Long-term variation of 90Sr and 137Cs in environmental and food samples around Qinshan nuclear power plant, China

Environmental radioactivity monitoring in the surroundings of nuclear facilities is important to provide baseline data for effective detection in case of any radioactive release in the region. In this work, we report for the first time the long-term monitoring data of 137Cs and 90Sr in environmental and food samples around Qinshan nuclear power plant in 2012–2019. The distribution levels, temporal variations and source terms of 137Cs and 90Sr in the investigated samples were discussed. The annual effective dose (AED) for the local population from the ingestion of foods was also evaluated. Peak values of 90Sr and 137Cs concentrations and 137Cs/90Sr activity ratio were observed in total atmospheric deposition in 2016 and some water and food samples in the following years. This seems to be associated to an additional radioactive input, mostly likely from the operational release of a local facility. This demonstrates that 90Sr and 137Cs, especially the 137Cs/90Sr activity ratio, are sensitive indicators for detecting potential radioactive releases. Nevertheless, overall 90Sr and 137Cs activity concentrations measured during 2012–2019 in this work were at the background levels with average AED far below the internationally permissible limit and recommendation.

. Map showing the geographical location of QNPP (A, scale: 1:42,000,000) and (B, scale: 1:4,000,000), and sampling sites in the surroundings of QNPP marked with filled circles and numbers (C, scale: 1:500,000). The map was produced using software MapInfo Professional using data from https:// bajiu. cn/ ditu/. The radioachemical analysis of 90 Sr in water and ash samples was according to the Chinese national standard procedures [12][13][14][15] . In general, Sr (100 mg) and Y (20 mg) carriers were added to water (adjusted to pH = 1.0 with concentrated HNO 3 ) and ash samples. For water samples (50 L of each), SrCO 3 /CaCO 3 precipitation was used to pre-concentrate Sr. The SrCO 3 /CaCO 3 precipitate was dissolved with 6 mol/L HNO 3 and the sample was adjusted to pH = 1 for further chromatographic purification.
For ash samples (5-30 g), after treated with 3-10 mL of concentrated HNO 3 and 3 mL of H 2 O 2 , and evaporated to dryness, a second ashing was performed at 600 ℃ to ensure the complete decomposition of organic substances. Each ash sample was digested with 30-83 mL of 6 mol/L HCl for two times, the combined leachate was proceeded with CaC 2 O 4 /SrC 2 O 4 precipitation for Sr pre-concentration. The CaC 2 O 4 /SrC 2 O 4 precipitate was ashed in a muffle furnace (600℃, 1 h), and dissolved in 1.5 mol/L HNO 3 for further chromatographic purification.
The processed sample solution for water or ash was loaded onto a chromatographic column (1 cm diameter × 15 cm length) containing di-(2-ethylhexyl) phosphoric acid (HDEHP), at a flow rate of 2 ml/min to separate 90 Y. The column was washed with 40 mL of 1.5 mol/L HNO 3 at 2 ml/min, and Y was eluted with 30 mL of 6 mol/L HNO 3 at 1 ml/min. The separation time was recorded to calibrate the decay of 90 Y. Yttrium was finally prepared as oxalate precipitation for gravimetric measurement of Y chemical yields prior to detection with a low background α/β counter (LB790; German Berthold Technology Company). Each sample was counted for 10 cycles, with each cycle for 100 min [12][13][14][15] .
The determination of 137 Cs in water and ash samples was according to Chinese national standards [16][17][18][19] . Prior to 137 Cs determination, the samples were kept for one month to allow for a complete decay of 131 I. Each dried food sample was screened with gamma spectrometry to eliminate the interferences of other short-lived radiocesium ( 134 Cs, 136 Cs, and 138 Cs) on 137 Cs measurement by beta counting. Cesium carrier (20 mg) was added to the water (50 L, adjusted with nitric acid to pH = 2.0) and ash samples (5-20 g). Cesium contained in each ash sample was leached with 1.5 mol/L HNO 3 for several times. Ammonium phosphomolybdate (AMP) was added to water or leachate sample (from the ash sample) to adsorb cesium. The precipitate was filtered and dissolved in NaOH solution. Cesium was finally separated as Cs 3 Bi 2 I 9 precipitate in citric acid and acetic acid medium for gravimetric determination of chemical yield and measurement of 137 Cs using the low background α/β counter. The measurement time was kept 100 min per cycle, in total of 10 cycles for each sample [16][17][18][19] . All the analytical results obtained in this work were summarized in Tables S2, S3, S4 in the supporting information. This might be associated to the different source input functions and transport processes between 137 Cs and 90 Sr. 137 Cs is particle reactive and readily attached to aerosols and dust, whereas 90 Sr is more water soluble and easier to be dissolved in rainwater 20 . The late occurrence of 137 Cs peak compared to 90 Sr peak in 2016 could be due to the association of 137 Cs to fine particles suspended in the atmosphere which prolonged the residence time of 137 Cs in the atmosphere; and the plum rain in the second quarter (mostly in May) of 2016 flushed the fine particles and thereby prompted the deposition of 137 Cs onto the ground. Nevertheless, both 90 Sr and 137 Cs activity concentrations in the atmosphere around QNPP were generally low in the past decade, and can be considered as baseline concentrations. This low background presents an opportunity to observe any future changes in the 90 Sr and 137 Cs concentrations and therefore can serve as baseline data for any future inputs from QNPP and other potential nuclear activities. Except for a few cases, no significant differences are observed for 90 Sr and 137 Cs activity concentrations in the two type waters between May and October in each year. 90 Sr and 137 Cs annual average activity concentrations in tap water (5.0-8.2 mBq/L for 90 Sr and 1.4-4.2 mBq/L for 137 Cs) are relatively stable compared to those in source water (4.3-11.1 mBq/L for 90 Sr and 0.9-7.0 mBq/L for 137 Cs) during 2012-2019. This should be a consequence of the additional water treatment process (e.g., filtration, disinfection) from raw water to tap water.
