The design and evaluation of a passively cooled containment for a high-rating pressurized water reactor

https://doi.org/10.1016/S0029-5493(99)00339-8Get rights and content

Abstract

An integrated pressurized water reactor (PWR) containment was conceptualized that allows heat to be rejected passively to the environment. The proposed containment is based on the demonstrated Ebasco Waterford 3 design. The secondary concrete shell was equipped with inlet and outlet vents that create an air-convection annulus. These vents also permit the submersion of the lower part of the primary containment into an external water pool. An internal water pool located at the bottom of the lower containment was added to increase in-containment heat storage. The performance of the proposed passively cooled containment was evaluated using a subdivided volume code, gothic version 3.4e; the relative novelty of subdivided volume analyses for containment performance evaluation requires experimental verification of principal code predictions. Two experiments were carried out; one to test the performance of the external moat, and one to verify the code’s ability to predict thermal-stratification inside the containment. To improve the subdivided-volume simulation of convection-related parameters, a modeling technique (boundary layer flow approximation) was devised. Finally, the behavior of the proposed containment was evaluated for the worst-case large break loss of coolant accident and the worst-case main steam line break accident. Peak pressures remained below 0.45 MPa during both transients; internal wall pressure differences, equipment qualification temperatures, pressure restoration time also remained below design limits. The mitigation capability of hydrogen recombiners was also evaluated.

Introduction

Passively cooled containments are designed with special features that ensure containment parameters remain within safety limits throughout postulated accident scenarios. For large periods of these accident transients, heat rejection and storage in passively cooled containments depends on the development of natural circulation. Therefore, design and analysis work for such transients has to deal with phenomena less explored in work for containments that employ active cooling systems, e.g. the development of thermal and concentration gradients.

This first question posed during this research work was: is passive heat rejection feasible for the containment of a high-rating; 1300 Mwe-reactor. While passively cooled containment concepts have been proposed for low- and medium-rating reactors (up to 900 MWe), no evaluations have been presented for high power ratings. The feasibility question was answered by evaluating a series of designated pressure-limiting features, first individually and then as a group integrated into a complete containment. The second question pertained to the establishment of a suitable performance-prediction method for the proposed containment concept. This involved the development of distributed parameter (3D subdivided volume) code models with special techniques devised to handle natural-convection-related parameters.

Individual features that enable passive heat rejection are discussed first. The integrated containment does not need to rely on active systems for containment cooling, and thus subscribes to the guidelines set forth for passive reactor systems. (EPRI, 1987) Passively cooled high-rating reactors are not only desirable because of economy of scale, but also because of siting constraints in areas with high population densities. Many experimental programs were devised to investigate the performance of passive containment cooling systems (PCCSs) suitable for LWR applications. A review of passive containment design and pertinent experimental programs can be found in referenced reports related to the current topic.

Reactor design criteria, such as power rating, free volume, surface area, design limits of equipment, and special heat rejection systems, determine the performance of a containment system during a design basis accident. Practical considerations (cost and constructability) also affect the selection of pressure-limiting features for a particular design. Economic considerations and constructability were largely beyond the scope of this research. However, using a demonstrated design (the Waterford 3 pressurized water reactor (PWR)) as a starting point, limits the need to evaluate cost and constructability to the proposed passive pressure-limiting features. Fig. 1 is a schematic of the proposed design.

Predicting the performance of the proposed containment concept involved a review of suitable computational tools, the selection of a tool that best suits the geometry/operating conditions combination for theproposed concept, and the verification of code predictions for this application. Table 1 lists some of the most commonly used containment analysis codes1. Most safety and licensing analyses for containments of currently operating reactors were carried out using lumped parameter codes.

Passive heat rejection that depends on in-containment flow, temperature, and concentration fields, can only be accurately predicted by distributed parameter methods (CFD-type calculations). Distributed parameter codes available for containment analyses were initially developed for specific investigations, and not necessarily the overall performance prediction in DBA scenarios. For example, the maap code was refined to handle peculiarities of flow through paths that have large temperature gradients, which affect the effective pressure drop through flow junctions. gasflow (which evolved from the hms code) was refined for the study of non-condensables convection and flame propagation in the presence of obstacles. Despite significant advances in computational power, the CFD-type mesh detail is still impractical for containment modeling. Furthermore, significant modifications/verifications of code logic have to be done before CFD-type codes can adequately simulate the gamut of phenomena important in a passively-cooled containment transient. The gothic code (version 3.4e) was used in the performance evaluation of the proposed containment. gothic has the models necessary to simulate containment systems and has been verified for a range of phenomena deemed essential to passive heat rejection from a containment.

