New laboratory methods

DETERMINATION OF MOISTURE CONTENT IN HYDRATING

CEMENT PASTE USING THE CALCIUM CARBIDE METHOD
JITKA PODĚBRADSKÁ, JIŘÍ MADĚRA, VRATISLAV TYDLITÁT, PAVLA ROVNANÍKOVÁ*, ROBERT ČERNÝ

Department of Structural Mechanics, Faculty of Civil Engineering,

Czech Technical University, Thákurova 7, 16629 Prague 6

*Institute of Chemistry, Faculty of Civil Engineering,

Technical University of Brno, Žižkova 17, 662 37 Brno

Submitted August 23, 1999; accepted December 10, 1999.

The calcium carbide method is used for measuring the moisture content in hydrating cement paste which could be considered
as a measure of the course of the hydration process. The experimental measurements show that the readings on the acetylene
manometer used for calculating the residual moisture content in the material increase with time significantly up to
approximately six hours which is remarkably different from the measurements on the most of other materials such as bricks,
gypsum, etc., where the final constant readings are achieved already after ten minutes. Two hypotheses for the explanation of
this fact are proposed. The first consists in the assumption that calcium carbide reacts not only with free water but also with
a part of water already bonded in the cement binder. The decomposition of ettringite in the reaction bottle accompanied by
partial loss of bonded water, which becomes available for the reaction with calcium carbide should be the most probable
mechanism in this case. The second hypothesis is based on the assumption that the delay in determining the final values of
residual moisture content is caused by a slow release of water from cement gel due to the high bonding forces of water to the
porous matrix in small gel pores. In any case, the relatively long time necessary for the achievement of final readings makes
the application of calcium carbide method for monitoring the hydration process questionable because its main advantage, fast
determination of moisture content, cannot be effectively utilized.

INTRODUCTION with ethylether, and subsequent gravimetric analysis.

Another possibility is the application of well known fast
Hydration processes of cement in concrete are and reliable methods such as microwave absorption
accompanied by fast incorporation of free water into the (e.g. [11, 12]) or g-ray attenuation (e.g. [13])
cement binder. Monitoring the not incorporated techniques.
(residual) water content in hydrating concrete mixture In the case we are interested only in laboratory
can therefore give useful information on the course of measurements and do not require a nondestructive
hydration starting with a defined content of water in the technique, the calcium carbide method could be
mixture. employed in determining the moisture content in
Due to the rapidity of the process, not every hydrating cement mixtures. The method is known for
common method is suitable for moisture measurements more than 40 years and some improvements were
in hydrating concrete. The classical gravimetric patented in 1989 [14, 15]. Currently is it commonly
treatment consisting in hot air drying at 105 °C cannot used for fast moisture field tests in soils. Its application
be employed because the reaction rates in hydration for building materials is described for instance in
processes strongly depend on temperature. Vacuum [16, 17]. Portable devices working on the principle of
drying at 20 °C is excluded automatically because calcium carbide method are usually presented as
moisture determination takes at least several days. ”Speedy Moisture Testers”, examples of such devices
Application of common electric conductivity methods are those of Mastrad, Soiltest or Riedel de Haën.
or capacitive methods at lower frequencies (e.g. [1 - 4])
is questionable due to the high ion content in the EXPERIMENTAL PART
hydrating mixture which makes it possible to obtain
principally only qualitative data whose quantification is The cement paste was prepared using Portland
doubtful. The well established neutron absorption (e.g. cement ENV 197-1 CEM I 42.5 R (Králův Dvůr, CZ)
[5, 6]) or NMR (e.g. [7-9]) methods cannot distinguish and water. The water to cement ratio w = 0.3 was chosen
between chemically bonded and free water. in our experiments. For moisture measurements, the
Therefore, we have basically two possibilities how CM device based on the calcium carbide method was
to determine the moisture content. The first is water employed (Janser Ltd., Brno, CZ).
removal from the sample by washing out by a fluid The calcium carbide method (see [16] for details)
which is soluble in water and nonreactive with cement, is a direct procedure for the determination of the
such as ethanol with ethylether [10] or isopropylalcohol moisture content where the present water is initially

Ceramics − Silikáty 44 (1) 35-38 (2000) 35

 J. Poděbradská, J. Maděra, V. Tydlitát, P. Rovnaníková, R. Černý1

chemically converted and afterwards the moisture

content is detected by a quantitative registration of the
gaseous product. The method is based on decomposition
of calcium carbide by water according to the exothermal
chemical reaction

CaC2 + 2 H2O → Ca(OH)2 + C2H2 .

