RESULTS OF INDOOR RADON MEASUREMENTS IN THE REPUBLIC OF MACEDONIA: – A REVIEW –

Radon and its short lived decay products accumulated in indoor environment are the main source of public exposure to natural radiations. The health effects as well as a great number of natural and artificial factors affecting the radon accumulation in indoor environments are some of the motives for the scientific interest in radon issue. Following this global trend, many studies of indoor radon in the Balkan region, including the Republic of Macedonia have been conducted in the last decade. This paper is an overview of the published papers regarding indoor radon concentration measurements with nuclear track detectors in the Republic of Macedonia. It gives basic information about the spatial and temporal variability of indoor radon over the territory of the country, following by a description of the some factors which affect its variations. This review attempts: to organize available indoor radon results in order to show clear picture of the so far conducted surveys; to highlight the need for continuation of more extensive radon investigation in workplaces; to motivate the building professionals to create as much as possible mitigation methods for indoor radon reduction, to motivate the health professionals for epidemiological studies etc.

Radon and its decay products accumulated in indoor environment are the main natural source of total public exposure to ionizing radiation (UN-SCEAR, 2000). Some epidemiological studies have proven the association between the chronic exposure to high indoor radon concentrations (CRn) and the incidence of lung cancer (Darby et al., 2005). Considering serious health effects, the preparation and implementation of a national radon program (NRP), which main goal is the reduction of the radon risk, becomes a primary task in most countries over the world. The NRP involves a complex organizational structure that includes components for radon monitoring, prevention and mitigation of increased CRn (WHO, 2009). Its preparation requires a multidisciplinary approach conditioned by the understanding of the radon processes in indoor environment.
The source of indoor radon is the radium present in soil and building materials (UNSCEAR, 2000). After its generation, radon emanates from solid grains and transported through porous media until its exhalation from the surface into atmosphere or into indoor environment. Diffusion and advection are the radon transport mechanisms driven by the concentrations and pressure gradients, respectively. In general radon dynamics is complex and depends on many factors resulting in the high variability of radon concentration in indoor air.
The factors that influence the CRn variation could be classified into three groups: geogenic radon potential (GRP), building characteristics and building users lifestyle (habits). GRP is a factor that relates building underlying geology and CRn. The GRP is explained through the soil gas radon concentration and soil permeability. The soil gas radon concentration depends on the 226 Ra specific activity in the soil, which is proven to be in relation with geology. Permeability of the soil is another key geological factor that primarily affects the dynamics of radon in the soil gas as well as its emanation from the soil surface. The density of infiltrated radon flux from the soil, as well as accumulated radon concentration in the building, depends on the characteristics of the building. In the following, different habits of building inhabitants which relate to diverse levels of room ventilation (windows opening) is in function of the inhabitant's lifestyle, but also with meteorological conditions (among other: Yarmoshenko et al., 2016;Nikolopoulos et al., 2014).
The meteorological conditions also significantly affect the radon dynamics resulting in temporal variability of CRn. Just like the meteorological parameters, the indoor radon concentrations are subject to daily, weekly, monthly and seasonal variation (Kolarž et  Because of that, the requirements are to present CRn in a building as annual CRn. The awareness of the potential health problems that could be caused by the increased CRn, led to gradual growing of the investigations of indoor radon over last decades. As in the rest of the world, a large number of scientific works the Balkans, including the Republic of Macedonia, have been done. The papers that appeared in the literature are mainly related to radon sources, radon measurements and are focused as well on factors that affected radon variability (such as: Žunić  This study summarized the results of indoor radon concentrations (CRn) in the Republic of Macedonia published in the literature. These results are based on radon measurements in 520 dwellings, 74 schools and 5 kindergartens. Discussion is oriented to radon spatial variations, the effect of buildings characteristics on radon variation, as well as to seasonal variations. The measured data is also compared with reported data for some countries in the Balkan region.

Study area
The  The area is described by complex geotectonic features of diverse relief and a complex geology. It is broadly divisible into four geotectonic units: Western Macedonian zone, Pelagonian massif, Vardar zone and Serbo-Macedonian massif and a separate Kratovsko-Zletovska volcanic rock area situated between the Vardar zone and the Serbo-Macedonian massif (five different colored in Figure  1). The entire territory has a transitional climate, from Mediterranean to continental.

