Smelter off-gas: compressor technology selection (Centrifugal vs screw) 2

[BernardE]

Good day

This is my first post to the forum.

I'm working on a project where are taking smelter off-gas to a fermentation plant for ethanol production. The off-gas has some moisture and dust in it (we are looking at removing the moisture and dust in a knock-out pot, expect to have around 20 mg/Nm3, max size 30 micron dust and no droplets after knock-out), and the composition is slightly variable (±15% on molar mass). The available gas can fluctuate greatly (50%), so the turn-down on the compressors is important.

The compressed gas needs to be oil-free. I'm looking at oil-free screw compressors or centrifugal compressors. Any advice on the selection would be appreciated!

The process conditions are:

Suction pressure: 83 kPa(a) [12.038 psia]
Discharge pressure: 700 kPa(a) [101.526 psia]
Suction temperature: 35 degC
Nominal flow: 16000 Nm3/h [13.6 MMscfd]
Nominal molar mass: 24.3 g/mol

Thanks for your help!

 

[zdas04]

This application is approx 12 compression ratios: (1) Three or maybe 4 stages with a centrifugal (with requisite intercooling); (2) 2 stages with a dry screw (with an intercooler); or (3) one stage with an oil flooded screw. So by eliminating the oil-flooded screw from consideration you have added the requirement for at least one intercooler to avoid the reasonably easy task of oil removal on discharge. Not sure this is a great approach.

However you feel about oil-flooded screws, do some hard evaluation of centrifugals. Look at the compression map and realize how small the compression envelope is. Basically, my rule of thumb is that the first compression that a raw stream should see will be positive displacement. I'm happy to use a dynamic machine after I've gotten Mother Nature's pressure/temperature "controls" out of the picture, but when I have a stream where "the composition is slightly variable (±15% on molar mass). The available gas can fluctuate greatly (50%), so the turn-down on the compressors is important." I want a PD machine.

Dry screws have the ability to do about 4 compression ratios in a frame, and temperature control is really important, they are far less robust than a recip and you'll find (if you go this way) that with raw, nasty gas they break more frequently than you'll want to deal with.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[rotw]

Further considerations for the centrifugal option...

One option is to target a single casing machine configured in back to back; of course inter-cooling is in order as suggested above.

If there is a significant swing in gas MW, it is going to be difficult; means that satisfying same operating conditions with a gas MW below 24 could be too demanding for a single casing.

If this is about coping with large flow rate fluctuations and MW is relatively stable, typically coping with turndown would involve investigating suction throttling or variable speed control method options. Heavy (and costly) continuous by pass / recycle could be anavoidable...

You need to be more specific about the alternative operating conditions ; for example if the fluctuations can be cast into two sets of {low head / high flow} and {high head/low flow} operating points - there may be also way to take advantage of the configuration I suggested above...

Edit: double checked MW and change within +/-15% variation (i.e. min. MW=20.6) could be okay for single casing / back to back.

 

[georgeverghese]

Feedback I got on the handling of a steel production electric arc furnace's exhaust gases at the downstream blower was that there were never ending problems with dusthandling equipment upstream of the blower, which included an electric precipitator and a huge baghouse filter - filter cloth at the filter was often pitted with burn holes from high temperature dust.

If you may face similar risks here with breakthrough of abrasive dust into the compressor suction, it would seem a VFD centrifugal compressor may be a far safer bet.

 

[zdas04]

While a VFD can improve the (miserable) turn down ratio of a centrifugal compressor, it is not enough. The gas composition of your application changes too much, the amount of particulates change too much, the total flow rate changes too much for a machine with a narrow operating envelope.

As I said above, I would use an oil flooded screw with fairly course suction filtration (maybe a duplex basket strainer), and very robust processing of the compressed gas. I didn't mention it above, but if you maintain the oil temperature high enough to cook water off of it, carry-over will be negligible and oil life will be very good. In addition, I would look at some sort of powered oil cleaner (like the ones from DeLaval) running continuously on the outlet oil and use that mechanism to get rid of the majority of the particulates.

[bold]David Simpson, PE[/bold]
MuleShoe Engineering

In questions of science, the authority of a thousand is not worth the humble reasoning of a single individual. Galileo Galilei, Italian Physicist

 

[BernardE]

Thanks for the great answers, this is very helpful.

The requirement for oil-free compression comes from our technology licencor. The process has very little tolerance for oil in the gas stream. This seems to leave me with the choice between centrifugal or oil-free screw. Do you have any experience on the relative cost of the two options? I have gotten a budget quote from MAN Diesel & Turbo for an oil-free screw that is a bit more expensive that what we budgeted for (around 25%).

Upstream of the compressors there is a gas purification plant that scrubs the gas to remove dust and particulates. The moisture in the gas is then removed via a knock-out system. We plan on installing an additional knock-out pot directly upstream of the compressors to further remove droplets and particulates.

In terms of redundancy, our original design called for a full 3 x 50% arrangement. I am thinking about looking at a 2 x 50%, with a spare bare block compressor held in store. Does this sound realistic for both centrifugals and screws?

 

[rotw]

I can answer with regard to centrifugals only. These machines especially to API 617 are considered to have high reliability/availability and are generally not spared (unless you want to waste money for that which is completely unnecessarily). It is common practice to purchase a spare compressor rotor mechanically tested with the main order. Furthermore, and I understand that you are looking at a basic screening, you cannot avoid to have a detailed analysis done before specifying redundancy / sparing philosophy etc. (e.g. based on criticality rating, type of service, total cost of ownership, etc.).

Anyway there is here another problem; be aware that if you cut the total flow by half, you decrease the overall compression efficiency, so your operating costs will increase. Also doubling the number of machines will definitely increase your capital costs and also footprint. Even more problematic is that possibly (actually it is quite probable given your nominal data) centrifugal compressor option could not be selected anymore because the flow coefficients on the high pressure stages will become too low for any centrifugal.

So short answer, it does not sound like a good plan for the centrifugal option to go 2 x 50%, let alone 3 x 50%.

NB. I assumed your nominal duty is the total duty (i.e.,1 x 100%).