Technical Market Support » Thermal Coal
Mercury is a naturally occurring metal and there is ample evidence that mercury emitted in industrial areas is transported globally. The UNEP Global Mercury Assessment concluded that there was sufficient evidence of global adverse impacts and that these warranted initiation of actions to reduce human generated mercury releases. Combustion of coal for power and heating is believed to be the largest contributor to ongoing anthropogenic mercury release to the atmosphere. Australian emissions of mercury from the fossil fuel electricity generation sector reported to the National Pollution Inventory (NPI) ranged from 970-1900kg in the years 1998-2007.
US studies have demonstrated that mercury capture is enhanced by the presence of a sorbent in the waste gas stream and an effective method to collect the sorbent. Unburnt carbon in the flue gas and chlorides in the coal both appear to enhance the potential for mercury capture in air pollution control devices such as electrostatic precipitators and fabric filters, although this is disputed by other workers.
A series of measurements of flue gas mercury concentrations were undertaken at seven different power stations on nine occasions. At two of these stations, Wallerawang and Mt Piper, continuous measurements of total and elemental mercury were made over an extended period of around seven days each. At all power stations batch measurements of speciated flue gas mercury emissions (total, oxidised and particulate mercury) using the Ontario-Hydro technique (O-H) were undertaken.
Chemical analyses of the feed coals to the power stations showed them to have mercury contents ranging from 18-52 μg/kg (ppb) which is low by international standards but in the range found in studies of Australian export coals. Chlorine levels of the feed coals were also low compared to coals used for power generation in North America and ranged from 250- 320 mg/kg.
Similar flue gas concentration results were achieved from both the continuous mercury monitoring and batch O-H measurements. Concentrations ranged from 0.5-5.6 μg/m3 and are generally low when compared with international measurements. Levels of mercury oxidation varied considerably at the different plants, ranging from 40 to 84%, indicating that at some plants, with high levels of oxidation, mercury capture might be facilitated were the mercury exposed to a suitable sorbent. That high levels of mercury capture were generally not achieved may be a consequence of both the low levels unburnt carbon in the ash (0.6 - 2.7%) and low chlorine availability (37-400 ppm). No relationship could be established between coal feed chlorine and mercury oxidation level or mercury emission rates.
Comparison of measured values and estimated mercury emissions using factors in the National Pollutant Inventory Emission Estimation Manual showed that there were often differences between the estimated and measured values. These differences ranged from an overestimation (ratio Measured/NPI) of 0.6 to an underestimation (ratio Measured/NPI) of 3.6. There was considerable variation apparent even within measurements made on the same plant.
The project points to the need for further on-plant measurements of mercury emissions if a more accurate estimation of mercury emitted by the Australian electricity generation sector is to be made. Outcomes from the project demonstrate that while it is possible to measure speciated mercury continuously from plants running under Australian conditions, the data obtained might not justify the expense of continuous measurements. An alternative and potentially more economical approach might be to measure mercury emissions more frequently by either the mass balance (solids chemical analysis) or O-H methods.