Organic Contaminants and Treatment Chemicals in Steam-Water Cycles

Abstract

Boiler feedwater and steam have to be of high purity, because of the susceptibility of the steam-water cycle to corrosion. Organic contaminants break down in boilers by hydrothermolysis, leading to the formation of organic acid anions, which are suspected to cause corrosion of steam-water cycle components. The impact and behavior of organic decomposition products in the steam-water cycle are not well understood. While guidelines for organic contaminants are becoming stricter, organic treatment chemicals are gaining popularity. Some alkalizing amines show potential for protecting steam-water cycles against corrosion, but their thermal stability is limited and acidic decomposition products are a concern. There are no official guidelines for the application of alkalizing amines in fossil-fired plants, because their thermal stability in the hottest sections of the plant is unknown. This thesis aims to contribute to well-founded guidelines for organic contamination and amine application in steam-water cycles to facilitate effective water treatment and chemical dosing. It does so by investigating the thermal decomposition of organic contaminants and treatment chemicals and using results to conduct corrosion experiments. To investigate hydrothermal reactions both batch and continuous flow reactors were tested, and it was concluded that the latter are better at investigating (hydro)thermolysis of organic treatment chemicals. A flow reactor gives precise control over retention time, temperature and pressure and these parameters can be changed much faster than with a batch reactor. The flow reactor was the basis for virtually all thermal stability and decomposition experiments described in this thesis. It was found that lower heating rates give more organic acid anions as degradation products of organic carbon, both in quantity and species variety. Thermal stability of the decomposition products determines which of these products is most prevalent. As boiler temperature increased, acetate became the dominant degradation product, due to its higher thermal stability. Shorter retention times led to more variety and quantity of organic acid anions, due to a lack of time for the thermally less stable ones to degrade. The absence of oxygen increased the thermal stability of decomposition products. The (hydro)thermolysis of monoethylene glycol and slurry oil produced up to a few hundred ppb acetate and formate. The (hydro)thermolysis of polyethylene glycol produced a high variety and quantity of organic acid anions, that could only partially be determined to be in the lower ppm range. The (hydro)thermolysis of the water dissolvable fraction of gasoil, naphtha and hydro wax did not lead to an increase in organic acid anions. Methyl ethyl ketoxime thermally degraded into several organic acid anions and nitrite in the higher ppb or lower ppm range. By using the flow reactor to investigate amine thermal stability, it was concluded that thermolysis under superheater conditions was more rapid than hydrothermolysis under boiler conditions. Anionic decomposition products increased linearly over time, while the thermal decomposition of morpholine followed first order kinetics. Metal catalysis of amine thermolysis caused by oxides on the inner surface of superheater tubes was investigated by using varying sizes and elemental composition. Kinetics of morpholine and ethanolamine thermolysis decreased as the tube size increased. The relation between the surface:volume ratio and the degradation rate constant was linear. Although results varied between the two applied tubing materials, there is no consistent trend that can link thermolysis kinetics to tube wall composition. The thermolysis of five alkalizing amines and two organic acids was comprehensively tested at superheater conditions. Morpholine, ethanolamine, cyclohexylamine, dimethylamine, 3-methoxypropylamine and acetic acid were shown to undergo thermolysis according to first order kinetics. The activation energy, prefactor and activation volume were obtained from the experimental data for all investigated amines. Dimethylamine did not fully degrade, in spite of longer retention times being applied, suggesting synthesis may occur. Formic acid is very unstable under steam water cycle conditions. It is still found in high temperature and pressure steam-water cycles, though, and therefore it could be hypothesized that it is synthesized in the condensing stages. Acetic acid has higher thermal stability than all other tested compounds and is therefore the dominant organic acid anion at high temperatures. Cationic degradation products were ammonia and some amines, meaning that the complete thermolysis of an amine does not necessarily lead to acidic conditions, as the formed ammonia also provides protection. A model was constructed to predict the thermal stability of the amines in steam-water cycle. More plant data is necessary to fully validate the model. Runs conducted with an experimental two-phase flow-accelerated corrosion loop showed a linear relation between liquid film pH and obtained corrosion rates for the same steam quality. The tested steam quality was not high enough to create the conditions in which ammonia provides insufficient protection against acetic acid, but expanding the liquid film pH model to higher steam qualities does give an idea of the dangers of high acetate concentrations in a steam-water cycle. The models for calculating the pH of the liquid film in two-phase flow and amine thermolysis (if validated) could be connected to assess if alkalizing amine application is recommended for a specific steam-water cycle. In general, it can be concluded from the results that ethanolamine provides better protection against two-phase flow-accelerated corrosion in the presence of organic acid anions, so there is a maximum superheater temperature for each amine at which it can be applied. When ammonia is the only volatile treatment chemical protecting the steam-water cycle, organic acid anions in the plant should be reduced until theoretical pH drop of the two-phase liquid film is acceptable.