Condenser Cleaning Pitfalls: Five Mistakes from Scale Diagnosis to Acceptance

Abstract: Condenser cleaning rework rarely traces back to the wrong chemical; it usually traces back to two stages that were skipped - the diagnosis and the acceptance check. A vacuum drop is treated as proof of scaling before the vacuum system and the circulating water side have been ruled out, and the job closes on a visual inspection alone. This article works through the five mistakes seen most often on site: how to judge scaling from end difference, circulating water flow and vacuum together; the fluoride and chloride limits for titanium and stainless tubes; isolation and vacuum control during half-side cleaning; why the water-box corners, tubesheet and division plate have to be cleaned; how rubber ball on-line cleaning is sequenced against the chemical clean; and how end difference and cleanliness factor are verified at handover. Process figures follow DL/T 957-2017 and GB/T 25146-2010, and an anonymised half-side clean on a 150 MW unit closes the article.

1. Fix the Diagnosis First: Vacuum Loss Is Not Proof of Scaling

The first mistake is made before any chemical is ordered. A falling vacuum with a rising end difference is read as proof that the bundle is scaled, but the same symptoms come from three different causes: scale on the tube bundle, air leaking into the vacuum system or an extraction stage losing performance, and circulating water flow or temperature drifting away from design. The symptoms overlap; the remedies do not, so the checks run from the cheapest to the most expensive.

The circulating water side comes first, because it can be read straight off the panel: pump configuration, cooling tower outlet temperature and circulation ratio against design. Then the vacuum system, with a tightness test and a look at the air leakage rate and the extraction stage. Only when both are clean is the tube bundle a suspect. At that point end difference - saturation temperature minus circulating water outlet temperature - and cleanliness factor decide: a unit in good order runs an end difference of 4-5 C and a cleanliness factor of 0.80-0.90 at full load, and scaling becomes convincing only when the end difference passes 8 C and the cleanliness factor drops below 0.70.

The last step is a deposit sample. Open the water box, record layer thickness and colour, and identify the deposit by acid solubility; carbonate scale under slime is the most common combination on the circulating water side, and a layer above 0.3 mm is reason enough to plan a clean. Skipping this step and ordering chemicals from experience is where most of the rework described here starts.

2. Mistake One: Cleaning Titanium Tubes on a Carbon Steel Programme

Condenser tube bundles are usually titanium, sometimes stainless steel or copper alloy, and the three tolerate very different cleaning chemistry. Titanium absorbs hydrogen in fluoride-bearing media and suffers hydrogen embrittlement, so a fluoride component is a veto item for titanium tubes. Stainless tubes are sensitive to chloride, which is why the cleaning liquor is held at 25 mg/L, and copper alloys need a copper inhibitor in the formulation. Carrying a carbon steel programme straight across to a condenser usually ends in pitting at the tube ends and, in the worst case, a scrapped bundle.

Temperature and concentration are held back for the same reason. Titanium tubes are commonly 0.5-0.7 mm wall, so the wetted surface is thin: the cleaning liquor is capped at 60 C, the concentration is calculated from the measured deposit thickness, and temperature, pH and iron are logged every 30 min. A sudden rise in iron means local over-cleaning; the response is to dilute the liquor or add inhibitor, not to keep running the clock. Sulphamic or citric acid systems are the usual choice, with same-material coupons in the circuit judged by weight loss under GB/T 25146-2010.

3. Mistake Two: Half-Side Isolation and Vacuum Control Done Halfway

Half-side cleaning exists to avoid a shutdown: one water box is isolated and cleaned while the other keeps the unit on load. The failures cluster around isolation. Closing the circulating water valves without a stop plate lets a passing valve leak circulating water into the cleaning circuit, diluting the liquor so that both the duration and the result slip; leakage the other way carries cleaning liquor into the circulating water system and disturbs the whole water chemistry programme.

