Sampling and Analysis for Mercury of Steam Collected from Geothermal Wells and Fumaroles
|Authors:||Ed Mroczek and Duncan Graham|
|Keywords:||mercury, sampling, steam, analysis|
|Conference:||World Geothermal Congress||Session:||8. Geochemistry|
|Abstract:||Elemental mercury (Hg0) is the predominant form of mercury found in chemical surveys of wells at New Zealand's geothermal fields. Mercury is an environmental pollutant and in New Zealand, resource consents limit the discharge to atmosphere. Unacceptable variation in analyzed "total" mercury in steam within duplicate samples and between sampling rounds at various geothermal fields has led to a reassessment of the sampling methodologies. However the cause of the problems was poorly understood. It was thought the variability was due primarily to sampling artifacts caused by the high volatility of Hg0 or perhaps to an inadequate sample preservation and digestion prior to analysis.
In this study we evaluate the effect of sampling methodology (sampling vessels) and preservative (permanganate and dichromate) and digestion procedures used on the final result. It appears that sampling methodology has the greatest effect on the variability observed.
The direct collection of steam into evacuated glass ampoules containing acid dichromate preservative with no subsequent hot digestion is the most convenient and simplest procedure for obtaining good analytical agreement between replicates. Samples collected using acid permanganate preservative, which is a stronger oxidizing agent than acid dichromate, with or without hot digestion prior to analysis, also gave comparable results. In this situation dichromate appears to be as equally effective as permanganate for converting Hg0 to a fixed but reducible form. The exception is in the presence of native sulphur,
Ampoule sizes can be increased to accommodate high NCG which would otherwise limit the sample size. Alternatively, where a sufficiently large sample cannot be conveniently collected into an evacuated ampoule, the steam/gas can be passed (natural over-pressure or a pump on the outlet) through a two flask train containing dichromate preservative. This can also give comparable results to samples collected in ampoules but requires much greater care in sample collection.
It is recommended that careful evaluation of sampling and analysis methodologies for mercury in steam, best suited to the steam composition under study, be undertaken prior to the commencement of an ongoing routine sampling and analysis programme. This will save time, effort and may also considerably reduce expensive analytical costs.