I See ICFs III: Alternate Realities

Once again I am a bit surprised to find how long it takes to cover a topic in some semblance of detail. Even with something as relatively mundane, perhaps, as concrete, there is much more to it than meets the eye.

In part one of this series, I walked down to one end of the proverbial see-saw, where I mentioned a visit to a local contractor’s residence, to see a passive-solar oriented, earth-bermed ICF (Insulated Concrete Form) house. There the friendly contractor, smitten with the ICF system apparently, regaled me for a couple of hours with glowing accounts of the numerous advantages to building structures from polystyrene foam forms and concrete. I related many of those pluses to you all, and mentioned that given a choice, based on what I know, between an ICF house and either a conventional stick-framed wooden house, or a trimberframe with SIP cloak, I would choose the ICF house. Reading that first post, it seemed perhaps that I was completely sold on the ICF way of building.

In the second post, I walked down to the other end of the see-saw, where I looked at the materials comprising the ICF system, namely the concrete and the polystyrene foam, in more detail, and concluded that they weren’t especially ‘green’ after all. I did omit a few things, for the sake of brevity – I didn’t consider every type of form used in ICF systems – some are made from concrete-polystyrene mixes, and so forth, however these account for a small slice of the market, and I would say are inferior in terms of insulative value. I also didn’t mention the plastic clips typical in ICF systems, clips which fix the foam panels to one another and allow for the attachment of other materials to the formed wall afterwards. These clips are either plastic, usually recycled plastic, so are relatively benign items, or they are made of steel, which are a somewhat high-embedded energy material and a little less benign. Still, given that the clips are a relatively minor component, they could be set aside in the big-picture consideration.

I also left out the topic of steel rebar in concrete, as that is an area I wanted to look at in a bit more detail today as I come to what I hope is a balanced view of ICF construction and concrete in general. Yes, I will see if I might be able to balance in the middle of that see-saw for a spell, a difficult task for me as I tend to be an ‘all-or-nothing’ sort of person most times.

One of the comments the contractor made upon my visit, after I asked him about durability of ICF construction, was that they “should last for millions of years” and that “the ancient Romans built with concrete, and that has lasted some 2000 years” – so apparently that’s all he needed to know. That wasn’t quite ringing true for me however, as I knew that Portland cement is hardly a product that stretches back into antiquity. Portland cement and Roman cement had to be different somehow.

In fact, the history of Portland cement goes back only to 1824, when a British stone mason named Joseph Aspdin invented it in his kitchen by heating a mixture of finely ground limestone and clay to create a hydraulic lime cement – that is, a cement that hardens with the addition of water. So it seems doubtful that the ancient Romans were using the same product, and if they were they were certainly remiss in their moving forward with the patent application :^). What were the ancient Romans using then?

To find out more on this topic means referring to a fellow named David Moore, a professional engineer who in his retirement devoted 10 years to the study of ancient Roman buildings and the concrete they were made of – buildings like this one for example, the Roman Pantheon:

Looking at the above structure, which I think quite beautiful, consider for a moment that it is made with un-reinforced concrete (i.e., no re-bar or other internal ties) , and that it stood as the largest dome ever constructed for over 1000 years, only being matched, size-wise, in Medieval Italy by Brunelleschi’s dome in Florence (which was actually not circular, but octagonal, and had iron chains for reinforcement). In truth, considering that modern concrete domes, like, say, the Tacoma Dome, are reinforced, the Roman Pantheon has not been equalled by modern man. It’s an amazing structure, and it is concrete.

Roman concrete, like the modern stuff, was composed of sand and aggregates, mixed with a cementitious paste which hardens. The Roman cement was based on limestone, which was heated in a kiln. Limestone contains calcium, carbon, and oxygen. The heating causes the limestone to undergo a chemical reaction in which the carbon and some of the oxygen are driven off, leaving behind a highly reactive product known as quicklime. A scientific term for quicklime is ‘calcium oxide’. Putting the quicklime into water causes another chemical reaction, much bubbling and heat, which yields a white paste known as slaked lime, aka ‘hydrated lime’. Mix the slaked lime with clean sand and you have a great mortar.

Unlike modern Portland cement, in which clinker is reground with a bit of gypsum to produce a powder (as detailed in the previous post) that can be conveniently stored in bags and activated with the simple addition of water, slaked lime has a somewhat shorter shelf life.

