Vibration Interlock on a Recip

wild guess here

The design of the switch and delay probably depends on what the pump is feeding, how critical it is, and how long it needs before it can safely be shut down. In other words the whole system is important. i.e It may be worth losing a cooling pump if it gives time for a reactor to safely shut down.

The fact they have a spare confirms, it is a critical component to some system. Not only are they willing to lose a pump, they have prepared for it.

The delay could also be put in to ensure it does not shut off for false alarms.

One would hope that it if this were the case, the system would provide excessive storage of compressed air as short-term reserve supply, Plus an auto start of the redundant unit.

With modern controls, this should be the norm.

D*mn, it should also be able to alert repairs w/out human intervention!

Knowing a bit about Betomachine from other posts, I doubt that this is an air compressor. It is most likely a large machine compressing hydrocarbon gas of some kind.

Auto start of the spare is probable in this case...

Truly, it is a two service machine compressing RFGA in two stages on throws 1 and 3 and compressing hydrogen in a single stage on throws 2 and 4. It is therefore one of the feed compressors to the hydrogen plant and the product compressor at the same time.

Sadly, due to the sensitivity of some of the other controls around the unit, the auto start functionality is not typically used since it tends to cause issues up and down stream of the compressors.

On most equipment, so called "safety devices" Jam relays, SMC's overload protection devices, limit switches etc. are there to protect the equipment by limiting the time damage may occur. Having a backup compressor, I would rather have the equipment shut down as soon as possible. I would then switch to the backup and proceed to test, inspect and monitor the failed device to identify if it needs replacing, calibration or repairs. If a critical unit shuts down and I can quickly and safely switch over to a backup, I will in most cases be able to facilitate repairs and get the unit back on line with less downtime and less expenses. When you kick up a stone while driving your car and your oil light comes on, do you pull over right away or do you drive home to your garage and check it out later? Better to replace an oil pan then the entire engine.

If the compressor needs to run to shut down the operation correctly and safely I would consider installing a failsafe device to switch over to an air storage tank, while still shutting down the compressor, that would supply enough air to the unit for the time needed to stop.

Alarms are always good to have to tell the operators "YOU HAVE A PROBLEM!! CALL MAINTENANCE!!" I would make sure the equipment could not be started until the tech could unlock the controls and preset the equipment after assessing the situation.

Then again, maybe the switch was wired incorrectly and not detected until it was needed. The proverbial OOOPS!

Believe you me... as some of my pal's used to told me long ago which I'd always found out very funny expression from their part definetly. I may tell ya' one or two things from few headaches with such puppies before. Let it alone the 'Full Moon' effect for the moment, the vibration stuff do play some kind of foolishment into this kind of troubleshot definetly, I agree. I did saw equipments sometimes with such troubles before couples of times and apparently after monitored the conditions for some time it came down to few different circumstances of operations by the way.

Equipment installed without vibrations reducing features on framing base and piping like say, And electrical signal conditioners systems around to minimized line noises, armonics and stuff. Also good to watch out for those neutrals / returns wirings since as they become loose or make bad contacts somewhere due corrosive effects or so forget it, it dont take much to cause some difficults as well.

Allset, Lock and Tag,

MC

Steve S

Aren't the new switches still referred to as earthquake switches, but just changed from the old mechanical to electronic. I guess they could be called different things for different circumstances. We do use the electronic for protective devices. Maybe were just used to the old vernacular and stuck in our ways.

Typically, when you speak of an earthquake switch, what it means is a mechanical switch, either of ball in socket type where the vibs get high, the ball falls out and the switch goes, or a simple spring loaded arm sort of switch like the one Betomachine is using.

A piezo accel based transmitter could I suppose be called an earthquake switch, but the reliability is better, the frequency response is better, and there are no moving parts. I would call it a vibration transmitter not an earthquake switch..

I call it an earthquake switch because it takes an earthquake to set it off and then it has to be reset in the field. The switch comes into the DCS as a discrete on/off. I would prefer a 4-20 signal so I can see things and I think that Steve S. is suggesting that as well. I've also seen similar switches "calibrated" with a heavy hammer hitting the frame next to them to be sure that they will actually trip.

I checked 618 and the way I would read it is that you need something a little more transparent than a switch as it calls for continuous measurement, alarm, and shut down features.

We have put online monitoring on a couple machines in our fleet with vibration monitors on the crossheads with some success/failure as well. One machine in particular tripped quite frequently but it was in a chlorine battery offgas service and saw a lot of moisture and particulates. This led to a lot of "spurrious" trips. The other one is in a clean gas service and does fairly well.

Yep this is a mechanical switch of the type specifically excluded in API 618.

Paragraph 7.6.7.1 in the 5th edition states: The device and its mounting shall conform to API 670. Ball-and-seat or magnetic-type switches are unacceptable.

So you would say that this meets the letter but not the spirit of 618?

No, no, your device is specifically not allowed by API..on a new machine anyway...

OK, thanks for the guidance on this, I appreciate it.

We have set all of ours to trip the compressor

Thanks for the Metrix link. I am trying to decipher what I need for an axial fan setup for fan vibration and bearing condition detection. Spec calls for "shock poulse measuring bearing damage" and "fan vibration monitoring . . . to trigger . . . two levels of vibration: alert and alarm".

Any cheap advice?

You're a gentleman and a scholar. Thanks for replying to a somewhat "off-topic" question!

Ried