atw Vol. 63 (2018) | Issue 2 ı February
operation as AM measure at approx.
50 h after the accident onset (Figure 9
right). The results show a significantly
increased hydrogen concentration in
the RB annulus in case of a failure of
the RB annulus exhaust air system
(Figure 9 left).
Further, in this case the use of
the RB annulus air supply/exhaust
systems is efficient to reduce the
hydrogen concentration and prevent
the formation of combustible gas
mixtures in the annulus rooms. With
the operation of the system the hydrogen
is removed from the annulus
quickly and the hydrogen concentration
remains below 1 vol.-% for the
long term. In that case, the use of the
emergency air filtration system of the
plant is needed in addition to limit
the radionuclide releases into the
environment.
In addition, a possible alternative
method for hydrogen reduction in the
annulus was investigated assuming
the installation of a small number of
medium size PARs in the upper RB
annulus (Figure 10 right). The results
are compared with a variant calculation
with a 10 times larger design
leakage and failure of the RB annulus
exhaust air system (Figure 10 left).
The results show that already the
implementation of PARs of medium
size can significantly reduce the
hydrogen concentration in the RB
annulus and keep it well below
lower combustible limits. The hydrogen
depletion starts at approx. 40 h
(150,000 s) after the accident onset
if the concentration exceeds about
1 to 2 vol.-%. Thus, an AM concept
with the installation of some PARs in
the annulus is considered a very
efficient mitigation measure for preventing
formation of combustible gas
mixtures in the RB annulus not just in
the case presented.
3 Conclusions
The behaviour of hydrogen as well as
aerosol and noble gases released into
the reactor building annulus of a
German PWR KONVOI reference plant
resulting from increased containment
leakages under severe accident conditions
was investigated using the
GRS code COCOSYS. Two representative
and different severe accident
scenarios – the base cases – have been
selected for the analyses.
The calculation results show no
formation of combustible gas mixtures
in the RB annulus during the observation
period for the base case with
containment design leakage and
operation of RB annulus exhaust
| | Fig. 9.
H 2 concentration in the RB annulus for variant cases with 10 times larger leakages and failure of RB annulus exhaust air system (left)
and with AM measure “operation of RB annulus air supply/exhaust systems” (right).
| | Fig. 10.
H 2 concentration in the RB annulus for variant cases with 10 times larger leakages and failure of RB annulus exhaust air system (left)
and with AM measure “PARs in the RB annulus” (right).
air system. It was identified that in
this case separate annulus rooms are
isolated at an early stage by the automatic
closing of fire protection doors,
thus preventing a further increase in
the hydrogen concentration in these
rooms.
In contrast, the variant calculation
with a 10 times larger containment
design leakage leads to formation of
combustible mixtures in the upper RB
annulus area. In this case, the RB
annulus exhaust air system is not
efficient enough to prevent formation
of combustible gas mixtures in the
upper RB annulus area.
Further, the variant calculation
assuming melt relocation into the
containment sump demonstrated that
the corium spreading into the sump
results in a higher steam generation,
which leads to a faster long-term
containment pressurization. After the
melt relocation into the sump, the
corium solidifies within a short time
and the generation of combustible
gases (H 2 and CO) coming from
MCCI is terminated. As a result, the
H 2 concentrations in the containment
as well as in the RB annulus are
significantly lower compared to those
in the case without melt relocation. In
this case, the lower combustible limit
of 4 vol.% in the RB annulus is no
longer reached.
Moreover, the results of the two
analyzed severe accident scenarios
(MBL and ND*) were compared in
order to investigate their effect on
the accident consequences. From the
comparison it was identified that in
the ND* base case, the filtered
containment venting starts about
16 hours earlier than in the MBL base
case. As a result, the maximum hydrogen
concentration in the RB annulus,
calculated for the ND* base case, is
lower than that in the MBL base case.
The comparison showed that in the
ND* base case the hydrogen concentration
does not exceed the lower
combustible limit of 4 vol.% until the
beginning of the containment depressurization.
Within the scope of the project, the
efficiency of different AM measures
for mitigation of accident consequences
in the reactor building annulus
was analyzed. The assessment
results show that the operation of RB
annulus air supply/suction system
significantly reduces the hydrogen
concentration and prevents formation
of combustible gas mixtures in RB
annulus. Therefore, the use of these
ventilation systems is considered as a
very promising accident management
measure for reducing the hydrogen
concentration in the reactor building
annulus. However, in that case the
ENVIRONMENT AND SAFETY 89
Environment and Safety
Investigation of Conditions Inside the Reactor Building Annulus of a PWR Plant of KONVOI Type in Case of Severe Accidents with Increased Containment Leakages ı Ivan Bakalov and Martin Sonnenkalb