stand by pump phelosophy

[upm]

Gents,
I am in the process to develop a swapping criteria for the stand by pump. When to swap , how and what frequency is required. I know it is not good to keep the s/b pump idle and it is not good to keep on starting it. So we are in between......but what to do????
the objective is to ensure the function of the s/b pump whenever is required .
Does any body though about it? any bodey has any thing in place???
upm

 

[pablo02]

How often does the main pump operate? What type of service? (what does it pump & how critical is the need?)

How large are the pumps and how large (Hp & voltage) are the motors?

What maintenance frequency do you try to maintain? Do you wish to keep the maintenance of each apart to keep the costs spread out?

We have sewage lift pumps that automatically alternate each time a pump cycle occurs and then we have some pumps that are manually "swapped" on a weekly basis, some on a monthly basis...

you need to look at your system closely, evaluate your needs (costs as well as availability), etc. and then choose a cycle that best suits your requirements and yet is easily fufilled by the people who operate the equipment [the first of the month or every Sunday is usually easier to get done than every 1000 hours]...

 

[PUMPDESIGNER]

We approach stand by pumps as follows:

1 Always avoid the stand by if that is appropriate.
We examine what will happen if there is a problem.
A If a totally new pump and motor can be obtained or any
repairs can be made in a reasonable amount of time that
is acceptable to the customer or application, then do
not put a stand-by pump in.
B We estimate how often problems will occur.

We have many customers decide against a stand by pump just on that first analyses.

2 If a stand by pump is provided, our first option will be to place that pump into full operation with alternating duty cycles achieved automatically. This ensures that no human invervention is required, the pumps and all valves and fittings will be exercised and maintained properly. When one unit fails, the other unit will be already in operation and functioning and thereby no down time will occur and repairs/replacement can be done in a comfortable manner with no rush or pressure on anyone.

2 If a standby pump cannot be placed into full time alternating duty, then the options multiply with almost limitless possibilities too numerous to consider. At this point we generally offer just a simple pump that they can change out, and thereby we are eliminated from that tangled mess about how often they do what to keep that pump in good condition. They can connect the pump but keep it turned off, they can leave the pump uninstalled and then install it, they can leave the pump uninstalled but run it anyway and maintain it in some fashion, and on and on it goes.

We usuall stress that a standby is best if in full operation alternating the work load, which then of course makes both pumps stand by pumps.

PUMPDESIGNER

 

[electricpete]

Let me see if I understand your question.
I assume what you are saying is that the system is already built. You have 2 100% pumps or 3 50% pumps. One is in standby and never required to auto-start during normal operation. You want to know if there should be some manually-initiated rotation between pumps. Your goal is to provide highest confidence that the standby pump will auto-start and function properly when called upon.

If I have understood your question it is the same one we have given a lot of thought to for balance-of-plant motors at our nuke plant.

The driving question: do you really have confidence that the standby pump will work when called upon if it has been sitting in standby for a year? (Would you trust your car to start on first attempt after sitting idle for a year?)

Things that can go wrong on auto-start after long idle period: sleeve bearing binding during long periods of inactivity, breaker binding, control circuit switch/relay binding, discharge valve auto-open and recirc valve auto-close failures etc etc.

My opinion is the best procedure is to start those standby's once a week or once a month. It should be under a controlled condition which tests as much of the auto-start circuit as possible, without jeapardizing reliability of the plant during the auto-start.

Now any failures are flushed out when you start the pump on your own time frame (all other pumps still available), rather than finding out it doesn't work too late (when the other pump just died).

In some of our larger motors there is plant risk involved in the swap itself. That has to be factored in.

I have asked this same question before on this board and other boards. Let me know if I have understand your question correctly and if so I will get a link to those threads.

 

[imok2]

Let's look at both sides of the discussion. We'll begin with reasons for alternating pumps:

If the pump is not run, the oil will drain away from the movable components, causing excessive wear and heat at startup.
Maybe the standby pump is frozen up. It's too late to learn that when you need the pump.
All standby equipment is subject to vibration. Bearings can be affected by this vibration with a condition known as "false brinneling," causing round, hard indentations in the bearing races.

