The CSI is Reliable for Plaster-based Pools

onBalance

TFP Expert
LifeTime Supporter
In The Industry
Jul 25, 2011
976
Utah
Using the Calcium Saturation Index (also known as the LSI in the pool industry) as a guide for maintaining proper pool water balance and to protect pool plaster, including quartz and pebble finishes, has become a mainstay in our industry for good reason.

Several experiments have been conducted on plaster coupons and we have determined that maintaining a balanced CSI (-0.5 to +0.5) helps prevents scaling and prevent the etching of plaster finishes.

Unfortunately and incorrectly, some plaster people are trying to claim that even when pool water is CSI balanced, if the alkalinity level is below 80 ppm (the industry’s Ideal minimum, not Operating minimum, which is 60), the pool water is aggressive and can discolor or damage pool plaster. Not so.

Anyone that is experienced in calculating the CSI (or LSI) knows that a low alkalinity can be offset by raising the calcium level and/or raising pH to keep the water balanced. To demonstrate and confirm that, we conducted several plaster experiments.

For example, two quality pool plaster coupons were made and cured in balanced water for 30 days. At that point, one coupon was placed into water that had a low Total Alkalinity of 40 ppm, well below the minimum standard. The Calcium level was set at 350 ppm and the pH was maintained from 7.8 to 8.0, which off-set the low alkalinity and achieved a balanced CSI of about -0.1 to +0.1.

The other plaster coupon was placed into water with a low Total Calcium level of 90 ppm. The alkalinity was maintained at 130 ppm and the pH from 7.8 to 8.2, which together off-set the low calcium level and achieved a balanced CSI of about -0.1 to +0.3.

After six months in the water, the plaster coupons were removed from the water containers, and the water was tested for the calcium content to determine if any dissolution and etching of plaster surface material occurred. (An increase in calcium from the water’s starting point would indicate that calcium had been dissolved from the plaster coupon, which contains the only available source of calcium in this experimental set-up).
The result? There was zero increase of calcium in either water container.

The results of this experiment are totally consistent with the concept of calcium carbonate saturation as published in the municipal water supply, as well as the swimming pool chemistry industries. The results demonstrate that even when the alkalinity or calcium level is low – below the APSP’s minimum standard – but the CSI calculation shows that the water is balanced, then the water IS balanced, and therefore, not aggressive. But understand, we aren’t recommending that maintaining low alkalinity or calcium hardness levels is the preferred way. This is just information to help pool builders, service techs, and pool owners confront false claims about water chemistry and various plaster problems.

The False Claim Regarding CSI Aggressive Water

It is very unfortunate that sometimes new pool plaster surfaces develop either gray mottling, white soft spotting and streaking discoloration, flaking (spalling), craze cracks, calcium nodules (from delamination), or other such defects, all of which are attributable to poor plastering workmanship, and that instead of advising the plastering company that improper plastering practices occurred, one NPC inspector (incorrectly) tells pool builders, service techs, and pool owners that aggressive water causes the above problems.

As mentioned above, the blame is often directed at low alkalinity or calcium hardness levels even though the actual CSI is balanced! But more importantly, the above listed plaster problems are NOT caused by aggressive water. Cement literature, empirical evidence, and even the NPC/Cal Poly/NPIRC study results demonstrate that.

Even when pool water IS aggressive, and etches plaster, it etches it uniformly, not in a mottled, spotted, or streaked pattern unless the plaster is non-uniformly defective as well.

It is our hope that pool builders, service techs, and pool owners will not be fooled by such non-sense. But it is happening. If your pool develops one of the above plaster problems, do not be fooled or intimidated, and be sure to challenge the concept that unbalanced water caused the problem.

Note: The above information does not apply to metal or mineral staining (copper, iron, or calcium, for example). That type of staining problems are not a plastering defect.
 
Last edited:

JamesW

TFP Expert
Mar 2, 2011
16,598
Great post. Thanks for doing the study! :goodjob:

The calcium carbonate saturation depends on the combination of calcium and carbonate levels.

The carbonate level depends on the carbonate alkalinity and the pH.

Bottom line, as long as the CSI is maintained in the correct range, the pool should not experience etching or scale.
 

bdavis466

TFP Expert
Platinum Supporter
Aug 4, 2014
5,062
San Clemente, CA
I particularly enjoy the claims from builders/plasterers that aggressive water chemistry caused the plaster discoloration but their suggested fix is an acid wash. So aggressive water discolored the plaster but I'll dump acid directly on it - that'll fix it.....:scratch:

I've also seen several claims that not enough brushing during the startup causes discoloration. They have these poor people filled with fear that they must brush their pools 4X a day during the startup and unfortunately some of that brushing nonsense gets suggested here. I'm not saying brushing is not important but if your going to fret over brushing something then focus it on your teeth :mrgreen:
 

JamesW

TFP Expert
Mar 2, 2011
16,598
I think that brushing is necessary for the first 2 weeks.

I would say at least daily.

Most startup guides require daily brushing.

4x per day is probably too much.
 

mas985

TFP Expert
LifeTime Supporter
May 3, 2007
12,800
Pleasanton, CA
Several experiments have been conducted on plaster coupons and we have determined that maintaining a balanced CSI (-0.5 to +0.5) helps prevents scaling and prevent the etching of plaster finishes.
Just curious if you were able to determine the relationship between the degree of scaling/etching vs CSI? Is it linear? What makes +- 0.5 the magical threshold? Is it based upon the time required to experience significant scaling/etching?

Logically, it would seem that you need a time element in conjunction with a CSI level to determine the extent of scaling/etching.
 

onBalance

TFP Expert
LifeTime Supporter
In The Industry
Jul 25, 2011
976
Utah
It is really nice and beneficial to have a forum where relevant and knowledgeable comments can be made for a better understanding of the issues.

The CSI (or LSI) is logarithmic, but that isn't the end of the story. It also depends on the quality of the plaster, and yes, a time factor before visible etching is seen.
Also, something that I should have mentioned in my write-up is that the CSI is not applicable to new plaster finishes under 30 days old. It is actually necessary, in order to achieve a smooth and dense surface, to have about a +0.5 CSI during the first 30 days. This is because the plaster (cement paste) surface contains about 20% calcium hydroxide, which is somewhat soluble in balanced and slightly positive CSI water and can be dissolved away. The plaster surface needs to be "carbonated" before the CSI should be lowered to the acceptable and balanced range. And that generally is achieved during the first month under balanced water.

The +-0.5 CSI was merely chosen because that is the range assumed (by the pool industry) to be in a relatively balanced condition. However, I am sure that if given enough time, 4 or 5 years at least, even a "constant" -0.5 CSI may start showing some signs of uniform etching of quality applied plaster. If I remember my numbers correctly, it requires about 40 pounds (250 ppm calcium hardness increase) of calcium carbonate to be removed from a 20,000 gallon plaster pool for visible etching to be seen. Of course, under magnification etching will be seen sooner.

But let's also understand that greater amounts of calcium loss (leaching in balanced water) can occur when the plaster mixing and application is done poorly.
 
Last edited:

mas985

TFP Expert
LifeTime Supporter
May 3, 2007
12,800
Pleasanton, CA
The CSI (or LSI) is logarithmic, but that isn't the end of the story. It also depends on the quality of the plaster, and yes, a time factor before visible etching is seen.
I understand that the CSI calculation has logarithmic components (e.g. TA & CH) and it is linear with PH. However, what I am interested in is the relationship between CSI and etching/scaling rates. Etching/scaling rates cannot possibly have a logarithmic relationship with CSI because CSI can be negative and logarithms are undefined for negative numbers.

So let me simplify the question, if CSI doubles, what happens to the etching/scaling rate? Does it double, more than double or less than double?
 

onBalance

TFP Expert
LifeTime Supporter
In The Industry
Jul 25, 2011
976
Utah
I have checked with a PhD chemist friend of mine, and he told me that since the pH scale is logarithmic, so is the CSI. He suggests that a CSI of -2.0 is ten times more aggressive than a CSI of -1.0 and of course, a CSI of +2.0 is ten times more scale forming than a CSI of +1.0, which correlates with the pH scale.
 

mas985

TFP Expert
LifeTime Supporter
May 3, 2007
12,800
Pleasanton, CA
Ok, then I think what you are really saying is that CSI is logarithmic to etching/scaling thereby making etching/scaling exponential with CSI which would make more sense than the inverse. Semantics but an important distinction.
 

setsailsoon

LifeTime Supporter
TFP Guide
Oct 25, 2015
1,592
Stuart/FL
OnBalance,

Great information and thanks! What's the source of this? I use pool math and I like the simplicity of using csi. But I've never been comfortable with the different ranges suggested by experts. I've seen +/- .6 to +/-.3 but I really only care about the change of calcium (etching/depositing). I find my pool is difficult to keep at ideal pH so I run total alkalinity lower and pH higher. I also seem to have no problem controlling csi to +/- .3 and it looks like that is a good place to be based on your data. Importantly my pool also is always sparkling clear... and all this for about $10 per month.

Thanks again for this important data!

Chris
 

ajw22

Gold Supporter
TFP Guide
Jul 21, 2013
10,794
Northern NJ
Also, something that I should have mentioned in my write-up is that the CSI is not applicable to new plaster finishes under 30 days old. It is actually necessary, to achieve a smooth and dense surface, to have about a +0.5 CSI during the first 30 days. This is because the plaster (cement paste) surface contains about 20% calcium hydroxide, which is somewhat soluble in balanced and slightly positive CSI water and can be dissolved away. The plaster surface needs to be "carbonated" before the CSI should be lowered to the acceptable and balanced range. And that generally is achieve during the first month under balanced water.
This is a very significant comment and should be added to:
Pool School - Start-up New Plaster
Pool School - Calcium Saturation Index (CSI))

I see people with new plaster immediately trying to adjust their TA to get their CSI negative. This is good guidance to leave TA high and not worry about PH in the high side of the 7's that gives a positive CSI.
 

onBalance

TFP Expert
LifeTime Supporter
In The Industry
Jul 25, 2011
976
Utah
Mark, although I have not performed a direct and specific experiment to address your question and determine the degree of etching for different CSI levels, I do think I have something that may shed some light on that.

Most of my experiments have focused on the effects of water balance on brand new plaster, as opposed to cured plaster, and also on poorly applied plaster which means they aren't applicable to your question. But I have found two experiments (that was performed about ten years ago) with good plaster that had been been properly cured and did a six month test with aggressive water. Also, another experiment where I performed a zero alkalinity acid treatment to determine the effects that could be used to compare with.

So here are the numbers: Experiment 1. Maintained an average CSI of about -0.7 for six months and etched about 45 ppm of calcium carbonate. Experiment 2. Performed a zero alkalinity treatment by lowering the CSI to about -3.5 and left for one week. After one week, I determine that about 40 ppm of calcium carbonate was removed which increased the starting calcium level from 200 to 240 ppm. At the end of the one-week period, the CSI was about -1.5. The beginning pH was 4.5 and had risen to about 6.5. The alkalinity started at zero and then raised to about 40 ppm. So the question is, which I unfortunately did not observe and determine at the time; what was the average CSI for the one-week period?

Obviously, once the plaster coupon was placed into the low CSI water, the pH and alkalinity would being to rise immediately as the plaster is etched, and more so at first, but then would slow down each day after that due to the rising CSI. If we assume that the average CSI was -2.0, running the math seems to indicate that the etching is logarithmic with the CSI. How do you see these results?
 

mas985

TFP Expert
LifeTime Supporter
May 3, 2007
12,800
Pleasanton, CA
I think the relationship you are showing is actually exponential and not logarithmic but these are inverse of each other so let me explain my reasoning with some definitions:

X = LOG10(Y) - Logarithmic relationship

inverse equation:

Y = 10^X - Exponential relationship

Logarithm - Wikipedia

Next:

PH = -LOG10[H+] where H+ is the hydrogen ion concentration

CSI is relate to PH by:

CSI = PH (measured) - PH (calcium fully saturated)

So when PH is less than the fully saturated PH, CSI is negative and the plaster may experience some etching or visa versa.

