In the previous post of this series, we introduced AVT as a treatment based on 100% volatile substances, and mentioned that it is divided into two variants, AVT(R), reducing, and AVT(O), oxidizing. The central difference lies in the presence or absence of copper alloys in the cycle. This post covers the first variant and why this requirement exists when copper alloys are present.
The goal of chemical treatment is not simply to prevent corrosion, but to make the protective oxide film covering each metal as stable and as insoluble as possible. Iron and copper have different electrochemical behaviors, so the redox potential condition that is optimal for one is not the same as the one that is optimal for the other.
While carbon steel forms a protective magnetite layer in a mildly oxidizing environment, copper is corroded precisely in that type of environment, forming cupric oxide (CuO), which is more soluble and less adherent.
Under reducing conditions, copper forms cuprous oxide (Cu₂O), a thin, adherent and protective layer. It is this inversion of behavior between the two metals that makes it impossible to treat a mixed cycle (steel and copper) with a single redox potential condition that is optimal for both at the same time.
That is why AVT(R) is an excellent compromise solution for Fe–Cu metallurgy, whereas AVT(O) is a more optimized solution for an essentially ferrous circuit.
AVT(R), in addition to the volatile alkalizing agent, also uses an oxygen scavenger that is likewise volatile. The role of this agent is to remove residual dissolved oxygen and keep the oxidation-reduction potential (ORP) of the feedwater low enough to favor the formation of Cu₂O instead of CuO.
The reducing agents most cited in the boiler water treatment literature include carbohydrazide and diethylhydroxylamine (DEHA), which have been replacing hydrazine, historically the most widely used, driven by the classification of hydrazine as a suspected carcinogen and the regulatory handling restrictions that came with it, not by any chemical limitation of the product itself.
The use of a reducing environment to protect copper alloys is not a recent innovation. The practice was consolidated over decades of operation of conventional fossil cycles, in a period when practically every condenser and every low-pressure heater used copper alloy as a design standard. AVT(R), in this context, was not created specifically for supercritical boilers; it was adapted to them when that generation technology became widespread, inheriting a metal-protection logic that already existed for subcritical units with the same mixed-metallurgy problem.
In the evolution of boiler treatment systems, copper alloys were adopted as a standard in low-pressure heaters in the 1930s-1940s, during the regenerative heating phase. The corresponding treatment program was named All-Volatile Treatment Reducing (AVT(R)).
In mixed-metallurgy cycles containing copper alloys in the feed circuit, AVT(R) normally operates in a narrower pH range, typically about 9.0–9.3 at 25 °C. This range is lower than the one that may be more favorable to minimizing FAC in predominantly ferrous systems, in which higher pH reduces the solubility of iron oxides. Thus, there is a trade-off between the chemistry needed to minimize corrosion and copper transport and the chemistry that gives carbon steel greater resistance to FAC. However, this trade-off depends not only on pH, but also on redox potential, dissolved oxygen, temperature, the alloy and hydrodynamic conditions.
Add to this the absence of buffering capacity, a characteristic of all AVT. A poorly dosed reducing agent, whether in excess or in shortage, has no chemical reserve to cushion the deviation, and the effect quickly shows up in the condition of the metal.
Unlike phosphate, where the central monitoring parameter is the Na/PO₄ ratio, in AVT(R) the parameter is the oxidation-reduction potential of the feedwater, together with the residual reducing agent content and pH. It is not enough to dose the product and assume the environment is reducing: an ORP above the expected range is the first sign that the agent was underdosed, decomposed thermally before performing its function, or that air ingress is compromising control.
AVT(R) exists because two metals with opposite electrochemical behaviors share the same water cycle. It is not the ideal variant for either of them, it is the possible compromise when copper remains in the design. In the next post of this series, we move on to AVT(O). What changes for a copper-free system, and why the absence of this restriction allows an operating regime more favorable to carbon steel.
Does your plant still operate with copper alloys in the cycle, and are you unsure about the ORP range or which reducing agent applies to your case? Talk to us for a technical assessment.
References
- Power Engineering. The importance of accurate dissolved oxygen condensate/feedwater monitoring. Available at: power-eng.com.
- Power Engineering. Re-emphasizing HRSG feedwater chemistry evolution. Available at: power-eng.com.