Cement and Concrete Research 130 (2020) 105992

Contents lists available at ScienceDirect

Cement and Concrete Research

journal homepage: www.elsevier.com/locate/cemconres

Discussion

Reply to the discussion of the paper “Earth Concrete. Stabilization T

Revisited” by A.T.M. Marsh, A. Heath, P. Walker, B.V. Venkatarama Reddy,
and G. Habert
⁎
Henri Van Dammea, , Hugo Houbenb
a
< MSE > 2, Multiscale Materials Science for Energy and the Environment, the Joint CNRS-MIT Laboratory, Cambridge, MA, USA
b
CRAterre-ENSAG, Grenoble, France

A R T I C LE I N FO A B S T R A C T

Keywords: Reply to the discussion of the paper “Earth Concrete. Stabilization Revisited” by A.T.M. Marsh, A. Heath, P.
Earth Walker, B.V. Venkatarama Reddy, and G. Habert.
Stabilization
Social perception
Hydro-thermal comfort
Norms

1. Introduction purpose of stabilization which, as they correctly point out, is often a

blend of several items like improving dry and wet strength, erosion
In their discussion of our paper devoted to the “stabilization” of raw resistance, complying with state regulations, with standardized tests, or
earth with Portland cement, Marsh and co-authors correctly point out a with insurance company requirements, or simply improving the socio-
number of discussion elements which were omitted or only briefly psychological perception of the property, especially in developing
discussed in our original paper. While we essentially agree with the economies.
substance of their message as wrapped in their abstract and in their We agree with Marsh et al. when they point out that high- and ultra-
concluding remarks – namely (i) a fair assessment of the effectiveness of high-performance concrete do not belong to the same material world as
cement use in building materials should be based on more than just raw earth and can hardly be compared with it. In the same vein, we
embodied carbon and dry compressive strength and (ii) there are si- may argue that considering a slurry (the zero-cement asymptotic limit
tuations in which cement-stabilized earth materials are the best or ac- for strength in wet conditions their Fig. 1) as a construction material
cepted solutions – we still believe that it was important to warn the doesn't really make sense.
builders that, considering the massive character of earthen architecture, On the other hand, we are fully aware that the perception of non-
mixing earth with even “moderate” amounts of cement is not an en- stabilized earth as a “material for the poor” is a serious handicap for its
vironmentally benign operation. It harms greatly its exceptionally low acceptance in developing regions and a strong driver for stabilization
environmental impact and its unique cradle-to-cradle recyclability, a associated with a more industrial and “modern” perception. As ex-
property which as we move toward a circular economy may well prove tensively discussed in two recent books advocating the use of virgin
to be as important as durability, if not more. Actually, the introduction earth without stabilization [1,2], the best way to modify this perception
of 5 to 10% cement in earth brings construction with stabilized earth in is to multiply and publicize exemplary projects in both developed and
the same environmental and technical discussion frame as construction developing countries. The Ricola Kräuterzentrum built by Lehm Ton
with ordinary concrete or mortar. This was the starting point of our Erde GmbH, a company founded by the Austrian architect Martin
work. Rauch [1], in collaboration with Herzog & de Meuron Architects, the
agency of two Pritzker prize-winning architects, is a very good example
2. Strength and moisture resistance in context (quoted by Marsh et al. also, later in their discussion). It is a large
110 × 30 m2 building with a slender concrete frame invisible from the
The first technical point raised by Marsh et al. is related to the very outside. The 11 m-tall non-load-bearing external walls are made of

⁎
Corresponding author.
E-mail address: [email protected] (H. Van Damme).

https://doi.org/10.1016/j.cemconres.2020.105992
Received 7 January 2020; Accepted 18 January 2020
Available online 28 January 2020
0008-8846/ © 2020 Elsevier Ltd. All rights reserved.
H. Van Damme and H. Houben Cement and Concrete Research 130 (2020) 105992

Fig. 1. Left: A finished façade of the Ricola

Kräuterzentrum (credit: Tobias Fritz). Note the white
“erosion lines”. Note also the concrete basement and the
narrow overreach of the roof. Right: A close look at the
unfinished façade, showing the erosion lines embedded
in the prefabricated rammed earth blocks and the joins
between blocks, to be filled with an earth mortar (credit:
Ricola).

