GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Guest Blogs

A Lesson From the Kranichstein Passive House

Applying optimizing tools can yield some surprising results

The Kranichstein Passive House in Darmstadt, Germany, where the International Passive House Conference is celebrating its 20th anniversary.
Image Credit: Photo courtesy of Bronwyn Barry

The global Passive House community is converging on Darmstadt, Germany, this week to celebrate the 25th anniversary of the Kranichstein Passive House and the 20th anniversary of the International Passive House Conference.

A great deal has changed in the world of building performance since the Kranichstein Passive House was first built. Much of that change we can attribute directly to the influence of this iconic building and the spread of the Passive House standard (which was derived from the efforts of Dr. Wolfgang Feist and others to model and optimize this particular building.) Triple-pane windows are now readily available in both Europe and North America, as are heat-recovery and energy-recovery ventilation units. A growing awareness of the benefits of airtight construction and detailing are also largely attributable to the global spread and influence of the Passive House standard.

My motivation for studying the Kranichstein building was largely coincidental. It grew out of my involvement in developing an optimization tool to help Passive House designers and energy modelers. The tool is called PDT-Passivhaus, or Predictive Design Technology for Passivhaus. This led me to dig a little deeper into how the Kranichstein project was designed and modeled.

My explorations revealed much about the evolution of the Passive House standard and my hope is that both the development of this optimization tool and information it has uncovered will be both interesting and helpful to others.

Why optimization?

Since beginning my journey into the complexities and nuances underpinning the Passive House standard, it became apparent that the habit of optimization for building design is not common practice. (Optimization is the search for the best combination of components, areas, and assemblies to achieve the comfort and performance criteria of the Passive House standard in an economic manner.)

Many projects that have met the Passive House standard have clearly done so by being “shoehorned” into fitting the criteria, often by throwing large sums of money into expensive assemblies and components, rather than by optimizing the design to creatively meet the performance criteria. I, too, am guilty of this. It’s easy to not take the time to review a project more carefully to search for opportunities to improve my designs, particularly in a benign climate like California. Once you’ve managed to make your building comply with the performance metrics of the standard, why explore further options?

When approached by a team of software engineers, looking for an opportunity to use their optimization algorithm for building energy modeling, I was intrigued. The PHPP had yet to offer an optimization process or a clear method for easily enabling building designers to quickly and efficiently explore thousands of iterations. The opportunity to help create this was highly appealing.

Why Kranichstein?

Quite conveniently, every copy of the PHPP software is supplied with a complete example project file of the Kranichstein Passive House building as the example project. We used this example file to develop the early versions of our optimization tool. It gave us a deep appreciation for both the complexity of the Passive House Planning Package and the effort that was taken in the design of the Kranichstein project.

Figure 1: Multi-variable optimization sample graph

The evolution of our optimization tool led us to develop two functions. The first is the ability to explore a range of options for a number of single variables whilst all other variables remain fixed. We’ve named this “multi-variable optimization.” This option results in a simple graphic output, plotting each variable’s parameters on the X-axis against four Y-axis graphs of Heat Load, Heating Demand, Cooling Demand, and Primary Energy.

Figure 2: Full optimization sample graph

The second user option enables the user to select a number of variables, define their individual parameters, and then have these all run simultaneously against each other to calculate the best performing combination of each of these variables against each other. We’ve called this option “full optimize.” The output of this run is a sine wave graph/plot, combined with a downloadable data table, listing their combinations numerically.

After looking at the option to allow users to select any cell within the PHPP to explore as a variable, we elected to provide a carefully curated a set of variables that we found to have the largest impact on building performance. These would help our users to quickly and efficiently optimize their designs without wasting time on selecting insignificant and variables that would not make improve building performance.

What we found at Kranichstein

When running the Kranichstein project through our optimization tool, it became immediately apparent that every single one of the variables we selected was perfectly and precisely optimized to the Passive House standard’s heat load target of 10 W/m².

It also became apparent that its more widely known certification alternate, the heating demand target of 15 kWh/m²yr, was not as significant. Despite there clearly being a correlation between the 10 W/m² target and the heating demand target of 15 kWh/m²yr, it was clear that heating demand played a minor role in the design and selection of the Kranichstein assemblies and components.

All assemblies and components used in the Kranichstein building hewed very closely to the 15 kWh/m²yr target, with the exception of the specific window areas for each elevation. While the sum total window surface area of the building readily tracked both the heat load and heating demand targets, individually the total window areas for the south, north, and west orientations nailed the heat load target, but were clearly well below the 15 kWh/m²yr heating demand metric.

Figure 3: Multi-variable outputs for Kranichstein assemblies, areas and components.

Looking at other projects

This exploration of Kranichstein in isolation wasn’t as interesting until I applied the insight it gave me to the design and optimization of other buildings. I’d always been troubled by the post by the owner of the Blue Heron EcoHaus in Saskatoon, published at Green Building Advisor.

In this post, Kent Earle generously shared the specifications and Hot2000 energy model predictions for his project. While it seemed like he may have originally hoped to meet the Passive House standard, his project did not meet the standard’s rigorous target metrics. He conceded that while building to the Passive House standard in Saskatoon may indeed be possible, “you’d be looking at making huge financial investments and sacrificing comfort” to achieve it.

From looking at his building design, I wasn’t so convinced. Given the fact that much of the research that undergirds the calculations within the PHPP has been derived from projects (for examples, the Saskatchewan Conservation House) located in right in his back yard and in other climates very similar to Saskatoon, I undertook some exploration of my own.

Figure 4: Blue Heron EcoHouse, first PHPP run, before optimization.

The first run of the PHPP model I developed for the Blue Heron EcoHaus confirmed similar results to what Mr. Earle had shared from the Hot2000 model run on his building. My first stop to find clues on how this building could have been optimized to meet the Passive House standard took me to the Energy Balance Graph. This gem is carefully hidden on the annual heating sheet of the PHPP. It readily revealed three obvious opportunities to optimize this design.

