Whether A Pump with Gear Box (to increase pump speed) is Feasible?

[ICECOOL]

Dear Sir,
I am trying to understand that whether a Pump who is flowing a High Temperature Fluid (Therminol) will be feasible with a Gear Box. The Gear Box will be used to increase the pump speed from 3000 rpm to 3600 rpm.
The reason behind increasing the speed is high Pump TDH and as per manufacturers it is not in their configurations. As per my understanding the increase in pump speed will help into get higher TDH. I have enough NPSH available.

Any help is highly appreciated. Pump parameter are mention below

Flow rate 1100 m3/s
TDH : 321m
Temperature : 300 Degree C
Discharge pressure : 36 bar
Country : India

Thanks,

 

[russ_watters]

Two more common options:

Install a different motor, that has a different nominal speed.
Install a variable frequency drive.

 

[ICECOOL]

Dear russ_watters

1. Installing a different motor is not an option as the frequency is 50 Hz and number of poles is 2.
2. Installing a VFD or any frequency convertor is an alternative but I concern about the O&M point of view. if VFD has a problem then pump is not workable. As per my understanding VFD is normally not used for the application I have mention above.
I will prefer to have a device more mechanical then electrical.

Thanks

 

[bigaggie]

Are you sure on that flowrate? There's only a handful of pumps in the world that can push that amount, and last I checked none of them ran at 3000 rpm.

In my experience, you have 3 options.

1) Gearbox. Yes, it is done, but normally not for something like this. I usually see it in instances where the motor is 3000 (or 3600 for 60 Hz), and the desired operating speed is in the 4-6000 RPM range. However, doing a retrofit is not easy, nor simple. If this is a vertical pump, I've never seen it done. If it is a horizontal pump, you will need to remove the current baseplate, design a new one, install the gearbox, and then tie it into whatever lubrication system is currently in place, or install a new lubrication system if one is not in place. Very expensive.

2) VFD: Your argument against a VFD is somewhat invalid. If the gearbox goes down, the pump goes down. And honestly, VFD's are probably more reliable and cost effective given the small speed change you're considering.

3) Hydraulic Re-Rate: I'm honestly not sure if this will work or not, as I don't know what kind of pump you have, nor its exact dimensions. But this is probably the lowest priced option, and I would submit that it's the easiest to implement. If your casing can handle it, a larger impeller (or impellers) would accomplish your goals. You should contact a pump manufacturer and ask if their aftermarket engineering group would be willing to investigate.

 

[Travis_King]

The most common way of increasing the speed of a pump from it's motor is with a sheave and belt arrangement. For this case, you only need a 1.2:1 ratio (big sheave on motor, smaller sheave on pump) to get to the desired speed. Likely this will be less expensive than a gearbox, as the desired change is so small.

Be sure to use proper engineering practices and calculations when designing the pulleys. You want to ensure the correct number of belts are used, and that they are properly tensioned. You'll also want to ensure that the belt arrangement is properly guarded so that no one can accidentally get caught in the rotating parts.

If you're set on a gearbox that will work too.

That's not to say that the pump will function as desired, though. At high temperatures you are always at a greater risk of cavitation; be sure to discuss this change with the pump vendor/manufacturer. Also, be sure to check whether your motor is sized adequately for the new operating point. Higher head means more power.

Some additional points:

1. VFD's are very good for slowing pumps down, but speeding them up using a VFD isn't usually done as they speed up the motor, not the impeller. Usually they're capable of overclocking the motor by 10%, after that you are running into the danger zone.

2. I agree with investigating options for a larger, or differently designed impeller for your pump casing. A vendor should be able to tell you whether what you want is possible or not fairly quickly.

 

[russ_watters]

1. VFD's are very good for slowing pumps down, but speeding them up using a VFD isn't usually done as they speed up the motor, not the impeller. Usually they're capable of overclocking the motor by 10%, after that you are running into the danger zone.

Most owners I've dealt with don't like it, but most vfds are capable of +100%. I've talked extensively with motor and VFD vendors about this and there really isn't an issue here much different than a pulley change: either way you have to ensure you haven't exceeded the nameplate power.