Radon measurements
In general, the CRn measuring methods are active or passive. There is also a consequent application depending on the goals of the measurements. So, the active methods are used to make short term measurements and they are applied in cases when quickly respond to CRn is needed in some building or to examine radon dynamics for a certain relative-ly short period of time. The passive techniques are time integrated methods in which CRn is measured by nuclear track detectors, exposed over longer period usually in 3, 6 or 12 months. Detector response (calibrated in Bq/m 3 ) presents the average CRn for a certain period. The main advantage of this technique is the possibility to perform simultaneous CRn measurements in many buildings in a particular region.
According to the goals of the four radon surveys performed in Macedonia, all measurements were made using passive nuclear track detectors. The survey code, study area, type of environment with the number of buildings considered in the survey, floor level, detector commercial name, as well as the period of detector exposure for each survey separately are given in Table 1.
The principle of detection, as well as the results processing procedures of the measured concentrations, are explained in detail in the cited references in Table 1. Design of surveys The purpose of the first survey, marked by S1 in Table 1, was to established the basic map of the annual CRn distribution across the whole country territory (Stojanovska et al., 2011a). Seasonal variations of CRn were also studied (Stojanovska et al. 2011a). To that end, the detectors were deployed in the dwellings on the whole territory of the country. The number of detectors in each municipality was determined in dependence of the population density. The CRn was measured in 4 consecutive seasons, so the annual CRn was calculated as arithmetic mean of the 4 measurements.
In order to investigate the CRn variability in primary schools, the S2 survey was conducted in 5 municipalities in the geotectonical Vardar zone. The measurements were done in the spring of 2012 and the annual concentration was assessed using seasonal correction factors (Stojanovska et al., 2011b).
The third S3 survey, was conducted in 3 municipalities, one in the geotectonical Vardar zone and the other two in the Kratovo-Zletovska area. The detectors were deployed in the dwellings, primary schools and kindergartens and exposed for a period of one year.
The fourth research (S4) deals with the seasonal variability of CRn and the effects of seasonal correction on the spatial variability of annual CRn and its uncertainty. The detectors were exposed during the winter period in 2013, in the dwellings throughout the whole territory of the country. The annual CRn was estimated according to the linear model developed in this study.
In all surveys and for each building, the measurements were made in one of the mostly occupied rooms: a living room or bedroom in the dwellings, classroom in the schools, and а playroom or bedroom in kindergartens.

RESULTS
The maximum value (Max), the arithmetic mean (AM), the standard deviation (SD), variation coefficient (CV = SD/AM), the geometric mean (GM) and geometric standard deviation (GSD) of the annual mean radon concentrations, as well as the number (percent) of buildings where indoor radon concentrations exceed national action level of 400 Bq/m 3 for existing buildings, are given in Table 2. The mean values in the S1, S2, S3 and S4 studies expressed through AM and GM were in intervals from 94 Bq/m 3 to 186 Bq/m 3 and from 82 Bq/m 3 to 131 Bq/m 3 , respectively. The CRn variations in each of them indicated by CV and GSD covered intervals: from 57% to 95% and from 1.7 to 2.34.

Building factors affecting radon variability
The GMs of annual CRn measured on the different floors in S1 survey in comparison with GM of annual CRn measured on ground floors in other three surveys are presented in Figure 2.
The GMs of CRn (with GSD in brackets) measured on ground, first and second floor in S1 were: 98 (1.9) Bq/m 3 , 66 (1.7) Bq/m 3 , 57 (1.6) Bq/m 3 , respectively. The GM values which represented S2, S3 and S4 surveys in Figure 2 are the same values given in the Table 2, respectively, since all measurements in that surveys were performed on the ground floors.

