Improving pump performance - Pump engineering (2025)

Hi,

I am working on improving efficiency of centrifugal water pump. My pump was designed at 20 lt/min running at 5500 rpm with 6 impeller blades(closed impeller). Do you think there is any benefit of changing from closed impeller to open impeller. I am looking in to different factors like 1. smoothening the blade profile 2. changing the length of the blade 3. Changing the number of blades to 5 and 4. modifying the impeller type.

I have not worked on pumps before. Your help will be appreciated.

Thanks.

Open impellers have their definite advantage but poor on efficiency when compared with closed impellers. Changing the impeller dia. (blade length?) and no. of blades effect the pump performance in other ways as well. There are many threads in this forum that discuss about pump geometry.

JRLAKE has given you a practical advice. I will not overlook it.

BTW, what kind of savings you foresee for a 1.2cu.mtr/hr pump?

You might want to give Discflo a call. I've seen several of their pumps in lime slurry as well as other slurries.

Better to select closed vs open impellers on what has to be pumped through the system. Its not an efficiency question.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

Its more important to know what,
1.) product you are pumping,
2.) what the system curve looks like,
3.) what the pump curve looks like.
4.) what the min and max demands are

5.) mfgr and model number, impeller of the pump you've got.

Then you can see if the impeller diameter, or rpm or something else needs to change.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

1. It is a centrifugal water pump
2. The pump is expected to operate at two operating points. i. 3000rpm with a flow rate of 5lit/m, head 5m(ii). 5500 rpm with a flow rate of 20lit/min expecting 7.6m of head.
3. Please find the attached pump curve.
4. The impeller is of 53mm in diameter.

Please also find the pump's picture.

This pump won't work for both flow points.

The flow capacity is much too small to reach 20 L/m at any speed. 5500 rpm seems high for this pump too.

If you could get it to run at that speed, you might expect a flow of 9.2 L/m and a head of 17m, which would have to be cut back significantly, losing lots of efficiency, noting that you are still nowhere near 20 L/m.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

3000 rpm, or just under, is pretty much flat out for an AC mains powered pump in a 50Hz part of the world. In North America and other 60Hz regions, flat out is about 3600rpm or so. A speed below these maximum values is relatively easy to contrive. How do you propose to reach 5500rpm?

----------------------------------

If we learn from our mistakes I'm getting a great education!

2:1 gears

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

Yes, that's obviously an option but why not design the impeller to run at the direct shaft speed if possible? Other than keeping mechanical engineers in work designing the extra gearing, and adding more parts to break...

----------------------------------

If we learn from our mistakes I'm getting a great education!

He can't get his flow requirements at normal speeds and still have an impeller that fits inside his tiny pump. He either needs a bigger casing or very fast speed. Bigger is apparently not an option.

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

I should have asked: what is actually driving the pump? Some types of motor are more suited for being overdriven to high speeds. Is the impeller designed for operation at such a high speed?

----------------------------------

If we learn from our mistakes I'm getting a great education!

Exactly Scotty, there's a whole lot to consider before "gearing up",
the bearings, NPSHR, pipe discharge pressure allowable ...

"If everything seems under control, you're just not moving fast enough."
- Mario Andretti- When asked about transient hydraulics

A pump of this size will be inherently inefficient anyway and if from a reputable manufacturer you could expect the design to be maximised to give the maximum efficiency available.

However the question was 1.)open / closed impeller and
2)changing blade shape, length and number etc.,
Open impeller depending on pump configuration can be as efficient as closed as open offers the opportunity for all blade surfaces be highly polish.
Changing blade shape and configuration will change performance and not necesarily efficiency.

This request should be sent to your pump vendor. There are many variables to consider when going from and closed to open impeller.

Typically axial (open) impellers have a higher specific speed, and the impeller/volute design allows for a longer, smoother transition. Also open impellers run against a liner, which also improves efficiency. Placing an open impeller into a volute or bowl designed for a low specific speed closed impeller would probably have the opposite affect you're looking for.

The number of vanes in an impeller determines the slope of the HQ curve. A 7 vane impeller will have a flatter curve (less % rise to shut off) then a 5 vane impeller. Be careful if you start experimenting with impeller vanes... industry standards like API-610 have a min rise to shut off of 10%.

I would recommend polishing the impeller, smoothing out the inlet vane profile, and backfiling (under filing) the impeller. You can also coat the bowl/volute for improved efficiency. Also, BEP flow is a function of impeller dia. A larger impeller will shift BEP to the right, smaller dia to the left. Know where your rated flow is relative to BEP, and that's something else you can play with.

dabluffrat

Looking at my vane profile, smoothening the blade profile and IGV profile and underfiling are the things I should do first and see if there is any improvement. Thanks a lot for your valuable suggestions.