Energy efficiency of filters, and too-often cleaning?

[dan1000]

As a newbie, I'm wondering if the gurus could help me understand how to get good energy efficiency while maintaining a well-filtered pool? What I'm really asking about is the resistance caused by different kinds of filters (cartridge, sand, D.E.).

Mine is a 14,000 gallon pool with salt water chlorine generator, and Pentair Intelliflo pump (which has quirks, but I like from an energy efficiency standpoint). Currently, I have a Hayward Stay Clear Plus cartridge filter with a 120 sq. ft. cartridge. I've set the Intelliflo to do one turn per day, and to take about 10 hours to do so. This works out to about 23GPM, apparently.

The pump tells me it is using about 250 watts of electricity when the filter is clean. However, after a few days, that energy usage climbs to about 650 watts. At this point, I've been cleaning the filter, and the enercy usage drops back to 250w. Obviously, this is too often for cleaning (it is becoming annoying), and my water looks nice and clean so I'm not sure where the dirt is all coming from. However, when I clean the filter, lots of dirt comes out, so perhaps it's not as clean as I think it is. Or perhaps what I think of as "dirty" is really just "not quite clean" (the filter is beige before cleaning, white afterward, and during cleaning a fair bit of what looks like fine garden dirt comes out).

Anyway, from reading this forum, I know about the filtering efficiency of cartridge, sand, and DE filters (ie: how small the particles are that they can filter - although why this is called "efficiency" I don't know). But what I haven't found discussed, is the energy efficiency of these filters, in terms of the resistance they provide to water flow. DE seems to stop the most dirt, but does that necessarily mean it provides the most resistance to the flow of water (in which case energy usage would go up to maintain the 23gpm)? Or is there a free lunch -- does the DE filter stop much more dirt while providing not much more resistance than sand or cartridge (or maybe even less)?

Bottom line is that while I of course want absolutely crystal clear water, I don't want the life-cycle cost to be a big hidden amount that is paid out over time from increased energy usage. So, for clean water with low energy usage, should I stick with cartridge, move to sand (possibly with a sprinkle of DE on top), or go with DE?

As for cleaning the cartridge so often, perhaps I just need a bigger cartridge to increase the cleaning interval? Is 120s.f. enough for a 14k gallon pool?

Thanks


Dan Freedman

[duraleigh]

Hi, Dan,

Welcome to the forum. Interesting post. What jumps into my mind immediately is a larger filter.

That would provide:

1. less frequent backwashing (convenient) and...

2. less energy usage because of the decreased resistance.

I'm not smart enough to calculate how long it would take to recoup (in energy savings) the increased cost of the filter but your entire system will work better and your time and labor savings would seem to make it worthwhile.

I suppose you could assume a sand filter, because of it's coarseness, would use less energy as well but may not give you the water clarity you require. That said, a well-balanced system with a sand filter provides water clarity that is visually almost impossible to distinguish between cartridge or DE.

[JasonLion]

As duraleigh said, getting a larger filter is by far the best way to be more efficient.

One thing to keep in mind is that when you are first starting up a pool the filter will catch small particles that you can't normally see in the water that came with the fill water. After a few weeks you will only be filtering stuff which falls into the pool, which is usually much less stuff (except during pollen season).

[mas985]

To give you a reference, my pool is about 21k gallons and I have a 420 sqft filter. The equivalent size for 14k gallons would be about a 280 sqft filter.

Normally, I don't see a pressure rise for 6 months or more. However, there have been several fires in No CA in the past few weeks and even a large filter is taxed by the large amount of ash being dropped in the pool. I saw a 4 PSI rise in just the last couple of weeks.

The Intelliflo adjusts the RPM automatically to keep flow rates constant so when the filter gets dirty, head loss increase (i.e. filter pressure rise) and RPMs go up. So you should actually see a faster rise in PSI than a standard pump would. For this reason alone, a pool with an Intelliflo pump would benefit even more with a large filter.

The sweet spot for the Intelliflo is about 1000 RPM. This is were the GPM/Watt is the highest and your costs are the lowests per turnover. One thing you could do is to set the GPM just below this value and allow the filter PSI to raise it to just above this value. At 650 watts, it sounds as though you are operating closer to 2000 RPM. So halving your GPM should give you a bit more energy efficiency.

