Speaker
Description
In Industry sizing of safety valves for two-phase flow is still a challenge. Hazard analysis to identify the worst case scenaio, mechanical and thermodynamic non-equilibrium conditions to estimate the mass flow rate and multiple critical flow conditions are among others topics that may lead to differences in sizing a safety valve of up to 1 order of magnitude. There are more than 20 models available to size a safety valve. All of them are based on ideal nozzle flow and corrected by an experimentally determined discharge coefficient.
API 520 recommend a homogeneous equilibrium flow model to conservatively estimate the mass flow rate to be discharged. Whereas ISO 4126-10 includes a method for condidering boiling delay and slip effects, which lead to much lesser valve sizes. The discharge coefficient for two-phase flow is part of a model and will not be measured.
Valve manufacturer certify only the capacitance and valve functioning under ideal laboratory conditions without inlet and outlet piping. Unfortunately, the piping and the operation conditions forces spring loaded safety valves to chutter and flatter. This phenomenon was intensively investigated during the past 10 years. Current static stability criteria will be substituted by more meaningful dynamic criteria. Serveral models are under investigation.
Within the presentation the main procedure for sizing safety valves for two-phase flow will be outlined with a focus on parameter possibily leading to large errors in practice. Basic differences of numerical and current analytical models will be presented and current activities of valve instability will be shown.