Comparison with literature
For comparison, the results from radon surveys performed in some European countries situated on the Balkan Peninsula are given in Table 3.  Table 2 summarizes the basic statistics of the CRn measurements results obtained in the four surveys implemented in the Republic of Macedonia. The measured CRn values were observed to be normally distributed. The main conclusions from S1 were that CRn values have shown spatial variability throughout the country territory. The variabilities of CRn between country regions (Figure1), as well as within them, were significant. In general, they were mainly affected by regional and local geology features, as well as by the building characteristics. The maximum value of 720 Bq/m 3 reported for S1 survey was measured in Pelagonia statistical region (Stojanovska et al., 2011a;). In order to clarify this high value, a small measuring champagn in Kruševo (a small town in the Pelagonia statistical region) was conducted after S1 survey. The elevated values for CRn and natural radionuclides content in top soil were related to Amphibole-biotite granodiorite lithostratigraphy (Stojanovska et al. 2012).
Certain differences between results in the surveys given in the Table 2 are noticeable. They originated from the differences in each study design. For example, the S1 survey was conducted in dwellings distributed throughout the whole country territory, just like S4, but in the S4 measurements were organized only on the ground floor rooms. For these reasons, the CRn values in S4 were higher than in S1 survey. It is interesting to note that in these two studies the percentage of dwellings that exceeds the National action level of 400 Bq/m 3 is the same and it is 2% in both cases. Subsequently, the S2 survey was done on the ground floor in the Vardar zone, where from the previous knowledge elevated CRn in that area was not expected. The results confirmed the same, the obtained CRn results in S2 were lower than those corresponding to S1, also the maximum measured value was < 400 Bq/m 3 . The higher CRn in Table 2 corresponds to S3 survey. In this case, the larger part of the investigated region lies down in Kratovska-Zletovska area, where bed rocks are from the volcanic origin. Detailed investigations of the CRn have been shown that the elevated CRn was in relation with tuff, tuffite of andesite and latite (Stojanovska et al., 2016). The existence of a connection between natural radioactivity and geology on the basis of broader research has been confirmed, not only due to indoor radon, but also to indoor thoron ) and 226 Ra  in spatial variability analysis of surface soils.
The effect of the building floor on annual CRn variations can be seen in Figure 2. The GMs values of CRn measured in the ground floor in all surveys were higher than this in the first and second floors. On the other hand, the difference between the first and second floor concentrations is not significant (Stojanovska et al., 2011a;). Also the GMs values that correspond to the measurements in ground floors over the entire territory of the country (S1 and S4) are practically the same (Figure 2). Opposite to this, the results from S2 survey are lower and from S3 survey are higher than the GMs in S1 and S4 surveys. The GMs of the CRn measurements performed in buildings without basements are lower than in buildings with basements ( Figure 3). Furthermore, the GMs of CRn in buildings with basements cover wider range than the GMs of CRn in buildings without basement. So, the spatial distribution of CRn in the buildings with basements is practically uniform, while in those without basements where the inhabitants are in direct contact with the ground the CRn varied considerably. These variations are associated with different influences of the geogenic radon over the different parts of the country.
From Figure 4, it can be concluded that the higher values of CRn refer to the stone buildings. On the same figure, the differences between the CRn measured in buildings made of brick and concrete are not noticeable for the S1 survey. On the contrary, the higher CRn in buildings built from bricks than these from concrete in the Kratovo-Zletovska area implicate the use of local building materials.
The fact that the global trend to improve energy efficiency in the buildings increase the CRn was confirmed in the S3 survey where the impact of new windows in buildings on CRn variations was studied. The analysis showed that the CRn in buildings with new windows was twice higher in comparison to CRn in buildings with wooden windows (GM = 110 (2.13) Bq/m 3 ). In the same survey, the environmental impact on CRn was examined, but it turned out that there was no difference between urban and rural areas (Stojanovska et al. 2016a).
That long term measurements of CRn in schools can be representative for a given region as well as for dwellings, even the different occupation time in schools was founded in the S3 survey. The results of annual CRn measurements in schools, dwellings and kindergartens were not different (Stojanovska et al., 2016a). There are also no differences between the results of CRn in schools when they are based on the annual measurements (12 months) and when they are in the period exempt from the summer holiday (9 months) (Stojanovska et al., 2016b).
The CRn seasonal variations are presented in Figure 5. The results clearly showed higher CRn for the winter and autumn periods compared to spring and summer. The latter could be ascribed to the fact that during the colder months the buildings are heated, which creates a higher difference in pressure between the soil and the building. On the other hand, in order to save energy in cold periods, the inhabitants keep windows closed. The seasonal measurements in S1 survey were used to developed seasonal correction factors for annual CRn assessment due to CRn measurements in one season. The regional variability of the correction factor was confirmed only for the measurements in autumn (Stojanovska et al., 2011b). Applying the seasonal correction on measurements in winter season: the relative uncertainty budget of the annual CRn increased only for 3% as well as the regional variability over the country is not affected significantly (Stojanovska et al., 2017). Finally we compared the obtained annual CRn in our research with that published in the literature. In general, the obtained GMs of CRn in all surveys ( Table 2) were higher than the worldwide GM of 30 Bq/m 3 reported in the UNSCER 2000 report. On the other hand, the both mean radon concentrations expressed with AM and GM, together with its deviations, are more or less typical as the radon levels reported for some other countries in the Balkan Peninsula (Table 3). CONCLUSION Spatial variations of CRn have been proven in all radon studies conducted in the Republic of Macedonia. In general, in the regions where bed rocks are from volcanic origin the CRn is higher in comparison with the other parts of the country. Also, the characteristics of the buildings have a significant effect on CRn variations. The highest concentrations were measured in buildings on the ground floor without basements built with stone. When new windows are installed in such a building, the CRn increases additionally. The results show that in the few percents of buildings in Macedonia the CRn exceeds the permissible level and they have to be remediated. Further research in this field should be include health and building professionals in order to create an effective radon protection program.