Conventional relief valve sizing usually over-estimates the relieving flow rate. This is mainly due to the fact that conventional calculations methods are very conservative, to poor estimation of relieve properties and because transient effects are not considered. Conventional methods are based on steady state simulations, considering an equilibrium condition during the relieving process. Many studies have shown that, in most cases, conventional methods were highly over-estimating relieving flows compared to dynamic methods.
Over-estimation of relieving flows can lead to overly expensive relief and flare system designs.
Using dynamic simulation one can consider transient effects like:
- effect on reboilers duty due to loss of level;
- effect on condensors duty due to reflux drum flooding;
- effect of process control and interlocks;
- effect of liquid hold-ups inside towers and vessels;
- effects due to piping hydraulics;
- effect of dynamic back-pressure of the relief valve on the relieving flow.
Once a dynamic model is built, different scenarios can easily be tested. Here a few examples:
- Partial power failure
- Total power failure
- Air instrumentation failure
- Failure in cooling