During this reaction, acetylene gas is produced whose

pressure in a steel bottle can be measured. Knowing the
acetylene pressure, the temperature and the volume of
the reaction vessel, the amount of water necessary for
the reaction can be calculated assuming the ideal gas
behavior. In order to achieve better accuracy of
moisture determination, producer of every CM device
gives a pressure-moisture conversion table determined
empirically which takes into account for instance the
change of reaction volume by different quantities of the
test material, deviations from the ideal gas law, etc.
Example of such an empirical conversion table between
the acetylene pressure and water content can be found in
[16].
Practical measurements are performed in the Figure 1. Dependence of residual moisture content in hydrating
following way. First, sampling is done using hammer cement paste (w = 0.3, i.e. 23.08% kg/kg of water) on the
and chisel, and the sample is crushed in an agate dish. duration of the original hydration reaction, beginning with the
In our measurements, we always used approximately 5 moment of adding water to cement. The pressure values used
g of cement paste. Then, several steel balls, an ampoule for the calculation of residual moisture content were measured
with calcium carbide powder and the crushed sample always 10 minutes after initiation of acetylene formation.
are put consecutively into the pressure bottle, and the
bottle is closed by the manometer head. The ampoule is
broken by intense shaking of the bottle, and the sample
is well mixed with the calcium carbide by continuous
shaking for 2-5 minutes. After 10 minutes, the first
pressure reading on the manometer is done. A better
accuracy can be achieved by keeping the system in
isothermal conditions. In our experiments, we used
water bath with temperature (25 ± 1) oC where the
bottle was placed.

RESULTS AND DISCUSSION

Figure 1 shows the residual moisture content in the

hydrating cement paste (w = 0.3, i.e., 23.08 wt.% of
water) in dependence on the duration of the original
hydration reaction for the first 3 days, determined by the
calcium carbide method. The readings of
acetylene pressure were always done after 10
minutes. The irregularities on the moisture vs. time
curve are slightly higher than it would correspond to the
precision of the manometer given by the producer (5 %
for our range of readings) but they are still in a
reasonable range taking into account possible additional Figure 2. The time dependence of moisture data calculated
errors due to weighing, interpolation in the pressure- from the continuous pressure readings on the manometer. The
moisture table and empirical character of this parameter in the key is the duration of the original hydration
conversion table. reaction between the moment of adding water to cement and
In our experimental measurements, we additionally the time of initiation of acetylene formation.
analyzed the reliability of measured data obtained by
the standard treatment recommended by the producer of
°
* - 24 hours, × - 72 hours, - 197 hours.

the CM device. We have left the sample in the pressure

bottle for a longer time, up to 40 hours, and recorded the

36 Ceramics − Silikáty 44 (1) 35-38 (2000)

 Determination of moisture content in hydrating cement paste using the calcium carbide method

