What Is Actually in the Water You Put Into a System

Why this matters

Water is the only material we install without a submittal. A shop that would never put an unknown alloy in a pump will fill a loop from whatever hose bib is closest and treat the result as an inert working fluid. It is not inert. Tap water is a dilute solution carrying dissolved minerals, dissolved gases, suspended particles and living organisms into the system, and when the water leaves as vapor, most of that cargo stays behind.

That is the whole idea. Almost every water-side failure that arrives looking sudden - a heat exchanger that plugged in a week, a heater that quit making hot water, a tower fill that collapsed under its own weight - is cargo that accumulated on a slow clock and crossed a threshold.

Before you open it or add anything to it

Opening or running a device that holds standing water releases an inhalable aerosol. A tower basin, a humidifier sump, a long-idle branch: the route is your lungs, so a glove does nothing for it. Where the work will generate aerosol from a system that has been standing, respiratory protection has to come from a written program with fit testing (29 CFR 1910.134; the fit-test interval at 1910.134(f)(2) is at least annually), and in a building with a water management program the disinfection-before-opening question belongs to that program (ASHRAE Standard 188).

Never dose one treatment chemical into a sump, drum or drain that holds another. A hypochlorite (bleach-type) product meeting an acid releases chlorine gas within seconds; the same product meeting an ammonia-bearing cleaner releases chloramine. Both are inhalation hazards in a mechanical room, so the control is ventilation and the respiratory protection named in Section 8 of the product's safety data sheet, plus the specific glove class that section names. Your employer has to make that sheet available to you (29 CFR 1910.1200). Do not pick a glove by feel.

Before cracking a sample point or a drain on a heated or pressurized system, isolate it, relieve the pressure to a gauge reading zero, and let it cool or route it through a sample cooler; hot water flashing out of a fitting scalds through clothing.

Any hose from a drum, test rig or chemical feeder to a potable line is a cross-connection. Treat it as a public-health event, not a plumbing detail: the backflow protection the local plumbing code requires for that chemical goes in before the connection is made, not after.

Four things ride in with the water

That last column is the useful split. Three of the four categories can only be added by putting more water in. The fourth can multiply from a trace, which is why a system that took almost no makeup can still have a biological problem, and why the fix for that one is never "use better water."

Only the water leaves

Evaporation removes H2O molecules and leaves everything dissolved in them. That single sentence separates every open evaporative system, which concentrates, from every closed system, which does not (a sibling article works that distinction out in full; see References). It also explains why a boiler, a humidifier, a tower and a steam table all fail the same way while nothing on the water report changed.

Hold on to the direction, because it runs opposite for two of the four categories. Dissolved solids concentrate as water leaves and deposit fastest where the surface is hottest. Dissolved gases do the reverse: they come out of solution as the water heats, so they are at their maximum where the water is coldest and freshest. Name both ends before you use either. At the cold makeup connection the water carries its highest dissolved oxygen and has essentially no deposition tendency. At the hottest heat-transfer surface it carries almost no dissolved oxygen and its highest deposition tendency. A tech who knows only the first half looks for scale at the fill valve and oxygen corrosion at the boiler tube, and finds neither.

Worked example: what a heating season actually delivered

A steam humidifier serving an office suite boils water to vapor and vents that vapor into the duct. Nothing recirculates. Take the building's supply at 300 mg/L total dissolved solids, of which 150 mg/L as CaCO3 is hardness (the "as CaCO3" convention is a reporting basis, not a statement that 150 mg of chalk is floating in the liter; a sibling article on reading the report covers the conversions).

The unit evaporates 4 liters per hour and runs 10 hours a day: 40 liters a day.

• Solids delivered per day: 40 L x 300 mg/L = 12,000 mg, or 12 grams a day.
• Over a 120-day heating season: 120 x 40 L = 4,800 liters, carrying 4,800 x 300 mg = 1,440,000 mg, or about 1.44 kilograms of dissolved solids.

Every gram of that entered as clear water and none of it left as vapor. Where it went is decided by which fraction is still soluble at the evaporating surface. The hardness fraction, half the load here, is the part with a solubility that falls as temperature rises, so it plates onto the hottest metal in the machine. The sodium and chloride fraction stays dissolved in the shrinking pool of remaining liquid and only leaves through a drain or a blowdown, which is exactly what the drain cycle on that machine exists to do. Defeat the drain cycle and that fraction crystallizes too, on the same surfaces, at the end.

What flips this answer: run the same unit on softened water and the mass delivered barely changes, because ion exchange trades calcium for sodium without removing much dissolved mass. What changes is the destination - sodium salts stay in the pool and leave through the drain instead of plating on the element. Run it on reverse-osmosis permeate at, say, a tenth the dissolved solids and the delivered mass drops roughly tenfold, which is why deionized feed is standard on humidifiers and why nobody bothers softening one.

The failure mode: a tech descales the element, hands back a clean machine, and changes nothing about feedwater or the drain cycle. The unit is on the same 12 grams a day it was on before, so the next teardown lands on the same calendar interval, and the customer learns that descaling is something this machine just needs. It was never a machine problem.

What changes the inventory

• Source. Groundwater is usually higher in dissolved minerals and lower in organics, and can carry dissolved iron, manganese, or hydrogen sulfide. Surface water is usually lower in minerals but higher in organics and suspended solids, and it swings seasonally with runoff. A utility that blends the two changes its own product through the year, and the annual consumer confidence report is an average, not the water arriving at that building today.
• Distance from the plant. Water picks up whatever the main and the building piping give it. A first-draw sample after a long stagnation is a different water from a flushed sample at the same tap, which is the entire reason the tap sampling protocol for lead and copper specifies a minimum six-hour stagnation before a first-draw sample (40 CFR Part 141, Subpart I).
• Anything installed upstream. A softener zeroes hardness and leaves alkalinity and pH untouched. A calcite neutralizer raises pH and alkalinity by dissolving calcium carbonate, which means it also raises hardness. Reverse osmosis strips nearly everything, including the mineral content that was holding a protective film in place.

How to verify you have this right

Sample the water actually entering the equipment, at the fill or makeup connection, and not the meter, the kitchen tap, or the utility's published average. Get the report to name a sample point and a date, because a number without those is not evidence.

Then check the two things a water report cannot tell you. Weigh or photograph the deposit. Its color, hardness and how it reacts to a drop of dilute acid says which fraction actually landed; a carbonate deposit fizzes, a sulfate or silica deposit largely does not. And find out how much water the system has swallowed, because the report gives you a concentration and the delivered mass is that concentration multiplied by volume. A separate article covers how to measure that when there is no meter.

References

• 29 CFR 1910.1200 (Hazard Communication) for the employer duty to make the safety data sheet available, and Section 8 of that sheet for the specific glove class and respiratory control for each product
• 29 CFR 1910.134, including the fit-test interval at 1910.134(f)(2), for respiratory protection where work generates an aerosol from standing water
• ASHRAE Standard 188, Legionellosis: Risk Management for Building Water Systems, for whether a standing system is disinfected before it is opened
• 40 CFR Part 141, Subpart I, for the six-hour minimum stagnation before a first-draw tap sample
• See related: The Difference Between an Open Loop and a Closed Loop Chemically; How to Read a Water Test Report; How Scale Forms and What It Actually Costs; The Water Chemistry That Attacks a System