CCGT’s ppb problem: why ultra‑pure steam makes or breaks turbine output

[Steam purity is critical](https://beta.co.id/en/blog/the-ppb-problem-why-steam-purity-makes-or-breaks-turbines-in-ammonia-plants) for reliable CCGT (combined‑cycle gas turbine) operation. Even [trace carryover of boiler water](https://beta.co.id/en/blog/the-ppb-economy-why-ultra-pure-steam-makes-or-breaks-hrsg-turbines) or condensate deposits solids on blades and turbine internals, increasing friction and reducing output. Analyses show that a mere 3 mil (0.075 mm) deposit layer on blades can raise fuel costs by 1–2% (researchgate.net). In one case a 30 MW turbine lost over 5% capacity before shutdown after only 3 months, due to deposits from poor steam‑water separation (watertechnologies.com).

That is why modern steam‑chemistry guidelines set very tight thresholds: sodium at ≲2 µg/kg (ppb, parts per billion) and silica at ≲10 µg/kg in superheated steam (studylib.net; chemtreat.com). IAPWS guidance recommends cation‑exchange conductivity (CACE, a conductivity measure after passing the sample through a hydrogen‑form resin) below 0.2 µS/cm and Na<2 ppb, SiO₂<10 ppb at the turbine inlet for high‑pressure, superheated steam (studylib.net).

Steam purity limits and turbine risk

Violating these thresholds leads to hygroscopic films and droplet condensation (“carrying over” solids) in cooler stages, causing accelerated blade corrosion and erosion. Notably, industry practice typically limits steam silica to <0.02 ppm (<20 ppb) — and often <0.01 ppm in stringent cases (watertechnologies.com) — since silica solubility drops as steam expands. Maintaining these limits protects efficiency and avoids unplanned outages: adherence to purity specs is directly linked to plant heat rate and availability (researchgate.net; watertechnologies.com).

In plant practice, discussion of these ppb‑level limits typically includes polishing of return condensate; a condensate polisher is a common way operators talk about maintaining low sodium and silica at the turbine inlet.

HRSG steam‑drum separation internals

HRSG (heat recovery steam generator) steam drums use multi‑stage separation to remove water and entrained solids and yield dry, pure steam. Drum internals typically include one or more centrifugal (cyclone) separators and layered demister pads (chevron or mesh scrubbers) upstream of the steam outlet. A recommended design couples cyclonic primary separators with chevron secondary scrubbers and a “dry pipe” collector, ensuring effective droplet removal (esteemprojects.com). One engineering guide states: “use of centrifugal primary separators with chevron secondary scrubbers … and a dry pipe on the steam outlets” (esteemprojects.com).

This multi‑stage approach can achieve removal of essentially all large droplets — often on the order of 99% of entrained moisture. The dry pipe (a takeoff located below liquid level) then collects only the topmost steam. Proper sizing is essential: guidelines suggest drum diameters ≥1.2 m and vapor velocities limited (e.g., horizontal ~0.65√((ρ_l–ρ_v)/ρ_v) ft/s) to prevent carryover (esteemprojects.com). These measures minimize mechanical carryover (liquid entrainment); failure of internals is a known cause of rapid fouling (watertechnologies.com; chemengonline.com).

Mechanical carryover as small as 0.02–0.05% of flow (typical spec for HP drums) can deposit enough solids to degrade turbine performance. Studies of steam generators confirm that with adequate cyclone/mesh internals and drum‑level control, carryover can be held to negligible levels, preventing turbine contamination (esteemprojects.com; chemengonline.com).

Online steam‑quality monitoring and diagnostics

Continuous monitoring of steam quality is standard to detect excursions in purity. Water/steam chemistry sensors watch key parameters in real time. Critical sample points include main/reheat steam and condensate/feedwater (chemengonline.com). Instruments include CACE analyzers, sodium analyzers, silica analyzers, and pH/oxygen meters in the feed/condensate loop.

