1. Introduction

Tritium and ¹⁴C are important radionuclides from a nuclear regulatory perspective in order to protect man and environment, since these radionuclides generally are the major contributors to the activity released to the environment from nuclear facilities. Especially, ¹⁴C is one of the dominating radionuclides released from nuclear facilities concerning dose to the public [1]. The methods for release limitation are difficult especially for tritium, and monitoring reports indicates large uncertainties in analyses due to the unreliable sampling and analytical methods.

Knowledge on source terms, analytical methods for monitoring their level in nuclear facilities and environment and method for release limitation of theses radionuclides is of vital importance for a reliable analysis, review and development of regulations.

For a nuclear power plant, the thermal power affects the amount of ¹⁴C and tritium produced and released. In addition to normal operation of nuclear power plants, ¹⁴C and tritium are also important radionuclides during decommissioning, handling, and processing of radioactive waste.

This paper aim to summarize and review the sources of tritium and ¹⁴C in the environment and the methods for sampling and determination of tritium and ¹⁴C species in nuclear facilities and environment, meanwhile to discuss the main challenges and forthcoming development in these areas.

2. Sources and ¹⁴C and tritium in the environment

Both tritium and ¹⁴C are nature occurred radionuclides. However, human nuclear activities since 1945 have released large amount of these radionuclides to the environment, causing significantly increased levels in the environment. The production of tritium and ¹⁴C and their pathway to the environment are presented below.

2.1 Production of ¹⁴C and its pathways to the environment

2.1.1 Production of ¹⁴C in the nature and nuclear weapons testing

In nature, ¹⁴C is mainly produced in the upper atmosphere by cosmic ray reaction with nitrogen through ¹⁴N(n, p) ¹⁴C. It has been estimated that the annual production rate of ¹⁴C by this process is 1.4×10⁶ GBq, and the total inventory of ¹⁴C in the atmosphere was estimated to be 1.4×10⁸ GBq, but much large in the deep ocean of 1.0×10¹⁰ GBq [2].

Large amount of ¹⁴C has been produced in the atmospheric nuclear weapons tests in 1945-1980, mainly by the reactions of neutrons produced in the nuclear weapons testing with the nitrogen in the atmosphere through ¹⁴N(n, p) ¹⁴C. In addition, a small fraction of ¹⁴C was also produced by ternary fission reaction of uranium and plutonium during weapons testing. It has been estimated that a total of 2.2×10⁸ GBq ¹⁴C was released to the atmosphere by nuclear weapon testing, causing a significantly increased concentration of ¹⁴C in the atmosphere in 1950s-1960s (Fig. 1), when a large number of atmospheric nuclear weapons testing was conducted [2].

2.1.2 Production and pathways of ¹⁴C in nuclear reactors

During the operation of a nuclear reactor, ¹⁴C is mainly produced by neutron activation reactions of ¹⁷O, ¹⁴N, ¹³C, and to some extent 15N and ¹⁶O, in the nuclear fuel elements, construction materials, moderator and coolant, as well as by ternary fission reaction of uranium and plutonium in the nuclear fuel. Table 1 lists the major production reactions of ¹⁴C in the nuclear reactors.

Presently the major anthropogenic source of ¹⁴C is the nuclear industry including nuclear power reactors and nuclear spent fuel reprocessing plants through atmospheric releases and water discharges. In the nuclear power reactors, a large fraction of ¹⁴C is produced in the nuclear fuel due to high neutron flux and high concentration of oxygen. But, it mainly remaines in the fuel elements and cladding material as solid waste or partly released during reprocessing. ¹⁴C produced in the other construction materials is mainly treated as solid waste during decommissioning. ¹⁴C produced in the moderator and coolant is therefore the major source of ¹⁴C released to the environment during the operation. For the graphite moderator used in the gas cooled and graphite moderated reactors (GCRs) and the light water cooled and graphite moderated reactor (LWGRs/RBMK), the major fraction of ¹⁴C produced in the graphite is presents as solid waste, and a small fraction might be released in gaseous form to the environment. ¹⁴C produced in the coolant (light or heavy water) is mostly released to the atmosphere or discharged as liquid waste. Some of ¹⁴C in the coolant water can be trapped in ion exchange resin and treated as solid waste. It has been estimated that about 1.1×10¹⁵ Bq ¹⁴C is produced yearly in the nuclear power plants all over the world, in which about 1.1×10¹⁴ Bq·y^-114C is released to the atmosphere as gaseous form from all operating nuclear power plants, meanwhile about 3.7×10¹⁴ Bq·y^-114C in both gaseous and liquid forms is released from the reprocessing plants [5]. The production rate and pathway of ¹⁴C in the nuclear power plants vary with the type of power plants related to the enrichment of uranium in the fuel, the concentrations of oxygen and nitrogen in the fuel, structural materials, moderator and coolant.

Light water reactors (LWRs) are the most popular nuclear power reactor, accounting for two third of the nuclear power reactors, including pressurized water reactor (PWR) and boiling water reactor (BWR). In LWRs, ¹⁴C in the coolant water is mainly produced through 18O(n, α)¹⁴C reaction due to the lower nitrogen concentration. Based on an assumption of 0.1 ppm N₂ in coolant water in the BWR and 0.054 ppm in the coolant in PWR, it was estimated that more than 99% of ¹⁴C is produced from 18O in LWRs, and a production rate of 580 GBq/GW_e/y for 2500 MWth BWR (core specific calculation) and 350 GBq/GW_e/y for 2775 MWth PWR (core-specific calculation) [6, 7]. It should be mentioned that application of N₂ as annular gas in some BWRs significantly increases the production of ¹⁴C in the coolant of the BWR because of ¹⁴N(n, p)¹⁴C reaction. ¹⁴C produced in the fuel is mainly retained in the fuel elements; however, it might be leached out to the coolant through the cladding material such as stainless steel. In this case, it will become an important source of ¹⁴C released to the environment. ¹⁴C produced in the coolant and fuel in the nuclear power plants is distributed as gas and water-soluble form. Leakage of plant systems will eventually lead to release of ¹⁴C to the environment. Therefore, the pathway and distribution of ¹⁴C in the system is important for establishment of control measure.

In the PWRs, the reactor coolant also acts as moderator and is circulated in a closed loop (primary circuit). Because the coolant water in the close loop is separated from the water/steam in the secondary circuit, ¹⁴C produced in the coolant water is not significantly dispersed to the secondary circuit. PWRs operate under reducing condition in the presence of excessive hydrogen, causing the major part of ¹⁴C produced in the coolant to be in the form of organic compounds such as methane, acetate and formate (75-95%) with a small fraction of inorganic form as ¹⁴CO₂ and ¹⁴CO (5-25%) [2, 4]. Due to boric acid addition and buffering in the coolant, a small portion of produced ¹⁴CO₂, ¹⁴CO and organic ¹⁴C remains dissolved, while most of ¹⁴CO₂, ¹⁴CO and ¹⁴C-methane leaks to the air space in the reactor, which are released from the coolant system at different steps. Most of the gaseous ¹⁴C is released by venting of the volume control tank (air ejector) to the atmosphere. An ion exchange system is connected to the primary circuit to purify the coolant by removal of most of corrosion products. A fraction of dissolved ¹⁴C produced in the coolant water is also removed in the ion exchange resin, mainly the organic compounds, supposed to be acetate, formate, and carbonate [4]. It was reported that a majority of the organic ¹⁴C compounds in the reactor coolant (60%) is gaseous compounds (i.e. methane), and the remaining 40% of organic ¹⁴C is dissolved non-volatile organic compounds (e.g. formate and acetate). It should be noted that ion exchange resin can only remove small fraction of dissolved ¹⁴C in the coolant (6-10%), therefore, some extent of ¹⁴C is discharged to the environment in liquid form. The majority of ¹⁴C release is the gasses from the primary off-gas treatment system (volume control tank), accounting for about 70% of the total release, other pathways include steam generator blowdown tank vent exhaust, fuel handling building ventilation exhaust, turbine gland seal condenser exhaust, etc. A normalized ¹⁴C release from PWRs was estimated to be 130 GBq/GW(e)/y [2].

In boiling water reactors (BWRs), the heat generated in the reactor core is transferred to the coolant. Due to no pressure, the coolant is boiled to form a mixture of water and steam. The created steam is directed to the turbines, and condensed in the condenser. After purification in the ion exchange column, the water is recycled back to the reactor vessel. Therefore, ¹⁴C and other radionuclides produced in the coolant reach to the turbines. Meanwhile, compounds of corrosion products are concentrated in the coolant water and deposited onto the surface of fuel elements (crud formation), causing a relative high production of ¹⁴C in the coolant of the BWRs. The ¹⁴C produced in the coolant is also the major source of ¹⁴C released from BWRs. Unlike PWRs, the single circuit in the BWRs generates a continuous releases of ¹⁴C through stack. Due to non-reductive condition in the coolant of BWRs, ¹⁴C produced in the coolant is mainly in ¹⁴CO₂ form, with a small fraction in organic hydrocarbon form (5-20%). Therefore, the ¹⁴C released from BWRs is mainly in ¹⁴CO₂ form. A small fraction of ¹⁴C dissolved in coolant (HCO₃ -) can be removed in the ion exchange purification system. However, it has been reported that only a very small fraction of ¹⁴C in the coolant of BWR (<1%) is removed in the ion exchange resin [7]. Therefore, some portion of ¹⁴C is also discharged in liquid form. Besides the major release pathways of gaseous ¹⁴C from BWR stacks in the form of CO₂, other systems including the turbine gland seal condenser exhaust, reactor building purge exhaust, and turbine building ventilation system exhaust also release small portion of ¹⁴C produced in coolant. The typical ¹⁴C release rate from a BWR was estimated to be about 260 GBq/y, which is higher than that in the PWRs [2]. ¹⁴C releases to the environment from the nuclear power production in LWRs are estimated to be about 50% of the total production in BWR and 30% in PWR, the remaining part remains in the nuclear fuel and other solid waste.

