SWG Chlorine Production vs Salt Level and Water Temperature

[mas985]

I thought it might be good to continue this discussion in another thread since I have some new data to share.

Last year I ran a few measurements which indicated that a fixed voltage, variable amp bipolar SWG cell such as mine would produce chlorine at rates proportional to the salt level in the water. Although I would like to confirm with a few more measurements, I am pretty satisfied with the results and how it seems to match theory thanks to Richard's (aka chemgeek) explanations.

However, measurements I did were only at a couple different water temperatures and it showed a very modest change in chlorine production so I wasn't sure temperature was influencing the results. I had read several articles indicating the electrolysis was more efficient at higher solution temperatures so it would make sense that an SWG should produce more chlorine as temperature rises in the water.

I recently measured SWG chlorine production at various temperatures and saw a clear correlation to chlorine production. The first graph of the data below shows why some SWG manufactures will shut off an SWG at low temperatures. The slope of the curve is getting pretty steep there so it probably makes sense not to run the cell.

However, what I didn't expect was that the SWG had a peak production around 80 degrees and then started to roll off again at higher temperatures. Even though the amps increase linearly with temperature, the chlorine production was decreasing. However, based upon the chlorine gas solubility graph show on this page, it would seem the chlorine gas solubility goes down with increasing temperature which would explain the decrease in chlorine production.

What I find most interesting is that water temperature seems to have a larger impact than salt level. Although for a given water temperature, salt levels do make a difference. Just not as much as water temperature.

The bottom graph shows the chlorine production per amp which I think might be a good proxy for cell life although other factors probably come into play as well. There is a small correlation with salt level so a low salt level should impact cell life by only a small amount. However, there is a much higher correlation with water temperature. Since most SWG cells are rated in amp-hours, a very high or very low water temperature could affect cell life much more than the salt level. The higher temperatures being worse than the lower ones.

Note too that the SWG recommended salt levels are between 2700 and 3600 ppm. So the 2600 ppm test was below recommended levels.

SWG Chlorine Production vs Salt Level and Water Temperature

SWG Chlorine Production per Amp vs Salt Level and Water Temperature

 

[chem geek]

The theoretical 2% per 1ºC increase in conductivity with temperature rise would be an 11% increase for every 10ºF. However, the grams per hour per Amp should not change with either temperature or salt concentration. The current should translate directly into production of chlorine gas. It is possible that at lower salt concentrations there is more production of oxygen gas rather than chlorine gas and this transition may be non-linear, but the total amount of chlorine and oxygen gasses should correspond to the current. That is, the electrons never just disappear -- they have to go somewhere towards some sort of chemical reaction.

Are you measuring "grams" by converting the rise in FC per hour into grams per hour given the size of your pool? If so, then any chlorine gas that does not dissolve in the water and instead outgasses would result in a lower measured FC. This should also cause a faster rise in pH. As for chlorine gas solubility, the output from the SWG might be as much as 10 ppm FC downstream, but that's 10 mg/liter or 10 mg/kg so the solubility curves shouldn't be limiting (that is, saturating) the water with chlorine gas. Also, I'm not clear if the engineering toolbox graphs take into account chlorine gas producing hypochlorous acid in the water. Unless the water is near saturation, the rate of chlorine gas dissolving into water should, if anything, occur faster at higher temperatures. Also, it doesn't really matter that much if the gasses are soluble -- they are still swept away by the water flow and even saturated water will still have electrolysis -- after all, the water may get fully saturated with hydrogen and yet hydrogen gas still gets produced (i.e. the cell does not shut down).

So I can understand the lower measured chlorine at lower salt (chloride) levels, even at the same current, if more byproducts (oxygen gas) are produced.

If you graphed the current as a function of temperature, then that would be useful to see if it has an 11% increase for every 10ºF, or thereabouts.

What I cannot explain is the change in chlorine production per amp as a function of temperature. It seems to imply that perhaps that the lower temperature has an effect similar to the lower salt concentration. Though this is clearly true in terms of conductivity and current, the byproducts (oxygen gas) should be more of a function of the chloride concentration and I find it surprising that apparently cooler temperatures favor byproduct production instead of chlorine and even more surprising that the same thing occurs at higher temperatures. The higher temperature drop could be explained by loss of chlorine through outgassing, but I didn't think this happened quite that fast -- but maybe the lower solubility does result in a slower dissolving of chlorine bubbles into the water.