standard 3.36 × 1,30 × 0.45 m3 prefabricated rammed earth elements. inventive solutions may also help avoiding stabilization. Actually, the
The elements are manufactured using non-stabilized soil taken from the situation of cement use in earth is not that different from its use in low
near vicinity (contrary to what Marsh et al. claim, lime and volcanic tuff or medium performance concrete. In concrete also, unnecessarily large
was not used for stabilization; see below). As it should, the building is cement content is too often used as a way to compensate for a poor mix
equipped with “boots” (a concrete base up to ~0.6 m above ground design [5]. The scrutiny of large data bases [6–8] show that the cement
level) and a rather discreet “hat” (a roof with only 40 cm overreach), in content of a ~20 MPa concrete may vary from a remarkably low
order to comply with contemporary aesthetics. However, in order to ~20 kg/m3 to a wastefully large ~400 kg/m3!
control erosion in the rainy climate of Basel, Switzerland, thin (a few
cm) layers of a more resistant sand/lime/pozzolan mixture were in-
corporated every 60 cm during ramming (Fig. 1). After some minor 3. Broadening the panel of sustainability indicators
initial erosion of the most exposed facades, these harder layers are
protruding, slowing down water dripping and protecting the façade A second point raised by Marsh et al. is the necessity to use a
from further erosion. broader panel of sustainability indicators. The embodied carbon of the
Another exemplary project is the EU-funded Cycle Terre project in binder is undoubtedly the primary indicator to consider for a correct
Sevran, a suburb city of Paris [3]. As a new metro line around Paris is environmental assessment of stabilized earth and other cement-based
built, about 40 millions tons of excavation material will be available material – at least, as long as the cement industry has not succeeded in
over a period of about ten years. The project intends to turn this ma- decreasing drastically the clinker content – but many other “second
terial into a useful construction material instead of treating it as a rank” indicators should be considered for a proper life cycle analysis,
waste. In its demonstration phase covering the 2018–2021 period, the including land and resource use, transportation, and human labor. This
Cycle Terre project aims to transform 8000 tons of material yearly into is correct. As far as natural resources are concerned, aggregates - in-
compressed bricks, plasterboard-like natural fiber-reinforced panels, cluding sand - are indeed not everywhere available, either from local
and ready-to-use renderings, mostly without any stabilization. In the sources or via low impact transportation means (boat). Sand is parti-
longer term, the project is to organize a comprehensive professional cularly critical for concrete, especially in regions where industry is not
sector covering all aspects of construction with earth. In the same vein, yet producing large quantities of industrial sand from crushing larger
the Manufacture-sur-Seine project of several architecture agencies in- aggregates [9]. As pointed out by Marsh et al., this can make stabili-
cluding Amateur Architecture Studio of another Pritzker prize-winning zation of local earth a better option than the design of a concrete mix
architect, Wang Shu, will lead to the construction of > 300 apartments with from-far-coming aggregates.
and 20,000 m2 of office space in low rise buildings along the Seine Another “second rank” eco-criterion wisely suggested by Marsh
river, starting in 2021 [2,4]. This type of project, in a highly priced et al. is human labor or, more precisely, the social and economical
urban environment, may radically transform the perception of stabili- benefits of the human labor involved in construction. It is indeed true
zation-free earthen architecture. that the question “Who is most benefiting from the selected materials
Returning to the possible motivations for stabilization, there is no and construction method?” should be considered on equal foot with
question that there are indeed situations where stabilization is fully technical and strictly environmental parameters. In developing econo-
justified like, for instance, unsuitable soil quality and lack of aggregates mies suffering from high unemployment, giving people the opportunity
for correcting the soil grading curve, or extreme climatic conditions to use inexpensive local materials and to have a rewarding activity, at
(monsoon rains, floods). Yet, too often, stabilization with cement is least temporarily, may be a decisive criterion. It is all the more im-
chosen systematically as the easiest way to get away from optimization portant that giving people the opportunity to participate themselves in
of the soil grading curve or from more demanding architectural choices, the construction process is a strong way to build community. As ex-
the first of which still remains compliance with the basic “boots and pressed by B. King, “To do it yourself means to do it with other” [10]. In
hat” rule. As illustrated above for the Ricola Kräuterzentrum, simple more developed economies, the increasing number of “self-builders”
using bio- or geo-sourced eco-materials pertains to the same logic.

