Fixed or Variable Speed Pool Pump?

[Alex Wolf]

My pump maker only gives the Head graph.
https://s3-ap-southeast-2.amazonaws.com/astralpools-au-2/MEPS/AstralPool_Pumps_CTX_MKII_190922_(1).pdf

This totally makes sense since the final efficiency depends on many parameters beyond the control of the pump manufacturer. I believe that the most reliable and fastest way to find it out is to test the exact system live, drawing up a flow to power ratio curve (l/s of pumped water per kWh of consumed energy) and finding its peak. However, this will require the installation of a VFD controller anyway.

At idle or under load? Still sounds like an oversight in calculations that try to violate the law of conservation of energy. If you produce the same amount of useful work with half the energy consumption, this can only mean doubling the efficiency and nothing else. Are centrifugal electric pumps so poorly designed that when their speed is halved, their efficiency doubles (leaving all other elements of the hydraulic system behind the scenes)?

That's my question. Ultimately people are saying that a variable frequency pump is (say) 4 times more efficienct than a single speed pump for a given amount of useful work. I can't buy that at face value.

If we assume that we are talking about replacing a single-phase asynchronous motor (with typical field slip) with a three-phase synchronous motor with a VFD controller, then we can talk about an increase in energy efficiency from approximately 65% to 95% in the best case. If we assume that the VFD controller is smart enough and equipped with the necessary sensors, then theoretically it can squeeze out a certain amount of additional efficiency, depending on the depth of the suboptimality of the current pumping system. But the claims about “4 times more effective by simply reducing the speed” sounds like a marketing sales policy for me.

[nctnico]

And there are likely some losses in the capacitors to create several speeds as well reducing the efficiency of an induction motor based solution even further.

[IanB]

At idle or under load? Still sounds like an oversight in calculations that try to violate the law of conservation of energy. If you produce the same amount of useful work with half the energy consumption, this can only mean doubling the efficiency and nothing else. Are centrifugal electric pumps so poorly designed that when their speed is halved, their efficiency doubles (leaving all other elements of the hydraulic system behind the scenes)?

Edit: I mean that usually the nominal speed is designed to be as close to peak efficiency as possible. Doubling the efficiency by simply reducing the speed sounds very doubtful in the general case.

This is not about inefficiency, this is about the performance characteristic of the pump, like the chart you showed in a later post. That chart was for a single speed, but if you included speed as a parameter on that chart you would see a family of curves that behaved in approximately the way I described.

Note that the required power is roughly proportional to flow times head, which naturally means that if you halve the speed the pump can only do about 1/8th of the work.

The main impact of lowering the speed is to match the optimum efficiency of the pump with a lower system demand of flow and head. If you only need half the flow, then running the pump at full speed will waste a lot of power.

[IanB]

All the comments above about the efficiency of different motor types are very important, and clearly show the impact of the motor choice on power consumption.

The pump efficiency is a hydraulic efficiency, and is a parallel and complementary aspect of the overall system performance.

Consider the previously posted diagram below.

Let's suppose the required flowrate is only 3 on the capacity axis (the flow is throttled to match demand). We can see the hydraulic efficiency would be about 35%, meaning that 65% of the input power from the motor is wasted. If the motor itself had an electrical efficiency of 60%, then the overall efficiency would be 35% x 60% = 21%, which is abysmal.

Now let's suppose we reduce the speed of the pump, so that peak efficiency point aligns with the required flow of 3 (we slow the pump and reduce the throttling). Now the hydraulic efficiency could be closer to 85%. If we also include a variable speed drive with a modern motor to achieve the speed reduction, then perhaps the electrical efficiency could be improved also to 90%. This would give us a new overall efficiency of 85% x 90% = 77%. This is clearly much better than the previous 21%, and is much less energy wasted.

Whether such an improvement is actually achievable in practice depends on the details of the system under consideration and the originally installed equipment.

[Peabody]

For dave it looks like the pump is already running quite long, much of the time that solar power is available. So there is not really an option to run longer and thus slower. It would be a thing only if not that much filtering is needed.

4hrs per day. I just lowered that to 3hrs.
Correct, running longer outside of solar periods defeats the purpose unless we had battery storage.

So you have circulation for only 3 hours a day?  When I had a pool that wouldn't be enough to keep all the airborne crud filtered out, or to keep the pool chlorinated properly.  But that was an old pool using trichlor, and with no automatic cleaner for the bottom.  I guess things are a lot different with a modern pool.

