What’s the best way to manage a home’s indoor climate?

As I type this blog post, our house’s fan is running behind me, circulating air – and directly in front of me is a lovely bank of fifteen-light windows through which I can see our back garden.

Why don’t I just turn off the fan and open the windows?

Wonder what my energy loss rate is?

Throughout history, people have craved windows in their dwellings – for air, for light, and for view.  Windows let in bugs and water.  They also leak heat out and cold in, or vice versa. 

And into the basement …

That’s why, for the last fifty years, the trend in evolving modern buildings (homes and otherwise) has been to seal the building’s airflow even as we increase its overall transparency.  

Is that wise? 

Windows are the bane of a homeowner’s existence – and even more the bane of a landlord’s existence.  They’re surprisingly expensive to make, assemble, and install.  If movable, they’re a fairly complicated installation that rattles and usually leaks.  Painting the window trim is a bitch. 

It’s a bitch of a job

For all these reasons, modern buildings have windows you cannot open.  In previous posts, I’ve talked about the architects’ tendency to design complicated structures, which tend to leak, or have unexpected consequences.  Via reader Matthew Healy, that led me to building airflow expert Tim Padfield and his paper, The Window in Context:

Abstract.  The ventilation requirement for a house is a compromise between the need to keep the indoor relative humidity down and the need to minimize energy consumption. 

* The first requirement is itself a short-circuited train of reasoning about how to prevent condensation in the structure with consequent mould growth. 

* The second requirement was until recently defined as the permitted energy used to keep the occupants physically comfortable.

Aside from being unsightly, mold is extremely toxic of a small subset of people, and that means litigation, lawyers, and money.

Got a little problem, have we?

Mould growth in the occupied part of the house is only indirectly a consequence of high indoor relative humidity- it is usually a consequence of uneven wall temperature gradients within the structure. 

I never knew that – it’s not the humidity by itself, but rather how temperature changes within a structure, because with a wide temperature gradient, somewhere is an environment that allows mold to grow.

Designing a structure that is robust enough to be badly built and intermittently maintained is just as important to the life cycle costs as is the energy consumption from day to day.

I absolutely love that line, for it’s the reality – maintenance is imperfect.  In Tthe risk of complicated structures, I’ve previously quoted Padfield’s Law of Construction:

Trends in building design have been towards complicated, lights structures optimized through the use of computer models but not always performing as expected when released into the care of the occupants.  In particular the intricate detailing and tendency to use non-water-absorbent materials brings considerable risk of condensation in the northern European climate, with subsequent mould growth.  Air barriers have not provided the expected security from condensation within the outer envelope of buildings.

We saw that in dreamers versus plumbers.  Then there’s this admission:

There are some rather fundamental holes in our knowledge of the dynamics of biological growth. In particular the ecology of microorganisms has been relatively neglected in favour of experiments based on sterile media inoculated with purified organisms in a constant environment.  The influence of air flow and air exchange on growth has been neglected.

It’s imaginative, but it’s not the world

Just as Korzybski said the map is not the territory, the model is not the reality. 

“This is not a pipe”

In finance, we love our models, but the word ‘model’ implies approximation, not the full complexity of reality.  Some things are hard to model, so we omit them – and sometimes that’s a mistake.  And when I hear phrases like “not well understood,” my mind immediately drifts to “tort litigation” and then “expert testimony” and even something called the “Daubert standard.”  Wikipedia’s summary is quite good:

In Daubert, the Supreme Court held that federal trial judges are the “gatekeepers” of scientific evidence. Under the Daubert standard, therefore, trial judges must evaluate proffered expert witnesses to determine whether their testimony is both “relevant” and “reliable”, a two-pronged test of admissibility.

And the UFO’s light were right along here …

The relevancy prong: The relevancy of a testimony refers to whether or not the expert’s evidence “fits” the facts of the case. For example, you may invite an astronomer to tell the jury if it had been a full moon on the night of a crime. However, the astronomer would not be allowed to testify if the fact that the moon was full was not relevant to the issue at hand in the trial.

The full moon had nothing to do with it

The reliability prong: The Supreme Court explained that in order for expert testimony to be considered reliable, the expert must have derived his or her conclusions from the scientific method (Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) 509 U.S. 579, 589.) The Court offered “general observations” of whether proffered evidence was based on the scientific method, although the list was not intended to be used as an exacting checklist:

* Empirical testing: the theory or technique must be falsifiable, refutable, and testable.

* Subjected to peer review and publication.

* Known or potential error rate and the existence and maintenance of standards concerning its operation.

* Whether the theory and technique is generally accepted by a relevant scientific community.

