Question 7 of 8: Coal Combustion

Question 7: Coal Combustion (10 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Part (a) — why the coal is pulverised. Combustion of a solid fuel is a surface process: oxygen must reach the carbon, and the rate at which it can do so is proportional to the exposed area. Grinding to a powder in which typically 70 per cent passes a 75 micrometre sieve multiplies the surface area of a given mass by three or four orders of magnitude compared with lump coal, and since the volume-to-surface ratio of a particle falls linearly with its diameter, burnout time falls with it. A pulverised particle burns out in one to two seconds, which is short enough to be completed within the residence time available in the furnace of a large boiler, so carbon-in-ash losses stay below about one per cent. Fine grinding also lets the coal be pneumatically conveyed and injected through burners, so that the fuel behaves like a gas: it can be metered continuously, its flow modulated rapidly with load, and it can be mixed intimately with combustion air in a turbulent jet where the flame is stabilised aerodynamically rather than sitting on a grate. That in turn permits the very large single-furnace units used in central-station practice, allows staged combustion for nitrogen-oxide control, and gives a fine ash that can be carried out of the furnace and collected in electrostatic precipitators rather than raked out as clinker. The price paid is the mill drying and grinding power of Question 5, and a fire and explosion hazard in the milling plant.

Part (b) — why excess air is required. Stoichiometric air is the amount that would exactly consume the fuel if every molecule of oxygen met its partner. In a real furnace mixing is imperfect and the residence time is finite, so if only the theoretical air were supplied there would inevitably be pockets running rich; the result is unburned carbon in the ash, carbon monoxide and soot in the gas, and a direct loss of the fuel's heating value. Supplying air in excess raises the local oxygen concentration everywhere, speeds the final stages of burnout when the remaining carbon is scarce, and gives a margin against transient mismatches between fuel and air as load changes. For pulverised coal the usual figure is 15 to 25 per cent excess air, which corresponds to about 3 per cent oxygen in the dry flue gas. Excess air is not free, however: every extra kilogram of air is heated from ambient to the stack temperature and thrown away, so the dry-gas loss rises roughly in proportion, and above about 25 per cent the loss of boiler efficiency outweighs the reduction in unburned fuel. Excess air also promotes thermal nitrogen oxide formation and sulphur trioxide, which raises the acid dew point and threatens cold-end corrosion in the air heater. Boiler operation is therefore a compromise, and the optimum is found by trimming the excess air until the sum of the dry-gas loss and the unburned-carbon loss is a minimum. To fix the magnitudes: burning pure carbon needs $32/12\div 0.2315=11.5$ kg of air per kilogram of carbon, so at 20 per cent excess the requirement is 13.8 kg/kg, and the resulting dry flue gas carries about 17 per cent carbon dioxide by volume.

Part (c) — proximate analysis. A proximate analysis is a standardised gravimetric test that divides the coal into four operational fractions without identifying any chemical element. A weighed sample is dried at about 105 °C and the mass lost is the moisture; it is then heated in a covered crucible to about 900 °C out of contact with air and the further mass lost is the volatile matter, the tars and gases driven off by pyrolysis; the residue is burned off in air and the mass consumed is the fixed carbon; what remains is the incombustible ash. The calorific value and often the sulphur are reported alongside. The test is cheap, quick and directly useful to the plant: the volatile matter governs ignition and flame stability and therefore burner design, the fixed carbon governs burnout time and furnace sizing, the moisture sets the mill drying duty computed in Question 5 and part of the boiler loss, and the ash sets the ash-handling and precipitator duty and the slagging behaviour of the furnace. Because the fractions are defined by the test procedure rather than by chemistry, results are only comparable when the same standard is used, and the basis — as received, air dried, dry, or dry ash-free — must always be stated.

Part (d) — ultimate analysis. An ultimate analysis is a true elemental analysis, reporting the mass fractions of carbon, hydrogen, nitrogen, sulphur and oxygen together with ash and moisture, the oxygen usually obtained by difference. Carbon, hydrogen and sulphur are determined by combusting a sample and measuring the carbon dioxide, water and sulphur oxides produced; nitrogen by a Kjeldahl or thermal-conductivity method. This is the analysis that combustion calculations actually require, because stoichiometry is done on elements: the theoretical air, the mass and composition of the flue gas, the carbon dioxide emission computed in Question 3, and the sulphur dioxide load on the scrubber all follow directly from it. It also supports calorific-value correlations such as Dulong's formula. It is slower and more expensive than a proximate analysis, so plants run proximate analyses routinely on every shipment and ultimate analyses occasionally, or on the design coal.

Part (e) — flue gas analysis, wet and dry. A flue gas analysis reports the composition of the products of combustion, conventionally by volume, and normally covers carbon dioxide, oxygen, carbon monoxide and nitrogen, with water vapour, sulphur dioxide and the oxides of nitrogen added when required. On plant it is measured continuously by in-situ zirconia oxygen probes and infra-red analysers, and historically by the Orsat apparatus, in which a measured gas sample is passed over successive absorbents — potassium hydroxide for carbon dioxide, alkaline pyrogallol for oxygen, cuprous chloride for carbon monoxide — and the contraction after each absorption gives that constituent's fraction.

The distinction between the two bases is simply whether the water vapour formed by burning the fuel's hydrogen and evaporating its moisture is counted. A wet analysis includes it and therefore describes the actual gas leaving the boiler; a dry analysis excludes it, the percentages being renormalised over the remaining constituents. The difference is not small: burning a bituminous coal produces a flue gas roughly 6 to 8 per cent water vapour by volume, and for natural gas nearly 19 per cent, so the carbon dioxide reading on a dry basis is correspondingly higher than on a wet one.

Which basis to use is decided by the measurement and by the calculation. Any sampled analysis is inherently dry, because the sample line and the analyser are below the acid and water dew points and the vapour condenses out before the gas is measured; the Orsat is dry for the same reason. Dry-basis figures are therefore what combustion calculations start from — excess air from the dry oxygen or carbon dioxide reading, the air-to-fuel ratio from a carbon balance, and the combustion efficiency trim on a running boiler — and they have the further virtue of being independent of the fuel's moisture, so readings can be compared across coals and across days. A wet analysis is needed whenever the water itself matters: computing the flue gas enthalpy and hence the boiler efficiency by the losses method, sizing the induced-draught fan and ducts on actual volume, finding the acid and water dew points to set the air-heater cold-end and stack metal temperatures, and evaluating heat recovery in a condensing economiser. The practical rule is to measure dry, then add the water back from the hydrogen and moisture in the ultimate analysis when the calculation calls for it.