In the States, FAR 121.195(b) requires you to plan to arrive at a weight which allows allows a full stop landing within 60% of the effective runway length. FAR 121.195(d) says that if runway conditions at the destination are expected to be wet or slippery, you must ensure a runway effective length that is at least 115% of the length you needed in paragraph (b). Since, presumably, the runway hasn't changed in length, this leads to a weight reduction. The change in terminology between the rules has always led to confusion. It would be much easier if para (d) simply repeated the requirement of para (b) and replaced 60% with 52%.
Operations Specification C054 used to say that the pilot could not begin an instrument approach if the reported visibility was less than RVR 4000 feet or 3/4 mile unless the runway length was equivalent to 115% of the length required by 121.195(b). This was more restrictive than the flight planning requirement, because it was effective at the time you began the approach.
C054 has been changed in recent years, at least at my airline (and since OpSpecs are essentially boilerplate, I suspect this is pretty standard). Now it says that if the forecast calls for a visibility of less than RVR 4000 feet or 3/4 mile
at the time of takeoff, then a runway length of 115% over that required in 121.195(b) must be available. This is interesting because I have never seen a forecast containing a reference to RVR. They have simply cut and pasted the original reference to basic turbojet landing minimums into a new planning requirement. In any event, now you need to consider both the wet or slippery criteria from 121.195(d) and the visibility requirement from OpSpec C054 as planning requirements; both trigger the 115% criteria.
But...you no longer need the 115% criteria to begin the approach. Nowadays, OpSpec C054 has a new requirement which says that:
If un-forecast adverse weather or failures occur, the PIC shall not begin the final approach segment of an instrument approach unless the runway length needed for landing is determined prior to approach. The runway surface composition and length, reported runway and weather conditions, AFM limitations, operational procedures, and aircraft equipment status must be considered.
We have a landing distance assessment table that essentially provides six columns separated by braking actions: Under the column titled "Dry" they provide AFM data to satisfy the requirement of 121.195(b). Under "Good" they provide AFM data to satisfy 121.195(d). Neither of these columns have a penalty for inoperative reversers, because, of course, reverse is not credited in AFM certification data for dry runways. The remaining four columns, "Good to Medium" through "Poor" provide manufacturer's advisory data for the respective braking actions. These number only allow for a 15% margin, as opposed to the 67% margin for 121.195(b) and and the 92% margin for 121.195(d).
So, as the runway conditions worsen, we just get rid of a big chunk of margin. No problem.
The old Advisory Circular for this, AC 121-195(d)-1A from 1990, states the following:
In determining safe operational runway lengths that provide for operational variables not included in type certification tests, FAR § 121.195(b) requires a runway length adequate to allow a full-stop landing (based on the type certification tests) within 60 percent variables include deterioration, of the runway atmospheric effective length surface conditions, instability of the piloting such as gusts runway. The techniques, operational tire and brake or windshear, crosswinds, others.
FAR 5 121.195(d) operations into wet or approach to touchdown, flightpath deviations, and requires an additional 15 percent runway length for slippery runways.
What we have done in response to the serious problem of landing overruns is to a) remove the requirement for a 92% margin in runway length in order to initiate a low visibility approach, and replace it with a requirement to evaluate the runway surface conditions, equipment limitations, etc., and then, b) provide non-certification data with no margins to protect against the same operational variables specified in the Advisory Circular.
I sense a serious loss of old, established knowledge in the re-writing of these requirements, but there you have it, as best as I can see. I'd certainly be interested in anyone else's interpretations. I realize this is not applicable to other parts of the world, but since we are all trying to harmonize, some of this may be bleeding across borders...