To this point, it is hard to see what the Romans had that we don’t; in fact, all in all, Portland cement seems far more convenient. Well, there’s more to it of course. While one type of mortar the Romans used is indeed slaked lime mixed with river sand, three parts sand to one part cement, they also had another type of cement, in which the slaked lime was mixed with pozzolan. Pozzolan takes its name from where it was first found, in the Pozzouli region of Naples. In the classic work by Vitruvius, “Ten Books on Architecture”, in Book II he devotes all of Chapter 6 to “Pozzolana for Concrete Masonry”. I’ll quote him directly:

“There is also a type of powder that brings about marvelous things naturally. It occurs in the region of Baiae and in the countryside that belongs to the towns around Mt. Vesuvius. Mixed with lime and rubble, it lends strength to all the other sorts of construction, but in addition, when moles (employing this powder) are built into the sea, they solidify underwater. Evidently this is why it happens: under these mountains are boiling earths and plentiful springs. These would not exist unless deep beneath there were huge fires, blazing with sulphur or alum or pitch. Therefore these interior fires and the vapor of their flames seep through veins in the ground and make this earth light, and the tufa created there has risen up without any component of moisture. Hence, when these three ingredients [lime, fired rubble, and pozzolana], forged in a similar fashion by fire’s intensity, meet in a single mixture, when this mixture is put into contact with water the ingredients cling together as one and, stiffened by water, quickly solidify. Neither waves nor the force of water can dissolve them.“

The ‘powder’ Vitruvius was referring to was volcanic ash, in case the reader hasn’t guessed. Pozzolan, in the the modern technical definition, is: A siliceous or aluminous material which in itself possesses no cementitious value, but will, in a finely divided form, and in the presence of moisture, chemically react with calcium hydroxide at ordinary temperatures to form compounds possessing cementitious properties.

So, the Romans used the volcanic ash, the pozzolan, in place of the river sand, and like the heating process used for Portland cement in which lime is chemically bonded with clay to form the clinker, with the pozzolan, the heating work had already been done – by the volcano of course! The volcano calcined the pozzolan (that is, the heat drove off the carbon dioxide and water), which enabled it to chemically react with the calcium hydroxide in the slaked lime. A further advantage of the pozzolan over the sand is that the sand crystals, the silica I mean, are so condensed that at the atomic level it is difficult for the calcium hydroxide in the lime to chemically attach and react with it – thus a hardened mix of slaked lime and sand is essentially a material in which the sand particles are merely suspended in the hardened lime. With the pozzolan, on the other hand, there is an amorphous silica atomic structure with many holes in the molecular network, and thus the molecules of calcium hydroxide have many places to attach themselves and subsequently chemically react with the pozzolan. Thus a hardened slaked lime-pozzolan mix is a permanently bonded, chemically altered material, similar to Portland cement.

So, it would seem then that, unless one considers volcanoes to be somehow more environmentally friendly than modern-day concrete plants – both after all use a lot of energy and raw material and produce a lot of greenhouse gases, dust, and so forth, there isn’t much to pick between them in terms how the Romans producing concrete cement and how we do it today- or is there? That Roman concrete still lasted an awfully long time, and judging by the volume of repair work I’m observing to various concrete highway underpasses in Massachusetts these days, with spalled concrete very much apparent, there must have been something the Romans were doing which were are not – what could that be?

According to David Moore, what set the Roman concrete apart were three things:

1) rigid quality control
2) low water to cement ratio
3) expert placement and compaction

It must be said that while cement is produced in plants with extremely high quality control these days, the same cannot be said for how it is mixed into concrete or placed however. The Romans were very picky about their lime quality- the whiter the better (which meant it was free from impurities). Even the stone that would be mixed into the concrete was finely graded and carefully chosen. As Vitruvius writes in Chapter 7 of Book II,

“When it is time to begin building, let the stone be extracted two years earlier, not in winter, but in summer, and lie about in an open place. Whatever stones have been touched and damaged by bad weather in the two years shall be thrown into the foundation courses. All the rest that have not been damaged, once they have passed the test of Nature, will be capable of enduring in construction above ground. These provisions shall be observed not only for squared stone but also in rubble structures“

It seems they weren’t in such a hurry as we seem to be these days. The low ratio of water to cement was an important key to the success of Roman concrete. The Roman product was akin to a zero-slump concrete, a mixture that is so stiff that it will not flow in between the rocks in the aggregate. The placed concrete was then carefully tamped down, the compaction driving out the air bubbles which would otherwise weaken the finished product. In short, unlike modern Portland Cement-based concrete, a wet liquid which pours readily and sets up to near full strength in some 28 days, the Roman concrete was a very dry mix, packed and tamped layer by layer. This type of concrete would not be the equal of Portland cement-based concrete after 28 days, however one or two years on it would have had strength far in excess of the modern stuff.

