Potential threat to human health during forest fires in the Belarusian exclusion zone

Abstract

Forest fires are of special interest for Belarus because of radioactive contamination caused by the 1986 accident at the Chernobyl Nuclear Power Plant. This work aims to determine the potential activity of long-lived radionuclides in surface air caused by forest fires and to estimate the potential health threat to firefighters. The methodology is based on measurements of radioactivity released by forest fuel materials using a combustion chamber. The emissions were combined with a simple dispersion model to estimate air concentrations and dose to firefighters. The inhalation dose from transuranium elements tend to be an order of magnitude greater than that from Cesium-137. Although there was variability among sites, about half of the total dose was caused by external radiation, as measured by dosimeters. Overall effective radiation dose ranged from 3 to 7 uSv for a 1-h exposure, far below the annual effective dose of 20 mSv for workers and 1 mSv for the public. Although, the risk of exceeding annual effective dose limits is low during small fires, such data are important to inform the population and reduce social and psychological stress caused by popular sources.

Copyright © 2018 American Association for Aerosol Research

EDITOR:

• Nicole Riemer

Introduction

Climate and land-use changes have increased the risk of large forest fires in the Eastern European region including the transborder territories of Belarus and Ukraine contaminated by radionuclides (Evangeliou et al. 2015; Zibtsev et al. 2015). According to the Regional Eastern European Fire Monitoring Center (REEFMC), between 1993 and 2013 over 1147 natural fires occurred in the territory of the Chernobyl exclusion zone (Zibtsev et al. 2015). About 263 forest fires occurred in the Belarusian part of this zone from 1996 to 2015.

Since 1986, the forestlands in the Chernobyl exclusion zone have experienced some severe fires (Usenia 2002). The latest large forest fire in the contaminated territories occurred in 2015 near the former settlement Kozhushki with ¹³⁷Cs contamination density over 1480 Bq km⁻² (current territory of Polesie State Radioecological Reserve). This fire was classified as a crown fire with the total burnt area over 60 ha. There were also two episodes (April and August) of a large fire in the Ukrainian part of the exclusion zone in the same year. This fire is a prime example of transboundary air pollution case because smoke moved to the north toward the territory of Belarus (Evangeliou et al. 2016).

After 30 years since the accident at the Chernobyl Nuclear Power Plant, radioactive contamination in the 30-km Chernobyl zone is determined by long-lived radionuclides of ¹³⁷Cs, and ⁹⁰Sr (with the half-life time 30.1 and 29.1 years, respectively) and alpha-emitting isotopes of ^238,239,240Pu and ²⁴¹Am, also known as transuranium elements (Sokolik et al. 2004; Kashparov et al. 2003). The transuranium elements have a high toxicity and are especially harmful to internal exposure pathway through inhalation. This can cause a local irradiation of lung tissue, lymphatic knots, etc.

The behavior of ²⁴¹Am in the environment is of specific interest for Belarus. The activity of this isotope is slowly growing with time through the beta-decay of ²⁴¹Pu (half-life 14.4 years). Calculations show that the contamination of transuranium elements increased by approximately four times between 1986 and 2006 and area of contamination reached 3.5 × 10³ km² (Knatko et al. 2005). The activity of ²⁴¹Am also increased beyond the 30-km zone but its values do not exceed permissible levels (contamination density by transuranium elements from 0.74 to 1. 85 kBq km² – so-called “areas entitled to resettlement”). By 2059, the activity of ²⁴¹Am will exceed the activity of ^239,240Pu by 2.5 times (Pazukhin et al. 1994). Currently, contribution of ²⁴¹Am to the total alpha activity is over 50%. Before the accident, contamination by ²⁴¹Am was formed through global deposition caused by Soviet nuclear weapon tests, such that the activity concentration in soil of the Berezinsky biosphere reserve in 1985 was only 0.12 ± 0.02 Bq kg⁻¹ (Kudrjashov et al. 2000).

