Running a SWG at 100% and Cell Life

[ajw22]

From ...

Have had to add 120 lb salt to pool in one year

I would like to add that a T-3 cell is absolutely way too small (undersized) for your pool volume. At a minimum your pool would have a T-9 cell or, since the incremental cost is not all that much, a T-15 cell. TFP recommends always getting a cell that is rated for 2X your pool volume and your...

www.troublefreepool.com

@JoyfulNoise said:

This is not true. Cells should never be run at 100% all the time. That will absolutely wear out the coating faster. Your assumption is that the electrolysis of the pool water and formation of chlorine is uniform across the cell plates. It is not. You’re welcome to move this post to Agree to Disagree or The Deep End if you want to have further discussion.

Explain that to me. Not so much the molecular chemistry of the cell plate wear, although others may be interested in that, as what the practical wear rate curve is at higher generation %?

A T-15 cell has a specified capacity of 580lb of chlorine and 8,688 hours at 100% . That is 0.067lbs of chlorine/hour.

We can call 0.067lbs of chlorine/hour the normal depletion rate.

Running it at 20% or 30% or 50% I assume the usage will deliver 0.067lbs of chlorine/hour until some %?

At what % does the depletion curve bend up and you are expending more then 0.067/hr?

Is 80% ok? 90%?

If 50% gives normal depletion does 80% use more relative capacity? Does 90% use more then 80%?

How much more then 0.067lbs of chlorine/hour is being taken from the T-15 cell when run at 100%?

 

[mas985]

I believe that is really the amp-hours of cell usage that determines the life of the cell. So not just the time it is energized but also the current going through the cell.

However, if salt levels are higher, the amps will be higher but so will the chlorine production. This is why some manufactures choose to specify the life of the cell in lbs. of chlorine production because in the end, that is really what matters.

 

[Steve-D]

This is how the SWG % was explained to me:

Just to make sure, here is how a SWCG works.
The cell either makes chlorine or it does not.
To adjust the output of the cell, we just turn the cell on or off for a period of time.

At 50% output the cell will be on about 50% percent of time. Different brands have different on/off times, but at 50% my IC40 is on for 2.5 minutes and off for 2.5 minutes. At 100% it will run continuously.

 

[ajw22]

I believe that is really the amp-hours of cell usage that determines the life of the cell. So not just the time it is energized but also the current going through the cell.

However, if salt levels are higher, the amps will be higher but so will the chlorine production. This is why some manufactures choose to specify the life of the cell in lbs. of chlorine production because in the end, that is really what matters.

I agree with what you said and let's stipulate for this discussion that the salinity is at the manufactures recommended level which should give the specified output.

 

[ajw22]

At 50% output the cell will be on about 50% percent of time. Different brands have different on/off times, but at 50% my IC40 is on for 2.5 minutes and off for 2.5 minutes. At 100% it will run continuously.

The question on the table is will the cell wear at 100% be more than 2X the wear at 50% and if so how much more?

 

[mas985]

I agree with what you said and let's stipulate for this discussion that the salinity is at the manufactures recommended level which should give the specified output.

There are usually two numbers used: Maximum CL lbs./day (production rate) and total lifetime CL lbs. I was talking about the latter. Over the range of allowed salt levels, the production rate changes but not the total lifetime CL lbs. That is fixed and a function of amp-hours.

 

[ajw22]

the production rate changes but not the total lifetime CL lbs. That is fixed and a function of amp-hours.

@JoyfulNoise is asserting that you reduce the total lifetime CL lbs if the cell is run at 100% and that is what I am trying to understand.

 

[SoDel]

What comes to mind is that some chlorine generators do actually limit current as a function of %. Apparently that method is common in Australia. Some chlorine generators are full on or full of for a % of time. That may explain some? If it’s the current limiting kind, 100% would (or could at least) have a different effect than 50%, assuming cell deterioration is not a linear function of current.

 

[ajw22]

Some chlorine generators are full on or full of for a % of time.

All of the SWG's from the major US manufacturers use an ON/OFF cycle time to control generation %.

