radon exhalation of hardening concrete: monitoring cement hydration and prediction of radon...

Journal of Environmental Radioactivity 86 (2006) 354e366www.elsevier.com/locate/jenvrad

Radon exhalation of hardening concrete: monitoringcement hydration and prediction of radon

concentration in construction site

Konstantin Kovler*

National Building Research Institute, Faculty of Civil and Environmental Engineering,Technion e Israel Institute of Technology, Technion City,

Haifa 32000, Israel

Received 22 July 2005; received in revised form 2 October 2005; accepted 23 October 2005

Available online 13 December 2005

Abstract

The unique properties of radon as a noble gas are used for monitoring cement hydration and micro-structural transformations in cementitious system. It is found that the radon concentration curve for hy-drating cement paste enclosed in the chamber increases from zero (more accurately e background)concentrations, similar to unhydrated cement. However, radon concentrations developed within 3 daysin the test chamber containing cement paste were w20 times higher than those of unhydrated cement.This fact proves the importance of microstructural transformations taking place in the process of cementhydration, in comparison with cement grain, which is a time-stable material.

It is concluded that monitoring cement hydration by means of radon exhalation method makes it pos-sible to distinguish between three main stages, which are readily seen in the time dependence of radonconcentration: stage I (dormant period), stage II (setting and intensive microstructural transformations)and stage III (densification of the structure and drying).

The information presented improves our understanding of the main physical mechanisms resulting inthe characteristic behavior of radon exhalation in the course of cement hydration. The maximum value ofradon exhalation rate observed, when cement sets, can reach 0.6 mBq kg�1 s�1 and sometimes exceeds1.0 mBq kg�1 s�1. These values exceed significantly to those known before for cementitious materials.At the same time, the minimum ventilation rate accepted in the design practice (0.5 h�1), guarantees

* Tel.: þ972 4 8292971; fax: þ972 4 8295697.

E-mail address: [email protected]

0265-931X/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jenvrad.2005.10.005

mailto:[email protected]

http://www.elsevier.com/locate/jenvrad

355K. Kovler / J. Environ. Radioactivity 86 (2006) 354e366

that the concentrations in most of the cases will not exceed the action level and that they are not of anyradiological concern for construction workers employed in concreting in closed spaces.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Concrete; Cement; Hydration; Radioactivity; Radon concentration; Radon exhalation rate; Back diffusion;

Ventilation; Porosity

1. Introduction

The present paper deals with radon exhalation during hydration of Portland cement. Radoncontained in trace amounts in hardening material, as other inert gases, does not interact with thesurrounding media chemically, and hence does not interfere with chemical reactions accompa-nying cement hydration. The unique properties of radon as a noble gas can be used for mon-itoring cement hydration and microstructural transformations in cementitious system.

Radon (222Rn) is one of the elements in the decay series of uranium (238U) that occurs asa trace element in almost all natural materials, including cementitious materials used in con-struction such as cement grouts, mortars and concrete. The direct predecessor of radon inthe decay series of 238U is 226Ra, which is often incorporated into the solid matrix of the cementparticles. Before radon can enter the pore space of the material and become available for trans-port to the indoor environment, it has to escape from the solid matrix of the particle. Radon isreleased from mineral grain via alpha recoil. As each radium atom decays by ejecting an alphaparticle, the newly formed radon atom recoils in the opposite direction. By recoiling, the radonatom may then enter the pore space between the grains in a fractured rock. At the same time,the radon atom could also recoil towards the interior of the grain and not enter the pore space. Ifthe material containing radium is stable (i.e. no structural changes and chemical reactions takeplace), the emanation of radon in the pore space is controlled by two main mechanisms respon-sible for this escape: recoil and diffusion through the solid matrix.

