Depends too on the required probability of success. If you have 5 van-loads of terrorists, you can attempt lower probability scenarios than if you only had one.
In the U.S., at least, we pretty much stopped embarking on building new nuclear plants after the Three Mile Island accident in 1979; the Nuclear Regulatory Commissioned issued no permits for new reactors from 1979 until 2012. While a number of new reactors went online during the 1980s and 1990s, it appears that all of them were projects which had already been approved prior to 1979.
The number of U.S. nuclear plants peaked in 1991, at 112; as of 2016, we were down to 100, with four under construction. The last nuclear plant to go online in the U.S. was Watts Bar #2, in Tennessee, which was completed in 2015; construction on that plant had begun in 1973, and that reactor is the only one completed in the U.S. since 1996.
Qwertol is correct, many new nuclear plants using new designs are under construction worldwide. Nowhere did the OP specify a US-only scenario.
There are 98 operating nuclear plants in the US, but there are 450 now operating worldwide.
About 100 new reactors are on order or planned, and 50 are now under construction, some using the latest Gen III+ design:
The basic issue with traditional civilian reactor designs is they cannot be immediately shut down from a thermal standpoint. After a scram, they will still produce up to 5% of the full thermal output for days. 5% may not sound like much but reactors are only about 33% thermally efficient, IOW for 1 gigawatt electrical, they will produce 3 gigawatts thermal. So for days after a scram shutdown they will be producing 5% * 3 gigawatts or 150 megawatts of waste heat. That is similar to the heat from 150,000 hair dryers running simultaneously for days.
During this period they require active cooling which in turn requires electrical power. Military power reactors using highly enriched uranium do not require this and can be instantly shut down to a low level.
However newer Gen III+ designs are not all “passively safe”, and in the OP scenario even a future Gen IV reactor could be compromised given sustained physical access.
http://www.westinghousenuclear.com/new-plants/ap1000-pwr/safety
The highly radioactive spent fuel pools are co-located with many current reactors. Those are not in a containment vessel and they also require active cooling.
But in that scenario the terrorists would only have caused a through d. That would destroy the plant, but not affect anything outside the plant. I wouldn’t call it a dirty bomb unless radioactive material is spread to unrelated areas. To make it a dirty bomb, they’d have to also destroy the containment shell- multiple feet of reinforced concrete-- which is not something a vanload of people could do in a few minutes.
Certainly, which is why I prefaced my post with “In the U.S., at least…”
What would make it a dirty bomb is a fault in the containment shell. The reason Fukushima released so much contaminated steam is apparently there are massive holes in the shell, caused by the heat and pressure from the meltdown. There are other reactors operating in the USA with the same faulty design.
Analogously, everyone knows RBMK-1000s will explode rather energetically if the conditions are wrong - the post Chernobyl upgrades have probably made them harder to make fail but it’s probably still readily achievable.
The control rods on one come up from the bottom…
So, the plan is: recruit enough people and weapons to win the initial shoot-out with the facility guards and hold off local law enforcement, and a couple engineers who can set charges in the right places, and sacrifice them all (since once bullets start flying at a nuclear facility, there will be such a massive response that there’s not much hope of getting away) in order to wreck one reactor and hope that the containment shell happens to have a fault so you can spread radioactive material over a random area near the reactor, probably not in a way that requires extensive off-site clean-up or remediation.
Honestly, if I had those resources I could probably find a better way of using them to create a dirty bomb.