Preamble

Vol. XX, No. X, XXX 20XX
NUCLEAR TECHNIQUES

Study on Water Chemistry Control Strategy for Ammonia-Containing Coolant in the Primary Circuit

Tianjun Luo¹, Yukun Yuan¹, Xiaoyuan Zhou¹, Zifang Guo¹, Yingzhe Du¹, Peng Lin¹, Guojun Hu²

¹China Nuclear Power Technology Research Institute Co., Ltd., Shenzhen 518028, China
²School of Nuclear Science and Technology, University of Science and Technology of China, Hefei 230026, China

Abstract

[Background] In some reactors, ammonia is added to regulate the pH of the primary coolant. The oxidizing species is effectively suppressed by radiolytic products of ammonia, thereby the coolant is maintained in a reductive chemical environment. [Purpose] This study aims to develop a predictive model capable of simulating water chemistry behavior under various control strategies to achieve coordinated regulation of both pH and dissolved hydrogen concentrations. [Methods] In this study, a transport model for radiolytic species in reactor coolant was developed based on the RETA reactor system analysis code, with applicability to a wide range of reactor types. The model demonstrated high predictive accuracy, with root-mean-square errors of only 1.79×10⁻⁸ for NH₃ and 5.69×10⁻⁸ for H₂ concentrations compared with experimental data. Taking the KLT-40S reactor as a reference, three ammonia injection strategies were established and comparatively evaluated: initial dispersion, constant-rate injection, and constant-rate injection optimized with hydrogen removal. The simulations were conducted by defining initial coolant parameters and radiation field conditions, followed by stepwise adjustment of ammonia injection rates and the timing of hydrogen removal. Each strategy was simulated until the system reached a quasi-steady state (1.6×10⁴ s), allowing assessment of its effectiveness in controlling pH and dissolved hydrogen levels. [Results] The results indicated that the initial dispersed injection strategy, while simple and effective in maintaining a reductive chemical environment, was only able to sustain pH regulation for less than 5 hours. Conversely, the constant-rate ammonia injection strategy enabled sustained pH control but resulted in excessive dissolved hydrogen concentrations, necessitating an appropriate hydrogen removal scheme. The hydrogen removal-optimized constant-rate ammonia injection strategy is capable of simultaneously regulating pH and maintaining dissolved hydrogen concentrations. The ammonia injection rate is 1.64 g·s⁻¹, and the hydrogen removal system is initiated 1200 s after the start of ammonia injection, with a controlled removal rate of 0.014 g·s⁻¹. When the system is stable, the coolant pHT stabilizes at 6.9, with dissolved hydrogen concentrations ranging from 30 to 35 mL·kg⁻¹ (STP). [Conclusions] This study is anticipated to provide valuable insights for the development of new reactor types and the optimization of water chemistry control strategies.

Keywords: Coolant, Reactor, Ammonia, Radiolysis, Water Chemistry

Small Modular Reactors (SMRs) have demonstrated promising application prospects in specialized fields such as icebreakers and floating nuclear power plants due to their compact structure, light weight, and flexible deployment. However, the highly integrated and compact design of SMR loop systems imposes more stringent requirements on radiation protection and material service safety compared to conventional reactors [1,2]. In particular, the generation and accumulation of radiolysis products, such as active species including oxygen (O₂) and hydrogen peroxide (H₂O₂), may adversely affect system chemical stability and operational safety [3-5]. Therefore, to gain a thorough understanding of the generation mechanisms of radiolysis products and their transport and transformation behavior in the coolant system, it is essential to obtain accurate thermal-hydraulic parameters and tightly couple them with water chemistry processes, enabling comprehensive analysis of radiochemical effects in SMR coolant systems.

The issue of irradiation in nuclear reactors has long been recognized [6-8]. Scott's research demonstrated that, in addition to direct radiation damage to materials, coolant radiolysis is also a critical factor contributing to material corrosion [9]. Radiolysis refers to the production of various chemical species in the reactor core region under high-flux neutron and gamma radiation [10,11]. Water chemistry issues primarily focus on the generation, transport, and reaction behavior of these radiolysis products. While most products have extremely short lifetimes, some relatively stable species (such as H₂O₂ and O₂) can persist in the system and may exert adverse effects on water quality [11,12]. As these oxidizing species migrate through the primary circuit, they can cause severe oxidative corrosion, posing potential threats to critical thin-walled structures such as stainless steel heat exchanger tubes [13]. Recently, Sultana et al. conducted experimental studies on radiolysis effects in supercritical water (SCW) and indicated that similar issues may exist in SMR systems [14].