Forest ecosystems were one of the major ecosystems contaminated as a result of Chernobyl fallout. Over 60% of forestland in southeastern region of Belarus is pine forest dominated by Pinus sylvestris L. Many studies carried out after 1986 have shown that the pine forest litter, which is the main forest fire fuel material, can concentrate up to 90% of radionuclides from the total amount in the forest ecosystem (Ipatyev et al. 1999; Bunzl et al. 1995; Agapkina et al. 1995). According to the classification of Kurbatskiy (1962), there are two groups of forest fuel materials. The first one includes grass, mosses, lichens, and the light fraction of forest litter. Forest litter with organic matter of various degree of decomposition, dry needles, leaves, and branches up to 7 mm in diameter belongs to the second group of forest fuel materials.

Through biomass burning, together with the typical combustion products (CO_x, NO_x, organic matter, etc.) radionuclides can be suspended and enter the atmosphere. This resuspension of long-lived radionuclides with smoke aerosols during forest fires can cause additional internal exposure for people in the area of fire (firefighters, forestry workers, local population, etc.) (Yoschenko et al. 2006).

The overall aims of this article are to determine the activity of long-lived radionuclides in the surface air by experimental simulation of forest fire emissions and to estimate the potential health threat through calculation of the effective doses for firefighters.

Materials and methods

Site description and sampling

Pine forests occupy more over 60% of the forestlands in Belarusian Polesie, with about 44% being a mossy type that is highly combustible. The sampling of forest fuel materials was carried out within the 30-km exclusion zone at sampling sites of this type dominated by P. sylvestris L. and Betula pendula Roth. The average age of trees was 60 ± 10 years. Sites differed by the density of radioactive contamination of both ¹³⁷Cs and the transuranium elements. A summary description of the sampling sites is presented in Table 1. Coordinates obtained with the GPS navigator were compared with images from Google maps (Figure 1).

Figure 1. Location of sampling sites.

Table 1. Characteristics of sampling sites.

Site	Location, distance from ChNPP (km)	Coordinates^a	Contamination density
			^137Cs, kBq m^−2	^238,239,240Pu, kBq m^−2	^241Am, kBq m^−2
MS	Masany, Khoiniky district, 15	N 51^o30,771′	4160 ± 210	38 ± 3	70 ± 14
		E 30^o01,236′
KR	Kryuki, Bragin district, 18	N 51°31′	3000 ± 140	12 ± 1	21 ± 5
		E 30°19′
LSK	Lesok, Khoiniky district, 24	N 51^o35,818′	910 ± 8	19 ± 1	27 ± 6
		E 29^o58,634′
ML	Malochki, Khoiniky district, 21	N 51^o33,214′	1790 ± 85	25 ± 2	21 ± 4
		E 29^o58,978′
KN	Kulazhin, Khoiniky district, 20	N 51^o33,728′	5840 ± 280	12 ± 1	18 ± 4
		E 30^o10,110′

^aNoted with the portable GPS navigator GARMIN GPS 12 (GARMIN Olathe, KS, USA).

ChNPP: Chernobyl Nuclear Power Plant.

An air dosimetric survey was conducted with portable dosimeter-radiometers model MKS-AT1125A equipped with scintillation detector NaI (TI) (ATOMTEX 2017). In situ measurements were performed at 10 points on a plot 50 by 50 m at 1 m above ground (for gamma-dose rate) and on the surface (5 cm above ground for alpha and beta particle field). The measurement uncertainty was up to 15%. The data on measured dose rate in the air were used to determine external exposure for firefighters.

Samples of fuel materials were collected at 10 points at each sampling site. Samples of forest litter were taken by a square sampler with sides of 50 cm. We separated forest litter into two layers: A₀L (light fraction of forest litter) and the combined A₀F + A₀H (where F—fermentative and H—humus) layer. Dry needles, leaves, and branches up to 7 mm in diameter (fuel materials from the second group) were collected over 1 m² area in 3 replicates. All samples were dried and weighed in the laboratory. The stock of fuel (biomass) for each site was calculated by summing the dry weight of all samples for the 1 m² area.