Perhaps seeing how stable the FC level is in the hotter (100F) water with the SWG turned off would provide a baseline to general chlorine loss from the water.

Richard

 

[mas985]

Production may be a bit of a misnomer as I am measuring the residual FC before and after the SWG 30 minute run in my spa. I can't measure production directly so I must infer it from the FC measurements and the volume of the spa. Perhaps a better title is "SWG FC Delivery". So the result depends not only on the production but how much gas gets dissolved in the water. Perhaps it might good to cover the methodology once more:

1. Replace spa water with pool water to start at the same conditions for each test
2. Heat spa water to desired temperature
3. Measure FC level (make sure there are no CC)
4. Run spa with SWG for 30 min (measure SWG volts, amps, temp during test)
5. Measure FC level
6. Calculate delivery rate (FCend – FCstart) * Spa Volume (grams) * 60 / 30

For the amps measurements, I calculate an alpha of 0.015268 using degrees C and 0.00848 for degrees F (close to 11% per 10 degF) and it is pretty linear from 56 degrees to 99 degrees. This is a bit less than the 2% per deg C but that number is for 10% salt content so I am assuming lower salt content should have a lower slope. Is that not correct?

I wasn't expecting the grams/hr/amp to change either over temperature but on the other hand, I wouldn’t necessarily expect constant efficiency (amps to CL) over a range of temperature either. Even if the rate of dissolving CL into the water does not change, other reactions probably do.

Also, I had assumed that because the solubility of chlorine gas in water decreases with increasing temperature, that this might also be true of the rate of dissolving into the water. If the rate of dissolving into the water is slowed down, then out gassing would occur in the spa. However, I agree that it would make more sense that the gas would dissolve faster in warmer water so the discrepancy must be due to production rates.

From what I understand, the concentration of dissolved oxygen and carbon dioxide reduce with increasing water temperature. Would this have any effect on the side reactions? Would it also be possible that the warmer water was speeding up some sort of organic process even though the water was clean and free from CC? I guess you can never be sure the water is free from everything.

Measuring the stability of FC in warm water may be difficult to do because I can’t guarantee that there would be no reactions with organic material getting blown into the spa unless the rate in reduction is quite fast. How fast do you think it would occur? I suppose I could cover the spa overnight.

 

[chem geek]

All good questions for which I don't have any answers. I think it would be a good idea to have a baseline where you do your procedure running the spa, but with the SWG off. See if there is any drop in FC over the 30 minutes and if this varies with temperature. My hunch is that you might see a drop at the higher temperatures.

The amps measurement does seem close enough to the predicted conductivity -- being a little less is not a big deal. I don't know what happens at lower salt concentrations, but the measurement is close enough already.

I really would like to get a handle on the chlorine gas dissolving / outgassing since that is a leading candidate for what makes the pH rise in SWG pools, after accounting for carbon dioxide outgassing. Unfortunately, the pH change in your short experiments probably won't be very much. In the most extreme case going from 70ºF to 100ºF with a roughly 30% drop in apparent chlorine production per amp, then if this loss were all due to chlorine outgassing, then the difference in pH at a TA of 80 and CYA of 80 would be the following:

FC increase of 10 ppm would result in pH rising from 7.5 to 7.96 (about 8.0)
FC increase of 7 ppm with 3 ppm chlorine gas outgassing would result in pH rising from 7.5 to 8.19 (about 8.2)
FC increase of 7 ppm with 3 ppm oxygen gas produced would result in pH rising from 7.5 to 7.78 (about 7.8)

This difference is probably too subtle to measure definitively though just raising the temperature without adding any chlorine has the pH drop a little from 7.5 to 7.41 so seeing a higher pH rise at higher temperatures would point to chlorine gas outgassing as a possibility whereas if the pH rise were less at higher temperatures, then this would point to byproducts such as oxygen gas being produced.

Richard

 

[mas985]

I need more FAS-DPD powder so if I get a chance to do this again, I'll make sure I test the PH before and after as well although I'm not sure I can get very accurate PH measurements. I'll also make a third FC measurement an additional 30 min after the test.