Since water chemistry issues have received widespread attention, the radiolysis process of coolants has been extensively studied. Ershov et al. pioneered and validated a kinetic model applicable to pure water radiolysis, systematically describing the generation mechanisms of major radiolysis products such as H₂, H₂O₂, and O₂ [15]. Subsequently, Joseph et al. further investigated radiolysis effects in pure water through extensive experiments and constructed a kinetic model capable of predicting the generation behavior of multiple species [16]. Elliot compiled and reported 60 key radiolysis reactions and their rate constants, providing an important foundation for water chemistry simulation calculations [17].

Subsequent research has focused on achieving better agreement with experimental results and expanding the simulation scope. Yakabuskie improved and refined the model by introducing mass transfer processes at the two-phase interface based on Joseph's work [18]. Palfi investigated the generation processes and yields of transient species in aqueous solutions under pulsed irradiation conditions using a combined experimental and computational approach [19]. Karditsas studied the radiolysis process in the ITER fusion reactor, proposing a time-dependent transport model and providing calculation examples [20]. Yousefi extended the application scope of radiolysis models to high-temperature and high-pressure conditions, providing a theoretical foundation for analyzing radiolysis behavior under more complex operating conditions [21].

Another category of research has focused on the influencing factors of radiolysis products and their control strategies. Roth and Daub analyzed the effect of pH on H₂O₂ generation and carbon steel corrosion behavior [22,23]. Dey reviewed the transformation processes of nitrogen-containing compounds in aqueous solutions under irradiation conditions, proposing that NH₃ has the potential to regulate oxidizing species and providing a new possible approach for corrosion control [24]. Building upon this foundation, Yakabuskie conducted studies on the radiolysis behavior of NO₃⁻ solutions [25], while Guo analyzed the radiolysis process of aqueous ammonia under various influencing factors [26]. Kumar evaluated the role of N₂H₄ in corrosion control [27], and Iwamatsu investigated the effectiveness of H₂ in suppressing H₂O₂ accumulation [28].

The aforementioned work has primarily focused on homogeneous systems in zero-dimensional frameworks, neglecting spatial distribution effects and lacking guidance for loop analysis. Considering that radiolysis products are generated in the core and transported through the primary circuit, their concentration evolution can be calculated by combining mass, momentum, and energy equations. In this work, we have integrated research in radiation chemistry, thermal-hydraulics, and numerical computation to develop a model for simulating the coupled thermal-hydraulic and water chemistry effects in SMRs. This model is based on the RETA system analysis code, with an added submodule for water chemical reactions and transport. Subsequently, the reliability and performance of the model were verified and evaluated, and the primary circuit of the KLT-40S reactor was modeled to analyze the regulatory effects and advantages/disadvantages of two ammonia injection strategies. This work is expected to provide valuable reference for optimizing ammonia-based water chemistry control strategies.

1.1 Radiolysis Process

The radiolysis process can be summarized into three stages: (1) radiation deposits energy in the coolant, leading to the production of a series of excited species such as H₂O*, H₂O⁺, and e⁻; (2) the excited species react to generate intermediate products such as ·H and ·OH radicals; and (3) the excited species and intermediate products interact to form final radiolysis products that diffuse out of the spur, including H₂O₂, O₂, and H₂ (as shown in R0).

For zero-dimensional calculations of coolant radiolysis, the G-value is introduced to convert the radiation dose rate into source terms for radiolysis product generation. The production rate of species i, denoted as $r_i$, is defined as:

$$\text{R0: } H_2O \rightsquigarrow H_2, H_2O_2, e_{aq}^-, \cdot H, \cdot OH, H_3O^+, HO_2\cdot$$

$$r_i = \frac{\rho D G_i N_A}{100}$$

where $r_i$ is the radiolytic production rate of species i (mol·L⁻¹·s⁻¹), $\rho$ represents the coolant density (kg·L⁻¹), $e$ is the electron charge (1.9×10⁻¹⁹ C), $G_i$ denotes the radiation chemical yield of species i (/100 eV), $D$ is the absorbed dose rate (Gy·s⁻¹), and $N_A$ is Avogadro's constant (6.02×10²³ mol⁻¹).

