Huge NH3 Compressors

Looking at a job that needs 38MW of refrigeration at 0C SST /39C SCT. So far we basing our system design on 5 x Mycom 400M or 5 x Howden 365-193. Aside from the Howden 510 series are there any bigger single compressors for ammonia? Customer would like to do the job in 2 machines (i.e 19MW each) is there a screw compressor available at this capacity anywhere?

Not sure if ammonia runs well in centrifugal compressors either due to lowish density and issues with tip speeds to generate the kinematic pressure conversion

Now that is a good sized NH3 system. 19 MW per compressor in a screw is asking for a lot. I have seen some centrifugal compressors used for ammonia, so that might be the only way to get this much capacity in only two machines.

The Howden 510 screws are the largest I am aware of, but they can be expensive. However, in a project this large cost is relative to installation costs also unless there are requirements for part load operation or efficiency.

Is this for a mine?

Not for mine - petrochem application.

Do you have any idea of who makes the ammonia centrifs?

The ones I have seen were York Turbomasters. Somewhere around 9-10 MW, but I'm sure they have larger ones. These are not what I would call petrochem duty compressors though as I don't think they would meet the API standards for centrifugal compressors with all of the special requirements.

I don't really have much experience with the large centrifugals, but one of our other members (TXiceman) does and he might be able to offer more details.

No matter what though, this sounds like a fun project.:)

The 510mm Howdens are fine for a base load machine, but I do like some duplicity in the compressors especially in an industrial application. Frick makes some big machines, but not a large as the Howden,

One thing you need to seriously consider is the turn down the machines will be required to follow. Multiple machines will follow the load in a much more energy efficient manor. The screw compressor is most efficient at full load.

Another issue will be the bearing design. For this type application, I would demand sleeve type journal bearings and tilt pad type thrust bearings.

As for centrifugals on ammonia, you will need to look to companies like Elliott. They are expensive to buy, long delivery and expensive to repair.

On the issue of repairs, the 510mm machines are HUGE and require specialized rebuild areas that can handle the weight. The bare compressor casing is 6 feet tall from the feet to the suction flange.

I would generally have a tendency to go with 4 - 321mm machines over 2 - 510mm machines. Repair parts are more readily available and they are easier to work on.

Ken

Quote:

---

Originally Posted by Cold Down Under

Not for mine - petrochem application.

Do you have any idea of who makes the ammonia
centrifs?

---

Hi Cold, Hitashi does

Norm

Hello Cold,

Sabroe has standard NH3 Chillers untill 6 Mw.
Above is at request.

Greetings, Bart.

Hi, Cold Down Under :)
welcome to RE forums...

Quote:

---

Originally Posted by Cold Down Under

Looking at a job that needs 38MW of refrigeration at 0C SST /39C SCT. So far we basing our system design on 5 x Mycom 400M or 5 x Howden 365-193. Aside from the Howden 510 series are there any bigger single compressors for ammonia? Customer would like to do the job in 2 machines (i.e 19MW each) is there a screw compressor available at this capacity anywhere?

Not sure if ammonia runs well in centrifugal compressors either due to lowish density and issues with tip speeds to generate the kinematic pressure conversion

---

I will suggest to go to 6 screws (5 for normal duty 20% of capacity and 6th for stand by) it is more easy for maintenance. Just to remove compressor from base and take it to workshop and meanwhile start up stand by one.

Quote:

---

Originally Posted by TXiceman

One thing you need to seriously consider is the turn down the machines will be required to follow. Multiple machines will follow the load in a much more energy efficient manor. The screw compressor is most efficient at full load.

---

Very simple calculation can show how much of production you can lost if you have only 2 compressors covering complete capacity and you stop with one compressor what means lost of 50% of capacity, even with 4 you can lost 1/4 (25%) of capacity if you do not have a spare one.

I was working in petrochemical plant and I know how expensive is to stop the plant in emergency i.e. sudden stop of cooling machinery and sending all products to torch....

....maybe I am not right, but for such plant I will always try to cover the worst scenario....

Best regards, Josip :)

Quote:

---

Originally Posted by Josip

Hi, Cold Down Under :)
welcome to RE forums...



I will suggest to go to 6 screws (5 for normal duty 20% of capacity and 6th for stand by) it is more easy for maintenance. Just to remove compressor from base and take it to workshop and meanwhile start up stand by one.



Very simple calculation can show how much of production you can lost if you have only 2 compressors covering complete capacity and you stop with one compressor what means lost of 50% of capacity, even with 4 you can lost 1/4 (25%) of capacity if you do not have a spare one.

