Introduction
Radiocarbon (14C) is a common radionuclide in the environment. It is the only radioactive isotope of carbon with a long half-life, 5700 ± 30 years (LNHB). Natural production of radiocarbon takes place at high altitudes in the atmosphere due to the nuclear reaction of thermal neutrons, generated by cosmic rays, with 14N nuclei (IRSN). 14C thus oxidizes to carbon dioxide and results in a flux of 14CO2 in the troposphere where it is incorporated into plants through photosynthesis, as well as in the meteoric and ocean waters through CO2 exchange reactions. This makes 14C an ideal tracer of carbon dioxide coming from the combustion of fossil fuels (Rakowski Reference Rakowski2011; Zazzeri Reference Zazzeri, Graven, Xu, Saboya, Blyth, Manning, Chawner, Wu and Hammer2023) or from other anthropogenic activities such as the nuclear industry (Jean-Baptiste Reference Jean-Baptiste, Fontugne, Fourré, Marang, Antonelli, Charmasson and Siclet2018; Kontul Reference Kontul, Povinec, Sivo and Richtarikova2018).
Burning of fossil fuels causes the emission of carbon dioxide into the atmosphere. The isotopic fingerprint of CO2 from fossil sources is slightly different from that from modern sources, mostly because of the very long formation period of fossil fuels, which therefore do not contain radiocarbon. This was first noted in the mid-1950s by the Austrian chemist Hans Suess (Suess Reference Suess1955) who proved that contemporary tree-sample radiocarbon activity was lower than in samples from the middle of the 19th century before the Industrial Revolution, based on fossil fuels (coal, natural gas, petroleum). This is called the Suess effect (Keeling Reference Keeling, Graven, Welp, Resplandy, Bi, Piper, Sun, Bollenbacher and Meijer2017; Sonnerup Reference Sonnerup, Quay, McNichol, Bullister, Westby and Anderson1999) and it is defined as a change in the ratio of the atmospheric concentrations of heavy isotopes of carbon (13C and 14C) by the admixture of fossil-fuel-derived CO2, which is depleted in 13CO2 and contains no 14CO2 (Tans Reference Tans, De Jong and Mook1979). There are studies in the literature that have shown that atmospheric Δ14C (Stuiver Reference Stuiver and Polach1977) decreased by about 20.0‰ between the years 1890 and 1950 from which 17.3‰ is attributed to the Suess effect and the rest is attributed to the variation in the cosmic ray production of radiocarbon (Stuiver Reference Stuiver and Quay1981). Even if 95% of the total carbon dioxide from the burning of fossil fuels is produced in the Northern Hemisphere due to air-mass mixing between the Northern and Southern Hemispheres, the Suess effect has a global character.
According to the literature, nuclear weapons testing in the 1950s and 1960s nearly doubled the atmospheric radiocarbon. The level of bomb radiocarbon was about 100% above normal levels between 1963 and 1965. The level of bomb radiocarbon in the Northern Hemisphere reached a peak in 1963, and in the Southern Hemisphere around 1965 (Currie et al. Reference Currie, Brailsford, Nichol, Gomez, Sparks, Lassey and Riedel2011; Hua and Barbetti Reference Hua and Barbetti2007; Levin and Kromer Reference Levin and Kromer2004; Levin et al. Reference Levin, Naegler, Kromer, Diehl, Francey, Gomez-Pelaez, Steele, Wagenbach, Weller and Worthy2010). Radiocarbon releases from nuclear facilities are of great interest because radiocarbon is the most important radionuclide from the point of view of the radiation doses received by the population due to the nuclear industry (UNSCEAR 2017). The radiocarbon produced by nuclear facilities is released mostly in the form of gaseous discharges (IAEA 2004). Currently there is no satisfactory technology for removing radiocarbon from effluents and its discharge, therefore radiocarbon releases should be monitored.
The results presented in this paper represent radiocarbon activity in atmospheric carbon dioxide and evergreen plants in Ramnicu Valcea, Romania. The aim has been to study the influence of a thermoelectrical power plant on radiocarbon variation in the vegetation and atmosphere of the Govora Industrial Area. This study is the continuation of the one previously carried out by the authors (Faurescu Reference Faurescu, Varlam, Vagner, Faurescu, Bogdan and Costinel2019) in which the radiocarbon level from the same location was compared with 14CO2 observations at Jungfraujoch, a high-altitude research station in the Swiss Alps conducted to monitor the background 14CO2 level over Europe and to define the reference for regional estimates of fossil fuel CO2 (Hammer Reference Hammer and Levin2017).