[dan1000]

Thanks for the comments so far. I agree that in general, bigger filter should equal lower resistance and also longer between cleanings. When I try to think of this like a physicist (which I'm not), I come up with the following:

The ideal filter would consist of a single thin sheet of exceptionally small holes, which in aggregate had an area equal to the area of the plumbing leading to and from the filter (ie: the area of a 2" or whatever pipe). That way, the filter would catch all the dirt, but let the water through with no additional resistance.

Of course, once the dirt starts to pile up, some of those small holes will become clogged, and the area will thus reduce. So, to avoid having to clean the filter out continuously, it should have an area much much much bigger than that of the plumbing. That way, the water will be able to find its way around the clogged holes to others that haven't been clogged yet. The holes should still be small enough to catch any dirt.

Now for the tricky bit: cartridge filters have a thin membrane as described above, and DE filters come close (ie: the water doesn't have to pass through much depth of material). But sand filters are different, I think. The water must pass through a LOT of material to get to the other side. Each grain of sand it passes must exert some frictional force on the water, so perhaps they shouldn't be as energy efficient?

I'm wondering if anyone has a sense as to whether one kind provides less resistance than another, for a given level of what is (for some reason) called "filtering efficiency" (ie: for a given level of ability to stop crud).

BTW: I'm in Honolulu with a pool only a couple of months old (new build). The chemistry seems fine. However, I did do one thing very wrong: I used flocculant to aggregate some extremely fine dust/dirt that was making the pool cloudy. But I vacuumed it through the filter instead of to waste. I'm wondering if perhaps I've ruined the filter by doing this.

Also, when I experimented with the Intelliflo, I initially thought that running the pump for 24 hours per day would be the most efficient solution (least back pressure). However, the unit won't pump less than 15GPM for some reason. Even if it should be pumping 10GPM (about right for me if running 24 hours/day), it would run at 15GPM. At that setting, it would do a turn in about 15.5 hours, but the display told me it was using about 175 watts. 175 * 15 = 2.7 KWH per day. At my current setting it uses about 250 watts * 10 hours = 2.5 KWH/day. Not a big difference, but a difference nonetheless. By the way, the original setup my PB left me with was with an "ordinary" sta-rite pump running 8 hours per day (many turns, not that he worried about that). That pump was using well over 1.2KW, or about 10KWH/day. So I do seem to be saving 75% by using the Intelliflo, but only if I can keep my filter clean !

Incidentally, I have a SolarAttic as well, but since the pool is already in the high 80's, I won't be able to truly evaluate it until the winter.


Thanks for any and all comments.

Dan

[mas985]

All three filter types usually have nearly the same pressure loss specification so as far as efficiency is concerned, that is less of an issue. A clean filter is usually less than 3 PSI loss so that is probably the smallest contribution to head loss.

However, a dirty filter, no matter what type, adds the most head loss and will reduce your efficiency more than the filter itself. So when choosing which type to go with, it is less about type of filter and more about size and maintenance required that will sway your decision.

Also, the Intelliflo has a minimum RPM setting which will determine the lowest GPM. I suspect you are running into the limits at the low end of 15 GPM. These low RPMs tend not to be as efficient as the mid range RPMs anyway so it is probably best where you are.

[dan1000]

Thanks, Mark. It would be interesting to learn that DE offers more or less the same resistance as sand or cartridge, for any given level of dirtiness. If true, then I suspect DE would provide the best energy efficiency, since it supposedly filters out a lot more crud.

Perhaps another related question would be whether all 3 kinds of filters require the same number of turns per day (currently I'm using 1 turn per day with my cartridge filter) in order to do their work.

At the end of the day, I suppose my true goal is to have a sparkling clean pool with little filter maintenance, without paying through the nose for energy or supplies.

There, not much to ask for

Dan

[Titanium]

This is a very interesting topic. I would not be surprised to learn that there has been little to no study on the most energy efficient way to run/clean pool filters. And as you guys have determined, of all the things that contribute to pump losses, all of them like pool piping, pump size, etc all stay the same, whereas the filter losses increase as debris loading increases.

I understand that Dan1000 is able to determine motor kw usage by means of the Intelliflo pump diagnostic display, but how is mas985 determining the motor kw usage?

