Dear all ,

I would like to share with this Ex to identify the best maintenance strategy

We have 2 pump when in duty and the other one standby

1st option is to apply PM or PDM on the duty one and functional ( detective ) maintenance for the stand by

2nd option is to keep swapping the pump ( change over ) and in this case apply the same strategy for both PM or PDM

which one do u think will be the right strategy .

Moreover, normal changing oil it stay under which strategy (somebody call it predetermine maintenance )

Regards,

RCM2,

In the duty-standby situation:
1. What happens when the duty pump fails? hence what are the consequences of the duty pump failing? Remember that pre-emptive maintenance is justified ONLY when the consequence matters.
2. What happens if the standby pump fails to start? If there is a consequence, you HAVE to do something to ensure it will start AND take full load.

Rather than asking us to asnwer, try doing it yourself. You may find it more useful to understand the reasoning better.

Dear Vee thanks for quick response,



In fact the given EX is not a real case but its an assumption and when I asked u not mainly to answer me because its a problem the reason is to share the experience and keep the discussion group more active

Any how

What I think and suggest is to keep the duty pump in service all the time and apply functional test in the standby in regular bases to insure its availability during real demand.

but I have seen other plants usage the swap as proactive approach and its works fine , since u will apply the selected strategy for both.

My question which one is more effective? , although I have seen that the first approach is the one I like but still the 2nd approach it sounds Ok as will.

By the way if we selected the 1st one will the stand by pumps have some maintenance activity such as replacing oil in fixed time period or not ? if yes what type of maintenance to be called?

regards,

RCM2,

I have embedded my answers into your text.

quote:

My question which one is more effective? , although I have seen that the first approach is the one I like but still the 2nd approach it sounds Ok as will.

Vee: In my experience, the MTBF of pumps wil increase about 10% if you run duty/standby, and the costs will come down 12-15%. This is mainly beacuse seal failures will drop significantly, as they are directly affected by the number of starts.

>> By the way if we selected the 1st one will the stand by pumps have some maintenance activity such as replacing oil in fixed time period or not ? if yes what type of maintenance to be called?

Vee: There are certain maintenance activities, usually of low cost, which are worth carrying out whatever operating policy we follow, without the need for any special analysis. These include oil top-ups or changes, cleaning and taking vibration readings. These will fall under housekeeping or PM tasks. But intrusive work is only called for if a fault is found during a test start.

quote:

This is mainly beacuse seal failures will drop significantly, as they are directly affected by the number of starts.

This is true for pumps with mechanical seals, pumps with packing glands need lubrication in most cases furnished by the process fluid.

In my experience the material costs for PM as defined in our equipment budgets can be lower then 10% of the total value, depending of the type of equipment.

RCM2, Svanels,

Svanels is absolutely correct in stating,

quote:

This is true for pumps with mechanical seals, pumps with packing glands need lubrication in most cases furnished by the process fluid.

In fact, with packed glands, we must start them fairly often so that the glands are wetted and lubricated by the process liguid; that is where the alternate running philosophy originated. However, most people carried on doing the same long after mechanical seals were introduced. These have a different failure mechanism, and are best kept running continuously as long as possible.

Rotation is my belief: P1 50% of the time in service P2 30% and P3 20%. this way no two should fail at the same time and you can analyze all 3 and have 3 reliable units 95% of the time with 100% availability still achieving planned scheduled maintenance.

Sam,

I am afraid I do not agree; in my view, rotation is a sure way to reduce reliability, increase costs and workload.

This message has been edited. Last edited by: Vee,

GUys,

Although I think we would all agree that 50% rotation is a sure fire way to reduce reliable performance int he long run, rotation slanted 60/40, 70/30 etc, can be the best option for certain cases.

Of course.. operating context is the driving factor always.

People,

Assuming we are not dealing with gland-packed items, a duty-standby situation produces the fewest number of failure modes in total, for a spared set of items. Any other situation, such as 70:30 replicates the failure modes for all the spared units, adding to maintenance effort.

There are specific situations where a 90:10 or even 70:30 option is well justified. Typically, the items are complex and difficult to start, with start-up procedures taking a few hours. One advantage with the 90:10 or similar policy is that 'fail-to-start' is not a failure mode that is of concern any longer.

In my view, the duty/standby situation produces the highest reliability, highest availability, lowest cost option,. It is not always possible to apply it, but when there is a choice, that would be my preference. The next best is 90:10 or similar.

people,

In certain cases the 70/30 or 60/40 is the wisest choice and this goes beyond mere considerations of the pump.

There needs to be some thought on the media being pumped, the liklihood of blockages occuring from not moving and a range of other factors. Again, as stated previously operating context is always the guide for any analysis and it is almost impossible to set out hard and fast rules.