Under continuous irradiation, various species react with each other. For species i, which is continuously produced and consumed during radiolysis, its net production can be expressed by equation E2:

$$\frac{dC_i}{dt} = r_i + \sum_{s=1}^{N} \sum_{m=1}^{M} (k_{sm}^i C_s C_m) - \sum_{s=1}^{N} k_{si} C_s C_i$$

where $k_{sm}^i$ is the rate constant for the reaction between species s and m (producing i) or between s and i (consuming i), $C_m$, $C_s$, and $C_i$ are the concentrations of species m, s, and i, respectively, N is the total number of reaction equations, and M is the number of species involved in the reactions.

1.2 Thermal-Hydraulic Model

For engineering applications, zero-dimensional radiolysis calculations are insufficient and should be extended to at least the primary loop and its major components. This necessitates coupled thermal-hydraulic mass transfer calculations in addition to radiolysis reactions. The simplified primary loop model and mesh division used in this paper are shown in Figure 1 [FIGURE:1].

Figure 1 Diagram of the simplified primary loop and computational mesh division

To numerically solve the equations and integrate radiation chemistry into reactor analysis, programming was performed based on the RETA software from the University of Science and Technology of China. The finite volume method (FVM) with fully implicit numerical discretization was employed, and the control equations were solved using the Preconditioned Jacobian-Free Newton-Krylov (PJFNK) algorithm. Based on RETA, an additional module called R-WRC was developed, dedicated to radiolysis process calculations and capable of coupling with RETA's core thermal-hydraulic functions. The implementation in R-WRC involves direct programming to include transport terms:

$$\frac{\partial \mathbf{C}}{\partial t} + \nabla \cdot (\mathbf{v}\mathbf{C}) = \nabla \cdot (\mathbf{D}\nabla\mathbf{C}) + \mathbf{r}$$

where $\mathbf{C}$ is the species concentration vector, $\mathbf{v}$ is the local velocity vector, $\mathbf{D}$ represents the diffusion coefficients of different species, and $\mathbf{r}$ is the source term for radiation-induced products.

Thermal-hydraulic calculations are crucial for parameters such as $\rho$, $\mathbf{v}$, and $\mathbf{D}$. When combining radiolysis processes with thermal-hydraulics, the following simplifying assumptions were adopted:

1. Only one-dimensional equations are considered, corresponding to the one-dimensional mesh in RETA system thermal-hydraulic simulations, similar to RELAP;
2. Diffusion terms are neglected, as the velocity in the main loop is sufficiently high to ignore species diffusion processes.

Based on these two assumptions, the simplified equation becomes:

$$\frac{\partial \mathbf{C}}{\partial t} + \mathbf{v} \cdot \nabla\mathbf{C} = \mathbf{r}$$

Due to the large number of mesh nodes that need to be considered in the solution process, resulting in extreme computational complexity, the following programming structure was established: (1) process each species separately on all mesh nodes; (2) combine all species into a unified vector; and (3) utilize a large sparse matrix for efficient computation.

1.3 Model Validation

To verify the reliability of the R-WRC module in radiolysis process calculations, its results were compared with simulation data from the dedicated radiochemistry software FACSIMILE, as shown in Figure 2 [FIGURE:2]. Both R-WRC and FACSIMILE were used to simulate the radiolysis behavior of aqueous ammonia solutions with initial ammonia concentrations of 1–3 mol·L⁻¹ at 25 °C, comparing the temporal evolution of H⁺ concentration (pH), ammonia concentration, and N₂ concentration under irradiation.

The results demonstrate that R-WRC can accurately describe the radiolysis process, with minimal differences between its simulation results and those from FACSIMILE, indicating good computational accuracy and reliability.

