Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Pump Recirculation line 2

Status
Not open for further replies.

Timewrap

Chemical
Dec 1, 2008
35
0
0
US
Dear Experts,

This question had been on my mind for a while, I require clafications in terms of the requiremnts of the PRL (pump recicrculation line)for a centrifugal pump. The basics I understand are 1) A small centrifugal pump can be deadheaded without any considerable damage 2) A big centrifugal pump can be deadheaded for a long time with considerable damage. 3) A PRL is required ONLY when there is risk of process cut-out or reduced flow (usually in a hot oil or a cooling water circuit when the demand can reduce to zero or to a maximum high).4) If there is no risk of deadheading the pump other than mistake of the process operator, than there is no need for the PRL.

If such is the case, then why some companies such as Suncor Canada had policy of installing PRL on every centrifugal pumps. OR what is your experience far on this particular topic??

Please comment.

Timewrap.
 
Replies continue below

Recommended for you


Timer:

I don't know where you learned your basics, but in 49 years I have never known an engineer or seen an installation that would allow a centrifugal pump to deadhead on a process application. ANY centrifugal pump will suffer resultant damage if submitted to an un-limited deadheading of its flow. In my opinion, you are asking for trouble if you design for allowing a centrifugal pump to be deadheaded.
 
Dead heading centrifugal is dangerous and foolish no matter size or service.

Pump will suffer seal damage, bearing damage, motor damage, shaft damage and numerous other things can go wrong.

The pump will continue putting energy into the trapped fluid which will result in blow out or possibly fire.
 
Dear Montemayor,

With full respect to your opinion, I'm sure you had misunderstood my question. I know that a centrifugal pump would be damaged if deadheaded for a long time and I myself never would allow such a situation to happen. I meant to ask that when would you consider to install a pump minimum flow line, does it depend on the size of the pump or does it depends on the process variable such as low flow condition due to low demand for a cooling water in the plant. You would not install a pump min. flow line just to prevent it from a operator closing the valve by mistake on the discharge line.

I've seen pump vendors listing min. flow / stable flow on the data sheet. But not have seen that every user installing the pump min. flow line.

Your valuable opinion will be most welcome.

Timewrap.
 
Depends on the pump, the discharge static head and the TIME that you want to operate at lower than BEP flow. I always think about low RPMs by looking at the process you're trying to run and feel it should approximate something like this,

At 10%, you're making a lot of heat and should never be there except when passing through on starts or stops.

If you want to run at 20% flow, it should be only for warm-ups followed by immediate ramp to higher flow ranges say within 15 minutes or so.

40% BEP flow should be for extremely limited time operations, such as an unusual upset low flow condition that should fully recover in a short period of time.

50% BEP would be a decision point, "Do we continue this process or do we shut down?" question.

60% BEP would be for the Lowest flow at which you would still try to keep your process going with an expectation of reaching 75% soon.

70% BEP would be the lowest recommended process flowrate.

80% BEP low average flowrate.

100% BEP would be average process flowrate.

120% BEP highest possible flow for very limited time (if you have power to get there).

If you need to be in low flow with high head regions for longer than the above % would suggest, you really need to think about splitting a big pump into two smaller pumps. If you need a lot of time at low flows without high head, you might consider a VSD.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)
 

Time:

I believe you are misunderstanding my comments.

You wrote: “The basics I understand are 1) A small centrifugal pump can be deadheaded without any considerable damage”. And that, I consider very foolish and something I have never seen anyone take as an accepted, specific fact.

Now, you state: “I know that a centrifugal pump would be damaged if deadheaded for a long time and I myself never would allow such a situation to happen.” I consider this contradictory to what you first stated and I find myself confused as to what you mean to say.

It is my hope that we now understand and accept the engineering criteria that NO centrifugal pump should be installed without a flow unloading device that allows for it to avoid the deadheading scenario. This can be manual or automatic – as long as there is a capability to relieve the pump.

Of course, there are some centrifugal pumps out there that have absolutely no relief to a deadheading scenario. I’ve designed and installed probably around 50 to 100 of these in my time. For example, some transfer pumps are designed on this basic scope – such as sump evacuation pumps. Others I can think of are fire pumps. Note that the end requirements of the fluid are such that one isn’t interested in maintaining a controlled flow rate. We are just interested in maximum flow under certain conditions and that’s it. The process or application is what dictates the needed control or lack of it. This last year I worked on some big pumps – 1,500 gpm at 3,500 psig pumping produced water downhole. These 12-stage beauties could suddenly confront a steadily plugged-up well – or the discharge control valve could fail in the closed position. We needed to protect these pumps from deadheading and couldn’t rely on the VSD to slow them down sufficiently. A recycle back to source had to be included – fully instrumented.

