BlogApplication notes
Why is scale building up in my cooling tower and heat exchanger?
- Type
- Application notes
- Published
- 20 November 2026
- Prepared by
- Technical team
- Product group
- Heat Transfer & Cooling
Acid cleaning and a higher inhibitor dose are the usual response to scale, but if the conditions that made it remain, it comes back. What actually builds scale can be cycles run too high, hard and alkaline make-up, a hot wall with slow flow, an inhibitor shortfall or silica, which acid barely touches, and in summer several can stack up at once.
As summer goes on, chiller condensing pressure creeps up and the process outlet runs above its set point for longer. Open an exchanger at shutdown and a white film lines the tubes, and white rings have formed on the tower fill and round the holes in the distribution deck. Acid cleaning helps for a while, yet sometimes the next shutdown finds scale back in the same places.
Crystallisation fouling, where crystals build up on an exchanger surface, is regarded as the most harmful type of heat exchanger fouling (Berce et al., 2021). Scale conducts heat far worse than metal, so even a thin film cuts heat transfer, and the chiller or process runs hotter and uses more energy for the same work (Berce et al., 2021; Kazi, 2020). The space under scale invites local corrosion and microbial growth, so the US Centers for Disease Control and Prevention treats scale, corrosion and sediment control as core operating items for preventing Legionnaires' disease (CDC, n.d.).
An open cooling tower rejects heat by evaporating water. Only pure water leaves and the dissolved minerals stay, so the circulating water grows steadily more concentrated. Bleeding some of it off to dilute it is blowdown, and the number of times more concentrated the circulating water is than the make-up is the cycles of concentration, or simply cycles (DOE, n.d.; EPA, 2012). Less blowdown saves water but raises the scale risk; more blowdown cuts scale but costs water and chemicals, so scale control means finding that balance. Calcium carbonate is the most common cooling tower scale (EPA, 2026), and it usually appears first, and thicker, on the exchanger wall where the water gets hottest rather than in the tower.
In such cases the technical team sets up the five hypotheses below and rules them out one by one. The first measurements are the same every time: take the conductivity of the make-up and circulating water on the same day to find the actual cycles, calculate a saturation index from the circulating water's calcium hardness, alkalinity, pH and temperature, and set both against the targets of the treatment programme.
Hypothesis 1: cycles too high
When scale grows evenly across several exchangers and the tower, and circulating conductivity sits higher than usual, the cycles are the first suspect. It shows up where blowdown is opened and closed by hand, where the conductivity controller has gone long without calibration, and in summer as evaporation rises. A tower that goes into summer on a conductivity controller last calibrated in spring is one such case. The controller screen shows conductivity near the set point, a hand meter reads higher, and the log shows the blowdown valve opening only now and then.
As the cycles rise, calcium and alkalinity in the circulating water concentrate by the same factor; at three cycles they are about three times those in the make-up. There is a limit to how much calcium carbonate can stay dissolved, so as concentration rises the product of calcium and carbonate ions approaches that limit, and past it crystals form. The US Department of Energy's Federal Energy Management Program warns that if concentration gets too high the solids form scale inside the system, and notes that many systems run at two to four cycles, with six or more possible depending on make-up quality and treatment (DOE, n.d.).
Blowdown usually falls short of what was intended in one of three ways: the conductivity controller's sensor is fouled or out of calibration and reads low, the blowdown valve is blocked or stuck shut, or summer evaporation rises while the blowdown setting stays put. The US Environmental Protection Agency's WaterSense guidance likewise says to check the conductivity controller, make-up valve and blowdown valve first when a tower misses its target cycles (EPA, 2012).
Check the cycles with two numbers. One is the conductivity of the circulating or blowdown water divided by that of the make-up; the other is make-up flow divided by blowdown flow. The two should match, and if they differ widely, suspect a leak or a metering error (EPA, 2012). Measure conductivity with a hand meter daily, or at least weekly at the same hour, and compare it with the controller display to catch sensor error. FEMP advises logging make-up and blowdown volumes, conductivity and cycles and watching the trend (DOE, n.d.); circulating conductivity that stays above the set point, or a record of the blowdown valve staying shut for long periods, is the sign of this hypothesis.