Figure 2 Comparison between the R-WRC in present work and FACSIMILE simulation results: (a) H⁺, (b) N₂, (c) NH₃

To further validate the model's reliability, radiolysis experiments were conducted using an aqueous ammonia solution with an initial ammonia concentration of 30 mg·L⁻¹, which was continuously irradiated for 72 h. The measured changes in NH₃ and H₂ concentrations were compared with simulation results, as shown in Figure 3 [FIGURE:3]. The simulated values for NH₃ and H₂ show good agreement with experimental data, with mean square errors (MSE) of only 1.79×10⁻⁸ and 5.69×10⁻⁸, respectively, indicating high accuracy in describing the main reaction pathways and radiolysis product generation. The pH simulation results show some deviation from experimental values, but the MSE remains low at only 0.03, which is speculated to result from continuous CO₂ dissolution during the experiments affecting the system acidity.

In summary, the model established in this work can effectively predict the evolution of major radiolysis products during aqueous ammonia solution radiolysis and demonstrates high reliability.

Figure 3 Comparison of simulated and experimental results for ammonia solution radiolysis

2 Model Establishment and Validation

The objective of primary circuit water chemistry control is to ensure the integrity of the primary system pressure boundary, fuel cladding integrity, and realization of fuel design performance, while minimizing the ex-core radiation field level. The core control strategy involves maintaining coolant pH and dissolved hydrogen concentration within acceptable ranges to ensure the coolant remains in an alkaline and reducing environment. For conventional pressurized water reactors using coordinated boron-lithium control, the pHT value during operation is typically controlled between 6.9 and 7.2, corresponding to OH⁻ ion concentrations of approximately 10⁻⁵–10⁻⁶ mol·L⁻¹; dissolved hydrogen concentration is usually maintained within the range of 2–4.5 mg·L⁻¹.

2.1 Simulation of Initial Dispersed Ammonia Injection Strategy

The initial dispersed ammonia injection strategy refers to injecting a given amount of ammonia solution into the primary circuit at the initial operating state, allowing rapid dispersion and uniform mixing throughout the coolant system. In the model, this is represented by assigning identical initial ammonia concentrations to all mesh nodes to simulate uniform distribution. As core irradiation continues and coolant circulates through the loop, species in each mesh node are continuously transported and updated, with concentrations dynamically changing under the influence of reaction and transport processes, exhibiting temporal evolution.

To investigate the effects of different initial ammonia dispersion concentrations, the model was used to calculate the temporal evolution of H₂ and OH⁻ concentrations during radiolysis of ammonia-containing coolant in the range of 0–100 mg·L⁻¹. The 0 mg·L⁻¹ pure water radiolysis data served as a control group, with 5 and 10 mg·L⁻¹ representing low concentration levels, 17, 34, and 51 mg·L⁻¹ representing medium concentration levels, and 60, 80, and 100 mg·L⁻¹ representing high concentration levels. To evaluate the chemical effects caused by the slow depletion of ammonia during long-term reactor operation and observe the establishment of concentration steady states, the simulation time was set to 1.6×10⁴ s, with results shown in Figure 4 [FIGURE:4].

Figure 4(a) shows that H₂ concentration quickly reaches chemical steady state as irradiation proceeds, and increasing the initial ammonia dispersion concentration significantly enhances H₂ production rate. At an irradiation time of 1.6×10⁴ s, the H₂ concentration in the 0 mg·L⁻¹ system is approximately 0.02 mg·L⁻¹, low ammonia concentration groups show H₂ concentrations of 0.62–1.25 mg·L⁻¹, medium ammonia concentration groups exhibit 2.12–6.34 mg·L⁻¹, and the high ammonia concentration group (100 mg·L⁻¹) reaches H₂ concentrations as high as 12.36 mg·L⁻¹. These results indicate that initial ammonia concentrations of 17 and 34 mg·L⁻¹ in the medium concentration dispersion strategy can meet H₂ limit requirements, while low and high ammonia concentration groups fail to satisfy the requirements.

From the perspective of coolant pH, ammonia dispersion strategies at all concentrations fail to maintain alkaline conditions over extended operation times, as shown in Figure 4(b). After initial ammonia injection and irradiation commencement, coolant pH rapidly decreases from weakly alkaline to neutral conditions (at 280 °C), departing from the alkaline control limit range within 1–5 h and eventually approaching neutral levels within 24 h. It is noteworthy that short-term pH changes do not directly exacerbate material corrosion behavior; however, the loss of alkaline environment may induce more unfavorable corrosion tendencies in metallic materials, leading to potential risks in actual reactor operation.