Not all radon produced in the particle enter the pore space. Some of the radon atoms decaywithin the pores and some reach the material surface. After being released into the pore space,the radon atoms start diffusing in the pore medium. When the size of the material is smallerthan the diffusion length (the case of a thin concrete slab, for example), practically all radonatoms released into the pores are also released into the atmosphere. In this case the radon pro-duction rate in the pore space equals its release rate in the surrounding air. Usually the emana-tion coefficient is measured by closing hermetically the material in a container. The emanationcoefficient in this case can be defined as the ratio between the radon activity outside the ma-terial Aout

Rn and the radium activity inside the material (ARa):

h¼ AoutRn

ARa

ð1Þ

The radon emanation coefficient (or emanation power) of concrete depends on a number offactors, including the dosage of gypsum (which is used as a set retarder in the clinker) andwhether or not coal fly ash (FA) is added to the clinker. Typical values for emanation powerof the constituents are w30% for gypsum, 5e7% for regular Portland cement and less than1% for fly ash (Bossew, 2003; Kovler et al., 2005; Stoulos et al., 2004). Bossew (2003) alsoreported emanation power data for quartz sand (6%) and mineral slag (0.8%).

356 K. Kovler / J. Environ. Radioactivity 86 (2006) 354e366

When the material has a well-defined surface S, it is convenient to express the exhalationeffect in terms of the exhalation rate E (Bq m�2 s�1):

E¼ lAoutRn

Sð2Þ

In the case of powders, one may express the exhalation rate per mass M of material(Bq kg�1 s�1):

E¼ lAoutRn

Mð3Þ

The works of Balek and his coworkers (Balek and Donhalek, 1983, 1992; Balek and Beck-man, 2002) attributed the changes in radon exhalation rate, mainly, to the changes in surfacearea and microporosity occurring in the hardening material during hydration. Their methodwas called radiometric emanation method (REM). The authors suggested labeling the sampleof cement powder before hydration by the parent isotopes of radon, such as 228Th or 224Ra.The labeling procedure increased radioactivity of their cement samples by 5e6 orders of mag-nitude, and obviously enhanced the sensitivity of their method significantly. At the same time,working with very high radioactivity concentrations requires specific safety precautions. In ad-dition, labeling can lead to uncertainty with the interpretation of the monitoring results, espe-cially when the effect of cement fineness is studied.

The goals of the present work were (1) to study radon exhalation of hydrating cement andcompare it with that of unhydrated cement, avoiding the labeling of cement particles; (2) toanalyze the back diffusion effect and recommend the most appropriate testing arrangement;and (3) on the basis of radon exhalation rates determined in the lab to estimate radon concen-trations, which can develop in construction sites with different air exchange rates.

2. Determination of radon exhalation rate via radon concentrations measured in theclosed chamber

2.1. Time-independent radon exhalation rate

The method of determination of radon exhalation rate E of building materials via radon con-centration C measured in the closed chamber is described by Kovler et al. (2005). This methodbelongs to the group of enclosed sample methods (ESM), according to the classification of Pet-ropoulos et al. (2001), when the sample is enclosed in an airtight container and the radon con-centration growth in the air volume is being followed. This method is based on the assumptionthat E does not depend on time.

It should be taken into account that the concentration is not a characteristic of the material,because it increases with time until achieving a constant valueCN corresponding to the saturationcondition. Maximum radon concentration in a hermetically closed space under saturation at in-finite time, CN (Bq m�3), is related to radon concentration C(t) at a given time t (s) as follows:

CðtÞ ¼ C0e�lt þCN

�1� e�lt

�ð4Þ

where, C0 is the initial radon concentration (Bq m�3) in the container at time t¼ 0, i.e. thebackground.


After measuring the radon concentration at a given time C(t) and calculating CN, the radonexhalation rate E (if the background is small) can be determined by the formula:

E¼ CNlV=M or E¼ CNlV=S ð5Þ

where

M and S e mass (kg) and surface area (m2) of the source, respectively,V e volume occupied by air (volume of the chamber minus volume of the specimen, m3),l e radon decay constant (2.1� 10�6 s�1),E e radon exhalation rate of the source, per unit surface area or mass (Bq m�2 s�1 orBq kg�1 s�1, respectively).