data from the manometer. Figure 2 shows the time recommended to add periodically cold water to keep
dependence of data measured by the manometer and temperature in the reaction vessel within the range of
recalculated to moisture content, the parameter in the 60 - 70 °C because otherwise it could increase too much
key being the time between the moment of adding water similarly as by reaction of burnt lime with water.
to cement and the time of initiation of acetylene Therefore, we can say that the hypothesis on
formation, i.e. the duration of the original hydration ettringite as a source of additional water for the reaction
reaction. Apparently, the measured moisture content with calcium carbide sounds logical. On the slightly
increases relatively fast up to 2 - 4 hours, and then the elevated temperatures and low water vapor pressure in
increase is much slower so that after 6 hours the the pressure bottle ettringite can loose a substantial part
changes of the measured moisture are practically of its bonded water which is at that moment available
negligible. for the reaction with the remainder of the calcium
We have done test measurements on several other, carbide.
more stable materials. We analyzed the behavior of The second hypothesis explaining the observed
bricks, for instance, and did not see any changes of delay in achieving the final constant pressure reading is
measured moisture content after 10 minutes. Also, we based on the assumption that the transport of water from
investigated some materials containing crystalline the cement gel to the calcium carbide is so slow (in
bound water, among them anhydrite and gypsum. other words the strength of bonds of water to the gel is
Again, we did not observe any changes in the recorded so high) that it takes these several hours.. We suppose
acetylene pressure for longer times than 10 minutes. that the rate of reaction between the calcium carbide and
We have two possible hypotheses for the observed water in the bottle primarily depends on the availability
increase of measured moisture content in hydrating of free water in this case. The main criterion of drying
cement paste. efficiency, namely the ratio of saturated vapor pressure
The first hypothesis consists in the assumption that to the actual vapor pressure in the drying chamber [20],
the increase should not be awarded to free water, and is valid for the condition in the reaction bottle as well as
that the remainder of calcium carbide began to react in any other drying environment. The presence of
with water already bonded in the cement binder. calcium carbide keeps the vapor pressure inside the
In this case, the most probable source of additional bottle at very low but still finite values which places
water for the reaction with calcium carbide would be in limitations to the maximum bond strength which can be
our opinion the decomposition of ettringite, an overcome. Therefore, also this hypothesis sounds
important crystalline sulphoaluminate hydrate of logical because the gel pores have very small diameters,
Portland cement. Its formation largely influences many and the bond strength of water to gel is high which
application properties of concretes made from Portland results in slower water removal from the pores.
cement such as setting, strength, deformation,
durability, etc. [18, 19]. The concerned calcium CONCLUSIONS
sulphoaluminate 3CaO.Al2O3.3CaSO4.31H2O is formed
during the hydration of cement by the reaction of The application of calcium carbide method for
gypsum with tricalcium aluminate, and it is also monitoring the hydration processes in fresh cement
produced by the action of calciumsulphate on calcium paste was found to be questionable because its main
aluminate solutions. It is also formed during the advantage, fast determination of moisture content,
destructive attack of sulphate solutions on cement cannot be effectively utilized due to the relatively long
mortars and concretes. time necessary for the achievement of final readings of
Ettringite formed in cementitious materials can be the acetylene pressure in the reaction bottle necessary
partially or fully destroyed if material is heated at for the determination of the moisture content.
elevated temperatures (above 70 °C) and that ettringite
can reform gradually in the material after a period of Acknowledgements
time delay when the material is destroyed at room
temperature in a moist atmosphere (hence the term This research has been supported by the Grant Agency of
delayed ettringite formation). The 31 moles of water are the Czech Republic, under grant No. 103/99/0024.
retained until a vapour pressure below 0.7 mm at 20 °C
is reached and it is reduced to 26 H2O and 18 H2O on
drying at room temperature over anhydrous CaCl2 and References
solid NaOH. Thermal dehydration leaves 7 – 8 H2O at
105-110 °C, 4 – 6 H2O at 145 °C and 2 – 3 H2O at 1. Berliner M.A.: Izmerenija vlazhnosti. Energija, Moskva
200 °C. The DTA thermogram of 3CaO.Al2O3.CaSO4. 1967.
.31H 2O is characterised by an endothermal 2. Oehme F.: Dielektrische Messmethoden zur quantitativen
decomposition at low temperatures with peak Analyse und fuer chemische Strukturbestimmungen.
temperatures ranging from 110 °C to 180 °C. Verlag Chemie, Weinheim 1958.
The reaction of calcium carbide with water is very 3. Kautsch R.: Elektrische Messtechnik zur Messung
fast and highly exothermal. In the industrial production nichtelektrischen Groessen. VEB Verlag Technik, Berlin
of acetylene from calcium carbide, it is even 1967.

