Drake posts:
The edge of the universe expands faster than the speed of light from our frame of reference because of dark energy.
In GR [and so cosmology] if you and I are in different gravitational fields our relative time passes at different rates. Our clocks tick at different rates. So if we observe a distant phenomena, we'll measure different, say, velocities! That does not happen in special relativity[SR]. We also measure different distances because of different spacetime curvatures for each observer.
Observations and calculations of distant velocities may exceed 'c' in curved spacetime. But LOCALLY, in flat spacetime, the speed of light is always 'c'. So in SR the speed of light is always 'c'; but in GR, and cosmology in general, distant observations of lightspeed may vary. There is no 'edge to the universe' but what you likely mean is that very distant galaxies are seen to be moving away from us faster than the speed of light'. Unlike flat space [called Euclidean] where everyone has the same shortest distance [a straight line between objects] in GR each observer has a different curvature between him and any object. So distance and time is rather vague; mathematically we say the Lorentz transform [for space and times] don't work in GR.
Saying ‘space’ expands is just a quick way of stating Hubble's Law. Relativity places the limit ‘c’ on local frames of reference, but there is no limit on global frames in an expanding universe: so distant galaxies may recede at superluminal velocities using our standard models and methodologies.
In other models, such speeds may NOT result. This takes some getting used to, just like in SR it is NOT obvious that space and time vary and that LIGHT speed is the constant. Our standard model is called the FLRW model. [See Wikipedia ] and it has a LOT of agreed upon conventions so scientists can speak from a common reference.
We use the cosmological time parameter of comoving coordinates, moving with the cosmological microwave background radiation, because it's convenient mathematically. That sets one universal frame of reference convenient to all because it is everywhere. There are other time measures that could also be used.
[See the Wikipedia ‘Metric distance’ and especially the illustrations. They REALLY helped me understand the 'expansion'.
For example, recession velocity is a coordinate-dependent number...what that means is if you use a different time coordinate [or a different model metric] you get a different answer. Some are greater than c, some are not.