To calculate the soil contamination density for the area, composite samples at each site were collected. Sampling was performed using a core sampler with 5 cm in diameter to a depth of 20 cm. Each soil core was split into 1–5 cm deep sections (up to 20 cm) using a spatula. Soil samples from the same depth were then mixed to get a site composite sample. The contamination density of surface (A_s) was calculated using data on the average number and weight of samples, sampler volume, the average amount of forest litter fallen branches, cones, etc. per 1 m².

Measurement of fire emissions from combustion

A combustion chamber was used to measure the activity of long-lived radionuclides released during the burning process (Dvornik et al. 2017). One working cycle of the device includes the following steps:

• sample preparation;

• sample burning;

• sampling of combustion products (aerosols and ash);

• chamber cleaning process.

On the first step, a sample for combustion was prepared. We took into account that dry forest litter is the main fuel type so our sample of fuel materials consisted mainly of forest litter. The weight of the sample ranged from 80 to 110 grams. The fuel was incinerated at temperatures above 600 °C, which created a concentrated stream of smoke. Sampling of aerosols was performed on perchlorovinyl fiber filters with a filtration surface of 10 cm² using a portable aspirator PU-3E. The airflow rate was 400 L min⁻¹ with a sampling duration of 10 ± 2 min. Samples of ash were taken with a spatula and placed into heat-resistant-glass vessels. The portion of aerosol and ash samples was 75% and 25%, respectively. To avoid cross-contamination of fuel materials, the chamber was thoroughly cleaned of residues between experiments.

Sample preparation for activity measurements

Measurements of ¹³⁷Cs activity concentration were undertaken in air-dried samples. Large pieces of plant materials (e.g., branches and pine cones) were homogenized before taking measurements. Soil samples were sieved to 2 mm for the removal of stones and small roots. Each dried sample was placed into a plastic cylindrical vessel with full geometry filling. We used two types of geometry: one with a diameter of 14.5 cm and 11 cm height (Marinelli vessel with volume up to 1 l); the other one with a diameter of 7 cm and 3.2 cm height. Each sample was measured in 3 replicates.

Quality control was performed by comparison of activity measurements of two samples, randomly taken from the main sample. If two measurements differ within the uncertainty, the main sample consider being homogeneous. Otherwise, the sample is rejected.

The activity concentration of ¹³⁷Cs was measured with a γ-spectrometer, equipped with a Ge-detector (CANBERRA (USA) model GX2018). The energy resolution of the detector was 1.8 keV for the ⁶⁰Co-line at 1.33 MeV. Detection efficiency of spectrum for energy 1.33 MeV was 22.4%.

The method of radiochemical analysis for ^238,239,240Pu and ²⁴¹Am involved several steps: preliminary ashing of samples to remove organic matter; radiochemical purification; and separation of plutonium and americium isotopes. ²⁴²Pu and ²⁵²Cf have been used as tracers (chemical markers) for determination of ^238,239,240Pu and ²⁴¹Am, respectively. Plutonium isotopes were separated from americium, uranium, and thorium by an anion-exchange procedure using the ion-exchange columns with ion-exchange resin. Under these conditions, plutonium sorbed on the resin and Am³⁺ entered the filtrate. Next, plutonium was eluted with 0.01 mol L⁻¹ HF solution in 0.3 mol L⁻¹ of HNO₃. Separation of ²⁴¹Am from Ca²⁺, Na⁺, and Mg²⁺ was carried out by using a cationite followed with elution by 4 mol L⁻¹ HCl. A thin-layer target of plutonium and americium was prepared by a coprecipitation procedure with microgram quantities of cerium hydroxide or lanthanum fluoride. A similar methodology is presented in Sokolik et al. (2004).

Alpha-activities of the radionuclides were measured using an alpha-spectrometer CANBERRA Alpha-Analyst with PIPS detectors. The detector’s efficiency was 20 ± 2% with an energy resolution of 15 keV. Measurements error ranged from 7% to 25% and depended on the sample’s activity and radiochemical analysis quality.

