stand by/duty pumps units 1

[rotw]

Hello,

Consider three motor driven pumps units of 1 MW each. Design is for 2 units in parallel operation and 1 in stand by.

My question is how do we preserve the unit which is in stand by ?

There are two scenario which I can think about:

1 - stand by machine is purged, and long term preservatives are added to the pump casing etc. Auxiliaries if any are also purged and preserved with chemicals or else as appropriate. So In case of routine maintenance on one of the two operating pumps, the stand by pump is prepared in advance, start up check list is checked then pump is put on line while the other goes on maintenance - service is never interrupted.

2 - stand by machine must be ready to start at any time. That means if one pump fail, service is critical so the stand by machine starts immediately. In this case, how do we make sure the pump is ready to start and long term preserved ?
or do we have to change over one pump between three pumps regularly to achieve that ? but then how often ? I was also wondering if there are no consequences due to this on machine lifetime due to repetitive start/stop cycles (e.g motor driver).

Thanks for any insight

 

[bimr]

The control system that starts the pump is typically configured so that the pumps are automatically alternated. Having two pumps alternating will help with the issue of repetitive starts. If only one pump was in service the repetitive starts would be of more concern. Alternate all three pumps.

The idea of a preservative is a bad idea unless your process can tolerate the preservative solution.

The possibility of two operating pumps failing at the same time is low, so you should not be concerned about this failure scenario,

 

[BigInch]

I'm not so sure about all of that. Even if you have a 98% reliability on your pumps, there is still 0.98 * 0.98 = 0.96 => 4% down time. That could cost you plenty, if you're down 4% of the year and you happen to be running a 200,000 BOPD oil pipeline.

I hate Windowz 8!!!!

 

[ccfowler]

rotary,

You haven't stated the nature of the liquid being pumped, so that is a significant unknown. Unless there are some particularly nasty fluid composition problems, I would not consider a pump filled with some preservative as being in standby duty. Except for the preservative, you are describing a very common "three 50% pumps" configuration. Usually, paying attention to hours of operation and how they are staggered gets a fair amount of discussion, and depending on the specific operating requirements different philosophies of choosing their duty cycles can make sense.

Usually, problems associated with the standby pump are most easily handled by taking care to run each of the pumps on a sufficiently frequent basis. Usually, wear and required period maintenance and repair are linked primarily to hours of operation. If operating time is evenly spread between the three pumps, then they can all be expected to require repairs at about the same time. You may want to give careful consideration to the specific characteristics of your system and schedule the division of pump duty cycles so that maintenance and repair activity scheduling can optimized for operating needs and process reliability.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

 

[JJPellin]

In our refinery, we would normally keep the standby pump full of product, hot and ready to run at all times. We would run the standby pump once per month for vibration data and to circulate the lube oil. If the pumps had identical drivers (all three electric motors), we would rotate which two were running so that all three got approximately the same run time. If there were different drivers with different energy costs (two motors and one steam turbine would be typical), we would run the turbine for at least four hours at least once every three months. This is a minimum requirement of our vibration program which is part of our OSHA 1910.119 Mechanical Integrity program. We do not see any problems associated with the start-stop. Our motors typically run 10+ years between overhauls. Our pumps typically run 4+ years between failures of any sort. We don’t see any significant difference between the reliability of the main pumps and the standby pumps. Many (maybe 25%) of our standby pumps are set up to automatically start on the loss of the main pump.

I would agree with the replies above in that putting some sort of preservative in the pump does not seem like a good idea.


Johnny Pellin

 

[ScottyUK]

Johnny -

10 years between overhauls on the motors seems a fairly long period. Are these white metalled bearings or rolling element? What interim maintenance do you typically do?

 

[LittleInch]

rotaryw, as ever "it depends" - is your process one that requires a "hot" standby ready to come on line at a moments notice or not?

What is the consequence, if any, of having your spare pump full of product and pressurised without rotating - do the seals like this?

In multi product pipelines, it is common to have auto vent and drain systems whereby product is drained and then filled when the pump is required to avoid contamination of a match with a pump full of a different product. Takes about 10 mins to get the spare on line, but in the mean time a parallel operation would pump at about 65% of max flow. If your process fluid changes over time, this might become a consideration.

where you have equal 3 x 50%, there is normally a system for changing operation of the duty and spare in order to equalise wear and ensure that your spare does actually work before you actually need it. What that system is varies widely from changing it every day to a week, a month, 3 months...

