Whether you use dynamic or static is situational, i.e. how quickly the water momentum is changed. If slowly, static with a dynamic multiplier (typically 2) is fine. If quickly, you're better served with a dynamic analysis.
The most conservative approach is to calculate surge pressure and multiply it by cross sectional area. This is usually unrealistic, however.
Instead, reduce it to a physics problem. You have a mass of fluid. You must stop this fluid. Calculate the average force required to do so in the time frame that you have. If it makes sense, multiply this average force by a factor if the mechanism is non-linear.
E.G. if your mechanism is a gate valve that closes at a linear rate, the maximum instantaneous change in cross sectional area is at the center, which is 27% higher than average, if I recall correctly.