Ceramics − Silikáty 44 (1) 35-38 (2000) 37

 J. Poděbradská, J. Maděra, V. Tydlitát, P. Rovnaníková, R. Černý1

4. Pande A.: Handbook of Moisture Determination and URČENÍ OBSAHU VLHKOSTI V HYDRATUJÍCÍ
Control. Marcel Dekker, New York 1974. CEMENTOVÉ PASTĚ POMOCÍ KARBIDOVÉ METODY
5. Van Bavel C.H.M. in: Humidity and Moisture, p. 95-102,
editor Wexler A., Reinhold, New York 1965. JITKA PODĚBRADSKÁ, JIŘÍ MADĚRA, VRATISLAV TYDLITÁT,
6. Boirat R.: Ann. Radioelec. 22, 87 (1967). PAVLA ROVNANÍKOVÁ*, ROBERT ČERNÝ
7. Rolwitz W.L in: Humidity and Moisture, p. 63-68, editor
Wexler A., Reinhold, New York 1965. Katedra stavební mechaniky, Stavební fakulta,
8. Pople J.A., Schneider W.G., Bernstein H.J.: High České vysoké učení technické, Thákurova 7, 16629 Praha 6
Resolution Nuclear Magnetic Resonance. Mc Graw-Hill,
New York 1959. *Ústav chemie, Stavební fakulta,
9. Pel L., Kopinga K., Brocken H: Magnetic Resonance Vysoké učení technické, Žižkova 17, 662 37 Brno
Imaging 14, 931 (1996).
10. Singh N.B., Singh S.P. in: Proceedings of the Metoda založená na rozkladu karbidu vápníku je použita
International RILEM Workshop on Hydration and Setting, pro monitorování obsahu vody v hydratující cementové pastě,
p. 35-41, editors Nonat A., Mutin J.C. RILEM, Dijon což může sloužit jako měřítko průběhu hydratačních reakcí.
1991. Experimentální výsledky ukazují, že hodnoty tlaku, naměřené
11. Summerhill S.: Instrum. Rev. 10, 419(1967). na manometru, které se používají k výpočtu obsahu vlhkosti
12. Taylor H.B.: AEI Eng. 3, 3(1965). v materiálu, rostou s časem až do přibližně šesti hodin po
13. Nielsen A.: Building Science 7, 257 (1972). zahájení produkce acetylenu, což je dosti odlišné od měření na
14. Jung D., Schneider H.: Patent DD 273684, Nov. 1989. většině jiných materiálů, jako je např. cihla nebo sádra, kde jsou
15. Jung D., Schneider H.: Patent DD 273685, Nov. 1989. konečné hodnoty tlaku dosaženy již po 10 minutách. V tomto
16. Bock P: ConChem-Journal 2, 91(1994). článku navrhujeme dvě možné hypotézy pro vysvětlení tohoto
17. Rapp A.O.: Parkettmagazin 3, 104 (1996). faktu. První spočívá v předpokladu, že karbid vápníku reaguje
18. Lea, F.M.: The Chemistry of Cement and Concrete. nejen s volnou vodou, ale také s částí vody již zabudované
Chemical Publishing Company, INC., New York 1971. v cementovém gelu. Nejpravděpodobnějším mechanismem by
19. Yang, R. Mechanism of Expansion of Cementitious v tomto případě byla dekompozice ettringitu v reakční nádobě,
Materials Due to Delayed Ettringite Formation (DEF). doprovázená částečnou ztrátou vázané vody, která tak může
Centre for Cement and Concrete, University of Sheffield jako volná vstoupit do reakce se zbylým karbidem vápníku.
1999. Druhá hypotéza je založena na předpokladu, že relativně dlouhá
20. Leroy R.: Chim Anal. 36, 294 (1954). doba pro stanovení konečné hodnoty tlaku je způsobena
pomalým uvolňováním vody z cementového gelu v důsledku
Submitted in English by the author. existence dosti velkých vazebných sil mezi vodou a pevnou
matricí v malých gelových pórech. V každém případě relativně
dlouhá doba, potřebná ke stanovení konečné hodnoty vlhkosti
v materiálu poněkud zpochybňuje použití této metody k
monitorování hydratačních procesů cementu v betonu, protože
metoda tím ztrácí svou hlavní komparativní výhodu, jíž je
rychlost stanovení obsahu vlhkosti.

38 Ceramics − Silikáty 44 (1) 35-38 (2000)