TonopahJoe
Industrial
- Aug 23, 2009
- 1
We have CNG filters for 170 Mwe combustion turbines with MWP's of 525 psig that are fed by either or both parallel pressure regulating stations, one of which takes in nominally 600 psig and the other 900 psig. The pressure set points of these two reg stations are 485 and 475 psig respectively and the regulators typically lock up at about 500 psig during no flow conditions.
The only PRV's for the filter vessels are small thermal reliefs on the tops of each filter vessel set at 525 psi. The 600/485 psi reg station has a gas powered isolation ball valve operated by a bi-stable solenoid valve that is controlled by a DCS about 2000 feet away that reads a single pressure transmitter at the reg station. It has a 15 second closure time.
The 900/475 reg station has the same control arrangement but it has a slam shut ball valve with a 2 second closure time. There is no other over pressure protection in the system.
We know from experience on the 600/485 reg station that the worker regulator fails as is on loss of control power, as it takes power to open or close the valve. Months ago the 15 second isolation valve had been inadvertently disabled at the DCS by operators in an effort to prevent unwanted CT trips, not realizing that it prevented valve closure on over pressure as well. When the worker valve's positioner lost its DC supply, the pilot PID controlled monitor upstream of it could not respond quickly enough during a load swing and with the isolation valve disabled, it could not close. So the pressure reached 545 psi at the filter vessels lifting their thermal reliefs before the monitor finally closed off.
It's been my experience at all other CNG fueled power plants I've been to that the lower pressure rated filter vessels are protected by calibrated and seal wired full port PRVs and that the isolation valves are there to shut off flow on over pressurization and not be the sole source of OPP.
It's obvious that there are numerous ways to prevent our isolation valves from closing--from operator error to DCS failure to cable failure to someone or some fault shutting off the power gas. I'm particularly concerned about the 900 psi reg stations, in spite of their 2 second closure time as they are just as vulnerable to failure as the 600 psi regs.
Does anyone know of a code requirement that mandates PRV's that I can cite to use as leverage to get real PRV's installed? Or can we "legally" modify the slam-shuts to prevent an OP using a more "fool proof" method for closing the valves on an OP?
The only PRV's for the filter vessels are small thermal reliefs on the tops of each filter vessel set at 525 psi. The 600/485 psi reg station has a gas powered isolation ball valve operated by a bi-stable solenoid valve that is controlled by a DCS about 2000 feet away that reads a single pressure transmitter at the reg station. It has a 15 second closure time.
The 900/475 reg station has the same control arrangement but it has a slam shut ball valve with a 2 second closure time. There is no other over pressure protection in the system.
We know from experience on the 600/485 reg station that the worker regulator fails as is on loss of control power, as it takes power to open or close the valve. Months ago the 15 second isolation valve had been inadvertently disabled at the DCS by operators in an effort to prevent unwanted CT trips, not realizing that it prevented valve closure on over pressure as well. When the worker valve's positioner lost its DC supply, the pilot PID controlled monitor upstream of it could not respond quickly enough during a load swing and with the isolation valve disabled, it could not close. So the pressure reached 545 psi at the filter vessels lifting their thermal reliefs before the monitor finally closed off.
It's been my experience at all other CNG fueled power plants I've been to that the lower pressure rated filter vessels are protected by calibrated and seal wired full port PRVs and that the isolation valves are there to shut off flow on over pressurization and not be the sole source of OPP.
It's obvious that there are numerous ways to prevent our isolation valves from closing--from operator error to DCS failure to cable failure to someone or some fault shutting off the power gas. I'm particularly concerned about the 900 psi reg stations, in spite of their 2 second closure time as they are just as vulnerable to failure as the 600 psi regs.
Does anyone know of a code requirement that mandates PRV's that I can cite to use as leverage to get real PRV's installed? Or can we "legally" modify the slam-shuts to prevent an OP using a more "fool proof" method for closing the valves on an OP?