Surprise-surprise: when you get down to it, modern Portland cement, in terms of the type usually employed in construction, is a product meant to satisfy convenience and economy over any other concern. To quote David Moore,

“We have to use a higher water content because, for economic reasons, we use automated machinery to mix our mortar and aggregate off-site to make concrete and then bring it to the site and pour it into forms that contain steel reinforcing. The mixture must be fluid enough to be worked by the machinery and flow around all the re-bar and forms with minimal manual interaction.“

Ah, yes, the good old ‘minimal manual interaction’ ploy. The modern concrete hardens quickly, and this brings with it a propensity to shrink and crack. Thus the need for re-bar. The cracks form regardless, especially if the mix was over-hydrated, and while the re-bar does hold things together over the short term, eventually, given its exposure to the air or water, the metal begins to corrode and eventually spalls the concrete. And so we spend millions repairing our concrete structures, and in fact are locked into a cycle of having to regularly do such things, with costs escalating with each round of repair. How wise, eh? Makes perfect sense if you are in the concrete business I suppose, but it is hardly a means of creating really durable architecture. And given the cost we so willingly pay in the production of Portland cement, in terms of the greenhouse gases that go along with that, it seems a shame we haven’t the wisdom to justify that profligacy by creating structures that have the potential to last for millennia. That would be a reasonable trade, but no, we’re evidently too efficient (greedy?) for such things.

It ought to be said as well that tamping concrete by hand is hardly easy afternoon work, and that the ancient Romans had thousands of slaves for such tasks – today this sort of process would certainly not be economical on a large scale or for large concrete structures.

The ancient Romans had no need for re-bar or other means of reinforcement because the low amount of water in their mix meant far less shrinkage, and fewer cracks as a result.

Some people are exploring options in terms of creating more durable cement, using the conventional infrastructure to place and pour liquid concrete mixes. A few modern materials exist which can take the place of the pozzolan – the volcanic ash – that the Romans used: fly ash and rice hull ash are two prominent choices. Fly ash is a by-product of coal power plants that otherwise gets trucked to landfills. Rice hull ash is the by-product of burning rice hulls for power generation. Either waste product is an excellent option to replace some of the cement in the concrete mix. Such ash by-products, as they serve to not only strengthen the concrete but to put material to use that would otherwise clog up the landfill, are definitely worthy of consideration. I came across a good read on the topic of using fly ash in concrete, in hope of producing a 1000 year foundation (please click on the .pdf link found on that page).

When employing conventional concrete, even with the addition of fly ash, ensuring that the contractor does not use any more water than necessary and uses sound practices for placement and finishing is very important. For instance, a frequent problem with modern concrete is that the contractor, frankly, does not know what they are doing, and they “overwork” the concrete when finishing the surface, which causes excess water to flow to the surface. When the concrete cures, the surface of the concrete is very weak, and will pop off in a few years, leaving a very rough surface. This is not a fault of the material, but it is the fault of the contractor.

Concrete is a big deal, especially when the lens is set to a wide view. Globally, some 1.6 billion tons of Portland Cement is produced, a figure expected to reach 2 billion tons by next year – the 7% of greenhouse gases mentioned the the previous post. Since the cement is but 12~15% of the content of a typical concrete mix, we are also staring at the fact that the aggregates from that mix amount to some 10 to 11 billion tons of material, and all the mining, trucking and processing that goes along with that. Then there is the water employed in mixing concrete too, some 1 trillion liters per year, potable water no less, and this is not counting the water used for cleaning and washing concrete mixing equipment and in slowing the cure of poured concrete. Add to that the environmental impacts of producing the steel rebar used for reinforcement of that concrete. It’s quite a horrific picture really, and when you think that virtually all that concrete is produced with specifications demanding a curing speed which means that it will crack and likely have an ultimate durability of perhaps 100 years, well, one has to consider the sanity of the whole process. In fact, I am being generous – according to P. Kumar Mehta, PhD (of UC Berkeley, now retired), one of the world’s leading experts on concrete, the designed service life of concrete is but 50 years, and experience has shown that in urban and coastal environments many concrete structures begin to deteriorate in 20 to 30 years. As a side note, the American Society of Civil Engineers gave the US infrastructure an average grade of ‘D’ and estimated that it would take $1.3 trillion to fix the problems. With the recent Stimulus spending, some of that work is now underway across the nation, but really, what’s the wisdom in how we are using this material, given the energy expended to make it, place it and then, gosh darn it, do it all over again?