You also state: “You would not install a pump min. flow line just to prevent it from a operator closing the valve by mistake on the discharge line.” Of course you would! --- for the obvious reasons given above in my example on the water disposal well pumps. The point here is that you can’t generalize on pump applications. You have to be specific regarding the application.

I hope I've been clear in stating what I mean to say and that I haven't confused you.
 
Timewrap, it seems my view my be different from others, but here is it anyway.

A spillback is required on a centrifugal pump if:

(a) The continuous turndown flow rate is below the pump minimum flow. The turndown flow is from your basis of design and the minimum pump flow is from the vendor.

(b) You have a high head multistage pump and damage/failure is likely shortly after an operator closes the discharge control valve fully. In this case you have to fully automate the spillback.

 
Dear All, Thanks for all your valuable comments.

I guess I got what information I was looking for especially from the comments provided by BigInch, Montemayor & CJKruger. I understand that the vendor data-sheet states Thermal flow and stable flow and the higher from both should be treated as the pump min. flow required to protect pump from damage.

Pl, commment.

In absence of the pump vendor data during initial design, the thumb rule to design the spill back line for 25-30 % of the pump rated capacity.

Thanks in advance,

Timewrap.
 
I'd go to 50%. No sense making any kind of design where a silly little recirculation line could potentially bottleneck your operations.

Neither of those two flowrates will protect the pump from damage, if you consider damage to include excessive maintenance directly attributable to sustained operation away from BEP.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)
 
We build small systems where low flow/high head single impeller pumps operating on the left-hand-side of the curve are the norm, and operating efficiency is not a concern. The luxury of operating at 50% or more of BEP is seldom available to us. BigInch's advice doesn't apply at our scale.

Under these conditions, spillback lines are the norm, and the pump becomes basically a pressure source.

No, we don't tear up pumps every day. They work fine- if minimum flow and heat accumulation are addressed in the overall design. On these small units, minimum flows of 10% or less of BEP flow are typical.

Spillback control usually consists of a manual globe valve, with or without the handle installed once the spillback setting has been adjusted. Sometimes it's just a piece of tubing sized to give the minimum flow, and long enough to give you the heat loss you require. Sometimes it's a self-operated back-pressure regulator.

Larger pumps, with minimum flows a significant fraction of BEP, get self-operated spillback control valves or some other automated control.
 
I don't think it was bad advice, given this is a chemical plant forum.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)
 
Or somewhat differently we may say that,

Almost all of the centrifugal pumps require removal of a minimum amount of kinetic energy and friction generated thermal energy;via a minimum flow line to the bulk discharge pumping system or back into the suction side bulk system.

However the damages and penalties of not doing this and their extent;varies from pump to pump according to the pump design,No. of stages,pumping fluids types,service and pump's characteristic curve normal operation range.

This has been my understanding. Hope this helps

Best Regards
Qalander(Chem)
 
BigInch: I never said what you were offering was bad advice, especially considering the OP talked about Suncor as an example of a client with a particular way of doing things. I can't imagine much that they have being "small" on the order I'm talking about.

It's just not a one size fits all world out there. People need to understand some fundamentals and do some engineering to get it right. I replied to the OP in that spirit.
 
Moltenmetal,

Just for information.
Once I did a project with Suncor Canada, all the pumps, big and small in terms of flow-rates and head require a min. flow line. This line was installed with a flow control valve with opens up when there is low flow in the discharge line.

Timewrap.
 
Biginch,

I've just checked my old P&ID's, it was a Yarway valve for small pumps and a flow control (control valve with a flow meter) loop for big pumps.

Timewrap.
 
Suncor's engineers are top shelf. The reason for a minimum flow recirc is to keep the pump from damage when the discharge flow falls below the minimum flow specified by the pump vendor. The minimum flow is not only crucial for turndown conditions, but it's also good for the standby pump when they need to take down the pump that is running.

Some companies cheap out on this for centrigugal pumps but it's short sighted when comparing the maintenance costs involved over the life cycle of the pump.
 
Biginch's comments look good. You might be able to adjust the ranges slightly based on the type of pump. i.e. Off BEP will add lateral loading to the shaft and vibration in addition to heat. Single volute are worst and will generate most lateral load. Min flows should be not much less than 60%. Double volute have less radial load. Radial diffuser are best. However, the bottom line is what the manufacturer will guarantee. Some are more fussy than others.
 
Status
Not open for further replies.
Back
Top