People that do not believe in alternating pumps also have some compelling arguments:

Every time you switch pumps you are causing a system upset than will probably affect the finished product. Temperatures and pressures change and some products become viscous or solidify when the system cools even a small amount. All of this translates to "off product" that will end up in the alcohol plant, have to refined all over again, or is sometimes shipped to the customer who refuses the shipment and asks you to initiate a program that will prevent this from ever happening again.
Each time you let the bearing cavity cool down you are producing moisture in the bearing case. The more often you start the pump, the more frequently this occurs. Moisture is a major cause of premature bearing failure
Startup torque is five times running torque meaning that it takes five times the power to start a pump than it takes to keep it running. These power surges can trip breakers or, in some cases, cause you to exceed your peak loads that will have an adverse affect on next month's electric bill.
Every time a pump starts the shaft thrusts towards the thrust bearing and then somewhere close to its 65% efficiency point, the shaft thrust in the opposite direction towards the pump volute. This axial shifting can cause seal and bearing problems.
Shutting down a pump will cause its internal temperature to change and that can cause a problem with many fluids. In some instances the shutdown pump has to be flushed out to prevent product from solidifying in the seal or on the surfaces of the impeller and volute. At startup, any of these solids that have not returned to their liquid state can cause the rotating shaft to go "out of balance."
Most knowledgeable people agree that pumps with long shafts should be turned over on a regular basis to prevent "sag." Jacking gear is often provided to do this.

If you decide that alternating pump makes sense to you, be careful that you do not run the pumps for the same amount of time, or they will both wear out with the same amount of hours.

 

[upm]

Gents,
Thankl alot for your thouts an dlet me answer some of your questions.
pablo02:
The pumps are used in a chemical plant (olefins) and all varies in size between 3 KW to 1 MW motors. Mainly any pump in the plant is critical as it will cause process upset when both pumps are not availiable.There is no measure for the pump running hours as we can run it to fail or swap to the standby pump. The main idea of alternating is to ensure the availibility of the stand by pump once is required.
electricpete :
The pumps are already built as 2X100% and your understanding of the questio is correct.i would like to get the links you have used in the past to assist in the program development.
This is all started when the main bottom pump failed and when swapped to the standby it did fail after some time, thus upset the plant opeartions and was about to shut it down. One factor i am aware of is bearing brinneling if it kept idle for long time and turnning the shaft manual might solve it.
The problem is how do we set the alternating frequency and do need to run both the same running hours????
If oil mist is supplied to the bearings does it take brinneling out???
upm

 

[PUMPDESIGNER]

All of our alternating is done via PLC programming rather than an alternating relay.
PLC alternation allows specific programming for each situation as required.

With PLC you can alternate pumps based on anything including:

Time
Starts and Stops
Pressure (Option of 1 pump only, or both pumps).
Flow (Option of 1 pump only, or both pumps).

You can also overlap the start and stop of the two pumps.
Automatically bring on lag pumps when lead pump fails.

You can also change any parameter when needed just through software.

Simple alternating relay can be used, no options hardly and most options would require field modifications to panel.




PUMPDESIGNER

 

[Parado]

Stand by pump, often refer to vital equipment with continuous service. It does not normally provide spare capacity, but more toward availability.
Pumps with intermetent service and having N+1 or N+2 should considered as spare pump because they provide spare capacity.
To generate swapping criteria, many factors should be considered. Sometime one criteria may not necessary suitable to every equipment. Following are some factors for consideration.

1) Equipment with rolling element bearing, especially vertically mounted - too long stand by can easily cause bearing false brinelling, normally "Test Run" once a month.
2) Heavy equipment with journal bearing horizontally mounted - avoid long stand-by which lead to rotor sagging or bow. Normally hand bar/turn by 90/180 deg once in couple of days. But if too heavy the rotor, barring with motor may be necessary but not for long due to slow-roll wipping of bearing.
3) Gear set or others - too long stand by can cause stand still corrosion, normally continue N2 purge or blanketing is applied during stand-by period.
4) Alternative service - equipment will accumulate close service hours and possible running into failure at the same time. This practice normally not adviseable, but applicable to equipment with precise operating statistic where running hours of equipment are well predicted. Stagard operation after all equipment reaching about 70 - 80% of predicted life.

Definetely there are more to consider. But what so ever, too frequent stop-start always harmful to equipment.

 

[upm]

i am trying to classify the pumps based on the criticality and services. I am stuck with product pumps and lube oil pumps as swapping might cause a process dip and cause a plant trip (e.g. low lube oil pressure).
i am thinking of running them for an hour once a month in circulation and check up the vibration and performance. These are PD screw pumps..............any concern???
upm

 

[electricpete]

Some more comments on that. Sleeve bearings do have a unique problem with becoming stuck after long idle periods which may be partly overcome by manual rotation.

