Richard, I know that TANK uses E.4.3-1 and E.4.3-1 to determine Ss and S1.
The point is that:
1) TANK will not use the computed values of Ss and S1 to continue the calculation when specifying a "mapped" earthquake type, with one exception.
2) TANK will not use the computed values of Ss and S1 to continue the calculation when specifying a "site-specific" earthquake type.

So, Sp is not affecting the results and this is why I'm believing that TANK seismic analysis procedures don't cover entirely what API prescribes.

Here is the only procedure I found to obtain any output variation related to changes on the Sp input value:
1) setting a "mapped" earthquake (this means that TANK will not use Sa0* and Sa*, but will try to use Ss and S1).
2) selecting an "F" terrain (TANK will not find any table reporting the values for the above said Ss and S1, so I'm assuming it will "read" Ss and S1 from the E.4.3-1 computation).
3) input the desidered Sp.

With these inputs, a change in Sp will affect the results in SOME way: I'm wondering if this is the way the program is supposed to work.

Please let me know if my explanation is not precise, sometimes I get stuck with English and the result is that what I'm trying to explain sounds confusing to me too