Windows were easily the biggest energy losers in this design, with 23.5 kWh/m²yr literally being thrown out the window. Ventilation losses via envelope leakage and recovery efficiency were next at 14.6 kWh/m²yr, closely followed by losses through the exterior walls to ambient of 14.3 kWh/m²yr.

Figure 5: Energy balance graph from Blue Heron EcoHaus PHPP

By running this project through a full optimization run, selecting the individual assemblies, total window surface area, average window U-factor, airtightness benefit, and ventilation recovery efficiency, I was able to quickly determine how to meet the Passive House standard for this project. The results indicated that my best opportunities lay in:

  • Reducing the total window surface area from 30.45 m² to half that amount.
  • Increasing the thermal performance of the windows from 1.24 W/m²K to 0.62 W/m²K.
  • Increasing the airtightness of this building from 0.6 ach50 to 0.2 ach50.

These three improvements allowed a reduction in the walls, roof and sub-slab insulation. This reduction in the thickness of the walls would have provided a few additional feet of interior usable space. Most significantly, the reduction in total window area would have halved the window costs on this project – an expense that is typically one of the highest component costs for any high-performance building. In addition, halving the window area would drastically increase the interior comfort of this building, and most importantly remove much of the very real risk of overheating for a project in this location with such a large number of unshaded east and west-facing windows.

Figure 6: Blue Heron EcoHaus before and after thermal values, areas, and airtightness

As with the Kranichstein building, once the three changes listed above were implemented, this project met the Passive House certification criteria via the heat load target metric of 10 W/m². It also hypothetically shared the same airtightness target of 0.2 ach50. (This more stringent airtightness metric makes good sense for buildings in climates with extreme temperature differentials such as Saskatoon, where sub-zero winter temperatures create a massive pressure drive through any holes in the building envelope.)

Figure 7: Blue Heron EcoHaus PHPP verification (after optimization).

Alternate design choices in Saskatchewan

The optimization options I found above that hypothetically allow the Blue Heron EcoHaus to meet the Passive House standard are only three of the many available to those who choose to explore their options. It was highly gratifying to discover that the early pioneers of superinsulated homes in Saskatchewan had recommended many of the same choices indicated by the optimization runs of both the Kranichstein Passive House and the Blue Heron EcoHaus.

In a paper submitted by Robert S. Dumont, Robert W. Bresant, Grant Jones, and Rod Kyle and presented at the SESCI Conference in 1978 in London, Ontario, just up the road from the Blue Heron EcoHaus, the authors make the following recommendation in their summary: “For conventional light-frame construction using gypsum wallboard as the interior finish, and no additional thermal mass, one should limit the south-facing window area to less than 8% of the floor area of the dwelling. Additional window area will only result in excessive heat gain during the day and too rapid temperature falls at night.”

They also offer an alternate to the quad-pane windows I’ve modeled in my hypothetical design above. “Thermal shutters can be of significant value in reducing both the heat loss from dwellings and in moderating the temperature falls at night in well insulated dwellings.” Notably, the Saskatoon project monitored in this study had quadruple-glazed widows of almost the same total 32.5 m² window area as the Blue Heron EcoHaus.

A great many lessons and conclusions may be drawn from the study of these three particular buildings. By virtue of the optimization tool that I’ve been able to utilize, I’ve isolated what I believe to be a clear thread of links that connect them:

  • The Saskatchewan Conservation House focused on managing losses rather than maximizing gains. It accomplished this by not over-glazing the house and by adding exterior insulating shutters. This resulted in a building that experienced very even interior temperatures – a hallmark of a well-designed Passive House.
  • When this same concept of managing losses was applied to the Blue Heron EcoHaus, it was able to hypothetically meet the Passive House standard via the 10 W/m² heat load certification metric. In Saskatoon, at this heat load target, the heating demand number fell between 25 and 27 kWh/m²yr, depending on the specific design choices made for the project.
  • The Kranichstein Passive House was clearly optimized and designed around the 10 W/m² heat load target. For that project location in Darmstadt, the heat load target resulted in a heating demand number of 14 kWh/(m²yr), indicating that the heat load target is the preferred target metric for optimization in more varied climates.

Bronwyn Barry is a Certified Passive House Designer and the co-president of the North American Passive House Network.


  1. GBA Editor
    Martin Holladay | | #1

    Response to Bronwyn Barry
    Thanks very much for your contribution to GBA. The optimization tool sounds like it will be a useful tool for Passivhaus designers.

    That said, I have a few comments on your observation that Kent Earle might have met the Passivhaus standard if he decided (among other changes) to alter his airtightness goal from 0.6 ach50 to 0.2 ach50.

    You're right, of course, that changing the airtightness number on your spreadsheet effects the changes you seek. But it's one thing to change a number on a spreadsheet and to observe the mathematical result; it's quite another thing to achieve 0.2 ach50 on the job site.

    Is it possible to achieve 0.2 ach50? Of course. Several builders have. Is it easy? Not particularly.

    If I were asked to bid (as a builder) on a job with a binding obligation to achieve 0.2 ach50, I would decline the job with a tip of the hat and a polite, "Good luck."

    Tracking down those last few leaks -- the leaks that prevent a house from achieving 0.2 ach50 -- is not an easy task, especially if the needle is stuck on 0.4 ach50.

    Pretty good house advocates would raise another objection, noting that your definition of "optimization" is skewed by your objective (namely, nailing the Passivhaus target). It's far from clear that insisting on a goal of 0.2 ach50 shows that the design is "optimized." You may have the plaque -- but the effort to go from 0.6 ach50 to 0.2 ach50 won't yield enough energy savings to justify the sweat of the builders hired to meet the goal.