Also, as another data point for cartridge filter sizing vs pool size, I have 420 sq. ft. of cartridge filters for a 24,000 gal inground pool.

It is fairly easy to determine the amp draw of the filter pump, but unfortunately, this amp draw is composed of both "real" and "reactive" current, and the utility only charges residential users for the "real" current. Only the "real" part of the current contributes to the kw usage of the motor and kw is only what the utility charges. As an aside, utilities make a very big deal of charging commercial and industrial customers for the "reactive" current.

It would not surprise me to learn that the majority of filter debris loading occurs after only a couple of weeks or so. I have been following the "tried and true" adage of cleaning the filter after the psi rises 10 psi over a clean filter, and, for me, this takes roughly 8-11 months. I have a big hunch that the "clean after 10 psi rise" rule is leading to a bunch of wasted energy.

And if the answer to saving energy is an increased frequency of filter cleaning, then it would be very useful to come up with ways to make this unenjoyable task 1) quicker, and 2) less messy. This increased frequency of filter cleaning might make devices such as this finally worth the investment, http://www.neoterics.com/

Also, there is a cartridge filter out there that claims to be able to go longer between cleanings because of a special flow pattern that somehow enables the filter material to be more evenly loaded. Unfortunately, I can not find the link where I saw this information, but I thought it was Sta-Rite?

Titanium

[Titanium]

I wonder if this could be adapted to be able to read the kw draw of pool pumps? It is very promising that the device can read power factor, as this is the critical element needed to determine what part of the raw amp reading of the pump motor is "real" and which is "reactive". My only concern is that this device is shown on a 120 volt receptacle, whereas most of our pool pumps are wired for 240 volts.

http://mitchellinstrument.net/p/WBMiTKA ... Meter.html

[web]http://mitchellinstrument.net/p/WBMiTKAW/Outlet+KWH+Meter.html[/web]

Titanium

[JasonLion]

The Kill-A-Watt electrical usage monitor is rated for 15 ams at 120 volts. It could be gotten to work with some small pumps, but most wouldn't work. It is simpler to watch the electric meter provided by the utility company with the pump on and then again with the pump off and subtract.

If you don't have an Intelliflo pump then the electrical usage of your pump will stay more or less the same as the filter gets dirty (it won't be exactly the same, but close enough). Normally, it is the flow rate that will vary as the filter gets dirty. With a fixed speed pump, a dirty filter means a lower flow rate. Only the full Intelliflo monitors the flow rate and turns up the power usage to try and maintain the flow as the filter gets dirty.

The resistance to flow of the filter varies far more dramatically with the size of the filter than it does with the kind of filter. A large sand filter will have less resistance than a small cartridge filter and a large cartridge filter will have less reistance than a small sand filter. No doubt one kind of filter or the other is "best" for energy efficency, but that difference is small compared to several of the other variables that affect the energy efficency of the whole system.

[dan1000]

Titanium - thanks for the comments. I have both the product you mention. The "Blaster 2000" cleans my filter ok, but not as clean as I can do by myself with a hose and nozzle. The big plus is that it works unattended (ie: set it up, leave, come back in 15 minutes). But you'd need a lot more water pressure than I have available in order to beat a hoze and nozzle.

I also have the Kill-a-watt, although I've not used it in a 240v scenario. It's great for figuring out how much electricity is being used by stuff, such as: computers, power "bricks" that are left plugged in all the time, and so on. It has readings for "VA" and "Watts" as well as KWH. I am not certain, but I suspect this means it gives both the "reactive" and true readings. Watts is always lower than VA (this is at least partly because of it being AC rather than DC -- in AC, the voltage comes and goes 60 times per second, so even though the voltage reads 125v, most of the time the actual voltage is less, cycling between 125v+ and 125v-).

I'll go back to my previous comment: I am looking to make a smart decision about having a clean pool with low costs (including energy cost caused by increased pump workload), and preferably with low labor too!