In heavy water reactors (HWRs), heavy water (D₂O) is used as coolant and/or moderator, the CANDU reactor is the commercial HWR .There are also many research reactors using heavy water as coolant and graphite as moderator for obtaining high quality thermal neutron spectra for experiment. In CANDU reactors, nuclear fuel elements in fuel channels are inserted in the reactor tank filled with heavy water moderator, the fuel is cooled by heavy water (coolant) pumped through the closed pressure tube (primary circuit) to transfer heat to steam generator where steam is produced in the secondary circuit containing light water. The annulus between each pressure tube and its outer tube is filled with CO₂ gas to insulate the cool moderator from the heat transfer system [4]. ¹⁴C in HWR is mainly produced in the moderator, coolant system, fuel elements and annulus gas system. Due to the enriched ¹⁷O in the heavy water, the production of ¹⁴C in HWR is normally higher than in LWRs. Most of ¹⁴C is trapped in the ion exchange resin in the heavy water purification system, causing a high ¹⁴C concentration of 1.7-7.9 TBq·m-3 in the ion exchange in the moderator purification system. The release of gaseous ¹⁴C is mainly from moderator system with a small fraction from the annulus system using CO₂. However, when N₂ is used as annulus gas, higher production rate and release of ¹⁴C is expected. The airborne ¹⁴C releases from two Canadian CANDU (600 MW(e)) reactors were reported to be 121-720 GBq/y in Point Lepreau and 277-2920 GBq/y from Gentilly 2 nuclear power reactors [8]. In the CANDU reactor, most of the airborne ¹⁴C is released in ¹⁴CO₂ form (65.5-97.5%), a relatively small fraction in the form of organic hydrocarbon (2.5-34.4%), and very small fraction in ¹⁴CO form (0.01-3.7%) [2]. Table 2 presents the forms of airborne ¹⁴C released from different type of reactors.

In Magnox, AGRs (gas cooled graphite moderated reactors) and RBMK reactors, graphite is used as moderator. In Magnox and AGRs, CO₂ gas is used as coolant, whereas light water is used as coolant in RBMK reactors. In RBMK, the reactor space is filled with helium-nitrogen mixture gases to prevent graphite oxidation.

In the graphite reactors, ¹⁴C is mainly produced in the graphite moderator through ¹³C(n, γ)¹⁴C and ¹⁴N(n, p)¹⁴C reaction. However, most of the produced ¹⁴C remains in the graphite during the operation of the reactor, and is not released to the atmosphere. During operation of AGR, some ¹⁴C produced in the graphite might be released as CO₂ coolant gas because of corrosion of graphite and isotope exchange between CO₂ coolant and graphite, which increases the ¹⁴C concentration in the coolant gas. Meanwhile nitrogen impurity in the gas coolant is another source of ¹⁴C in the coolant, the concentration of nitrogen in the CO₂ coolant gas is a key issue for the ¹⁴C content in the coolant gas. A large portion of gaseous ¹⁴C releases from gas cooled reactors come from the purification of the CO₂ circuits by leakage or during periodic routine purge. Of the totally produced ¹⁴C in Magnox reactors and AGRs, only 3% and 6% is released, respectively. Gaseous ¹⁴C released from a Magnox type reactor in France (Chinon 2 reactor, 200 MW(e)) was reported to be 370 GBq·y^-1. In RBMK reactors, most of the released gaseous ¹⁴C is produced by the ¹⁴N(n, p)¹⁴C reaction, caused by the large amounts of nitrogen in the graphite block. It is assumed that the major part of ¹⁴C released from RBMKs is in ¹⁴CO₂ form. It is estimated that in total 1000±300 GBq/GW(e)/y ¹⁴C is released from RBMK reactor.

Table 3 summarizes the ¹⁴C release from different types of nuclear power reactors, it can be seen that HWR has the highest gaseous release of ¹⁴C, and only a very small fraction of ¹⁴C is discharged in liquid.

2.2 Production of tritium and its pathways to the environment

2.2.1 Production of tritium in the nature and nuclear weapons testing

In nature, tritium (³H) is mainly produced by cosmic ray reaction of nitrogen (¹⁴N(n, ³H)¹²C) and oxygen in the upper atmosphere, the produced tritium is converted to water (HTO) and reaches to the earth’s surface through precipitation. An estimated annual production rate of tritium is 1.48×10⁸ GBq. With the radioactive decay of tritium, a steady state inventory of tritium in the earth surface is 2.59×109 GBq [2], causing a concentration of tritium in precipitation of 0.6-1.2 Bq·L^-1 before nuclear era. A large amount of tritium was produced and released to the environment during atmospheric nuclear weapons testing in 1945-1980, it is estimated that 2.96×10¹¹ GBq of tritium was released to the environment by this way, causing the tritium concentration significantly increased on the earth. This tritium is mainly produced by neutron activation reaction of Li through ⁶Li(n, α)³H and 7Li(n, nα)³H) in hydrogen bombs. The contribution from ternary fission reactions of ²³⁵U and 239Pu is minor because of very small fission yield of tritium. The reaction of neutrons with nitrogen and deuterium (¹⁴N(n, ³H)¹²C, ²H(n, γ)³H) in the atmosphere also contribute the tritium during explosions . Underground nuclear weapons tests seem to add little tritium to the atmosphere. Due to the relative short half-life of tritium (12.32 y), the last fraction of the tritium released during nuclear weapon tests has been decayed, and the present tritium concentration in the precipitation without direct contamination has also ready reach to about 1 Bq·L^-1. Fig. 2 shows the variation of tritium in the atmosphere from 1950’s to 2016 in Japan.

2.2.2 Production and pathways of tritium in the nuclear reactors

In nuclear reactors, tritium is mainly produced by the following neutron activation reactions:

2 H ( n, γ ) 3 H σ th = 5.2 × 10 − 4 b 3 He ( n, p ) 3 H σ th = 5330 b 6 Li ( n, a ) 3 H σ th = 940 b 10 B ( n, 2α ) 3 H σ th = 3835 b

Tritium is therefore mainly formed in the following components in the nuclear reactors:

Among these components, tritium produced in fuel, boron control, graphite and reactor core materials is mainly retained inside the fuel elements or materials and not released to the environment until reprocessing or decommissioning/ melting. Small portion of tritium produced in the fuel might diffuse through the cladding material, such as stainless steel into the coolant. Transfer of tritium from the graphite moderator to the coolant may occur because of corrosion of the graphite.

Tritium produced in the coolant is partly or entirely releases in the discharge streams to air or discharged water depending on the practical management of the gas releases and waste water in the nuclear power plants. The tritium released to the air might be in tritiated water (HTO or 1H³HO) vapor or hydrogen gas (HT or 1H³H), and tritium discharges in liquid are mainly in water form (HTO). The typical discharge rates of tritium from various types of reactors are presented in Table 4. The production and release pathways of tritium in major types of reactor are discussed below.

In light water reactors (LWRs), besides the production of tritium in the fuel by ternary fission, which is remained in the fuel elements until reprocessing, most of tritium in LWRs is produced in the coolant water from boron and lithium. In PWRs, the production of tritium in the coolant is mainly from boron because of addition of boric acid for reactivity control. In BWRs, boron control rods are an important source of tritium. In LWRs, tritium is often discharged to the environment without additional trapping due to less production rate. Table 5 shows the tritium releases per energy production from PWRs and BWRs in 1980-2002. The different methods for reactivity control causes the higher tritium releases from PWRs compared to the BWRs.

Compared to light water reactor, large amount of tritium is produced in heavy water coolant and moderator in HWRs through the ²H(n, γ)³H reaction, with an estimated production rate of 8.9×10⁷ GBq/GW(e)/y [2]. Because of the high thermal neutron flux in the reactor core compared to the fuel channels, the fraction of tritium formed in the heavy water coolant is smaller than that in the moderator. The estimated concentration of tritium in the moderator is 3640 GBq/kg after 40 years operation, while only 81 GBq/ kg tritium exists in the coolant for the same operation time.

The major part of the tritium produced in the heavy water moderator and coolant remains in the heavy water, but a minor part of heavy water might be released from the system to the reactor building or escape during detritiation process of heavy water. The HWRs are the reactor type with the highest releases of tritium among all types of reactors (Table 5) due to high production of tritium in the heavy water moderator and coolant.

In graphite reactors, tritium is mainly produced in the graphite from lithium impurity through ⁶Li(n, α)³H because of high cross section of this reaction (940b). Tritium produced in the graphite might be released to the atmosphere by corrosion of the graphite. In addition, tritium might also be produced in the water vapor present in the reactor core. The vapor is transferred from the reactor core by the coolant gas (CO₂) and finds its way to the liquid discharges.