In summary, the initial dispersed ammonia injection strategy can maintain dissolved hydrogen concentration within target limits by controlling ammonia concentration in the range of 17–34 mg·L⁻¹. However, due to the difficulty in maintaining alkaline conditions long-term and the rapid pH reversion to neutral after irradiation begins, this strategy struggles to simultaneously achieve the two primary objectives of coolant chemistry control: meeting dissolved hydrogen concentration requirements and maintaining a weakly alkaline environment.

Figure 4 Dissolved H₂ and OH⁻ concentrations as a function of time under the initial ammonia dispersion strategy at 280°C and absorbed dose rate of 1×10⁴ Gy·s⁻¹: (a) H₂, (b) OH⁻

2.2 Simulation of Constant-Rate Ammonia Injection Strategy

Since the initial dispersed ammonia injection strategy struggles to meet primary circuit water chemistry control requirements, this paper adopts a constant-rate ammonia injection strategy to achieve long-term stable pH control in the coolant system. This strategy involves continuously applying a stable ammonia source term before the coolant enters the core, simulating the continuous ammonia injection effect of the charging pump in the primary circuit chemical volume control system. Compared with the initial dispersion method, this strategy can maintain relatively stable ammonia concentration in the core over longer periods. This section calculates the evolution of dissolved hydrogen and OH⁻ concentrations under continuous ammonia injection rates of 0–16.4 g·s⁻¹ and square-wave intermittent injection at 1.64 g·s⁻¹, with results shown in Figure 5 [FIGURE:5].

Figure 5 Dissolved H₂ and OH⁻ concentrations as a function of time under continuous ammonia feeding strategy at 280°C and absorbed dose rate of 1×10⁴ Gy·s⁻¹: (a) H₂, (b) OH⁻

Figure 5(a) shows that excessive ammonia injection rates under the constant-rate strategy can cause dissolved hydrogen concentration to rapidly exceed limit values. For example, at an injection rate of 1.64 g·s⁻¹, H₂ concentration exceeds 20 mg·L⁻¹ at 8000 s, far above the upper limit. Even at lower injection rates (0.164 g·s⁻¹), H₂ concentration may exceed limits if irradiation time is sufficiently long. Although stopping ammonia injection can suppress further H₂ accumulation, the concentration does not rapidly return to normal range and may instead increase further due to continued radiolysis of residual ammonia in the system. These results indicate that the constant-rate ammonia injection strategy carries potential risks of H₂ concentration exceeding limits.

An improvement was attempted based on the constant-rate strategy by adopting a square-wave intermittent injection mode, i.e., injecting ammonia at 1.64 g·s⁻¹ for 1 h, stopping injection, then resuming constant-rate injection after a 1 h interval, repeating this cycle. Simulation results show that this improved strategy still carries the risk of dissolved hydrogen concentration exceeding limits within a relatively short time (approximately 2000 s). Although pH can be maintained relatively stable initially, as shown in Figure 5(b), the cumulative effect of ammonia in the system with increasing cycles may cause pH to continuously rise, creating potential overrun risks.

Furthermore, as shown in Figure 5(b), square-wave intermittent injection demonstrates stronger pH controllability compared to constant-rate injection. The latter only shows short-term pH controllability at injection rates of 0.164 and 1.64 g·s⁻¹, while both too low (0.0164 g·s⁻¹) and too high (16.4 g·s⁻¹) injection rates fail to effectively regulate coolant pH.

In summary, square-wave intermittent injection has certain effectiveness in pH control, allowing small fluctuations of system pH within the target range by adjusting injection frequency and concentration. However, regardless of the control method employed, H₂ concentration carries the risk of exceeding limits, making it difficult for this ammonia injection strategy to fully meet established water chemistry control objectives. To address the continuous rise in dissolved hydrogen concentration under this strategy, hydrogen removal measures can be further introduced to simulate their effectiveness in controlling H₂ concentration within limits, enabling comprehensive optimization and regulation of water chemistry status.