[Alex Wolf]

I would add to the above that in addition to the simple double conversion of electrical energy into mechanical torque into hydraulic flow (with its own efficiencies and their product), we must also take into account the efficiency of the entire pumping system, taking into account the diameter, length and height difference of the pipes, the characteristics of the filters, and even intake and exhaust grilles. All this will ultimately affect the position of the peak of the final efficiency on the chart. And as far as I know, there is no such rule that the slower the more efficient, since the optimal flow rate for maximum efficiency directly depends on the characteristics of the pump impeller under a certain load - flow resistance, which is comprehensively determined by the entire pumping system.

[IanB]

At idle or under load?

There's really no such thing as "at idle". If the pump is running, it is always somewhere on the head-flow curve (see the diagram you posted).

If the discharge pipework has a very low resistance then the pump will be to the right of the curve, with a high flowrate and low required head.

If the discharge pipework has a high resistance the pump will be to the left of the curve, with a high head and low flowrate.

(This is for a single, constant speed. If we make speed a variable parameter it adds a third dimension to the chart.)

[IanB]

I would add to the above that in addition to the simple double conversion of electrical energy into mechanical torque into hydraulic flow (with its own efficiencies and their product), we must also take into account the efficiency of the entire pumping system, taking into account the diameter, length and height difference of the pipes, the characteristics of the filters, and even intake and exhaust grilles. All this will ultimately affect the position of the peak of the final efficiency on the chart. And as far as I know, there is no such rule that the slower the more efficient, since the optimal flow rate for maximum efficiency directly depends on the characteristics of the pump impeller under a certain load - flow resistance, which is comprehensively determined by the entire pumping system.

It turns out that system of pipes and equipment is like a "load resistance", and the pump itself is like a "supply". The operating point occurs where the load is balanced with the supply. This can be done graphically with a load line and a pump curve. Where the two lines intersect is the operating point.

There now can be a consideration of "impedance matching", where the capacity of the pump is matched to the pumped system (either by changing the size, characteristics or speed of the pump). A best system efficiency is achieved with a good match that puts the pump at its most efficient operating point.

[Alex Wolf]

At idle or under load?

There's really no such thing as "at idle". If the pump is running, it is always somewhere on the head-flow curve (see the diagram you posted).

"At idle" in this case means without any significant resistance to flow (no height difference, maximum diameter of the shortest possible pipes without bends or contractions). That is, a pump driven by an asynchronous motor doesn't produce any useful work as soon as it reaches operating speed, and the slip angle is optimal, and the rotation speed is maintained not only by the torque of the engine, but also by the pressure of the flow itself (in a looped system).

[IanB]

"At idle" in this case means without any significant resistance to flow (no height difference, maximum diameter of the shortest possible pipes without bends or contractions). That is, a pump driven by an asynchronous motor doesn't produce any useful work as soon as it reaches operating speed, and the slip angle is optimal, and the rotation speed is maintained not only by the torque of the engine, but also by the pressure of the flow itself.

OK, with that clarification, let's suppose that that is a flow somewhere around 15 on the horizontal axis of the chart. We can see from the brake horsepower curve that the power required from the motor is high and increasing. So the demand on the motor is actually higher than at lower flowrates. Far from idling, the motor is actually working very hard.

[johansen]

My fixed speed CTX360 vs the variable speed Astral pump.
Both similar head at say 300L/m if you use the orange "overdrive" mode.

And their own claimed power consumption figures shows
Variable: 514kWh/yr for 303L/min for 8m head
Fixed: 1353kWh/yr for 360L/min for 8m head.

So correcting the variable one for 360L/min gives 610kWh/yr
So better than half the energy consumption at the same flow rate, why

there are a lot of folks here with mild reading comprehension skills.
its not the same flow rate. its 308/360.
personally, i don't believe the energy savings you quoted above are possible.. unless the motor is so saturated, that the VFD can optimally drive the motor and suddenly the motor's efficiency increases from 65% at full speed (yes, a 1hp motor is almost always that bad) to 80% at 2/3rds speed, pushing 80% the flow rate.

if you notice, IanB suddenly changed his mind and now agrees with me that pumps have a cubic power curve, which is why i said half speed is 1/7th the power. when you account for half the flow rate, multiplied by 1/8th the power, you get back to what i said, moving the same volume of water for 1/4th the energy.

so you could argue the energy savings are 75% on that point alone. who cares if the pump has to run twice as long.

also keep in mind it doesn't. the slower flow rate through your filter means the chealating and binding agents work better on the filter.