Purged and preserved with chemicals is a bit of a long term mothballing exercise and would normally not be regarded as a standby, but a "spare".

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[JJPellin]

ScottyUK,

Sorry, I didn't reply sooner. I have been on 14 hour night shift for the past 23 days. The majority our motors are NEMA frame motors smaller than 250 HP with ball bearings. We grease our motor bearings based on speed, load and bearing size. I think we have a very good program. We do not usually do any other time-based maintenance on the NEMA motors. The larger motors get pulled for major overhaul every 10 to 15 years. These are generally sleeve bearings with circulating lube oil. I am not a motor guy, so I don’t know what work is done during these major overhauls.

We may have an advantage for motor reliability because of our location. In Minnesota, it is very cold and very dry for a good portion of the year. That may help with insulation life and bearing life, also.


Johnny Pellin

 

[rotw]

Thank you very much all for your inputs.

Really helpful and most appreciated.

 

[milltir]

I don't agree with the comment by BigInch above.

If the individual availability of a pump is 98% then the availability of the system will always improve with standby pumps in parallel. In the instance described in this thread of 3 x 50%, the system availability will be 99.74% if an individual pump availability is 98% (Binomial Distribution theorem).

 

[milltir]

99.94% not 99.74% if I had read my spreadsheet correctly (eyes going with age).

(1 x p³) + (1 x 3p²q) + (0.5 x 3pq²)

 

[ScottyUK]

Johnny,


Thanks for that information. I think NEMA motors are generally better-built than the majority of IEC types and are more generous in both the iron and copper in them. I guess larger bearings with lower percentage load probably helps. Your climate sounds the opposite of ours, other than it isn't especially warm here.

 

[rotw]

Guys

What about centrifugal compressors ? how do you deal with a compressor motor driven which has purpose of stand by unit.

My guess is that machine shall be designed for long term stand still. Scenario 1 (hot start up): in this case the machine will have auto-start capability and the lines will be kept pressurized, still machine will be started once in a while (to determine) in recycle mode (means inlet and outlet blocked and gas recirculated via the anti-surge by pass). During mechanical stop I don't know if I have to supply oil to the bearings. Certainly the sealing shall be buffered.

Scenario 2 (cold stand by) : If the machine has no auto start feature then the casing can be depressurized, no need to buffer the seals. When the compressor has to go online, operators will check procedure and start up according start up protocol whereby the lines will be pressurized so the machine can start up. During machine stop rotor shall be turned. I am thinking about slow roll or do you think it requires to have the complete train started on which case pressurizing the lines or using some inert gas will be needed. I am also not clear if the oil shall be supplied to bearing when machine is standing idle.

Whether it is scenario 1 or 2 is a matter of plant philosophy and maintaining production.

Sorry to post in the pump forum but I would like to continue on the same thread.
Thank you if you can share your experience on this issue. What type of operational issue do you foresee?


"If you want to acquire a knowledge or skill, read a book and practice the skill".

 

[BigInch]

milltir,
You are right. I should have been working with the 2%, 1-0.98

you must get smarter than the software you're using.

 

[pennpiper]

The original question stated:
"Consider three motor driven pumps units of 1 MW each. Design is for 2 units in parallel operation and 1 in stand by. My question is how do we preserve the unit which is in stand by?"

I read an article sometime back (I don't remember where I found the article or the title) which gave what sounded like a very good and reasonable advice. In general the article stated the following.
Process Plants such as refineries and Chemical Plants where you find multiple pump sets (described in our question) are an environment rich in Vibration. Some of the vibration is low velocity and low in frequency not noticeable to humans. Some vibrations are very strong and will permeate and damage machines that are shutdown. These vibrations appear to concentrate at small point to point contacts and cause damage at these points.
To demonstrate this phenomenon the author of the article included close-up photographs of pitting in the bearing raceway where the Ball came in contact with the raceway. According to the article, this pump had been allowed to set idle for a year.
To prevent this type of damage, the author recommended a "Run-Rest" protocol with all pumps having equal "Run" periods shorter "Rest" periods.
Example(Pumps "A", "B" & "C")
- Run A & B - Rest C
- Run A & C - Rest B
- Run B & C - Rest A
- Repeat - - -
This can be done on a 24 hour cycle or a 48 hour cycle or any thing in between.




prognosis: Lead or Lag

 

[EdStainless]

We do not let our spare pumps sit longer than 72hrs between runs.
I can see the benefit of cycling through the series more often.
This needs to be tied into your monitoring system. Often you can learn a lot about a motors condition by watching it at start up.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube