Is it better to throttle back a pump or presure relieve it?

[Harry_Fish]

I have a Pan World 150PS that has too much flow for my needs

it it better for the pump to:

Turn close down the output with a ball valve

or

To "T" off the output and send the Excess flow back to the sump?
(using a T and a second valve to controll the excess flow?)

[MeuserReef]

I have always been told that if you throttle the pump back, it will reduce the wattage consumed by the motor on the pump and will not harm anything on the pump.

Perhaps some others will chime in to reaffirm this.

[doug6644]

I have to believe, throttling a pump would create more "head " pressure on a pump, therefore consuming more "wattage", or amp draw. An electric pump is not built to throttle, it is made to force water, no matter what the circumstances are, it is going to fight the head pressure, and in turn, using more energy to push the water. On all the pumps I have used (at work), we bypass water before we throttle it back, it helps with pump life at the least. The less a pump has to fight to pump, the better off it will be.

[Harry_Fish]

I guess that is my thought as well.

It seems like more resistance would be more stress

Since I am building up a manifold to run other items.

I will have two points I can adjust the flow

One by turning it down the other as a relief so I can do it both ways

[biecacka]

some pumps like the darts, snappers or others actually perform better when throttled back according to the manufactuer. they also use less wattage this way i have been told. i am not sure with the pan world though i would contact them to make sure. hth

[doug6644]

And not to start a huge debate or argument, if I was a manufacturer. Yes sure throttle our pump, and then don't say it creates more heat and stress on the pump, causing premature failure, and we can sell another pump.

[Mike31154]

The only way I'm aware of to throttle a pump back so that it consumes less power is to have an electrical control to slow it down. I'm of the same mindset as previous posters on this thread that increasing resistance against the rated flow output will cause more stress on the dynamic components (bearings, shaft, armature) of the pump, thereby consuming more power.

I suggest that you're on the right track by teeing off the output and returning the excess flow to the sump or using it to run other components.

[doug6644]

And let me clarify, I would guess the stress would be placed on the motor not the pump(impeller). The impeller will only bring in so much water, then it has to fight to push the water that it is bringing in behind it. Therefore, if the water is not leaving the pump unrestricted, it will create more "head". I guess I do not see how a restricted pump, can be more efficient than an unrestricted pump.

[doug6644]

TY Mike, the only way I now how to reduce the amount of energy used in a pump, would be to use a VFD (variable frequency drive). And the only way I am aware a VFD comes into use, is when the flow itself increases and decreases.

[AcroSteve]

Doug, you are not correct. Look at many of the performance curves on some pumps and you will see that the wattage drops as head increases.

From my post in this thread.

http://archive.reefcentral.com/forums/showthread.php?s=&threadid=352943&perpage=20&pagenumber=3


As far as a pump drawing less power at a greater head, I initially had a hard time believing the curves untill I asked around work related to much larger pumps and came to this explination.

At the maximum output, I consider the pump "hooked up" and not exceeding the shear capacity of the water. Therefore the motor sees a greater load.

With increased head pressure, the pump must "slip" and therefore the shear capacity of the water is exceeded. This is similar to the difference between static and dynamic friction. Once the static friction is overcome by exceeding the shear capacity of the water, the dynamic friction of the water bypassing is less than the static friction was. Therefore the load on the motor is reduced.

I realize I have kind of jumbled the terminology here, by this is how I explained it to myself.



So... I say use a valve and cut it back. Gate valves work great for this.

[kcress]

Restricting centrifugal pumps does not stress them in any way. Fear not!

Yes you increase the head as you restrict their output but, BUT, at the same time, you reduce the flow. A pump is a device that takes electricity and moves water with it. If you reduce the amount of moved water you restrict the power required. The load on the pumps motor is reduced, the current drops, the power consumption drops, the pump works less. It will last longer.

This is the same thing that happens with a vacuum cleaner and is a common reason for the mistake of believing the motor works harder. If you restrict a vacuum's inlet the vacuum gets much louder and more frenetic. The assumption is "it's working harder". Nope! It is suddenly unloaded and so the motor speeds up - sounds louder.

In a perfect world your pump is not throttled at all and provides precisely the perfect amount of flow you need. This is the most efficient situation. Throttling a pump is less efficient, but it still reduces the power consumption.

So, remember, a centrifugal pump is a 'mass flow " device. It moves water,(mass). If you reduce the water that flows you reduce the 'mass flow' and hence the power requirements.

The one caveat is that you must not block the flow completely or over time you can over heat the water that is stirring around the impeller and so damage the seals.

[Memnoch21]

I don't mean to Hijack the thread here, but I've been wanting to do the same thing as Harry_Fish for a while now.

Both sides of the argument here are making sense to me, at least to some degree... Does anyone have any *real*, real world numbers? ie: plugged into a Kill-o-watt or the likes. Also, would a different style of pump change anything said? For example a QO 6000?

Thanks!
Chris

[Reefugee]

Here's an excerpt from one of my other posts. The first picture was the pump running from my sump, through a chiller, then up about 4 feet through about 6 ft of 3/4" flex hose back to my tank.

The second one was the pump in the sump just pumping water with a very short section of hose back into the sump. Basically - no head pressure.

This was one quick test, so by no means should it be considered a controlled test. If anyone is interested, I am willing to try the same test again.



Ocean Runner 3500 return pump. Since there is quite a bit of head room, I considered this a "throttled" pump.
http://i16.photobucket.com/albums/b37/mn95616/120GRebuild/oceanrunner3500throttled.jpg


I wondered what the watt usage would be if the pump just ran with nothing to throttle it back:
http://i16.photobucket.com/albums/b37/mn95616/120GRebuild/OceanRunner3500notthrottled.jpg
Hmmm.. according to AquaMedic, these pump on consume 65 watts.

[Memnoch21]

Thanks! That is EXACTLY what I was looking for. :)

Chris

[daveonbass]

wow that's awesome news. I have been wonderign this same question. I'm glad that someone has finally answered it. Now I don't feel bad putting valves in at each of my CL's out flows.

thanks for all the info.

[kcress]

To restate it: Throttling down a centrifugal pump does not harm it. Nor does it wear out its impeller. Nor does it over heat it. Nor does it damage its bearings, chemically or otherwise. This is absolutely true as long as the flow is not completely blocked.

Onward to the demo.
Here is the simple straight forward setup. It's a magnetic drive pump. An 1100gal/hr Little Giant pump made specifically for aquarium service.

A standard schedule 80 bulkhead was used and all piping and the ball valve are 1 inch which is the same as the pump inlet and outlet.

http://i32.*******.com/20pvoet.jpg




Close up of the pump.

http://i27.*******.com/v65heg.jpg




A close up of the excellent wattmeter that was used.

http://i28.*******.com/2i27l1x.jpg





The unrestricted bulkhead fitting.

http://i30.*******.com/fp7b86.jpg





Here we go. The system running with the ball valve wide open.
The pump is consuming about 100W of power doing maximum work, which is
moving about 1100gals/hr of water. Note the solid column of water returning to the tank. 103.8W

http://i32.*******.com/122gh0h.jpg






Now for some throttling. The ball valve is about half closed.
Notice the flow has dropped substantially. Notice also the power has not risen. The pump is not working harder.. It is working LESS. The power has decreased. 72.7W

http://i31.*******.com/o08at.jpg




Now the valve has been closed about 25%. Power has dropped still further.
57.9W

http://i32.*******.com/29px4py.jpg




Here I have closed completely, blocked, the flow. Notice the power has dropped further yet! Of course it is not zero as there is a fan being run, bearing friction is present. Windage is occurring in the motor. There is resistance in the windings and there is some recirculation occurring in the pump head. Over time - minutes - the pump head would get warm. If there was the slightest trickle of water allowed thru the pump there would be no measurable heating as water can carry a great deal of heat away.
47.7W

http://i28.*******.com/2nrkj11.jpg


I hope this helps you shed a very common misconception.

