A lot of times stuff is not replicatable (suck it spellchecker, i just invented the word) because it's fucking difficult. I mean I have spent thousands of dollars and even worse wasted many hours in the lab on getting something I thought should be straightforward, obvious, and simple to work. Sometimes you want things to work so badly, you might even see things (usually fluorescence) where there is none. It's like how Percival Lowell saw canals on Mars. As a scientist you have to fight hard against your own bias, and not take it personally when someone attacks your work. Biological systems are unreliable (or not easily modeled), it's not like a computer program where everything follows a known deterministic path. In biology, the conditions in which something happens may not be known. It may work in one lab because they are using a reagent with a trace contaminant of salt whereas in another it won't work because the conditions are too pure.
So anyway, I reckon we have 3 reasons why studies are not reproducible (here they are in order of unethicalness/immorality):
1. The actual conditions are not what the researcher thinks it is. (The reagent constituents are not normal for example).
2. The researcher wants to believe a result so badly that they see an effect that doesn't exist. (Nowadays you have to photograph your results and/or use software, so this *should* get caught in peer review).
3. The research was published due to pressure to get grants combined with confidence that a particular hypothesis is real and should work -- in spite of lab failure (which the researcher ignores, telling themselves somebody in their lab made a "pipetting error").
Obviously, #3 is the most evil of the above. None of these are an excuse for publishing bad science. In terms of mitigating effects, #1 is the hardest to avoid. #3 should be very avoidable if you have scruples.