Electric potential energy is similar to gravitational potential energy, except it can be positive *or* negative since the electric force can repel as well as attract. Electric potential energy is given by the formula

`E_"el" = k(q_1q_2)/r`.

From this we can derive a formula for a change in electric potential energy:

`Delta E_"el" = kq_1q_2(1/r_2 - 1/r_1)`.

*Electric pontential* is, confusingly, not the same as electric potential energy. While the latter is measured in joules, electric potential is measured in joules per coulomb. It represents the `E_"el"` that a unitary point charge would have at a particular point. It is denoted with `V` and calculated with

`V = E_"el"/q_2 = (kq_1)/r`.

`V` is to `E_"el"` as `vec epsilon` is to `F_"el"`. They differ only by an `r` in the denominator:

`|vec epsilon| = V/r`.

Most of the time, we talk not about electric potential but about electric potential *difference*, also known as *voltage*:

`Delta V = kq(1/r_2 - 1/r_1)`.

In general, we can relate work to voltage with

`W = Delta E_"el" = q_2Delta V`.

For parallel plates, we have a more specific equation. The work done by the electric force to move a charge `q_2` from one plate to the other is given by

`W = Delta E_"el" = q_2|vec epsilon|d`,

where `d` is the distance between the plates in metres (m). Why were we able to replace `Delta V` by `|vec epsilon|d`? Because, for parallel plates,

`|vec epsilon| = (Delta V)/d`.

This is *not* true for point charges. It looks similar to the general equation `|vec epsilon| = V//r` mentioned above, but that `Delta ` makes a big difference. With point charges, the amount of work done is strictly related to the values of `r_1` and `r_2`. Not so with parallel plates—since the electric field is uniform, it doesn’t matter where the particle is between the plates.