Get the closest sized pump you can.
If it over runs your return system, you can put in a valve and throttle the flow. Pay less to the power company and put less carbon into our atmosphere.

Or, divert to your sump which marginally adds a little more heat to your aquarium's water and stirs your water a little more.

It's your choice, but now you can make it based on facts.

Cheers.

[jdieck]

As you trottle the pump, the head increases and the flow decreases following the pump's performance curve...

http://www.engineeringtoolbox.com/docs/documents/310/pump-regulation-throttling.png

... and as you trottle the pump (This is for all centrifugal pumps) the amount of power will be reduced as posts above demostrate.

http://www.engineeringtoolbox.com/docs/documents/310/pump-regulation-power-efficiency.png
Power consumption comparison between bypassing, trottling and changing the pumps's RPM

But... Just a note of caution.
Depending on the pump's performance curve, most pumps have a point or combination of head pressure flow for which efficiency is higher, in other words the power consumed per unit of water pumped is the lowest. (Best efficiency point or BEP)

http://www.engineeringtoolbox.com/docs/documents/635/pump_system_curve.png
The concentrical lines are the efficiency curves for a pump with a given impeller diameter. In this case the center of the circles is the highest efficiency. (BEP)

When the pump runs at a higher point in the curve by trottling (Higher head lower volume) the efficiency lowers as the dinamic loss (Increased impeller/water slippage) is increased.
When you move in the direction of lower flow higher head although the total power is reduced, the efficiency drops faster than the power consumption.

See this picture that shows a typical power drop when moving to the left by trottling (Red) versus the efficiency drop (Blue)
Note that the same happens when you move to the right, like when running the pump with no head at all.
http://www.engineeringtoolbox.com/docs/documents/313/pump-power-efficiency.png

This means that the power consummed per unit pumped will be higher which means increasing the temperature. This shows what happen to temperature when trottling.
http://www.engineeringtoolbox.com/docs/documents/313/pump-power-throttling-temperature-increase.png

Usually this increase in temperature is not an issue as the pumps are designed for it but specially if the body is cooled by the water (Basically magnetic driven pumps) the water temperature increase is more significant so if after trothling you notice a temperature rise in your aquarium water and this is higher than what you can manage with your evaporation or your chiller then a bypass will be an option or a combination of bypass and throttling.

[Fiziksgeek]

What an amazing thread...haha. You got not only a detailed theoretical answer, but experimental results to back it up as well.

Just a thought, but can you T-off the return for other purposes? Feeding a skimmer for example? It solves your excess flow problem and perhaps allows you to eliminate another pump in your system.

[Harry_Fish]

Thank you!

This is a far more detailed answer than I expected.

[jdieck]

:thumbsup:

[spsfreak]

Very good thread, nice to see some info\tests to back up the scenario we have all heard of before - that is throttling back a pump.

[psteeleb]

generally speaking (volute with impeller type pumps) you should have some back pressure, or the pump would be unconstrained, and will spin faster then it is optimally designed. With no back pressure, spinning faster then it's optimal design, will create excess heat, use more energy and leads to seals wearing out faster. Generally speaking the higher the pressure the pump the more this comes into play.
Our aquarium pumps are generally low pressure so it probably doesn't really matter

Power head type pumps and propeller pumps - let them run full blast.

[kcress]

jdieck; Thanks for that nice efficiency round up! Very nice.

Fiziksgeek; Of course! What you need to get from this is that centrifugal pumps for aquariums are very forgiving. Messing with the head pressure won't hurt them it just changes flows and efficiency.

psteeleb; Close - but not entirely correct... With some larger pumps the pump and impeller are designed to provide a certain flow and hence require the appropriate horsepower to run at that "point on the curve". If the flow is allowed to increase for any reason, (larger pipe, reduced head, etc.), more work is being done as more 'mass is being moved' this will require more horsepower. The motor will be overloaded and the motor will fail.

