The Measurement Of Trace Element Emissions At Sour Gas Plant Incinerator Stacks

Abstract

Abstract A method for stack sampling and estimating trace element emission rates from sour gas plant incinerator stacks is described. These stacks are often of large diameter (up to 6.1 m), and the gases are usually low in particulate matter load and of high temperature (420 °C to 680 °C). Results from a survey of 29 elements at ten sour gas plants in Alberta are discussed. Introduction In the past, there have been sporadic reports of the presence of metals in some natural gases, and Leeper has discussed the presence of mercury as a cause of corrosion in LNG plants(1). In Alberta, the question of possible emissions of trace elements from sour gas plant incinerator stacks has arisen on several occasions. Hsu and Hodgson(2) reported that the selenium content in stack gases from 14 Alberta plants ranges from 1.05 to 2.70 g/m3, but give no indication of the sampling method. Available consultants' reports suggested the presence of a variety of metals, but no attempts were made to sample representatively the particulate fraction of the stack gas. Furthermore, the quality of acids used to trap trace elements was not described in terms of trace element concentrations. A search of North American governmental reference methods indicated that a code for measurement of trace element emissions, analogous to the Alberta Environment Reference Method (AERM)(3) for estimating particulate emissions, does not exist. It was, therefore, necessary to design sampling and analytical methodologies which would:collect a representative sample of stack gas,minimize sample loss in and sample contamination by sampling train components,assure quantitative recovery of the sample from the train, andensure adequate analytical results. Sampling Requirements and Method In a steady-state process, the gaseous phase of a stack effluent may be taken as uniform in composition. The problem of collection of a representative sample, therefore, reduces to sampling entrained particulates or liquid droplets representatively (although the latter are not likely to exist under the conditions at hand). For this purpose, the AERM is an established method, whoseessential features could be retained basic sampling train for particulate matter, consistent with the AERM is shown in Figure 1. The objective in the use of this train is to sample isokinetically, that is, to sample in such a manner that the velocity of stack effluent entering the sampling nozzle is equal to the local velocity of the stack gas stream. It is readily shown that the condition for isokineticity is: Equation (1) Available In Full Paper For a given survey, equation (1) reduces to: Equation (2) Available In Full Paper where K is constant for all local velocities in the stack. In the evaluation of K, DN is known from the choice of nozzle size, Cp and Ko are determined by calibration, and Ps, Pm, Md, Ms and Bwo are determined initially from a preliminary, usually non-isokinetic, survey. These parameters are updated after each test.