I don't think oil mist will make a difference in preventing either false brinneling or sticking of sleeve bearings.

Here is thread on this subject that I initiated on another site:

http://www.vtab.se/PHP-NBoard/html/images/materiali/Forum2/HTML/002473.html

There is a similar thread I initiated at eng-tips in the same time-frame but I can't find it yet.

 

[abeltio]

upm
now your question is more clear...my experience is from the power plant side...
there are as you mentioned 2 positions...
run the same amount of hs on both pumps
run one pump and exercise the stand-by
let's save ourselves the pros/cons of each position...
i have seen each one supported vehemently with a 50% split.
personally i adhere to: run one pump and excercise the stand by.
the routine i have implemented in a couple of sites is:
run the lead pump continuously, exercise the stand-by 1hr every other day which gives the following hs for a 24K interval (major plant overhaul):
52x7/2 = 182 hs every 52 weeks or... approx 500hs between major overhauls.
Which according to the OEM manual in those cases required only a minor check-up of the motor+pump of the stand by pump... and a complete overhaul of the pump that accumulated most of the hours...
Why we chose this approach? Of course... learning the hard way...
We had the same amount of running hours on both pumps... the OH was due... all the spares were in stock... the contractor picked up the parts, unfortunately the truck had an accident (slid off road and gave a couple of turns until it stopped against a tree - just minor bruises to the driver) and all the parts got some damage (murphy's law).
Needless to say, this was somewhere in the middle of nowhere... and it took 3 weeks to re-stock all the spares + 25% premium for emergency delivery
We learned two things:
- never release ALL the spares at the same time.
- never burn the candle by both ends (i.e. run both pumps until the maintenance is due on both at the same time)

Had we decided to run just one pump and have the stand by with just a few hs we would not have had any delay... we would have re-started the plant with the stand by pumps and happily waited for the spares to arrive for the other pumps.

This is for lube oil/cooling water/liquid fuel pumps... indoor, with moderate temperatures (between 20C and 80C)

it is very important to have a reliable lead/lag sequence...
1. to change between pumps with a time overlap
2. so that the stand by takes over in case the lead pump fails.
if the lead lag sequence does not work properly, what we have is a system with only one running pump and a spare pump already installed.

Hope this helps.
Saludos.
a.

saludos.
a.

 

[jet1749]

Some excellent replies. Our experience with duty/standby pumps is lead/lag cycles of 60%/40% This ensures that ordinarily both pumps should not expire at the same time. False brinelling of the standby unit can be taken out of the equation if manual pump rotation can be carried out periodically (weekly?)and also if the pumps do not share baseplate or foundations in common, perhaps install flexible pipe joints to prevent transmission of vibration to standby unit.

 

[electricpete]

One thing about manual rotation of a standby pump:
That will require the pump to be tagged out for personnel safety, correct?

In that case the standby function would be jeapardized during the brief period when the equipment is tagged out. Do you guys agree or is there a way around it?

I think it is probably not a big concern as long as attention is paid to ensure the tagout does not remain in effect for any longer than necessary for the manual rotation.

 

[Parado]

Stand by pump normally are auto cut-in. Majority of which are base on low or low-low pressure. Of course minority are base on level and capacity.
It is definetely not adviseable manually or hand turning the stand by unit in any case when the main pump is in operation due to safety reason, which should be the prime factor to put in consideration when deloping of any criteria or program.
False brinelling not necessary cause by pumps installed in the same skid, but vibration sourced from equipment install around the plant can induce vibration energy even through the grouting. More often happen to vertically mounted pumps or motors. Reason for that is because the vertically mounted equipment do not have rotor weight acting radially to the bearing. Control bearing radial internal clearance to zero can help prolonging bearing life before false brinelling take place, but still do not totally avoid bearing from this problem. Monthly test run for a couple of hours do help a lot in this case.

 

[upm]

Gents,
very wide subject and if the answer to the question i raised is simple, there will never be such detailed response and discussion.
Turning the pump manually is not adviced in our plant as we need totaly to isolate and tag out the pump to do such whsich is not prefered as the main pump or/and process might requires the stand by pump during this period. i think i will approach it by asking my self what can go wrong if the pump is idle? or frequently started? then i need to evalute these two questions for each pump application and set the frequency.
Guys, we need to remember askining manit. to turn 200 pumps once a week is a headach and they will never do? therfore, we should make it simple but effecient.
upm