This may give the impression the seals were over worked but what really happens are the seals are cooked by the burning up motor.

On larger pumps the motor will often be carefully sized as a larger motor doing less than nameplate work is badly inefficient. This opens the way for overheating if more flow is allowed.

This is very acute in air flow situations. If you look at any furnace or air conditioning air circulating blowers they will almost always state that "you must have x back pressure", otherwise the blower motor is toast.

[stugray]

kcress & jdieck,

Thank you! Great experimntal results & explanations.

In my basement Sump using a Seq baracuda, I use BOTH methods ( recirc back to sump through manifold ) & a throttling valve.

Now IF I just had one of those fancy Kill-A-Watt meters, I could find the pumps "sweet spot" by adjusting both valves and adjusting for my setup.

As long as you do not throttle a pump back beyond the rated "max head pressure" you will be fine ( a closed valve on the output LOOKS like extra "head" to a pump ).


Oh and on a side note: I personally watched as a neighbor turned on a Hottub with the gate valves closed. After only 3-4 minutes, The pump BLEW a 2" Pipe fitting OUT OF THE INTAKE SIDE OF THE PUMP without breaking the fittings.

My only explanation at the time was that excessive cavitation actually BOILED the water. There was a cloud of steam when it popped.

Stu

[jdieck]

<a href=showthread.php?s=&postid=14264517#post14264517 target=_blank>Originally posted</a> by kcress
psteeleb; Close - but not entirely correct... With some larger pumps the pump and impeller are designed to provide a certain flow and hence require the appropriate horsepower to run at that "point on the curve". If the flow is allowed to increase for any reason, (larger pipe, reduced head, etc.), more work is being done as more 'mass is being moved' this will require more horsepower. The motor will be overloaded and the motor will fail.

This may give the impression the seals were over worked but what really happens are the seals are cooked by the burning up motor.
Building on this explanation, in addition to the potential for motor overload there is a second potential effect that adds to the problem.
As the flow increases the pump requires a higher inlet pressure (With the fancy name of NPSH, Net Positive Suction Head). If this additional pressure is not provided a vortex of negative pressure (Vacuum) forms just at the inlet of the impeller. (This is what is called cavitation). Cavitation creates bubbles that will hit the tips of the impeller and the volute at an incredible speed wearing and pitting the impeller and volute very quickly. On top of this this cavitation creates a lot of vibration which tied to overheating of the motor will destroy seals and bearings in a snap.

This effect in aquarium seems to be very common in closed loops were the inlet and outlet pressures are almost the same so the pump has an almost negligible head at the discharge. To prevent cavitation in this cases it is important to insure that the inlet pressure is not diminished by using the largest practical piping diameter for the pump suction and by having a valve at the pump's discharge that can be partially closed as required to prevent this cavitation effect by increasing the pump's discharge head.

[jdieck]

<a href=showthread.php?s=&postid=14264611#post14264611 target=_blank>Originally posted</a> by stugray
Oh and on a side note: I personally watched as a neighbor turned on a Hottub with the gate valves closed. After only 3-4 minutes, The pump BLEW a 2" Pipe fitting OUT OF THE INTAKE SIDE OF THE PUMP without breaking the fittings.

My only explanation at the time was that excessive cavitation actually BOILED the water. There was a cloud of steam when it popped.

Stu
Stu:

Did he also turned on the in line heater? I think with no flow the pressure build up by the heater boiling the trapped water might also be a posibility.

[stugray]

jdieck,

"Did he also turned on the in line heater?"

No..... Power was applied to the system for the first time ( full of water of course ), pumps turned on & blew in less than 4 minutes.


