Licentiate seminar

Modeling of Electrolytic Pickling

Defendant Main Advisor Extra Advisor Date
Nulifer Ipek Fritz Bark Noam Lior 2002-03-06

Göran Lindbergh, KTH, Mekanik, Faxén Laboratory

Evaluation committee


Pickling is the most common of several processes used to remove scale from steel surfaces. Dissolution of the oxide scale by dipping in several successive baths containing strong acid mixtures, has been the traditional way of removing it since the late 1700s. This method, however, has many negative effects, such as creation of a corrosive environment, attack of the base metal, and NOx emissions and nitrate effluents which pollute the air and water. A much more benign method, studied here, is neutral electrolytic pickling, which uses a neutral solution of sodium sulphate. Descaling is carried out by electrolytic action, achieved by application of electrical current. The efficiency of neutral electrolytic pickling processes in industrial use was reported to be only 20-30 %, partly due to the evolution of gas by the accompanying water electrolysis, and partly because much of the applied current does not reach the steel being pickled. The primary objective of this study is to provide a better fundamental understanding of the process and to thereby optimize it. To the best of our knowledge, this is the first attempt to model the process comprehensively. The improvements sought are to increase the line output, and reduce the consumption of energy and chemicals. Supported initially by Avesta Polarit AB, the modeling was focused on the electrolytic pickling tanks for stainless steel bands at their Avesta plant. A critical literature review was conducted, to elucidate the state of the art of the electrolytic pickling method mainly, but also of the acid pickling method. Neutral electrolytic pickling is reported to provide a fast descaling rate due to the conversion, at high anodic overpotentials, of the otherwise hard to dissolve chromium oxide into more soluble Cr(VI)-compounds. The dissolution of the oxide scale is achieved by electrochemical reactions, which require participation of electrons. The electrons are supplied by the externally applied current, which is induced in the steel band by the imposed electric potential on the electrodes. The rate of scale dissolution is proportional to the applied current density, and thus by controlling it also the pickling process can be controlled. In addition both electrochemical and chemical dissolution of the scale and base metal occurs. No complete set of dissolution reactions, however, and reaction rates is provided in the literature. The dissolution behavior of oxides formed on different type of steels, has however not been clarified yet. A progressive approach to understand and model the electrolytic pickling process was adopted. An electrostatic analytical model, to study the potential distribution in an electrolytic pickling cell without flow, scale dissolution reactions, and electrolysis, was a first step in the study. The model, even at this highly simplified stage, showed from the current fields that much of the applied current short-circuits between the electrodes instead of being transferred to the steel band for descaling. This model was used to investigate the effects of geometry, configuration and electrolyte and steel properties on the current and potential distributions, and ways to increase the process efficiency were found and verified experimentally. The second step in the study was the development of a mathematical model that now added consideration of the electrolytic gas evolution on the electrodes and on the steel band. The model was solved numerically, using the finite element program FEMLAB. The bubble velocity was found to be of the order of 0.1 m/s, of the same order as that of the steel band, and the conclusion is that it therefore needs to be considered when computing hydrodynamic effects and transport (which will be accomplished in the next stage of the project). Experiments were conducted in a small experimental electrolytic pickling cell to obtain better understanding of the process and for model validation. The analytical model shows trends very similar to those observed in the experiments. It shows, for example, how the current efficiency decreases as the electrode-band distance increases, and how it increases with the band-to-electrolyte conductivity ratio. Bubbles have been observed to cover about 30 % of the electrode and band surfaces, with especially strong coalescence on the bottoms, but the coverage was significantly reduced when the electrolyte was made to flow. This is believed to yield a more even current distribution and homogeneous pickled surfaces, but needs further studies. The numerical model thus shows an encouraging increasing agreement with the experimental results. Another encouraging result from the numerical model is that the prediction of overpotential at the steel band, being the driving force of the electrochmical reactions, agrees reasonably with that reported in the literature.