How to Work Out Whether Blowdown Is Set Right

Why this matters

Almost nobody checks blowdown. The controller has a setpoint, the setpoint came from somebody, and the loop runs on it for years. The check is a one-page calculation against one timed measurement, and when the two disagree, the disagreement is almost never in the setpoint. It is in the valve, the probe or the discharge point, and each of those fails in a direction that leaves the panel reporting a healthy loop while the water quietly concentrates past the ceiling its own chemistry sets.

Before you take a single measurement

• Blowdown discharge is hot water under system pressure. Confirm the line is piped and restrained before you crack a manual valve, stand clear of the outlet, and check the discharge temperature limit your adopted plumbing code sets for a sanitary sewer connection with the authority having jurisdiction, since that number is local.
• If any part of this takes you inside the tower, lock and tag the fan motor disconnect under 29 CFR 1910.147 and confirm the fan has stopped by eye before crossing the plane of the access door. Panel work is a different standard: 29 CFR 1910.333(b)(2), with live-dead-live per NFPA 70E-2021, 120.5.
• Pulling a conductivity probe means opening a pressurised, hot line. Isolate it, relieve pressure to a vented point and verify at a gauge that pressure is actually zero before you break the fitting.
• Cleaning a fouled probe usually means an acid. Take the glove class and any respiratory control from that product's safety data sheet, keep it physically away from the hypochlorite or bromine feed line, and dilute acid into water rather than water into acid.
• Do not dry-brush a silted basin or pressure-wash fouled fill while you are gathering numbers. That converts a contact exposure into an inhalation one.

The one-page check, filled in for one tower

Each line below is a field, what it comes from, and the value this tower produced. Fill yours the same way and the arithmetic does the rest.

A. Circulating flow: 900 gpm. From the design documents, confirmed against the pump curve at the measured head. A design figure that has never been verified is the most common bad input on this page, because a worn impeller or a throttled balancing valve moves it without anyone knowing.

B. Range: 12 F. Measured, supply minus return at the tower, at steady load. Take it at the same time of day you take everything else on this page.

C. Evaporation: 10.8 gpm. The standard approximation is that about 1 percent of circulating flow evaporates per 10 F of range, and it is derived assuming essentially all of the heat rejection is latent, so on a cold, dry day when a larger share leaves as sensible heat the real evaporation is less than this line gives. Here: 900 times 1.2 times 0.01, which is 10.8 gpm.

D. Limiting species and its ceiling: silica, 6.0 cycles. From the makeup water report and the treatment supplier for this water and this equipment. Record the two numbers it came from, not just the ratio: makeup silica in mg/L from the water report, and the supplier's silica limit in mg/L for this loop's operating pH and temperature. Silica solubility rises with temperature and rises sharply above about pH 8.5, so a ceiling quoted for one condition does not transfer to another, and a loop that runs hotter in summer does not have the ceiling it had in spring. This is the field that cannot be guessed and cannot be copied from another site.

E. Target cycles: 6.0. Set at the ceiling here because the loop is stable and monitored. On an unmonitored loop, set it below the ceiling to leave margin for a makeup source that changes.

F. Required liquid loss: 2.16 gpm. Evaporation divided by cycles minus one: 10.8 divided by 5.

G. Drift: 0.045 gpm. From the tower's own data sheet, which quoted drift as 0.005 percent of circulating flow: 900 times 0.00005. Take this from the equipment, never from a general figure, because eliminator design has changed a great deal and an old tower and a current one are not close.

H. Other liquid losses: 0 gpm. Filter backwash, a side stream, a process bleed. This tower has none. If yours does, it counts against required loss the same way drift does.

I. Required intentional blowdown: 2.1 gpm. F minus G minus H, which is 2.115, called 2.1.

J. Controller setpoint: 1,800 microsiemens. Makeup conductivity, measured at 300, times target cycles of 6.0. Set the deadband from the controller's own manual rather than inventing one; a deadband too tight cycles the valve constantly and a deadband too wide swings the loop either side of the ceiling.

K. Measured blowdown: 1.3 gpm. Use an installed meter if one exists; that is the preferred method and it removes the exposure below. Where you must time it, hose the discharge into a receptor and time the fill of a graduated vessel inside the receptor rather than holding a container in the stream, keep your face out of the plume, and treat the discharge as biologically active tower water rather than as hot water only. Three runs, averaged, taken while the controller was calling for blowdown.

Field K against field I is the whole test.

Reading the disagreement

Here it is 1.3 gpm measured against 2.1 gpm required, and the gap is worth converting into the unit that matters: cycles.

