Mercury contamination of natural gas is not an issue for most fields. However, when mercury (Hg) is present in production streams, it can cause significant problems if it is not removed.
Problems Caused by Mercury
Mercury in natural gas, condensate and crude oil can adversely affect hydrocarbon production and processing in a variety of ways.
- Hg forms amalgams with a variety of metals, including aluminum, copper, brass, zinc, chromium, iron, and nickel. When these amalgams form with metal components of processing equipment, corrosion of the equipment results, because either the amalgams are weaker than the mercury-free metal (Leeper, 1980; Bingham, 1990), or, as is the case with the aluminum amalgam, the amalgam reacts with moisture to form a metal oxide plus free mercury, which then can continue the corrosion process (Crippen and Chao, 1997). Specific examples include corrosion of aluminum in cryogenic heat exchangers, gates and stems of wellhead valves. Such mercury-induced corrosion of aluminum heat exchangers resulted in catastrophic failure of exchangers at the Skikda LNG plant in Algeria (Kinney, 1975)
- Hg poisoning of catalysts reduces the catalyst life (e.g., Hennico et al., 1991)
- Hg amalgam formation on steel pipe walls can result in classification of production and processing equipment as hazardous waste.
Occurrence of Mercury in Hydrocarbon Production Streams
Hg concentrations in natural gas are typically reported as µg per "normal" cubic meter, where normal (N) indicates standard temperature and pressure. Hg concentrations in natural gas vary from 0 to > 300 µg/Nm3 with some of the highest concentrations occurring in the Indonesian Arun and Dutch Groningen fields (see table below). In condensate, Hg concentrations vary from 10 to 3,000 ppb (Sarrazin et al., 1993). Published concentrations for various natural gases and oils include:
In natural gas, mercury occurs in the metallic form. In contrast, condensate associated with natural gas contains Hg in a variety of forms, including elemental, ionic and organometallic (Sarrazin et al., 1993). As a result, Hg is not limited to a particular boiling fraction of a condensate. For example, Sarrazin et al. (1993) report the following mercury distribution between the boiling ranges of a Southeast Asian condensate:
Geologic Origin of Mercury in Hydrocarbon Accumulations
There is no published, comprehensive investigation of the origin of Hg in hydrocarbon accumulations. Bailey et al. (1961) suggest a hydrothermal origin of mercury in oils from San Joaquin Valley, California, based on the intimate association of hydrothermal mercury ores and the mercury-containing crude oils in that area. The source of mercury in Rotliegand sandstone gas in northern Germany is thought to be underlying volcanic rocks (Zettlitzer et al., 1997). In contrast, Frankiewicz et al. (1998) attribute mercury found in Gulf of Thailand gas and condensate to "coal and carbonaceous shale in or near the producing reservoirs".
Mercury Removal from Natural Gas, Condensates and Oils
The most commonly used method for removing Hg from natural gas streams is chemisorption on sulfur-impregnated carbon (Leeper, 1980; Bingham, 1990). This is the method used at the P. T. Arun LNG plant (Muchlis, 1981). A variety of other methods are also available, including molecular sieves and sulfur-treated zinc oxide (e.g., Table 1 in Bingham, 1990; Mussig and Rothman, 1997; Crippen and Chao, 1997).
Chemisorption on sulfur-impregnated carbon is not effective for treating condensates or crude oils (Leeper, 1980), where other methods must be used (e.g., Hennico et al., 1991). One such method is adsorption onto a solid ion-exchange resin containing chemically bound active -SH groups (a patented process, Duisters and Van Geem, 1990).
Measurement of Mercury in Production Streams
A variety of methods have been described for measuring Hg in natural gas (e.g., Crippen and Chao, 1997; Zettlitzer et al., 1997; Lewis, 1995; Chao and Attari, 1995; Ceccarelli et al., 1993; Bingham, 1990; Muchlis, 1981). Direct measurement at the well head with disposable Drager tubes is effective only for gases with relatively high Hg contents (Muchlis, 1981). Other on-site measurement techniques are also limited in applicability to high-mercury gases (> 1mg/m3; Crippen and Chao, 1997).
To achieve better detection limits, samples must be analyzed off-site. However, mercury must be collected from a gas sample before the gas enters a pipeline, due to potential loss of mercury to the pipeline walls. For example, Grotewold et al. (1979) report a 60% reduction in mercury content (30 µg/Nm3) during gas transit through a 68-mile pipeline. Similarly, accurate Hg measurements cannot be made on stored natural gas samples or on bottomhole samples, since loss of mercury to container walls and small gas sample volumes render such analyses spurious.