2.3 Simulation of Hydrogen-Removal-Optimized Constant-Rate Ammonia Injection Strategy

Since both initial dispersion and constant-rate ammonia injection strategies have limitations, this section introduces a hydrogen removal mechanism based on the constant-rate injection strategy. Corresponding hydrogen removal rates were matched for ammonia injection rates of 0.164 and 1.64 g·s⁻¹ (for 0.164 g·s⁻¹ ammonia injection: −0.00148 g·s⁻¹; for 1.64 g·s⁻¹ ammonia injection: −0.0140, −0.0150, −0.0160 g·s⁻¹). The hydrogen removal process was simulated by setting a negative H₂ source term at the top of the core in the mesh division. It should be noted that earlier activation of the hydrogen removal device results in lower steady-state levels of hydrogen concentration and pH, while later activation leads to higher steady-state levels. Since the initial H₂ concentration in the coolant system is zero, to avoid computational errors and based on comprehensive considerations of pH control after ammonia injection and hydrogen removal rates, this negative source term was set to activate only after reaching specific times (1200 s, 10800 s). Simulation results are shown in Figure 6 [FIGURE:6].

Figure 6 Dissolved H₂ and OH⁻ concentrations as a function of time under the hydrogen-removal-optimized continuous ammonia feeding strategy at 280°C and absorbed dose rate of 1×10⁴ Gy·s⁻¹: (a) H₂, (b) OH⁻

Figure 6(a) shows that higher hydrogen removal rates can cause dissolved hydrogen concentration to drop rapidly, potentially leading to adverse consequences such as oxygen removal failure. For example, under ammonia injection rate of 1.64 g·s⁻¹, when hydrogen removal rates are 0.016 and 0.015 g·s⁻¹, H₂ concentration reaches the lower limit at approximately 2000 s and 7200 s, respectively. In contrast, appropriately reducing the hydrogen removal rate (e.g., 0.014 g·s⁻¹) can maintain H₂ concentration within a reasonable range. Similarly, at lower ammonia injection rates (0.164 g·s⁻¹), further reducing the hydrogen removal rate (−0.00148 g·s⁻¹) also helps achieve H₂ concentration compliance.

pH evolution is shown in Figure 6(b). In constant-rate ammonia injection strategies without hydrogen removal, pH continuously increases with irradiation time, eventually exceeding the pH limit upper bound. When ammonia injection rate is 1.64 g·s⁻¹ with hydrogen removal rates of 0.016 and 0.015 g·s⁻¹, pH can only be maintained for 2 and 4 h, respectively. Further analysis of two ammonia injection strategies that meet H₂ concentration requirements reveals that both approaches satisfy pH limit ranges, with the general principle that lower ammonia injection rates result in lower hydrogen removal pressure and lower steady-state pH and dissolved hydrogen levels.

In summary, through optimization studies of hydrogen removal for the constant-rate ammonia injection strategy, two reasonable control schemes were obtained:

1. Constant-rate ammonia injection at 1.64 g·s⁻¹, with hydrogen removal device activated 1200 s after ammonia injection begins, controlled at a removal rate of 0.014 g·s⁻¹, corresponding to pHT = 6.9;
2. Constant-rate ammonia injection at 0.164 g·s⁻¹, with hydrogen removal device activated 10800 s after ammonia injection begins, at a removal rate of 0.0014 g·s⁻¹, corresponding to pHT = 6.5.

Both schemes yield dissolved hydrogen concentrations of approximately 2.7–3.1 mg·L⁻¹, equivalent to 30–35 mL·kg⁻¹ (STP). However, considering that higher coolant pH can effectively reduce corrosion of metallic materials in the primary coolant system, the first scheme is superior.