In RBMK, the reactor space is filled with helium-nitrogen mixture gases to prevent graphite oxidation. Because of the very high cross section of 5330b for reaction ³He(n, p)³H, a high amount of tritium in the RBMK is produced, causing a high release of tritium from RBMK reactors.

Table 5 and 6 summarized the releases of tritium and ¹⁴C from different types of reactors, and the estimated radiation doses. It can be seen that radiation dose from ¹⁴C is much higher than tritium in most types of nuclear power reactors.

2.3 Releases of tritium and ¹⁴C from nuclear fuel reprocessing plants

Tritium and ¹⁴C produced in the nuclear fuel are removed during the reprocessing. All ¹⁴C and tritium in the fuel are released during the shearing and dissolution step, while most of tritium and ¹⁴C in the cladding materials still remain in the undissolved cladding. Around half of ¹⁴C releases as off gas during dissolution of the fuel, most of it is trapped in scrubbing solution, and small portion is released to the atmosphere. About half of tritium enters into the dissolution of the fuel, and is separated as liquid waste. The distribution of ¹⁴C and tritium in the reprocessing plants is shown in the Fig. 3.

2.4 Releases of tritium and ¹⁴C from nuclear accidents and other nuclear facilities

Up to now, two major nuclear accidents in Chernobyl in 1986 and in Fukushima in 2011 have releases large amounts of radioactive substance to the environment, including tritium and ¹⁴C. The damaged nuclear reactor in Chernobyl accident is a RBMK type reactor, due to the high production of tritium and ¹⁴C in this type of reactor and burning of graphite during accident, large fraction of tritium and ¹⁴C was released to the atmosphere. It was reported that large fraction of ¹⁴C released during the accident were in gaseous form (CO and CO₂) due to the burning of graphite at high temperatures (up to 3000℃), which was dispersed to a large area in the world. Total ¹⁴C released to the environment was estimated to be 44 TBq (about 24% of the total inventory of reactor core) [10]. which is relative small compared to the nuclear weapons tests, but comparable to an annual releases from all PWRs in the world. Significantly increased ¹⁴C level in the air was observed in Finland after the Chernobyl accident [11]. In addition, aerosol ¹⁴C (mainly in fine graphite particles) was also released, which mainly deposited in local region of Chernobyl, which had been changed into organic carbon by assimilation via microorganisms in the past decades [12]. There is no report on the tritium releases from the Chernobyl accident. Since high production of tritium in the graphite and He coolant in the RBMK type reactor, it is expected that a relative large amount of tritium was released to the environment. Based on the fission yield of ¹³⁷Cs and ³H and the released ¹³⁷Cs, a release of 37 TBq tritium from Chernobyl accident was estimated [13]. This might be a high underestimation, because tritium in RBMK is mainly produced through neutron activation of lithium impurity in graphite and He in cooling tube, which was released through explosion and burning of graphite. A low (20 Bq·L^-1), but about 7 times increased tritium level was observed after the Chernobyl accident in the precipitation in Finland which received high Chernobyl fallout [11], but no remarkable increased tritium level in precipitation was observed after the Chernobyl accident in Germany, Ireland and Japan [13]. However, it might be still very small compared to the releases from the nuclear weapons tests, and even to the releases from HWRs.

Fukushima accident in 2011 has released large amount volatile radionuclides (e.g. ¹³¹I, ¹³⁷Cs, ¹³⁴Cs, radioactive Xe) to the environment; tritium and ¹⁴C should be also released. Since the damaged reactors in Fukushima Daiichi NPP are boiling water reactors, the production and inventory of tritium and ¹⁴C in these reactors are much low compared to the RBMK reactor in Chernobyl accident, the released tritium and ¹⁴C should be much less compared to that from Chernobyl accident. No report on the estimated amount of tritium and ¹⁴C releases from the Fukushima accident is available. The measurement of ¹⁴C in tree rings showed a slightly increased ¹⁴C level only in the local region (<15 km from the damaged reactor in Fukushima Daiichi NPP) [14]. Increased tritium concentration of up to 6 Bq·L^-1 was measured in the precipitation in Japan within a few weeks after the Fukushima accident. However, this is very quickly decreased to background level after 2 month [15].

Tritium was widely applied in medical research by labelling tritium to different organic compound, it might also release some amount of tritium to the air in forms of tritium gas (HT), organic gas tritium, and to rivers and marine system in liquid form, but it is only a small contribution compared to the total tritium on the earth [45]. The potential huge source of tritium will be production and application of tritium in fusion device (reactors) in the near future. In the fusion reactor, up to a few tens of kilogram of tritium (and deuterium) will be used. The release of a small fraction of these tritium during production, storage and operation of the power reactor might cause significant increased concentration of tritium in the environment (1g = 360 TBq tritium), which might be a big challenge in the limitation of its releases in the future.

In summary, tritium and ¹⁴C are naturally produced in the atmosphere by cosmic ray. In the human nuclear activities, tritium is mainly produced through neutron activation reactions of ³He, ⁶Li, ¹⁰B and ²H in the nuclear reactors, and ¹⁴C is mainly produced from neutron activation of ¹⁴N, ¹⁷O and ¹³C. Tritium and ¹⁴C released to the environment from the nuclear power reactors are mainly produced in the reactor coolants, and gases. The nuclear fuel and cladding materials contains large fraction of tritium and ¹⁴C produced in the reactors due to high neutron flux, but they mainly remain inside these materials. The nuclear weapons tests in the 1950’s and 1960’s have been the major source of tritium and ¹⁴C in the environment, which have declined significantly and is close to the natural level in normal environment until present time. The releases from the nuclear power plants are now become one of the major sources of tritium and ¹⁴C in the environment, especially in the vicinity of the NPPs. Both tritium and ¹⁴C are released from nuclear facilities to the environment in gaseous releases and liquid discharges. The majority of tritium is discharged to the environment through liquid discharges, while gaseous ¹⁴C is the major way to be released to the environment from the nuclear power plants. Due to the different production ways, HWRs release much higher tritium per energy production, especially in liquid discharges, to the environment compared to other types of nuclear power reactors. While, light water reactors (PWRs and BWRs) release more gaseous ¹⁴C to the environment per power production compared to other types of power reactors.

3. Approaches and setup for monitoring ¹⁴C and tritium in nuclear facilities and environment

Monitoring of ¹⁴C and tritium released from the nuclear facilities is important to the managers and operators of the nuclear facilities, as well as to the regulators and public to demonstrate that the radioactive emission is within acceptable limits. Tritium and ¹⁴C are released to the environment in both gaseous and liquid discharges and in different forms. Reliable methods for sampling/collection of tritium and ¹⁴C are critical for accurate monitoring the releases and environmental level of these two radionuclides.

There are normally two systems for monitoring ¹⁴C and tritium releases from the nuclear facilities, discharge and environmental monitoring systems. The discharge monitoring systems are used to directly measure tritium and ¹⁴C in airborne and liquid discharges before release to the environment; environmental monitoring systems are intended to measure the environmental level of these two radionuclides by analysing environmental samples, such as atmosphere, surface water, and environmental samples surround the nuclear facilities; the two systems are complementary. Environmental monitoring provides the accurate assessment of radiation exposure to the public, discharge monitoring provides direct data on the total amount of radionuclides and their species released from the nuclear facilities. Therefore, both systems are often operated in parallel.

3.1 Monitoring of ¹⁴C in air, concerning collection efficiency of different forms of ¹⁴C

Gaseous ¹⁴C is normally released from the nuclear facilities through stacks in different species, ¹⁴CO₂, ¹⁴CO, and hydrocarbon (C_nH_m). Table 2 presents the relative percentages of different species of ¹⁴C released from different nuclear power reactors. In PWRs, most of ¹⁴C was released as hydrocarbon, mainly CH₄ and C₂H₆, while in other types of power reactor (BWRs, HWRs, graphite reactor), ¹⁴C is mainly released as ¹⁴CO₂. Releases in ¹⁴CO form accounts for very small fraction for all types of power reactors. The species of ¹⁴C in the atmosphere are also mainly in three forms ¹⁴CO₂, ¹⁴CO and hydrocarbon. However, the species of hydrocarbon in the atmosphere are more complicated than that in the stack of nuclear facilities. The percentages of different forms of ¹⁴C in the atmosphere depends on its sources and the environmental conditions. Besides the direct releases of ¹⁴C from the nuclear facilities, ¹⁴C in the atmosphere also include the recycling of ¹⁴C in the ecosystem, i.e. re-emission of ¹⁴C accumulated in the environmental media including vegetation as well as soil and water. Meanwhile, constant production of natural ¹⁴C in the upper atmosphere through cosmic rays is also a major source to the atmosphere in the present.

The methods for collection of ¹⁴C from the stack of the nuclear facilities and from the atmosphere therefore depend on the requirements and purposes. There are three common used methods: (1) only collection of ¹⁴CO₂; (2) collection of all forms of ¹⁴C in one sample; (3) collection of different species of ¹⁴C.

3.1.1 Collection of ¹⁴CO₂ from the stacks and atmosphere

¹⁴CO₂ in air and stack of a nuclear facility can be collected by two types of methods, active and passive collection. Passive sampler is just simply exposure of a collector to air, ¹⁴CO₂ is absorbed in the regents in the collectors, NaOH pellet, NaOH solution and filter paper impregnated with NaOH as absorbent have been used for passive collection of ¹⁴CO₂ in the air. 4 mol·l^-1 NaOH solution putting into plastic tray, covered with nylon mesh to prevent from drop of suspending particles in the air into the NaOH solution, and filter paper wetted with 0.8 mol·l^-1 NaOH solution have been used at Chalk River Laboratory, Canada for collecting ¹⁴CO₂ from air. This system is suitable for short term sampling because the NaOH solution will be saturated if it is exposed for a long time [27, 61]. Compared to NaOH impregnated filter paper, NaOH pellets as absorbent is easy to handle and transport and have been applied for collect ¹⁴CO₂ from stack of nuclear reactors [16, 17]. However, it should be mentioned that the passive method only provides a relative level of ¹⁴C, and cannot give a precise concentration of ¹⁴CO₂ in the air or stack gas. Some calibrations are needed to convert the measured ¹⁴C value into the concentration of ¹⁴CO₂ in the air or stack gas, but normally with a relative high uncertainty. It is therefore not often used.

The active sampling is the most often used method for collecting ¹⁴CO₂ from the stack of nuclear facilities and from air. In this method, gas/air is pumped from the sampling site through a bubbler filled with absorbent or tube filled with molecular sieve. The most often applied absorbents are NaOH solution (0.8-4.0 mol·l^-1), NaOH pellet and organic amine (such as CarboSorb provided by PerkinElmer). The sampling efficiency depends on many parameters, such as the design of the bubbler, flow rate, absorbent components and concentration, as well as volume of the absorbent solution and gas/air collected. The flow rate should be kept low to allow a sufficient contact time of the gas/air with the absorbent solution. Bubbler with micro-pores sintered glass on the bottom is often applied to distribute the input gas/air into small bubbles for a better contact of the gas/air with absorbent. Two sequentially connected bubblers are often employed to ensure a quantitative collection of ¹⁴CO₂. Compared with NaOH solution, CarboSorb solution has better compatibility with organic scintillation cocktails, and therefore low quenches effect in liquid scintillation counting (LSC) measurement of ¹⁴C. Molecular sieve tube has been reported to be an effective approach for collecting CO₂ from air and stack gas with a good collecting efficiency, it is also easy to operate in the field and transport afterwards. In addition, it has a high capacity and can be operate for a long time, therefore suitable for sampling a large volume of air samples [18]. The main drawback of this technique is that ¹⁴CO₂ adsorbed on the molecular sieve cannot be directly measured; it has to be released from the molecular sieve and trap the ¹⁴CO₂ into CarboSorb or NaOH solution again for measurement using LSC, or converted to other forms such as graphite for accelerator mass spectrometry (AMS) measurement of ¹⁴C. For collecting ¹⁴CO₂ in stack gas in nuclear facilities, NaOH solution and CarboSorb absorbers are the effective and the most often used methods, whereas, molecular sieve method is an effective method for collecting ¹⁴CO₂ from large volume of air sample.

3.1.2 Collection of all species of gaseous ¹⁴C from the stacks and atmosphere

The major fraction of gaseous ¹⁴C released from PWRs is in hydrocarbon form, while the fraction of gaseous ¹⁴C from other types of nuclear power reactors released as hydrocarbon is in 2-35%. Measurement of only ¹⁴CO₂ will underestimate ¹⁴C level in the stack gas and air, although the radiation risk of ¹⁴C in hydrocarbon form is smaller than that of ¹⁴CO₂, due to less absorption of hydrocarbon by vegetation and intake by humans through inhalation. Therefore, all species of ¹⁴C are often collected from stack gas in many nuclear power plants, especially PWRs. The major method for collection of all species of ¹⁴C is based on the conversion of all ¹⁴C to ¹⁴CO₂, and then trapping the formed ¹⁴CO₂ using the same methods as for collecting ¹⁴CO₂ from air. Hydrocarbon is normally converted to CO₂ by oxidation combustion with O₂ flowing and catalyst, in this case ¹⁴CO is also converted to ¹⁴CO₂. In practice, this is implemented by pumping stack gas or air through a tube which is filled with catalyst and heated in an oven to 400-800℃, where all species of hydrocarbon and ¹⁴CO are converted to ¹⁴CO₂. The off gas from the tube is then trapped in a bubbler filled with NaOH or amine. The most often used catalyst is Pt/ Pd material, it has been reported that more than 99.7% of hydrocarbon can be converted to ¹⁴CO₂ using Pd as catalyst and heated at 600℃ [19]. CuO has also been used as catalyst/oxidant to convert hydrocarbon and CO to CO₂ under heating at 600℃, and the formed CO₂ is then trapped in monoethnolamine. The oxidation and CO₂ collection efficiency by this method are >99.3% and >97.5%, respectively [20]. In this method, O₂ flow is not necessary, since CuO provides oxygen for the oxidation. Some commercial instruments for collecting all ¹⁴C (e.g. V3H14C (VF, Czech Republic) have been applied in some nuclear facilities [17, 21, 22, 23, 24, 25, 26, 27, 62].

3.1.3 Collection of different species of gaseous ¹⁴C from the stacks and atmosphere

For better understanding the components of ¹⁴C species released from the nuclear facilities and in the atmosphere and estimating the impact of ¹⁴C to the ecosystem and public, different species of ¹⁴C are collected and measured. In many systems for monitoring ¹⁴C from stack of nuclear facilities, gas or air is passed through two devices separately, one is used to trap ¹⁴CO₂, and the other for all species of ¹⁴C. The difference between the values obtained by the two devices is the content of ¹⁴C in hydrocarbon and CO forms. The results obtained by this method are general reliable if the ¹⁴C fraction of hydrocarbon and CO forms is not very small compared to ¹⁴CO₂. However, if the sum of the hydrocarbon and CO fractions are less than 5% of the total ¹⁴C, the uncertainty of the analytical results for these two species will be relative big.

A sequential collection system has also been proposed and applied for collecting different species of ¹⁴C from the stack of nuclear facilities or atmosphere. In this case, gas or air first passes through a ¹⁴CO₂ bubbler where CO₂ is completely trapped, the off gas from the first bubbler passes through a tube filled with catalyst (e.g. Hopcalite, Cab-OSil, Packard Instruments, The Netherlands) at ambient temperature where ¹⁴CO is converted to ¹⁴CO₂, which is then trapped in the second CO₂ bubbler; afterwards the off gas from the second bubbler passes through anther tube filled with Pd/Pt catalyst which is heated in an oven at 400-800℃ to convert hydrocarbon to ¹⁴CO₂, which is trapped in the third CO₂ bubbler. By this setup, three species of ¹⁴C can be sequentially collected. In this system, it is important that the trapping efficiency for ¹⁴CO₂ has to be sufficiently high to avoid any cross over from the front one to the following one, especially for ¹⁴CO fraction, which normally accounts for a small percentage (<1%). By optimization of bubbler design and adjustment of flow rate, a trapping efficiency of >99.9% have been obtained for ¹⁴CO₂ [19].

Because the ¹⁴CO species accounts for a very small fraction of ¹⁴C, it is often collected with hydrocarbon. In this case, the gas or air first passes through a ¹⁴CO₂ bubbler where CO₂ is completely trapped, the off gas from the first bubbler passes through a tube filled with Pd/Pt catalyst and heated in an oven at 400-800℃ to convert both hydrocarbon and ¹⁴CO to ¹⁴CO₂, which is trapped in the second CO₂ bubbler. By this setup, two fractions of ¹⁴C, ¹⁴CO₂ and hydrocarbon plus ¹⁴CO are collected. Fig. 4 shows a commercial available system (Hague 700 ¹⁴C sampler, SDEC, France) for sequentially collecting ¹⁴CO₂ and ¹⁴CO plus hydrocarbon- 14 .

For low-level air samples, a sequential collection of ¹⁴CO₂ and ¹⁴C_nH_m is often implemented by passing through the air to a collection column and a furnace followed by the second column. The first column is filled with molecular sieve for trapping CO₂, the off gas from the first column passes through a tube filled with platinum catalyst, which is heated in a furnace to convert hydrocarbon and possible CO to CO₂, which is then trapped in the second molecular sieve column. The two columns are transported to laboratory for further separation of trapped CO₂ by heating and re-trapping the released ¹⁴CO₂ in NaOH or CarboSorb solution for LSC measurement of ¹⁴C.

3.2 Monitoring tritium in air concerning separation of different species of tritium

Many types of tritium monitors have been developed to measure gaseous tritium in stack exhaust gas from the nuclear facilities. Ionization chambers (e.g. Tyne Room Tritium- in-air monitor, Tyne Engineering Inc. Canada) is often used for such monitors, proportional counter (e.g. LB 110 tritium monitor, Berthold Technologies GmbH & co. KG, Germany) and solid scintillation detector are also employed in this types of monitors. However, these systems are not sensitive, with a detection limit of 0.1-10 kBq·m-3. In addition, such monitors cannot well separate other radionuclides such as ¹⁴C, radioactive noble gases, etc. therefore not very accurate when the concentrations of the other gaseous radionuclides are high. These monitors are not suitable to measure low-level tritium in the gas discharge from nuclear facilities and environmental air. Collection of tritium from large volumes of discharge gas or air has to be implemented for determination of low-level tritium and its species.

In the discharge gas from nuclear facilities and environmental air, tritium occurs as tritiated water vapor (HTO, DTO, T₂O), tritium gas (HT, DT, T₂) and hydrocarbon. The gaseous tritium released from most of the nuclear facilities is mainly in tritiated water form (HTO), especially in HWRs. Tritium in tritium gas (HT, T2) and hydrocarbon species has a slow exchange with hydrogen in the vegetation, and are accumulated in vegetation and humans to a less degree. These species of tritium are consequentially less important than tritiated water (HTO) in view of radiation protection. Therefore, tritiated water (HTO) is more often measured in the reactor building and discharge gas in the stack of the nuclear facilities. For this purpose, the discharged gas from the stack or air in the reactor building is passed through a condenser or a dryer such as silica gel, tritiated water vapor is collected in the condenser or adsorbed in the silica gel. The water adsorbed in the silica gel is further converted to liquid water by heating it and condensing the released water vapor. The separated tritiated water is directly used for LSC measurement of tritium.

The total tritium in the discharge gas is also often collected and monitored. In this case, the discharge gas or air first passes through a tube filled Pt catalyst, which is heated in a furnace to 400-800℃ to convert tritium gas and hydrocarbon tritium to HTO. The tritium in the tritiated water form is then trapped by a condenser or dryer. Bubblers has also been used for trapping tritiated water vapor, an ethylene glycol and tritium free water mixture is better than only water as trapping solution in the bubbler to minimize evaporative losses in longer term sampling. Since the trapping efficiency of water bubbler is normally not very high, three sequential bubblers are used to obtain a better recovery of tritium. For air samples with high content of other radionuclides, a purification of the trap solution is needed. Distillation or purification through an ion exchange cartridge (mix bed ion exchange resins) is sufficient good to remove most of interfering radionuclides.

For determination of different species of tritium in the discharge gas or air, the discharge gas or air first passes through a condenser or dryer, the off gas passes through a tube filled with platinum catalyst at ambient temperature to convert HT to HTO, which is then trapped in the second condenser or dryer. The remaining hydrocarbon species in the gas is passed through a tube filled with Pt/Pd catalyst and heated to 400-800℃ in a furnace to convert the organic species of tritium (hydrocarbon) to HTO, which is then trapped in the third condenser or dryer. The collected three fractions of tritium are then measured using LSC. HT can also be converted to tritiated water by passing gas over a heated CuO catalyst [2]; in this case, no oxygen is needed. Since HT and organic gas tritium species normally accounts for a very small fraction of tritium in the discharge and atmosphere, they are often combined as one fraction, i.e. the first tube filled with catalyst and heated following one trapper to collect the converted HTO from both HT and organic gas tritium. A commercial tritium sampler (Fig. 5) is available for collection of two fractions of tritium: tritiated water and HT.

For analysis of air with low concentration of tritium, such as the environmental air, a large volume of air is needed. In this case, molecular sieve is often used to replace the condenser or dryer to collect tritiated water [28]. The collected three fractions of tritiated water in molecular sieve columns are transported to laboratory to separate the trapped tritiated water in each column by heating followed by condenser for tritium measurement [29].

For on-line measurement of different species of tritium in the stack discharge with high level of tritium concentration, two ionization chamber monitors can be used. In the sampling line, a desiccant cartridge is inserted between the two ionization chamber monitors for trapping the tritiated water from the gas stream. The second monitor measures the concentration of tritium gas and hydrocarbon tritium, and the first monitor measure all the species of gaseous tritium. The difference between two monitors gives the concentration of tritiated water [2].

The tritium samplers discussed above require the use of an air pump and flow meters to delivery certain volume of air/gas to the sampler in order to give concentrations of the total tritium or different species of tritium. A passive diffusion tritium sampler has also been developed and used for monitoring tritium level [30, 31]. This type of tritium sampler consists of a 20 ml LSC vial with a lid modified to contain a stainless steel insert with a diffusion orifice in the centre. HTO vapor in the air diffuses through the orifice and is trapped in the water in the vial. The amount of water vapor trapped into the vial is controlled by the diffusion rate of the orifice, exposure time and other parameters of the air components. The trapped tritiated water in the vial is directly measured using LSC after added scintillation cocktail. This method is less expensive and simple in operation, which is very suitable for field use for monitoring the outdoor air tritium level. The detection limit of this method is reported to be as low as 1 Bq·m-3 for over onemonth exposure [32, 33]. However, the results obtained are not quantitative since the level of tritiated water trapped is influenced by many parameters.

3.3 Combined system for monitoring both tritium and ¹⁴C in air

Gaseous tritium and ¹⁴C in air can be sampled simultaneously in a combined system. In such a system, air is pumped through sequentially connected collectors or tubes. The first collector is a condenser/dryer/bubbler filled with ethylene glycol and water mixture/or a molecular sieve for trapping tritiated water vapor, the second collector is a bubbler containing NaOH or CarboSorb solution or another molecular sieve to trap ¹⁴CO₂. The off gas then passes through a tube filled with platinum with oxygen gas flow and heated in a furnace, where HT is converted to tritiated water, hydrocarbons are converted to HTO and ¹⁴CO₂, and ¹⁴CO is converted to ¹⁴CO₂, the off gas passes through the following collectors, the third collector is a condenser/ dryer/bubbler filled with ethylene glycol and water mixture/ or a molecular sieve for trapping the formed tritiated water vapor, and the fourth collector is a bubbler contain NaOH or CarboSorb solution or another molecular sieve to trap the formed ¹⁴CO₂. The tritium collected in the first and third collector is used to measure tritiated water and organic tritium plus tritium gas after separation of the tritiated water from the collector. The trap solutions in the second and fourth collectors are used to measure ¹⁴CO₂ and hydrocarbon- 14 plus ¹⁴CO, respectively, after a further separation or purification. Some commercial systems are available for combined collection of different species of tritium and ¹⁴C (Fig. 6), which have been used in some nuclear facilities for monitoring gaseous tritium and ¹⁴C.

3.4 Monitoring of ¹⁴C in water, speciation analysis of ¹⁴C in water

Large amount of ¹⁴C is produced in the reactor water (coolant and moderator), which forms the major source of ¹⁴C in the discharge gas from the nuclear facilities. However, only a small fraction of the volatile species of ¹⁴C (¹⁴CO₂, ¹⁴CO and volatile hydrocarbon-¹⁴C) escapes from the reactor water and is released through the stacks. Other fractions of ¹⁴C is stable under reactor operation condition and remains in the reactor water. Some of the reactor water might be discharged into rivers or seas from the nuclear facilities, which cause an elevated concentration of ¹⁴C in the environmental water samples. In principle, the chemical species of ¹⁴C in the environmental water contaminated by the discharges from the nuclear facilities should be the same as its sources. However, some new species of ¹⁴C might be formed during natural processes involving biological activities, and the distribution of the ¹⁴C species in the environmental water might be changed. In general, ¹⁴C in the reactor water and environmental water normally exist as dissolved gaseous ¹⁴C (volatile hydrocarbons), inorganic carbon (mainly carbonate/biocarbonate) and water soluble organic compounds [2, 6].

Some methods have been reported for the determination of ¹⁴C and its species in the reactor water [34, 35, 36, 37, 38, 39]. In general, three types of methods were used to separate ¹⁴C and its species: combustion, wet oxidation using strong oxidative reagents (K₂S₂O₈ with AgNO₃ in acidic media or CrO₃ and H₂SO₄) and acid digestion (H₂SO₄). The combustion method is normally used to determine the total soluble ¹⁴C in the water after evaporating the water to dryness; acid digestion is used to separate ¹⁴C in carbonate/ bicarbonate form; while wet oxidation for decomposing organic compounds is used for separation of ¹⁴C in both organic and carbonate forms. The released ¹⁴C is adsorbed in NaOH or CarboSorb solution for LSC measurement. For separation of volatile species of ¹⁴C in water samples (e.g. CH₄, C₂H₆), a combustion method can be used by purging volatile species of ¹⁴C from the water, which pass through a tube filled with catalyst (Pt/CuO) and heated in a furnace at 400-800℃ to convert all volatile organic carbon (hydrocarbon) to ¹⁴CO₂, which is trapped in a NaOH bubbler [27]. For determination of all soluble ¹⁴C using combustion method, the water is first adjusted to pH 8-9 using Na₂CO₃ and/or NaOH followed by evaporation to dryness. The residue is then transferred to a boat that is placed in a tube furnace for combustion or combusted using sample oxidizer (Fig. 7). The water sample is first evaporated to a small volume (<5 ml), and then transferred to a boat to avoid that residues attach to the surface of the beaker, causing loss of some residue during transferring the entire residue to the boat. There are two types of commercial combustion systems available for separation of total ¹⁴C in the residue: Pyrolysis based tube furnace (Raddec International, Southampton, UK) [40] and Sample Oxidizer (PerkinElmer, USA) [41]. Sampler Oxidizer can rapidly separate ¹⁴C from residue in 2-3 min/sample compared to Pyrolysis of about 4 hours for each batch (4-6 samples). However, the tube furnace system (pyrolysis based system) can treat a large size sample up to 20 g, compared to only less than 1 g samples in Sample Oxidizer system. Schematic diagram of two systems (Fig. 7) shows the separation principle and structure of the system. In wet oxidation method for separation of organic ¹⁴C or all soluble ¹⁴C, some ¹⁴C might be released as ¹⁴CO or volatile organic species during the wet oxidation process. A combustion furnace with catalyst can be applied before trapping ¹⁴CO₂ in order to ensure the collection of all soluble ¹⁴C.