I havent jumped a fence since I was a teenager until then ;-)

Stu

[jdieck]

<a href=showthread.php?s=&postid=14265527#post14265527 target=_blank>Originally posted</a> by stugray
I havent jumped a fence since I was a teenager until then ;-)

Stu
:lol:

[~sp0ok~]

good info

[kcress]

Stu; I never actually thought of tuning an aquarium setup for maximum efficiency with a wattmeter. Great idea.

jdieck; Thanks for that further detail. I was always focused on not toasting the motor - not stopping to consider the inlet issues of an 'overloaded' pump.

[doug6644]

So, the theory still comes back to the fact that, too much head does create overextension of the pumps ability. It is better for a pump to run unrestricted than to throttle it, but tests also prove that fewer watts are consumed. Is the amp draw on the motor also reduced? Amp draw x volts = watts consumed, therefore, you can use more energy by throttling a pump too far. But, you may also, throttle it to the "sweet spot", and consume fewer watts. Wattage versus pump replacement.

[doug6644]

You stirred the pot Harry!!?? I tried to avoid this, hehe.

[jdieck]

<a href=showthread.php?s=&postid=14267731#post14267731 target=_blank>Originally posted</a> by doug6644
So, the theory still comes back to the fact that, too much head does create overextension of the pumps ability.
Nope, the fact is that throttling consumes less power and this means the motor require less cooling despite the water temperature increase by a couple of degrees, bad for the aquarium that is limited in cooling but in reality the pump and motor could care less as they are still way below the maximum design limits.
Of course there is a limit to everything. You do not want to run a pump completely throttled (closed=no or very limited flow which BTW defy the purpose of pumping). In this situation the water inside the pump will not get replaced and will keep on heating up and build up heat until evaporating and drying up the water lubricating the ceramic bearings on magnetic drive pumps or the seal on shafted pumps with the equivalency of running them dry.

It is better for a pump to run unrestricted than to throttle it
Define unrestricted. As mentioned before, totally unrestricted (no head at all) can overload the motor because the power consumption increases. Some motors may not take the overload kindly while on the other hand properly sized motors may not be afected at all and just run a bit hotter but, remember they may still be subjet to potential cavitation.

but tests also prove that fewer watts are consumed. Is the amp draw on the motor also reduced? Amp draw x volts = watts consumed, therefore, you can use more energy by throttling a pump too far.
How can you use more energy by reducing the power? Not possible.
In electrical systems energy consumption is measured in watts, as the power (watts) drops, consumption (measured in kilowatt-hours) necessarily drops as the hours are constant and the watts drop.
As the watts drop also the amperage drops as the voltage remains constant.
In summary there is no way you can consume more overall energy by throttling the pump.
Having said that, your efficiency drops which makes the pump use more power per gallon pumped this only means that although the total power drops, it does not drop as much as the volume drop because of the loss of efficiency.

But, you may also, throttle it to the "sweet spot", and consume fewer watts. Wattage versus pump replacement.
Not necessarily. The power consumption at the sweet spot will most probably be higher.
To determine the "sweet spot", defined as the point at which the efficiency is maximum you need to measure not only power but also flow.
Measuring power and flow at different points of head pressure and dividing the power by the gallons per hour, you can determine the point at which the consumption per gallon pumped is the lowest, again it is power per volume pumped not power alone what defines the sweetest point. In other words, lowering the head from one point to another may increase the volume by 30% but may increase the power by only 20%, in this case you will be using more total power but the pump will be running more efficiently because you will be using less power (Only 20% increase) per volume pumped (30% more)


There are many other variables that I have not mentioned for the sake of simplicity things like motor efficiency, power at the shaft of the motor vs power taken from the power lines vs power output from the pump in the form of flow and head; impeller diameter and RPM, open impeller vs closed impeller, volute design etc.

Also effects like an increase in the water temperature inside the pump can lead to calcium carbonate precipitation making it necessary more frequent cleaning to prevent the incrustations ceasing the pump.