I was working in petrochemical plant and I know how expensive is to stop the plant in emergency i.e. sudden stop of cooling machinery and sending all products to torch....

....maybe I am not right, but for such plant I will always try to cover the worst scenario....

Best regards, Josip :)

---

I could not of put it better myself!
How many of us have seen beautiful plant installs, only to realise there is no spare capacity.

Also Having large or very large plant is great under stable and continuous duty.
But start to unload that duty or introduce a fluctuating load!

Most the large installs I work on have NEVER struggled to operate and full capacity.

But boy have we had issues when the load drops off or fluctuates!

It's like a Juggernaut at full speed suddenly trying to stop.

Besides as everyone knows nothing runs forever
everything needs servicing occasionally.

Correct me if I am wrong but most Screws start to become inefficient below approx 80% load.
OK slightly more "range" when a variable speed option is used.

More and more often Screws are being installed without due consideration to these constraints.
Grizzly

Yup as per Bart, there is from Sabroe with 9MW. Not sure about Frick though. I gotta to check

Quote:

---

Originally Posted by Josip

I was working in petrochemical plant and I know how expensive is to stop the plant in emergency i.e. sudden stop of cooling machinery and sending all products to torch....

---

This is an important point. Lost process time is extremely expensive, much more than the equipment itself. Even as TXiceman suggested multiple Howden 321's with a 1.93 rotor LD ratio is a good starting point for discussions. The longer LD ratios have more rotor flex in them, so I might suggest the short LD ratios for an improved reliability issue.

Even if it took five compressors, having one available (or a simple bare shaft compressor in stock) might be a reasonable alternative.

Kobelco 32 size (5,9 MW)

I have applied a number of 321/1.93 machines and no problems. They all had the tilting pad thrust bearings. IS this a 50 or 60 hz application? If the pressure ratio is not too high, the longer L/D ratios do OK.

You might also look at the larger Frick and Mycom machines.

The first cost will increase for the installation with all of the extra switch gear and wiring.

Just how critical is the application? How long can they afford to be shutdown?

I had one customer with special steel case compressor that sold for about $140,000 US for a bare compressor. Delivery of a new machine was 5 to 6 months and the accountants could not justify a spare compressor. They claimed that if the machine was down that 1/2 of the plant was down and it was costing them about $100,000 to $150,000 per day in lost production and possible EPA fines. Operations wanted a warehouse spare and the bean counters said that they would need to do proper maintenance on the compressor. Even if they had to come down for a minimal overhaul, they would be down 2 to 3 days. They would not consider sparing anything other than a shaft seal.

From a plant operations stand point I have put in a number of systems with 3 - 50% or 4 - 33% compressors. They rotate the spare into service on a monthly basis. When they have problems with a machine, they simply take it out of service and bring up the spare.

But Bean counters always know more than engineers and operators.

Ken

I agree with the 321/1.93 sizes since the operating conditions are only 0C SST /39C SCT on ammonia. That's not too bad at all. Howden used to have a 321/2.2 if I remember right, but I'm not sure if they still do or not.

I know on the 255/2.2 there was a fairly low differential pressure you could not exceed and I always attributed that to the long rotor LD ratio.

The job I worked on with the spare compressor was for a chemical plant and the project manager said he wanted a spare no matter what because of the cost of missed production. In this case the bean counters lost or probably agreed with the logic.

Thanks guys - we are zeroing in on a solution - we tend to agree with the multiple machines theory - this customer's system basically has three loads, two are variable and the third 70% of the load wants as much refrigeration as available so as the other two loads back off - more cooling is available for the big load - sounds like and easy life for the compressors.

I would choose the Mycom 400 compressors, ... they have a lot of expercience with petrochemical plants, ...

I think the plant should be with 3 or 4 compressor package as it is easy to handle on part load/ spares availability.

raj

Years ago Carrier built packages the size of the average house, for petro-chem work, I think chilled water and centrifs. But the interesting thing they had gas turbine drives, the gas is readily available at petro-chem sites. Has your client looked at the option.
Magoo

Further to earlier post, I agree with Haroldmycom's point that Mycom have experince in the petrochem business, pumping all sorts of nasty stuff, plus their compressors run up to 4500 rpm which would suit a gas turbine application. Cost savings in that client cuts the electricity grid out of project and running costs.
Talk to Mycom.
magoo

A Mycom 400 @3550rpm NH3 0/39C economised will do about 10MW and 9.3Mw uneconomised @3550rpm. This would give you your total capacity with just 4 machines. You will need to speed a bit for a VSD to drive the 2MW motor to run then.