Effective dose due to inhalation

The combustion chamber provides a measurement of the radioactivity released per unit area of fuel (Bq m⁻²), with the assumption that the combustion conditions are representative of fuel materials consumption during a typical forest fire. It is difficult to provide exact calculations of the atmospheric concentration that may result in the exposure of fire personnel. For instance, Viner et al. (2015) used a Gaussian puff-plume model to estimate atmospheric dispersion for their conditions, whereas Amiro and Davis (1991) used a parameterization of a near-ground Lagrangian trajectory model.

For the current application, the goal was to provide a reasonable estimate of atmospheric dispersion to estimate exposure and inhalation risk in the near-field (i.e., about a 1.5 m height within a fire environment). We first created an emission rate Q based on the combustion chamber emission data (E, Bq m⁻² fuel) and the rate-of-spread of the fire (R, m s⁻¹), such that (1)

This represents a line source at ground level that moves with the rate of spread of the fire, and assumes that the line is of infinite width (i.e., all personnel on the line are exposed equally). Fire rate-of-spread is highly variable, but most fires in Belarus are relatively small and of low intensity. For surface fires spreading from moderate to fast range, the head-fire rate of spread is normally less than 10 m min⁻¹ (0.17 m s⁻¹) under a pine canopy (Taylor et al. 1997), and we used this rate as a conservatively high value.

We then calculated the atmospheric activity concentration A_v using a dispersion parameter that characterizes the fire as an infinite line source (D, s m⁻²): (2)

We used the trajectory model of Wilson (1982) to estimate D for a continuous line source. We drive the model with an estimate of friction velocity of 0.2 m s⁻¹ (relatively windy for the under-canopy domain) and a roughness length of 0.1 m (short under-canopy vegetation) under neutral atmospheric stability. This gives D = 0.1 s m⁻² at a height of 1.5 m and a downwind distance from the line of 100 m (this is the closest distance to a line source in the Wilson model with the assumed parameters). Note that the concentration near the ground is largely insensitive to height over about two orders of magnitude (with regard to height), and that different atmospheric stabilities change the concentration by less than a factor of three (Wilson 1982). We assumed that the fire personnel would be continuously exposed to this concentration as the fire front moves.

To evaluate internal exposure related to inhalation process we used the following equation (Methodical Instructions 2007): (3) where E_inh is the expected effective dose due to inhalation in mSv; A_vⁱ is the concentration of radionuclide i in the surface air expressed in Bq m⁻³; e_i is the dose coefficient for radionuclide i expressed in Sv Bq⁻¹ (ICRP 2012); V is the inhaled volume in m³ h⁻¹; and t is the time of exposure (hours). The following assumptions were used for calculation:

• the subject is an adult man at heavy work (fireman) without protective equipment;

• the solubility class of the aerosol particles is F (for ¹³⁷Cs) and M (for ^238,239,240Pu and ²⁴¹Am);

• the operating time in the conditions of radioactive contamination is 1 h;

• effective dose coefficients for radionuclides were taken from ICRP (2012, Annex A);

• the activity median aerodynamic diameter (AMAD) was considerate to be 5 µm for ¹³⁷Cs and 1 µm for the transuranium elements.

Following Publication 66 of the ICRP (1994), the ventilation rate during heavy exercise was set to 3 m³ h⁻¹. Analysis of operational data for forest fire suppression in Belarus showed that the mean time of fire suppression within the period 2012–2013 was 61.9 ± 7.4 min, hence the selection of the 1-h exposure.

Statistical analysis

The data statistical analysis was carried out by using MS Excel 2013 for Windows and tools provided by Statistica 10.0 64 Trial version. Data were described by the range, mean, standard error, and coefficient of variation. p-Value <0.05 was considered statistically significant.

Results and discussion

Soil and biomass contamination

Table 2 shows a summary of the experimental data on biomass radioactive contamination at different sites. It is shown that values of dry forest floor density ranged from 3.11 to 3.80 kg m⁻² at different forest sites. The highest values of fuel material stock were observed in the pine stands near the former settlements of Kryuki (KR) and Masany (MS) and were estimated at 3.80 ± 0.2 kg m⁻² and 3.50 ± 0.3 kg m⁻², respectively.

