Platinum Metals Rev., 1967, 11, (3), 100
Efficiency of Platinum Gauzes in the Manufacture of Nitric Acid
A Method for Determining the Frequency of Pickling
In an ammonia oxidation plant for the manufacture of nitric acid the rhodium-platinum alloy catalyst gauzes must be pickled at intervals to remove accumulated impurities such as iron oxide and so to avoid a reduction in catalytic efficiency. This article outlines a method recently devised to determine the optimum frequency of pickling required to ensure consistent operation at high efficiency.
The largest contribution to the costs of operating a modern nitric acid plant is made by the cost of liquified ammonia. Because of this, it is incumbent on plant operators to reduce to an absolute minimum all ammonia losses at each stage of the process. One stage at which significantly less than 100 per cent efficiency is generally obtained is the oxidation of the ammonia to nitric oxide over the rhodium-platinum gauze pad. Conversion efficiencies are generally about 94 to 95 per cent, but vary with the age and condition of the platinum alloy gauzes and may drop to 90 per cent if precautions are not taken to maintain a high level of gauze activity. Pickling the gauze pad in hydrochloric acid to remove impurities carried over in the feed gas (mainly iron oxide scale) must therefore be carried out as frequently as production requirements permit in order to maintain the gauzes in their active state and to prevent an economically undesirable drop in conversion efficiency.
Because the determination of ammonia oxidation efficiency is highly dependent on the analytical methods used in measuring gas concentrations, and since these are not always easy to keep constant, apparent converter efficiencies may vary with time, making it difficult to decide on the optimum time at which a gauze should be pickled. This article reviews work that has been undertaken by Imperial Chemical Industries Limited, Nobel Division, to examine the cumulative sum (cusum) technique when applied to converter efficiency determination (1). The nitric acid plant used for these tests operates with the converters at atmospheric pressure and with the subsequent absorption stages at 3 to 4 atmospheres pressure. The plant has been described elsewhere (2).
Analytical Methods for Determining Converter Efficiencies
The efficiency of the ammonia oxidation is measured by the chemical determination of the ammonia concentration in the feed gas and the nitric oxide concentration in the gas immediately below the platinum alloy gauze pad. Ammonia is generally determined by titration with standard acid, and the oxides of nitrogen by back-titration with standard acid after absorption in an aqueous alkali solution. Two analytical methods were compared during the initial experiments:
In the first instance, about 4 litres of the ammonia/air mixture were drawn off above the gauze pad during a period of about fifteen minutes, and ammonia titrated simultaneously with the sampling. Hydrogen peroxide was used for the nitrous gas oxidation.
In the second method, 0.8 litres of the feed gas were taken off during thirty seconds, and the oxygen remaining in the residual air was used for the oxidation of the nitrous gas sample.
The Cumulative Sum Technique
It is not always possible to determine at daily intervals the data required to plot a cumulative sum chart for the converter efficiencies. A corrected cusum was calculated therefore for each of the two analytical methods described. The datum is the conversion efficiency below which it had been decided that gauze pickling would be carried out.
The differences between these duplicate determinations of conversion efficiencies and the datum line were calculated and their arithmetic mean (x ) was taken. These values would normally be added to each other, and plotted daily to give a cusum chart. Since this was not always possible, the arithmetic mean of the successive values of x was calculated and multiplied by the time interval in days. The results were added to each other to give a corrected cusum chart. This method enables such a chart to be compiled with results obtained at irregular intervals and the results were found to be almost identical with similar charts obtained using data from regular, daily determinations. The example given in the table was based on a datum of 93 per cent.
|Day||Conversion Efficiency (per cent)||Difference from Datum||Mean x||Corrected Cusum|
The corrected cusum for the fifth day, for example, was obtained by taking the mean of 4.6 and 4.9 multiplied by the time interval in days and added to the original 4.6.
The Cusum Method in Use at Ardeer
Fig. 2 illustrates a cusum plot for one of the plant’s ammonia converters, but in this instance a datum of 94 per cent had been chosen. From day 12 to day 50 this burner operated at a conversion efficiency greater than 94 per cent, and from day 50 to day 78 at less than 94 per cent.
To maintain a conversion efficiency equal to the datum already chosen it is therefore necessary to pickle the gauzes after approximately eighty days’ service.
It is obvious that if it is possible to pickle when the cusum plot has just passed through its peak, the converter efficiency will be maintained at a value considerably in excess of the datum. The dotted line in Fig. 3 is a typical cusum plot for the number two burner; the other curves are obtained by pickling the gauze pad at intervals of one to two months. The cusum plots in Fig. 3 show that the conversion efficiency can be maintained at a value above the datum, in this case 94 per cent.
Use of the cusum method for determining the optimum time for pickling the converter gauze pads has enabled the Ardeer nitric acid plant to increase significantly its overall nitrogen efficiency.
- 1R. H. Woodward and P. L. Goldsmith, “Cumulative Sum Techniques”, I.C.I. Statistical Methods Panel, 1962
- 2D. T. Austin and J. G. Sloan, Ind. Chem., 1961, 37, 159 ; J. G. Sloan and H. C. Staats, Platinum Metals Rev., 1961, 5, 54