Materials and methods
The sampling was done in the Govora industrial area placed about 10 km south of Ramnicu Valcea city, Romania (Figure 1). This location has some particularities. The first particularity to take into account is the fact that near the sampling location operates a nuclear installation, namely “Experimental Pilot for Separation of Tritium and Deuterium” (PESTD) designed for the detritiation of heavy water moderator of CANDU reactors. Until now, PESTD’s normal operation was with heavy water and tritiated water below the exemption level approved by Romanian legislation. Heavy water reactors emit significant amounts of tritiated water and 14C, the 14C being a by-product resulting primarily from neutron activation of 17O from heavy water molecules. Because one of the important releases of PESTD is gaseous radioactive effluent, the monitoring of atmospheric 14C was included in the environmental monitoring program.
Figure 1. Location of the sampling point in the Govora industrial area, Romania (adapted from Google Earth).
Another particularity is that the Govora industrial area operates a 315 MW coal-fired thermoelectric power plant. Due to the Suess effect, a relative decrease of the 14C activity on a local scale is expected as a result of the dilution of the carbon isotopic mixture by fossil carbon. To determine the thermoelectrical power plant’s influence on environmental radiocarbon levels in the Govora industrial area, the radiocarbon activity in the atmosphere and vegetation was determined by liquid scintillation counting and the direct absorption method. The samples were collected for two years as follows: air samples collected every two weeks by active absorption of CO2 into sodium hydroxide (NaOH 3M) with a Raschig tube, and monthly evergreen vegetation (Thuja occidentalis L.) samples.
For the air samples, the sampling device used was made in our laboratory and it is similar to the one developed by the Central Radiocarbon Laboratory affiliated with the Institute of Environmental Physics of Heidelberg University (Hammer Reference Hammer and Levin2017). To ensure a fast and reliable method for radiocarbon measurements in air the schematic set-up of the sampler (Figure 2) is as follows.
Figure 2. The schematic set-up of the atmospheric CO2 sampler by active absorption.
The air is pumped by an air pump with variable flow and passed through an air flow meter that allows adjusting the flow of pumped air. To have enough CO2 to saturate the liquid scintillation cocktail as described below, a sampled air volume of approximately 25 m3 is required. Therefore, for a sampling time of two weeks, this volume of air is reached if air is sampled with a flow rate of 75 L/hr. After the flow meter, the air passes through a gas meter where the sampled air volume is monitored. Then, the air passes through a gas-washing bottle to be humidified. This prevents the NaOH solution from being too concentrated and solid deposits of sodium carbonate in the Raschig tube (Figure 3). The gas-washing bottle is connected to the Raschig glass tube, where the atmospheric CO2 is absorbed by the NaOH solution. The tube is rotated slowly by a motor so that the NaOH solution will be permanently renewed on the Raschig rings. To provide a large surface area within the volume of the tube for interaction between liquid and gas were used 6 mm × 6 mm × 1 mm borosilicate glass Raschig rings (Lenz Laborglas GmbH, Cat. No. 5 1270 06).
Figure 3. (a) Raschig tube and (b) glass Raschig rings.
For vegetation samples, evergreen vegetation samples were chosen i.e. thuja leaves (Thuja occidentalis L.). Evergreen vegetation can photosynthesize through the winter as long as they are not frozen and have access to water (Oquist Reference Oquist and Huner2003). Although maximum photosynthetic rates of evergreen species decrease in winter, the photosynthetic ability is still maintained (Miyazawa Reference Miyazawa and Kikuzawa2005), and CO2 assimilation takes place. Since Romania has a temperate continental climate, the mean annual temperature is 11°C in the south of the country and 8°C in the north of the country (MEW 2005). Even so, the mean temperature in the winter period falls below –3°C, but below-zero periods are short, and evergreen plants’ photosynthesis is stopped only in these periods.
