«Item 7b Severe Accidents Related Issues Preliminary Monitoring Report Report to the Federal Ministry of Agriculture, Forestry, Environment and Water ...»
The strategy of Temelín NPP should therefore prevent re-volatilization of the fission products that have already been deposited on containment and piping surfaces, which could result in case of violent air turbulence. This includes prevention of massive hydrogen deflagration and detonation after basemat melt-through and mixing of hydrogen in the containment with the air from outside atmosphere.
ETE Road Map - Preliminary Monitoring Report – Item 7b: Severe Accidents Related Issues 119 Moreover, the pressure in the containment should be reduced before the basemat meltthrough to avoid sudden air expulsion from the containment to the environment and carryover of re-volatilized fission products. If the pressure in the containment is lowered in time, it will be also possible to minimize leakages through the rooms below the containment, between the basemat and the environment. During the Prague meeting Czech specialists mentioned these issues, but no detailed information was obtained on the approach to be followed.
The development of SAM strategy addressing specifically the issue of minimization of radiological releases in case of basemat melt-through should be monitored. Further discussion of this issue is provided in section 5.6.
5.4.3 Long term over pressurization of the containment
VLI No. VLI title / description 8.7.1 What is the overpressure assumed to be the failure pressure of the containment?
Is the strength of ventilation duct valves equal to that of the containment walls?
8.7.2 Is a filtered venting system available to be used in the case of loss of heat removal capability from the containment?
8.7.3 If there is a purge system, how is it protected against leaks outside the containment?
8.7.4 Do inherent passive features – like PARs – or by active technical equipment, provide containment mixing?
8.7.1a What are the differences between the containment design at Temelín and that at typical WWER 1000 e.g. in Balakovo?
8.7.1a CEZ-ETE's presentations in April 2001 and September 2001 address a station blackout sequence, and conclude that the maximum pressure in containment is 830 kPa (0,83 MPa) and report that containment integrity is "not challenged". The Temelín containment failure curve provided in the September 2001 viewgraphs however indicates a th 5 percentile failure pressure of 0,8 MPa for the "wall-upper ring junction" failure mode.
Given that the station blackout analysis shows a 0.83 MPa maximum pressure and there is a 5% chance of containment failure at about this pressure, what is the basis for concluding that containment integrity is "not challenged"?
8.7.2a Is there a deliberate containment venting strategy (for pressure management, hydrogen concentration management, or other purposes) as part of the Temelín SAMGs, and if so, what is that strategy, what cues cause its implementation, and how does that strategy account for competing accident progression processes (i.e., how does the venting strategy consider reactor cavity melt-through to avoid melt-through under pressurised conditions)?
8.7.5 What is the physical and experimental basis for the conclusion that parallel hydrogen burning and DCH are not predicted for Temelín in case of HPME/DCH due to vessel failure at high pressure in a severe accident?
120 ETE Road Map - Preliminary Monitoring Report – Item 7b: Severe Accidents Related Issues State-of-the-art requirements and practices Large dry containments are recognized to have considerable safety margins, which provide effective protection against over pressurization both in short and long term. Nevertheless, in case of long-term failure of heat removal systems, the pressure inside the containment can grow to the values exceeding design strength limits. Therefore, several European countries have decided to install filtered venting systems, which can serve for controlled release of gases from the containment with effective retention of volatile fission products. Such filtered venting systems are installed e.g. in Sweden (FILTRA system), France (sand filters), and Germany (HEPA and charcoal filters in the venting system).
In the US the generic severe accident management guidelines developed by each NPP supplier owners group include either purging and venting, or only venting of the containment to address combustible gas control. On the basis of the industry wide commitment the NRC is not proposing to require such capabilities, but continues to view purging and /or controlled venting of all containment types to be an important combustible gas control strategy that should be considered in a plant’s SAMGs. [NRC 02, NRC 03].
