The function of the pulverising mills is to dry the coal, to reduce the coal from stockpile size to pulverised fuel size, typically 70% passing 75 micrometres (200 mesh), and to transport the pulverised fuel (PF) to the boiler. Coal is continuously fed to the pulverisers from the power station bunkers. Hot air enters the pulverisers and continuously conveys the PF to the boilers.
This transport air is generally known as "primary air". Figure 12 shows a general schematic of a mill. Hot primary air is blown through the mill by the primary air fan while coal is fed in at the required rate by a feeder. The temperature of the air/coal mixture exiting the mill is normally controlled (typical temperature 70 °C) by adjusting the temperature of the inlet air. This is done by balancing the mix of air from the air-heater and ambient air. The wetter the coal fed to the mill, the higher must be the inlet temperature.
Figure 12: General Schematic of a Pulverising Mill
Figure 13: Schematic Layout of a High-Speed Attrition Mill
Figure 14: Schematic Layout of a Medium Speed Vertical Spindle Mill
Figure 15: Schematic Layout of a Low Speed Ball or Tube Mill
Pulveriser Performance
Poor performance of mills can have the following forms:
* Capacity problems because coal is too difficult to grind.
* Capacity problems because coal has low energy content.
* PF fineness problems.
* Mill temperature problems, which could lead to capacity problems or mill fires.
* High maintenance requirements because of abrasive nature of mineral matter in the coal.
Relevant Coal Properties:
The coal properties that have the most impact on pulverising performance are:
Surface Moisture Content
Power stations typically operate pulverisers to achieve a target mill outlet temperature, normally of 70-80˚C. High surface moisture levels cause mill inlet temperatures to rise, so as to maintain the desired mill outlet temperature. Increases in mill inlet temperatures may impose limitations on primary air fan capacity, which in turn impose limitations on achievable boiler load. Additionally, increases in mill inlet temperatures also enhance the risk of fires inside the pulverisers.
Specific Energy
The specific energy of the coal governs the rate (t/h) at which coal must be pulverised to achieve a required boiler load. Low specific energy coals are more likely to cause boiler load limitations to be imposed because maximum pulveriser capacity has been exceeded.
Hardgrove Grindability Index
The Hardgrove Grindability Index (HGI) test is a laboratory procedure designed to predict mill power consumption and/or product fineness. Coals that have lower values of HGI are generally more difficult to grind than coals which have higher HGI values ie. more mill power is required and the PF product is coarser. HGI does not always produce reliable predictions. Coals of different rank having the same value of HGI do
not behave in the same way in the pulveriser. Generally, lower rank coals require less pulveriser power to grind, than higher rank coals of the same HGI, however they generally produce a coarser PF product.
Nature and Concentration of Hard Minerals
The nature and concentration of hard minerals (especially quartz and pyrite) affect the potential of the coal to cause wear of pulveriser components. Estimates of the wear potential of coal are made by the Abrasion Index test, using the apparatus of Yancey, Geer and Price, however Abrasion Index values do no necessarily correlate well with actual mill wear results. The wear potential of a coal is primarily governed by the nature and concentration of hard minerals in the coal, with increases in the angularity, the coarseness and concentration of hard minerals causing a corresponding increase in mill wear potential.
Tendency of Coal to Ignite in Pulveriser
It is generally felt in the utility industry that the tendency of a coal to ignite in pulverisers is governed by the volatile content of the coal. Although studies of mill fires and mill explosions do no necessarily support this association, pulveriser manufacturers generally specify mill air outlet temperature limits based on coal volatile content, with 80-90˚C being specified for medium-volatile coals, and 60-70˚C being specified for high-volatile coals. A more plausible factor that would influence the tendency of a coal to ignite in the pulveriser is the surface moisture content, which influences air inlet temperatures as previously discussed.
Pilot-Scale Results versus Coal Properties
From pilot-scale measurements, an evaluation of a particular coal's performance can be made against the performance of a range of reference coals. Figures 16 and 17 show correlations between HGI and mill performance (fineness and specific power consumption respectively)for a range of coals. These correlations are somewhat scattered, indicating that factors other than HGI play an important part in the grinding process. When power consumption is adjusted for PF fineness (kW.h per tonne of the PF fraction less than 75 mm) the scatter is reduced (Figure 18).
Abrasion Index is designed to predict mill wear but, as Figure 19 shows, the correlation is not strong when the Abrasion Index is high. The explanation may lie in the fact that the Abrasion Index mill is a batch process, whereas the vertical spindle mill has continuous throughput, allowing hard but fine mineral particles to be swept out of the mill.