2.2. Time-dependent radon exhalation rate

The method described before is valid only under the assumption of constant radon exhala-tion rate of the sample with time. For hydrating materials, which can drastically change theirmicrostructure within a few hours, transforming from liquid to solid, along with temperatureand humidity changes occurring in the hardening material, which are triggered by chemical re-actions of cement hydration, the radon exhalation rate E should be time-dependent. Therefore,the formula (5) is modified:

EðtÞ ¼ V

S

�dC

dtþ lCðtÞ

�or EðtÞ ¼ V

M

�dC

dtþ lCðtÞ

�ð6Þ

2.3. Back diffusion

If the duration of the test is relatively short (hours), the testing container is completely air-tight, and the measurement circuit is relatively airtight (this depends on the length of air pipesand their permabilities, and also on the permeability of the connectors), the leakage can be rea-sonably neglected.

As far as back diffusion is concerned, the theoretical and experimental work reported by Sa-muelsson (1990) suggests that choosing a chamber with a free volume 10 times larger than thepore volume of the sample may acceptably minimize the back diffusion effect. It means that thedepression of the free exhalation rate after closing the accumulator is less than 10%, a valueacceptable in most experimental situations.

In view of this, it has to be emphasized that Eq. (6) is valid, if (a) there is no leakage of radonout of the container; and (b) the activity concentration in the container air is low compared tothe activity concentration in the pore air of the sample. Leakage and back diffusion effects maybe taken into account if the effective decay constant l*¼lþ lb, l, where lb, l is a decay constantcorrecting for first-order removal of radon by back diffusion and leakage, is further introducedinstead of l.

3. Experimental

The cement used throughout this work was ASTM Type I, ordinary Portland cement, gradeCEM I 52,5 N (the former Israeli grade 300). The concentration of radium, 226Ra, in the


Portland cement was 64.2� 2 Bq kg�1, as determined in cement powder sample by gamma-rayspectroscopy, using an NaI detector (NPP Doza). The sample was dried at 105 �C for 24 h be-fore testing. Cement was measured as a sample of 200 cm3 volume in sealed cylindrical poly-ethylene container after 30 days, to achieve secular equilibrium of the 226Ra progeny.

Cement paste specimens were mixed in a pan mixer, cast in the plastic forms of 1.5 L andintroduced uncovered into the radon chamber for testing radon (Fig. 1) immediately aftercasting.

The method applied in this study, does not require any addition of radioactivity. Only naturalradon is measured. The sensitive continuous radon gas monitor (solid-state alpha spectrometerRAD-7, Durridge Company Inc.) was used to measure radon concentrations. The experimentalset-up allowed simultaneous testing of two samples enclosed in two chambers separated by themetal partition. The maximum duration of the test was 3 days.

The volume of air in each radon chamber (the volume of air in the measurement circuit in-cluded, but the volume of specimen subtracted) was 6.14 L. Such a small chamber made it pos-sible to obtain high radon concentrations within a short time of the test, which was importantfrom the viewpoint of improving sensitivity and reliability. At the same time, the effect of rel-atively small volume of the radon chamber, in comparison with relatively large specimen en-closed had to be estimated from the point of view of possible back diffusion. This part ofthe study is described in the next section.

4. Analysis of back diffusion effect

The effect of back diffusion was studied experimentally on different set-up arrangements.The test was duplicated for each experimental arrangement, and the radon concentrations

Fig. 1. Experimental set-up for simultaneous measuring of radon exhalation rate in two adjacent chambers.


were averaged. Totally, three different testing arrangements were prepared, including the mainexperimental set-up described before:

1. M¼ 5.0 kg; V¼ 5.80 L (M/V¼ 862 kg m�3);2. M¼ 2.7 kg; V¼ 6.14 L (M/V¼ 439 kg m�3);3. M¼ 2.7 kg; V¼ 85.84 L (M/V¼ 31 kg m�3).

For the preparation of these testing arrangements, three combinations from two differentspecimen masses (5.0 and 2.7 kg) and two volumes (6 and 86 L) were used. The large radonchamber was used in one of our previous works (Kovler et al., 2004). The results of the meas-urements of radon concentration versus time for these experimental arrangements are shown inFig. 2. As expected, the highest concentration was achieved in the chamber with the maximumM/V ratio (862 kg m�3), and the lowest one in that having the minimum M/V ratio (31 kg m�3).

The cement paste samples were made at water to cement ratio of 0.33. The tests were ex-ecuted in the experimental room with a temperature of 30 �C.

In order to estimate the pore volume of such sample, let us assume the following, accordingto the Powers model of cement paste (Powers and Brownyard, 1948):

(1) the maximum degree of cement hydration at the age of 3 days (the end of testing) is ex-pected to be around 30%;

(2) coefficient of volume expansion of hydrated products is 2.2 (relatively to the volume of un-hydrated cement);

(3) specific volume of cement is 0.32 cm3 g�1.

Fig. 2. Radon concentrations measured for hydrating cement paste made at water to cement ratio of 0.33 at three dif-

ferent specimen mass to dead air volume ratios: M/V¼ 862 kg m�3, M/V¼ 439 kg m�3 and M/V¼ 31 kg m�3.


With these assumptions, we will have the following results:

(1) initial capillary porosity of the cement paste (in the beginning of hydration, az 0%)p0¼ 51%;

(2) porosity calculated for 3 days age p3d¼ 33%.

These estimations of the cement paste porosity did not include the microscopic gel pores ofthe size of w2e3 nm, which were considered as an integral part of the cement gel.

Density of the cement paste is 2000 kg m�3. Consequently, the absolute volume of pores inthe samples of M¼ 2.7 kg, which were used in the experimental arrangements #2 and #3,changes in the course of hydration from 0.685 to 0.445 L, and in the sample of M¼ 5.0 kg (ar-rangement #1) e from 1.272 to 0.825 L. Finally, the ratios between the dead air volume (vol-ume of the chamber minus volume of the specimen) and the pore volume change during the first3 days of hydration is as follows:

(1) arrangement #1: from 4.6 to 7.0;(2) arrangement #2: from 9.0 to 13.8;(3) arrangement #3: from 125 to 193.

From this calculation it follows that the main testing arrangement (#2) shown in Fig. 1 moreor less satisfies the criterion suggested by Samuelsson (1990) that a chamber free volume 10times larger than the pore volume of the sample acceptably minimizes the back diffusion effect.The arrangement #3, in which the dead air volume is an order of magnitude larger than in thearrangement #2, meets this criterion perfectly, with a very high safety factor, so we can assumethat the back diffusion can be neglected for this geometry. In contrast, the arrangement #1 doesnot meet this criterion, and we can expect a significant back diffusion effect.

Multiplying the concentrations C by V/M ratios for each chamber, we can estimate the sig-nificance of the back diffusion effect. Ideally, if there is no back diffusion at all, the productC(V/M ) should be the same for all experimental arrangements, M/V. However, Fig. 3 demon-strates that it is not the case. The back diffusion is strongest for the arrangement #1, when theCV/M-curve deviates from that of the arrangement #3 almost from the beginning of the test.The curves for the arrangements #2 and #3 are much closer, and the deviation from the ‘‘ideal’’curve is observed at approximately 12 h after the commencement of the test. At this momentthe concentrations measured in the chamber #2 achieve rather high values, 5000e6000 Bq m�3. The difference between the arrangement #2 and arrangement #3 may be attrib-uted to the different leakage rate of the two chambers. It can be also seen that the maximumexperimental scatter of the measurements has been obtained for the ‘‘ideal’’ (from the pointof view of minimum back diffusion) geometry, for the largest chamber of 86 L.

We can conclude that the arrangement #2 represents a reasonable compromise to fulfill bothrequirements: to have acceptably small back diffusion and small experimental scatter.

5. Radon exhalation of hydrated cement paste versus unhydrated cement

The radon exhalation of unhydrated cement was compared with that of hydrated cementpaste, made with water to cement ratio of 0.25. Masses of the cement and cement paste samplesin this test were 1420 and 2481 g. The radon concentrations of both materials were measured inthe laboratory under temperatures of 21� 2 �C. In order to compare the measurement data, the


relative radon concentrations (per 1 kg of cement) were calculated. These values arepresented on a logarithm scale in Fig. 4. The low-value range is a characteristic of unhydratedcement. The radon concentration curve for hydrating cement paste starts from zero (moreaccurately e background) concentrations, similar to unhydrated cement. However, the relativeradon concentrations developed within 3 days in the test chamber containing cement pastewere w20 times higher than those of cement. This fact clearly proves the importance ofmicrostructural transformations taking place in the process of cement hydration, in comparisonwith cement grain, which is a time-stable material.

To explain this enormous difference in amount of radon atoms exhaling from the sameamount of radium contained in cement, we have to take into account several important mech-anisms occurring in the course of cement hydration. For this reason, it would be convenient todistinguish between three main stages, which are readily seen in Fig. 4 and correspond withstages in cement hydration and microstructural development: stage I (dormant period), stageII responsible for setting, when intensive microstructural transformations occur, and stage III(the densification of the structure and drying).

6. Kinetics of cement hydration

6.1. Stage I (dormant period)

In the very beginning of the test, when cement paste is still plastic, radon atoms are easilytrapped by water surrounding cement particles, transport to the surface of the sample (mainly,by diffusion) and exhale in the air. It is known that the radon emanation coefficient under sat-urated conditions is smaller than that under dry conditions, whatever the building material

Fig. 3. The product of radon concentration C and dead air volume to sample mass ratio (V/M ) versus time, for different

experimental arrangements (M/V¼ 862 kg m�3, M/V¼ 439 kg m�3 and M/V¼ 31 kg m�3).


(Fournier et al., 2005). This behavior stems from the fact that the recoil distance of the radonatoms in air is about 900 times higher than in water. Thus, water traps the recoiled radon atomsmore easily than air. A thin water film, which continuously engulfs cement particles, would besufficient to stop the recoiled atoms in the water. At the same time, the presence of water ina porous medium is known to dramatically influence radon diffusion. For example, Fournieret al. (2005) found recently that radon diffusion coefficient in cement decreases fromw1� 10�2 (dry material) to w2� 10�6 m h�1 (saturated material).

These two tendencies seem to be counteracting in the fresh cement paste, because there wasalmost no difference in radon concentrations observed between cement and cement paste in thefirst 3 h of the test.

6.2. Stage II (setting, intensive microstructural transformations)

Approximately 3.5 h after mixing cement with water, the radon concentration curve of thecement paste deviates from that of the unhydrated cement, and the concentrations dramaticallygrow. This fact can be a result of the fast microstructural formation occurring in the hydratingpaste. It is also well-known that calcium silicate hydrates (CeSeH) formed by cement hydrationhave very fine structure. CeSeH is not a well-crystallized material. In fact, it is very nearlyamorphous. As a result, it develops as a mass of extremely small irregular particles of indefinitemorphology. As a consequence, hydrated cement paste has very high surface areas. For example,measurements using physical adsorption of water vapor on D-dried calcium silicate pastes indi-cate that CeSeH has surface area of 250e450 m2 g�1, which is three orders of magnitudehigher than in the unhydrated cement (Mindess and Young, 1981). Winslow and Diamond(1974) used low-angle X-ray scatteringe the technique, which allows surface area measurement

Fig. 4. Dependence of relative radon concentration in the chamber (in Bq m�3 per 1 kg of cement) versus time, for

cement and hydrating cement paste made at water to cement ratio of 0.25.


at different moisture states; they revealed areas of about 800 m2 g�1 for saturated samples; theseareas are reduced at lower relative humidity. In other words, the extremely high surface area ofthe newly formed CeSeH, while porosity is still high, should significantly promote radon ex-halation. It is readily observed from the measurements on hardening cement paste.

6.3. Stage III (hardening, slow microstructural transformations and drying)

Finally, one might expect a reduction in radon exhalation rate at later ages. Indeed, as can beseen from Fig. 4, the slope of the radon concentration curve significantly reduces in further. Itcan happen due to the following reasons: reduction in the porosity and changing moisture. Wecannot exclude some influence of moisture change, although this factor does not seem to bea decisive in slowing down the exhalation rate. At the same time it is well-known that the radondiffusion coefficient depends on the porosity. When the cement paste structure is consolidatingand densifying, radon exhalation rate should decrease.

