Salt water generators are closed systems... what?

[jmastron]

In previous posts it was mentioned that splash out will lower salt levels. I know evaporation will increase salt levels and dilution will reduce levels. I wouldn't think splash out would change anything.

It's not the splashout itself that lowers levels; it's adding unsalted water to replace the splashout that dilutes/reduces the salt level. Same with rain-induced draining, backwashing, etc where you replace water.

IRight, but the chloride ions go somewhere, or else the pool would continuously increase in chloride saturation.

As I understand, CYA can bind to chloride ions so whatever process causes them to leech from the water doesn't occur, but what actually is the process it's preventing? Do the positive and negative chlorine ions in the water reform into chlorine gas if CYA isn't there and UV light acts as a catalyst?

As others have said (much more scientifically), that's exactly what happens. I've been using liquid chlorine since we bought the house in 2014. In December 2016 I bought a K1766 test kit and measured the salt level at 1600ppm. I recall testing occasionally since then at higher levels (~1900-2000 ppm a couple years later) but didn't write the values down. This summer I bought and SWCG and tested the sale level -- 2400 ppm. Using Pool Math that actually matches what adding 86 gallons of liquid chlorine should do. I've used more than that, but accounting for winter rain drain (no summer rain here), splashout, etc, that seems in the ballpark. I only had to add 4 bags to get to 3500ppm and make the cell happy.

 

[phonedave]

The efficiency isn't what matters to chloride ion concentration. It's actually simpler than that - the resistance of the electrochemical cell is inversely proportional to the chloride concentration. If the chloride concentration is too low, then the cell resistance will be too high. This will limit the current flow which is what determines how much chlorine is generated. And if the resistance of the cell is too large, the voltage needed to make current flow will be so high that you will parasitically produce more oxygen and hydrogen gas (water electrolysis) than chlorine gas. So ultimately the required chloride ion concentration vis-a-vis salt additions is dominated by the required electrical conductivity of the solution.

I was going to go that route and mention that cell requirements dictate the conductivity of the water, which is based on the concentration of free ions in solution, but then I thought that was MORE complicated than just saying it's not 100%

 

[Abnaxis]

Cell voltage is bounded, you can’t apply whatever voltage you want to fix a specific current. Cell voltage is what controls the electrolysis reactions and if the voltage gets too high you will generate parasitic chemical species like oxygen gas and chlorates. If the voltage is too low, you won’t generate any chlorine. These power supplies have to operate in a bounded voltage-current domain or else they won’t do what you want them to. Whether you pick current or voltage as the control point is somewhat arbitrary. Temperature also affects what reactant are formed - too low and you’ll get more oxygen than chlorine; too high and you’ll start to form chlorates (although not usually at pool/spa temps). This is one reason why cells cutoff below a certain temp.

I understand all that, BUT hear me out. Like I said, I did a bunch of research work in electrolysis of material that ISN'T chlorine, but AFAIK there's nothing fundamentally different about chlorine than what I dealt with, so here's my thought process.

If you draw a (idealized, not accounting for bubbling effects and other nuisance physics things) graph between cell voltage and rate of gas production, it looks something like this:



(Sorry if this seems condescending. From the posts here, it obvious most people in the thread understand all this but I'm going through it in minute detail so I can get my thought process out more explicitly)

For this graph, the threshold voltage Vt and the slope of the purple line are what I'm focused on, because those change with concentration, water temp, and water chemistry (pH definitely has an effect, though it might not be much for the relatively narrow pH band pools stay in).

The reason why I'm honed in on preferring a system that is designed to control current (note: as JN is saying what you supply to the cell is voltage, but you can design a dumb circuit that will modulate voltage and reach an adjustable "steady state" current within the limitations of its power supply) is because the relationship between actual grams of chlorine produced and current running thruogh the system is linear, while the relationship between voltage and chlorine mass generated is not. In other words, if I double the current running through the cell I will always get double the chlorine, whereas if I double the voltage I will get more chlorine, but without knowing what the starting voltage is, what Vt is, and what the cell "resistance" (inverse of the slope of the purple line) is I can't tell you how much more chlorine is being generated--and there's no practical way to know Vt and resistance ahead of time since temperature and chemistry will change them.

From what I'm hearing others say, that's is OK to them because the conditions where you need more chlorine are also conditions where Vt and resistance are low. Vt and resistance drop when temperature rises so chlorine production will rise when temperature rises, but that's good because you need more chlorine at higher temps anyway so it's a nice little auto-adjustment to the environment. My own attitude toward it is I prefer knowing exactly how much chlorine mass I'm adding, because I'm a control freak

For the curious, the above curve also explain a lot of the other things SWCG-related that JN has sort of mentioned off-hand in this thread. Oxygen also has a curve exactly like the one pictured with its own values for Vt and resistance, the only reason why you don't produce oxygen in your cell is because the catalyst layer in the cell keeps Vt and resistance low enough for chlorine that you can make enough chlorine for sanitizing without going above Vt for oxygen. A low temperatures, however, Vt and resistance increase enough that you have to go over the Vt for oxygen to produce enough chlorine, which is why cells have a temperature safety cutoff.

 

[Orion7319]

I see you are in Tucson, AZ and I'm in Bradenton , Fl. Both are very warm locations. Are there any advantages in having a slightly higher salt level to lower the amount of voltage needed by the cell in warm locations (my water temp under screen can hit 90)? I guess the real question I'm asking is if there is anyway to fine tune the work done by the SWG to get its maximum life or are the differences so minuscule that it doesn't matter

In the case of the aquarite if the salt is on the high side the system works too hard and will eventually burn out a specific component on the board. It’s an easy and cheap fix if you know how to solder and have a decent soldering station. Point is that if you run these on the low or high side of the ranges you could be asking for trouble. Just keep the salt in the middle of its range and don’t worry about it. As I mentioned before, you can maximize its life by doing things like keeping the CYA a bit higher, which keeps the chlorine from burning off as quickly so you don’t have to run the salt cell as hard to replace that extra chlorine that you lost. Also if you have a VSP you could consider running it 24/7 at a lower level. This makes it easier to build up a surplus of chlorine because your running it at night.

 

[Orion7319]

I understand all that, BUT hear me out. Like I said, I did a bunch of research work in electrolysis of material that ISN'T chlorine, but AFAIK there's nothing fundamentally different about chlorine than what I dealt with, so here's my thought process.

If you draw a (idealized, not accounting for bubbling effects and other nuisance physics things) graph between cell voltage and rate of gas production, it looks something like this:

View attachment 359275

(Sorry if this seems condescending. From the posts here, it obvious most people in the thread understand all this but I'm going through it in minute detail so I can get my thought process out more explicitly)

For this graph, the threshold voltage Vt and the slope of the purple line are what I'm focused on, because those change with concentration, water temp, and water chemistry (pH definitely has an effect, though it might not be much for the relatively narrow pH band pools stay in).

The reason why I'm honed in on preferring a system that is designed to control current (note: as JN is saying what you supply to the cell is voltage, but you can design a dumb circuit that will modulate voltage and reach an adjustable "steady state" current within the limitations of its power supply) is because the relationship between actual grams of chlorine produced and current running thruogh the system is linear, while the relationship between voltage and chlorine mass generated is not. In other words, if I double the current running through the cell I will always get double the chlorine, whereas if I double the voltage I will get more chlorine, but without knowing what the starting voltage is, what Vt is, and what the cell "resistance" (inverse of the slope of the purple line) is I can't tell you how much more chlorine is being generated--and there's no practical way to know Vt and resistance ahead of time since temperature and chemistry will change them.

From what I'm hearing others say, that's is OK to them because the conditions where you need more chlorine are also conditions where Vt and resistance are low. Vt and resistance drop when temperature rises so chlorine production will rise when temperature rises, but that's good because you need more chlorine at higher temps anyway so it's a nice little auto-adjustment to the environment. My own attitude toward it is I prefer knowing exactly how much chlorine mass I'm adding, because I'm a control freak

For the curious, the above curve also explain a lot of the other things SWCG-related that JN has sort of mentioned off-hand in this thread. Oxygen also has a curve exactly like the one pictured with its own values for Vt and resistance, the only reason why you don't produce oxygen in your cell is because the catalyst layer in the cell keeps Vt and resistance low enough for chlorine that you can make enough chlorine for sanitizing without going above Vt for oxygen. A low temperatures, however, Vt and resistance increase enough that you have to go over the Vt for oxygen to produce enough chlorine, which is why cells have a temperature safety cutoff.

I think your concentrating too much on the potential efficiency of the salt cell alone and almost completely ignoring the overall efficiency of the pool system as a whole.

 

[Abnaxis]

I think your concentrating too much on the potential efficiency of the salt cell alone and almost completely ignoring the overall efficiency of the pool system as a whole.

I almost couldn't care less about efficiency, I care about knowing exactly what's going into my water to the best approximation I can get the information. If I know how much current my cell is drawing, I know exactly how much chlorine is being generated. I can't say the same if all I know is the voltage across it. Like I said, I'm a control freak

This kind of brings it back to the original gist of what motivated me to make this thread in the first place, which is that I'm almost constantly asking "but then where does it go?" when I'm reading about water chemistry treatments and different systems for generating the chlorine to sanitize the pool. If I add an acid or a base to modify pH, it's not like that chemical is just turning into gas to float off out of the water (except in the case of sodium bicarbonate), it's part of the chemistry now. And my research into the academic literature on the subject gives me the impression that nobody really knows what's going on in the water (organic chemistry is hard, yo), we just kind of keep doing what's worked for that last however-many centuries and hoping it keeps working as intended without fully understanding all the underlying processes.

This makes me uneasy because I'm getting ready to inherit a pool from previous owners who were very much not conscientious about maintaining ANYTHING about the house we've bought. I'm about thiiiiiiiis close to doing a full water exchange on the thing out of paranoia.

 

[Orion7319]

I almost couldn't care less about efficiency, I care about knowing exactly what's going into my water to the best approximation I can get the information. If I know how much current my cell is drawing, I know exactly how much chlorine is being generated. I can't say the same if all I know is the voltage across it. Like I said, I'm a control freak

This kind of brings it back to the original gist of what motivated me to make this thread in the first place, which is that I'm almost constantly asking "but then where does it go?" when I'm reading about water chemistry treatments and different systems for generating the chlorine to sanitize the pool. If I add an acid or a base to modify pH, it's not like that chemical is just turning into gas to float off out of the water (except in the case of sodium bicarbonate), it's part of the chemistry now. And my research into the academic literature on the subject gives me the impression that nobody really knows what's going on in the water (organic chemistry is hard, yo), we just kind of keep doing what's worked for that last however-many centuries and hoping it keeps working as intended without fully understanding all the underlying processes.

This makes me uneasy because I'm getting ready to inherit a pool from previous owners who were very much not conscientious about maintaining ANYTHING about the house we've bought. I'm about thiiiiiiiis close to doing a full water exchange on the thing out of paranoia.

The pool math app seems to be able to calculate the salt water cell output fairly precisely. It has information for all sorts of salt cells and it knows how many pounds of chlorine they can produce in a 24 hour period. I don’t know how they came up with those numbers, probably from the manufacturer. I can tell you they seem to work. A lot of the folks here have a pretty good understanding of the chemistry and have really drilled down into it. That’s how they came up with the methodology they use here. I don’t claim to understand it all, but I can say that if you understand the basics and the methodology maintaining the pool is super simple. I certainly understand your concern about buying a house with a pool. I’ve been there done that.

 

[mas985]

I almost couldn't care less about efficiency, I care about knowing exactly what's going into my water to the best approximation I can get the information. If I know how much current my cell is drawing, I know exactly how much chlorine is being generated. I can't say the same if all I know is the voltage across it. Like I said, I'm a control freak

It isn't that easy to go from Amps to FC. The geometry of the T-15 cell is such that it is really 2 parallel cell sets with 6 cells each.

For example, the T-15 cell is said to generate 1.45 lbs/day. However, if you do the math from current to lbs/day, it doesn't add up. It is about 50% short. These values are taken from actual measurements:



Why is this? Two things come to mind. This only assumes the current travel between each of the 6 plates in series but at the ends of the cells, current could easily bypass some of the plates. Also, this assumes all the current goes into producing chlorine but we know that oxygen has a lower redux potential so could be produced under the correct conditions

 

[Orion7319]

For example, the T-15 cell is said to generate 1.45 lbs/day.

That number works, that number is accurate, it’s simple to work with that number to maintain my pool. If I run my cell 100% for 24 hours it’s going to make that much chlorine. I don’t need to know how much current it’s using to know how much chlorine it’s going to make.