Table 2. Biomass radioactive contamination at experimental sites.

	Biomass density^a, kg m^−2	Activity concentration
		^137Cs, kBq kg^−1	^238Pu, Bq kg^−1	^239,240Pu, Bq kg^−1	^241Am, Bq kg^−1
MS
Forest litter (total)	3.21 ± 0.20	180 ± 3.4	851 ± 72	1960 ± 170	3220 ± 930
Fallen branches	0.1 ± 0.021	12 ± 1.0	15 ± 1.0	24 ± 2.1	36 ± 7.8
Bark + needles + cones	0.2 ± 0.042	33 ± 1.0	22 ± 2.0	48 ± 5.2	68 ± 10
Total	3.50 ± 0.26
KR
Forest litter (total)	3.40 ± 0.15	103 ± 2.0	390 ± 30	670 ± 60	1590 ± 320
Fallen branches	0.10 ± 0.01	11 ± 1.0	14 ± 3.0	18 ± 3.0	33 ± 10
Bark + needles + cones	0.30 ± 0.05	25 ± 1.0	34 ± 5.0	40 ± 10	72 ± 14
Total	3.80 ± 0.21
LSK
Forest litter (total)	3.20 ± 0.20	90 ± 2.0	270 ± 22	630 ± 50	1650 ± 230
Fallen branches	0.1 ± 0.021	9.3 ± 0.2	13 ± 2.0	20 ± 2.0	27 ± 6.0
Bark + needles + cones	0.24 ± 0.10	18 ± 1.0	24 ± 2.0	42 ± 3.0	60 ± 11
Total	3.50 ± 0.31
ML
Forest litter (total)	2.90 ± 0.10	80 ± 2.0	260 ± 20	600 ± 45	1300 ± 210
Fallen branches	0.2 ± 0.022	8.5 ± 0.3	12 ± 1.0	20 ± 4.0	30 ± 6.0
Bark + needles + cones	0.3 ± 0.031	16 ± 1.0	30 ± 1.5	40 ± 3.0	50 ± 12
Total	3.1 ± 0.20
KN
Forest litter (total)	3.0 ± 0.27	170 ± 4.0	680 ± 40	800 ± 50	2060 ± 510
Fallen branches	0.1 ± 0.011	7.5 ± 0.4	12.0 ± 0.85	8.2 ± 0.7	25 ± 5.0
Bark + needles + cones	0.3 ± 0.032	38 ± 1.0	28.0 ± 2.0	14.0 ± 1.3	60 ± 12.0
Total	3.4 ± 0.31

^aDry weight.

In comparison with pine forests situated in the territories without radioactive contamination, the stock of forest fuel materials exceeded normal values by about 0.4–0.6 kg m⁻². The primary cause for this difference is the lack of sanitary cutting (thinning) in the contaminated areas for more than 30 years, leading to a significant accumulation of dead organic matter (Usenia 2002; Mousseau et al. 2014). Forest litter made up over 90% of the fuel materials. Bark, needles, and cones or so-called woody debris ranged from 6% to 9% while the percentage of fallen branches did not exceed 3%. As it is shown in Figure 2, the sampling sites had a similar distribution of radionuclides in organic matter and along the soil profile.

Figure 2. Vertical distribution of long-lived radionuclides in the organic matter and soil profile.

Contamination density of long-lived radionuclides was the highest in the near-surface layers (A₀F + A₀H). The contents of ¹³⁷Cs in A₀F + A₀H varied from 58% (LSK) to 93% (KN) of its total amount in organic matter. Almost the same percentage was observed for total plutonium (²³⁸Pu + ^239,240Pu) and americium. The highest values of ¹³⁷Cs contamination density in near-surface organic layers were 450 ± 12 kBq m⁻² and 470 ± 13 kBq m⁻² at MS and KN sampling sites, respectively. In general, the activity of ¹³⁷Cs is two orders of magnitude higher than for the transuranium elements in the organic matter and through the soil profile.

