Reciprocating compressors or centrifugal compressors? 2

[Sabergg1981]

Hi Guys
I stated before my question about designing a field compressor station to boost pressure of a gas field upstream a gas refinery. The condition of our reservoir varies in years. The Compressor Suction pressure changes from 70 barg to 25 barg after 5 years and then remain constant for 10 years at 25 barg. discharge pressure shall be constant 91 barg and also our capacity falls down from 70 MMSCMD to 4 MMSCMD after 15 years. our designers insist to use LP&HP centrifugal compressors instead of reciprocating compressors. But as you konw these type of compressors are sensitive to suction condition but they believe not much sensitive !!! and also say recip compressors are not reliable for continuous operation and using VFD drivers or gas turbine and also minimum turn down ratio of 30% and more we can handle variation in suction condition. they also believe recip manufacturer dont recommend using reciprocating compressors over 4MW for continuous service.
i want ask you to state your approach and your type selection
i want you if you know any gas field reciprocating compressor station ( Name, Location,...)

Thanks

 

[TD2K]

Recips are higher maintenance but I would definitely be looking at them for this configuration. You have a major increase in the compression ratio required and a significant decrease in the flow, those to me would eliminate centrifugals. Towards later in the field's life you will be recycling a significant amount of flow with an attendant cost and potentially a changing MW with its effect on centrifugals.

Have your designers got any preliminary information from vendors showing how an HP/LP configuration could be make to work with actual curves and numbers, not just a salesperson spiel?

 

[Sabergg1981]

@ TD2K

LP compressors gonna to increase pressure 25 barg to 45 barg and HP compressors gonna 45 barg to 91 barg not considering losses and i must mention no vendors have a talk on performance maps of these compressors before purchase. i want to know what happen for turn down ratio of 2 centrifugal compressors working in parallel?

 

[Montemayor]

As TD2K infers, this is a "no-brainer" type of application. Obviously your "designers" were not even born (nor are they aware) when all natural gas field producers used reciprocating compressors. And this was done successfully in the USA through 2 major world war production efforts - without any centrifugal compressors, gas turbines, PLCs, laptop computers, or software. All we used was a slide rule. And unlike today, we did our projects on time, under budget, and with performances as predicted.

I do not agree with TD2K on his opinion of maintenance. Reciprocating compressors do not require higher maintenance in my opinion - IF you are experienced and know what you are doing with them. The problem today is that probably there isn't a graduating engineer that can take a recip apart and put it back together. This, I learned very early on to do in the field. Centrifugal compressors have their place in industry, but what you describe is clearly not one of those places. Thank god for reciprocating compressors; otherwise we wouldn't be able to vary capacities, pressures, or gas compositions - all areas where a centrifugal fails miserably --- and with a LOT of expensive maintenance (by outside, specialized contractors)!

 

[zdas04]

Lucky I was on the road today, I'd have been rude, where TD2k and Montmayor are trying to be helpful. Doing raw gas compression with a centrifugal should only be considered when there is zero options. The only place I see this is off shore where: (1) the energy density of a centrifugal paired to a gas turbine is very high (minimizing the deck space required); and (2) the deck space is worth more than the tail-end reserves so they are willing to abandon significant gas in place to keep from having to set less energy-dense equipment. If the ground you are considering for these compressors is on 5th Ave in Manhattan, your designers have made an excellent choice. If it is anywhere else onshore your designers are lacking in a basic understanding of equipment, economics, and equipment maintenance (centrifugal compressors generally require one full time factory tech rep per 3 compressors)--in other words they are acting like fools. Turndown ratios, flexibility, initial cost, and ongoing costs all favor recips for this application.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[Sabergg1981]

@ Zdas04

You were very helpful to me and guide me but as you convincing fool persons is hardest thing in the world. as i mentioned above they claimed reciprocating compressors are not reliable in continuous service and also they didn't recommend above 4 MW in power for continuous service. as a comparison i need to show our designers an upstream compression station in the world by name ti prove my words and show them they are in the wrong side. they believe variable condition of suction easily controlled by VFD engine and using LP&HP scheme and thrn down ratio. but i investigate turn down ratio is usually about near 50% and manufacturer doesn't guarantee the operation range of compressors but rated point. this field is onshore with lots of space.