SWG Generation Cycle Times - Further Reading

www.troublefreepool.com

 

[SoDel]

I’m not arguing the point, just figuring what was possibly the reason.

For chlorine generators that use % of time, my way of figuring is the cell will last so many hours; deterioration of the cell is per hour; any hour has to be the same as any other hour or the calculated output would have to be adjusted as a function of %; running it at 100% does not reduce the number of total hrs it will last; buying a cell that needs to run 100% of the time to sufficiently chlorinate a pool is insanity because there will be days when more than 100% is needed and the settings only go up to 100% (the dial needs to go to eleven if you remember This is Spinal Tap).

 

[Asherk74]

Perfect timing, as I am trying to modify my pump run time, rpm and SWG %.
right now I run it at 40-50%, but considering dropping my pump run time and increasing my % as I had been under the impression a lower % increases life span. but I question that.

 

[SoDel]

Perfect timing, as I am trying to modify my pump run time, rpm and SWG %.
right now I run it at 40-50%, but considering dropping my pump run time and increasing my % as I had been under the impression a lower % increases life span. but I question that.

This is a digression from the topic at hand (sorry) but one observation when switching from 12 to 24 hr pump run time: simply doubling SWG % resulted in increasing FC every morning. My theory is that UV caused FC disassociation (I hope I’m using that term right) increases with increased FC, so higher FC levels during the day when using a 12 hr schedule result in more FC loss overall. It may be that the slow and steady FC wins the race. Lowering pump run time may result in the need to increase SWG total run time and reduce cell life. (Caveat: this is all a vague notion at this point)

 

[JoyfulNoise]

Late to the party … love the discussion so far. In my defense, I was making meatballs all day for an evening party so that took precedence. Being of Italian heritage, cooking meatballs and sauce takes precedence over all else. If you need CPR, it’ll have to wait until the meatball mix is finished.

My comment and assertion is based on my experience (many years worth) in electrochemistry. You are assuming a very idealized process where the plates wear down uniformly. I can tell you from direct experience that they don’t. You are also assuming that pool water is a good medium to generate chlorine from. It is not at all. Chlorine generators for pools operate far from ideal conditions and so the more stress you put on the plates by running them at a 100% duty cycle, the worse off the catalytic coating is going to fare.

There are two main factors that affect the electrolysis of chlorine gas - current density distribution across the plate and cell AND mineralization of the plate surface. In a bipolar electrolysis cell, the current “crowds” mainly at the edges of the plate and at any “hot spots” on the surface. Current crowding is a problem in any electrolytic setup. What it typically does is causes the catalytic coating to degrade along edges first and anywhere there are hot spots. At these high current density areas you also get greater precipitation of minerals, mainly calcium, from pool water. The calcium mineralization is non-conductive for the most part so it also exacerbates the irregular current distribution.

To abate some of these problems, cells are designed with a three electrodes where the outer plates are often ground and the center plate is charged. The plates are also set into non-conductive dividers to help keep the current from flowing around the edges where the electric fields are the least uniform. One then does periodic current reversal to alternate which side of the plates are anodes and which are cathodes. This helps to more evenly distribute the inherent damage across all plates as well as reduces the amount of scale formed on any one side. The plate spacing and voltages used are set to minimize as much current variation as possible but there will always be current bulge due to a flowing electrolyte.

My admonition is this - the more you can let a cell “rest” by keeping the output % low, the more chance you give for mineralization to be redissolved into solution. The more you drive the plate with no OFF time, the more those plates sit in a high pH, highly oxidative solution. That chemistry can not only cause the precipitation of scale but also the dissolution of ruthenium metal from the plate. The Pourbaix diagrams for the stability of ruthenium species in chloride solutions gets very complex with multiple different soluble species being possible. Ruthenium is a weird metal in that some of its oxides are insoluble while other oxides can actually exist as a gas phase. Ruthenium hydroxides and chlorides are very soluble in water. It’s complex behavior is what gives rise to its usefulness as a catalyst.

I’m not saying an SWG owner can’t run a cell at 100% when it’s necessary. But it’s far better to design a system where that is unnecessary. You can certainly take the position that you should just drive it as hard as you can and hope for the best, but I guarantee that will not get you the published lifetimes on these cells. I don’t think there’s any way to a priori know how long a setup will last, but prudence is always the safest bet.