With illness due to mold rising as a source of class-action litigation and tort liability to owners, the cause of mold growth can be a multi-million-dollar issue.

The movement of water in hygroscopic materials is not understood, to judge by the failure at the many models to account convincingly for water and dissolved salt movements in porous materials and their difficulty in dealing with lateral irregularities in real walls. 

Water runs along cracks.

Water leak inside a ceiling (not Photoshopped!)

There is continued reliance on mathematical analogy between heat and material transport through porous materials, even though the movement of water molecules through the mesh of different sized pores is unlikely to be purely diffusive phenomenon.

The erratic ventilation caused by undisciplined use of windows is a considerable further inconvenience for these modelers. 

Programmers and economists tend to ignore what we cannot model, and of these things we wish to ignore, human behavior is among the most complex. 

It is perhaps easier to believe that one has control when all the windows are sealed shut and a computer controlled ventilation system is installed, but experience does not always support this conviction. 

People, bless their illogical hearts, do odd things; sometimes the wrong thing, sometimes the right thing.  Can we construct incentives to do the right thing?  Why, yes, we can – individual metering is the most basic example.  Further, people are observant: recent pilots that have installed usage meters in houses – or hooked them up to computer monitors – have shown that people reduce their energy use when they can see where it comes from.

These material investigations, together with physiological studies of people’s tolerance to draughts and uneven temperature indoors, promise to re-activate the discussion on the use of traditional ventilation through windows as a perfectly reasonable way to control to indoor climate without unreasonable waste energy. (Pg. 2)

In other words, the biological thermostat might just be better than the mechanical or electronic ones.

One way to keep the temperature where you want it

Another approach

If houses could be designed to work well with windows that are opened when it is convenient for the occupier, one could minimise energy loss and unpleasant draughts. For example, a bedroom with water absorbent and carbon dioxide permeable walls will hold a good microclimate throughout one night with windows closed. The windows can be opened for a short period during the day to regenerate the buffer capacity of the wall.

That morning sense of needing fresh air isn’t your imagination. the air does need recirculation.

This is also the time when the ambient temperature is highest, so the energy loss is less than if the windows were slightly open all night to ventilate away moisture and carbon dioxide. (Pg. 37)

When I was kid, you were supposed to sleep with a window open, even when it was cold.  Maybe they knew what they were doing.

If only I could open windows too

Send post as PDF to

“Missed it by that much.”

As I wrote in that post:

Architects should never have the final say in designing any building.

Property development is susceptible to what I call the ‘ribbon-cutting trap’ – the tendency to declare that completion of the building is a successful finish.  In reality, it’s only a successful beginning, because building is easy, operating is hard.  Maintenance is unforgiving to wishful thinking […] where I come from, water is a property’s greatest enemy. That’s why water and sewer systems are so critical to building integrity, and more fundamentally, why it’s inconceivable to us that housing could exist without water and sanitation.

I got a puckish email from faithful reader Matthew Healy, pointing me to some articles on condensation in vapor barriers by Tim Padfield, whom I presume to be the world’s foremost expert on the physics of building walls.  [If Harvard or MIT hasn't signed him to be an expert witness in their architect litigation, both should.]

A man who knows how glasses do or don’t fog up: Tim Padfield

Mr. Healy pointed me to this great line, which might be said to be the epitaph for Harvard’s Otto Hall: “complicated structures will never be built accurately enough to realise the intentions of their designers.”  In fact, I’m going to christen it Padfield’s Law of Construction Complexity:

Padfield’s Law of Construction Complexity ought to be graven on every foundation stone, pre-printed at the bottom of every blueprint.

“Don’t design what you can’t maintain”

Like a certain blogger we know, Dr. Padfield is a prolific author on a narrow topic, publishing pieces with titles like The interaction of water with paper in small spaces –

Figure 10: The warm wall evaporates water from the paper of the menu, which condenses on the cold glass, runs down and is re-absorbed by the paper. Because the sides of the enclosure are also exposed to the cold, condensation is worse towards the edge of the glass, obscuring the price while revealing the culinary delights on offer to the hungry traveller.

– and How to design climatically stable museums, as well as what reader Healy found, his paper Condensation in the Walls of Humidified Buildings, which describes Otto Hall with painful clarity:

Many museums in temperate climates are humidified in winter to attain the unofficial standard of about 50% RH at a comfortable temperature around 20C [68º Fahrenheit – Ed.]. Air conditioned museums are also usually pressurised, to prevent unauthorised entry of air that has not been filtered and humidified. The inside air will tend to flow and diffuse out through the walls.

