Nuclear fission

11.7Example Spent fuel pool in Ukraine (SSTC)

Initial analyses of severe accidents in SFP were performed using MELCOR 1.8.3 code. The code was developed for analysis of severe accidents in the reactor core; therefore its application for evaluation of SA in SFP requires to utilize special modelling techniques and assumptions. This may influence correctness of predictions made by the code thus number of case studies with different modelling options applied was performed to get consistent results.

As an example, the main results of total station blackout severe accident analyses for VVER1000/V320 SFP are provided below. To evaluate SA progression in SFP and estimate the consequences of SFP injection recovery (2 kg/s injection rate) at various SA stages the following total station blackout cases were evaluated:

• 
  no operation recovery case;
  
• 
  SFP injection recovery at a decrease of SFP level down to 3.4 m (approx.30% of fuel height is uncovered);
  
• 
  SFP injection recovery at fuel cladding temperature 1200 C;
  
• 
  SFP injection recovery at the moment of fuel collapse down to SFP bottom;
  
• 
  SFP injection recovery at MCCI start.
  

The results of analyses indicate that:

• 
  early SFP injection restoration (at SFP level of 3.4 m) allows to avoid fuel damage and to restore successful spent fuel assemblies cooling;
  
• 
  when fuel cladding temperature of 1200 C is reached, SFP injection terminates further fuel assemblies’ damage. About 50% of FA height remains undamaged. Total hydrogen production is 2.5 times lower than in no-actions case;
  
• 
  start of SFP injection after FA collapse allows to terminate further debris heat-up (due to relatively low initial debris temperature and its porosity), to establish its long term cooling and to prevent SFP liner melt-through. Total hydrogen production is 15% lower than in no-actions case;
  
• 
  MCCI cannot be terminated by late SFP injection; however concrete melt-through occurs for 1.52 times slower than in no-actions case.
  

Correspondent analysis plots are provided on Figures 10.7.–10.7..

Figure 10.7. – Water level in SFP (without operator actions – left, with operator actions – right)

Figure 10.7. – Claddings temperature in SFP (without operator actions – left, with operator actions – right)

Figure 10.7. – Cumulative hydrogen production from fuel assemblies and racks (left) and in the cavity (right)

Figure 10.7. – Debris and lower head temperature in SFP

Figure 10.7. – Damaged concrete level

Based on the analysis results it can be concluded that SFP injection restoration at various SA stages contributes to a decrease of radioactive releases, hydrogen generation and concrete structures melting speed.

Considering the limitations of code version used these results and conclusions shall not be treated as the final ones. More recent code version model with more precisely defined assumptions (including the ones on SFP fuel decay heat) is now under development to evaluate SFP SA termination possibility at the late SA stages. It shall also be noted that some of the uncertainties associated with SFP SA progression (e.g., potential criticality evaluation) are selected for further in-depth analysis to be performed in the framework SA phenomena evaluation program initiated by the Utility.

11.8EXAMPLE FROM BULGARIA (INRNE)

(It has been used Risk Engineering and the INRNE work in this field)
In Bulgaria, presently only two VVER-1000/V320 reactors (Units 5 and 6, respectively) at the Kozloduy NPP are in operation. In order, to evaluate their safety, especially with respect to the risks associated with accidents occurring in the SFP, some analyses have been carried out. These analyses have been performed by Risk Engineering ltd. by using the version 2.1 of the MELCOR code [117], [118].

The SFP is situated in the containment and are used for storing of the spent fuel (until the residual heat of the spent fuel is reduced to the admissible levels) and also for temporary storage of control rod absorbers and dummy fuel assemblies [119].

The MELCOR 2.1 model that was used for the analysis consists of the following parts [118]:

Containment model – the same of for the reactor SA analyses with slight modifications;

SFP pools model (TG21B01-TG21B06, including drainages and wet refuelling shaft);

Primary side model (reactor without internals and 4 loops which are lumped into a single loop).

The study is limited only to the SFP into containment, closed to the reactor pit.

It is developed specific guidance for fuel behavior in SFP during severe accident conditions and they are under review.

1^ This might be true for ‘spontaneous pipe breaks’ due to lower pressure in primary system. However other human induced ‘drain down’ events need to be considered.

2^ Structural failures may not only lead to loss of coolant inventory, it can also lead to loss of cooling depending on what structure fails.

3^ Failure of emergency power supplies - the failure of a mitigation system (i.e. emergency power supply system) shall be modelled separately from the Initiating event. In this case the IE for the reactor PSA as well as for the SFP PSA shall be LOOP, and the availability of the emergency power supply system shall be taken into consideration as a mitigation system in the event tree.

4^ k-eff (or k-effective) is the effective neutron multiplication factor (ratio between neutron production and neutron loss in a system containing fissile material). This factor represents the possibility for a system to undergo a sustainable fission chain reaction, in which case k-eff ≥ 1

5^ The objective is also to be able to predict FP releases for new fuel types in the reactor core.

6^ Mo is of special importance due to the formation of CsMoO₄ which prevents the formation of CsI and favors the formation of gaseous iodine.

7^ See public documents at https://gforge.irsn.fr/gf/project/passam/

8^ http://cordis.europa.eu/project/rcn/198668_en.html

9^ The first entrance criterion is a thermocouple temperature at the core outlet above 650°C. The second criterion postulates the opening of the SAMGs one hour after loss of shutdown cooling when the thermocouples are not available. The entrance to SAMGs in this study conservatively assumes that both criteria are requested while in real accidental situation only one of these two criteria is sufficient to open the SAMGs.

Report IRSN/PSN-RES-SAG 2017-00005

Technical report ASAMPSA_E/ WP40/D40.7/2017-39 volume 4 /