2
H. Van Damme and H. Houben Cement and Concrete Research 130 (2020) 105992

water or another substance. They are all based on the solid-liquid

equilibrium of some compound with a large latent heat of melting. A
widespread example is plain concrete with encapsulated paraffin wax
beads [12].
The requirements for the evaporation-condensation process to occur
reversibly and rapidly are rather stringent. The material should not only
have a pore population with the right dimensions for capillary con-
densation in the comfort temperature and moisture range, it should also
have a pore network with much larger pores connected to the smaller
pores and to the surrounding in order to allow for rapid exchange of
relatively large volumes of vapor with the moist atmosphere (1 L of
water is producing 700 L of vapor in standard conditions). This kinetic
condition is much better fulfilled in raw earth – especially non stabi-
lized earth – than in ordinary plain concrete (just think of the difference
in time required to reach equilibrium after a RH change).
The temperature buffering properties stemming from the liquid-
vapor equilibrium of capillary water is most important in some highly
hydrophilic bio-sourced materials, like hemp for instance [13]. In hemp
Fig. 2. Evolution of the temperature in the middle of a rammed earth wall concrete (or “hempcrete”), short pieces of defibrillated hemp stems are
separating two climatic chambers. One chamber is kept at 20 °C and 50% RH. used as aggregate in a mortar matrix. The binder is chosen and the
The other chamber is following a sequence of four stages for 24 h each: [20 °C; concrete is formulated according to the target properties (mechanical,
50%RH], [10 °C; 80%RH], [40 °C; 45%RH], [20 °C; 50%RH]. The continuous acoustic, thermal). Provided the matrix is not too air-tight (a lime
line is the measured temperature. The dashed line is the temperature predicted mortar for instance), a strong thermal regulation capacity is obtained,
using a model which takes into account the thermal conductivity of the wall which goes well beyond what is expected from the “intrinsic” (phase
and its heat capacity, but not the vaporization-condensation of capillary water. change-free) thermal mass. Using a modified coupled mass and heat
Adapted from [14]. transfer model, the contribution of the liquid-vapor capillary water
phase change has been quantitatively described and verified experi-
4. Other material functions mentally [13].
Our understanding of the thermal properties of raw earth has not
The third and last point raised by Marsh and coauthors is the need to yet reached the same level, but there is clear experimental evidence
consider material functions other than strength, in particular those re- that a phase change effect is at work [14]. As shown in Fig. 2, the
lated to hydrothermal living comfort. Once more, Marsh and coauthors temperature buffering capacity of crude earth is much better than one
correctly point out – as we did – the potential benefits of the high can predict using software that neglects phase changes (the type of
thermal mass and large moisture buffering capacity of raw earth walls. software unfortunately used in climatic modeling of buildings and
Provided the architecture has been correctly designed to take advantage certification procedures).
of them, these are properties that should definitely enter into an overall This leads us to reflect on our criteria for certification and for giving
environmental assessment. the “high environmental quality” label to dwellings. These criteria are
Actually, the thermal benefits of using earth as building material are still largely founded upon a few selected physical properties which were
larger than anticipated, thanks to a synergetic hydro-thermal effect chosen in some historical context for a given class of materials owing
which turns raw earth into a natural phase-change material (PCM). The their performance to those physical parameters. The least that can be
thermal mass of a wall may be defined as its heat content per unit wall said is that this manner of proceeding is not very well adapted to the
area. The thermal mass acts as a source of heat when the indoor air introduction of materials working differently, or materials in which
temperature falls below the wall temperature, and as a heat sink in the there is a strong coupling between properties. In fact, our world is an
opposite situation, flattening the diurnal and seasonal temperature extraordinarily diverse showroom of resourceful and frugal answers to
fluctuations [11]. Quantitatively, the thermal mass depends on the wall local material and environmental constraints. This intrinsically verna-
thickness, the density of the wall constituents, and their heat capacity. cular character of traditional dwelling construction materials and
The largest contribution is coming from the mineral constituents, with techniques – to use with measure what is locally available and recycl-
some minor contribution from the water adsorbed on the fine-grained able and invent locally-optimized solutions – is contrasting with the
constituents (clay) or condensed in the capillary pores. The contribu- evolution of the modern habitat construction sector, privileging a lim-
tion of the entrapped moist air is negligible. Earth and plain concrete ited choice of materials and constructive solutions among which
behave very similarly as far as this “intrinsic” thermal mass is con- Portland-type cement-based mortars and concrete occupy a dominating
cerned, with a higher density but a smaller wall thickness for concrete. position. This out-of-context character of the present habitat construc-
What makes raw earth different from concrete is an extra con- tion sector (at least, for low rise buildings) is reinforced by the norms
tribution to the thermal mass stemming from the evaporation and and standards system which in its time was a useful way to inject
condensation of capillary water and its exchange with the surrounding technical knowledge, reliability and safety in construction practice, but
moist air. As the wall temperature is increasing, some capillary con- which is also a confession of our inability to deal with contextual di-
densed water is evaporating and the vapor is escaping from the wall. versity and constitutes an exceedingly rigid straitjacket for innovation.
The enthalpy change associated with this is the very large latent heat of In agreement with Marsh et al., we believe that it is about time to
vaporization of water, ~2.2 103 kJ.kg−1, which is taken from the broaden the lens through which we look at our living space.
surrounding. On the contrary, when the temperature is decreasing,
more water vapor is condensing, now releasing the associated latent
heat. Thus, the reversible liquid-vapor phase change of capillary water Declaration of competing interest
is adding an extra contribution to the wall thermal mass. This is the
principle of all PCMs used for thermal comfort [12]. However, there is The authors declare no conflict of interest.
no industrial PCM which is based on the liquid-vapor equilibrium of

