Sizing Non-critical Plant Air Receiver Buffer Vessels 1

[Sawsan311]

Dear All,

In some application we deploy dedicated plant air supply from air compressor packages for use as utility cleaning stations such as purging equipment. This non-critical supply can be fed directly after the air compressor and upstream the dryer package. In this case, is there any requirement to consider sizing this buffer plant air receiver vessel for the time gap from the moment one compressor stops until another one starts (loading/unloading) i.e. 5 minutes?
Please note that another dedicated instrument air receiver vessel located downstream the dryer package which is sized for the 20 minutes for peak design dry instrument air demand (critical).
I believe the final critical receiver vessel is adequate to absorb the pressure fluctuations in case of loading and unloading and there is no requirement to oversize the plant air receiver vessel for one compressor capacity loading/unloading for 5 minutes.

Appreciate your design views
Regards,

 

[LittleInch]

Up to you really.

How important is it to your operations to perhaps run out of compressed air or only get a reduced pressure / flow in the 5 minutes(?) between compressor unload and load.

how often does that happen or what are the relative flows and capacities here?

Is the compressor set(s) under or oversized for your air requirments?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Sawsan311]

Each compressor is as high as 50% of the total load which is considerable, maintaining minimum pressure supply to critical consumers is evidently captured by the sizing of the IA dry air receiver .

In case the operational intent is to maintain steady constant minimum disturbance to the IA supply header pressure, isn't the pressure fluctuation in IA supply header gets automatically absorbed by the IA main receiver vessel downstream the dryer or the design preference is to mitigate this fluctuation upstream the dryers to overcome the pressure drop across the beds.

I am discussing from design aspects.

 

[Heaviside1925]

OP,
I would also consider recover time of the IA receiver and any flow pulled away for non-critical use would increase the recovery time of the IA receiver. Are there any critical max /large flow and subsequent recover time scenarios that need to be considered?

 

[pierreick]

Hi,
Consider this PPT presentation to support your query.
Pierre

 

[georgeverghese]

There is usually a self operated backpressure PCV on the utility air line upstream of UA receiver that would close when pressure upstream (in this case air to dryer) drops below PCV setpoint. By this time, the lag compressor would have auto started and loaded up within say 2minutes max. However, if it doesnt, and the operator runs out of the main control room on seeing the PAL on the DCS alarm panel, till the time he manually starts up the lag machine could well be 5minutes..? Note that the setpoint of this backpressure PCV is LOWER than autostart trigger for the lag machine.

 

[Sawsan311]

Thanks all for your great inputs!

From what is discussed above and as georgeverghese advised, the automatic start-up of the standby compressor upon the loss of the air supply pressure (at a set point higher than the PAL acting on the PCVs closing the less priority users, accommodating the buffer time for load/unload would only come into picture when manual switching among compressors applies and for the case of non-automatic switching over of compressors as in diesel generated engine unlike electric driven motors.

 

[georgeverghese]

@sawsan, the scenario painted here IS for the case where the air compressors are configured for autostart of the lag machine. There can be many reasons why a lag machine may not have started up or load up when expected:
a) Operator has left the field auto/manual mode selector switch for the lag machine at manual
b) Lag machine discharge valve left closed
c) Lag machine starts but fails to load
amongst some others - the full list of possibilities may be apparent when you go through the many start inhibit interlocks you'll find in the compressor technical manuals.

 

[Sawsan311]

Agreed George, the main intent of our discussion is whether we need to provide exclusive buffer vessel downstream the air compressor which would pass full demand and handle 1-2 minutes load/unload time gap. Attached snap from presentation sent by Pierre shows plant air receiver upstream the dryer package which refers to the same query in this venue whether we should have the receiver vessel located directly downstream the air compressor to exclusively handle the loading/unloading time gap or we can take credit of the fact that the auto-start happens very quickly that minimum disturbance will take place in the network and hence we need not to install a full flow receiver upstream the dryer.

Regards,

 

[LittleInch]

Without any knowledge of the air supply rates and how they vary over time no one can really form a view here.

If you have periods of low flow and hence the compressors unload and / or the motor stops followed by a sudden high demand that lowers the air pressure in your main receiver too low before the compressor starts / loads then you might want to have a big volume to avoid disruption to the air supply or loss of air to non essential consumers. Air in a plant is a utility. Personnel just want it to be avialable at all times and not stop for a few minutes when they do want it or run out of pressure. Your critical instruments will be ok as they have min 20 minutes and probably a lot more as this would be sized on a high instrument air demand which likely won't happen.

But if the rate of change of air demand in slow then it becomes an economic and maintenance issue about having a air receiver which is too big.

Hence there is no one answer - It's called design....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Would suggest that UA receiver be sized for 5minutes usage of the most frequent user of UA, not for the largest.

 

[pierreick]

Hi,
another resource.
I've seen quite often additional buffer tanks at different location to overcome potential issue on air supply (hieratic).
Reference: Projects in SE Asia and China.
my view.
Pierre

 

[Sawsan311]

LittleInch, thanks. There is evidently no doubt in the need to have a sufficiently larger receiver vessel that accommodates any downtime or cut-off int he supply to the critical IA users besides the prioritization shifting from the less-critical to the more electrical loads using pressure control. In fact this receiver buffer vessel is typically sized for 20 minutes of the design air demand, yet it is located downstream the dryer unit.

The discussion here is whether another additional buffer vessel [u]is required directly downstream the air compressor to be sized for the load/unload time gap roughly 2 min-5 min upstream the dryer[/u] IN ADDITION to the IA receiver vessel mentioned above. This is what the discussion is about.

 

[LittleInch]

Is it?

Perhaps a block diagram will help?

Sounds to me like you have what I will call Plant air coming off the compressor so wet, dirty air?

If your compressor isn't running this supply dies correct?
Your clean dry instrument air is ok as the inlet to this vessel will either have an NRV or will close if it senses loss of air in reverse.

So tis still up to you. Do you install a wet air receiver u/s the drier to power your plant air system or don't you.

No one here will tell you what is good for you.
I am a bit surprised you don't have a wet air receiver though?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MJCronin]

Only 12 -17 more questions before the OP finally gives us either a block-diagram or PID !!

Stay the course, my friends .... it will only be a couple of more weeks !!!

MJCronin
Sr. Process Engineer

 

[LittleInch]

We're going round in circles.

I refer you to my first answer.

And don't diss MJC. He's a bit grumpy at times and doesn't like people with MBAs, but if you don't like what he says just ignore him.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.