Now regarding hydronic recirculation pumps which are used to push water through heat pumps, ground fields, water heated flooring (pex pipe, ugh)

I would have to get the numbers again, but i have a number of these pumps and at 110 watts at 120vac, one might push 10 gallons a minute with a 3 foot head. (the water shooting out of a 1/2" pipe will lift the water 3 feet out of a 5 gallon bucket).
at 60vac, 60hz, the flow rate is about half that volume, and the motor only sucks up maybe 40 watts. this is a capacitor run motor, impedance limited, 10uf cap. no vfd, just a variac.

[johansen]

And then in the same veriable speed datasheet they show even less of an advantage vs fixed speed 

367kwh vs 414kwh is not exactly selling a huge advantage?

again, reading comprehension skills are a huge variable....

the screen shot you showed is showing the power consumption of the variable speed pumps compared to a conventional 2 speed and 3 speed pump! not a single speed.

there is, in my opinion, NO significant advantage of a VFD over a 2/4/6 pole pump motor. if you need to fine tune the flow rate, use a valve.

the 4 and 6 pole windings take up hardly any space in the motor, they don't blow up, they don't burn out. if they are made a little bigger than needed you can get the motor's efficiency up. the savings pay for the vfd many times over, because it will blow up, short out, the capacitors will dry out, a power surge will kill it. etc.

[ejeffrey]

(All things being equal, a pump running twice as long at half the flowrate will consume the same amount of energy overall. All things are not equal in reality, but the differences may not be that much. It all depends...)

Not really.  Or at least it depends on what is equal.  Flow resistance rises with velocity, so of the pumping is limited by flow resistance (as opposed to elevation rise) the lower speed pump will need to supply much less work to move the same volume of water.  How much less depends on the flow velocity and amount of turbulence but it is likely a big difference.

However if the pump itself runs less (or more) efficiently below it's ideal speed the energy consumed by the pump won't track the work delivered to the water.  I have no idea how pool pumps are sized but I'd guess that at full speed they are running above their peak efficiency.

[IanB]

if you notice, IanB suddenly changed his mind and now agrees with me that pumps have a cubic power curve, which is why i said half speed is 1/7th the power. when you account for half the flow rate, multiplied by 1/8th the power, you get back to what i said, moving the same volume of water for 1/4th the energy.

I really didn't change my mind. I merely pointed out that if the pump is required to move, say, 1 m³/h, and you halve the speed, it will now be moving about 0.5 m³/h. If this is less than you need, the saving is moot.

so you could argue the energy savings are 75% on that point alone. who cares if the pump has to run twice as long.

In real world scenarios, time is usually a variable. Otherwise you could drive a car with a maximum speed of 30 km/h and obtain really good economy.

[Alex Wolf]

Otherwise you could drive a car with a maximum speed of 30 km/h and obtain really good economy.

This is highly dependent on the transmission ratios. Typically, for road cars, the optimal cruising speed is a little more than 100 km/h or 60 mph. 

[IanB]

(All things being equal, a pump running twice as long at half the flowrate will consume the same amount of energy overall. All things are not equal in reality, but the differences may not be that much. It all depends...)

Not really.  Or at least it depends on what is equal.  Flow resistance rises with velocity, so of the pumping is limited by flow resistance (as opposed to elevation rise) the lower speed pump will need to supply much less work to move the same volume of water.  How much less depends on the flow velocity and amount of turbulence but it is likely a big difference.

This is right. I did oversimplify a bit, and perhaps a bit too much. But when efficiencies are as bad as the numbers quoted in this thread, I suspect what I said is good to a first order approximation.

To be sure, turbulent flow resistance in piping is with the square of the flowrate, and therefore reducing the flowrate (or using fatter pipes) will reduce the energy cost. The trade off with fatter pipes is they cost more to install and there is an economic optimum.

[SeanB]

AFAIK the best reason to use a VFD drive on a pool pump is that you go from a single phase motor to a 3 phase one. The single phase motor is generally a lot lower efficiency in converting input AC power to actual shaft power over the 3 phase unit, so the overall efficiency of having a fixed 3 phase drive, and a 3 phase motor, of the same RPM and power output, is a motor that runs a whole lot cooler, so thus less of the input power is going out as heat in the windings and core losses. Your typical pool pump motor runs around anything from 600W to 1.1kW, though the efficiency is not always stated, and they run really hot. Swap out to a 3 phase motor in the same frame, and you find efficiency goes up from around 70 odd percent to well over 90 percent, just from the lower losses in the core and windings.