The highest 90 Sr and 137 Cs activity concentrations in source water are observed in 2016, with 137 Cs in May and 90 Sr in October, respectively. Whereas the peak values for both radionuclides in tap water, as observed in May 2017, seem to appear one year later compared to the source water. This again reflects the human intervention in the water treatment and delivery process. It is also noted that the annual average of 137 Cs activity concentration in source water decreased rapidly after 2016 to reach similar level as in 2012, while 90 Sr activity concentration was decreasing slowly and still nearly two times higher than the level in 2012. All the measured values for 90 Sr and 137 Cs activity concentrations in water samples around QNPP during 2012-2019 were below the concentration limits recommended by WHO and Chinese national standards 21,22 and also comparable to the reported values for waters collected around other NPPs in China (Table 2).  (Table 2), and far below the Chinese regulatory limit for general foodstuffs. Though no general trends in the temporal variations of 90 Sr and 137 Cs activity concentrations in the four food species can be seen during 2012-2019, the lowest activity concentrations for both radionuclides are mostly observed in rice and the highest values in crucian carp. Interestingly, 90 Sr activity concentration is often higher in mullet compared to salsola, whereas the opposite is obtained for 137 Cs. Even between the two fish species, the difference in 137 Cs activity concentration (up to 10 times) is much higher than that in 90 Sr activity concentration (max. 1.5 times). These features are linked to 90 Sr and 137 Cs levels in the environment where different biota are growing and their different physiological mechanisms for the incorporation and excretion processes of Sr and Cs 28,29 . For example, mullet inhabits salt water and brackish waters whereas crucian carp is a freshwater fish that inhabits lakes, ponds, and slow-moving rivers.
In the freshwater systems, the high particle/colloid loads may effectively scavenge 137 Cs, whereas 90 Sr is kept stable in the water due its highly conservative behavior. Freshwater in some rivers has been found to include high concentrations of 90 Sr 30,31 . Besides, higher salinity in brackish water may facilitate the uptake of 137 Cs (similar to potassium) into the fish. The values of concentration factors for Cs and Sr in marine fishes are 100 and 3, respectively 32 . Therefore, it is reasonable to expect a larger difference in 137 Cs activity concentration between mullet and crucian carp compared to that in 90 Sr activity concentration.  where I Cs and I Sr is the annual intake of 137 Cs and 90 Sr in foods (Bq/y), respectively. e(g) was the dose conversion coefficient of ingested radionuclide 137 Cs and 90 Sr in foods. Dose conversion coefficients of 1.3 × 10 -8 Sv/Bq for 137 Cs and 2.8 × 10 -8 Sv/Bq for 90 Sr were used in this study according to the Chinese national standards .
The obtained AED through ingestion of 137 Cs and 90 Sr in foods for the local population around QNPP was calculated to be in the range of 1.3 × 10 -4 -1.1 × 10 -3 Sv/y during 2012-2019, with a mean value of (3.9 ± 4.0) × 10 -4 Sv/y. In this calculation, the annual per capita human food consumption (0.19, 0.085, 67.84 and 0.40 kg/year/person for mullet, salsola, rice and crucian carp, respectively) by the local public were according to the survey in 2019 and assumed to be constant during 2012-2019. The estimated mean AED for these eight years is considered negligible in respect to the recommended limit (1 mSv/y) established by International Commission on Radiological Protection (ICRP). This reveals that the foods around QNPP were at a safe level during the period 2012-2019.
Source term identification via 137 Cs/ 90 Sr activity ratio. The temporal variations of 137 Cs/ 90 Sr activity ratios in the investigated samples during 2012-2019 are presented in Fig. 5. For total atmospheric deposition, the annual average 137 Cs/ 90 Sr activity ratios ranged from 0.24 to 0.81, with notable peaks observed in 2016. Narrower distribution ranges of 137 Cs/ 90 Sr activity ratios were observed in the water and food samples compared to the total atmospheric deposition, varying within 0.085-1.18 for source water, 0.11-0.80 for tape water, 0.055-0.13 for mullet, 0.069-1.14 for rice, 0.078-1.08 for salsola, 0.052-0.72 for crucian carp, respectively.
