This is not intended to be a complete answer, but is for your consideration. You did not say what diameter the pipe was, which matters. The larger the diameter, the more difficult the fix, if required. For rough load values, take the temperature change and calculate the change in length of pipe. Convert this strain to a force, and moment on the pump. Pump mfgs will sometimes give you allowable forces/moments combinations; API has some allowable values for API pumps as well, etc. Vessel mfgs may give you some general info over the phone if the vessel is already in service re allowable nozzle loads, otherwise, the best practice is to calculate these loads for the vessel manufacturer and include in the specifications. If you have no other alternative and must make a determination as to whether the calculated nozzle loads and subsequent nozzle/vessel stresses are adequate, there are several methods available. Review the applicable piping code such as ASME B31.1, or B31.3, etc., and vessel codes ASME Section VIII, or applicable code. Get a good understanding of primary and secondary stresses. The piping should generally be designed such that the loads are carried mostly by the piping supports. Don't forget to consider the piping displacements as the temperature changes. There are many references as indicated in the above replies that speak to generally accepted good engineering practice for pipe routing such as this example. Once you are sure that the mechanical integrity issues are satisfactorily addressed and then look at the flow issues (safety first). Further modifications may be made to improve performance, decrease life cycle cost, etc., based on published data re standared good engineering practices. As usual, there are always exceptions that usually stem from the pumped material (or fluid) properties.