The activity of radionuclides in soil decreased with depth. Most of the ¹³⁷Cs and transuranium elements were concentrated in the top 0–5 cm soil layer. Among sites, the percentage of ¹³⁷Cs in this layer varied over the range of 41–76%. Similarly, 36–68% of the total plutonium and americium was also associated with 0–5 cm soil layer. The highest values of ²³⁸Pu and ^239,240Pu contamination density in mineral soil were observed at MS and ML sampling sites.

There was a significant difference between total plutonium and americium contamination density. The total inventory of ²⁴¹Am was 1.5–2 times greater than that of total plutonium in both mineral soil and organic matter. For example, at the MS site (the most contaminated area), the contamination density of ²⁴¹Am was 70 ± 14 kBq m⁻² compared to 40 ± 3 kBq m⁻² for total plutonium.

The total content of cesium and transuranium elements in mineral soil is higher than in organic matter, and consequently, in fuel materials. However, mineral soil does not participate in the burning process during forest fires, but it has a significant influence on radionuclide redistribution within the ecosystem. Up to 90% of radionuclides can be associated with forest litter (Teramage et al. 2014; Yoschenko et al. 2006). As it was mentioned above, the content of ¹³⁷Cs was much higher than of the transuranium elements. Distribution of these isotopes in combustion products was expected to be similar.

Radioactive contamination of combustion products

It should be noted that there are two types of combustion products: solid products (ash, products of incomplete combustion) and aerosols. Figure 3 shows that the activity of long-lived radionuclides in solid combustion products (A₁) exceeded the activity in the fuel materials (A₀) by 1.5–4 times. The greatest A₁/A₀ ratio was 4.2 ± 0.4 for ¹³⁷Cs at the LSK sampling site. Through the burning process, forest materials lose up to 90% of their organic matter, and the radionuclides concentrate in the mineral part of the combustion products and ash. The activity of combustion products highly depends on the degree of fuel combustion, and consequently, on its moisture, burning intensity, etc.

Figure 3. Ratio of radionuclides activity concentration in solid combustion products (A₁, kBq kg⁻¹) to its activity concentration in fuel materials (A₀, kBq kg⁻¹).

In terms of radiation safety, the solid combustion products are an open source of ionization with potentially a high activity of long-lived radionuclides. In some cases, ash and the products of incomplete combustion could be considered as radioactive wastes. These would be very low-level waste (VLLW) at our sites based on the measured activity of combustion products (IAEA 2009). The management of this type of waste requires consideration from the perspective of radioprotection and safety.

In the Republic of Belarus, radioactive waste management is regulated by National health standards “Requirements for radiation safety of personnel and population on handling radioactive waste” (OSP 2015). Current approaches to radioactive waste classification in Belarus follow IAEA requirements closely and the ash sampled during the combustion experiments are also categorized as VLLW with activity concentration less than 10³ Bq g⁻¹.

The results of the emission experiments are presented in Figure 4. Each point represents an average value of emitted amount based on three measurements. Some samples during the measurements showed a departure of over 40% and were counted as artifacts (marked with “cross” points at the graphs). Emissions clearly increase with the activity concentration in the combustible fuels. A similar relationship was mentioned in our latest work (Dvornik et al. 2017).

Figure 4. Emission rate of ¹³⁷Cs (Bq m⁻²) and transuranium elements (^238-240Pu, ²⁴¹Am in mBq m⁻²). “Cross” marks indicate artifacts with measurements error over 40%.

The average airborne activity concentrations are presented in Table 3. During the combustion experiments, airborne activity concentration of ¹³⁷Cs was reached 7.3 ± 0.60 Bq m⁻³ (for MS sampling site). Minimum activity was observed while burning of materials from the LSK sampling site. The airborne concentration of ¹³⁷Cs amounted to 0.64 ± 0.031 Bq m⁻³. The average value for that site is 0.6 ± 0.18 Bq m⁻³.

