The spent fuel (or partially spent) is still rather radioactive and gives off decay heat, even in a cold shutdown. This is managed by a continuous flow of coolant.
The following is a dissertation from JT Claassen,
Pressurised water reactor cold shutdown transient analysis"
https://repository.nwu.ac.za/bitstream/handle/10394/19156/Claassen_JT_2016.pdf?sequence=1
It simulates the failure of the residual heat removal system of a Westinghouse CPR 1000 (A PWR, of which the VVER series is a type of)
"
In reference to graph 4.5 below, where a RHR (Residual Heat Removal) failure was injected at time t=0 we note that there is a
very gradual increase in the average core coolant temperature starting at 100.38C and rising to
205.75C in slightly over 9 hours. More importantly the fuel peak temperature also shows a gradual
increase in roughly 5 hours, tracking at the same rate of rise as the coolant temperature. However
after 310 minutes the peak fuel temperature increases from roughly 151.87C to 2211.17 C over a
period of 2hours and 27 minutes, or at a rate of 14 degrees per minute. Thus if corrective action has
not been forthcoming at this juncture, rapidly escalating events going forward is very likely to negate
any mitigating actions taken.
"
It should be noted that a fuel cladding temperature of 2000 is sufficient for significant Zr oxidation, which releases hydrogen gas.
So, as one can see, the failure of the cooling system of a reactor, even one in cold shutdown, proves catastrophic.
That simulation presumes no attempt is made to provide cooling via low pressure coolant lines, which could happen if engineers were forced to leave due to shelling.