Conclusions

This paper established a water chemistry control model for ammonia-containing coolant in the primary circuit and investigated the radiolysis behavior of ammonia-containing coolant using the KLT-40S reactor as an example. The effects of initial dispersed ammonia injection, constant-rate ammonia injection, and hydrogen-removal-optimized constant-rate ammonia injection strategies on dissolved hydrogen and pH were compared. The results demonstrate that:

1. The initial dispersed ammonia injection strategy can achieve dissolved hydrogen concentration compliance and maintain coolant reducing control, but cannot effectively regulate coolant pH;
2. The constant-rate ammonia injection strategy has shortcomings in dissolved hydrogen concentration control; while square-wave intermittent injection can achieve pH regulation, it cannot maintain dissolved hydrogen concentration within limits for extended periods;
3. The hydrogen-removal-optimized constant-rate ammonia injection strategy can simultaneously achieve coolant pH regulation and dissolved hydrogen concentration control by matching appropriate hydrogen removal rates;
4. This paper also provides a reasonable water chemistry control scheme: ammonia injection rate of 1.64 g·s⁻¹, hydrogen removal initiated 1200 s after ammonia injection begins, at a removal rate of 0.014 g·s⁻¹.

Author Contributions

Tianjun Luo and Yukun Yuan were responsible for code development, data analysis, and manuscript drafting; Xiaoyuan Zhou handled data processing; Zifang Guo contributed to manuscript revision and final proofreading; Yingzhe Du and Peng Lin provided the research platform and secured funding; Guojun Hu was responsible for quality review and content oversight.

References

[1] Hussein E M A. Emerging small modular nuclear power reactors: A critical review[J]. Physics Open, 2020, 5: 100038. DOI: 10.1016/j.physo.2020.100038.

[2] LV Xin, LI Shu, YIN Junlian, et al. Multi-component radiolysis equilibrium calculation model for nitrogen-water system[J]. Atomic Energy Science and Technology, 2024, 58(11): 2310-2317. DOI: 10.7538/yzk.2023.youxian.0862.

[3] ZHANG Yang, GUO Zifang, LIU Chunyu, et al. Radiolysis process simulation of coolant containing ammonia for pressurized water reactor's primary circuit[J]. Journal of Harbin Engineering University, 2024, 45(12): 2377-2383. DOI: 10.11990/jheu.202307039.

[4] GUO Zifang, ZHANG Yang, LI Yifan, et al. γ-Radiolysis of the aqueous ammonia solution saturated by N2[J]. Journal of Radiation Research and Radiation Processing, 2024, 42(1): 30-39. DOI: 10.11889/j.1000-3436.2023-0088.

[5] CAO Qi, DU Xiangyi, GUO Zifang, et al. γ-radiolysis of boric acid-lithium hydroxide-ammonia coolant[J]. Journal of Radiation Research and Radiation Processing, 2023, 41(3): 15-20. DOI: 10.11889/j.1000-3436.2023-0020.

[6] WU Shifa, XIE Jingyao, WAN Jiashuang, et al. Control characteristics of the steady-state operation scheme for coolant temperature in the primary loop of the Lead-Bismuth cooled fast reactor[J]. Nuclear Techniques, 2025, 48. DOI: 10.11889/j.0253-3219.2025.hjs.48.240386.

[7] ZHANG Weihao, LIU Maolong, ZENG Chen, et al. Experimental and numerical simulation study on the solidification behavior of molten salt reactor coolants[J]. Nuclear Techniques, 2025, 48(4): 040603. DOI: 10.11889/j.0253-3219.2025.hjs.48.230405.

[8] LUO Hui. Systematic measurement method for primary coolant flow rate in advanced passive nuclear power units[J]. Nuclear Techniques, 2023, 46(10): 100602. DOI: 10.11889/j.0253-3219.2023.hjs.46.100602.

[9] Scott P. A review of irradiation assisted stress corrosion cracking[J]. Journal of nuclear materials, 1994, 211(2): 101-122. DOI: 10.1016/0022-3115(94)90360-3.

[10] Draganić I G. Radiolysis of water: a look at its origin and occurrence in the nature[J]. Radiation Physics and Chemistry, 2005, 72(2-3): 181-186. DOI: 10.1016/j.radphyschem.2004.09.012.

[11] Uchida S, Katsumura Y. Water chemistry technology – one of the key technologies for safe and reliable nuclear power plant operation[J]. Journal of Nuclear Science and Technology, 2013, 50(4): 346-362. DOI: 10.1080/00223131.2013.773171.

[12] Boyd A W, Carver M B, Dixon R S. Computed and experimental product concentrations in the radiolysis of water[J]. Radiation Physics and Chemistry, 1980, 15(2-3): 177-184. DOI: 10.1016/0146-5724(80)90129-6.

