Leakage estimation of the high-pressure and high-temperature natural circulation helium loop

Highlights

• Natural circulation system is capable of removing the decay heat from the reactor.

• The leakage rate is severe, even if the joints on the construction are welded.

• The leakage dependence on the temperature is too significant to omit.

Abstract

Natural circulation loop systems are promising a reasonable option to provide the emergency decay heat removal from a nuclear reactor during an electrical power outage. The present paper deals with an experimental study on the influence of leakages on the behaviour of a single-phase natural circulation loop. The research shows the thermodynamic and hydraulic processes in a large scale natural circulation loop during several measurements. The experimental device was designed to provide heating and cooling power of 250 kW with the maximum operating temperature and pressure 520 $°$C and 7 MPa, respectively. The continuous loss of coolant is very dangerous because it can result in coolant overheating.

Introduction

One of the reactor concepts selected in the frame of the Generation IV International Forum is the helium cooled Gas Fast Reactor (GFR). ALLEGRO, the medium-power GFR demonstrator, will be the first GFR ever built (Mayer and Bentivoglio, 2015). The first commercial application of Gen. 4 is expected no sooner than 2030 (Locatelli et al., 2013). Until then, the massive research of GFRs, as well as their support subsystems is being conducted on model devices (Miletic et al., 2014). An important and often discussed topic, the safety of the nuclear reactors, is a feature closely related to the cooling systems. Natural circulation loops seem to be a reasonable option to provide emergency decay heat removal when the electric supply of a facility is compromised (Pioro et al., 2013).

The idea of a natural circulation loop (NCL) transmitting heat between the hot and cold points was introduced by Keller (1966), who provided experimental results as well as the mathematical description of such a device. Later works were oriented to NCL stability Welander (1967) in the meaning of flow-rate and pressure oscillations when working with a constant heating power. The research of NCL was further developed by the theoretical work of Zvirin (1982). The theoretical results, mostly obtained from the small scale NCL model, could not be adapted to industrial-scale NCLs without proper scaling criteria, provided first by Zuber, 1980, Heisler, 1982. Scaling laws, proposed by Zuber, are also known as the philosophy of power/volume ratio. This philosophy has some inherent flaws which surpass some specific phenomena in the NCL, such as flow instability Nayak et al. (1998). Furthermore, the main problems with the scaling laws was that they contained a number of dimensionless factors and conversion to an NCL with another geometry layout was complicated. The problems were solved in the work of Vijayan et al. (2000), who proposed new scaling laws which described the instability nature of NCLs. Later the model was expanded Vijayan et al. (2001) to the state as it is used to date. The mathematical model is capable of calculating the Reynolds number at one specific point in the loop during turbulent or laminar flow. The results were verified by several experiments (Misale, 2001, Lakshumu et al., 2016, Saha et al., 2015). These experiments were conducted on small scale experimental NCLs and the results were in conformity with the calculations. There are no records of experiments on large-scale experimental loops.

The presented research was conducted on a model of the natural circulation helium loop built by the Faculty of Mechanical Engineering STU in Bratislava (Fig. 1). The test device is the model of the cooling system, to remove the GFR decay heat. It was designed to provide cooling power of 250 kW with maximum operating temperature and pressure 520 $°$C and 7 MPa. Experiments on the large-scale model provided valuable data. However, we observed disagreement between our experimental results and theoretical models. The NCL did not achieve the desired operating pressure calculated from the initial conditions-pressure and temperature measured before heating-up of the system, which motivated us to investigate the reason for the mismatch of theory and experiment. It was found that the helium mass in the NCL continuously decreased over time. The leakage of the coolant has not been mentioned, and it is not included in any of the reported mathematical models. This can be due to three reasons. The first reason is the coolants used, which were mostly CO₂ or tap water. The second is that the measurements in small scale devices usually did not last long enough for leakages to be significant. The third reason is the design of those devices (Vijayan et al., 2001). In all small scale loops, the heat was supplied by the external heating of the pipe, as was the cooling. This method of heat input/output does not correspond to actual application, where the heating rods and pipes for cooling water have to be inserted into the NCL. Therefore, the NCL has to have openings which may not be sufficiently sealed, resulting in coolant leakage. The loss of cooling medium is hazardous because even at steady state flow, one volume unit of coolant has to transfer more and more energy from the heater, which results in increased coolant temperatures (Berka et al., 2015). It is important to find a relationship between leakages and operating parameters as the geometry didn’t change between measurements. Our premise is that the initial pressure and/or the achieved pressure at steady state have a significant effect on the leakages.