What is best practice for design of dewatering systems on pressurised LPG storage spheres and how is this affected (if at all) if the installation is located in an extreme cold climate? Is manual or automatic draining recommended? Automatic drain systems are likely to be a more complex design.
For example, a new-build design might be fully automated and comprise a 2" nozzle with a remote-operated, accessible fire safe primary isolation valve at minimum distance followed by a dewatering pot with an interface level sensor. The interface level sensor would throttle a control valve with bypass on the dewatering pot drain line and an independent lo lo interface level sensor would trip closed the primary isolation valve and de-energise the solenoid on the downstream interface level control valve to force it closed. The trip system would also impose an output high limit of 0% on the interface level controller forcing its output to zero to avoid the valve bumping open when the trip is reset. The dewatering pot would have a hard-piped connection to flare and a hard-piped connection to a closed drain. The interface level sensor would have a local repeater indicator visible from the bypass valve. The drain line would reduce to 3/4" diameter downstream of the control valve (this serves as a restriction orifice on the drain system). It would be designed to be self-draining with no pockets and would be well-braced to minimise vibration while in operation. The number of flanges in the system would be minimised and the majority of joints would be socket-welded (screwed fittings not allowed except perhaps for instruments). Cold climate considerations (also mitigating Joule-Thomson expansion effects) might include electrical tracing on the dewatering pot drain line with a common alarm to DCS on tracing circuit failure and provision for methanol (anti-freeze) injection into the dewatering pot flare connection.