I am the gamma ray shielding...

In any system filled with water, there will be corrosion.  It's inevitable.  Water is a highly polar molecule, and so it has a tendency to dissolve/corrode almost everything.  All you can do is attempt to minimize the amount of corrosion through careful chemistry control.  Even so, you will still get some corrosion.

In the primary coolant system of a nuclear reactor, what this means is that you will have a few (hopefully not many!) metal atoms suspended in the coolant.  These metal atoms will pass through the reactor core and absorb a neutron, and sometime become radioactive, depending on the metal.

Some components in the primary coolant system are made from Stellite, a Cobalt-Chromium alloy that has excellent wear and corrosion resistance.  Unfortunately even Stellite corrodes, and it also erodes over time due to wear.  For this reason there will always be some of these Cobalt, Chromium, and Iron atoms carried through the reactor core, suspended within the primary coolant.

The biggest nuisance among these metal atoms is Cobalt.  Cobalt-59 atoms that flake or corrode off the Stellite can absorb a neutron, and become radioactive Cobalt-60.  Co-60 has a half-life of 5.3 years, and decays by emitting two gamma rays at energies of 1.17 and 1.31 Million Electron Volts.

Eventually these now-radioactive atoms will find an eddy and come to rest in a low-flow point and settle out, much like silt behind a large dam in a river.  These low-turbulence points where the radioactive atoms settle out will now emit quite a bit of gamma radiation, due to internal contamination.  This stuff has a name: CRUD.

On the submarine, there were a number valves connected to the primary coolant system which were located outside of the reactor compartment, so that they could be opened or closed if necessary without having to go near the reactor.   One such system on the ship was for emergency cooling; if all electrical power were lost, we still had to be able to remove decay heat from the reactor core.

As a result of performing the required testing on this emergency cooling system over the years, several of the valves outside of the reactor compartment had accumulated a fair amount of CRUD in them - enough to deserve posting signage and discouraging loitering.

On to the story... one day in port, some routine maintenance inside the reactor compartment was being performed, and I was the control point watch.  That job meant I that I had to control access in and out of the reactor compartment, to ensure nothing bad happened to the guy in there, and to ensure he didn't bring any nasty radioactive crap out that was clinging to his suit.

To ensure the last, I had a very sensitive detector for sensing contamination.  It looked like this:
 
This meter is a Geiger type detector, with an attached "Pancake Probe".  The probe is heavily shielded on all sides but one, so that it will directionally detect radiation.  The open side has a thin mica window, protected from puncture by a mesh screen.  The radiation can pass through the mica and into the detector, which will provide a reading on the meter.  A selector knob allows you to choose a multiplier for the level of contamination you are dealing with, to keep the needle on-scale.  It reads out in counts/minute (times the multiplier).

Geiger-Mueller detectors work by maintaining a very high voltage across a low pressure gas.  This voltage is just slightly less than would be needed to cause all the gas to ionize and begin conducting continuously like a neon lamp. 

If a gamma ray or beta particle enters the tube, and reacts with a gas molecule, it will ionize it that molecule.  Next, the very high voltage immediately accelerates the positive and negative ions, causing them to bang into other atoms, ionizing them in turn.  Eventually the entire tube becomes ionized in a massive (but very brief) cascade, and you get one audible "click".  The gas in the Geiger tube immediately resets (because the voltage is slightly too low to keep the gas ionized) and the process repeats for every radiation event inside the tube.

I was sitting at control point, and got bored.  I decided to point the pancake probe (a very sensitive instrument for detecting minute amounts of contamination) directly at the CRUD-filled valves that were emitting a modest gamma radiation field.

The instrument went from clicking every second or so, to screaming.  I had place it on the highest range while pointing pancake probe at the internally contaminated valves, even from across the room.

Out of idle curiousity, I placed my body between the pancake probe and the valves, and the meter stopped screaming.  I was absorbing the gamma radiation that was making the meter swing!  On that day I shielded the sensitive probe from 10 mr/hr worth of gamma rays.  It was a little unnerving realizing that I was absorbing so much gamma radiation, just by being in the room with it. 

Meh.  I didn't even get an image for that exposure...