3
H. Van Damme and H. Houben Cement and Concrete Research 130 (2020) 105992

References of cement use, Cem. Concr. Compos. 32 (2010) 555–562.

[9] A. Torres, J. Brandt, K. Lear, J. Liu, A looming tragedy of the sand commons,
Science Mag 357 (2017) 970–971.
[1] A. Heringer, L. Blair Howe, M. Rauch, Upscaling Earth. Material, Process, Catalyst, [10] B. King, The New Carbon Architecture. Building to Cool the Climate, New Society
gta Verlag, ETH Zürich, Zürich, 2019 (151 p.). Publishers, Gabriola Island, BC, Canada, 2017 ((159 p.), chap. 10).
[2] J. Dethier, Habiter la Terre. L’art de bâtir en terre crue, Traditions, Modernité et [11] C. Beckett, D. Ciancio, A review of the contribution of thermal mass to thermal
Avenir, Flammarion, Paris, 2019((511 p.) English edition: The Art of Earth comfort in rammed earth structures, 2nd Int. Conf. on the Sustainable Built
Architecture, Thames and Hudson Ltd., 2020). Environment, Kandy, Sri Lanka, 2012 Accessible on www.researchgate.net/
[3] Cycle Terre, https://www.cycle-terre.eu , Accessed date: 27 December 2019. publication/290899729.
[4] Manufacture-sur-Seine, . [12] T.-C. Ling, C.-S. Poon, Use of phase change materials for thermal energy storage in
[5] H. Van Damme, Concrete material science: past, present, and future innovations, concrete: an overview, Constr. Build. Mater. 46 (2013) 55–62.
Cem. Concr. Res. 112 (2018) 5–24. [13] L. Arnaud, D. Samri, E. Gourlay, Hygrothermal behavior of Hempcrete, chap. 6, in:
[6] B.A. Young, A. Hall, L. Pilon, P. Gupta, G. Sant, Can the compressive strength of S. Amziane, L. Arnaud (Eds.), Bio-Aggregate-Based Building Materials, ISTE,
concrete be estimated from knowledge of the mixture proportion? New insights London, 2013, pp. 179–241 Wiley, Hoboken.
from statistical analysis and machine learning methods, Cem. Concr. Res. 115 [14] M. Moevus-Dorvaux, L. Couvreur, B. Cloquet, L. Fontaine, R. Anger, P. Doat, Béton
(2019) 379–388. d’Argile Environnemental, CRAterre Editions, Villefontaine, France, 2016, pp.
[7] P. Ziolkowski, M. Niedostatkiewicz, Machine learning techniques in concrete mix 40–45 accessible on www.amaco.org.
design, Materials 12 (2019) 1256.
[8] B.L. Daminelli, F.M. Kemeid, P.S. Aguiar, V.M. John, Measuring the eco-efficiency