So instability occurs when ELR bigger tan DALR.
Yes. That is absolute instability. Conditional instability (or conditional stability?) is when ELR is between DALR and SALR.
That means ELR cools faster tan 3 per 1000 correct?
Yes, in the case of absolute instability.
But how is that posible knowing that for air to rise has to be warmer than its surroundings.
The ELR is that of the surrounding air.
I mean if lets says ELR is 3.5 this is gonna be stable because it is cooling at a faster rate and therefore would sink again right?
The ELR is not the lapse rate of the rising parcel of air. It is the lapse rate of the airmass. It is what you get when you measure the temp of the airmass at different levels as opposed to the lapse rate you get if you track the temp of a parcel of air as it rises through the airmass.
The parcel of air that begins to rise may cool at the DALR or SALR dependent upon whether it reached its dew point temp. Whatever, if the ELR is greater than the lapse rate of the rising parcel, then the parcel will be warmer than the surrounding air at each new level it passes through by an increasing amount the further it rises. So it will keep going up.