And it is the result of a test measurement under a specific set of conditions.
We must remember the purpose is to find the transformer equivalent series impedance. To do that, we rig a test that gives us a voltage value that is numerically equivalent to the percent impedance of the transformer.
Kind of like rigging an Archimedes' lever type test where the number of cars on the lever equals a person's mass. We could say the count is given in "mass cars" to indicate how the test was conducted, but what we really mean is the person's mass. We could use some method to calculate the "mass cars" but since "mass cars" is a measurement tool, we really are just calculating the person's mass.
So using that moniker, we CAN calculate the impedance voltage. More correctly, we can calculate the equivalent series impedance (at least reasonably close). You can special-order a transformer with an impedance specification and the manufacturer will build one pretty close. The impedance voltage test confirms the design and gives you the precise value.
The problem with the calculation is that we need a bunch of design criteria about windings, spacing, materials, thickness, grain, orientation, stacking, dimensions, clearances, housings, and I can't remember what all but it is a long list. If the OP does not have the %Z, then it is doubtful the OP will have the design and material specifics to calculate the %Z either.
What the OP may be able to make use of is an impedance range for typical transformers of the same type as the transformer in question.