Differential Pressure Permissive

DP sounds like a poor surrogate for flammable gas concentration.

I believe this also is the attitude taken by I&E techs. This thing is ridiculously sensitive, and apparently trips at 0.12" H₂O. Heck we don't even have a calibrator in that range. Vibration, or a healthy (or unhealthy) flatus will set it off, but no one should be in compartment during light-off, or start-up steps anyway.

Agree, I believe an in-situ or extractive gas monitor would be a better measurement.

Methane has a lower flammability limit of 5% of atmosphere. So, if it were simple and homogenously mixed, roughly anything under a 5% increase in pressure would not pose a flammable hazard.

But, things aren't simple and it isn't going to mix homogenously reliably.

Depending on the pressure and temperature of the gas and the atmosphere into which it is leaking, the location and size of the leak, obstacles and partitions in the space, and the oxygen concentration, much smaller changes in pressure could correspond to a flammability hazard. Just imagine a cold small size leak low in a compartment with many pockets to trap heavy cold gas; 5% flammability limit could be reached in localized areas pretty quickly without much change in pressure if the local pocket is small compared to the entire volume.

I'm with the other commentors in the belief this is not a great tool for the intended job.

<...So, if it were simple and homogenously mixed, roughly anything under a 5% increase in pressure would not pose a flammable hazard....> Nonsense. A flammable atmosphere can have the same pressure as that on the other side of the differential pressure measurement and the instrument won't indicate there is anything wrong.

The assumption of course is that the compartment is a closed volume, with initially normal atmosphere.

If the compartment is open, and they are trying to monitor for leaks, it needs to be very sensitive and will probably still only catch sizeable ruptures under considerable pressure.

.

Then again, if timing of the indication isn't that critical, even the slowest leak can cause a large momentary change in pressure upon reaching the effective lower frammability limit

....so long as there is reliable ignition source in the vicinity.

Except to complicate matters, each compartment is under full exhaust flow as a prerequisite by top of compartment fans. I think dP was a horrible idea.

Um, yeah. I concur.

<...have an idea what that amount should be where there could be an ignition hazard inside the generator compartment?...> Ignition hazards are largely independent of minor variations in pressure. Further, just because there is a change in pressure does not necessarily mean that an ignition risk has arisen, for it could be something else. A natural variation in local barometric pressure would be typically +/- 30mBar due to changes in the weather, which would swamp the <...1mbar...> setting of such a device. So the <...lead instrument tech...> is correct, as is the concurrence.

The origin of the <...logic...> needs to be determined and the source consulted for the reasoning behind it. Once the true record of the reasoning behind the <...logic...> has been established, the matter is best addressed in a team situation between the facility's Engineers, the original equipment manufacturer and the facility's fire insurers before any action is taken. The IChemE promotes the HazOp methodology to cover proposed changes to plant. The technique is applicable to a broad range of installations and its practice is highly recommended.

Good luck with it.

<unsubscribes>

The real question is that presented a choice between meeting dispatch (ever) and never meeting dispatch, there is pressure to do something about this quickly.

I concur that discussions with insurance, engineer on site, I&E technicians (that know the system better than the engineer on site), and the OEM should already have solved this issue. I suspect there was a mix-up in units by whomever programmed the branch in PLC safety/interlocks at the OEM level.

That question should only be answered by the manufacturer. Any changes made to a safety related subsystem without the manufacturer's explicit approval will most likely result in the liability falling to your organization and its insurers.

Of course this is a point well taken, and taken seriously. What we have here is a "failure to communicate", I suspect.

If you are concerned about methane concentration, why not get manufacturers clearance to use a ZYMQ 2 or equivalent methane sensor? Since the technology is available to directly measure the gas of concern, why not use it? They are also cheap enough to scatter a few around in areas of concern and potentially act as a PM trigger for sealing systems.

Actually, I believe there is a gas sensor already in compartment if I am not mistaken.

We have never had leakage issues anyway, and that is a good thing. We operate on fairly high pressure gas.

There are thousands of historical threads on gas turbine generators on control.com

One of the gurus there might shed some light on the DP setpoint. As always, specific info on which model, generation, etc. is important.

"... each compartment is under full exhaust flow..." which means that the compartment is under negative pressure. I suspect that your DP sensor is there to ensure that the compartment is never pressurized, a condition that can occur if there is low/no flow through the fans, a seal failure allowing exhaust gas to enter the compartment, an open compartment door/panel, etc., etc.

You didn't say if this was an open or combined-cycle unit; if there is a HRSG (Heat Recovery Steam Generator) in the exhaust path then there is also the possibility of the DP trip to guard against high backpressure due blockage or fouling of the HRSG.

Another thought, it's possible that there's some form of timing circuitry that allows for a ride-through during a Texas sized fart.

Bottom line: you can't make any modifications until you understand why the "trip/permissive" is there in the first place, and how it interacts with the rest of the protection logic.

This is a combined cycle unit, so yes, there is an HRSG. Pretty sure the duct where the heat recovery tubes are located is under slight positive pressure due to jet engine exhaust.

There is a seal boot around where the exhaust flanges to the duct.

I think the vacuum pressure inside the compartment is more than this trip level, and it is a differential sensor.