BTW:Here in Honolulu, electricity is priced at a constant (roughly) 20 cents per KWH. So, at 250w times 10 hours, I'm using 2.5KWH/day, and spending about $0.50/day (~$180/yr) running the pump. At 650w times 10 hours (when filter dirty), I'm using 6.5KWH/day, and spending about $1.30/day (~$475/yr). With my old (non-Intelliflo) pump, I was using about 12KWH/day and spending $2.40/day (~$876/yr). I haven't measured the SWCG's usage yet -- perhaps better if I don't

Dan

[dan1000]

Jason

Thanks for the additional input. What I'm hearing so far is that all 3 kinds of filter are similar in resistance, and that size matters (I always suspected, darn it), and that the dirtying of filters is likely the largest component of resistance when it comes to what the pool motor feels.

Also, for those with non-Intelliflo pumps, the flow will gradually slow down as the filters get dirty, whereas for those with Intelliflo (or similar), flow will remain the same, but energy use will go up.

I will begin to do a chart on how long it takes for the energy usage on my pump to increase, and correlate that with the PSI increase. In that way, I may be able to insert some hard data for others to use when evaluating their own energy efficiency.

Thanks

Dan

[mas985]

This is a very interesting topic. I would not be surprised to learn that there has been little to no study on the most energy efficient way to run/clean pool filters. And as you guys have determined, of all the things that contribute to pump losses, all of them like pool piping, pump size, etc all stay the same, whereas the filter losses increase as debris loading increases.

I understand that Dan1000 is able to determine motor kw usage by means of the Intelliflo pump diagnostic display, but how is mas985 determining the motor kw usage?
Pentair publishes energy consumption as a function of RPM level which I modeled. Chemgeek did this as well.

Also, as another data point for cartridge filter sizing vs pool size, I have 420 sq. ft. of cartridge filters for a 24,000 gal inground pool.

It is fairly easy to determine the amp draw of the filter pump, but unfortunately, this amp draw is composed of both "real" and "reactive" current, and the utility only charges residential users for the "real" current. Only the "real" part of the current contributes to the kw usage of the motor and kw is only what the utility charges. As an aside, utilities make a very big deal of charging commercial and industrial customers for the "reactive" current.

The electric meter on the side of your house is the most accurate you can get since that is what the power company actually charges you by. It is a bit tricky to remove the other stuff in the house but if you don't mind reseting a few clocks then the best thing to do is turn off all the breakers except the pool.

It would not surprise me to learn that the majority of filter debris loading occurs after only a couple of weeks or so. I have been following the "tried and true" adage of cleaning the filter after the psi rises 10 psi over a clean filter, and, for me, this takes roughly 8-11 months. I have a big hunch that the "clean after 10 psi rise" rule is leading to a bunch of wasted energy.

That's one reason why I never wait that long. In fact I usually clean it before a 1 PSI rise but we are getting a lot of ash from nearby fires so I am waiting for that to clear up.

And if the answer to saving energy is an increased frequency of filter cleaning, then it would be very useful to come up with ways to make this unenjoyable task 1) quicker, and 2) less messy. This increased frequency of filter cleaning might make devices such as this finally worth the investment, http://www.neoterics.com/

I clean mine twice a year and for the most part it isn't that hard or take that long, maybe a half hour.

Also, there is a cartridge filter out there that claims to be able to go longer between cleanings because of a special flow pattern that somehow enables the filter material to be more evenly loaded. Unfortunately, I can not find the link where I saw this information, but I thought it was Sta-Rite?

Titanium

See above responses

[JasonLion]

and that the dirtying of filters is likely the largest component of resistance when it comes to what the pool motor feels.

That isn't exactly right. There are many different ways the plumbing may be setup, in some cases the dirty filter resistance can be very significant and in other cases it can be fairly minor.

The 10 psi increase rule of thumb corresponds to 23.1 feet of dynamic head. The total feet of dynamic head (TDH) with a clean filter varies from pool to pool, so the percentage increase will also vary. Commonly clean filter TDH is between 40 and 60. Including all pools the range might be more like 20 to 90 TDH with a clean filter. Obviously adding 23 feet to a system that runs at 20 feet clean is going to make a huge difference, while adding 23 feet to a system that starts at 90 feet isn't going to make nearly as much of a difference. On a typical pool a 10 psi increase in filter pressure will increase the TDH by a bit over 1/3. That is a large effect, but not "the largest component of resistance".