Radiocarbon measurements were done by direct absorption method and liquid scintillation counting (Leaney Reference Leaney, Herczeg and Dighton1994; Varlam Reference Varlam, Stefanescu, Varlam, Faurescu and Popescu2006). The method involves capturing carbon dioxide from the sample in the scintillation cocktail as carbamate and measuring it in an ultra-low level liquid scintillation spectrometer QuantulusTM 1220. The scintillation cocktail is homemade and contains CarbonTrap (Meridian Biotechnologies Ltd), fluorescence substances PPO and bis-MSB (PerkinElmer), and solvents methanol and toluene. Depending on the type of sample, carbon dioxide is obtained either by combustion as in the case of vegetation samples, or by acidification with HCl in the case of the NaOH solution in which atmospheric CO2 was captured. Before combustion, vegetation samples were dried to constant weight at 60ºC, ground, and 10 g of sample was pelletized. Combustion was done in an oxygen atmosphere (17 atm.) in a Parr 1121 combustion vessel (Moghissi Reference Moghissi, Bretthauer, Whittaker and McNelis1975). It can accommodate samples weighing up to 10 grams using oxygen charging pressures up to 20 atm. The vessel is useful for determining trace amounts of tritium, carbon-14, or heavy metals in vegetable matter. In the case of sodium hydroxide solution, pure CO2 was obtained by acidification with HCl, and the pure CO2 was collected in a gas bag (Supelco, product no. 24655). The direct absorption of CO2 into the scintillation cocktail was done by bubbling according to the set-up described in the literature (Faurescu Reference Faurescu, Varlam, Vagner, Faurescu, Bogdan and Costinel2019). This setup contains a pump for gas vehiculation, two purification traps containing an aqueous solution of AgNO3 to retain water vapors and chloride, a flowmeter, and a bubbler with the scintillation cocktail. Gas bubbling for 10 min with a flow rate of about 0.2 L/min through the scintillation cocktail was enough to ensure saturation of amine with the CO2 as carbamate.
Next, the amount of CO2 captured in the scintillation cocktail is determined by weighing. The mass of CO2 is the weight difference between the bubbler with the scintillation cocktail before and after bubbling. For the measurement by liquid scintillation counting method, the entire scintillation cocktail was transferred in 20-mL low-potassium glass vials (PerkinElmer, product no. 6000128) and then counted using an ultra low-level liquid scintillation spectrometer, Quantulus 1220. The counting efficiency was established with the internal method. This method involves the measurement of a standard sample together with a background sample. The background sample was prepared in a similar way as the unknown samples with CO2 obtained by acidification with HCl from marble. Marble was chosen for a background because it does not contain radiocarbon due to the very long time it takes to form. The standard sample is a background sample in which a standard capsule with known activity (PerkinElmer, product no. 1210-122) was dissolved. The standard sample, the background sample, and the unknown samples were counted for 10 cycles of 100 minutes (1000 min counting time). The counting efficiency at the best factor of merit was around 65% with a background of around 2.3 CPM (counts per minute). A double check of the CO2 saturation of the scintillation cocktail was done by checking the spectral quench parameter of the external standard, SQP€. We obtained the same SQP€ values around 630 ± 1%. Data acquisition was performed by using WinQ Windows workstation software, and for spectra processing 1224-534 EASY View software was used.
Results and discussion
Validation of LSC radiocarbon measurement of environmental samples was performed through an intercomparison exercise. The 14C concentration were determined on five environmental samples (thuja leaves, Thuja occidentalis L.) by liquid scintillation counting and accelerator mass spectrometry (AMS) (Table 1). LSC measurements were performed at the National R&D Institute for Cryogenic and Isotopic Technologies – ICSI Rm. Valcea and AMS measurements at the Horia Hulubei National Institute for R&D in Physics and Nuclear Engineering (IFIN-HH). The results of the radiocarbon measurements are reported in pMC according to Stuiver and Polach (Stuiver Reference Stuiver and Polach1977). The δ13C ratio was measured by isotope ratio mass spectrometry on a Delta V IRMS on dried thuja leaves samples. The comparison of results was performed with the ζ-test calculated according to the following equation:
$$\frac{{\zeta = {A_{AMS}} - {A_{LSC}}}}{{\sqrt {U_{AMS}^2 + U_{LSC}^2} }}$$
where AAMS is activity reported by ASM lab and ALSC activity reported by LSC lab and UAMS and ULSC are the appropriate combined uncertainties with coverage factor k=1. Results for which the absolute values of the ζ-test are less than 1.64 are in agreement. Results between 2.56 and 1.64 are questionable.
Table 1. Results obtained from the intercomparison exercise LSC–AMS
The results were satisfactory and showed that radiocarbon measurements by LSC provide reliable values although the LSC uncertainties are much higher than that of the AMS measurements.
To establish the influence of the thermoelectrical power plant on environmental radiocarbon levels in the Govora Industrial Area we measured the radiocarbon activity in the atmosphere and the thuja leaves sampled from this location.
The results have a decreasing trend, but due to local influence caused by the continuous production of fossil CO2, we cannot observe a seasonal variation. For the radiocarbon activity in the air samples, the minimum value recorded was 0.211 ± 0.014 Bq/gC in a winter month, the maximum value of 0.233 ± 0.016 Bq/gC in a summer month, and an average value of 0.226 ± 0.016 Bq/gC. Similar values were recorded for thuja leaves, sampled in the same period as air samples, with a minimum of 0.216 ± 0.014 Bq/gC, a maximum of 0.237 ± 0.018 Bq/gC, and an average of 0.228 ± 0.016 Bq/gC. The Pearson correlation coefficient between the radiocarbon activity in air and the vegetation was 0.92 which suggests a strong correlation between them. The R2 value of 0.842 indicates also a pretty good fit of the estimated trendline values to the measured values (Figure 4, b).