 

[mas985]

That number works, that number is accurate, it’s simple to work with that number to maintain my pool. If I run my cell 100% for 24 hours it’s going to make that much chlorine. I don’t need to know how much current it’s using to know how much chlorine it’s going to make.

Exactly, normally it is close enough. Plus if you are testing regularly, what you really need to know is what the residual is. That is what really matters and not how much the SWG is putting in. If your FC target is 4 ppm but the residual is 2 ppm, you just need to double the output to hit target. It doesn't really matter what the SWG is actually producing. It is all relative.

 

[Orion7319]

Exactly, normally it is close enough. Plus if you are testing regularly, what you really need to know is what the residual is. That is what really matters and not how much the SWG is putting in. If your FC target is 4 ppm but the residual is 2 ppm, you just need to double the output to hit target. It doesn't really matter what the SWG is actually producing. It is all relative.

Right! All you need to know is how much chlorine the pool is loosing per day and how much to run the salt cell to replace what the pool is loosing per day. If your target is 4 just shoot for 6 - 8 and don’t worry about it. And as I was saying you can manipulate your CYA and the runtime and the output of the salt cell and you won’t run into any trouble unless something totally out of the blue happens.

 

[JoyfulNoise]

It isn't that easy to go from Amps to FC. The geometry of the T-15 cell is such that it is really 2 parallel cell sets with 6 cells each.

For example, the T-15 cell is said to generate 1.45 lbs/day. However, if you do the math from current to lbs/day, it doesn't add up. It is about 50% short. These values are taken from actual measurements:

View attachment 359639

Why is this? Two things come to mind. This only assumes the electrons travel between each of the 6 plates in series but at the ends of the cells, electrons could easily bypass some of the plates. Also, this assumes all the electrons go into producing chlorine but we know that oxygen has a lower redux potential so could be produced under the correct conditions

@mas985

You are quite correct, the SWG is definitely not a simple linear system where one has to simply account for the current and then easily determine the number of moles of chlorine produced. Heck, the system doesn't even operate in the linear Tafel portion of the i-E electrolysis curve but way out in the mass-transport limited region.

SWGs are what are known as "bipolar" electrolysis cells. Basically, the two outer plates are charged and the innermost middle plate is common (or it can be vice versa depending on how many connections they want to make). The rest of the plates are all unconnected but become charged up through coupling to the electrodes via the solution (sort of like capacitive charging). Bipolar cells are used because they are very easy to build, don't require a bus bar to connect plates and can be made very compact. However, bipolar electrodes suffer from very high parasitic electrical currents that are modified by the solution flow through them. In other words, not all of the current flowing into the cell is used for electrolysis. This is why your data shows a very low efficiency - it's due to the parasitic flow of current outside and around the electrodes.

So the operating voltages and currents have very little to do with optimized electrochemical efficiency but rather are set to produce a given chlorine production figure. Chlorine output is typically modified by adjusting the number of plates in the cell and/or changing the surface area of the cell as well as setting the current and voltage limits.

 

[Abnaxis]

It isn't that easy to go from Amps to FC. The geometry of the T-15 cell is such that it is really 2 parallel cell sets with 6 cells each.

For example, the T-15 cell is said to generate 1.45 lbs/day. However, if you do the math from current to lbs/day, it doesn't add up. It is about 50% short. These values are taken from actual measurements:

View attachment 359639

Why is this? Two things come to mind. This only assumes the current travel between each of the 6 plates in series but at the ends of the cells, current could easily bypass some of the plates. Also, this assumes all the current goes into producing chlorine but we know that oxygen has a lower redux potential so could be produced under the correct conditions

What I care about isn't whether the cell is "ideal" or 100% efficient. I care that the relationship between the "%" setting and the chlorine produced is linear, and that the production doesn't swing wildly with water conditions. In other words, if a T-15 produces 1.45lb per day at 100% setting at pH 7.0 and 80°F water, I want it to produce 0.72lb when it's set to 50%, 0.29lb when set to 20%, etc, and I want it to keep producing close to those levels even if it's working on 7.7pH and 71°F water.