  2. Expert Member

    What is at stake?
    If Kent had decided to alter his designs to meet Passive House standards what are we actually taking about in benefits? it would be interesting to know what the expected savings would be over his quite aggressive Pretty Good House approach, so an owner could decide whether living in a house where the design choices, like window size, are compromised is worth the sacrifice.

  3. GBA Editor
    Martin Holladay | | #3

    Anticipated savings
    Without undermining Malcolm's question, which is quite relevant, I'd like to say that I'm looking for a more specific number than Malcolm is looking for.

    I'd like to know how many kWh of heating energy are saved annually (according to PHPP) by making just one of your suggested changes -- the change from 0.6 ach50 to 0.2 ach50.

  4. STEPHEN SHEEHY | | #4

    If you reduce the window area by almost 50% but increase the U factor from U.22 to U.11, would you really cut the window expense in half? Don't much more efficient windows cost a lot more?

    Even more importantly, one of the Blue Heron house's best features is the amount of light the windows let in. I suspect the house is a pleasure to live in, which is more important than whatever meager savings that might be realized with only half as many windows.

  5. GBA Editor
    Martin Holladay | | #5

    Response to Stephen Sheehy
    I think that you are right about the likely cost of U-0.11 windows. (You'll notice that near the end of her blog, Bronwyn refers to these 0.11 windows as "the quad-pane windows I’ve modeled in my hypothetical design.")

    Concerning window area, there are two possibilities that affect whether Bronwyn's suggestion would make sense to hypothetical homeowners (homeowners considering a design like the one developed by Kent Earle). Possibility #1 is that the homeowners thought that the large area of south-facing windows was necessary for energy performance but wasn't needed for aesthetic reasons. If this is the case, the homeowners (hypothetically) should consider Bronwyn's suggestion about reducing the window area. (In fact, many passive-solar homes have too much south-facing glazing -- Bronwyn is right about that.)

    Switching from U-0.22 windows to U-0.11 windows is a whole different story; I think that it would be hard to justify the cost of quad-panes.

    Possibility #2 is that the homeowners have exactly the amount of glazing that they want -- they love the light and enjoy living in a light-flooded space. If that's the case, then the window area reduction suggested by Bronwyn isn't worth considering.

  6. vensonata | | #6

    insulated window panels
    I see a mention of insulated panels in the conservation house. It would be interesting to see if Bronwyn simply added R 30 panels to every window and left everything else as is. I think the house would have come very close to passivhaus. Especially because of the long winter nights of 16 hours corresponding to the coldest time of year, the panels do an amazing job. I have been tinkering with my own panels and find soft open cell gasket to be best with a tight mechanical latch to uttterly prevent frosting. After more than 3 years they seem very worthwhile and no trouble to close. Exterior shutters on bathroom windows can simply be left closed for the winter.

  7. Dana1 | | #7

    Then there's the unstated site-sourced energy factor.
    Is the lifecycle cost of the energy demand reduction of moving to very-high performance windows financially rational against the (still falling) lifecycle cost of energy from rooftop photovoltaic solar at 2015 pricing?

    Unlike window and insulation cost to performance ratios, the installed-cost/performance of rooftop solar keeps dropping 8-12% per year. It's a technology with a significant learning curve that becomes cheaper as production volumes grow. It's arguable that super-performance windows would have a comparable curve, but it's not at all clear that it's competitive with PV now (I suspect it isn't) or that it's learning curve will ever catch up. Insulation is a mature industry already in very high production volumes, without a lot of room for major installed cost reductions. So it's really PV vs. super-windows. My gut tells me PV is ahead by something like an order of magnitude, even without being leveraged by heat pumps.

  8. STEPHEN SHEEHY | | #8

    Repky to Dana
    Once a window is good enough to sit beside on a cold day and not notice a chill, that's probably a pretty good window, appropriate for a pretty good house. Investment in PV after that point may make more sense than spending for "better" windows. Obviously the sweet spot depends on climate.

  9. jackofalltrades777 | | #9

    Triple Pane Windows
    Some ballpark figures for economical PVC windows from a company like Intus is around $20 per square foot for the triple pane fixed window (Arcade Series). These are heavy duty well-designed and German engineered windows. While a cheapo double pane window from Jeld-Wen runs around $17 per square foot. Quality is night and day difference between a German-Engineered window like Intus vs a big-box store Jeld-Wen window. The $3 difference is money well spent for a heavy duty triple pane window like Intus.

    There are plenty of double pane windows in the $50+ per square foot range when you option them out. I've seen plenty of Pella and Marvin quotes for double pane windows in the $50 per square foot range. Pella double pane window costs run from $30 - $100 per square foot. All depending on the options you choose.

    So one has to be careful to compare triple pane window pricing with double pane window quotes. There are homes being built right now with double pane windows that cost 2x to 5x more than a triple pane window but the homeowner chose the more expensive, less efficient window because of brand loyalty. Most homeowners in the USA today don't look at efficiency of the materials being used. They look at the eye-candy and brand names their grandparents heard of. It's an uphill battle to convince homeowners otherwise.

  10. user-5975040 | | #10

    Question for Peter L
    @Peter - how would an Intus @ $20 per sq ft compare to an Andersen 200 or 400 in terms of pricing and performance?

  11. jackofalltrades777 | | #11

    The one issue is that Andersen doesn't like to broadcast its energy efficient specifications. One has to dig and dig and dig to eventually find a U-Value somewhere. It's all about "Low-E" this and Low-E that. So after much digging it appears the 200 & 400 series comes in around a U-Value of 0.30 average. By comparison an Intus Window clearly and openly lists its U-Values and they are easy to find. An Intus Window comes in around 0.08 - 0.10 for Ug (center of glass) with the PVC window framing, it comes in around 0.12 - 0.15 total U-Value (window + frame).