Figure 4. Variation of radiocarbon activity in air and vegetation during (a) the observation period, and (b) the correlation between radiocarbon activity in air and vegetation.
Despite the good correlation between the radiocarbon activity in the atmosphere and thuja leaves sampled in the same period, we wanted to see if the same correlation is maintained for a delay in the assimilation of the radiocarbon concentration by the vegetation for one month, two months or three months. Figure 5 shows the correlations between radiocarbon activity in air and vegetation in the three scenarios.
Figure 5. The correlation between radiocarbon activity in air and vegetation: (a) one month delay, (b) two months delay, and (c) and three months delay.
It can be observed that the correlation between the two parameters is getting worse as we consider that the assimilation of 14CO2 in the vegetation takes place after one month, two months, or three months. For a one-month delay, the Pearson correlation coefficient between the radiocarbon activity in air and vegetation was 0.14 with an R2 value of 0.0189. For a two-month delay, the Pearson correlation coefficient was 0.23 and R2 equal to 0.0543, while for a three-month delay, the Pearson correlation coefficient was −0.10 and R2 equal to 0.0093.
Summary and conclusion
The presented study assesses the influence of a thermoelectrical power plant on environmental radiocarbon levels in the Govora Industrial Area. Due to the Suess effect, a relative decrease of the 14C activity on a local scale was expected as a result of the dilution of the carbon isotopic mixture by fossil carbon. At the same time in the studied location, a nuclear facility designed for the detritiation of heavy water from CANDU reactors operates. Atmospheric 14C is included in the environmental monitoring program due to the gaseous radioactive effluents of this installation.
The sampling and preparation procedures presented have proven to be simple procedures both for 14CO2 sampling into sodium hydroxide solution and for sample preparation by direct absorption method. The method involves capturing carbon dioxide from the sample in the scintillation cocktail and measuring it in an ultra-low level liquid scintillation spectrometer. LSC radiocarbon measurement is less expensive than other techniques but with the disadvantage of higher uncertainties compared with other techniques like accelerator mass spectrometry.
The radiocarbon results in the atmosphere had a decreasing trend, but due to local influence caused by the continuous production of fossil CO2, we cannot observe a seasonal variation. The average value for the atmosphere was 0.226 ± 0.016 Bq/gC. 14C values for vegetation were in the same range as those observed for the atmosphere. The minimums, maximums, and average of 14C activity encountered in the air are found also in the vegetation.
A strong correlation between radiocarbon activity in air and vegetation was highlighted, with a Pearson correlation coefficient of 0.92. A good fit of the estimated trendline values to the measured values was highlighted also by an R2 value of 0.8419.
Despite this good correlation, we considered three scenarios to see if the same correlation is maintained for a delay in the assimilation of 14CO2 in the vegetation of one month, two months, or three months. The correlation between the activity of radiocarbon in the air and vegetation worsens for all these scenarios, which demonstrates that the radiocarbon in the atmosphere is immediately assimilated by the vegetation.
Author contribution
Faurescu Ionut and Varlam Carmen drafted the manuscript and did the interpretation of the experimental data. Faurescu Denisa, Vagner Irina, and Bogdan Diana performed the sampling and the radiocarbon measurements. Costinel Diana performed the δ13C ratio measurements.
Funding
This work was financed by the Ministry of Research, Innovation, and Digitization through Program 1 – Development of the National Research and Development System, Subprogram 1.2 – Institutional Performance – Projects for financing excellence in R&D, Contract no. 19PFE/2021 and PN 23 15 03 01 part of Core Program ICSI Lab2Field. This paper was prepared in connection with the work done for the monitoring program of the Experimental Pilot Plant for Tritium and Deuterium Separation, PESTD. This work has been carried out using infrastructure funded by the European Union via POC/448/1/1 Program (Grant No. 127931 – Extinderea PESTD pentru dezvoltarea de aplicatii de cercetare-dezvoltare în domeniul tritiului – TRIVALCEA. AMS measurements were made free of charge by the Horia Hulubei National Institute for R&D in Physics and Nuclear Engineering (IFIN-HH), thanks to our long collaboration.
Disclosure statement
No potential conflict of interest was reported by the authors.
Open access
