Minor quantities of certain sulphur compounds can cause natural gas liquids (NGL) to fail the copper strip corrosivity test. At Esso Resources Canada Limited's Quirk Creek Gas Plant, it was found that there were periods, during high plant throughput, when NGL quality excursions were experienced. To gain control of the NGL quality and to ensure that the product could be shipped by pipeline, a program was undertaken to test various ways of treating the NGL product. In this paper, the experimental work involving slip stream testing and analysis of sulphur compound removal capacity at ambient temperatures is discussed. The resulting design of a full stream catalized zinc oxide treating unit installed to assure the NGL product meets the copper strip corrosivity test specification is described.
Mixed NGL's and condensates produced in Alberta are gathered and transported by pipeline to Edmonton. Some of these liquids are then batched through the Cochin or Interprovincial Pipeline System to the Sarnia, Ontario area. Refined white products, such as aviation fuels, are also batched in these pipelines. Because of the possibility of pipelines. Because of the possibility of contamination of these refined products by the mixed condensates, a rigorous batching procedure has been established and fairly strict product specifications have been imposed on the mixed condensates. In particular, the mixed condensates must contain no more than 0.15 wt. % sulphur and must be able to pass the copper strip test with a number 1A or 1B rating.
The copper strip test is the standard test for corrosivity of a LPG such as fractionated propane and butane. The test procedure is covered by ASTM D-1838 and involves immersing a polished copper strip in the liquid product sample for one hour at 100 degrees F. The test strip is then rated by colour comparison with the ASTM Copper Corrosion Standards. In this comparison, the strip will be rated as a No. 1 (slight tarnish), No. 2 (moderate tarnish), No. 3 (dark tarnish), or No. 4 (corrosion). Unfortunately, the rating of the strip is often very subjective. Furthermore, the test is very labour intensive and yet gives the operator very little information as to how to modify his process to correct any problem.
There is no good correlation between the level of sulphur constituents in the product and the results of the copper strip test. This is particularly true when the copper strip test is applied to a mixed condensate which contains a significant quantity of pentane pluses. In these heavier condensates, it is reasonable to expect that elemental sulphur, various mercaptans, and other sulphur components would be present. With this potential mix and the complexity of sulphur potential mix and the complexity of sulphur chemistry, it is not surprising that a reliable correlation of the copper strip test results with product quality has not been found. product quality has not been found.
In October of 1987, the mixed condensate from Esso Resources Canada Limited's Quirk Creek Gas Plant was found to be failing the copper strip test. This problem occurred soon after restarting the plant following a major turnaround, and at a time plant following a major turnaround, and at a time when the plant was running at high throughputs. Earlier in 1987, a similar incident had occurred but was short lived.