The background 137 Cs and 90 Sr levels on the world surface are mostly due to global fallout of atmospheric nuclear weapons testing during 1945-1980. The activity ratio of 137 Cs/ 90 Sr from the global fallout deposition was estimated to be about 1.6 34,35 . Our results indicate that all the 137 Cs/ 90 Sr activity ratios are lower than the global fallout signal. This might be associated to faster removal of 137 Cs compared to 90 Sr originated from global fallout and/or additional input of 90 Sr in the surrounding environment of the study region. The latter is very unlikely, because 90 Sr concentrations measured in this work are in the background level and there is no record of significant radioactive releases in the study region. Besides, as any potential release of 90 Sr (such as civil nuclear industry) would normally be accompanied by 137 Cs release and often with much higher 137 Cs/ 90 Sr activity ratios compared to global fallout, owing to the lower volatility of Sr than Cs 36 . For example, very high 137 Cs/ 90 Sr ratios have been reported in the atmospheric fallout from Fukushima accident (∼1000) 37 and Chernobyl accident (∼250) 38 . In seawater, 137 Cs/ 90 Sr ratios reached 39 ± 1 beyond the coast of Japan due to massive liquid releases of (1) AED = (I Cs + I Sr ) · e(g) www.nature.com/scientificreports/ cooling water in spring 2011 37 . The 137 Cs/ 90 Sr activity ratios before the Fukushima accident were reported in the range of 2.6-18 in Japanese fish whereas the 137 Cs/ 90 Sr activity ratios ranged from 98 to 480 after the accident 39 .
As 137 Cs is more readily adsorbed and immobilized on clay minerals while 90 Sr exhibits a higher mobility, it is reasonable to foresee a decrease in global fallout derived 137 Cs/ 90 Sr activity ratios in some freshwater and food samples. This has been approved in earlier studies where activity ratios of 137 Cs/ 90 Sr in wheat and polished rice from Japan increased by nearly 20 and 3 times, respectively, from 1959 to 1995 40 . 137 Cs and 90 Sr in the dry atmospheric deposition was reported to reflect their signature in the ground-level air, which is mostly from resuspension of the deposited 137 Cs and 90 Sr in soil 41 . Sr is chemically easier to elute than Cs in the soil column by rainwater 20 , therefore 137 Cs/ 90 Sr activity ratio in the typical surface soil usually becomes higher than the typical global fallout value 30,31 . However, lower 137 Cs/ 90 Sr activity ratios compared to the global fallout level were observed in the total atmospheric deposition in this work. This might be related to the relative high annual precipitation rate in the study region, therefore the total atmospheric deposition signal are predominated by wet precipitation associated to lower 137 Cs/ 90 Sr activity ratios than global fallout, but are comparable to those measured in fresh waters in this work.
The peak 137 Cs/ 90 Sr activity ratio (1.14 ± 0.18) in the total atmospheric deposition in the second quarter of 2016 was ca. 2.5 times higher than the average values (0.46 ± 0.15) in 2015. This might suggest an additional radioactive input with higher 137 Cs/ 90 Sr activity ratio in the study region in 2016. As a potential consequence of this additional radioactive source, maximum 137 Cs/ 90 Sr activity ratios were observed in the source water (1.18 ± 0.05) in 2016, rice (1.14 ± 0.26) in 2017 and salsola (1.08 ± 0.09) in 2018, respectively. The occurrence of peak 137 Cs/ 90 Sr activity ratio following the sequence of atmosphere-water-biota pinpoints the additional radioactive source might be a direct atmospheric fallout either from local nuclear facilities (e.g., QNPP and other local NPPs) or global sources (e.g., Fukushima accident and others).
It is virtually impossible that the Fukushima fallout arrived to the study region in 2016, five years after the accident. At the first several days of the accidents, air transport in the mid-latitudes was dominated by prevailing westerly winds, which could circle around the globe in 2 weeks 42 . For example, several pulses of radioactive emission from Fukushima were observed in Northern Taiwan 14 days after the accident 43 . We suspect the potential additional radioactive input in 2016 is from a local source. It was reported that the two units in Fangjiashan NPP (FJSNPP) as an expansion of Phase I in QNPP, were put into operation in December 2014 and February 2015, respectively 10 . The commencement of the two units could potentially introduce increased release of 137 Cs and 90 Sr to some extent. However, this does not support the decrease of 137 Cs and 90 Sr concentrations and 137 Cs/ 90 Sr activity ratios in the total deposition for the following years. Therefore, further confirmation is needed due lack of operational and discharges data from QNPP in this work.

Conclusions
This study presents the first long-term systematic study of levels, variations and sources of 90 Sr and 137 Cs in environmental and food samples around QNPP in 2012-2019. The concentrations of 90 Sr and 137 Cs obtained in this work represent the background level, with all the values below the recommendations by WHO and Chinese national standard. Moreover, the peak concentrations of 90 Sr and 137 Cs appeared in 2016 were suspected to be related with an additional input from the local facility, but it requires further confirmation. This study indicate the high sensitivity of 90 Sr and 137 Cs, especially the 137 Cs/ 90 Sr activity ratio for detecting any radioactive release in the region. In the future, 90 Sr and 137 Cs monitoring is recommended as regional safeguard measure against accidental release from the local nuclear power plant.