Also a quality motor has to be properly sized to the pump capacity along the performance curve from zero head (maximum volume) to maximum head (zero volume). Some pumps will run hotter than others. Some manufacturers, taking advantage of the better insulating materials and varnishes for the motor windings, select motors that are undersized with the confidence that they can take certain amount of overload of course in this cases the life of the motor will be reduced although probably still beyond the length of time we will use them.

Finally something I have always be weary off is the claims of many manufacturers about power consumption.
In my book there is no valid claim unless they publish a performance chart that not only include the flow versus head pressure but also the power consumption at the different points along the curve as well as the overall pump motor combined efficiency. Motor horse power capacity or wattage taken means nothing by itself.

[liveforphysics]

The less water pumped, the less load on the motor.

The power used by a motor is directly related to phase angle. The greater the torque, the more retarded the phase angle, and the more power the motor draws.

A blocked outlet allows water to simply spin in the pump volute. This spinning water maintains a good degree of momentum, which causes it to not need as much work from the pump as it would require to get fresh incomming water spun up to speed.

Because of this, less torque is placed on the motor, which advances the phase timing, and reduces current draw.


You are unloading the pump as you restrict the input. It seems counter-intutive to think about, but that doesn't change how it is.

The golden rule:
Never restrict an inlet, only the outlet.

[Harry_Fish]

<a href=showthread.php?s=&postid=14267859#post14267859 target=_blank>Originally posted</a> by doug6644
You stirred the pot Harry!!?? I tried to avoid this, hehe.


Yes, but look at all the good info!

[MeuserReef]

<a href=showthread.php?s=&postid=14271752#post14271752 target=_blank>Originally posted</a> by Harry_Fish
Yes, but look at all the good info!

Agreed ;)

[biecacka]

info is good on this topic. if you still have questions you can always ask the maker of the pump for the "sweet"spot to run their pumps on.

[doug6644]

Okay, that was fun. I believe you jdieck, looks like you squished my theory. I'll sit back and study more, ho-hum.

[liveforphysics]

Doug6644- Huge respect to you. This forum would be so much nicer if more folks would take your very mature and classy way to deal with learning something new that conflicts with your previous thoughts. You have my respect.

Best Wishes,
-Luke

[jdieck]

<a href=showthread.php?s=&postid=14276716#post14276716 target=_blank>Originally posted</a> by liveforphysics
Doug6644- Huge respect to you. This forum would be so much nicer if more folks would take your very mature and classy way to deal with learning something new that conflicts with your previous thoughts. You have my respect.

Best Wishes,
-Luke

+1 for 6644 :thumbsup:

[doug6644]

Thanks for the indepth, and wholesome education guys. I think alot of people benifitted from our ranting, including myself. And, untill our paths meet again, good day.

[BeanAnimal]

How did I miss all the fun?

Anyway just an addendum to the information:

If the output of the pump is severly restricted for long periods of time (close to 0 flow), the wet end may not be able to shed heat (caused by friction) and could damage the seals. I.E. don't deadhead (fully close the output) of a centrifugal pump for long periods of time.

[Harry_Fish]

Just an FYI and to clear up any confusion: I really like my overflows!

They are small, so they don't intrude into the tank.

I was just pushing way too much water through them since I
orginally planned to use hard pvc pipe and a lot of 45s and 90s.

In the end I used Spa-Flex end to end (was far easier to run)
so I had too much flow.

Once I turned down the flow I can hardly hear the overflows
(the cooling fans in my light fixture is louder).

You really only want about 3x the tank volume running through
your sump so it had time to skim and heat the water (contact time) I was pushing about 5 and a half times my tank volume.

I would really encourage anyone to run Glass-Holes overflows,
Just ensure you have the proper sized pump.

If I was to do it over I would have gone to a smaller pump.

Have fun!

[BeanAnimal]

Harry,

The the turnover rate does not affect the heating, a BTU is a BTU. A submerged heater will transfer all of its energy to the water, no matter what the flow rate, and do so with the exact same efficiency :)