There may be an even easier way out though: he said he's on a 50hz electrical system: so install a 60hz motor and clock the VFD to it. Heck, a multi-rated motor may already be rated for both.

Also regarding redundancy: you can do redundant vfds, but on a non-redundant pump and motor it seems rather silly.

 

[Travis_King]

I was thinking the same thing when I read 50 hz, if you get a motor and VFD rated for 60 hz (3600 synchronous speed) you'll be golden.

I still feel like a sheave setup is a much easier and cheaper solution though.

 

[Travis_King]

I'm also curious as to this application. If we assume it's 1100 m3/hr and not m3/s (which, as bigaggie suggested is approaching 17.5 million gallons per minute), then at 321 m you're looking at around the 1250 hp range for an SG of 1. This isn't small potatoes, any pump that can handle this will likely draw a lot more power when you run it at 3600 rpm as opposed to 3000.

 

[russ_watters]

I still feel like a sheave setup is a much easier and cheaper solution though.

I don't think I've ever seen a belt driven pump.

 

[Travis_King]

I don't think I've ever seen a belt driven pump.

We use them all the time in mining and minerals processing. Processes often change, and sheaves are easy and super cheap (especially for the bigger pumps) to replace and re-size. In fact, I'd say the vast majority of pumps and fans that I've worked with in the last few years were belt driven.

Though, none of those pumps were of the size I would expect this behemoth to be...

 

[ICECOOL]

Dear All,

Thanks for the knowledge sharing.

Basically, This is the pump which will be installed in future project. The pump is Horizontal, Single stage and flow rate is not in Millions :P
The pump motor is approximately 900 kW.

one of the option suggested by you guys is "Hydraulic Re-Rate"..It doesn't work because even in the max limit pump will give only 280m TDH with 1100 m3/hr flow rate.

I have the same thought about VFD as Travis_King says. But I may need to investigate it more.

The plant where I am installing the Pump is Solar thermal plant. and there is recycling everyday. Therefore the chances the pump should be very robust.

The last option which I am thinking of using 2 pumps in parallel instead of using 1 pump with such a high flow rate. But that will give some tough time to control guys for balancing the flow rate.

I am still searching and finding a resolution. But I am bit interested to see if somebody else faces this kind of problem before..

Thanks

 

[Travis_King]

I just want you to be clear that if you want to pump that same volume at 321 m as opposed to 280 m you will be increasing your power by roughly 120 kW. It's not merely a matter of speeding the pump up, you need the power too.

As far as parallel pumps go, are you saying that the pump you have is able to operate at the desired 321 m TDH but not at the correct capacity? If the pump can't reach the 321 m at any point on the curve, then parallel pumps won't help you.

 

[ICECOOL]

Yes, increasing the TDH will increase the motor rating as well.

Yes, pump will be available when I decrease the flow rate. But I am not agreed with your comment that parallel pump won't help me.

The pump curve suggest that the pump can have better TDH at lower flow rates. Therefore, if I half the flow rate then it might be in my spec. But it will increase the number of pumps which create additional problem in controlling.

I am open for your comments and suggesitions

 

[Travis_King]

Yep, re-read my comment, maybe it wasn't clear. I was just making sure you knew that adding two pumps wouldn't help if neither pump can develop enough head on it's own at some point on the curve. As you suggest that your pump can develop the necessary head at a reduced capacity, parallel pumps are a viable option.

As you mention, though, adding an additional pump adds complexity to the system for controlling and balancing flow. Remember also that if they are pumping into the same discharge line, the piping will have to be designed to reduce friction losses (and thus TDH) at the full, combined flowrate.

For something of this size though, I really recommend simply finding a pumping solution that fits your needs. Trying to operate a 1000 hp pumpset outside of it's design conditions isn't exactly good practice...

Just fyi:
In mine dewatering we sometimes use numerous multi-stage centrifugal pumps to get high flows and high heads. Instead of one or two pumps, we may use 4-5 in parallel with discharge piping balanced by design for easier control. High head and high flow make for costly pumps. As with anything in engineering, it's about optimizing all the variables. If you've got 4 small multistage pumps operating at 1/4 capacity at the design head, and one goes out, you've still got ~3/4 capacity. If you have one gigantic pump and it goes down, you better get it fixed again. Sometimes it's worth it, sometimes it isn't.