Because heat rejection/storage features have a common heat source; the containment atmosphere, and their simultaneous operation changes the properties of that heat source, the evaluation of features must be approached in an integral manner; Fig. 2 shows the heat transfer mechanisms that are part of an integrated analysis. An analysis that covers all these mechanisms goes beyond tests included in the gothic 3.4e qualification manual (George et al., 1991b). Therefore, an input model assessment methodology was tailored for the proposed containment concept to improve code-prediction confidence.

Section snippets

Preliminary gothic models

Pressure-limiting features were examined successively using a coarsely-noded bare containment. (Gavrilas et al., 1996) Heat rejection through the steel shell was enhanced by using an air-convection annulus on the upper portion, and an external moat on the lower portion. The in-containment pressure rise was also limited by incorporating a large internal pool. The gothic 3.4e code logic was modified, where possible, to better represent these features; for example an equivalent

Performance analyses of the proposed containment concept

Tailored analyses for the proposed containment concept include: experimental verification of essential phenomena, special modeling techniques, and sensitivity studies. The moat performance was verified experimentally: the heat transfer coefficient between the shell and the moat was determined, and compared to the heat transfer coefficient calculated by the gothic code. Modeling techniques were devised to ensure that conservatism is preserved by the proper positioning of heat sinks and steam

Conclusions and recommendations for future work

This research has opened avenues for future investigation in two directions. The first is the optimization of the proposed passive containment concept for a large rating PWR, and the second is further code development for distributed parameter containment analysis. The design optimization recommendations were discussed in reference (Gavrilas et al., 1996).

The verification and performance predictions presented here were tailored to demonstrate a conceptual design. Therefore, they represent a

Acknowledgements

This research was sponsored by the Electric Power Research Institute.

References (12)

  • 10CFR50, 1993. Appendix A, Criterion 50, Containment Design Basis, Federal...
  • EPRI, December 1987. Advanced Light Water Reactor Requirements Document, The Electric Power Research Institute, Rev. 0,...
  • Gavrilas, M., Hejzlar, P., Driscoll, M.J., Todreas, N.E., October 1994. gothic Code Evaluation of Alternative Passive...
  • Gavrilas, M., July 1995. Alternative Passive Cooling Concepts for a Large Rating Pressurized Water Reactor Containment,...
  • Gavrilas, M., Todreas, N.E., Driscoll, M.J., October 1996. Containment Passive Cooling Design Concepts, Second...
  • T.L. George

    gothic Containment Analysis Technical Manual, Version 3.4e

    (July 1991)
There are more references available in the full text version of this article.

Cited by (15)

  • Three-dimensional transient analysis of coupled RCS-containment integral system

    2020, Nuclear Engineering and Design
    Citation Excerpt :

    A containment cooling system is generally installed in the containment, such as the AP1000 steel containment (Schulz, 2006) and the economic simplified boiling water reactor (ESBWR) condensation system (Hinds and Maslak, 2006). For a LOCA in the primary loop, the containment cooling system will act as the final heat sink of the accident (Tujikura et al., 2000; Kim, 2005; Byun et al., 2000; Gavrilas et al., 2000). Therefore, the importance of the thermal–hydraulic phenomenon in the containment is more obviously.

  • SGTR analysis on CPR1000 with a passive safety system

    2018, Nuclear Engineering and Design
    Citation Excerpt :

    Unlike AP1000, the APR1400 adopts a secondary passive heat removal system, which is also used to improve the CPR1000 technology (Zhang et al., 2011). As for the passive containment cooling system, Byun et al. (2000) designed a semi-passive containment cooling system for a large concrete containment and Gavrilas et al. (2000) designed a passively cooled containment for a high-rating pressurized water reactor. CPR1000 (as shown in Fig. 1) is a Chinese GEN II PWR with 3 loops.

  • Preliminary assessment of a combined passive safety system for typical 3-loop PWR CPR1000

    2017, Nuclear Engineering and Design
    Citation Excerpt :

    Some experimental researches (Wu et al., 2015) have already been conducted on this heat removal system for CPR1000. As for the passive containment cooling system, Byun et al. (2000) designed a semi-passive containment cooling system for a large concrete containment and Gavrilas et al. (2000) designed a passively cooled containment for a high-rating pressurized water reactor. Although the passive systems have become a remarkable characteristic in the design of advanced nuclear power plants, they face the risk of function failure as well.

  • Comparison of SGTR accident with two different passive residual heat removal systems

    2019, International Conference on Nuclear Engineering, Proceedings, ICONE
  • Research about the influence of passive residual heat removal system on SGTR accident

    2019, International Conference on Nuclear Engineering, Proceedings, ICONE
  • Comparison of SGTR accident for CPR1000 with two different passive residual heat removal systems

    2017, 17th International Topical Meeting on Nuclear Reactor Thermal Hydraulics, NURETH 2017
View all citing articles on Scopus
View full text