One industry source recommends continuous monitoring of water makeup with ≤0.1 µS/cm conductivity, ≤10 ppb silica, ≤2 ppb Na (chemtreat.com). Where makeup desalinization is part of the architecture, operators frequently reference reverse‑osmosis trains such as brackish‑water RO when discussing how upstream quality control supports those limits.

At the condensate pump discharge, guidance cited in industry sources targets CACE ≤0.2 µS/cm, Na ≤2 ppb, DO (dissolved oxygen) ≤20 ppb (chemengonline.com). Some plants describe pairing continuous monitoring with continuous ultra‑pure water production; systems such as electrodeionization (EDI) are often mentioned in that context.

Monitoring steam specifically (e.g., on main or reheater lines) is crucial: impurities in attemperator sprays or carryover would appear here first. In practice, operators sample superheated steam and reheat lines for conductivity/Na to catch any turbine‑bound contamination; saturated steam is sampled periodically only to verify separator efficiency (chemengonline.com). Modern analytics even include particle counters or spectrometers for droplet detection. Compared to offline isokinetic sampling (slow lab tests), online monitors can catch rapid transients and predict trends (researchgate.net).

Integrated analytics (combining feedwater and steam sensors) allow diagnostics: if both condensate Na and CACE rise in tandem, a condenser leak is indicated, whereas a steam‑only spike points to separator failure (chemengonline.com; chemengonline.com). Proactive steam‑chemistry surveillance has demonstrable ROI: detecting a condenser leak via Na before it spikes CACE or silica can avoid major deposits, and trending conductivity can prevent corrosive lay‑ups. As a complement to monitoring, final polishing steps like a mixed‑bed demineralizer are often part of plant discussions about maintaining consistent, ultra‑low impurity levels.

Final‑stage design for wet‑steam erosion

The LP (low‑pressure) — and sometimes IP (intermediate‑pressure) — turbine stages operate with the highest moisture fractions and thus face the worst erosion. Condensation in expanding steam forms droplet impact, causing “honeycomb” or serrated erosion on trailing edges (huirui-laser.com; watertechnologies.com). Operation off‑design (low load or below design inlet steam temperature) exacerbates this: “operation below design inlet steam temperature or at low load can cause condensation… leading to erosion” (watertechnologies.com).

To mitigate, turbine OEMs employ specialized materials and blade designs. Modern last‑stage blades are often high‑alloy steels or superalloys with tough trailing edges. Hard‑surface coatings or claddings are applied: laser‑cladded Ni/Co‑based alloys (e.g., Stellite‑type or WC/CoCr) create very dense, well‑bonded wear layers that greatly slow droplet erosion (huirui-laser.com; huirui-laser.com). Coating studies show these layers are denser and more erosion‑resistant than older electrodeposited or brazed hardfacing (huirui-laser.com).

Beyond materials, engineering strategies include shaping blade profiles to shed water and adding moisture “separation screens” upstream. In one case, an operator replaced turbine bypass and control valve trims with “wet‑steam erosion” trim designs, doubling the number of allowable cold starts from ~145 to 245 before any erosion was seen (imi-critical.com). That modification saved over $250k in one year (imi-critical.com). Applying erosion‑resistant alloys/coatings on the final turbine stages significantly extends blade life.

Evidence base and operating payoffs

Authoritative guidelines and engineering studies consistently stress ultra‑low impurity limits — e.g., Na<2 ppb and silica<10 ppb (studylib.net; chemengonline.com). Case analyses quantify effects: tiny deposits reduce output by several percent, while improved internals/coatings dramatically extend operational life (researchgate.net; imi-critical.com). The discussion above is drawn from laboratory studies, industry handbooks, and plant‑case reports (researchgate.net; studylib.net; watertechnologies.com; chemtreat.com). In the background of many of these programs are full makeup and polishing trains; for example, operators frequently reference membrane systems when discussing how water production integrates with online monitoring and HRSG separation hardware.