You'll have to be a whole lot clearer about what you are asking about. "Radiation seep" isn't a defined technical term. Radon seeps up out of the ground. Radioactive (contaminated) water seeps from containment into the ocean. Radionuclei seep into ground water. Even assuming you meant radioactive fallout, you haven't provided enough details of what type of radioactivity needs to "seep" down into their bunker. The deepest water I know of was found at a depth of 100 km. At 100 km, you'd expect the temperature to be 2500°C or 4500°F. In other words, to get below where water might possibly "seep", you'd have to get to temperatures above what we could live at. (Temps increase about 25°C per km, which means we're not going to be able to live below 2 km. Perhaps you should reframe your question since it is the dose of radiation which causes damage and dose depends on concentration, type, and energy of the radiation. BTW as you get deeper into the Earth, radon levels increase. Anyway, if we assume very long lived radioactive isotopes, with very energetic (lethal, damaging) decay modes, then the question naturally arises whether such isotopes will be transported downward at a sufficient rate to be a hazard to some sub-surface habitat. Of course, the answer is it depends on the geology (and hydrology) at that specific location. There are several minerals which are natural barriers to water. Salt is one, another is certain types of clay. I suggest you take a look at the proposed design of the Yucca Mountain repository (defunct) because they attempted to design it to prevent water "seeping" in (and out). It would be a pretty good start for a design for your habitat's "shield". You can't really build "a wall" (meaning thin (several feet, say) (sorry Mr. Trump)) to keep the water out. The barrier needs to be of sufficient thickness so that even if a earthquake moves a fault by several meters, the barrier is still sufficient to prevent significant seepage at the "crack". That's really not too difficult to do, lets just say 20 meters of clay and 10 meters of salt. The trick is to keep this ceiling compressed so that cracks don't develop, (salt is self-repairing). You'd have to be careful in digging the tunnels to live in. Caves might not be the answer. You want to keep the barrier under compression and caves, even if the rock ceiling is strong enough to support the material above it, could cause decompression (as well as other types of movement). You could ask a question on the Engineering on a different forum. Anyway, I suspect tunnels rather than big caves would be safer, but it's pretty much a wild-assed guess. FWIW

You'll have to be a whole lot clearer about what you are asking about. "Radiation seep" isn't a defined technical term. Radon seeps up out of the ground. Radioactive (contaminated) water seeps from containment into the ocean. Radionuclei seep into ground water. 

Even assuming you meant radioactive fallout, you haven't provided enough details of what type of radioactivity needs to "seep" down into their bunker. The deepest water I know of was found at a depth of 100 km. At 100 km, you'd expect the temperature to be 2500°C or 4500°F. 

In other words, to get below where water might possibly "seep", you'd have to get to temperatures above what we could live at. (Temps increase about 25°C per km, which means we're not going to be able to live below 2 km. Perhaps you should reframe your question since it is the dose of radiation which causes damage and dose depends on concentration, type, and energy of the radiation.

By the way as you get deeper into the Earth, radon levels increase. Anyway, if we assume very long lived radioactive isotopes, with very energetic (lethal, damaging) decay modes, then the question naturally arises whether such isotopes will be transported downward at a sufficient rate to be a hazard to some sub-surface habitat. 

Of course, the answer is it depends on the geology (and hydrology) at that specific location. There are several minerals which are natural barriers to water. Salt is one, another is certain types of clay. I suggest you take a look at the proposed design of the Yucca Mountain repository (defunct) because they attempted to design it to prevent water "seeping" in (and out). It would be a pretty good start for a design for your habitat's "shield". 

You can't really build "a wall" (meaning thin (several feet, say) (sorry Mr. Trump)) to keep the water out. The barrier needs to be of sufficient thickness so that even if a earthquake moves a fault by several meters, the barrier is still sufficient to prevent significant seepage at the "crack". That's really not too difficult to do, lets just say 20 meters of clay and 10 meters of salt. 

The trick is to keep this ceiling compressed so that cracks don't develop, (salt is self-repairing). You'd have to be careful in digging the tunnels to live in. Caves might not be the answer. You want to keep the barrier under compression and caves, even if the rock ceiling is strong enough to support the material above it, could cause decompression (as well as other types of movement). You could ask a question on the Engineering on a different forum. Anyway, I suspect tunnels rather than big caves would be safer, but it's pretty much a wild-assed guess. For what it's worth.