 

[bigaggie]

Yes, increasing the TDH will increase the motor rating as well.

Yes, pump will be available when I decrease the flow rate. But I am not agreed with your comment that parallel pump won't help me.

The pump curve suggest that the pump can have better TDH at lower flow rates. Therefore, if I half the flow rate then it might be in my spec. But it will increase the number of pumps which create additional problem in controlling.

I am open for your comments and suggesitions

I cannot begin to tell you how bad of an idea this is.

Really. Very. Bad.

If you run a pump like this back on the curve that far, you will almost certainly experience higher vibrations, most likely due to discharge recirculation, and that could have a significant impact on pump reliability, which is crucial in power plants.

I'm assuming that this is a pump you already own or possess. It's a little hard for me to tell because your wording is odd (no offense meant, I know English is not your first language, nor the easiest to learn). You said that in its "max limit" it can't reach it. I'm assuming (again) that this means that with the maximum diameter impeller it will not reach the TDH you want at the desired flow.

Here's what I'm suggesting by "re-rate". The manufacturer has curves that are standard to that pump model. He knows what impeller it has in it, and he knows what it will do. He has other impellers as well. There are probably multiple impellers for that model line with a wide range of flow and head rates. You should consult with the manufacturer and ask if they have any other impellers for this model. If they do not have another existing impeller for this model, they may be able to design a brand new impeller that will accommodate your needs, most likely by increasing the width of the discharge of the impeller.

It looks like this is a single stage, double suction impeller, which in general is not that difficult to design. Just from my rough calculations, your target specific speed at 3000 RPM is around 1200, which is about what I'd expect for that pump style. The real challenge will be designing it to hit the desired operating point while not raising the NPSHR above what is available, and also fitting within the existing casing, though normally these casings can be easily modified to accept larger impellers. Hard for me to give much more recommendations without knowing specifics of the pump.

 

[Travis_King]

I cannot begin to tell you how bad of an idea this is.

Really. Very. Bad.

If you run a pump like this back on the curve that far, you will almost certainly experience higher vibrations, most likely due to discharge recirculation, and that could have a significant impact on pump reliability, which is crucial in power plants.

First, I agree.

To be fair though, it really depends on the curve itself and where the pump sits on it now. If he halves the flow rate and it brings him a little to the left of the BEP (Best Efficiency Point, for any wondering), there's nothing wrong with that. If it brings it well up the curve toward the minimum continuous flow point, then yea, you're operating terribly.

OP, if you want to use two pumps in parallel, then they need to be spec'd for that purpose.

 

[bigaggie]

If the curve is relatively flat, then yeah you really shouldn't have recirculation running back a little bit (though I've seen it before). Most manufacturers allow for you to run about 70-110% of BEP. But he's talking all the way back to 50% (assuming he's at 100% now). Even on a flat curve you're looking at 10-20% efficiency loss, if not more. And vibrations will still increase because the flow won't match the inlet geometry of the impeller. Maybe not as bad as a steeper curve, but they will go up.

 

[Travis_King]

If the curve is relatively flat, then yeah you really shouldn't have recirculation running back a little bit (though I've seen it before). Most manufacturers allow for you to run about 70-110% of BEP. But he's talking all the way back to 50% (assuming he's at 100% now). Even on a flat curve you're looking at 10-20% efficiency loss, if not more. And vibrations will still increase because the flow won't match the inlet geometry of the impeller. Maybe not as bad as a steeper curve, but they will go up.

Yep, agreed.

I've done it before, modifying existing pumps to fit a new service just to "get this $%!# thing working", even where they were running at 50-60% of BEP (still safely to the right of MCF). They were inefficient, the tended to grumble and make a bit of noise here and there, but they worked.

On a 1000 hp pump? I wouldn't dream of it.

OP, This is a time when you've got to get with the vendor and see if you can work out a solution. You may well be able to run that bad boy at 3600 rpm through a gargantuan gearbox, but that is something you'll need the vendor to help with. Pumps (and equipment in general) of this size are not something you can cobble together an hope for it to work.

I hope we've offered some useful information, anyway.