In light of these constraints, preferred Hg sampling methods include passing a well-head gas stream through (1) a cartridge containing gold or silver filaments or gold-coated silica beads with which the Hg amalgamates, or (2) a solution from which the mercury precipitates as a mercury salt, or (3) activated charcoal to which the mercury is adsorbed. Adsorbed or precipitated Hg is then shipped from the well to a laboratory for analysis (e.g., by atomic absorption or atomic fluorescence spectroscopy). Mercury concentrations at the ng/m3 level can be measured with these techniques (Crippen and Chao, 1997).
Difficulty in measuring Hg in natural gas results in wide differences in measured values for the same samples. For example, in one case, Dutch and German oil companies reported factor of two differences in Hg concentrations in the same natural gas streams.
For more information on the techniques described here, or to discuss a specific project, e-mail us at email@example.com, or
call us at U.S. (214) 584-9169.
Bailey E. H., Snavely P. D. and White D. E. (1961). Chemical analysis of brines and crude oil, Cymric field, Kern County, California. United States Geological Survey Professional paper D424. In: Short Papers in the Geologic and Hydrologic Sciences, Articles 293-435. Ed.), D306-D309.
Bingham M. K. (1990). Field detection and implications of Mercury in natural gas. SPE Production Engineering May 1990, 120-124.
Ceccarelli, C. M., A. R. Picon, A. F. Paris, I. Romero, R. Rodrigues, L. Nogueiras, and L. Guerra Montilla, 1993, Determination of mercury content in Venezuelan natural gas streams, Proceedings of the Seventy-second GPA annual convention, p. 116-121.
Chao, S., and A. Attari, 1995, Methods developed for detecting hazardous elements in produced gas: Oil & Gas J., Jan 16th issue, p. 41-45.
Crippen, K., and S. Chao, 1997, Mercury in natural gas and current measurement technology: 1997 Gas Quality And Energy Measurement Symposium, February 3-5, Orlando Florida, p. 1-16.
Duisters H. A. M. and Van Geem C. (1990). Process for removing mercury from a non-polar organic medium. United States Patent # 4,950,408 ,
Frankiewicz, T. C., J. A. Curiale, and S. Tussaneyakul, 1998, The geochemistry and environmental control of mercury and arsenic in gas, condensate, and water produced in the Gulf of Thailand: ABSTRACT: Annual AAPG Convention (Salt Lake City, Utah, 5/17-20/98) Extended Abstract No. A209, v. 1, p. 3.
Gijselman P. B. (1991). Presence of mercury in natural gas: An occupational health program. SPE paper #23200: Society of Petroleum Engineers,123-130.
Grotewold G., Fuhrberg H. D. and Philipp W. (1979). Production and processing of nitrogen-rich natural gases from reservoirs in the NE part of the Federal Republic of Germany. In: Proceedings of the 10th World Petroleum Congress, Bucharest, 4, 47-54.
Haselden G. G. (1981). The challenge of LNG in the 1980's. Mech. Eng. 103, 46-53.
Hennico A., Barthel Y., Cosyns J. and Courty P. (1991). Mercury and arsenic removal in the natural gas, refining and petrochemical industries. Oil Gas European Magazine 17, 36-38.
Kinney, G. T., 1975, Skikda LNG plant solving troubles: Oil & Gas J., Sept. 15th issue.
Leeper J. E. (1980). Mercury - LNG's problem. Hydrocarbon Processing, November 1980, 237-240.
Lewis, L., 1995, Measurement of mercury in natural gas streams: Proceedings of the Seventy-fourth GPA Annual Convention (San Antonio), p. 104-108.
Morrison J. (1972). NAM recovers mercury produced with Dutch natural gas. Oil & Gas Journal April 17, 1972, 72-73.
Muchlis M. (1981). Analytical methods for determining small quantities of mercury in natural gas. In: Proc. of the Annual Convention of the Indonesian Petroleum Association, Jakarta (May 25-27, 1981), 401-421.
Mussig, S., and B. Rothman, 1997, Mercury in natural gas - Problems and technical solutions for its removal, SPE #38088: Society of Petroleum Engineers, p. 559-569.
Sarrazin P., Cameron C. J., Barthel Y. and Morrison M. E. (1993). Processes prevent detrimental effects from As and Hg in feedstocks. Oil & Gas Journal January 25, 1993, 86-90.
Situmorang M. S. M. and Muchlis M. (1986). Mercury problems in the Arun LNG Plant. In: Proceedings of the 8th International Gas Union - Liquified Natural Gas International Conference, 1, Session #2, Paper #6, 9 pp.
Zettlitzer, M., H. F. Scholer, R. Eiden, and R. Falter, 1997, Determination of elemental, inorganic and organic mercury in north German gas condensates and formation brines, SPE #37260: Society of Petroleum Engineers, p. 509-516.