Regards
Mike W

Mike W.
Thanks for correction on top end speed for a mycom 400. I should have checked before posting.
magoo

Cold Down Under.
Were are we all, with your post? Is it dead on the water or what ever.
magoo

A job of this magnitude would take a few years to happen.Petro chem like to talk it to death before making any decision and this would be only one part of the process.

If Frick did the job, they will definitely prefer Centrofugal compressors.

Quote:

---

Originally Posted by RANGER1

A job of this magnitude would take a few years to happen.Petro chem like to talk it to death before making any decision and this would be only one part of the process.

---

Thats a fact, process clients are hell ... , we have to follow all kinds of spec's.

:rolleyes:

The petro chemical and refinery companies are the ones that separate the boys from the men. Dealing with the spec from Royal Dutch Shell, Exxon, BASF and Dow is not much of a problem. You just have to read the specs, sort out what applies and what does not apply and understand ASME Section VIII, ANSI B31.1, B31.5 and the British and EU codes and standards.

There are a lot of "industrial refrigeration" companies out there that will try to take on the projects with the heavy specs and go down in a hurry and give the rest of the true industrial refrigeration companies a bad name.

Before you jump off with a company that claims they can handle big specs, ask for a users list with reference names and phone numbers.

Ken

Quote:

---

Originally Posted by TXiceman

There are a lot of "industrial refrigeration" companies out there that will try to take on the projects with the heavy specs and go down in a hurry...

---

That is very true. I've seen companies who think process work is just like any other ammonia project and find themselves rudely faced with reality of process spec.'s.;)

I think that there are far too many variables when it comes to applying the right compressor package for a petrochem/oil&gas application to even list half of them on a forum. However the first client requirement I would look for is API619. This seperates the low spec from the medium to high spec packages. Many compressor manufacturers claim to be compliant but does compliance include a large list of deviations? There are very few compressors out there that can provide close compliance to API619 with those being Howden, Kobe and Mycom.

To pick up on a few points earlier in the thread, gas turbines, Yes petrochem plants will have fuel gas available but the gas consumption of these turbines is quite high and I would assume that the HV induction motor would be a better solution. A 2MW Exd motor coupled to a speed increasing gearbox would cost approx £180K. A turbine approx £1M+

Bearing types - I only know of sleeve and tilt pad assemblies being available at these sizes, what manufacturers provide different?

Maintenance - the Howden 510 is a big compressor but it is generally applied in environments where the is a substantial amount of engineering available. Yes it weighs the best part of 10 tonnes and you need a crane to move it and a large workshop area to work on it but these Petrochem sites have this. Generally the compressor is small compared to other equipment on site.

Thanks,

Chopper

Chopper, API619 still has a lot of exceptions by Howden, Mycom and Kobe. The majority of the problems are around the lube oil system.

Some of the tings that API619 requires are:
-Stainless Steel lube oil resevoir.
-Dual oil Coolers
-Dual oil filters
-SS lube oil piping
-5 minute oil retention time.
-radial and axial proximity probes
-bearing temperature sensors
-SS lube oil filter housings
-Extensive oil system flushing
-In general an API 614 lube system

Most of these can be met, but some border on impossible with the oil flooded design compressors.

The oil spearators are very large and are a pressure vessel as opposed to a typical atmospheric API-614 lube oil system. I have installed radial probes and RTDs in a Howden WRV 255 compressor and it is not easy. Howden will not assist and will not warranty the casing as it does not have bosses in which to drill.

I have supplied a WRV 321 with a SS oil separator, oil filter housing, compressor discharge line and lube piping. It more than doubled the price of the unit.

I have also provided the oil system flush per API614.

Ken

The ones I worked on we provided dual filters, the radial & axial probes from Bentley-Nevada, bearing RTD's, and the normal cast iron compressor housing. The cast steel housings are expensive!

And for the pressure vessels, they were built to the customers specification and the oil coolers were typically TEMA-C.

Everything else was as you say; a clarification or deviation.

I have talked with people that have serviced 20,000 HP turbomasters cooling natural gas fields. Did not get to see the job or aplication unfortunatly. :(

Quote:

---

Originally Posted by US Iceman

The ones I have seen were York Turbomasters. Somewhere around 9-10 MW, but I'm sure they have larger ones. No matter what though, this sounds like a fun project.:)

---