Otto Hall had both.

Otto Hall: you wouldn’t know the skin was crumbling

As the air, together with its burden of water vapour, passes through the wall it will cool down. The relative humidity will consequently rise, if one assumes that water vapour moves at the same pace as the other constituents of air. We can use the atmospheric moisture calculator to show that air at 50%RH at 20C will reach 100% RH [Dew point --- Ed.] at 9C [50º Fahrenheit – Ed.].

On a cold winter day this temperature will be reached at some point within the wall. Water condenses there and may later freeze. The water will also dissolve salts and move them around in the wall and will accelerate corrosion of metal fittings within the wall. The process is shown in the diagram. The temperature gradient is drawn diagonally down through the wall section. The picture also hints at a process that partially counteracts this process: water can evaporate from the outer surface of the wall.

Deposition of water in the wall is therefore dependent on several competing processes: diffusion of air through the porous wall, flow of air through cracks in the wall, absorption of water into the hygroscopic brick and finally, mixing of the humid air originating from inside the building with outside air in the outer layer of the wall.

Engineers have responded to this threat to the durability of the wall by inserting a barrier to air flow through the wall, usually a polyethylene foil. Other professionals in the building trade then make holes in the barrier to insert electrical cables, bolts, pipes and just nails.

Just what happened in Otto Hall.

This has resulted in some spectacular failures of buildings that were designed with an attention to detail appropriate to constructing a jumbo jet but built in the more relaxed tradition of the building industry

You sure about the assembly here?

As he continues, low-tech means low-maintenance-risk:

I have never seen comparable phenomena in the simple, solid walls of uninsulated, porous brick that are typical of older houses that are used as museums. This must be because the fine balance of processes described [above] results in a wall that is seldom wet, even in very cold weather.

It seems to me that museum designers should be investigating the advantages of walls of homogeneous, porous material with thermal insulation distributed evenly right through the wall section. This will avoid the cracks, steep temperature gradients and impermeable layers that cause water to accumulate. It will also tend to make walls that have a large thermal inertia, which provides extra stability to the inside climate.

In another essay, The Window in Context, Dr. Padfield observes

“Designing a structure that is robust enough to [work even if] badly built and intermittently maintained is just as important to the life cycle costs as is the energy consumption from day to day.”

These principles apply equally well to financial structures.  When I was young and foolish, I liked very complex financial rules, incentives, and compensation structures.  As I’ve had more experience [you mean, aged -- Ed.], I’ve simplified rulemaking extensively, believing that simple rules not only after a better chance of being followed – fuzzy boundaries are bad boundaries – but also, they serve as better motivation.  That’s led me to this rule of thumb on joint ventures:

Rule number 1: measure twice, cut once

The best one says, “what’s mine is yours.”

That joint-venture-rule-of-thumb works well in my marriage, too J.

Send post as PDF to

“Imagine, one day we’ll write blog posts on devices like this”

“What’s a blog, honey?”

Long ago (1977), when I was just beginning to learn the business, our company was doing a repositioning and workout of a troubled affordable property in North Cambridge (as it happened, less than half a mile from my grotty rent-controlled apartment), which needed an enormous makeover.  Trying to put lipstick on the pig, our management company was sprucing up the elevators and common areas, and repainting the exterior.

Adds to my attractiveness, don’t you think?

The building was as plain as plain could be: a twelve-story high-story, the facade a rectangular grid of windows and concrete walls. 

It looked something like this

The manager, with a smile, showed me the two designs: a uniform light blue, and a Mondriaan pattern of blues, oranges, reds, yellows, greens, vibrant and randomized.

“Which one,” said Georgia with a twinkle, “do you think the architects wanted?”

The architects proposed something like this

“What’s wrong with the colors?” I asked.  (Back then, I was naïve.)

“How are you going to repaint it?”

And that, dear readers, was when I realized that:

Architects should never have the final say in designing any building.

It’s a pragmatic principle that Harvard University wishes it had followed more closely, in its development of what will be one of the shortest-lived museum wings ever: Otto Hall as part of the Fogg Museum, as reported in this little story from The Boston Globe’s Robert Campbell (also a Cantabrigian) that’s funny if it isn’t your money they’ll be spending:

Werner Otto Hall is attached to the rear of Harvard’s Fogg Art Museum. The 17-year-old building, which houses the Busch Reisinger Museum, is slated for demolition and removal.

You can think of it as a medical detective story, except that the patient is a building, not a person.

It’s a building that broke out with a skin disease that, at first, nobody was able to diagnose.