If buildings and other structures were to be designed to have, say, 150 years of useful service life, that is to say if the life-cycle cost were actually factored in instead of being largely ignored in a rush to get stuff done and make money, then we would be truly maximizing the return on investment, both from a capital perspective but also in terms of natural resource use.

Now, how to reduce the environmental impact of concrete? Here’s a few ideas, gleaned and drawn from my readings thus far:

-reduce the amount of cement in the concrete mix by using pozzolanic by-products such as fly ash to amend the mix. This product will set more slowly than the regular mixes and thus the construction schedules ought to be modified in light of this. It is simply the question of maximizing resources over labor productivity. Less Portland cement means less greenhouse gas production.

-employ construction and demolition wastes (i.e., concrete and masonry rubble, a billion tons of which are produced worldwide each year) for a component of the coarse aggregate in concrete. Additionally, dredged sands and mining waste can serve as fine aggregate components. Some amount of recycled material in the aggregate is far better than the continued expansion of mining virgin aggregate sources.

-according to Professor Mehta, the 1 trillion liters of water used each year in the making of concrete could be cut in half by better aggregate grading and expanding the use of mineral admixtures and superplasticizers. Recycled water is to be preferred to potable water. There are large savings to be had in the use of water for curing concrete – saturated burlap covers with a plastic over sheet, for example, can be employed, as they were on the 1000 year foundation project linked to above.

-employ concrete mixes that use less water, and are enriched with fly ash and superplasticizers, (to 50~60% of the cement mix) to produce a concrete that hardens slowly. This produces a concrete more like the Romans employed, and eliminates all or most of the shrinkage – when you eliminate the cracks which lead to the premature degrade of the material you produce something with far greater durability. Such concrete would also largely preclude the use of rebar, saving money and material. If reinforcement is called for, fiber reinforcement is a viable option that is both quicker and cheaper than steel re-bar.

For those, like me, seeking to build really durable architecture, one has to think carefully about materials, and it is hard to get away from using concrete. While I greatly prefer clay-based walls to concrete walls, ICF type or otherwise, given the superior handling of moisture clay provides, its easy fixability and capacity of being altered down the line with ease, and the low embedded energy of such a material, one is still faced with the foundation and how to make it. The foundation is second only in importance to the roof. Dry laid, mortarless stone is very nice, and when carefully fitted can approach art, as with this classic Peruvian example:

That said, if I don’t have years to spare for such a laborious process, and want a watertight basement, then I am really needing to go with some form of concrete. ICF makes a lot of sense, given the incorporation of both form and insulation with the system. I wouldn’t use regular Portland cement however, but a mix with a good volume of fly ash, some recycled content in the aggregate, fiber reinforcement, superplasticizers, and with as low a slump (minimal water content) as I could get away with. I would specify a 90 day period for curing to usable strength, and use a burlap wrap to keep the temperature of the cure down low and the hardening chemical reaction moving along slowly. I realize I would need an engineer to cooperate with my plans in that regard as most building departments and their codes are not quite with such radical ideas quite yet. It does seem though, if one takes the steps, it is possible to have a really durable foundation, something upon which would be worth putting a truly durable building.

As I said, I’m not so keen on a full ICF house, just ICF for the foundation for me, except where my choice would be limited to choosing between ICF and stick framed, or ICF and SIP/timberframe. I consider those systems even worse for the environment than ICF, both in initial production and in long term recyclability, and poorer uses of a resource which I consider most precious: wood.

On balance, one has to weigh both the environmental impact of producing and obtaining a material, along with any benefits that material may convey over the long term, especially in regards to improved energy use, re-useability, convertability, and so forth. There are some definite negatives to the production of Portland Cement and polystyrene, and the conventional mixed product of concrete is not all that durable. With amendments and modifications to the mix, however, a durable insulated wall would offer convenient construction, excellent strength, along with better thermal performance which conserves fossil and other fuels over the long term. The incorporation of the foam as both form and insulation also means a relatively streamlined building process and less waste. Concrete walls, however, once cast are not so easy or convenient to modify (as will inevitably happen as ‘buildings learn’ – for more, see the book by Stewart Brand linked in the sidebar at the right of this page). The difficulty of structural modification is a major drawback in my mind.

As this blog rolls along, I plan to post more in the future about sustainable building and some of the systems I have developed, and while you can’t go wrong with clay and other low embedded-energy materials generally speaking, it’s nice to now to feel a little less anxious, personally, when it comes to using concrete, so long as it is the right concrete.