To my mind, that's a lot harder to do if you're trying to modulate voltage to maintain your set-point compared to modulating current, unless there's more going on in the details of the control circuitry of these things. Which brings me to:

@mas985

You are quite correct, the SWG is definitely not a simple linear system where one has to simply account for the current and then easily determine the number of moles of chlorine produced. Heck, the system doesn't even operate in the linear Tafel portion of the i-E electrolysis curve but way out in the mass-transport limited region.

SWGs are what are known as "bipolar" electrolysis cells. Basically, the two outer plates are charged and the innermost middle plate is common (or it can be vice versa depending on how many connections they want to make). The rest of the plates are all unconnected but become charged up through coupling to the electrodes via the solution (sort of like capacitive charging). Bipolar cells are used because they are very easy to build, don't require a bus bar to connect plates and can be made very compact. However, bipolar electrodes suffer from very high parasitic electrical currents that are modified by the solution flow through them. In other words, not all of the current flowing into the cell is used for electrolysis. This is why your data shows a very low efficiency - it's due to the parasitic flow of current outside and around the electrodes.

So the operating voltages and currents have very little to do with optimized electrochemical efficiency but rather are set to produce a given chlorine production figure. Chlorine output is typically modified by adjusting the number of plates in the cell and/or changing the surface area of the cell as well as setting the current and voltage limits.

THAT'S what I was interested in--essentially, it doesn't matter so much how the power supply to the circuit is modulated, because that isn't what controls the production rate of chlorine. Paraphrasing myself from earlier, "voltage modulation control is terrible unless there's something smarter happening in the background"--well, there's something smarter happening in the background .

I guess they're enabling and disabling banks and/or physically moving plates around to control the production rate? I'm curious how they're changing the surface area of the electrodes, but as long as there's SOMETHING ensuring a consistent linear calibration of the cell I'm happy. I still think current modulation sounds better on the face of it for a bunch of other reasons, but my inner control freak who wants to know how many pounds of chlorine I'm producing when I fiddle with stuff is placated.

 

[JamesW]

production doesn't swing wildly with water conditions.

The water conditions don't swing wildly. The salinity should stay relatively constant.

The water temperature can change quickly with a heater and that can have some effect.

In the real world, the problems that you worry about really do not happen.

Keeping the FC within a good zone, takes some trial and error anyway to keep it constantly dialed in.

Conditions that use up chlorine, like sunlight change constantly, so thinking that you can find one exact production rate that will provide the perfect levels is unrealistic.

In the real world, all good SWGs work just fine including the Aquarite.

I would consider the Aquarite a good quality and reliable SWG with none of the problems that you are speculating about.

Maybe you need to get some actual experience with some SWGs and then come back with some real world data vs merely hypotheses and speculation.

 

[JamesW]

Also, if the SWG is producing more than the pool needs, it does not cause a runaway condition because the amount of loss is proportional to the amount of chlorine.

So, overproduction causes the chlorine to increase by a few ppm until the loss rate reaches an equilibrium with the gain rate.

If the pool is covered, then you can get a runaway condition where the FC continues to climb.

In some cases, the FC went over 100 ppm due to using a SWG with an opaque safety cover.

 

[JimMarshall]

Right, but the chloride ions go somewhere, or else the pool would continuously increase in chloride saturation.

As I understand, CYA can bind to chloride ions so whatever process causes them to leech from the water doesn't occur, but what actually is the process it's preventing? Do the positive and negative chlorine ions in the water reform into chlorine gas if CYA isn't there and UV light acts as a catalyst?

This chloride buildup is one of the things that the sorcerers at the pool store report to you as "TDS"

 

[jmastron]

What I care about isn't whether the cell is "ideal" or 100% efficient. I care that the relationship between the "%" setting and the chlorine produced is linear, and that the production doesn't swing wildly with water conditions. In other words, if a T-15 produces 1.45lb per day at 100% setting at pH 7.0 and 80°F water, I want it to produce 0.72lb when it's set to 50%, 0.29lb when set to 20%, etc, and I want it to keep producing close to those levels even if it's working on 7.7pH and 71°F water.

To my mind, that's a lot harder to do if you're trying to modulate voltage to maintain your set-point compared to modulating current, unless there's more going on in the details of the control circuitry of these things. Which brings me to:



THAT'S what I was interested in--essentially, it doesn't matter so much how the power supply to the circuit is modulated, because that isn't what controls the production rate of chlorine. Paraphrasing myself from earlier, "voltage modulation control is terrible unless there's something smarter happening in the background"--well, there's something smarter happening in the background .

