9 replies to this topic

[Fr3dd]

Good Morning everyone,

I'm working in the conceptualization of the utilities and offsites of an Used Lube Oil Recovery Plant. Two of the involved units are the Nitrogen Generation System and the Plant & Instrument air system.

After an alternative study, it was decided to propose a Nitrogen Generation Plant with Pressure Swing Adsorption (PSA) technology as well as a standard Instrument air system with an air compressor and a Dryer package.

Since both systems require an air compressor, is it feasible (and reliable) to specify a common air compressor for both systems?.

What are the advantages and disadvantages of this type of configuration?.

Thanks in advance!.

[Absolute Zero]

Hi,

First of all you need to develop the basis of your project, for example you need to identify the air pressure required for your instruments, usually the instrument air system is operated at lower pressures and the pressure swing adsorption systems are operated at elevated pressures, air consumption is higher in PSA plant as compared to instrument air plants I think you should first define your pressure requirements for the PSA system the instrument air system then decide whether to use the same feed compressor of using different feed compressors for both the systems.

[ankur2061]

Fr3dd,

What you are proposing is perfectly feasible and has been done before in many plants. However you would need to preapre a feasibility checklist with the following points:

1. Consequences of common air compressor failure scenario. Does instrument air failure (control valves and any other pneumatically operated instruments) and simultaneous nitrogen generation failure effect the safe operations of the plant / unit consuming both instrument air and nitrogen?

2. Is buffer storage (instrument air and nitrogen receivers) available to ensure adequate response time for switchover to an alternative source of compressed dry air for supply as instrument air as well as to the N2 PSA plant?

3. What capacity of buffer storage is required for dry compressed at inlet of Nitrogen PSA plant?

4. Air Compressor standby philosophy for higher availability due to breakdown. For example, if I have two compressors running at 2X50% with one standby (2 working + 1 standby) I have higher availability compared to 1X100% (1 working + 1 standby).

5. Is it possible to have critical process isolation and emergency shutdown valves to be electrical motor-operated (MOVs) in order to prevent a hazardous condition due to simultaneous non-availability of instrument air and nitrogen due to failure of the common air compressor system?

These are some of the points that need to be considered and answered before you decide to go for a common air compressor system for your instrument air and N2 generation PSA plant.

Hope this helps.

Regards,
Ankur.

Edited by ankur2061, 30 November 2011 - 03:21 AM.

[Fr3dd]

Thank you both for your responses,

First, required air pressure is not that determinant since both pressure are sufficiently close (maybe a control valve will be required).

I did some research based on the considerations given by Ankur and it is actually feasible if enough previsions are taken into account in order to guarantee a safe operation of the plant and enough time of response in case of a compressor failure scenario.

Thank you both, your comments were really helpfull.

[kkala]

Another issue regarding the topic is that plant air need not be oil free or dry, at least according to local cases for fertilizer plants & refineries. On the contrary instrument air has to be oil free and dried to strict limits (see http://www.cheresou...ument-air-dryer). Similar air specification is assumed for the feed of PSA nitrogen unit.
According to my information, there are compressors (even reciprocating) able to produce oil free air, on the condition that air inspired is oil free. Their capital cost is higher.
It is also possible to remove oil (droplets) from compressed air ( http://www.processi...ressed-air-mean). This can be the (feasibility study) option in case of common compressor (s), of course if "oily" plant air is acceptable (as it is here).

Edited by kkala, 03 December 2011 - 03:13 PM.

[kkala]

...
4. Air Compressor standby philosophy for higher availability due to breakdown. For example, if I have two compressors running at 2X50% with one standby (2 working + 1 standby) I have higher availability compared to 1X100% (1 working + 1 standby)....