    It's not just the energy performance of the glass but it's the entire package. Compare a European window like Intus to an Andersen window and it's like comparing a Bentley to a Chevy Cavalier. The Euro windows feature heavy duty hardware, steel reinforcements, triple gaskets, superior interior air sealing, etc. They are built like a tank

  12. jk96 | | #12

    Peter we ended up with
    Peter we ended up with Anderson 400 windows. No argument from me on the quality difference but I didn't find that Anderson was trying to hide its specs as you suggest. Our bid clearly stated the U-value and specs for each window in the bid. I also went to their website out of curiosity just now and only took a couple of clicks to locate their performance specs.

  13. user-5975040 | | #13

    For Peter L
    Thanks Peter. If Intus windows are in the $20 per sq ft range and Jeld Wen is around $17 - where would Andersen land per sq ft? One of my frustrations with building a new home is how little transparency there is in pricing. I get why it is that way - but it's a major roadblock in planning.

  14. sasksaab | | #14

    Attempt at humour? (Cdn spelling)
    "London, Ontario, just up the road from the Blue Heron EcoHaus". Saskatoon is approximately 2700km from London, Ontario. London is south, so should it read "down the road"?

  15. GBA Editor
    Martin Holladay | | #15

    Response to Jonathon Zacharias
    Bronwyn lives in San Jose, California.

    If you live in California, you tend to assume that Saskatchewan and Ontario are near each other, because after all, they are both in Canada. But it turns out that if you are driving, the distance from Saskatoon to San Jose (1,696 miles) is less than the distance from Saskatoon to London, Ontario (1,716 miles).

    In other words: Saskatoon is just up the road from San Jose, California.

  16. user-1089777 | | #16

    Airtightness benefit
    Thanks for all the great comments on this post. I’m visiting my family in the UK after the conference in Germany and don’t have great access to internet, so please forgive the sporadic responses. I’ll post replies as I am able.

    Martin – the airtightness benefit is as much about durability as it is about energy savings, particularly in a climate like Saskatoon. Whether your aim was just above average, Pretty Good, or Passive, anyone in this region would be well advised to get their airtightness reading as low as possible. The extreme temperature differentials in Saskatoon cause huge pressure drive between assemblies and can cause serious moisture damage in a very short space of time. As to the cost, going from 0.6 ACH to 0.2 ACH is more a matter of good design detailing than good contracting. If the junctions on a project have not been well detailed to enable a continuous, accessible path for the builder to install a continuous air-barrier around the entire building, it will be both expensive and difficult to achieve. I noted that the builders of Kranichstein managed to achieve 0.2 ACH back in 1990. Given all the advances in tapes, membranes and paint-on air-sealing systems, there’s not much excuse for designers and builders of today to not meet this target cost-effectively.

    As to the performance benefit of this air-tightness recommendation, attached is the graphic output from PDT-Passivhaus that illustrates the benefit of improving the ACH air-tightness reading. Note that this graphic shows the ACH reading on the x-axis and the metric values for Heating Demand (kWh/m2a) and Heating Load (W/m2) on the y-axis. Reducing the ACH from 0.6 to 0.2 ACH reduces the Heat Load from 17.5 W/m2 (5.5 BTU/hr.ft2) to 14 W/m2 (4.4 BTU/hr.ft2.)

  17. GBA Editor
    Martin Holladay | | #17

    How easy is it to hit 0.2 ach50?
    As an exercise, I decided to look up the blower-door test results of the green homes in our case study section here at GBA. These are all homes built with attention to air sealing. I wanted to see how many of them hit 0.2 ach50.

    None of them did. (David Posluszny claims that his Massachusetts house was tested at 0.09 ach50, which is remarkable. However, that result was achieved by installing a wrong-side vapor barrier -- Ice & Water Shield on the exterior side of the wall sheathing. The house claiming the title of "the tightest house in the world" -- a 600-square-foot cabin in Alaska with only three windows -- came in at 0.05 ach50.)

    2.1 ach50: Denver Developer Focuses on Zero-Energy Homes

    2.03 ach50: Large Connecticut Home is ‘Zero-Energy-Ready’

    1.96 ach50: A Net-Zero Home in Massachusetts

    1.0 ach50: Malcolm Isaac's house

    0.97 ach50: Modern Dream Home is Energy-Positive

    0.96 ach50: Luxury Home Earns Gold NAHB Energy Value Housing Award

    0.92 ach50: Matt Risinger's SIP house

    0.63 ach50: A Modest New House Proves Green Doesn't Mean Expensive

    0.6 ach50: Passivhaus on a Budget

    0.59 ach50: Riverdale Net Zero House

    0.58 ach50: Rideau Residences

    0.56 ach50: A Net-Zero-Energy House for $125 a Square Foot

    0.54 ach50: Massachusetts Owner-Builders Complete a Superinsulated Home

    0.51 ach50: NewenHouse Passivhaus in Wisconsin

    0.5 ach50: A Net-Zero-Energy Home in Rural Tennessee

    0.5 ach50: A Contemporary Passivhaus Design in Seattle

    0.49 ach50: WolfWorks house in Farmington, Connecticut

    0.49 ach50: Freas house

    0.47 ach50: Rob Dumont house

    0.46 ach50: Michael Trolle house

    0.45ach50: Andrew Michler's house in Colorado

    0.44 ach50: Matt and Laura Beaton's home in Massachusetts

    0.42 ach50: West Paris, Maine, Passivhaus

    0.41 ach50: Dan Whitmore's Passivhaus duplex

    0.40 ach50: Mini-B Passivhaus

    0.40 ach50: Chris and Zoe Pike's house in Ripton, Vermont

    0.4 ach50: A Superinsulated House in Rural Minnesota

    0.4 ach50: A Higher Standard

    0.4 ach50: Maura and Kurt Jung's house

    0.38 ach50: Lehto house in Killingly, Connecticut

    0.38 ach50: North residence

    0.38 ach50: Brian Post and Kyra Salancy's house in New Hampshire

    0.38 ach50: Paul Honig's house

    0.36 ach50: Mill Creek Net Zero house

    0.34 ach50: The Potwine Passivhaus

    0.33 ach50: Farmhouse Style Meets Passive House

    0.33 ach50: Newry Passive House

    0.3 ach50: Vahid Mojarrab's house in Santa Fe

    0.286 ach50: Chris Corson's house in Maine

    I'm sure that many builders will report that it's fairly easy to hit 0.6 ach50. It is, if you pay attention. But 0.2 ach50 is much harder to hit. (In a comment posted on GBA, Steve Toomey reported that his builder hit 0.2 ach50. Bravo!)