For a pool pump, where all parts, automatic cleaner, filter, heat pump, skimmer and returns are all optimised for a very small variance from the full power of the standard motor, and you will probably see around 10% savings in power from having the fixed frequency inverter and a 3 phase motor there. Pool pump motor you can retrofit, they mostly use a standard IEC flange size, though often they use a special shaft, but that is easy to fix just by swapping in the 3 phase motor body, and toss the end caps, rotor and the old pump single phase windings. Reuse the rotor, it likely will be a perfect fit, as they do use standard sizes in the motors, to gain advantage of industrial production. Take a thermal image of a single phase motor versus a 3 phase one, running any load, from zero to full, and you will always see the single phase one runs a lot hotter, especially with lower load.

[bdunham7]

I'm a bit astonished at some of the replies here.  Pool pump efficiency is a well-known and documented thing.  I have a pool/spa combo and I have a pump/filter system that I set up myself.  I've roughly measured flow rates, pressures and energy consumption, but that was almost a decade ago.  However, I can still tell you with 100% assurance that running your pump slower for longer will save you energy.  This is clearly documented in charts and graphs from various pool pump manufacturers and I'm not talking just about the more recent advertising for the variable-speed versions.   It's just a question of how much.

First, we're talking about a pool system here, and I'm going to assume that the pumps are approximately at the level of the pool.  This is important when you start discussing the head pressure, as a fixed head pressure componenent that would result from an elevation change--like a well pump--will result in a much different result than a variable head pressure which is dependent on the pumping losses through the pipes and filters.  In a pool system, there is essentially zero fixed head and the variable head pressure is entirely dependent on the pumping volume.  In such a system, doubling the pump volume over some baseline requires 8 times the mechanical input power, thus halving the volume will reduce the input power to 1/8.  This is true over the range of volumes, pressures and speeds that you'd see in any reasonable pool pump system.

So from there it is simple--with most typical pool pumps, cutting the pump speed in half will reduce the mechanical input power required to 1/8.  The other major factor is the motor efficiency.   I have a two-speed pump, not a variable.  Unfortunately, the typical Pentair two-speed pool pump was much less efficient on the low speed, but I built my own using a specialized motor that has separate windings, capacitors, etc that  are switched in by relays and retains most of its efficiency on the lower speed.  A variable speed ECM pump will gain 10-15% efficiency over even a good induction motor and will be more efficient over its usable range as well.  I'll probably install one if and when my current setup wears out.

The bottom line for those of you trying to adapt generic pump efficency concepts to a pool is that there is no net work being done--the water just returns to the pool.  Thus pumping losses--including motor efficiency--are 100% of your energy consumption.

[ejeffrey]

The 80% savings from cutting the rpm in half is legit, in theory you can run a pump at half speed on 1/7th the power..  The claims about real energy savings are dubious because the waste heat just heats the water

I'm sorry, but as a chemical engineer who knows something about pumps and hydraulics, I know of no such theory. If you could make such huge savings by running a pump slower, then there are lots of accountants who would like a word with you.

Process pumps are designed and intended to be used with 100% duty cycle, at least while the system is running.  They provide the flow rate  needed for the system.  There is no way to lower the speed without increasing the size of the pump.  Dave is taking a out a pool pump that runs 3 hours per day.  That is pretty much guaranteed to be several times less efficient than  an optimally sized pump that runs the same system 6 hours a day at half the flow rate.  What fraction of the savings are achieved by running the same pump at lower rpm depends on the pump efficiency curve but it's for sure a real thing.

[IanB]

Dave is taking a out a pool pump that runs 3 hours per day.  That is pretty much guaranteed to be several times less efficient than  an optimally sized pump that runs the same system 6 hours a day at half the flow rate.  What fraction of the savings are achieved by running the same pump at lower rpm depends on the pump efficiency curve but it's for sure a real thing.

I understand it's a real thing, but the hidden question is why wouldn't you just install a smaller (and cheaper) pump in the first place, rather than running a bigger pump and slowing it down?

It seems to be the case that sometimes the high flowrate from the big pump is needed, for example to backflush filters (?). In which case installing a smaller pump would not work.

Therefore the best solution seems to be to use a three phase motor with a variable speed drive for maximum efficiency, and run the pump slower for longer when the high flowrate is not needed.

An alternative solution is apparently to use a motor with different windings for different fixed speeds, and to switch speeds as required.

[Alex Wolf]

why wouldn't you just install a smaller (and cheaper) pump in the first place, rather than running a bigger pump and slowing it down?