The first fix is the sensor. Clean the conductivity sensor, calibrate it against a standard solution and recalculate the actual cycles; if they are above target, lower the blowdown set point and confirm the valve works. If blowdown is manual, automate it with a conductivity controller, which measures circulating conductivity continuously and bleeds only when the set point is reached (DOE, n.d.). Set the target cycles and the matching conductivity set point from the make-up analysis and the treatment programme together. Raising cycles from three to six is calculated to cut make-up by about 20% and blowdown by about 50% (DOE, n.d.), but only where water quality and treatment can hold the scale back.
Hypothesis 2: hard, alkaline make-up
If cycles and inhibitor are unchanged but scale grew after the water source changed or groundwater was blended in, turn to the make-up quality. Typical cases are plants using groundwater or hard mains water as make-up, areas where the source changes with the season, and sites that have sunk a new well for an extension. Picture a plant that starts blending groundwater into a tower that used to run on mains water alone. The cycle setting has not been touched, yet before long a white film shows in the exchanger tubes and white rings form round the holes in the distribution deck.
The raw materials of scale are calcium and carbonate ions. Hardness is the amount of calcium and magnesium dissolved in water; alkalinity is its capacity to absorb acid, carried in cooling water mostly by bicarbonate ions, HCO₃⁻. Calcium carbonate scale forms when calcium ions meet carbonate ions, CO₃²⁻, and the amount of carbonate depends on pH: as pH rises, bicarbonate turns into carbonate, so the same calcium scales more easily. EPA names calcium carbonate the most common cooling tower scale and lists adjusting pH below 8.6 and softening to remove calcium as treatments (EPA, 2026).
The tower raises pH by itself. When water meets air over a large surface, dissolved carbon dioxide escapes, the carbonic acid balance shifts towards carbonate and pH rises. A power plant cooling study used data from three pilot cooling towers to model this carbon dioxide stripping, and the model predicts loop pH from make-up and loop alkalinity and the operating conditions (Safari et al., 2014). With highly alkaline make-up, concentration and rising pH add up, so scale can form even at low cycles.
Write calcium hardness, total hardness, total alkalinity and pH for the make-up and circulating water side by side in one table. Hardness can be split into calcium and magnesium by EDTA titration (ASTM International, 2025), and alkalinity is measured by acid titration (ASTM International, 2016). If the make-up values have risen against past analyses, this hypothesis is the strong one. If calcium cycles in the circulating water run clearly below the conductivity cycles, the gap is calcium already dropping out as scale somewhere.
There are two lines of fix: treat the make-up, or lower the pH. If make-up hardness is what limits the cycles, consider softening the whole make-up or a side stream (EPA, 2012, 2026). If alkalinity is the issue, acid or carbon dioxide feed can lower the pH, but overfeed turns straight into corrosion, so pair it with automatic pH control and an alarm. Where make-up quality shifts with the season or the source, reanalyse the make-up every quarter and reset the target cycles.
Hypothesis 3: hot wall, slow flow
When only some of the exchangers scale, and thickest towards the outlet, the cause may be the temperature and velocity in those units. Cooling water flow cut at part load, a blocked strainer that has dropped the flow, process exchangers with high heat loads and layouts where only some tubes run slow all fit this pattern. Open a condenser that has been run at reduced flow for part load, and the inlet tubes are fairly clean while a hard white to off-white film thickens towards the outlet.
Unlike most salts, calcium carbonate dissolves less as it gets hotter. This property of calcium salts such as calcium carbonate and calcium sulfate is called inverse solubility, and when these calcium ions meet a hot heat transfer surface they precipitate right there (Kazi, 2020). EPA also describes calcium carbonate precipitating as water heats up (EPA, 2026). The hottest water in an exchanger is not the bulk but a thin layer against the metal, so the wall can be above saturation even when the circulating water as a whole is below it. The crystals on the wall are mostly calcite, with aragonite and vaterite under some conditions (Berce et al., 2021).