 

[jseyfert3]

Current crowding is a problem in any electrolytic setup

Is this also why batteries driven with high charge/discharge currents have shorter lifespans vs an identical battery driven at more modest charge/discharge currents?

 

[JoyfulNoise]

Is this also why batteries driven with high charge/discharge currents have shorter lifespans vs an identical battery driven at more modest charge/discharge currents?

That’s more related to the intercalation of the charged ions into the electrode material (carbon or phosphate based materials). If you try to “shove” the ions into the material faster than the transport kinetics can handle, you’ll swell and damage the crystalline structure. It then makes it harder in subsequent cycles to discharge the ions. Memory effects are related to microscopic damage to the electrodes.

 

[Steve-D]

So looking at all of the above, is it fair to say that the total time that the SWG is in an "on" state would be the more useful means of estimating the life of an SWG against its rated lifetime hours? That is, If my SWG is rated for 10,000 hours of use and I run my pump 24/7 continuously with the SWG set at 50% it "should" last for close to 20,000 hour of pump runtime?

 

[ajw22]

I’m not saying an SWG owner can’t run a cell at 100% when it’s necessary. But it’s far better to design a system where that is unnecessary. You can certainly take the position that you should just drive it as hard as you can and hope for the best, but I guarantee that will not get you the published lifetimes on these cells. I don’t think there’s any way to a priori know how long a setup will last, but prudence is always the safest bet.

So accepting that 100% is not good for the life of the cell where is the threshold % between good and not good? Or do you assert that lower % is better. 10% is better then 30% and 70% is better then 100%?

Do we know what % generation the manufactures lifetime chlorine specifications are based on?

Since the manufacturer specs a T-3 cell for a 25K pool and we know it would take running close to 100% 24/7 to generate that much chlorine then maybe the manufacture takes into consideration the cell degradation at 100% that you are describing.

 

[mas985]

My admonition is this - the more you can let a cell “rest” by keeping the output % low, the more chance you give for mineralization to be redissolved into solution.

This part seems reasonable and logical. However, it would be useful to put a number on the effect. Is there any empirical evidence that would suggest how much the cell life degrades with duty cycle of operation? I would think that the chlorine generation industry would have such numbers. Is the degradation significant by any measure or is it in the noise?

Does the mineralization actually dissolve when off or does the cell just accumulate scale faster per on time unit? The later would seem more likely to me depending on chemistry conditions.

The more you drive the plate with no OFF time, the more those plates sit in a high pH, highly oxidative solution. The more you drive the plate with no OFF time, the more those plates sit in a high pH, highly oxidative solution. That chemistry can not only cause the precipitation of scale but also the dissolution of ruthenium metal from the plate. The Pourbaix diagrams for the stability of ruthenium species in chloride solutions gets very complex with multiple different soluble species being possible. Ruthenium is a weird metal in that some of its oxides are insoluble while other oxides can actually exist as a gas phase. Ruthenium hydroxides and chlorides are very soluble in water. It’s complex behavior is what gives rise to its usefulness as a catalyst.

But isn't this part a linear function of "on" time? I understand the point on mineralization and how that could be a non-linear effect but this seems to me more likely to be a linear effect and would be highly dependent on not only on time but also current.

I’m not saying an SWG owner can’t run a cell at 100% when it’s necessary. But it’s far better to design a system where that is unnecessary. You can certainly take the position that you should just drive it as hard as you can and hope for the best, but I guarantee that will not get you the published lifetimes on these cells. I don’t think there’s any way to a priori know how long a setup will last, but prudence is always the safest bet.

I think this is the crux of the situation. Upsizing a cell is certainly going to be worthwhile for many reasons but if a PO is at the limit of the SWG production, is it more cost effect to run at 100% with some life degradation or to purchase a second unit and run both at 50%? Again, degradation numbers would be useful.