Due to cavitation at high impeller speeds. High performance with high efficiency requires a large impeller and high torque.

[IanB]

why wouldn't you just install a smaller (and cheaper) pump in the first place, rather than running a bigger pump and slowing it down?

Due to cavitation at high impeller speeds. High performance with high efficiency requires a large impeller and high torque.

No, that is not factual at all. Pumps come in many sizes, from tiny to enormous. Whatever the pumping duty required, an efficient pump can be selected from a catalog to suit that duty.

[Siwastaja]

So better than half the energy consumption at the same flow rate, why?
Just because of the way the motor is driven?

Just as expected.

In which case why not simply make the fixed speed pump with the variable drive control but fix it at one speed? 

Because the better motor type (probably permanent magnet synchronous AC motor also known as BLDC), which needs the inverter circuit anyway, trivially allows varying the speed once the circuitry is there: it's only about a user interface to set the speed. Given added value in higher efficiency, cost is higher, and people also expect more features. Besides, these motor types have larger sweet region so it would be stupid not to allow users to gain even more savings. If you only need the single speed, just don't adjust it. But even if you predominantly want a certain speed, you don't know exactly what the optimum is, before adjusting it. It's like selling a car with gas pedal glued in place; one would then control speed by applying variable amount of braking, which would be highly ineffective.

It makes no sense for a company to make a fixed speed pump at half the efficiency if it can make it by just changing the motor drive type

Of course it does, market is full of stuff where old legacy model, which is cheaper to manufacture, especially given economics of scale and existing production lines, is being manufactured even when it is inferior in every other way except price.

[johansen]

I'm a bit astonished at some of the replies here.  Pool pump efficiency is a well-known and documented thing.  I have a pool/spa combo and I have a pump/filter system that I set up myself.  I've roughly measured flow rates, pressures and energy consumption, but that was almost a decade ago.  However, I can still tell you with 100% assurance that running your pump slower for longer will save you energy.  This is clearly documented in charts and graphs from various pool pump manufacturers and I'm not talking just about the more recent advertising for the variable-speed versions.   It's just a question of how much.

First, we're talking about a pool system here, and I'm going to assume that the pumps are approximately at the level of the pool.  This is important when you start discussing the head pressure, as a fixed head pressure componenent that would result from an elevation change--like a well pump--will result in a much different result than a variable head pressure which is dependent on the pumping losses through the pipes and filters.  In a pool system, there is essentially zero fixed head and the variable head pressure is entirely dependent on the pumping volume.  In such a system, doubling the pump volume over some baseline requires 8 times the mechanical input power, thus halving the volume will reduce the input power to 1/8.  This is true over the range of volumes, pressures and speeds that you'd see in any reasonable pool pump system.

So from there it is simple--with most typical pool pumps, cutting the pump speed in half will reduce the mechanical input power required to 1/8.  The other major factor is the motor efficiency.   I have a two-speed pump, not a variable.  Unfortunately, the typical Pentair two-speed pool pump was much less efficient on the low speed, but I built my own using a specialized motor that has separate windings, capacitors, etc that  are switched in by relays and retains most of its efficiency on the lower speed.  A variable speed ECM pump will gain 10-15% efficiency over even a good induction motor and will be more efficient over its usable range as well.  I'll probably install one if and when my current setup wears out.

The bottom line for those of you trying to adapt generic pump efficency concepts to a pool is that there is no net work being done--the water just returns to the pool.  Thus pumping losses--including motor efficiency--are 100% of your energy consumption.

Thanks for the words of reason here.

I am not familiar with pentair but i am familiar with two different kinds of 2 speed motors that use the same windings. There is two ways to connect the windings. in parallel for half speed, which maintains the same flux density, or series for half speed which cuts the flux density in half which produces 1/4th the torque on the rotor.

pumps typically use the latter. lathes and machine tools, the former.

because the windings are not optimal for either use case, they aren't very efficient on slow speed.


one experiment i did successfully, i took a 4 pole, 1750 rpm shaded pole fan motor. I cut the windings apart and changed the polarity, and put a capacitor in series with two of them. the motor ran fine at 60 vac, at 3500 rpm.

salient pole induction motors are not common above 1/4hp. but you could build a 2 speed pump motor with such a method.

[ahbushnell]

Finally got my pool circuit monitoring enabled.
The pool pump takes twice the energy per day than my heat pump hot water tank.

I think pumping losses are not linear with flow rate.  If so it seems like a lower flow rate over longer time would be better.  But there is probably a knee on that.