With 1.3 gpm of intentional blowdown plus 0.045 gpm of drift, total liquid loss is 1.345 gpm. Cycles is evaporation divided by liquid loss, plus one: 10.8 divided by 1.345 is 8.03, plus one is 9.03. This loop is running at about 9 cycles against a silica ceiling of 6.0.

Confirm before acting, because a single method is not a finding. System conductivity measured 2,650 against makeup at 300, a ratio of 8.83. Two independent methods, one from a water balance and one from a tracer, landing at 9.03 and 8.83. They agree, so the loop really is over-concentrated and the controller really is not reporting it.

Now find why. The controller setpoint was 1,800 and the loop was sitting at 2,650, so the controller was not calling for blowdown at a conductivity well above its own setpoint. That points at the probe, and the probe was coated. A coated conductivity probe reads LOW, because the film insulates the electrodes, so the controller believes the water is less concentrated than it is and under-blows. The loop concentrates, deposition accelerates, the coating thickens, and the error grows in the same direction. It is a runaway, not a drift.

Name the other end of that mechanism so the direction is established rather than asserted: a probe reading high, whether from a poor temperature compensation setting or from a location that sees a pocket of concentrated water rather than the bulk loop, produces the opposite failure. The controller over-blows, cycles run below target, makeup and treatment chemical consumption both rise, and nothing looks wrong because water quality tests clean. Under-blowing scales and fouls equipment; over-blowing wastes water and chemical and, held long enough, dilutes the inhibitor below its working residual while continuously replacing cycled water with fresh aerated makeup, so it buys a corrosion problem in place of a scaling one. Neither is the safe direction. Both are the same instrument being trusted.

What the correction is worth, and where it stops being worth much

Run the makeup arithmetic across three targets on this tower, using the same 10.8 gpm evaporation.

• At 3 cycles: liquid loss is 10.8 divided by 2, or 5.4 gpm, and makeup is 16.2 gpm.
• At 6 cycles: liquid loss is 2.16 gpm and makeup is 12.96 gpm.
• At 12 cycles: liquid loss is 10.8 divided by 11, or 0.98 gpm, and makeup is 11.78 gpm.

Going from 3 cycles to 6 cuts makeup from 16.2 to 12.96 gpm, a 20 percent reduction. Doubling again from 6 to 12 cuts it from 12.96 to 11.78, which is 9.1 percent. The first doubling buys more than twice what the second one does, and the second doubling is the one that pushes the water past its scaling and corrosion limits. That is the case for setting cycles at the limiting species rather than as high as the controller will allow: the water savings flatten out long before the chemistry does.

The drift line behaves the same way and it is the reason field G exists. At 6 cycles, drift at 0.045 gpm is 2.1 percent of the 2.16 gpm liquid loss, so ignoring it changes the answer by about two percent and nobody would notice. At 20 cycles, liquid loss is 10.8 divided by 19, or 0.57 gpm, and the same 0.045 gpm of drift is 7.9 percent of it. A shop that learned to ignore drift on a moderately cycled tower carries that habit onto a high-cycle loop where it is a real error, and it errs in the direction of over-blowing.

When the page says the setpoint is right and the loop still fails

Three causes that this calculation cannot see, all of which have produced the same complaint.

The blowdown is taken from the wrong point. A bleed line teed off the basin, or off a line upstream of the tower where the water has just been diluted by makeup, discharges water that is less concentrated than the loop. The volume is correct and the mass of dissolved solids removed is not, so the loop concentrates on a correct-looking blowdown rate. Take blowdown from the recirculating line downstream of the pump, before the tower, where the water is at loop concentration.

The valve does not pass what the calculation assumed. A solenoid partially blocked with debris, or a manual bleed valve someone throttled to stop a noise, both hold the loop above target while the controller's log shows the correct number of open minutes. That is why field K is a timed measurement and not the controller's own report of itself.

Makeup changed and nobody re-derived field J. The setpoint is makeup conductivity times target cycles, so if the utility switches source or a softener upstream goes into bypass, the same setpoint produces different cycles the next morning. Re-measure makeup conductivity quarterly and any time the loop's behaviour changes for no visible reason, and treat a moved makeup number as a reason to redo this page rather than as a curiosity.

References

• Tower manufacturer data for design flow, range and drift rate
• Water treatment supplier documentation and the makeup water quality report for the limiting species and its ceiling
• Controller manufacturer manual for setpoint, deadband and probe maintenance intervals
• 29 CFR 1910.147 for fan motor isolation; 29 CFR 1910.333(b)(2) for electrical work; NFPA 70E-2021, 120.5 for the proving sequence
• See related: Cycles of Concentration in Plain Terms; What an Evaporative Tower Does to the Water in It