Isolation has three layers: valves closed and tagged, stop plates fitted where the pipework allows, and the cleaning circuit run on a dedicated temporary pump and mixing tank, physically separated from the circulating water system. A temporary baffle goes over the open manhole and the opening is kept short, so that air cannot leak into the steam side and pull the vacuum down - noticeable on units whose tightness margin is already thin. Running-side vacuum is held above -95 kPa with exhaust temperature inside limits, and one side, including drain-down, cleaning, rinsing and closing up, is finished inside a day. The two sides run in separate windows, about 2 days in total.

4. Mistake Three: Tubes Jetted, Water Box Left Dirty

Tube-by-tube water jetting is the most visible stage of condenser cleaning, and pressure follows the material: 20-30 MPa for titanium and stainless tubes, 30-50 MPa for carbon steel, with the nozzle centred before the lance advances to avoid gouging the tube ends. The mistake is treating this stage as the whole job. Slime and scale left in the water-box corners, on the back of the tubesheet, behind the division plate and deflector, and on the secondary strainer and ball collection screen is carried back into the tubes within weeks, and the end difference climbs again.

Dead zones are cleaned differently from tubes: low-pressure washing with manual work, using soft tools rather than a wire brush on the tube holes, since the 3-5 mm chamfer at each tube hole holds hard scale and has to be checked hole by hole. A flow check before the manhole goes back on confirms that nothing is left lying in the water box, and photographs with point numbers go into the handover file.

5. Mistake Four: Rubber Ball Cleaning and the Chemical Clean Collide

Rubber ball cleaning is a maintenance measure, not a replacement, and that much is settled at the selection stage. What still goes wrong on site is sequencing: the chemical clean and the ball system are run as if they could overlap. Balls have to be removed and the collection screen emptied for the duration of the chemical clean, otherwise they take up cleaning liquor and jam at the tube inlets, disturbing flow distribution in the circuit and damaging the balls themselves; once the clean is finished, the ball system should go back into service quickly so that the tube surface keeps being wiped.

Check the ball system parameters at the same time: wet ball diameter normally 1-2 mm above the tube bore, recovery rate not below 90%. A recovery rate that stays low points to a diameter mismatch or a blocked collection screen, and in that state the bundle re-scales quickly even after a good clean. Put the ball system back within 3 days of the chemical clean and log the recovery rate weekly, alongside the unit's water and steam chemistry records.

6. Mistake Five: Acceptance Stops at a Bright Tube Bundle

Acceptance has two tracks: cleaning quality, judged by scale removal, corrosion rate and passivation state, and operating performance, judged by end difference, cleanliness factor and vacuum. A torch shone down a tube proves nothing measurable, and it cannot explain why vacuum did not recover - which usually points at the vacuum system itself rather than at the cleaning work.

Performance figures are read after 72 h at stable full load; readings taken just after start-up are not admissible either way. The clean has produced measurable value when the end difference is back within 0.5 C of the pre-shutdown baseline, the cleanliness factor is back above 0.83 and vacuum has recovered by 1 kPa or more. On a 150 MW unit, 1 kPa of recovered vacuum corresponds to roughly 1 g/(kW.h) of supply coal consumption, which is the economic case for keeping condenser cleaning in the annual outage plan.

7. Case Study: Half-Side Cleaning on a 150 MW Unit

In May 2026 Blue Star Cleaning cleaned the condenser of a 150 MW unit at a thermal power plant in East China, working half-side with the unit on load over two windows. The condenser is a two-pass surface type with about 9,000 m2 of cooling surface, some 12,600 titanium tubes of 25 x 0.5 mm, and once-through circulating water. Before the work, vacuum sat about 1.5 kPa below design, the end difference had risen from 4.2 C to 8.8 C, and the cleanliness factor was 0.68.

The diagnosis came first: circulating water flow and inlet temperature were on design and the tightness test passed, so the water box was opened and samples taken. The layer measured 0.3-0.6 mm, grey-white hard scale under slime, identified as mainly carbonate. The route followed from that: stop plates fitted in the water box, cleaning circuit on a temporary pump, a sulphamic acid system circulated at 40-50 C for 4 h with chloride held below 25 mg/L and no fluoride component at any point, the tube bundle jetted at 20-30 MPa tube by tube, and the water-box corners and tubesheet back cleaned by hand.