The building is called Werner Otto Hall. It’s a small building, attached to the rear of Harvard University’s famous Fogg Art Museum.

The Fogg, with five Rembrandts

Otto Hall houses Harvard’s Busch Reisinger Museum, a collection of Germanic art mostly of the 20th century.

Indoors, it’s a delightful set of galleries. Outside, it’s a rotting mess.

Otto Hall opened in 1991. Today, only 17 years later, its exterior walls have deteriorated so badly that Harvard says the only way to repair them would be to take them off and start over.

Obviously something’s mightily wrong.

Yet this disaster was created by the best and the brightest. The client was Harvard, or more specifically the Faculty of Arts and Sciences. The architect was Charles Gwathmey of the firm Gwathmey Siegel, known for – among other buildings – his superb 1992 addition to another museum, the Guggenheim in New York.

Hey, the maquettes always look great!

The general contractor was Walsh Brothers, a Boston firm now in its fourth generation that has long been regarded as one of the best in the region. Walsh Brothers built the American Academy of Arts and Sciences in Cambridge and much of Massachusetts General Hospital.

All-star teams work well in some areas where performance is primarily individual – like baseball – and not so well in others – like basketball or soccer – where smooth flowing teamwork beats individual brilliance.

Otto was much praised, by this column among others, when Otto Hall opened in 1991. From outside, it was a clean example of modernism, bold but not so bold as to upstage the more celebrated architecture of Le Corbusier’s Carpenter Center next door.

Really fits in an Ivy League setting, doesn’t it?

However dynamic it may be in the maquettes or the models, the Carpenter Center is totally out of place adjacent to its ivy-covered brick neighbors – and inside, it’s a terrible place to watch a movie.

Inside, it offered a set of galleries that intertwined with one another in unexpected ways.

Property development is susceptible to what I call the ‘ribbon-cutting trap’ – the tendency to declare that completion of the building is a successful finish.  In reality, it’s only a successful beginning, because building is easy, operating is hard.

We’re just here for the credit

That’s why it’s hard to believe that today the whole building – not just the walls, the whole thing – is slated to be completely demolished and removed. Tomorrow [June 29 – Ed.] the Fogg will close, and eventually new construction will engulf Otto’s former site. The new work will be part of a major addition to the Fogg designed by Pritzker Prize-winning Italian architect Renzo Piano, due to open in 2013.

Harvard would like to direct your attention away from Otto’s sick walls by claiming that the building just didn’t fit in with the plans for the new, larger museum, so it had to go anyway. But Harvard also admits, when pressed, that Otto’s exterior was incurable.

If only words would be wished away with hortatory titles

So what happened? What’s the diagnosis?

To put it simply, the guys who worried about the museum’s art were not the guys who worried about the weather. We’ll call them the art guys and the weather guys.

I love the plainspoken way in which Mr. Campbell frames the problem, all too reminiscent of my own experience with ungainly old Walden Square Apartments thirty years ago.

The art guys, applying a conventional standard, decided that Otto’s interior should be kept at a temperature of 70 degrees, with 50% humidity. Those numbers would be best, they believed, for the health of the artworks.

I’m sure they would be.

Not only that, but the curators asked that the interior be pressurized, like the fuselage of an airplane. They didn’t want cold, dry Cambridge winter air slipping in and damaging the precious artworks. The museum owned world-class works by such artists as Joseph Beuys and Max Beckmann – the self-portrait of Beckmann in a tuxedo is one of Harvard’s unforgettable treasures – and also housed frequent touring exhibitions.

I expect to live forever, free from humidity

Here’s the first worm of doubt in letting the ‘art guys’ have their head.  If there were works particularly sensitive to dry air, one could encase the works themselves rather than encase the whole building.

If there were going to be any air leaks through the exterior walls, the curators wanted to be sure the air would leak out of the building, not into it.

Taking advantage of positive pressure

It’s important to understand that this kind of sophisticated climate control was still fairly new at the time Otto was designed. Art curators were making demands that neither the world of architects nor the world of engineers and contractors had quite caught up with.

Second warning signal: why design a new technology for irreplaceable objects when a proven technology (encasing the paintings in climate controlled glass) would work well?

OK, that’s the art guys.

The weather guys – the architect, his engineering consultants, and the builder – created pretty much the kind of exterior they’d always built. The primary purpose was not to nurture the art but to keep out the weather.

Anyone who’s watched a ceiling drip water will be with the weather guys on this. 

“I don’t know much about art, but I know what a leak is.”