The situation is further complicated by the pump efficiency curve. Any given pump will vary in energy efficiency at at different TDH values. Pumps are most efficient near the center of their range. Depending on how your systems clean TDH matches up with the center of the pumps range you might be getting more or less efficiency out of the pump with a dirty filter. Normally the efficiency change isn't that large, but if you are operating near the edge of the pumps range the change can sometimes be dramatic.

[dan1000]

Jason:

Interesting. In my case, the filter obviously is a large portion of the resistance, or else my energy usage wouldn't go up 250% when the filter gets somewhat dirty. I have three return lines going back to the pool. Two are 2" (one for the Polaris 360 - which also incidentally makes the pump's energy usage go up by about 500w when I use it, and another for the eyeballs around the pool), and one is 1.5" (which feeds a dedicated "sand pool" area that pours over into the main pool (for foot washing when coming in off the beach -- no sand in it so far).

So, presumably, my dynamic head is kept low by there being a lot of pipe for the water to return through. Therefore, my filter has become the most critical element. Hmm. I should run the pump without the filter, and see what energy usage that gets me. I will do that now, and post shortly.

Dan

[dan1000]

Just did some measurements.

With no filter at all, my pump uses 185 watts to pump 23GPM, with the pump at around 1300RPM.

With a newly cleaned (once again, and this time I did it thoroughly) filter, the pump then uses 205 watts to pump 23GPM, with the pump at around 1365RPM. Pressure on the Hayward filter goes up by just a fraction of a PSI with the filter.

So, an empty cartridge filter doesn't seem to be adding much at all over no filter whatsoever.

Can anyone else with an Intelliflo or with a clamp-on ammeter post some numbers for their situation? Presumably for non-cartridge filters, you'll need to use some kind of "bypass" setting, if available.

THanks

Dan

[dan1000]

Sorry to do this to death, but I just thought I'd mention that the other reason my filter is an important part of my dynamic head is that my flow using the Intelliflo is (deliberately) very low - 23GPM, and goes through big pipes.

However, I'm not sure if "filter friction" goes up/down by the same amount as "pipe friction" as the flow rate increases/decreases.

The pool's not cloudy, but my brain is beginning to be


Dan

[Titanium]

Dan1000,

I have a clamp-on ammeter, but - unfortuneately - measuring the motor amps with this device will not give me kW. The motor amps that I read from the clamp-on ammeter will give me kVA, though, which combines the effect of "real power" (i.e. kW) and "reactive power" (i.e. kVAR). For a purely resistive load, like an incandescent light bult, kW is the same value as kVA. For other loads with a reactive component in addition to the resistive portion, like induction pool motors that we're talking about, kW will always be smaller than kVA.

For a motor that is running at 75%-100%, a typical power factor might be something like 80% or 85%. So a motor with a reading of 1 kVA and a power factor of 80% will equal 0.8 * 1kVA = 0.80 kW



I know it is hard to see in the above image, but notice that the amps in the middle curve start out high to the right and ends up leveling off at 50% or so as you go left. This is the behavior of measured motor amps versus the motor loading. The reason the motor amps behaves this way can be seem from the leftmost curve. The leftmost curve shows a high power factor at high motor loadings, but the power factor goes to **** as you follow the curve to the left. In the field, I have measured motor amps on a motor which was not attached to a load - in other words, the motor shaft will literally spinning in free air. I STILL measured the motor amps to be around 50% of the motor nameplate's Full Load Amps (FLA). Now obviously, this motor was doing very little real work (i.e. drawing very little kW) and the only kW being drawn was from the motor cooling fan and motor losses. The majority of the measured amps on this motor were due to reactive power, and this was due to the fact that the motor was extremely lightly loaded and hence the power factor was so very low.

Also notice in the top picture where you see a Volts sine wave and an Amps sine wave? All power factor means is how far away that the two sine waves are from being "in sync" with one another. At 100% power factor, the two sine waves would have their peaks and valleys matched up. As the power factor goes down from 100%, the two sine waves get further and further apart.