Having wondered in the past whether 2x100% or 3x50% installed capacity is better, following theoretical elaboration may be useful (concept similar to http://www.cheresour...ity-calculation), even though it does not concern the main topic of the thread.
1. Assume two pieces A, B, of equipment installed; one operating, the other standby (2x100% option).
Alternatively three pieces A, B, C are installed; two operating, one standby (3x50% option).
x=probability of a piece to be non operable (failure on demand or failure in operation or under maintenance, also called single failure); then probability of a piece to be operable is 1-x.
Note: If installed capital cost of one 50% piece is y, cost of (3x50%) option is 3y, versus cost of (2x50%) option being 2y2^0.6=3.03y (six tenth factor rule applied). So difference in capital cost does not seem significant.
Let (A+)=probability of piece A to be operable =1-x, (A-)=probability of A to be non operable =x. Same for B,C.
2. Probability of 100% total capacity
Option (2x100%): P=(A+)(B-)+(A-)(B+)+(A+)(B+) =2(1-x)x+(1-x)^2 =(1+x)(1-x) = 1-x^2 (note α).
Option (3x50%): P'=(A-)(B+)(C+)+(A+)(B-)(C+)+(A+)(B+)(C-)+(A+)(B+)(C+)= 3x(1-x)^2 +(1-x)^3= (1+2x)(1-x)^2.
P'/P=(1-x)(1+2x)/(1+x)= 1-2x^2/(1+x) <1, thus always P'<P. Option (2x100%) favorable.
3. Probability of 50% total capacity.
Option (2x100%): P=same as in para 2 above = (1+x)(1-x).
Option (3x50%): P'=(A-)(B+)(C+)+(A+)(B-)(C+)+(A+)(B+)(C-)+(A+)(B+)(C+)+(A+)(B-)(C-)+(A-)(B+)(C-)+(A-)(B-)(C+) = (1+2x)(1-x)^2 +3(1-x)x^2 = (1-x)(1+x+x^2) = 1-x^3 (note β).
P'/P= (1+x+x^2)/(1+x) = 1+x^2/(1+x) >1, thus always P'>P. Option (3x50%) favorable.
4. Conclusion. When usual operation is close to 100% capacity, or a bit less, best option seems the (2x100%). E.g. Boiler feed water pumps, boiler fuel oil pumps or fuel oil heater.
When usual operation is close to 50% capacity, or less, best option seems the (3x50%). E.g. fire water pumps.
This for equipment of same reliability, otherwise conclusions can be different.

Note α: verifies that this probability is actually 1-(A-)(B-).
Note β: verifies that this probability is actually 1-(A-)(B-)(C-).

Editing note, next day: Phrase in red is added, inadvertently cut in the sent post.

Edited by kkala, 04 December 2011 - 11:22 AM.

[kkala]

Referring to the previous post, attached "opavailability.xls" contains a diagram to show "Probability to cover 100% or 50% of the capacity" for a given probability x of single failure, for 2x100% and 3x50% options. The curves graphically indicate the conclusions of previous post.
x is supplementary to operational availability of a single piece of equipment (e.g. A, B, C).
Please read the previous post (by kkala) on the subject for explanations.

Attached Files

•  opavailability.xls   28KB   114 downloads

[katmar]

A question that hasn't been asked by the OP, but which I believe is very important and should be considered, is the provision of breathing air during maintenance. This air is often supplied by taking the plant or instrument air and putting it through a good oil removal and filtration system. If you need breathing air then it becomes very dangerous to have cross connections between the air system and the nitrogen system.

In any case, the nitrogen system piping should be very clearly color coded and labelled.

[breizh]

Hi ,

Very important point from Katmar . Make sure you will not have "bridge" between the networks in case some one has to use breathing equipments ! Must be separated network .

Note : lived experience !


Breizh

Edited by breizh, 06 December 2011 - 03:15 AM.

[Fr3dd]

Thank you guys for your always very illustrating responses.

I'll check the recommendation given by kkala, in this case, theorethicaly, we will operate near to 100% capacity most of the time, it could be checked if 2x100% is better than 3x50%, I guess it may be decided based in capital costs. We're proposing as control phylosophy the start of spare compressor in case of excesive pressure drop in distribution header (i.e. high air consumption), however, is it logical also specify a "lead/lag" phylosophy?.

I'll make sure that the recommendation given by katmar and remarked by breizh is taken into account.

Thank you.