Two methods have been proposed for determination of different species of ¹⁴C. One method use two systems; one is used to separate the entire ¹⁴C employing wet oxidation at acidic media, and another one use acid digestion to separate carbonate. The difference between the ¹⁴C results from the two systems is the fraction of organic ¹⁴C [4]. If the fraction of organic ¹⁴C is relative small (<5% of the total ¹⁴C), the results of organic ¹⁴C from this method might be inaccuracte (e.g. with an uncertainty more than 100%). Another method is to sequentially separate each species of ¹⁴C and collecte them individually in an integrated system [6, 7].

Fig. 8 shows the schematic diagram of the system for sequential separation of ¹⁴C species in water samples [6]. The system consists of 6 parts: (1) the gas supply to purge gas (N₂) to the system; (2) Reaction vessel where the water sample is put in a flask which can be heated by a hot plate and stirred, a separatary funnel is connected to the sample flask to add acid to the water, and a cooling/condensor to convert the evaporated water vapor to liquid and return to the sample container funnel; (3) three bubblers, the bottle No.1 contains diluted H₂SO₄ for trapping tritiated water and other volatile radionuclides (e.g. radioactive isotopes of iodine, chlorine, technetium, etc.), bottle No. 2 and 3 are filled with NaOH solution for trapping ¹⁴CO₂, the bottle No. 3 is a backup of the bottle No. 2 to ensure completely trapping of ¹⁴CO₂; (4) catalytic furnace, a tube filled with catlyst/oxidant (CuO) which is heated in a furnace, ¹⁴CO and organic carbon compounds are converted to ¹⁴CO₂ in this part; (5) two bubblers for trapping ¹⁴CO₂, among them the bottle No. 5 is a backup of the bottle No. 4 to ensure complete trapping of ¹⁴CO₂; (6) a vacumm pump to suck gas from the system.

The separation of ¹⁴C species in water is carried out in three steps: 1) for volatile species of ¹⁴C, the water sample is purged from the bottom of the sample container using a flow of N₂, the volatile carbon compounds are purged out from the water and transferred through the system, oxidized forms of carbon (CO₂) are absorbed in the bottle No. 2, while other forms of carbon compounds pass through the catalytic furnace where organic carbon is converted to CO₂, which is then trapped in the bottle No. 4, the solutions in bottle No. 2 and 4 are used for measurement of volatile species of ¹⁴C, and solution in all bubblers are replaced with new trap solution before the second step; 2) acid stripping for separation of carbonate species, the system is first evacuated to 0.2 bar using the vacumm pump, and carrier gas (N₂) is then purged through the system, H₂SO₄ is added from the separatary funnel to the water sample, the acidified water is purged for 1 hour. CO₂ released from the sample is trapped in bottle No. 2 and 3, which is used to measure ¹⁴C in carbonate fraction. (3) wet oxidation for soluble organic species of ¹⁴C, after the previpus two steps, the gas flow from the sample flask is switched to the connection to the catalytic furnace, and 5% K₂S₂O₈ and 4% AgNO₃ solution are then added to the water sample flask, the water is then heated to 95℃ and stirred for 2 hous, gasses released from the sample as CO₂ and CO flow through the catlystic furnace to convert all species of ¹⁴C to ¹⁴CO₂, which is trapped in bottle No. 4 and 5, and used for measurement of soluble organic ¹⁴C.

This system can also be used for separation of ¹⁴C species in environmental water or wastewater discharged from the nuclear facilities.

3.5 Monitoring of tritium in water, concerning organic and inorganic tritium

In reactor water, liquid discharge and environmental water samples, tritium exists dominantly as tritiated water (HTO or T₂O). While, in the liquid discharge and the environment, a small fraction of tritium might occur in organic forms. The determination of tritium in water is straightforward and easy by LSC. However, because the decay energy of tritium (maximum energy of beta particle is 18.6 keV), quench effect has to be considered before measurement, such as peroxide, alkaline or acidic medium and high salt content. A specific quench correction can be used for the determination of the specific samples when the tritium concentration is sufficiently high, otherwise purification is normally needed. Distillation is the most often applied and simplest method for purification of water samples for tritium measurements. The water sample is first neutralized before distillation to remove acids and volatile compounds. For the water sample with relative high content of organic compounds, active charcoal is added to adsorb the soluble organic compounds, which is then removed by filtration, the filtrate is further purified by distillation. The tritium in the organic form, which is absorbed in active charcoal, can be further separated by oxidation combustion (Fig. 7) to convert the organic compounds to tritiated water for measurement. This method has also been used for speciation analysis of tritium and ¹⁴C in urine samples [42].

The tritium concentration in environmental water is normally too low to be directly measured by LSC, an enrichment system is normally used to enrich tritium from water samples. The enrichment system of water is based on the electrolysis of water, in which hydrogen water (H₂O) is electrolyzed to H₂ and O₂ gases before tritiated water (T₂O and HTO), and HTO is remained and enriched, which is used for measurement by LSC. By this way, an enrichment factor of 30-50 can be reached depending on the initial volume of the water and the final water volume [43, 44].

3.6 Combined system for monitoring both tritium and ¹⁴C in water

In the most cases, both tritium and ¹⁴C in the liquid discharge and environmental water are required to be monitored. The system used for ¹⁴C can also be used for tritium. For the measurement of total tritium and ¹⁴C, the water sample can be treated by wet oxidation; ¹⁴C in both inorganic and organic forms is converted to CO₂ to be trapped for ¹⁴C measurement, while tritium remaining in the liquid sample is further purified by distillation for measurement of tritium in the distillated water. Meanwhile the speciation analysis of tritium in the water can also be carried out by using the same systems as for speciation analysis of ¹⁴C in water. Tritium in tritiated water form is obtained by measurement of the distillated water from the original water sample. The total tritium is obtained by the further purification of the wet oxidation treated water samples by distillation. The difference between the total tritium and tritiated water fraction is the organic tritium. It should be mentioned that the organic tritium fraction in the liquid discharge and environmental water is normally too small to be significant in view of radiation protection.

3.7 Separation of tritium and ¹⁴C in solid waste and environmental samples

For the determination of tritium and ¹⁴C in solid samples such as ion exchange resin, water evaporates, concrete, graphite, metals, soil, sediment and vegetation, oxidation combustion is often used. In this technique, the samples are combusted at high temperature (850-1000℃) under O₂ flow and with Pt or CuO catalyst. Regardless of its species, all tritium in the sample is converted to water vapor, which is released from the sample and collected by a condenser or water/diluted acid trapping. Meanwhile, ¹⁴C in both inorganic and organic forms is converted to CO₂, which is released from the sample and trapped in an alkaline solution such as CarboSorb or NaOH. The separated tritium in water and ¹⁴C in trap solution are measured by LSC after mixing with scintillation cocktails, respectively. Pt or CuO catalyst is used to convert any CO to CO₂, since CO cannot be effectively trapped in alkaline solution.

Two combustion systems have been used for this purpose: Pyrolysis based tube furnace (Raddec International, Southampton, UK) [40] and Sample Oxidizer (PerkinElmer, USA) [41]. Sampler Oxidizer (Fig. 7) can rapidly separate tritium and ¹⁴C from solid samples in 2-3 min/ sample compared to Pyrolysis of about 4 hours for each batch (4-6 samples). However, the tube furnace system can treat larger samples, up to 20 g for concrete, soil and metal and, 5 g for biological samples, compared to only less than 1 g concrete, soil and metal, and 0.5 g biological samples in Sample Oxidizer system. Besides these samples, the Sample oxidizer has also been used to analyze oil, bitumen and plastic samples. The results showed that both inorganic and organic forms of tritium and ¹⁴C could be quantitatively separated from samples using these two systems.

In ion exchange resin, ¹⁴C exists in both inorganic (mainly carbonate) and organic forms (i.e. organic compounds). A sequentially extraction/digestion method similar to the method used for reactor water has been used to separate inorganic and organic ¹⁴C in ion exchange resin and processing water from nuclear reactor [6, 7].

In the reactor bio-shield concrete, tritium might be present as strong bound species, which is produced inside the crystal of minerals in the concrete during exposure to neutrons. Meanwhile tritium water (HTO) released from the reactor, especially the reactors which use heavy water as coolant, might also bind to concrete in HTO form. Based on the different volatility of tritium species, they can be separated by heating/combustion at different temperatures. Tritium water (HTO) associated to concrete can be liberated at 100℃. While strongly bound tritium produced by neutron activation of concrete cannot be released at temperature <300℃, and have to be combusted at >700℃ [40].

In the Danish research reactor DR-3, due to leakage of heavy water from the reactor tank, the reactor bio-shield concrete was contaminated by heavy water with high HTO content in addition to tritium produced in the bio-shield concrete during reactor operation. A concrete core from DR-3 was sampled under cooling using nitrogen gas from liquid nitrogen container, in order to investigate the distribution of tritium species. The sampled concrete core was immediately cooled and stored at -20℃ until analysis. Before the analysis, the concrete core was cooled by immersing the concrete in a plastic bag into liquid nitrogen, and then sliced to about 1 cm slice. The slice was then immediately crushed to powder (<0.5 mm), and transferred to a round bottom flask. The concrete powder in the flask was quickly connected to a heating and condensing system. The flask was heated using a heating mantle at 200±20℃ for 2 hours. The released tritium water (HTO) was condensed in the condenser and receiver by cooling, and transferred into a vial by rinsing the condenser tube and receiver using tritium free water for tritium measurement. The remained concrete powder was further analyzed using the Sample Oxidizer system for strongly bound tritium by combustion at >1000 ℃ [41].