  18. user-1089777 | | #18

    Dana 'hearts' PV, Bronwyn 'hearts' good windows
    I’m encouraged by all the comments regarding window performance generated by this post. Sourcing locally made windows that offer even decent performance numbers are still our biggest challenge here in North America.

    Dana – your beloved solar panels may certainly be cheaper to purchase currently, but when you sit next to junky windows on a cold day and you feel chilly because your body is now the radiator, or when your wife has to constantly clean up the condensation rotting your interior sill, or when you feel a draft off the windows because they’re a cold surface and that temperature asymmetry moves the air in the room, you may wish you’d chosen a better performing window for almost, if not the same, amount of money!

    While I can’t argue that solar panels are currently not cost competitive with PV, please consider that not too long ago PV was also not a cost-competitive investment. Only years of federal and local subsidies finally pushed this component over the economic viability threshold. The same incentives will likely be needed to shift the North American window industry out of its ‘SUV-window’ rut.

    I also noted in my article that the passive solar project monitored as a comparison to the Saskatchewan Conservation House had quad-pane windows (this was back in 1978!) Even with those highly insulated windows that project still experienced significant interior temperature fluctuations. A good quad-pane window is still barely 0.11 or R-9. Over a certain area, it will LOSE much more heat than what it gains for much of the year, so the recommendation of Bob Dumont (et al) to keep the floor to glazing ratio of a building in this climate to 6% is as much a comfort one as it is an economic one. To my knowledge, nobody complained that there was not enough light in the Saskatoon Conservation House.

  19. user-1089777 | | #19

    Response to Stephen Sheehy
    Stephen – as Peter L confirmed, windows with great performance numbers are now relatively cost competitive and readily available on the US and Canadian markets - as imports. I’ve used the Intus windows mentioned in my own Californian Passive House projects. Their performance, cost and quality beat the pants off of any local products - bar one. This is not to say I wouldn’t prefer to purchase a local product, but only one US company is able to deliver anything close to what I’d like to see installed on all my projects (and I design and build in a very benign climate.)

    The one local company I like is a small manufacturer out of California selling heavily into the Canadian market: Synergist Windows. They make an all-wood product that is both beautiful and kicks butt on performance, but is still a little pricey due to lack of being able to source materials at scale. On the whole, I’ve been underwhelmed by the poor performance and quality of most US and Canadian windows. Their inability (or lack of interest) to update their profiles to improve performance is disappointing. I’m predicting we’ll be losing even more manufacturing jobs to China since I just learned that China has produced over 10 Certified Passive House windows in the past year alone. I’m betting Dana Dorsett also buys his solar panels from China. I wonder how the economics of that stack up when we include loss of local manufacturing jobs?

  20. STEPHEN SHEEHY | | #20

    Reply to Bronwyn
    When looking for windows for our new house, we really would have liked to buy windows made in the USA, but, as you point out, the Intus windows we ended up using were much better than we could get from Marvin or Andersen or anyone else. Maybe they are content with their existing market share, but they risk losing that if they don't get with the program. Some of us remember when Toyotas didn't worry the big three US automakers.

    I think many consumers aren't familiar with energy efficient windows and how much more comfort they can bring. They just use what the builder recommends. Whenever my builder stops by the new house, he invariably comments on how much better constructed the Intus windows are than what he usually installs.

  21. GBA Editor
    Martin Holladay | | #21

    Anticipated savings from going from 0.6 ach50 to 0.2 ach50
    If I read your graphs correctly, the saving attributable to the switch from 0.6 ach50 to 0.2 ach50 is 5 kWh per square meter per year (44 kWh - 39 kWh).

    There are 226.4 square meters in this house, so the annual savings is 1,132 kWh of heat.

    Using an air-source heat pump with a COP of 2.5, that is a savings of 453 kWh of electricity each year. If electricity costs 15 cents per kWh -- in most of Saskatchewan, electricity costs less than that -- the annual savings is $68.

    Whether it's worth the trouble to reach that goal depends on how hard it is to go from 0.6 ach50 to 0.2 ach50. (And if reaching 0.2 ach50 is "darn near impossible," then this entire discussion is unrealistic.)

  22. charlie_sullivan | | #22

    Evaluating options in an underserved market
    Bronwyn makes a really interesting point about cost effectiveness: Solar panels have gotten over the hump to where they are mass market items that you can get for bargain prices. Some other technologies, perhaps including some varieties of high-performance windows, have not. Maybe we should all go buy these not-yet-cost effective technologies in order to help them get to the point where they become mass market items and become cost effective.

    That's a valid argument that I suspect is central to a lot of why many of us do what we do. At the same time, it's a slippery slope that can lead to very poor decisions, wastes of resources, and reinforcement of the idea that green buildings are inherently expensive and are playthings for the rich rather than being solutions to real world problems.

    If only we had a good way to distinguish between those two categories. But perhaps we can't. And perhaps it takes both kinds of demonstration projects to move the field along: Some projects that cost more than is justifiable to prove what's possible, and other projects where what's most impressive is the cost effectiveness. I think people doing each kind of project can benefit from having the other type proceeding as well.