 

[zdas04]

4MW is 5300 hp. In the San Juan Basin of Northern New Mexico and Southern Colorado, Williams Field services set something like 150,0000 hp consisting of various stations with 2-stage Ariel JGK/4 recips driven by Waukesha 7042 GL engines (about 1200 hp/compressor). Most stations had 10-15 compressors in parallel. These machines have been running in this basin on raw field gas since 1989.

I visited several facilities in East Kalimantan, Indonesia that each had a number of recips in the size you are talking about that have been running on raw field gas for decades. I'm currently working on a project in Queensland, Australia where all the raw gas stations are big hp recips.

A Cat 3616 is in your hp range, and Ariel JGD/4 will accept that much hp. Also Ariel has a number of frames that mate to even larger hp drivers.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[Sabergg1981]

@ Zdas04

Thanks about useful information you gave me. I am going to prepare a document about the reason of using reciprocating compressors than centrifugal one in the gas reservoir applications. i listed parameters below:
1) Maximum Efficiency of reciprocating compressors
2) Flexibility of recip than centrifugals to variation in suction
3) Lower Capital cost of recip to HP&LP configuration
4) Dependence of range of operation in centrifugal suction to variation in suction
5) Not using VFD driver to control

can you endorse my approach and if you can add to my list other parameters you think may be helpful.

Thank you

 

[zdas04]

I would drop #5, I use variable speed on recips all the time.

Design point on a centrifugal is usually 1.8-2.5 compression ratios/stage. For a recip it is usually 4-5 compression ratios/stage. Recips are not "rubber compressors", but they are considerably more flexible than centrifugals. I generally design recipst to deliver about +/-5% of design suction pressure (i.e., a 70 barg machine should have its suction between 66.5 barg and 73.5 barg). For a centrifugal I want suction pressure to be within 2% of design so the 70 barg would have a suction range of 68.6 and 71.4 barg. The centrifugal is "worse" (i.e., requires more precise control), but not enough worse to drive a decision. For your application I would design recip skids with one slash two stages. 70 barg into 91 barg is an easy single stage recip application with 1.3 compression ratios. 25 barg to 91 barg is still an easy single stage application, but the transition from 70 barg to 25 barg would require a couple of valve changes and clearance adjustments to get the stiffness right. If the field decides they really need 6 barg suction (which is nearly a certainty over that time frame) then you can rotate a few spectacle blinds and turn it into a two stage application. You can do the same thing with a centrifugal, but you must have designed some really unusual interstage cooling.

I see a lot of stations with one big compressor and a lot of stations with a series of smaller compressors. I find that with a series of compressors: (1) a failure of one machine does not shut the field down; and (2) if I have 10,000 hp installed and find I need 12,000 hp, I can add two more 1,000 hp machines, or if I find I need 8,000 hp I can turn two off (or remove them if it is a long term change). With one big machine, if I have to change oil then the field is shut in. To me this is the main reason why modular recip stations out-perform monolithic recip or centrifugal stations every single time.

An example. In several federal units in the Fruitland CBM "fairway" in the San Juan Basin, one operator (call them "Standalone") was a major working interest owner. At the Central Delivery Points all operators except for Standalone contracted with Williams Field Services to collect the gas and Williams used the nominal 1,000 hp compressors mentioned above. Standalone installed one big compressor at each station for their share. Every day, Standalone was supposed to take their portion of the gas and Williams was supposed to take everyone else's. Taking more or less than your share created an imbalance that could be made up at a time agreeable to all parties. At 10 years this imbalance was on the order of 30% of cumulative production--due entirely to the fact that Standalone had frequent compressor outages that took their gas to zero. Williams also has outages, but on a 10 compressor site, the outage caused the production to drop to 90% (actually, it usually didn't drop at all as there was a little room in the other 9 compressors to make up the slack most of the time) instead of 0.

This modularity is just not feasible with centrifugal machines, the cost/hp is just prohibitive at small hp.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.