Velocity acts through both wall temperature and shear. Slow flow leaves water against the wall longer, so it gets hotter, and the shear, the sweeping force of the flow that strips crystals off, is weak. Fast flow cools the wall and washes deposits away. Fouling design guidance for heat exchangers names keeping tube-side velocity high as a basic way to reduce fouling (Nesta & Bennett, 2004).
The operating log shows this hypothesis first. For each exchanger, record the cooling water inlet and outlet temperatures and the approach temperature, the difference between process outlet and cooling water inlet. An approach that widens at the same load signals fouling of the heat transfer surface; on a chiller, check whether condensing pressure or temperature rises even after allowing for the season. Measure cooling water flow and exchanger pressure drop together to confirm velocity is not below design, and if at shutdown the tubes carry a harder, thicker white or off-white film at the outlet, the hottest end, give this hypothesis weight.
Restoring design flow comes first. Check the design flow for each exchanger and clear blocked strainers and half-closed valves. If flow is cut at part load, set a minimum velocity and do not go below it. Calculate the saturation index at the highest water temperature, at the exchanger outlet, not at the average loop temperature. Remove existing scale by mechanical cleaning or inhibited acid cleaning, and record whether the approach recovers afterwards so the next cleaning can be timed (Kazi, 2020).
Hypothesis 4: inhibitor shortfall
If water quality and flow are unchanged but scale jumps one day, open the inhibitor records. It is more likely in systems whose chemical pump often stops, where the dose is fixed instead of tied to make-up volume, and where cycles are pushed up to save water. Take a system whose chemical pump lost prime over a weekend when air filled its suction line. On Monday the tank level is exactly where it was on Friday, and the exchanger opened that week has a thin coat of fresh scale on its tubes.
A saturation index shows in one number whether water tends to dissolve calcium carbonate or to deposit it. The Langelier Saturation Index, the LSI, is the actual pH minus the saturation pH, pHs; above zero the water is supersaturated and tends to scale, below zero it tends to dissolve (Langelier, 1936). pHs is calculated from calcium hardness, alkalinity, total dissolved solids, the TDS or total of dissolved minerals, and temperature. The Ryznar Stability Index, the RSI, is 2×pHs − pH, and Ryznar held that values below 6 are clearly scale-forming (Ryznar, 1944). The LSI shows a tendency only and does not give the amount or rate of scale (Langelier, 1936), and the RSI, which Ryznar put forward to gauge the amount, is still read alongside actual results on site (Ryznar, 1944).
Much cooling water is run deliberately supersaturated to allow higher cycles, and a scale inhibitor holds that supersaturation in check. Inhibitors such as phosphonates, phosphorus-based organic chemicals, stop crystal growth at doses far below what would react with all the calcium. A molecular simulation study proposed a mechanism in which diphosphonates bind at growth steps on the calcite surface by taking the place of two carbonate ions, making it hard for the next calcium carbonate unit to attach (Nygren et al., 1998). EPA also describes phosphonates holding calcium carbonate in solution in the tower basin so it leaves with the blowdown (EPA, 2026). But there is a limit to the supersaturation an inhibitor can hold, so when the saturation index passes that line, or the pump stops and the dose falls, scale forms all at once.
Measure the LSI and the inhibitor as a pair on the same day. Calculate the circulating water's LSI weekly at the exchanger outlet temperature, and measure the inhibitor, or the tracer fed with it, that day too. There are three warning signs: a tank level that has not dropped or a pump that stopped, an inhibitor reading below target, and days when the LSI passed the upper limit set by the treatment programme. CDC advises checking scale and corrosion inhibitor levels as often as water quality measurements indicate, and automating chemical feed and monitoring (CDC, n.d.).
Put the pump and feed line right first, and tie the dose to make-up or blowdown volume. Write the LSI upper limit set by the treatment programme down as an operating rule, and lower the cycles or adjust pH when it is passed. If water quality has changed a lot but the same inhibitor is still fed at the same dose, revisit the inhibitor choice and dose with a water treatment specialist. Since the LSI shows only a tendency, judge it together with outcome measures such as exchanger approach temperature and calcium cycles (Langelier, 1936).