  23. Expert Member
    Dana Dorsett | | #23

    "junky window"?? Response to #18
    A U0.20 window is hardly "junky", but it's not a U0.11 window, but I'd hazard it's less than half the cost. At mean mid winter temperature of -12C in Moose Jaw or -15C in Saskatoon a U0.20 window it's not a frost/condensation on the window type problem, and is not a comfort problem to anybody except perhaps the princess annoyed by the pea.

    Was it was Joe Lstiburek who is credited with the statement that "Five percent of all people are never comfortable."?? :-)

    I'm sure in a side by side blindfold test one could tell the difference at -15C but whether the difference is worth paying extra for isn't necessarily clear.

    I certainly don't love solar panels- cheaper low-carb energy is worth having, and as an investment are worth more to the US economy than the premium paid for very high performance windows. It's not about love, it's about expectations: I'm much optimistic that solar will continue to be cheaper year on year than I am that high performance windows will. At some market established minimum comfort threshold high performance windows and insulation have to compete on lifecycle energy cost, which is a losing proposition for windows at the learning curve of windows relative to the learning curve of solar.

    The premium we pay in New England for first-world fabbed 20% efficiency PV vs. first tier Chinese 15% efficiency panels is about 25 cents per installed watt, less than a 10% mark-up. When Solar City ramps up it's Buffalo NY high-efficiency panel plant it's likely that the difference will shrink. (They're hiring now, if you're interested: ). There isn't a huge cost advantage to building high-end panels in China compared to the US, but the Chinese have invested more capital in production facilities over the past decade and have a large market share. The assertion that Chinese PV is putting high end US window manufacturing or US PV manufacturing workers on the street is an odd argument that needs something that looks like to evidence to support it.

    The learning curve on silicon PV is still running 20-25% installed-cost reduction for every doubling in production volume- it's getting cheaper faster now than ever, and that's without technology breakthroughs, and there are nothing that will inherently slow that any time soon. In the US about half the installed cost of rooftop PV is in soft cost exclusive of installation labor. Panel costs comprise only about 20% of the total installed cost, installation labor about 10%. (see: ) The solar business employs more local labor than the coal business, and most of that is installation labor. The labor content of a PV panel is actually quite small relative to the installation labor, whether manufactured in the US or offshore.It's a highly automated process, and becoming more so. Until there is more onshore manufacturing of PV closing the doors on Chinese PV to the US market would result in the loss of skilled installer jobs. Closing the door on Chinese PV would result in a ZERO increase in US window manufacturing jobs.

    Sure, PV used to be a lot more expensive than it is now, but it's manufacturing learning curve is at least an order of magnitude faster than the learning curve on better windows. I'd be thrilled if high performance window manufacturing was increasing at such a rate that it's learning curve could catch up, but short of a manufacturability breakthrough on vacuum insulated glass I don't see how incremental technology improvements can get around the basic commodity and craftwork constraints of building decent windows. It's just a lot harder to achieve a significant financial learning curve on products as mature as windows at any volume increases. The performance levels of windows in the US is driven by building code mandates. Until/unless codes mandate higher performance windows, the volumes will remain small, and whatever learning curve there is on windows will still take a long time to play out.

    If people in the US are sufficiently satisfied with their U0.32 code min windows (and there is reason to believe they are), codes won't be mandating U0.15 or whatever windows at any significant mark up unless there is a financial rationale for the higher performance. And without the higher volumes, the will be premium priced for a long time to come. I'd LIKE to be more optimistic about the prospects for better windows becoming more cost effective and ubiquitous, but I'm just not.

  24. GBA Editor
    Martin Holladay | | #24

    Response to Bronwyn Barry
    You wrote, "The airtightness benefit is as much about durability as it is about energy savings, particularly in a climate like Saskatoon. ... The extreme temperature differentials in Saskatoon cause huge pressure drive between assemblies and can cause serious moisture damage in a very short space of time."

    The idea that an air leakage rate of 0.6 ach50 can cause "serious moisture damage in a very short space of time" is absolutely without foundation. There is no need to aim for 0.2 ach50 in Saskatoon.

    Of course it's important to avoid gross defects in the building envelope. But the idea that 0.6 ach50 isn't good enough -- that one has to aim for 0.2 ach50 to avoid moisture damage -- is unsupported by any evidence.

  25. GBA Editor
    Martin Holladay | | #25

    Window U-factor
    In your suggested improvements for Kent Earle's house, you showed that his windows have a U-factor of U-0.22.

    But in his blog post on windows, Kent Earle told us that his triple-glazed Duxton windows have whole-window U-factors ranging from 0.15 to 0.18. Can you explain why your U-factor assumption is so different from the U-factors that Kent reports?

  26. peteramerongen | | #26

    Great job Bronwyn! This optimization tool is badly needed. Does the tool allow you to enter your own costs?

    What I didn't find (and what must be in there somewhere) is any discussion of the dollar cost of buying the capacity to save that last kWh per year needed to hit the target. On the simpler optimizations I've done trying to hit the target here in Edmonton, Alberta that cost can be as high $10/kWh/a.

    I agree that better windows, air tightness, and better HRV's are typically more cost effective than higher R value walls and ceilings.

    The reason we have so much trouble hitting the PH target up here in the frozen north is that the cost benefit curve gets so darn flat.

  27. Expert Member
    MALCOLM TAYLOR | | #27

    Maybe this is unfair, but the optimizing to me resembles the five year plans the Soviet Union used to put out. Goals are established, and inputs tweaked, while the reality of what is actually possible or desirable is dismissed. There always seems to be an unacknowledged whiff of ideology surrounding the discussion of Passive House standards under the purely scientific overlay of graphs and tables.

  28. david_white | | #28

    cfm50 per SSF and window areas code
    First: For Kranichstein to hit 0.2 ach50 is not a useful comparison. The difficulty of an air leakage target has mostly to do with leakage per area of shell (e.g. cfm50/ssf). For the multifamily Kranichstein PH to meet 0.2 ach50 is much easier than for the single family detached Blue Heron house, because the former has far fewer square feet of shell per interior volume than the latter.