One side took a day: 3 h drain-down and isolation, 4 h circulation, 2 h rinse and displacement, 2 h water-box cleaning and closing up, with the two sides taken in separate windows and no visible effect on unit load. Results: 96% scale removal, coupon corrosion rate 0.15 g/(m2.h), and a go-through check with no blockages or tube-end damage; after 72 h at full load the end difference was back to 4.5 C, the cleanliness factor to 0.84 and vacuum had recovered 1.2 kPa. The plant then wrote the ball system start-up requirement and water-box dead-zone cleaning into its maintenance work instructions. For route selection and non-stop cleaning options, see condenser cleaning methods compared and on-line non-stop chemical cleaning.

8. Frequently Asked Questions

How much vacuum drop and end difference justify cleaning a condenser?

Rule out the circulating water and the vacuum system first. With circulating water flow and inlet temperature on design and a satisfactory vacuum tightness test, an end difference that has risen from 4-5 C to above 8 C and a cleanliness factor below 0.70 is enough to treat the tube bundle as scaled. Inspect the water box or pull a sample tube, and only order chemicals once the deposit exceeds 0.3 mm, rather than dosing on a single vacuum reading.

Can a titanium tube condenser be cleaned with a fluoride-bearing chemical?

No. Titanium absorbs hydrogen in fluoride-bearing media and suffers hydrogen embrittlement, so fluoride is a veto item for titanium tubes. Titanium systems normally run a sulphamic or citric acid system at 40-50 C, with chloride in the cleaning liquor held at 25 mg/L, a limit that also protects the tubesheet and any stainless parts, and the temperature ceiling kept at 60 C.

How is vacuum on the running side held during half-side cleaning?

By water-box isolation and by controlling the window. Close and tag the circulating water valves on the cleaned side and fit a stop plate where possible, so a passing valve cannot dilute the cleaning liquor; add a temporary baffle over the open manhole and keep it open briefly so air cannot leak into the steam side. Hold running-side vacuum above -95 kPa, keep exhaust temperature within limits, and finish one side, including drain-down, cleaning and rinsing, inside one day.

End difference climbs back two or three months after cleaning - where does it come from?

Usually from dead zones and from on-line maintenance. Slime left in the water-box corners, on the back of the tubesheet, behind the division plate, and on the secondary strainer or ball collection screen is carried back into the tubes within weeks; if the rubber ball system is not put back in service after the chemical clean, or the recovery rate sits below 90% or the ball diameter does not match the tube bore, the bundle gets no continuous wiping. Fix both and the cleaning interval on similar units usually returns from 8 months to 12-18 months.

Which figures are used to accept a condenser cleaning job?

Two sets. For quality, scale removal, not less than 95% for carbonate scale and slime, and the corrosion rate of same-material coupons, judged by weight loss under GB/T 25146-2010. For performance, end difference, cleanliness factor and vacuum, all read after 72 h at stable full load. On a 150 MW unit, end difference came back from 8.8 C to 4.5 C, cleanliness factor from 0.68 to 0.84 and vacuum recovered 1.2 kPa, figures that say more than a bright-looking tube bundle.

All cleaning processes and acceptance criteria in this article follow GB/T 25146-2010 Quality acceptance specifications of chemical cleaning for industrial equipment and DL/T 957-2017 Guide for chemical cleaning and film formation of condensers in fossil fuel power plants.

Service coverage: Jiangsu (Nanjing, Suzhou, Wuxi, Changzhou, Zhenjiang, Yangzhou, Taizhou, Yancheng, Suqian, Nantong, Huai'an, Lianyungang, Xuzhou) · Zhejiang (Hangzhou, Ningbo, Jiaxing) · Shanghai · Anhui (Hefei, Ma'anshan, Wuhu, Xuancheng)

© 2026 Danyang Blue Star Anti-Corrosion Cleaning Co., Ltd. All rights reserved | Source: www.lanxingqingxi.com