Hypothesis 5: silica-type scale
A hard, glassy film that acid cleaning barely removes calls for a fresh look at what the scale is. Suspect it first in systems fed with silica-rich groundwater, in systems run at high cycles, and where magnesium-rich water runs at high pH. Look into exchanger tubes after an acid clean and most show bare metal, yet some stretches still carry a translucent grey film. Compare the approach temperature before and after cleaning, and the recovery is smaller than expected.
Silica scale differs in how it forms in the first place. Groundwater and some mains water carry dissolved silicon dioxide, or silica, which concentrates in the tower like other ions. Unlike calcium carbonate, which grows as crystals, silica scale forms by polymerisation, with dissolved silicic acid molecules linking to one another. Below roughly pH 8.5 amorphous silica forms more readily, and above it magnesium silicate, which like calcium carbonate dissolves less when hotter and so sticks to hot walls (Demadis, 2003).
A common guideline keeps silica below 200 ppm at pH under 7.5, and below 100 ppm at pH 7.5 or higher (Demadis, 2003). Silica is not crystalline, so calcium carbonate inhibitors do little, and a separate approach that blocks polymerisation or keeps particles dispersed is needed (Demadis, 2003). Silica barely dissolves in ordinary acids, so acid cleaning can remove the calcium carbonate yet leave the silica film behind.
A drop of dilute acid is the first fork in the road. If a piece of scale fizzes and dissolves, it is calcium carbonate; if it barely reacts, it is silica or a silicate. Then measure silica and magnesium in the make-up and circulating water. If circulating silica is near the guideline above, silica is what limits the cycles. A composition analysis of a scale sample gives the firm answer.
If silica is near the guideline, set a separate cycle ceiling based on circulating silica. Do not push pH too high in magnesium-rich water, and consider whether a silica-specific inhibitor or dispersant is needed. Existing silica scale does not come off well with ordinary acid cleaning, so choose the cleaning method with a specialist contractor based on the composition analysis.
Testing the hypotheses
Begin by finding out what the scale is. If a sample fizzes and dissolves in dilute acid it is calcium carbonate, so look among hypotheses 1 to 4; if it barely reacts, go to hypothesis 5, on silica-type scale (Demadis, 2003). That single check splits the five hypotheses into two groups.
For calcium carbonate, look at two numbers. Actual cycles from conductivity above target mean hypothesis 1 (DOE, n.d.; EPA, 2012). Calcium cycles clearly below conductivity cycles mean calcium is already dropping out somewhere, so find where among the remaining hypotheses. If the make-up analysis shows hardness or alkalinity up against past records, it is hypothesis 2 (EPA, 2026; Safari et al., 2014).
Next, where the scale has built up. An even rise across all exchangers points to the water-side hypotheses 1, 2 and 4; scale thick only at the outlet of particular exchangers points to their velocity and heat load, hypothesis 3 (Kazi, 2020; Nesta & Bennett, 2004). Setting flow, pressure drop and approach side by side for exchangers fed by the same water is itself a control test. The water is the same and only the conditions differ, so if one unit alone is bad, its conditions are the candidate.
If scale rose suddenly, check the chemical feed log and the pump. Days when the LSI at exchanger outlet temperature passed its limit, or the inhibitor ran short, mean hypothesis 4 (CDC, n.d.; Langelier, 1936). Whatever the hypothesis, judge the job done only when the approach temperature recovers and calcium cycles line up with conductivity cycles again.
The hypotheses can also move together within one season. Summer evaporation raises the cycles, heat load climbs at the same time and lifts the wall temperature, and the LSI ends up past the inhibitor's limit, so hypotheses 1, 3 and 4 can all stack up at once. In that case fix what is quickest and easiest first, blowdown and chemical feed, then look at exchanger flow and make-up treatment.