Figure 16: PF Fineness versus HGI for Pilot-Scale Mill Tests
Figure 17: Mill Power Consumption (kW.h/t coal) versus HGI for Pilot-Scale Mill Tests
Figure 18: Mill Power Consumption (kW.h/t coal) versus HGI for Pilot-Scale Mill Tests after adjusting for the Fraction passing 75 mm
Figure 19: Mill Roll Wear versus Abrasion Index for Pilot-Scale Mill
Coal Heat Release: Flame Behaviour
The primary purpose of the coal/air path is to produce heat from the coal. The ease with which the heat available from the coal is released may be termed “combustion reactivity”; a coal that is easy to burn may be described as “free burning”. One aspect of combustion reactivity is the coal’s ability to produce a flame that is strong, stable, free of pulsation and situated close to the burners. In this way, there is no danger of flame extinction and the rapid heating of the coal provides the maximum time for the coal to burnout in the boiler furnace. A related aspect of flame stability is flame behaviour at part-load: the inherent instability of flames at part-load determines the limits of turn-down capability.
Relevant Coal Properties
The coal properties that have the most impact on boiler performance are:
Moisture Content
Moisture content affects combustion temperatures, with higher moisture contents resulting in lower temperatures.
Volatile Matter (VM) Content
Flame ignition and stability are to a large degree maintained by combustion of the VM, since these gaseous products are able to burn much faster than the residual char. Therefore high VM coals may be expected to display superior flame behaviour.
Pilot-Scale Results versus Coal Properties
When VM is considered to be a significant factor, an alternative index, Fuel Ratio, is often utilised. This is defined as follows:
Fixed Carbon
Fuel Ratio (FR)= ----------------------
Volatile Matter
A low Fuel Ratio corresponds to a high VM content.
In the pilot-scale Boiler Simulation Furnace, burner turn-down capability is measured by operating the burner at full-load (150 kW) as well as a number of part-loads. At each load the required swirl setting to achieve a stable flame is recorded. If a high swirl setting is required this signifies poor turn-down capability. In the limiting case there is simply not enough swirl available and the flame cannot be stabilised. In Figure 20, the coals that required the highest swirl settings tended to have the highest Fuel Ratios (lowest VM contents) as expected, but the correlation is not perfect. The VM determined by proximate analysis is therefore an imperfect predictor of flame stability.
This may be partly because Figure 20 did not take into account moisture content which is also a factor. Another explanation is that the VM released under PF combustion conditions can be much different, usually higher, than the VM yield determined by the proximate analysis. In addition, the calorific value of the volatile matter is more relevant than the quantity released.
* Some coal technologists use VM (% daf) to correlate with coal reactivity data. It is easily shown that there is a one-to-one correspondence between VM (% daf) and FR. For example, a VM of 40% daf corresponds to a Fixed Carbon of 60% daf, which gives a FR of 1.5.
Figure 20: Burner Turndown Capability (from the required swirl to stabilise the flame as firing rate is reduced)
Coal Heat Release: Char Burnout Efficiency
The second aspect of coal combustion reactivity is the burnout efficiency, defined as:
Total Mass of Combustibles Burnt
Burnout Efficiency (%)= ------------------------------------------------ x 100
Total Mass of Combustibles inCoal
The unburnt combustibles (100 – Burnout Efficiency) represent a waste of coal. In addition, the unburnt combustibles remain in the fly ash; too much in the ash renders it unsuitable for use in concrete. The ability to use fly ash in concrete is a convenient way of avoiding disposal costs and environmental effects.
Relevant Coal Properties
The coal properties that have the most impact on boiler performance are:
Volatile Matter (VM) Content
A high VM yield during combustion leaves a smaller percentage of char to be burnt, which is a good start to achieving favourable burnout. In addition, fundamental experimentation has shown that char remaining from high volatile coals tends to be more reactive than char from low volatile coals.
Maceral Analysis and Vitrinite Reflectance
In an effort to remedy to imperfect predictions provided by VM content, researchers have undertaken studies to correlate petrographic characteristics of coal with burnout efficiency.
PF Fineness
Finer PF burns more quickly than coarser PF. Fineness is very much a result of a coal’s inherent milling behaviour.
Pilot-Scale Results versus Coal Properties
Figure 21 shows the correlation between burnout of coals in ht pilot-scale BSF and FR. Though the coals had all been pulverised to the same nominal fineness of 70% passing 75 mm and furnace conditions were the same, there was considerable scatter in the trend. Efforts to include data from the petrographic analysis in a correlation have brought only partial success.
From this we must conclude:
· Char properties (size, shape, porosity, chemical reactivity, variability) may to too complex to be predicted or characterised by petrographics and VM determinations,
· The full size distribution of the PF is never the same for two coals, even though the percentage passing 75 mm may be the same.
Figure 21: Burnout versus Fuel Ratio for Coal Burnt in Pilot-Scale Boiler Simulation Furnace