In an ideal world all would be able to run things in a duty stand by situation, which I agree is by far the best for overall reliability, but this is not often the case.

It would be nice if someone who has some sets of pumps in parallel ran a trial over a year or two to find out.

In general, I favour running a duty pump most of the time, but giving the standby one/s a run every week or so, so I guess that is about 90/10. I think it most likely that the duty one will wear sufficiently to justify maintenance attention, whereupon the standby takes over as duty.

A way of helping decide overhaul value is in theh attached paper, and as many repliers give their book titles, I will give mine too:

Ray Beebe
author of lots of papers and Predictive maintenance of pumps using condition monitoring ISBN 1856174085

Clarion2002_pumps.pdf (413 KB, 108 downloads)

Ray,

quote:

It would be nice if someone who has some sets of pumps in parallel ran a trial over a year or two to find out.

I have done such trials, and my advixe is based on the results of such trials. Regards.

It's not clear to me why 50/50 is a bad thing?

Someone assumed this meant more starts? I don't know how you come to that conclusion... depends on what drives you to start and shift the machine. My thought is that there is a limit on the length of time a machine sits idel before I become uncomfortable with its reliability. Therefore I would reduce the shifting interval in the 90/10 scenario from my concern for the idle pump and would have less shifting per unit time in the 50/50 scenario.

Where am I mistaken?

Hullo Pete,

50:50 produces more failure modes in total to resolve, that is why it is a poor choice, not because of number of starts. 60:40 has the same problems, and in my view, is equally poor.

90:10 or any higher ratio is fine, as the number of failure modes to resolve is less.

You are quite right when you say the number of starts has nothing to do with 50:50 operating mode. That is only affected by frequency of starts, as well as with duty/standby operations. To that extent 90:10 is better than duty/standby. However the latter has many other advantages, principally one of getting a discipline and structure in place. This in turn helps get PdM readings for BOTH units every time. Else it can be a hit or miss.

My comment on number of starts is simply to explain the rate of deterioration of seals, no more.

I hope all of this makes sense to you. Regards.

Just to ensure my understanding about this issue. CMIIW.

I believe when we choose duty and standby mode, we have to choose one pump as a duty (main) pump which will operates almost entire operation and one as standby which have to replace the main pump when have problem.

For 50:50, I believe this option is not a duty and standby mode. because both pump is active. 90:10 situation, My opinion is the duty pump operates for 90% of operational time and 10% is used by standby pump which is a functional check to ensure that the standby pump is available when emergency situation is happened.

We can use 100:0 for this option, but we have to still functional check the standby pump to ensure its availability. There will be an additional cost for the inspection, which can be avoided if we used 90:10 or similar. I believe many operator will prefer the 90:10 for the efficient cost.

Regards, Rosmana.

Guys,

While I fully agree from the pojnt of view of the pump, this ideal situation is not often available in my view.

The media being pumped also has a large impact on the rotation strategy that is employed. As indicated in an earlier post:

quote:

In certain cases the 70/30 or 60/40 is the wisest choice and this goes beyond mere considerations of the pump.

There needs to be some thought on the media being pumped, the liklihood of blockages occuring from not moving and a range of other factors. Again, as stated previously operating context is always the guide for any analysis and it is almost impossible to set out hard and fast rules.

In some of the trials that we have run on various pumping applications over the last three years we have found that where there is Lime to be pumped, for example, the rotation is generally best at 70/30 or the system will clog and not start.

When there is just water involved, or similar media, then 90/10 or greater (to 99/1 if you like) is by far the better option. Depending on the properties of the chemical or liquid being pumped, this chanegs dramatically.

Not that I disagree with whats being said, just trying to add the additional dimension to the debate.

What is the specific gravity of the lime? Is the duty-standby policy applicable to crude oil & slurry applications?

Josh.

Fair question, I dont have the answer with regard to the specific gravity question.

I have done a lot of work in the areas of slurry and crude oil and my recommendation is that yes a duty stand-by regime is effective. It may not be a 90-10 (the ideal) due to practical considerations of how long it takes for the slurry to block the lines if left unmoved, but D/SB will still be the best option.

Josh,

Yes, for crude oil service duty/standby is fine. The policy does not have anything to do with the specific gravity of the fluid.

Slurries introduce a more severe form of fouling. Similarly, if the crude oil carries a lot of sludge, similar problems may be encountered. Specifically, if the pump is idle, the solid particles(oe sludge)may deposit and plug the suction line or even the pump. This is a special case relating to this failure mode, and should not be used to justify moving away from a duty/standby mode in the general case.