Table 3. The average airborne activity concentration, during the simulation of forest fire.

Site	Av (^137Cs), Bq m^−3	Av (^238Pu) mBq m^−3	Av (^239,240Pu), mBq m^−3	Av (^241Am), mBq m^−3
MS	2.3 ± 0.30	2.8 ± 0.21	4.4 ± 0.53	11.0 ± 0.40
KR	1.4 ± 0.10	1.6 ± 0.13	2.6 ± 0.13	7.6 ± 0.36
LSK	0.6 ± 0.18	1.0 ± 0.25	1.6 ± 0.43	5.3 ± 1.0
ML	0.5 ± 0.19	0.8 ± 0.052	1.0 ± 0.18	3.5 ± 0.50
KN	1.6 ± 0.10	1.7 ± 0.31	2.4 ± 0.44	7.4 ± 1.0

The concentration of the transuranium elements in smoke aerosols was three order of magnitude lower than for cesium. The airborne activity of ²³⁸Pu ranged from 1.7 to 14.3 mBq m⁻³. The highest concentration of ^239,240Pu in smoke aerosols was observed at the MS sampling site at 20 ± 4.0 mBq m⁻³. However, this value may be an outlier caused by high uncertainty during the measurements.

A secondary air contamination by the pollutants previously deposited on the surface can be estimated by the ratio between atmospheric activity concentration (A_v) and contamination density (A_s). Among transuranium elements, ²⁴¹Am has a higher release potential in comparison to Pu isotopes. The average ratio for ²⁴¹Am was (30 ± 9) × 10⁻⁸ m⁻¹ while for ²³⁸Pu and ^239,240Pu it was (20 ± 5) × 10⁻⁸ m⁻¹ and (24 ± 7) × 10⁻⁸ m⁻¹, respectively. In general, from site to site a high variability of this parameter could be observed.

The activity of ²⁴¹Am in aerosols was 1.5–2 times higher than the activity of total plutonium. A similar value was observed in organic matter and soil. However, it should be noted that the transuranium elements have high boiling points in comparison with cesium. The fire temperatures during forest fires are usually not high enough to release transuranium elements through the gas phase. Therefore, Pu and Am would be released in small amounts only as aerosols, but they can be transported by ash and water particles.

Numerous studies have investigated the behavior of radioactive aerosols during wildfires (Kashparov et al. 2000; Yoschenko et al. 2006; Commodore et al. 2012; Baeza et al. 2016). We compared airborne levels of radionuclides obtained from experiments with the data presented in Yoschenko et al. (2006). Although experimental conditions differ, our data on ²³⁸Pu and ^239,240Pu volume activity in smoke aerosols are of the same order of magnitude.

Atmospheric conditions within a fire are highly variable, and it is difficult to estimate a full range of conditions that would encompass periods of very high concentrations (poor dispersion) and those of low concentrations (very high dispersion). However, near-ground concentrations under forest canopies tend to have less variability than those in the far-field, with specific meteorological conditions being a weaker driver.

Effective dose assessment

Firefighters are at increased risk during forest fire suppression activities in radioactively contaminated zones. The total effective dose (D_total) is the sum of external and internal irradiation. External exposure is from the field caused by radioactive soil and vegetation. During forest fires, the radioactive smoke plume can be an additional source of irradiation (internal).

The dose caused by external exposure was determined by dosimetric measurements in situ. The effective dose during a 1 h stay at each contaminated site is presented in Table 4. The calculation of external dose exposure neglected the influence of alpha-emitting isotopes due to their low-penetrating ability of α-particles through the short range in air. We also neglected the dose from external exposure in the smoke cloud, because of its minor value compared to the external exposure from surrounding surfaces (Dvornik et al. 2016).

Table 4. Expected additional effective dose for firefighters.

Isotope	ei, Sv Bq^−1	Internal dose of exposure (inhalation), mSv
		MS	KR	LSK	ML	KN
^137Cs	6.7 × 10^−9	4.6 × 10^−5	2.8 × 10^−5	1.2 × 10^−5	1.1 × 10^−5	3.3 × 10^−5
^238Pu	4.3 × 10^−5	3.7 × 10^−4	2.1 × 10^−4	7.9 × 10^−5	9.8 × 10^−5	2.2 × 10^−4
^239,240Pu	4.7 × 10^−5	6.2 × 10^−4	3.7 × 10^−4	2.2 × 10^−4	1.3 × 10^−4	3.4 × 10^−4
^241Am	3.9 × 10^−5	1.3 × 10^−3	8.8 × 10^−4	6.2 × 10^−4	4.1 × 10^−4	8.7 × 10^−4
External dose, mSv^a		2.4 × 10^−3	4.8 × 10^−3	1.8 × 10^−3	3.1 × 10^−3	5.0 × 10^−3
Total effective dose (Dtotal), mSv		4.8 × 10^−3	6.3 × 10^−3	2.7 × 10^−3	3.7 × 10^−3	6.5 × 10^−3

^aMaximum value for 1 h stay at contaminated site.

The contribution of transuranium elements to the total effective dose (D_total) can reach 50% (for MS sampling site). The duration of inhalation intake during the fire (1 h) is lower than the duration of personnel staying in the contaminated area. Thus, external irradiation from radioactive soil and vegetation is a greater contributor to the total dose. However, inhalation of radionuclides could be significantly dependent on the intensity of breathing. It should be noted that during the hard work, the intensity of breathing increases on average by a factor of two and during very hard work up to four to five times.

Although firefighters and forest workers involved in fire suppression are at the risk of aerosols inhalation, to clarify internal dose, other internal exposure pathways should be taken into account (e.g., ingestion). For these specific professional groups, it is also important to evaluate the average number and duration of fires in contaminated forests, number of shifts. The real values of internal dose are expected to be higher.

According to the hygienic standards in Belarus (OSP 2012), values of annual effective dose for workers and members of the public are 20 mSv and 1 mSv, respectively. During small forest fires at radioactively contaminated areas, the risk of exceeding these annual effective dose limits for firefighters should be considered low.

Conclusions

Although we conclude that Belarus forest fires have a low health impact on forest fire fighters, the methodology presented in this article allows quantification of airborne activity and resuspension of long-lived radionuclides. Additionally, the use of the combustion chamber makes it possible to control the combustion conditions and activity concentration of fuel materials. However, we did not have direct measurements of air concentration or deposition following fire. Near the source of ignition, the airborne activity of radionuclides is greatest and the transuranium elements contribute to the total effective dose. This is confirmed by results mentioned in Yoschenko et al. (2006) and Kashparov et al. (2000). Although, the risk of exceeding the annual effective dose limits is low during small fires, such data are of great importance in informing the population and reducing the social and psychological stress often caused by the influence of mass media.

It is necessary to analyze modern methods of prevention, detection, and firefighting technologies and to develop scientific recommendations to improve the efficiency of prevention of wildfires (especially large ones) including in the border territories of the Chernobyl exclusion zone.

Currently, there is no early warning system for the populations of villages located among forested lands to prepare them for protection of houses or to provide them with knowledge about personal protective equipment that would help to avoid additional doses. A major objective for further research is to develop a special decision supporting system based on the experimental data of long-lived radionuclide behavior in the air during forest fires, spatial information on fire risks, and exposure dose prediction. Such a system would address the problem of protection of firefighters, the local population, and the environment through different implementation strategies at administrative and regional levels.

Acknowledgments

Materials were presented at the 18th Radiochemical Conference (RadChem 2018) held in Mariánské Lázně, Czech Republic, on 13–18 May 2018. The authors would specifically like to thank Dr. Brian Amiro for critical comments and helpful suggestions on the article.

Additional information

Funding

The work reported in this article was conducted under the project GR_20160509 “Analysis of stock of transuranium elements in forest fuel materials and assessment of its potential radiation threat during forest fires” supported by Belarusian State Program for basic research.