Limits and cautions
Figures such as the cycle range, pH 8.6 and silica at 100 to 200 ppm are general values from the cited public bodies and literature; operating limits for a given plant come from water analysis and the manufacturer's recommended range. Closed loops, boilers, systems fed with seawater or reclaimed water, and cases dominated by calcium sulfate or calcium phosphate scale need a different assessment. The LSI and RSI are indices for gauging a water's tendency (Langelier, 1936; Ryznar, 1944), so in inhibited water follow the limits set by the treatment programme.
Acids, acid cleaners, biocides and scale inhibitors harm skin and eyes. Read the safety data sheet, the SDS, before handling, and wear safety glasses and chemical-resistant gloves. Acid cleaning carries corrosion and gas risks, so leave it to an experienced contractor. When servicing a tower, plan cleaning and disinfection together, as CDC guidance does, to prevent Legionella exposure (CDC, n.d.). If scale grows sharply in a short time, the approach does not recover after cleaning, or corrosion appears alongside, get an on-site assessment from a water treatment specialist.
Sources
- ASTM International. (2016). ASTM D1067-16, Standard test methods for acidity or alkalinity of water. https://www.astm.org/Standards/water-testing-standards.html
- ASTM International. (2025). ASTM D1126-17(2025), Standard test method for hardness in water. https://store.astm.org/d1126-17r25.html
- Berce, J., Zupančič, M., Može, M., & Golobič, I. (2021). A review of crystallization fouling in heat exchangers. Processes, 9(8), 1356. https://doi.org/10.3390/pr9081356
- Centers for Disease Control and Prevention [CDC]. (n.d.). Controlling Legionella in cooling towers. Legionella Control Toolkit. https://www.cdc.gov/control-legionella/php/toolkit/cooling-towers-module.html
- Demadis, K. D. (2003). Water treatment's "Gordian knot". Chemical Processing. https://www.chemicalprocessing.com/environmental-protection/water-wastewater/article/11380085/water-treatmentsgordian-knot
- Kazi, S. N. (2020). Fouling and fouling mitigation of calcium compounds on heat exchangers by novel colloids and surface modifications. Reviews in Chemical Engineering, 36(6), 653–685. https://doi.org/10.1515/revce-2017-0076
- Langelier, W. F. (1936). The analytical control of anti-corrosion water treatment. Journal of the American Water Works Association, 28(10), 1500–1521.
- Nesta, J., & Bennett, C. A. (2004). Reduce fouling in shell-and-tube heat exchangers. Hydrocarbon Processing, 83(7). https://www.hydrocarbonprocessing.com/magazine/2004/july-2004/extra-report-heat-transfer/reduce-fouling-in-shell-and-tube-heat-exchangers
- Nygren, M. A., Gay, D. H., Catlow, C. R. A., Wilson, M. P., & Rohl, A. L. (1998). Incorporation of growth-inhibiting diphosphonates into steps on the calcite cleavage plane surface. Journal of the Chemical Society, Faraday Transactions, 94, 3685. https://doi.org/10.1039/A806585C
- Ryznar, J. W. (1944). A new index for determining amount of calcium carbonate scale formed by a water. Journal of the American Water Works Association, 36.
- Safari, I., Hsieh, M.-K., Chien, S.-H., Walker, M. E., Vidic, R. D., Dzombak, D. A., & Abbasian, J. (2014). Effect of CO2 stripping on pH in open-recirculating cooling water systems. Environmental Progress & Sustainable Energy, 33(1), 275–282.
- U.S. Department of Energy [DOE], Federal Energy Management Program. (n.d.). Best management practice #10: Cooling tower management. https://www.energy.gov/node/849241
- U.S. Environmental Protection Agency [EPA]. (2012). WaterSense at work, section 6.3: Cooling towers. https://www.epa.gov/system/files/documents/2023-05/ws-commercial-watersense-at-work_Section_6.3_Cooling_Towers.pdf
- U.S. Environmental Protection Agency [EPA]. (2026, February 25). WaterSense cooling tower webinar slides. https://www.epa.gov/system/files/documents/2026-03/watersense_webinar_coolingtowers_feb_25_2026.pdf