In metals used in the nuclear facilities, tritium mainly originates from adsorption from the air as HT or HTO or local production through neutron activation reactions of B (¹⁰B (n, 2α)³H, Li (⁶Li(n, α) ³H) and N (¹⁴N(n, ³H) ¹²C). It is proposed that when the metal is exposed to tritium (HT or HTO), tritium enters metals by diffusion with a rate controlled by the metal composition and surface feature, and tritium is normally trapped in hydrated oxide. While in irradiated metals, tritium forms internally through neutron capture reactions with boron and lithium inside the metal bulk. It was found that the adsorbed tritium can be released at temperature of lower than 400℃, while release of internally produced tritium in the metal needs to be heated at higher temperature of 600-900℃ [63]. For the sample containing both adsorbed and local produced tritium, the distribution of tritium might be highly heterogeneous, a careful sampling and sample preparation should be given to obtain a representative sample for tritium analysis.

Tritium species in biological samples including food are also an interesting issue, which is related to the estimation of radiation exposure through food consumption, which is also a key issue in the investigation of the transfer pathway of tritium from soil and air to vegetation. Species of tritium in sediments is also an interesting issue regarding to sources, transfer and accumulation of tritium in the environment [45]. Speciation analysis of tritium in biological samples and sediment mainly focus on tritiated water, water exchangeable and non-exchangeable organic tritium. Tritiated water in fresh biological and sediment samples can be separated by freeze-drying or distillation. Water exchangeable tritium is separated from dried samples by leaching using tritium-free water [46]. The non-exchangeable organic tritium remains in the leached samples is finally separated by a combustion method using sample oxidizer or tube furnace (Fig. 7). The tritiated water can also be extracted by water leaching using tritium -free water, based on quick exchange of tritium water in the biological samples and sediment with leaching water. Experiment has shown that tritiated water adsorbed in silica gel, concrete and brick can be quickly leached out by water [46, 47, 48]. However, the water exchangeable organic tritium will also be extracted, and a combined fraction of water exchangeable tritium is obtained [45].

In the environment around some of the NPPs, tritium and ¹⁴C in air and foodstuff are monitored in the routine monitoring program. Table 7 shows reported levels of tritium and ¹⁴C in air and tea leaves collected in the agriculture land around the Qinshan NPP, China (with 9 units, among them 2 units are CANDU type reactors) [49].

In summary, three methods have being used for monitoring ¹⁴C and tritium in the gaseous releases and atmosphere, i.e. ¹⁴CO₂, total ¹⁴C, and different species of ¹⁴C and HTO , total tritium and different species of tritium. Because tritium is mainly discharged as HTO vapor in the gaseous form, HTO is often determined in the gaseous discharge from nuclear facilities in routine monitoring program. The species of ¹⁴C in gaseous discharges vary depending on the type of reactor, monitoring of different species of ¹⁴C, mainly ¹⁴CO₂ and hydrocarbon-14, is in general implemented in the gaseous discharges from nuclear facilities. ¹⁴CO₂ is mainly trapped in NaOH solution or organic amine in a bubbler. A trapping efficiency of >99% for ¹⁴CO₂ is obtained in this method. HTO is often collected using a condenser or silica gel dryer. For speciation analysis of ¹⁴C, the gas from the stack first passes through a bubbler with NaOH to trap ¹⁴CO₂, and then the off gas passes through a furnace tube filled with CuO or Pt/Pd catalyst and heated to 400-800℃ to convert hydrocarbon and CO to CO₂, which is then trapped in the second bubbler. A combined system is also used to monitoring both ¹⁴C and tritium in the gaseous discharges using the same principles. For atmosphere samples, the concentrations of ¹⁴C and tritium are normally low and molecular sieves are often used for collecting ¹⁴CO₂ and HTO, respectively. In this case, a long sampling period is applied to collect sufficiency amount of tritium and ¹⁴C for measurement. In the reactor water and liquid discharges, ¹⁴C mainly exists as dissolved gaseous ¹⁴C (volatile hydrocarbons), inorganic carbon (mainly carbonate) and water-soluble organic compounds. Three types of methods have been used to separate ¹⁴C and its species in the water samples, combustion, wet oxidation using a strong oxidative reagents and acid digestion. The combustion and wet oxidation methods are used to determine the total soluble ¹⁴C in the water after evaporating the water to dryness. Acid digestion is used to separate the ¹⁴C in carbonate form. By using acid digestion followed by wet oxidation, both inorganic and organic ¹⁴C can be separated. Systems with a tube furnace or a sample Oxidizer with CuO or Pd/Pt catalyst has been used for separation of ¹⁴C and tritium in solid samples. In the liquid discharges and environmental water samples, the water form of tritium dominates, with only a small fraction of organic tritium forms. Tritium in water form is separated by distillation and measured by LSC. The organic associated tritium can be separated using tube furnace or sample Oxidizer.

Although a number of methods for speciation analysis of tritium and ¹⁴C in the discharges from the nuclear facilities and environment have been developed and some commercial instruments are available. The methods developed a few decades ago are still used. In these methods, only ¹⁴CO₂ and HTO in gaseous discharges and total ¹⁴C and HTO in liquid discharges are monitored. It is a challenge to update and replace the old methods by the state of the art methods of speciation analysis of tritium and ¹⁴C in order to provide a more useful and comprehensive information of tritium and ¹⁴C released from the nuclear facilities for precise estimation of radiation dose to the critical population.

4. Analytical methods for measurement of ¹⁴C and tritium

The most often used techniques for off-line measurement of tritium and ¹⁴C is liquid scintillation counting (LSC). Among other techniques, accelerator mass spectrometry (AMS) is the most sensitive technique for the detection of ¹⁴C, and is often used for analysis of non-nuclear samples but normally not used for samples from nuclear facilities such as discharges, resin and waste.

Noble gas mass spectrometry is the most sensitive technique for measurement of tritium in very low levels (< mBq), which is based on the measurement of ³He, the decay daughter of tritium. Helium in the sample is first removed by vacuum and the sample in a tight container is kept for a few months for ingrowth of tritium from tritium decay, the produced ³He is separated and measured using noble gas mass spectrometry. However, this method is not used for analysis of discharge and waste samples from nuclear facilities. Ionization chamber and proportional counter are often used for directly measurement of gas tritium (HT, T2) and ¹⁴C (¹⁴CO₂, CH₄), and they are often used in the most of on-line monitors for tritium and ¹⁴C.

4.1 Measurement of tritium using LSC

Tritium in water form is normally prepared as purified water for measurement using LSC. To improve the counting efficiency, water is filtered, neutralized and distilled to remove the chemicals to reduce the quench effect. A scintillation cocktail is mixed with the purified water for LSC measurement, a better compatible cocktail is normally selected to use bigger volume of water and to obtain a better counting efficiency, and finally a good detection limit. For determination of tritium in environmental samples, 10 ml water sample is often used and mixed with 10 ml scintillation cocktail for LSC measurement. In this case, a detection limit of 1-2 Bq·L^-1 can be obtained using ultra-low level LSC (e.g. Quantulus 1220, PerkinElmer, USA). For improvement of detection limit, a high capacity LSC (e.g. AccuFLEX LSC-LB7, HITACHI ALOKA, Japan) with 145 ml vial can be used. In this case, up to 70 ml water sample can be used, and a detection limit down to 0.3 Bq·L^-1 can be reached.

4.2 Measurement of ¹⁴C using LSC

¹⁴C collected or separated from samples is normally absorbed in NaOH or CarboSorb trap solution, solid absorbent such as Ascarite (NaOH immobilized in a solid support), NaOH pellets or molecular sieve. CarboSorb and diluted NaOH solution can be directly mixed with scintillation cocktail for LSC measurement, but a relative high quench level is normally obtained when directly mixing 2-4 mol·l-1 NaOH with scintillation cocktail, and therefore a less counting efficiency will be obtained. Therefore, ¹⁴C absorbed in NaOH solution, NaOH pellet and immobilized NaOH solid is normally further separated. This is normally achieved by two methods. NaOH pellet is dissolved in water, excessive BaCl₂ is added to the NaOH solution, ¹⁴C is precipitated as BaCO₃, which is separated by centrifuge and washed by water. The BaCO₃ precipitate is dispersed in scintillation cocktail for LSC measurement, which reduce the quench level, meanwhile ¹⁴C can also be purified by removal of other radionuclides in the solution during precipitation [50, 51, 52]. However, the BaCO₃ powder mixed with scintillation cocktail can cause some level of physical quench. Another method is to release ¹⁴C absorbed in NaOH as CO₂ by addition of acid (H₂SO₄) under N₂ purging, the released ¹⁴CO₂ is trapped in CarboSorb (an amine solution), which is then mixed with specific scintillation cocktail for LSC measurement. ¹⁴C adsorbed in molecular sieve and Ascarite can be released as ¹⁴CO₂ by heating, and the released gaseous ¹⁴CO₂ is trapped in CarboSorb for LSC measurement after mixed with scintillation cocktail.

