The separation between coolant and moderator in CANDU reactors provides a backup safety system through the moderator cooling system. The moderator cooling system is used to remove heat deposited in the moderator during normal operation due to gamma and neutron heating. This heat, which is approximately 5% of reactor thermal power, is of the same order of magnitude as decay heat shortly following reactor shutdown; therefore, this system can be used to remove decay heat if the emergency core cooling (ECC) system fails.

The moderator cooling system used in existing CANDU reactors uses a pumped loop (see Figure X-2), and its role as a backup safety system can be significantly enhanced if a natural circulation loop is used. A passive moderator cooling system (PMCS) concept for CANDU reactors has been under investigation for the past several years and is described in [4]. In the PMCS concept, the moderator operates close to saturation so that two-phase flow is generated by flashing in a riser that is connected to a heat exchanger (see Figure X-3).

FIG. X-2. Current CANDU moderator cooling system.

The passive loop design in Figure X-3 ensures the following: 1) single-phase exists in the calandria (for effective neutron moderation), and 2) two-phase exists in the riser so that sufficiently high flowrates are achieved for decay heat removal following certain postulated accident scenarios. The stability of the PMCS loop was verified experimentally and compared to code predictions using a scaled loop [4]. For the CANDU-SCWR, the objective is to operate the passive moderator loop during normal operation to ensure that this loop is available following accidents without operator intervention. This is made possible by the design of the CANDU-SCWR fuel channels, which is different from the design used in existing CANDU Reactors. Existing CANDU reactors use a fuel channel design that consists of a pressure tube that is insulated from the cool heavy water moderator by a gas gap and a calandria tube (Figure X-4). In this configuration, the pressure tube operates close to the coolant temperature while the calandria tube operates close to the moderator temperature.

During normal operation, heat is deposited directly in the moderator by direct gamma and neutron heating (~5% of the total thermal power). Heat transfer from the coolant to the moderator is negligible because of the presence of the gas gap between the pressure tube and the calandria tube. The moderator heat load is of the same order of magnitude as the decay heat shortly after reactor shutdown. This makes the moderator an attractive heat sink for decay heat removal following certain postulated accident scenarios provided that heat can be efficiently transferred from the fuel channel to the moderator.

Existing CANDU reactors rely on pumps to remove heat deposited in the moderator during normal operation, and to remove decay heat during certain accident scenarios. Furthermore, existing CANDU reactors require the moderator to operate with a certain degree of subcooling to avoid film boiling on the calandria tube if the pressure tube balloons into contact with the calandria tube following a loss of coolant accident (LOCA) combined with loss of Class IV power [5]. This restriction on the moderator temperature has to be removed in order to implement the passive moderator cooling loop during normal operation. This is possible by the use of an alternate fuel channel design for the CANDU- SCWR, which is required to protect the pressure tube from the high temperature coolant. This alternate fuel channel design is shown in Figure X-5 and utilizes an internally insulated pressure tube.

FIG. X-5. CANDUSCWCR fuel channel.

The main differences between the fuel channel design shown in Figure X-4 and that shown in Figure

VIII — 5 are: 1) the insulator in Fig. X-5 is on the inside of the pressure tube to keep the pressure tube cooler (close to the moderator temperature), and 2) the calandria tube is eliminated. Elimination of the calandria tube removes the subcooling restriction on the moderator because this restriction is mandated by the design shown in Figure X-4 where the pressure tube could balloon into contact with the calandria tube following certain accident scenarios [5]. Therefore, the moderator could operate close to saturation and the PMCS concept can be used during normal reactor operation. This results in two important advantages: 1) the PMCS is demonstrated to be functional at all times, and 2) there is a potential for cost reduction because of the elimination of the moderator pumps.