    Second: the proposed window area of 15.2 m2 is TINY for a house with a TFA of 226 m2. Note that the gross floor area of this house is a bit bigger, say around 280 m2. 15.2/280 = 5.4% of GFA, i.e. it fails by a wide margin to meet NYC code requirement for natural light (10% of GFA).

  29. GBA Editor
    Martin Holladay | | #29

    Response to David White
    Thanks for your perceptive comments.

    Changes on a spreadsheet give mathematical results... but those mathematical results aren't necessarily applicable to an actual job site.

  30. Tim C | | #30

    Five percent of all people are never comfortable
    I'm in southern Wisconsin, with 20 year old dual pane, low-E, aluminum spacer, aluminum/wood frame windows. I don't have any comfort complaints with them and only minor condensation problems. I'd certainly benefit from a better frame design & warm spacers, but I'm skeptical that there are significant, meaningful comfort gains to be had by adding even a third pane, much less a fourth.

  31. Expert Member
    Dana Dorsett | | #31

    Southern WI is the warm edge of zone 6 (response to Tim C)
    I would agree that in your climate a U0.35 ish window isn't usually a comfort issue- it's just not that cold there. Up in Saskatchewan or Manitoba where the average mid-winter temps are 15-20F colder than southern WI that third pane is both a comfort factor and solves the window condensation issue.

    The binned hourly mean temp in Madison, WI for a typical January is about +18F.

    In Saskatoon SK it's about +4F.

    In Winnepeg, MB it's about +6F

    In Kenora, ON it's about +2F.

    The average January daily HIGH in Kenora ON is about a degree F cooler than the average daily LOW in Madison WI.

    A U0.20-ish triple pane really IS a comfort factor in those places, but I'd be hard pressed to make the argument that stepping up a 4-paner is going to be "worth it" on the comfort factor for most people. Maybe if the upcharge was $25 /window it might be worth it for some, maybe not for others.

    In balmy central Massachusetts the cost of U0.20 window is worth it on Net Zero houses, if only to meet code for being able to keep doored off bedrooms at the mandated 68F code min at the 99% outside design temp when heating the place with a point source such as a mini-split. With some design tweaks to other factors it can meet Net Zero Energy with a code-max U0.32 window, but it's hard to guarantee the 68F with reasonably sized windows at that performance level. The cost of adding electric baseboards or something else to solve the heating system code compliance problem is comparable to the upcharge for a ~U0.20 window, and the higher performance window allows you to relax a bit on other parameters. There is a noticeable comfort difference, but it's not significant enough to be a "must have".

  32. GBA Editor
    Martin Holladay | | #32

    More on "optimization"
    "Optimization" is a funny word. It implies gradual progress toward perfection. But the results of an optimization exercise depend on the program's goals.

    Let's say that I want to optimize the GBA web site, with the goals of increasing page views and the maximizing the number of visitors. This type of optimization would quickly result in a web site that was 50% cat videos and 50% photos of topless women. GBA is now "optimized."

    Similarly, a Passivhaus optimization program shows that Kent Earle's house would be improved if the builders targeted an air leakage rate of 0.2 ach50; if the area of windows were cut in half; and if Earle specified quadruple-pane windows. Earle's house is now "optimized," according to the goals of the computer software. But does the advice make any sense?

  33. Expert Member
    Dana Dorsett | | #33

    I like the BeOpt better.
    BeOpt is designed to cost optimize the building performance by being able to compare the bang/buck ratio of alternate assemblies, but doesn't constrain the target performance to the rigid Darmstadt specification.

    That's something different from tweaking an existing design to cost-optimize it into PassiveHouse compliance, rather than letting it remain non-compliant. ( Getting to compliance is something of a Feist-ian bargain, mayhaps? :-) )

    Whether the cost of compliance is "worth it" is purely subjective. Getting PassiveHouse certification could be viewed in the same vein as marble shower surrounds or granite countertops- nice stuff to have, looks good in the architects' & builders' porfolios, but in the grand scheme of things not as important to the world as retrofitting existing homes up to something akin to IRC 2015 code minimum or better, or simply not heating your house (PassiveHouse or other) with resistance heaters on a coal-fired grid.

  34. brightbuildings | | #34

    Excellent study
    Thank you Bronwyn. This is very exciting to be shown a path to the Passive House standard in Saskatoon requiring only R35 walls on a single family home!

    Certification under heating load criteria of 10 W/m2 means you can use ventilation air heating, allowing a truly simple and cost effective design. The improved windows meet the comfort criteria without supplemental heat (we need 0.65 W/m2K in a cold climate).

    I imagine you will have good daylighting and views even with reduced glazing.

    This is an important demonstration showing an approach to passive house that is far less dependent on passive solar gain and therefore orientation. Far more sites are suitable for this approach.

    It's true that space heating demand is almost 30 kWh/m2a but that is still a 90% reduction from average heating energy in Saskatchewan of 300 kWh/m2a.

    The key is the windows, which would likely need to be imported. I would really like to try VIG, vacuum insulated glazing, in a passive house certified cold climate frame. I imagine it is costly, but may not be that much in the end if you are going with somewhat smaller windows.

  35. user-1089777 | | #35

    Window comments for Charlie, Martin & Dana
    Charlie Sullivan – Thanks for your philosophical perspective on the balance between beacon projects that reach beyond the confines of what is considered possible (such as the Saskatchewan Conservation House and Kranichstein) and those that opt for a pragmatic approach and remain within the confines of current economic viability (such as the Blue Heron Haus.) Both ends of this spectrum serve as useful benchmarks. This is the same for any industry or product development trajectory. We can use the example of flight: it’s now often cheaper and faster to travel by air than most train or motor travel, yet we still look to space flight and ideas like Musk’s Hyperloop as the next frontier. What I’ve shown here is that the physics of the Passive House standard is absolutely possible, even in a very challenging climate such as Saskatoon.