Normally to measure kW, one has to know the power factor, voltage, AND Amps. And this normally takes an electrical multimeter which is fairly advanced (and expensive). This is why I was so excited to find such a low cost device which could measure amps and power factor earlier today. I think I am going to buy a Kill-A-Watt device (which I found on Amazon for around $21 - how can I refuse?!?), and play around with one and see if I can either:

1. modify the internal guts in order to run on 240 V
2. use a 240 Volt to 120 volt transformer and apply some kind of multiplier to the displayed reading, or
3. attach one hot and the neutral of my 240 Volt feed to the motor, which will only be 120V into the Kill-A-Watt, and see if I can apply some kind of multiplier to the displayed reading. I am making an assumption that this little trick won't affect the power factor.

Titanium

[Titanium]

mas985,

Re: "I clean mine twice a year and for the most part it isn't that hard or take that long, maybe a half hour."

Wow, you are really doing good to get your cartridge filter done in only 30 minutes. By the time I dismantle the lid, remove four filters, hose all four filters really good, reinstall the filters, lube the o-ring, and reinstall the lid, I am usually into the darn process most of two full hours. Maybe I am spending too much time getting the filters uber-clean??

Re: "I usually clean it before a 1 PSI rise"

You are cleaning your cartridge filter before a 1 psi rise and yet you are only cleaning twice a year?

Titanium

[mas985]

mas985,

Re: "I clean mine twice a year and for the most part it isn't that hard or take that long, maybe a half hour."

Wow, you are really doing good to get your cartridge filter done in only 30 minutes. By the time I dismantle the lid, remove four filters, hose all four filters really good, reinstall the filters, lube the o-ring, and reinstall the lid, I am usually into the darn process most of two full hours. Maybe I am spending too much time getting the filters uber-clean??

Re: "I usually clean it before a 1 PSI rise"

You are cleaning your cartridge filter before a 1 psi rise and yet you are only cleaning twice a year?

Titanium

After the pool was built, I cleaned the filter every month until most of the plaster was gone but since that time, I haven't had the need to clean it very often. Usually every 6 months or so. I guess I don't get much dirt in the pool. I know that there are a few others with large cartridge filters who don't clean them that often as well. The worst time of year is when the leaves fall which starts in November and usually ends in January and I clean the filter at that point. Most of the rain we get is between December and April so I suspect that is why I don't need to clean the filter that often. Pollen dumps seem to be the worst offenders and that happens in the fall.

This year we have had quite a few forest fires in No Cal and the ash falling in the pool and everywhere else has been extensive. You can actually see the ash on the ground. So far I have seen a 4 PSI rise but like I said before, I will probably wait another week just to make sure all of the ash has fallen. The last time I cleaned the filter was in February.

When I first started cleaning the filter, I have to admit it took me about a hour or about 15 min per filter. I don't spend as much time cleaning them as I use to because I am getting better at it and I don't try to get them perfectly clean. Also, it has been a least a year since I soaked them because I found that there really wasn't a need to do so. However, the next clean I probably will be soaking them which will take longer to do.

I think the disparity of views regarding cartridge filters stems from the variability in conditions caused by usuage, weather and landscaping. If I had to clean my filter every month because of a substantial PSI rise, I would go nuts and be on the side of those who hate cartridge filters. But since I don't have to clean mine very often, I don't mind cartridge filters at all. Also, since I only clean it twice a year and other filter types require regular backwashing, I am probably saving some water.

BTW, I too have a clamp on meter and found that it comes close to to the kw usage of the pump but it is an analog scale so it is very difficult to get an accurate reading. I used my house's power meter and got a reasonable reading from that as well but it was higher than the amp meter but it did have the controllers power included which should not have been more than 100 watts. The amp meter gave me about 1.8 kw while the house meter gave me something over 2 kw depending on head loss. Probably the difference can be explained because the amp meter accuracy is not so good. Given the power company uses the house meter, that is the reading that is most important.

Also, I believe most pumps are designed to operate within about 10% of full load to keep efficiency high. Based upon some measurements I have done over a range of head losses, the power draw on the pump decreased less than 10% with a head loss range of 50-70 feet (I used a return side valve to increase the head loss of my plumbing while reading the power off the house meter). So unless you are operating outside the normal range for the pump, it is probably pretty close to full load and I suspect that the power factor for most pumps will be fairly high (~ 0.9) and can be ignored if measuring amp draw. Given the amp meter reading I did and the power measured on the house meter, it would seem to be the case. But I would encourage you to measure the power with the house meter and see if you get the same results.