Of the information provided, a few points come to mind from my days as a reactor operator in the civilian world. A cooling tower is not needed for a 5 MW facility provided the owner has access to sufficient water for a once through cooling system-out of the river/lake/reservoir and back in-as much larger nuclear powered vessels do. We and developed nations have environmental concerns that normally prohibit that, especially for much power reactors. I don?t know if the NoK had an available unlimited water source for their facility. PU-239 is dangerous if ingested or much worse inhaled; it is first of all a heavy metal similar to lead; i.e. producing something similar to lead poisoning which is very much in the news now. PU-239 is an alpha emitter which can be stopped by the dead layer of skin; thus internally it is dangerous. If the weapons grade material is 100% pure, then there is no problem handling it provided respiratory protection is used as a precaution or by use of a glove box to prevent airborne material, but I seriously doubt if it would be 100% PU-239. If less pure it will contain other radioactive material; maybe enough that would require remote handling equipment (very specialized equipment) not just glove boxes and the warhead may be a dud when ?exploded?. The major problem would be processing the PU-239 from the rods. That requires more than glove boxes because of the other radioactive material that may be present because it wasn?t removed in the separation process and would be releasing highly penetrating gamma radiation. This all supposes you care about your workers. As to the chimney (or stack per US terminology), that is to disperse the ventilation exhaust from the containment and to monitor for radioactive releases in more modern facilities. Modern release limits don?t permit direct releases of radioactive gases. With a 5 MW facility, carbon filters and other measures can be taken to limit the release (to disguise the facility and limit airborne detection methods) while not requiring a stack assuming again you don?t care about releases to the workers and local population. Also GE BWR power reactors do not have a visible dome; instead of the dome being over the entire reactor and subsystems, it is only over the reactor within a rather large conventional looking structure. Another point is that a low enriched fuel, with a neutron source to trigger the U-235 fission process, interspersed with-in a natural uranium, U-238, matrix ?breeds? fissionable PU-239. That is what power reactors do to operate. The U-238 absorbs a neutron to form the fissionable isotope PU-239. It is a little more complicated than that but one does not to have a lot of U-235 for this. Otherwise why produce PU-239 and go through all of the processing necessary to obtain a weapons grade purity. I guess my main point is that when assessing non-Western facilities, one can not judge them on the basis of what the West/developed nations do now because government regulations and industrial codes have been brought on by public pressure and commercial experience coupled with some CYA risk assessments.
from the article: Inside Intel / Not a reactor - something far more vicious