    Martin – the reason I used a different U-value for the Blue Heron Haus than what was posted for the Duxton windows is twofold. First is due to a difference in testing protocols. Values for windows tested using NFRC protocols differ from those tested to the ISO protocols required for accurate PHPP inputs. I’ve presented and written about that extensively. Here’s a presentation that covers the basics:
    Second is due to variation in window sizes. The number I’ve shown is an average u-value for all the windows. PHPP adjust the u-value per window size since the stated u-values are for standard testing sizes. As we know, window sizes vary drastically from one project to another. The PHPP calculates this value per the project design.

    Dana – I encourage you to be more optimistic with regards to the development of high performance windows for the US and Canadian markets. Consumers may currently be “sufficiently satisfied with their U 0.32 code windows” but don’t underestimate them. I’m banking on consumer’s ability to demand better once they’re aware that better comfort, quality and performance are all possible and available. By your thinking, the rotary dial-up phone would still be in use. I acknowledge that our market sadly isn’t there – yet.

  36. user-1089777 | | #36

    Air-tight comments for Martin and David
    Martin – I’m disappointed that you weren’t excited about how I reduced the sub-slab insulation on Blue Heron Haus. (Sub-slab insulation has previously been your Passive House bête noire.) You’ve obviously moved on to airtightness. The list of projects you assembled all hit very good numbers for projects with no pre-posted targets. (If you don’t aim for something, you’ll fall/settle for anything.) Frankly I’m delighted to see how this conversation has shifted over the past eight years. Previously people like yourself were highly skeptical of the 0.6 ach target. It is now universally considered a reasonable and achievable target.

    My analysis of the Blue Heron Haus suggests a higher bar. Kranichstein hit this number a long time ago on a building half the size, so I’m not asking for the impossible here. I’m betting this number will be quite attainable on a regular basis in the near future. I do concede it’s not easy and will take a focused effort.

    By the way, you missed these projects. Please add them to your list:

    0.05 ach50:
    0.065 ach50:

    David White – I’m waiting to receive confirmation from Dr. Feist via Twitter that the airtightness test for Kranichstein was measured for his unit only and not the whole building, which I believe is the case. (I’ll post the confirmation when I receive it.) At 156 m2 TFA, Kranichstein is about half the size of the Blue Heron Haus of approximately 226 m2, so by your account would be much harder to meet the 0.2 ach number.

    I'm not current on code minimum requirements for minimum day-lighting in Saskatoon, but judging by the Saskatchewan Conservation House numbers, it’s clearly much lower than that of NYC.

  37. user-1089777 | | #37

    Cost for Peter
    Peter – thanks for the compliments on the optimizer tool. I sincerely hope that it’s useful to folks like you who work in climates that are extremely challenging (not just for Passive House!) Hats off to you. My investigation of the Blue Heron Haus was prompted by a conversation we had back in 2013 in Vancouver. You told me how difficult it was and I needed to see for myself if there was any possibility to meet the Passive House Standard in Saskatoon. I was honestly surprised to be able to do so. I was prepared to have to enlarge the project into a duplex, but found it can be done with better, fewer windows and a tighter envelope – all of which are possible. There is now a growing cohort in your region who are aiming for this high bar, including Mike Nemeth. I'm sure he'll break the ice ceiling!

    The cost question is always a challenge. I don’t have localized cost figures for Saskatoon and from my own project experience in California I know that increases in assemblies, components or better air-tightness are not always linear. We’d love to be able to add a cost element to the PDT-Passivhaus optimizer in the near future. We’ve developed this tool on our own dime without the support of any government subsidy or a large number of programmers. It’s exclusively for PHPP users, which is currently a small market here in the US and Canada. At the moment we’re beta-testing this tool to find out if the market finds what we’ve done is useful. From your response and the response to my presentation in Darmstadt at the International Passive House Conference, I’m encouraged. It bodes well for us to add the cost element to the service in future.

  38. Andrew_C | | #38

    Usefulness of simulation
    There have been a few comments here that seem to be dismissive of the usefulness of simulators and optimization. Certainly there should be a discussion about the allowable range of various model parameters (e.g. air leakage). But just because the combination of parameters on a particular simulation run seem unrealistic (to you, now, based on current state and current understanding) doesn’t mean the simulation isn’t worth running. The reason for testing is to gain insight into the effects of various combinations of factors. The great thing about computer models is that you can test hundreds and thousands of cases without the cost, time, and safety issues of physical testing.
    Yes, of course mathematical models are simplifications of real world physics. That’s why we use the Scientific Method: hypothesize, test, compare results with real world data, tweak, re-test, etc. Eventually, you should have a tool that gives you insight, direction, and useful approximations.
    And once you have a good tool (i.e., known to provide decent correlation with real world results, e.g. PHPP), now you can create optimizer tools that test many combinations to come up with a range of “best” solutions. Again, you can have discussions of the realistic/useful range of individual variables (insulation levels, % glazing, etc.), and also about the optimization goal(s), but having a optimizing tool to use with a known model is potentially incredibly useful.

  39. Tedkidd | | #39

    What a great post Bronwyn! That it elicited so many really fine comments kinda blows my mind! People are clearly seeing the value of design iteration, of setting really high goals, and recognizing that even getting close means a fantastic house!

    Now how do we drive this perspective of investing time into iterative design into the "Home Performance" industry? This approach can yield really amazing outcomes when applied to the retrofit marketplace.

  40. Andrew_C | | #40

    "Feist-ian bargain"
    Wrong. But funny. :)

  41. user-1089777 | | #41

    Confirming Kranichstein Blower Door area/volume
    Follow up to comment #37: I received confirmation that the Kranichstein blower door measurement is for a single unit and not for the entire 4-unit building. Based on a recent tour of the property, I realize it would be impossible to test the entire building as there are no inter-connecting doors between the units.

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |