By SubcontractorHub Editorial Team·Published September 2026

Short version: evaporated pellets are the default — highest purity, least residue, lowest bridging risk. Solar crystals are fine and cheaper on municipal water in a dry space. Rock salt saves a couple of dollars a bag and slowly fouls the brine valve. Potassium chloride costs three to four times as much, is about 10% less efficient, and is only worth it for a medical sodium restriction or a septic reason — and it needs the valve salt dose raised by roughly 10% when you switch.
Every bag in the softener aisle is sodium chloride except one, and they all soften water identically. What separates them is purity — what else is in the bag besides salt, and where that residue ends up.
The salt never enters your water supply as salt. It dissolves into brine, the brine strips hardness minerals off the resin bed, and the whole lot goes down the drain. What stays behind is whatever would not dissolve, and that is what accumulates in the brine well until the float or the valve stops working properly.
Prices are indicative US retail as of 2026 and vary considerably by region and bag count.
The gap between 99.8% and 95% sounds trivial. Run the arithmetic on a year of salt and it is not.
A household using 10 bags (400 lb) a year:
Evaporated pellets at 99.8% → about 0.8 lb of insoluble residue
Rock salt at 96% → about 16 lb of insoluble residue
Sixteen pounds of fine mineral sediment settling into the bottom of a brine tank every year is what eventually clogs the brine well screen and jams the float. That is a service call and a tank clean-out, and it costs more than the few dollars a bag saved.
Potassium chloride is marketed as the healthier, greener salt. It does not soften water any better, and it has genuine trade-offs that get left out of that pitch.
The two real reasons to use it:
The costs, stated plainly:
If you switch, change the setting. Dropping potassium chloride into a valve still programmed for sodium chloride leaves the softener under-dosed by about 10%, so it runs out of capacity early and the customer reports hard water on the most expensive salt on the shelf. Run the tank down first, then raise the salt dose.
This is the question behind most potassium chloride purchases, and it deserves a number rather than a reassurance.
Softening adds roughly 7.5 mg of sodium per litre for each grain per gallon of hardness removed. On 10 gpg water that is about 75 mg/L. Two litres a day is around 150 mg — roughly one slice of bread, or slightly more than half a cup of milk.
On 25 gpg water it is about 190 mg/L, which is more meaningful. Even then, the usual professional answer is not potassium chloride: it is to leave the kitchen cold tap unsoftened and, where drinking water quality is also a concern, add a point-of-use RO unit there. That gives unsoftened, low-sodium drinking water at a fraction of the running cost, and the comparison is set out in water softener vs reverse osmosis.
Two failure modes happen in the brine tank itself, and both present as “my water is hard but the tank is full of salt.”
A salt bridge is a hard crust spanning the tank with an empty void beneath it. The water below never touches salt, so no brine is made and regeneration does nothing. Push a broom handle straight down: if it hits a hard layer and then drops through into a gap, that is a bridge. Break it up with the handle, not with anything metal that could puncture the tank.
Mushing is the opposite — salt recrystallising into a sludge at the bottom that blocks the brine well. It usually needs the tank emptied and cleaned.
Both are made worse by humidity, by overfilling, and by low-purity salt. Keep the tank no more than two-thirds full, let it run down close to empty every few months rather than topping up indefinitely, and use higher purity salt in a damp basement. Full detail is in the maintenance schedule, and if the water is already hard the troubleshooting guide works through the causes in order.
Salt use is set by hardness, household size and the valve's salt dose setting, not by which bag you buy. A family of four on 15 gpg water typically gets through 8 to 10 bags a year. The regeneration calculator works out salt per cycle and per year from your own figures, and it also shows the efficiency trade behind the dose setting — running at 6–8 lb per cubic foot instead of the 15 lb nameplate dose cuts salt use substantially for a modest reduction in capacity.
Salt is the lowest-margin thing a water treatment business sells and one of the most valuable. It is the only reason to be in a customer's basement several times a year, and the dealer carrying the bags is the dealer who spots the bridged tank, the leaking valve, the resin that is past it and the softener that is now eight years old.