So by extension:

CSI = -LOG10[H+ (measured)] - -LOG10[H+ (calcium fully saturated)] = -LOG10(H+ (measured)/H+ (calcium fully saturated))

The inverse of this becomes:

H+ (measured)/H+ (calcium fully saturated) = 10^-CSI

The excess H+ ions determine how much etching will occur as confirmed by this paper: http://www.poolhelp.com/wp-content/uploads/2017/05/JSPSI_V4N1_pp54-59.pdf

So while I do agree that PH & CSI are logarithmic with respect to H+ concentration, the inverse makes H+ concentration exponential with respect to CSI. So the etching/scaling rates should be related to the CSI by an exponential. This is the 10x factor you were referencing.

Also it is difficult to gauge if 40 ppm etching is all that significant. For your first experiment, what was the size of the water vessel, the surface area of the plaster and the plaster density? It could actually be a very small percentage of the total surface.
 

onBalance

TFP Expert
LifeTime Supporter
In The Industry
Jul 25, 2011
976
Utah
That is complicated stuff, Mark.
I placed 40 square inches of plaster into 4.5 gallons of water. Essentially, a 4 inch by 4 inch square by 3/4 inch thick plaster coupon.
I do not know the density question.
 

mas985

TFP Expert
LifeTime Supporter
May 3, 2007
12,800
Pleasanton, CA
Do you know how much the coupon weighed? Or did you weigh the coupon before and after the experiment?

I placed 40 square inches of plaster into 4.5 gallons of water. Essentially, a 4 inch by 4 inch square by 3/4 inch thick plaster coupon.
Also, 4x4 is 16 square inches, not 40 square inches. Was it 40 or 16?
 

JamesW

TFP Expert
Mar 2, 2011
16,598
CaCO3 <=> Ca2+ +CO32-

The above reaction is reversible. Solid calcium carbonate dissolves and calcium and carbonate ions combine and precipitate. When the water is saturated, dissolution equals precipitation.

The saturation of calcium carbonate is when the product of the molar concentrations of calcium (Ca2+) and carbonate (CO32-) are equal to the solubility product constant Ksp.

Ksp = [Ca2+][CO32-]

For example, at a pH of 7.8 and CH = 105 and TA = 105, the molar concentration of calcium x the molar concentration of carbonate is equal to the Ksp.

The molar concentration of carbonate is determined by the molar concentration of bicarbonate and by the pH because a percentage of the bicarbonate will become carbonate depending on the pH.

pH……….% CO32-

8.0……….0.5
7.9……….0.4
7.8…….….0.315
7.7……....0.25
7.6…….…0.2
7.5……...0.16
7.4………0.126
7.3…..….0.1
7.2……….0.08
7.1…….…0.06
7.0….....0.05

The amount of carbonate at a pH of 8.0 is 10 x the amount of carbonate at a pH of 7.0. So, a pH change of 1.0 point will make a difference of 10 times the amount of carbonate. So, I think that it indicates that the loss rate will be 10 x as much at a pH of 1 point below a balanced CSI.

The kinetics of dissolution and precipitation of calcium carbonate are complex and are dependent on many factors, such as pH and water temperature.
 

JamesW

TFP Expert
Mar 2, 2011
16,598
Do you know how much the coupon weighed? Or did you weigh the coupon before and after the experiment?

Also, 4x4 is 16 square inches, not 40 square inches. Was it 40 or 16?
Total surface area = 44 square inches counting front, back and 4 sides.
 

onBalance

TFP Expert
LifeTime Supporter
In The Industry
Jul 25, 2011
976
Utah
James is correct about the surface area.
I know the approximate weight of the coupon, but did not determine the precise weight. Not nearly accurate enough to be helpful for calculations.