PDF of the Bauteilekatalog by clicking here.

[2] There are some decent free online tools with which one can calculate the U-values of different building components. I used this one here. Available in German and French. But wait, there's more! The tool also estimates the grey-energy embodied in the structures and 'environmental impact points' to help in the decision making process. More about these at some later date.

[3] Steinwolle is made from mostly natural mineral raw materials while Glaswolle has about a 70% recycled-glass content.

Limits on the overall heat transfer coefficient through the building envelope

Figure 3. A simplified representation of the heat transfer coefficient limits for a few specific building components.

In general, the first step towards reducing energy consumption in a house is to use a lot of insulation. The lower the heat transfer through the building envelope, the better. One limitation to this is that the thicker you make your shell, the smaller you make your interior area. In the case of our house, we're probably looking at a minimum of 35 cm (≡ 14") thick exterior walls. This is for timber construction[1], construction in masonry or concrete usually ends up being thicker for the same insulating properties.

---

[1] Actually, it is more appropriate to say "mixed" construction. About a quarter to a third of the shell will probably be of concrete because our building land is on a slope.