I guess they're enabling and disabling banks and/or physically moving plates around to control the production rate? I'm curious how they're changing the surface area of the electrodes, but as long as there's SOMETHING ensuring a consistent linear calibration of the cell I'm happy. I still think current modulation sounds better on the face of it for a bunch of other reasons, but my inner control freak who wants to know how many pounds of chlorine I'm producing when I fiddle with stuff is placated.

I get you, I'm an engineer and like things to fit into clear specs and calculatable boxes. But I don't think the chlorine generation for any of these cells is quite that precise -- "rated 1.45lbs/day" doesn't mean it generates exactly 1.93oz per hour no more no less. I'm sure water temperature, salt levels, manufacturing variances play a part. I think it's more like automobile MPG ratings.

I'm guessing the 12.5% liquid chlorine I used to buy wasn't exactly that every time either -- the production plant probably has a minimum/maximum concentration range, then the transportation and storage results in variances too. And in my pool the amount of sun each day, when or if the cover gets taken off, water temp, how much fill water I add all change the chlorine consumption so knowing I produce exactly X.XX pounds vs +/- 10% wouldn't help much.

That's what I love about the TFP methods -- we do pool math calculations to get pretty close estimates of the chemicals to add and what they'll do, but always rely on good testing for feedback. If the Muriatic Acid I put in yesterday is a bit weaker than expected, I'll just notice the pH be at my upper limit a day earlier next week and will add accordingly.

I'm very happy with my $625 Calimar SWCG (a generic equivalent to the Aquarite) so far and am happy to fiddle with the settings a bit to keep things where I want them (I actually leave it at 100% and use a WiFi timer that I can remotely control).

 

[mas985]

I'm very happy with my $625 Calimar SWCG (a generic equivalent to the Aquarite) so far and am happy to fiddle with the settings a bit to keep things where I want them (I actually leave it at 100% and use a WiFi timer that I can remotely control).

How long have you had the Calimar cell?

 

[jmastron]

How long have you had the Calimar cell?

Bought in May (the whole unit including control center, in case that wasn't clear), and have been using it since early June. Seems to be working well, once I raised the CYA level to ~70 running 100% for 5-6 hours is enough to maintain FC with 88-90 degree water and the cover off (I reduce it when the cover is on during the day).

 

[riny]

You know, I just found this thread after asking a very similar question about Cl- build-up. Everything makes sense but I'm hung up on something:

Salt is NaCl. The salt water chlorine generator splits the NaCl into Sodium (Na) and Chlorine (Cl) molecules.

To be pedantic, I'm not sure this is correct. I'm pretty sure the water does that. Only the Cl- participates in the reaction in the salt sell; the Na+ is needed to make the water conductive. You don't actually need the sodium to make hypochlorous acid but the electrolysis would be impossible without it. In theory, you could use other chloride salts to get the same effect. I hope someone will correct me if I'm wrong about this.

Many people who convert to SWGs make the mistake of thinking their water has zero chloride ion in it and calculate their salt need based on that when, in fact, they are almost 1/3rd of the way to a salt water pool. Every gallon of 10% LC added to 10,000 gallons of pool water adds 16ppm salt which comes from both the excess salt water in liquid chlorine from the manufacturing process as well as the breakdown of hypochlorite to chloride ion.

Ok here we go, continuing the thought from above. When you say "salt" here (and when we test salt levels in the pool) most people probably think "table salt" like @JJ_Tex said, but I assume we're using the chemistry definition and if my understanding is correct, the sodium doesn't really matter. If you use bleach or chlorinating liquid, that's NaClO (sodium hypochlorite) and it will provide enough Na+ to make this work like you describe. But if you use dichlor or trichlor pucks, those don't contain any sodium. Is there some other positively-charged ion the Cl- will bond with in this case?

Because if not, if you run your pool on things like tabs and cal-hypo for a long time and never introduce any sodium, then you decide to convert to salt... is the excessive amount of existing Cl- going to cause a problem?

I know we don't like those products here (for good reason) but I come across so many people who swear by them. And these days with the national shortage and everyone converting to salt, I'm thinking this could be a very plausible scenario.