Grid Parity is commonly used to refer to the point in time when an alternative power source becomes “cost competitive” with power delivered through the grid. You probably don’t need me to point out that this definition could be regarded as rather vague, and open to interpretation and misuse!
Grid parity is analysed by working out the levelised costs of energy (LCOE) of new power projects of a specific type (e.g. solar PV) built at different points into the future. This LCOE is generally modelled to fall for projects built further in the future, on the assumption of falling (capital) costs over time. So plotting LCOE against time shows a downward curve. To calculate grid parity now requires comparing that trend with a similar one for the future costs of energy available through the grid: on the same chart it would be where the two lines cross. (Which they will if it’s assumed that grid energy costs don’t fall at the same rate or, better still, rise).
While that sounds straightforward enough, there are obviously lots of uncertainties regarding the future costs of both the alternative power source and its grid alternative. How quickly will alternative power costs actually fall? What discount rate, analysis period and methodology is being used to calculate LCOE? What supply mix is future grid power cost being based on? In this supply, what is the future cost of fuels assumed to be? And so on… Changing the slope of each line by a small amount can result in big shifts in the resulting “grid parity” cross-over point.
It’s also important not to mix up wholesale costs with retail costs. For example the LCOE of a new power project positioned as a centralised source within the grid does not reflect the price that an end-user will pay when that power reaches their premises. Whereas if the calculation is based on a PV system attached to their rooftop, comparing the LCOE of such a system with the end-user’s retail cost of power is reasonable. That distinction matters, because the LCOE calculated for a utility-scale PV project (for example) will be lower than that of a residential-scale one; so comparing the former (a wholesale cost) with the cost of retail power to an end-user is meaningless.