In the PV analysis process, "voltage angle" is not an output variable. To get voltage angle, the PV analysis needs a full AC load flow process. It can be done using python.
PV study algorithm - Given a base case, simulate power transfer between two study zones (i.e., increase generation dispatch at zone A and scale up load in zone B) and at each incremental transfer step (within a for loop), apply contingencies one at a time (second for loop), followed by a load flow solution (use psspy API functions). The transfer level and associated contingency is flagged as a voltage collapse scenario for those scenarios where the load flow does not converge. Record monitored variables and transfer scenario parameters just before the voltage collapse is reached. The process is repeated for higher power transfer level until the base case voltage collapses under no contingency, or the source zone reaches its maximum specified exporting generation capacity.
Voltage violations and branch overloads are monitored, and while they do impose operational limitation of the study network, they may not affect the voltage collapse point. Additional care may be required when voltages below 0.80 per unit (pu) are found, because such under-voltage levels could trigger the stalling of motor loads, leading to a voltage collapse scenario, but this condition cannot be studied in a PV analysis.
Voltage collapse is reached when the load flow fails to converge within the specified error tolerance and number of iterations. Mathematically a non-convergent solution is a set of equations with no numerical solution, a singularity. Note that non-convergence could be the result of numerical instability due to cumulative error or oscillatory control actions. Only solutions that diverge are identified as a voltage collapse outcome; other non-convergent solutions should be investigated to ensure that voltage stability limits are determined only by truly divergent scenarios.