Dealers who treat salt as a scheduled recurring delivery rather than a call-when-you-run-out errand see the replacement sale coming and get the service work that goes with it. Dealers who do not, lose the account to whoever does. Keeping delivery intervals, service visits and equipment age on one schedule across a whole book of customers is exactly what water treatment business software is for, and when that eight-year-old softener finally needs replacing, financing beside the cash price is what closes it on the visit.
Evaporated salt pellets are the best general choice. At around 99.8% sodium chloride they are the purest common option, which means the least insoluble residue building up in the brine tank and the lowest risk of a salt bridge or mushing. Solar salt crystals are cheaper at roughly 99.6% purity and work well on moderate water use, but they dissolve less readily and bridge more often in humid basements. Rock salt is the cheapest and the worst choice for most installations, because its insoluble mineral content leaves sediment that eventually fouls the brine valve. On municipal water with average hardness, solar crystals are fine; on hard well water with a high regeneration frequency, pellets are worth the difference.
It is not better at softening, and it costs roughly three to four times as much. Potassium chloride exists for two specific reasons: it adds no sodium to the water, which matters for people on a medically restricted sodium intake, and the discharge is potassium rather than sodium, which some septic and irrigation situations prefer. The trade-off is real. Potassium chloride is about 10% less efficient at regenerating the resin, so the valve's salt dose setting must be raised by roughly 10% to get the same capacity, and it is more prone to bridging because it absorbs moisture more readily. Use it when there is a specific reason, not as a general upgrade.
Roughly 7.5 mg of sodium per litre for every grain per gallon of hardness removed, which means about 75 mg per litre on 10 gpg water. To put that in context, a slice of bread contains around 150 mg and a cup of milk about 120 mg, so two litres of softened water at that hardness adds roughly the sodium of one slice of bread. For most people this is not a meaningful contribution. For someone on a strictly sodium-restricted diet on very hard water it can be, and the two usual answers are a potassium chloride softener or, more commonly, leaving the kitchen cold tap unsoftened and adding a reverse osmosis unit there.
A salt bridge is a hard crust of salt that forms across the brine tank with an empty void underneath it. The tank looks full from above, but the water below never contacts the salt, so no brine is made and regeneration accomplishes nothing. The water goes hard while the homeowner is certain the tank has plenty of salt in it. Humidity, overfilling the tank, low-purity salt and potassium chloride all raise the risk. To check, push a broom handle straight down through the salt: if it hits a hard layer and then breaks through into a gap, that is a bridge. Break it up carefully with the handle rather than anything metal, and keep the tank no more than two-thirds full.
You can mix pellets and crystals of sodium chloride without harm, though there is little reason to. Do not mix sodium chloride and potassium chloride, because the valve's salt dose setting has to be calibrated for one or the other and a mixture leaves the softener set wrong for whatever is actually in the tank. If you are switching between the two, run the tank down as close to empty as possible first, then change the valve setting at the same time. Mixing rock salt with anything else is also a poor idea, since its insoluble residue ends up in the brine well regardless of what it is mixed with.
You need salt sold for water softeners, but not a particular brand. Never use table salt, ice melt or road salt. Table salt contains anti-caking agents and iodine that leave residue in the tank, and ice melt products frequently contain calcium chloride, magnesium chloride and dyes that will foul the resin bed you are paying to protect. Rust-remover or iron-fighting salts contain a citric acid additive that helps on water with low levels of iron, but they are not a substitute for a dedicated iron filter above about 0.3 ppm. Any bag labelled water softener salt from a hardware or grocery store is suitable.
30-minute demo. See how water treatment dealers keep salt deliveries, filter changes and equipment age on one schedule across the whole customer book — and quote the replacement with financing before someone else does.
Book a DemoWater Treatment, Water Softeners, Plumbing
SubcontractorHub
AI proposals, sales pipeline, and project management — all in one platform.
30 minutes. No commitment.