The analyses performed within TACIS programme for WWER 1000 NPPs showed that the maximum pressures achieved in the course of severe accidents reach from 0,425 MPa [Schoels 02] to 0,56 MPa [TACIS 02] for LB LOCA 850 mm and to 0,515 MPa for SB LOCA 80 [Schoels 02]. In case of virtual combustion of hydrogen the peak pressure in the containment can reach up to 1,3 MPa [TACIS 02].
Current plant status The comparison of the containment cumulative failure probability functions for typical WWER 1000 [Morozov 03] and for Temelín NPP containment [POSAR] indicates that the Temelín containment has much better strength properties than the typical WWER 1000 containment considered in TACIS programme. The fifth percentile of Temelín containment cumulative failure probability (CFP) is exceeded at the pressure of 0,8 MPa (absolute), and the median CFP is reached at the pressure of 1,1 MPa (absolute). The strength of all containment elements including ventilation ducts is designed and tested to be the same as the strength of containment walls. Thus the Temelín containment can successfully withstand pressure increases that would be dangerous to other WWER 1000 NPPs. However, the calculated containment pressures in the case of blackout reach the value of 0,83 MPa, so they exceed the 5th percentile failure pressure. CEZ stated that the containment integrity “is not challenged”, but no explanation of this was provided during the meeting.
In the long term the main hazards are due to loss of heat removal and to gas generation from molten corium concrete interaction. As the concrete used in Temelín for the base layer of the containment basemat includes no carbon, there is no long term generation of CO and CO2 and therefore the rate of gas pressure increase inside the containment is much slower than in other NPPs. Czech analyses include the case of CSS failure and evaluation of the rate of containment pressure increase due to residual heating of fission products. They show more than 24 hours are available before containment pressure reaches values dangerous for containment integrity. According to NPP Temelín, it is unthinkable that the containment spray system would not be recovered, at least in one train, during such a long time. Moreover, there is a possibility of containment venting, although this is regarded as the last measure to be applied [Sykora 03].
Temelín containment has no venting system designed specifically for severe accidents, and normal ventilation system has large pipes (400 mm diameter), which are not qualified for accident conditions. Therefore, this normal ventilation system would not be used in case of severe accidents. On the other hand, there is a system qualified for 1,6 MPa, provided with filters of high efficiency (99% for aerosols), with double lines and throttling valves, designed to be used for testing containment strength, but available in severe accident conditions. According to Temelín NPP, this system will be used as filtered venting system as a last resort, in ETE Road Map - Preliminary Monitoring Report – Item 7b: Severe Accidents Related Issues 121 case when other means of pressure control fail [Sykora 03]. No details concerning venting implementation threshold and its effectiveness were given in the Prague Workshop, nor are available in open literature, but the general features of the venting strategy indicate that it can be implemented and would resolve any problems with long term pressure increase inside the Temelín containment.
The containment mixing is provided by the containment spray system, and in the case of spray failure, by PARs, which generate considerable amount of heat and thus provide local natural circulation of the air. There is no other technical equipment that could be relied on to mix the containment atmosphere under accident conditions. The mixing of atmosphere in the SG boxes with the atmosphere in the containment dome is less efficient than in large containments of PWRs. This issue has been studied within PN7 project by means of 3-D GASFLOW code. The results will be discussed in section 5.5.
Evaluation The measures listed above in conjunction with appropriate SAMG strategies are sufficient to consider the hazards of late confinement failure as being of negligible importance. The capacity of the filtration system, which is relied on in venting to reduce containment pressure, needs further attention in the context of further monitoring or in bilateral meetings between the Austrian and Czech governments.
5.5 Hydrogen Control
VLI No. VLI title / description 8.8.1 What is the capacity and layout of PARs?
8.8.2 What is the status of PAR qualification?
8.8.3 Which poisons have been considered in PAR qualification?
8.8.4 Is there a regular PAR-testing programme to avoid their deterioration with time?
8.8.5 Have been any PAR tests done under realistic conditions of severe accidents?
8.8.6 Has the risk of PAR introducing a hydrogen burn been evaluated?
What is the safe operating range? Has the consequences of hydrogen combustion ignited inside the catalytic recombiner been evaluated (missile effects)?