Another scenario comes to mind. When I had a single speed pump, I would run at high SWG % for short run times. BTW, that cell was my longest lasting out of 3 cells. So my question would be, do you see a difference between a 100% SWG with a 6 hour continuous run time and then 18 hour off time vs. 24 hours run time at 25% SWG setting? Technically, they both have the same on/off time just at different frequencies.

 

[JoyfulNoise]

So accepting that 100% is not good for the life of the cell where is the threshold % between good and not good? Or do you assert that lower % is better. 10% is better then 30% and 70% is better then 100%?

Do we know what % generation the manufactures lifetime chlorine specifications are based on?

Since the manufacturer specs a T-3 cell for a 25K pool and we know it would take running close to 100% 24/7 to generate that much chlorine then maybe the manufacture takes into consideration the cell degradation at 100% that you are describing.

This part seems reasonable and logical. However, it would be useful to put a number on the effect. Is there any empirical evidence that would suggest how much the cell life degrades with duty cycle of operation? I would think that the chlorine generation industry would have such numbers. Is the degradation significant by any measure or is it in the noise?

Does the mineralization actually dissolve when off or does the cell just accumulate scale faster per on time unit? The later would seem more likely to me depending on chemistry conditions.


But isn't this part a linear function of "on" time? I understand the point on mineralization and how that could be a non-linear effect but this seems to me more likely to be a linear effect and would be highly dependent on not only on time but also current.


I think this is the crux of the situation. Upsizing a cell is certainly going to be worthwhile for many reasons but if a PO is at the limit of the SWG production, is it more cost effect to run at 100% with some life degradation or to purchase a second unit and run both at 50%? Again, degradation numbers would be useful.

Another scenario comes to mind. When I had a single speed pump, I would run at high SWG % for short run times. BTW, that cell was my longest lasting out of 3 cells. So my question would be, do you see a difference between a 100% SWG with a 6 hour continuous run time and then 18 hour off time vs. 24 hours run time at 25% SWG setting? Technically, they both have the same on/off time just at different frequencies.

I wish I had a good answer for how the catalyst fairs under different duty cycle conditions but I don't ... and I sincerely doubt the pool equipment industry has the data either. Bleach and chlorine manufacturers do not use bipolar cell designs because they are too inefficient. Instead, they use compartmentalized cells where the anode and cathode are separated and they also use feed water that is completely demineralized. Their cells are also designed so that the catalyst panels are easily inspected and removable so that they can keep a continuous batch operation running. Their cells likely operate at 100% duty factor simply by virtue of the process and manufacturing requirements for continuous operation.

The pool industry likely just picked up the chlor-alkali process from the bleach industry and decided to run with whatever design they could easily package and sell. I doubt they have done any exhaustive analysis of plate lifetimes because they don't really need to. It's not like their warranty covers lifetime issues. Their warranties simply attach a fixed time period (1-3 years) to the device regardless of how it is used and then they add lots of restrictions on to it so that claims can be minimized. I don't know of anyone that has ever submitted a claim against failed coatings ... it's either the cell works or it doesn't work within the given warranty period and it's replaced on that basis alone. The 10,000 hour operational life is likely a guesstimate on their part based on chlorine gas output and whatever specs they are told by the plate manufacturers (these are all made by one or two companies in China). That 10,000 hour spec is most likely just marketing and I doubt they really know how long these cells will last.

The only way to truly know what operational parameters have the most effect would be to run an accelerated test on various cells and then do an analysis of the coating afterwards. It would be interesting to run three cells for a fixed amount of chlorine generation (not ON time but actual chlorine generation) at different duty cycles and then inspect the cell and coating to see what differences pop up. My guess is that there would be a noticeable difference.

Someday the TFP Institute for Advanced Studies will have to investigate the issue and give a definitive answer. Until then, I will always caution prudence when operating SWG equipment ...

 

[Flying Tivo]

The more you drive the plate with no OFF time, the more those plates sit in a high pH, highly oxidative solution. That chemistry can not only cause the precipitation of scale but also the dissolution of ruthenium metal from the plate.

So can one assume that if you NEED to run your cell at 100%, then Higher flow is recommended to diminish this oxidative state around the cell? We usually recommend minimum flow for KW/hr savings.