[13] Bérerd N, Moncoffre N, Martinet P, et al. Influence of water radiolysis on the passive properties of 316L‐stainless steel[J]. ChemPhysChem, 2024, 25(18): e202300785. DOI: 10.1002/cphc.202300785.

[14] Sultana A, Meesungnoen J, Jay-Gerin J P. Effect of very high dose rates on the radiolysis of supercritical water at 400°C and 25 MPa[J]. Canadian Journal of Chemistry, 2022, 101(5): 284-296. DOI: 10.1139/cjc-2022-0210.

[15] Ershov B G, Gordeev A V. A model for radiolysis of water and aqueous solutions of H2, H2O2 and O2[J]. Radiation Physics and Chemistry, 2008, 77(8): 928-935. DOI: 10.1016/j.radphyschem.2007.12.005.

[16] Joseph J M, Choi B S, Yakabuskie P, et al. A combined experimental and model analysis on the effect of pH and O2 (aq) on γ-radiolytically produced H2 and H2O2[J]. Radiation Physics and Chemistry, 2008, 77(9): 1009-1020. DOI: 10.1016/j.radphyschem.2008.06.001.

[17] Elliot A J, Bartels D M. The reaction set, rate constants and g-values for the simulation of the radiolysis of light water over the range 20°C to 350°C based on information available in 2008[R]. Atomic Energy of Canada Limited, 2009. No. AECL--153-127160-450-001.

[18] Yakabuskie P A, Joseph J M, Wren J C. The effect of interfacial mass transfer on steady-state water radiolysis[J]. Radiation Physics and Chemistry, 2010, 79(7): 1009-1016. DOI: 10.1016/j.radphyschem.2010.02.001.

[19] Pálfi T, Wojnárovits L, Takács E. Calculated and measured transient product yields in pulse radiolysis of aqueous solutions: Concentration dependence[J]. Radiation Physics and Chemistry, 2010, 79(11): 1154-1158. DOI: 10.1016/j.radphyschem.2010.06.004.

[20] Karditsas P J. Water radiolysis in fusion neutron environments[J]. Fusion engineering and design, 2011, 86(9-11): 2701-2704. DOI: 10.1016/j.fusengdes.2010.12.040.

[21] Yousefi N. Gamma-radiolysis kinetics of liquid, vapour and supercritical water[M]. The University of Western Ontario (Canada), 2014.

[22] Roth O, LaVerne J A. Effect of pH on H2O2 production in the radiolysis of water[J]. The Journal of Physical Chemistry A, 2011, 115(5): 700-708. DOI: 10.1021/jp1099927.

[23] Daub K, Zhang X, Noël J J, et al. Gamma-radiation-induced corrosion of carbon steel in neutral and mildly basic water at 150°C [J]. Corrosion science, 2011, 53(1): 11-16. DOI: 10.1016/j.corsci.2010.09.048.

[24] DEY G R. Nitrogen compounds' formation in aqueous solutions under high ionizing radiation: An overview[J]. Radiation Physics and Chemistry, 2011, 80(3): 401-410. DOI: 10.1016/j.radphyschem.2010.10.010.

[25] Yakabuskie P A, Joseph J M, Stuart C R, et al. Long-term γ-radiolysis kinetics of NO3⁻ and NO2⁻ solutions[J]. The Journal of Physical Chemistry A, 2011, 115(17): 4270-4278. DOI: 10.1021/jp200262c.

[26] Guo Z, Zhang Y, Jiao L, et al. Radiolysis of aerated aqueous ammonia solutions[J]. Journal of Radioanalytical and Nuclear Chemistry, 2023, 332(12): 5059-5070. DOI: 10.1007/s10967-023-09239-y.

[27] Kumar P S, Mohan D, Chandran S, et al. Evaluation of nitrogen containing reducing agents for the corrosion control of materials relevant to nuclear reactors[J]. Materials Chemistry and Physics, 2017, 187: 18-27. DOI: 10.1016/j.matchemphys.2016.11.031.

[28] Iwamatsu K, Sundin S, LaVerne J A. Hydrogen peroxide kinetics in water radiolysis[J]. Radiation Physics and Chemistry, 2018, 145: 207-212. DOI: 10.1016/j.radphyschem.2017.11.002.