8.8.7 Has there been an evaluation of the form and temperature of the gas mixture at the PAR outlet performed from the standpoint of protection of safety related components?
8.8.8 Has the system of PAR management over the lifetime of the plant been developed?
8.8.9 Is the hydrogen concentration measured? How?
What is the measurement range? Are the monitors environmentally qualified?
Are the monitors classified as safety related components?
8.8.10 Is there a strategy developed for the late stage of the accident when the fraction of hydrogen is high and the steam provides inertization of the containment?
How is the stable safe state to be achieved?
8.8.11 Has the danger of deflagration to detonation transition been considered for the ventilation ducts connecting reactor cavity with the containment?
8.8.12 Has the possibility of local hydrogen concentrations exceeding average values been considered? E.g. after the core cavity door failure and local mixing of hydrogen with air?
8.8.13 Have local hydrogen deflagrations in reactor cavity been considered?
8.8.14 What are the sensors and their measurement ranges installed in Temelín NPP for SA management purposes?
122 ETE Road Map - Preliminary Monitoring Report – Item 7b: Severe Accidents Related Issues State-of-the-art requirements and practices
• Requirements US NRC and some regulatory bodies in EU countries consider hydrogen hazards in NPPs with large dry containment to be negligible, but other EU countries regulatory bodies require technical means for hydrogen control.
After more than 20 years of studies of hydrogen hazards, NRC staff has recently published the results of analyses of hydrogen control issue for severe accidents that led the NRC to the conclusion that large dry containments have a significant capacity for withstanding pressure loads associated with combustion. In Individual Plant Examination (IPE) Program [NUREG 1560] and in [NUREG 1150] it was found that H2 combustion is not a contributor to early failure for large dry containments. The conclusion was: it is not required to install a hydrogen control system in large dry containments. However, mixing was still considered necessary to prevent creation of local pockets of high hydrogen concentration.
According to the estimates quoted in Ref. [NRC 2000, p. 4-9], conditional large early release probability (CLERP) before the vessel breach is 0,1 [1/a], CLERP at vessel breach is less than 0,1 [1/a] for high pressure path and 0,1 [1/a] for low pressure path. After vessel breach CLERP is less than 0,1 [1/a]. The analyses made for Zion showed that CCFP before and at vessel breach was about 0,01 and the contribution to this low probability from hydrogen combustion was very small [NRC 2000, p. 4-10]. The results for Surry were similar to those for Zion. The NUREG-1150 study did develop uncertainty distributions and the 95th percentile for Surry was predicted to be 0,1 [1/a], and for Zion 0,05 [1/a]. The contribution of hydrogen combustion to these two estimates was again predicted to be small. This implies that even when the uncertainties are taken into account hydrogen combustion is not a major cause of containment failure before or at the time of vessel breach for large dry containments.
NRC has introduced rule changes that eliminate requirements of hydrogen control system in NPPs with large dry containments for design basis accidents. The Commission believes that accumulation of combustible gases beyond 24 hours can be managed by the licensee implementation of the severe accident management guidelines (SAMGs) and has therefore eliminated the requirements that necessitated the need for hydrogen recombiners and the backup hydrogen vent and purge systems [NRC 04].
NRC has also considered the request to eliminate the existing requirement in § 50.44(b)(2) 10CFR50 to ensure a mixed atmosphere in containment. However, NRC did not agree with this request and the final rule also retains the § 50.44(b)(2) 10CFR50 requirement that containments for all currently licensed nuclear power plants ensure a mixed atmosphere. [NRC 04] German Reaktor-Sicherheitskommission (RSK) studied the hazard of hydrogen burning under severe accident conditions for 5 years, and in 1994 published its recommendations [RSK 94] developed in co-operation with the experts from USA, France and Japan. RSK assumed that in the case of a severe accident all safety systems could fail. If later on the core cooling is recovered, water would come in contact with zirconium and it should be assumed that all zirconium present in the core would be oxidized with corresponding production of hydrogen.