They built what is called a cavity wall. A cavity wall is like a sandwich. There’s an outer layer of one material, to keep out the rain, then an empty layer of air, then an inner layer of some other material. Somewhere in the sandwich – the position can vary – there’s also a layer of insulation, plus a sheet of something called a vapor barrier.

How to build a cavity wall

In the case of Otto, the exterior was finished in porcelain enamel panels, limestone panels, and glass windows. The interior was finished in ordinary drywall.

Otherwise known as Sheetrock, familiar to office dwellers around the world.

Now let’s focus on the vapor barrier. This is a thin sheet of some kind of pliable fabric, usually plastic. Despite the name, its purpose is not to keep moisture out of the building but exactly the opposite. The vapor barrier is supposed to keep moist indoor air from leaking into the cavity wall.

Why would that be harmful? Because when it’s cold outside, as it is in a Cambridge winter, the cavity will be cold, too, and any moist air leaking into it will condense into water, or perhaps even freeze into ice.

Oh, poopie

Where I come from, water is a property’s greatest enemy. That’s why water and sewer systems are so critical to building integrity, and more fundamentally, why it’s inconceivable to us that housing could exist without water and sanitation.

The effect is like that of water condensing on a cold highball glass on a humid summer afternoon. Eventually, the water begins to damage the insides of the walls, creating rust and rot.

Lovely clear explanation.

At Otto there were times when the walls were soaked through.

Uh-oh.  Wet walls are a bad thing.

Otto, remember, had a pressurized interior. Jim Collins, a Boston architect who is working with Piano on the new museum, puts the situation eloquently. “You’ve got an engine pumping moist air into the wall,” he says. If the vapor barrier were perfect, it would stop the moisture. But in building construction, few things are perfect.

Architect Gwathmey says his wall design was just fine.

Architects always believe their designs are just fine.  Harvard doesn’t.

Harvard sued the architect and the contractor in 1996. As usual in such legal matters, neither side will talk for the record, but word on the street is that the parties split the cost of repairs – repairs that proved, in the end, not to make any difference.

Frank Gehry, who designed MIT’s Stata Center to great architectural acclaim, evidently thinks that leaks are so bourgeois; MIT disagreed.

In a recent interview from his New York office, he suggested other possibilities. “Somebody must have cut holes in the vapor barrier after it was installed,” says Gwathmey. “Maybe a subcontractor installing plumbing or telephone connections.”

Maybe – and who would have warned them that the vapor barrier had to be pristine, totally airtight?

Vapor barriers do, in fact, get punctured. Andy Sebor is a Connecticut engineer who is a recognized national expert in this field. He says failures of this kind are common in art museums of Otto’s vintage. He notes that at the Davis Museum at Wellesley, a building by another Pritzker-winning architect, Rafael Moneo, the curators themselves caused problems. They ruptured the vapor barrier by drilling holes to hang artworks.

As I say, who would know?

“You have to keep the vapor barrier away from the owner’s drill bit,” says Sebor wryly.

Hope that wasn’t a vapor barrier

Sebor says that both construction methods and curatorial demands were changing in the years leading up to the Otto. “We engineers let architects and museum people go off on their own,” he says. “There was a lot of wishful thinking.”

Maintenance is unforgiving to wishful thinking.

He notes many oddities, such as the fact that at architect Louis Kahn’s gallery at Yale, heating elements were installed in the wall cavity, keeping it dry in winter. Kahn thus solved a problem he may not have known existed. Architects today, says Sebor, are more sophisticated.

There are a couple of other lessons to be learned from Otto.  

Now, Otto, never let the art guys have the final say

One is that long-term institutions like Harvard should build durably. They’re short-sighted when they indulge in the cost-cutting that’s common in the commercial world. At Otto, the exterior limestone and metal panels were connected to the framing structure by what are called “ties,” small metal connectors. Otto’s ties were made of galvanized steel, a material that eventually rusts when subjected to moisture, as it was at Otto. They should have been stainless steel.

You mean, you have to tell architects to specify steel that doesn’t rust?

Another lesson, perhaps, is that architects should be more wary of new ways of building. In Boston we’ve seen two other costly cases of architectural skin disease, the failure of windows at the Hancock Tower and of granite panels at 28 State St., the former Bank of New England – both of which were designed by noted architects.

The plywood wasn’t pretty, but it didn’t fall out

The John Hancock, a soaring parallelogram of dark-blue-tinted glass, was an inspiring sight when it rose in 1974.  Less inspiring was the sight of its enormous glass panels falling out of their frames to explode on the Copley Square pavement, sucked out by the airlift wing effect – another example of the ultimate lesson:

Architects should never have the final say in designing any building.

And that’s my final answer!

Send post as PDF to