7. Prediction of radon concentrations at concreting in construction site

The dramatic increase of the radon exhalation rate up to the maximum observed in a fewhours after mixing with water is one of the main findings reported and analyzed in the previoussection. The maximum value of radon exhalation rate observed for hydrating cement pastes canreach 0.6 mBq kg�1 s�1 and sometimes exceeds 1.0 mBq kg�1 s�1. Such extremely high valuesof radon exhalation rate significantly exceed all E-values known from the previous literaturedealing with radon exhalation from cementitious materials. Potentially, enormous radon exha-lation rate when concrete sets may lead to the development of high radon concentrations in con-struction sites at closed spaces. This phenomenon may be of radiological concern forconstruction workers, who are employed in casting concrete in a routine form (by daily orweekly basis), and consequently are exposed to enhanced radon concentrations in work places,especially where ventilation is poor.

Assuming that the time dependence of radon exhalation rate in situ remains similar to thatdetermined in the laboratory, the radiological consequences of the extremely high values of ra-don exhalation rate developing a few hours after mixing cement with water, should be evaluatedin terms of predicted radon concentrations for construction site. The result of this analysis isreported hereafter.

For deriving the expression for radon concentrations developing during concreting in closedspaces, let us replace radon decay constant l with effective decay constant l*¼ lþ lv; wherelv is ventilation rate (s�1), in formula (6) and solve it as ‘‘First order linear differentialequation’’:

CðtÞ ¼ e�l�t S

V

Zel

�tEðtÞdt or CðtÞ ¼ e�l�t M

V

Zel

�tEðtÞdt ð7Þ

The ratio of M/V for the real construction conditions is different from those tested in the lab.Let us consider the case of casting concrete floor each square meter of which contains 50 kg ofcement paste. The rest of the concrete composition is assumed to be aggregates, entrained andentrapped air and admixtures; which almost do not contribute to radon exhalation of concretefloor. Let us assume that the height of air column above each square meter of concrete floorin the space, where concreting is executed, is 2.5 m. It means that M/V ratio for this case is


50/2.5¼ 20 kg m�3. This ratio is even less than the lowest ratio applied in the laboratory tests(31 kg m�3). In addition, the space should have some air exchange that is why the back diffu-sion is neglected.

The results of the calculation by formula (7), assuming that the radon exhalation rate devel-ops according to the curve obtained in the laboratory (the case ofM/V¼ 439 kg m�3), but in thesmoothed and approximated form, makes possible to predict the development of radon concen-trations in construction site. The curves of radon concentration calculated for different ventila-tion conditions (from the space closed hermetically to the cases of air exchange rate of 1.0 h�1)are shown in Fig. 5.

Let us assume that the action level in dwellings is 200 Bq m�3, according to ICRP. This limitis shown by dashed line parallel to the time axis. It can be seen that in the cases of extremelypoor ventilation (lv¼ 0.1 h�1 and less) the radon concentrations exceed 200 Bq m�3. More-over, the peak concentrations can increase, when taking into account the background of the givenplace. In this case, the curves shown in Fig. 5 will shift up. However, from the radiologicalpoint of view, such shifting is also not of any concern, because the local maximum of radonconcentration is observed only when concrete sets (4e10 h after casting). Then the concentra-tions drop down to a level which depends on the background and usual radon entry rate fromhardened building materials and other possible sources.

Increasing ventilation a little, from lv¼ 0.1 h�1 to 0.2 h�1 only, improves significantly thesituation. The difference in radon concentrations between lv¼ 0.1 h�1 and lv¼ 0.2 h�1 is ratherimpressive. In any case, the minimum ventilation rate accepted in the design practice is 0.5 h�1,which guarantees that the concentrations in most cases will not exceed the action level and

Fig. 5. Predicted radon concentrations for the case of casting concrete floor under different ventilation conditions (from

hermetically closed space to the air exchange rate of 1.0 h�1).


that they are not of any radiological concern for construction workers employed in concretingin closed spaces.

This conclusion is drawn with a high safety factor. The elevated radon concentrations devel-oping in a few hours after mixing cement with water are very short-lived. In addition, construc-tion sites are usually very draughty, and have much activity going on which will increase airexchange, so the true values are likely to be on the lower side. Also, the workers move fromone construction site to another, so it is unlikely that they are always exposed to the low-air-exchange conditions.