NaOH or CarboSorb trap solution for ¹⁴C cause a relative high quench in the LSC measurements, relative small samples can be analyzed with this method. Therefore, the detection limit of this method is not sufficient low for the determination of ¹⁴C in low-level environmental samples. The measurement of low level ¹⁴C using LSC can also be carried out by converting ¹⁴C to benzene. In this way, a large sample can be used and the quench can be overcome. To obtain this, ¹⁴C in CO₂ gas is reduced by lithium under heating to obtain lithium carbide (Li₂C₂), the hydrolysis of Li₂C₂ produces acetylene (C₂H₂), which is trimerised to benzene (C₆H₆) on vanadium catalyst [53]. This method has a much better detection limit for ¹⁴C, and therefore can be used for measurement of ¹⁴C in low-level samples. However, the analytical procedure is very time consuming and therefore more expensive.

4.3 Measurement of tritium and ¹⁴C using ionization chamber (on-line measurement)

The on-line monitoring of gaseous tritium and ¹⁴C normally use ionization chambers as detectors. In this instrument, the gas sample passes through the chamber, the beta particles emitted from ¹⁴C or tritium ionizes the surrounding gas in the ionization chamber, and produced ions are measured by the ionization current, which is proportional to the concentration of tritium or ¹⁴C in the gaseous sample passing through the chamber. In general, this type of monitors could not discriminate between radionuclides, when used for specific radionuclides, i.e. tritium or ¹⁴C. Therefore, ¹⁴C or tritium has to be separated from each other and from other radionuclides before measurement. For ¹⁴C, any gaseous forms such as ¹⁴CO₂, ¹⁴CO, ¹⁴CH₄, and ¹⁴C₂H₆ can be measured. While for tritium the monitoring works best for tritium gas such as HT (T₂). Although tritiated water vapor can also be measured, but the high risk of adsorption and deposition of tritiated water in the ionization chamber generates a high risk of contamination. Hence, a high memory effect, and the detector requires frequent decontamination. Most of ionization chambers are the flow through type, which requires pumping of the gas sample through the system. An open window type of ionization chamber is also used. These chambers employ a cover; air or gas penetrate through the window to the inside chamber for measurement. This type of instrument works without pump, and are often used to monitor rooms, hoods, glovebox, etc.

Because HTO can be easily uptake by humans through inhalation compared to HT (10,000 to 25,000 times greater), it is desirable to know the relative amounts of each species following a significant release. In the case of stack monitoring, discrete samples of the stack discharge should be taken using bubblers or desiccants with a catalyst for oxidizing the HT. Fig. 9 shows a tritium monitor using ionization chamber as detector applied in nuclear power plant in Canada for monitoring tritium releases from stack [54]. This system is used to measure both total tritium and tritium gas (HT) by using two ionization chambers and a dryer between them. The gaseous sample passes through the first ionization chamber, in which the concentration of total tritium is measured, the off gas then passes through a desiccant to remove HTO vapor/moisture, leave the tritium gas (HT) which passes through the second ionization chamber to measure concentration of tritium gas (HT).

Another technique uses a semipermeable membrane tube bundle in the sampling line to remove the HTO (preferentially over the HT), which is directed to a HTO monitor. After removing the remaining HTO with another membrane dryer, the sampled air is directed to the HT monitor. Although this technique is slower than the one requiring a desiccant cartridge, it does not require a periodic cartridge replacement. Furthermore, it can be adapted to measure tritium in both species in the presence of noble gases or other radioactive gases by adding a catalyst after the HTO dryers, followed by additional membrane dryers for the HTO. However, because of its slow response, it is more suitable for discharge or stack monitoring than for room monitoring. Because significant releases into a room are quite rare, it is easier to treat any such release as one of HTO than using complicated techniques for continuous differential monitoring [55].

4.4 Measurement of ¹⁴C using AMS

The most sensitive measurement method for ¹⁴C is AMS. In this method, the sample in CO₂ gas is reduced by H₂ in the presence of powdered iron or by Zn metal to elemental carbon (graphite). The carbon/graphite is deposited on the powdered iron or zinc and the mixture is pressed into a target for measurement by accelerator mass spectrometry. This method can analyze small sample of microgram carbon and very low level of ¹⁴C down to μBq [56]. Because of high sensitivity, it is very often used to measure pre-nuclear samples for ¹⁴C dating purpose. It is very seldom to be used for measurement of ¹⁴C from nuclear facilities, because of high risk of contamination of laboratory by high level ¹⁴C samples.

In summary, ionization chamber is the often used instrument for on-line monitoring of tritium or ¹⁴C in gaseous form. This type of monitors could not discriminate between radionuclides. ¹⁴C or tritium has to be separated from each other and from other radionuclides before measurement. Two ionization chambers and a dryer between them are used for monitoring of gaseous tritium in HT and HTO. The first chamber is used to measure both HT and HTO, while the second chamber measure only HT after HTO is removed by a dryer. In general, this type of monitor is used to measure tritium or ¹⁴C in the stack with high concentration because of its low sensitivity. LSC is the most often used method for off-line measurement of tritium and ¹⁴C. ¹⁴C trapped in NaOH solution or pellet is normally converted to BaCO₃, which is suspended in scintillation cocktail for LSC measurement. Tritium in condensed water is directly measured using LSC after mixed with scintillation cocktail, while tritium trapped in dryer or molecular sieve has to be released and converted to water form for LSC measurement. The most sensitive method for tritium is noble gas mass spectrometry, which takes a long time for analysis because of ingrowth of ³He from tritium. AMS is the most sensitive method for ¹⁴C measurement; it can measure μBq level ¹⁴C and just need a small sample. But the sample preparation is time consuming and measurement is high cost. Therefore, these two techniques are seldom used for monitoring tritium and ¹⁴C in nuclear facilities.

5. Challenges, problems and methods in monitoring of tritium and ¹⁴C

In the past decades, a numbers of methods for the monitoring of tritium and ¹⁴C in air and liquid discharges has been developed and applied in the nuclear facilities including many nuclear power plants. The on-line monitor using ionization chamber is widely used for monitoring gaseous tritium in the nuclear facilities, especially heavy water reactors and reprocessing plants [29, 52]. While such an on-line monitoring system is still not well developed and widely used for monitoring gaseous ¹⁴C in the nuclear facilities because of low activity concentration of ¹⁴C compared to other radionuclides in the gas discharge. Separation and enrichment of ¹⁴C from the discharge is required before measurement, therefore off-line measurement is often applied. Development of a rapid and on-line monitoring system will be helpful for rapid monitoring of the releases of gaseous ¹⁴C, and therefore avoid accidental releases of elevated amounts of ¹⁴C to the environment.

The collection of representative sample is one of the major challenges and sources of analytical uncertainty in monitoring gas releases and atmosphere. A robust sampling system is critical, whereby the sample is extracted at a homogeneous location within the point source. This requires an evaluation of the sample environment, transport mechanisms, and collection materials. The criterion for the homogeneous sampling location includes a determination of the angular or cyclonic flow, uniformity of the air velocity profile, gas concentration profile, etc. Modeling techniques such as computational fluid dynamics may be used to validate a well-mixed location without the necessity of field tests conducted in the stack or vent [57].

A number of methods and instruments have been reported for separating different species of ¹⁴C and tritium in the gaseous releases from the nuclear facilities [35, 52, 54, 58, 59]. While, many facilities still measure only ¹⁴CO₂ and tritiated water in the gaseous releases, this might not be reasonable for PWRs, because of the high fraction of hydrocarbons in releases from PWRs. It is still a challenge to establish a reliable and easy operated system, which can be applied for monitoring different species of ¹⁴C and tritium released from nuclear facilities.

Due to the major contribution of radiation dose to the public, ¹⁴C monitoring in the environment surrounding the nuclear facilities becomes more important. Some countries have started to monitor the ¹⁴C level in the atmosphere and environmental media such as grass and tree ring [49, 59, 60]. However, because of relative low concentration of ¹⁴C in the atmosphere and environmental samples, its accurate determination is a challenge. A long sampling time or a big sample is needed to be able to measure the ¹⁴C concentration in the environmental level. The present method for determination of ¹⁴C in environmental samples is time consuming; normally take a few days to a week. It is a challenge to develop a more effective sampling instrument and rapid or on-line measurement instrument for prompt and accurate determination of ¹⁴C level.

The concentration of tritium in the environment is also very low compared to the nuclear facilities. The monitoring of tritium level in the atmosphere and other environmental samples needs a long sampling time, especially for different species of tritium. The measurement is also often carried out in the laboratory after purification of the collected sample. It is still a challenge for on-line and rapid monitoring the tritium in the atmosphere.

Although dominant ¹⁴C is released as gaseous discharge from the nuclear facilities, liquid discharges also contain various amount of ¹⁴C depending on the sources of ¹⁴C and trapping technology used in the nuclear facilities. ¹⁴C in the liquid discharges can be rapidly accumulated into food chain, causing a radiation exposure to the public. However, the monitoring of ¹⁴C in the discharges and recipients is still limited in the surrounding area of nuclear facilities. It is still a challenge to develop a rapid analytical method to monitor ¹⁴C in the environmental samples.