The New Plaster That Etched on Soft Fill Water

Why this matters

A replaster is the biggest single ticket most pool customers ever pay, and the first month decides how it wears for the next fifteen years. When a new finish turns rough and dull in weeks, the first question anyone asks is whether the plasterer or the service company owns it, and the answer is usually sitting in the startup log. This case is reconstructed from one: the readings were taken, written down and kept, every pH and chlorine number was in range, and the water was dissolving the surface the whole time. Nobody had done the one piece of arithmetic that would have shown it.

The complaint, five weeks in

A 15,000 gallon residential pool, replastered with a standard white marble-aggregate finish in late spring. The plastering contractor filled it from the house hose bib and handed startup to the service shop on the day the fill finished, with the shop's startup procedure to follow, per 'Pool Plaster Startup Brushing Protocol'. The shop's tech ran the startup chemistry and brushing for four weeks and then put the pool on the weekly route.

At week five the homeowner called: the floor felt like sandpaper on bare feet, the walls had lost their sheen, and a shallow-end step had gone gray and blotchy where the kids sat. Run a hand across the floor and the marble grains stood proud of the cement around them. That is etching: cement paste dissolved away from the surface, leaving aggregate exposed.

Starting from the record

The tech's first instinct was the plaster: a wet mix, late troweling or a calcium chloride accelerator in the mix all show up as finish defects in the first weeks. Before anyone called the plasterer, the service manager pulled the startup log and read it end to end. It was, by the usual standard, a good log: dated, initialed, every entry with pH, total alkalinity, calcium hardness and chlorine. The temperature column was empty, and there was no column for the saturation index at all. Water temperature was taken from the homeowner's floating thermometer for one visit, which read 76 F, and the manager used that for every row.

STARTUP LOG, 15,000 gal replaster    water ~76 F
LSI added later: pH + 0.6 + CF + AF - 12.1
-------------------------------------------------
day       pH    TA   CH   CYA  FC   LSI (later)
0 fill    7.6   50   70    0   0    -0.76
7         7.4   50   85    0   2    -0.87
14        7.5   45   95   30   3    -0.88
28        7.5   40  100   30   3    -0.92
-------------------------------------------------
TA, CH, CYA in ppm.  CF = log CH - 0.4
AF = log (TA - CYA/3)
day 0 note: "fill water soft, CH 70"
no calcium chloride add recorded

Working out what the water was doing

The manager computed the index for every row the way 'How to Calculate the Saturation Index on the Pool Deck' lays it out: pH, plus the temperature factor (0.6 at 76 F), plus the calcium factor (log of hardness minus 0.4), plus the alkalinity factor (log of carbonate alkalinity, which is TA minus a third of the CYA), minus 12.1.

Day 0, the fill water. 7.6 + 0.6 + 1.45 + 1.70 = 11.34, minus 12.1 is -0.76. Calcium at 70 ppm and alkalinity at 50 ppm is soft water by any measure, and the note in the log said so. The startup protocol calls for calcium chloride at half-fill whenever the fill tests below 150 ppm hardness, brought up toward 200 to 300 ppm for the final volume. The log shows no add.

Day 7. 7.4 + 0.6 + 1.53 + 1.70 = 11.23, about -0.87. Calcium had risen from 70 to 85 on its own, which is the new plaster leaching calcium hydroxide into the water. That rise is the surface giving itself up, and in the log it reads like good news.

Day 14. Stabilizer went in, and the carbonate alkalinity fell to 45 minus 10, or 35 ppm. 7.5 + 0.6 + 1.58 + 1.54 = 11.22, about -0.88.

Day 28. TA was down to 40 after four weeks of daily acid to hold pH at the low end, which the plaster was pushing up. Carbonate alkalinity 40 minus 10, or 30 ppm. 7.5 + 0.6 + 1.60 + 1.48 = 11.18, about -0.92.

Every row sat well below the -0.3 floor of the band used in 'Calcium Hardness and Scale Management', from 0.46 below it at the fill to 0.62 below it at day 28, and it got worse each week rather than better. A bicarb-style startup deliberately keeps the water slightly hungry so plaster dust dissolves rather than setting on the surface, but no startup method leaves new plaster near -0.9 for a month.

Why the readings looked fine

The tech was not careless. He read every number against its own range and acted on every one that was out:

• pH read high most mornings (fresh plaster drives it up hard), and he brought it down to 7.4 or 7.5 with acid each visit, exactly as the protocol says. Each of those doses also knocked total alkalinity down, which is why TA drifted from 50 to 40.
• Chlorine sat at 2 to 3 ppm once it was allowed in.
• Calcium was below the range in 'Pool Chemistry Quick Reference' on every row, and he noted it, but it was rising every week without his help, so it looked like the pool was correcting itself.

Three range checks, three plausible stories. The saturation index puts the four numbers into one, and that one said the water had been taking calcium carbonate out of the cement every day since the fill. The rising calcium was not the pool correcting itself; it was the evidence.

Ruling out the plaster

The plasterer still deserved a fair look, because a defective mix etches too. Three things pointed away from the finish:

• The pattern was uniform. Workmanship faults (a wet mix, late troweling, a cold joint) tend to show in patches, trowel arcs or the area finished last. This surface had lost paste evenly on the floor and walls, worst on the step where bathers sat and water stays agitated.
• The timing matched the chemistry. The damage built steadily from the first weeks, while the log shows corrosive water from day 0.
• The fill water explained it without any help. Soft, low-alkalinity water on fresh cement is the textbook cause, and the log recorded the soft water in writing on day 0.

The manager did not accept the conclusion on his own say-so: he sent the log, with the index column added, to the plasterer and asked whether anything in the mix or the finishing would change the picture. The plasterer pointed to the same day 0 note.

What a computed index would have done

Run on the fill water at day 0, the index says -0.76 before the first brush stroke. The fix is two adds, sized from 'The Dose Per 10,000 Gallons Factor Card':

• Calcium hardness 70 to 250 ppm. 180 x (15,000 / 10,000) x 0.12 lb per ppm = 180 x 1.5 x 0.12 = 32.4 lb of the 77 percent product, added at half-fill as the startup protocol directs, method per 'How to Raise Calcium Hardness With Calcium Chloride'.
• Total alkalinity 50 to 90 ppm. 40 x 1.5 x 0.14 = 8.4 lb of sodium bicarbonate, method per 'How to Raise Total Alkalinity With Sodium Bicarbonate'.

At pH 7.5, 76 F, no stabilizer yet, that water reads 7.5 + 0.6 + 2.00 + 1.95 = 12.05, about -0.05. Inside the band, and still slightly on the dissolving side, which suits a startup that wants dust dissolved rather than scaled.

The weekly index mattered as much as the fill index. Even with the adds, daily acid through startup walks TA down; computing the index each week shows when the walk-down has taken it back below -0.3 and a bicarbonate add is due. The day 28 row would have flagged -0.92 in time to stop another month on the route at the same balance.

The temperature nobody recorded

The 76 F used for every row came from one reading on a floating thermometer near the surface on a sunny afternoon, and the replaster went in during late spring, when nights were still cool. That one assumption flatters the log. At 60 F the temperature factor is 0.4 rather than 0.6, so each row drops by 0.2: the day 28 water reads about -1.12 instead of -0.92. The corrected water from the section above holds up better, landing near -0.25 at 60 F, still inside the band. So the missing temperature column did not hide a borderline case; it hid how much worse a cold week was than the log already shows, and it is the one input no later reconstruction can recover. A tech who reads the thermometer at elbow depth on every visit gives the next person a log that can be worked backwards without guessing.

What the shop changed

• The startup log grew two columns: water temperature, measured at elbow depth on every visit with the kit's thermometer, and the saturation index, computed on site. A row with an index below -0.3 is not signed off until the correction is planned and written beside it.
• The fill water gets an index before the fill is half done. The half-fill calcium add in the startup protocol is now triggered by the index as well as by the 150 ppm hardness line, because soft alkalinity can take a pool negative with hardness right at 150.
• Weekly route pools get the index once a month and on every season change, because the temperature factor drops as water cools and a pool that balanced in August can be corrosive in October.

What the customer heard, and what was left

The etched surface does not come back. The service manager told the homeowner that plainly: the chemistry the shop held during startup dissolved the top of the finish, the shop had stopped it by bringing calcium and alkalinity up that week, and the roughness that was there would stay unless the surface was treated. A light acid wash or a surface polish can take the proud aggregate down on a rough finish; whether it suits this one is a decision for the plasterer and the procedure in 'Drain and Acid Wash Standard', not a route call. What the shop did own was the next fifteen years of water: an index logged every month, and the homeowner shown the number.

The log that convicted the shop was also what let it settle the question in an afternoon instead of an argument. A log with no hardness column could not have been read backwards at all.

References

• See related: Pool Plaster Startup Brushing Protocol (startup sequence, half-fill calcium add)
• See related: How to Calculate the Saturation Index on the Pool Deck (factors and arithmetic)
• See related: Calcium Hardness and Scale Management (the -0.3 to +0.3 band, etching background)
• See related: The Dose Per 10,000 Gallons Factor Card; Drain and Acid Wash Standard