8. Conclusions

1. The radon concentration curve for hydrating cement paste enclosed in the chamber in-creases from zero (more accurately e background) concentrations, similar to unhydratedcement. However, radon concentrations developed within 3 days in the test chamber con-taining cement paste were w20 times higher than those of unhydrated cement. This factclearly proves the importance of microstructural transformations taking place in the processof cement hydration, in comparison with cement grain, which is a time-stable material.

2. The testing arrangement #2 with the sample mass to dead air volume ratioM/V¼ 439 kg m�3 represented a reasonable compromise to fulfill both requirements: tohave acceptably small back diffusion and experimental scatter.

3. Monitoring cement hydration by means of radon exhalation method makes it possible todistinguish between three main stages, which are readily seen in the time dependence ofradon concentration: stage I (dormant period), stage II (setting and intensive microstructuraltransformations) and stage III (densification of the structure and drying).

4. The maximum value of radon exhalation rate observed, when cement sets, can reach0.60 mBq kg�1 s�1 and sometimes exceeds 1.0 mBq kg�1 s�1. These values exceed signif-icantly to those known before for cementitious materials. At the same time, the minimumventilation rate accepted in the design practice (0.5 h�1), guarantees that the concentrationsin most of the cases will not exceed the action level and that they are not of any radiologicalconcern for construction workers employed in concreting in closed spaces.

Acknowledgements

This work was supported by B. and G. Greenberg Research Fund (Ottawa). The authorthanks Eng. Andrey Perevalov, Eng. Pavel Larianovsky and Eng. Aviel Levit for their valuablehelp in the experimental work. The useful comments and advices of Dr. Victor Steiner and Prof.Eugen Rabkin are highly acknowledged.

References

Balek, V., Beckman, I.N., 2002. Theory of emanation thermal analysis. X. Characterization of morphology changes dur-

ing hydration of cementitious binders. J. Therm. Anal. Calorim. 67, 37e47.Balek, V., Donhalek, J., 1983. A new method for investigating cement hydration by radioactive indicators. Cem. Concr.

Res. 13, 1e6.

Balek, V., Donhalek, J., 1992. Radiometric emanation method for monitoring morphology and porosity changes during

radwaste cementation. Cem. Concr. Res. 22, 459e464.Bossew, P., 2003. The radon emanation power of building materials, soils and rocks. Appl. Radiat. Isot. 59, 389e392.


Fournier, F., Groetz, J.E., Jacob, F., Crolet, J.M., Lettner, H., 2005. Simulation of radon transport through building ma-

terials: influence of the water content on radon exhalation rate. Transport Porous Med. 59 (2), 197e214.

Kovler, K., Perevalov, A., Steiner, V., Rabkin, E., 2004. Determination of the radon diffusion length in building materi-

als using electrets and activated carbon. Health Phys. 86 (5), 505e516.

Kovler, K., Perevalov, A., Steiner, V., Metzger, L.A., 2005. Radon exhalation of cementitious materials made with coal

fly ash: part 1 e scientific background and testing of the cement and fly ash emanation. J. Environ. Radioact. 82,

321e334.Mindess, S., Young, J.F., 1981. Concrete. Prentice-Hall, Englewood Cliffs, NJ, USA.

Petropoulos, N.P., Anagnostakis, M.J., Simopoulos, S.E., 2001. Building materials radon exhalation rate: ERRICCA

intercomparison exercise results. Sci. Total Environ. 272, 109e118.

Powers, T.C., Brownyard, T.L., 1948. Studies of the Physical Properties of Hardened Portland Cement Paste. Bulletin

22. Research Laboratories of the Portland Cement Association, Chicago, USA.

Samuelsson, C., 1990. The closed-can exhalation method for measuring radon, 1990. J. Res. Natl. Inst. Stand. Technol.

95 (2), 167e169.Stoulos, S., Manolopoulou, M., Papastefanou, C., 2004. Measurement of radon emanation factor from granular samples:

effects of additives in cement. Appl. Radiat. Isot. 60, 49e54.

Winslow, D., Diamond, S., 1974. Specific surface area of hardened Portland cement as determined by SAXS. J. Am.

Ceram. Soc. 57, 193e197.

radon exhalation of hardening concrete: monitoring cement hydration and prediction of radon...

Documents