With increasing concern over the emission of hazardous air pollutants, measures are being considered to limit their discharge. While proposed legislation is aimed mainly at the power generators, because of their potentially large mass emissions, specific installations such as incinerators are also being targetted.
Although most international particulate legislation is aimed at reducing the overall mass of material, the World Health Organization and other bodies have been promulgating a push to control specific species, particularly the heavy metals- for example, lead, antimony, nickel, chromium, cadmium, mercury, etc. – which are usually present as trace impurities in the feed stock.
These heavy metals are initially volatilized in the high temperature zone of the process and upon reaching the cooler zones some of the volatiles are condensed from the gas stream as submicron sized particles. In this size range, the particles are respirable and can pass from the lungs to the blood stream; and being accumulative can lead to various health problems.
Legislation already exists governing the emission of heavy metals, etc., from various types of incineration process and regulations controlling the emission of pM 2.5 material are proposed for the USA. Currently the US proposal, covering utility installations, is to limit the average emission of pM 2.5 micron particles to 0.025 mg/Nm3.
For coal, the quantity of material which can be classified as hazardous, is usually very small,
At the high temperatures within the combustion zone, these materials, having a melting point much lower than the silica and alumina components, tend to be volatilized, initially into a gaseous phase and later, to condense in the cooler regions of the process plant.
Many compounds can be found in particulate form at the normal process back-end temperatures, or the gases can be further cooled to condense additional material. Mercury, however, because of its low melting point, -40°C (-40°F), can exist in a number of different forms, for example, solid particulate through to gaseous elemental mercury. Consequently all forms and states must be accurately assessed and measured, at both the inlet and outlet of the plant, if stringent emission levels are to be met.
Analytical determinations made on fly ash samples, isokinetically collected at the inlet and outlet of power plant electrostatic precipitators designed for a total mass removal efficiency of some 99.5 per cent, have indicated that the particulate heavy metals can be collected and removed from the gas stream at efficiency levels exceeding 95 per cent. This data, although not considering any gaseous phase materials, indicates that with prior knowledge of the quantity of materials present in the coal or feed stock, the current particulate technology could be used to obtain the required emission levels.
Where some of the components may be present in a gaseous phase, then by further cooling of the waste gas stream it may be possible to condense sufficient particulate material to satisfy the required emission levels.
If the gas stream needs to be cooled to approach the dew point temperature, then a wet rather than a dry precipitator can be effectively used to collect the finer particulates. An advantage of the wet precipitator following a wet scrubber used for desulphurization purposes, is that any sulphuric acid mist formed as a result of dropping below acid dew point temperature will be readily collected. The d50 size of sulphuric acid mist is around 0.4 micron, so its removal will greatly assist in reducing the pM 2.5 micron fraction.
Another approach for cooling the waste gases has been investigated, whereby the collectors of a conventional dry outlet field have been changed to water-cooled plates, thereby enabling the cooled gas to be maintained above any potential dew point temperature.
An alternative system is where a conventional dry outlet field has been converted to a wet field by the addition of water injection sprays in front of and above the field area. Provided the water is injected in an atomized form, there is little likelihood of electrical breakdown between the electrode system and the overall precipitation efficiency. In all these approaches, corrosion resistant materials must be used as there is a risk of acid dew point corrosion, and in the latter case, water treatment must be considered.
Until fairly recently, electrostatic precipitators have been the preferred, cost effective method of particulate removal from most coal fired plants. With the demands of reducing sulphur dioxide emissions, rather than fitting acid gas removal equipment, there has been a move towards firing low sulphur fuels.
This type of fuel can produce a high resistivity fly ash giving rise to precipitation difficulties, which demands a large and costly precipitator to satisfy the target efficiency. To counter the problem of large precipitators, some installations have been retrofitted and others designed with bag filters, as the preferred method of particulate collection.
The successful precipitation of particulates within an electric field is dependent on the particles receiving sufficient charge so that migration across the field can occur. The amount of charge that can be placed on a particle is proportional to its surface area, so that the larger particles are more readily precipitated than the smaller ones.
The charging mechanism for particles greater than 2 micron diameter is by collision, but as the particle size reduces, the charging mechanism alters to induction or diffusion charging as the ions pass close to, but do not collide with, the particles.
For extremely small particles, Brownian motion assists in both charging and migration phenomena and high collection efficiencies are obtainable. The effect of the Brownian motion results in an increase in the fractional particle size efficiency relationship below 0.5 micron diameter. The effect of this relationship is to preferentially collect the inert or larger particles in the inlet fields and the finer materials towards the rear of the precipitator. This means that if a higher efficiency is required to capture additional finer sized hazardous materials, then an extra precipitation field could be added to an existing plant, or built into a new plant if sufficient data is known beforehand.
In the case of bag filters/fabric filtration, particulates are collected by impaction and interception with the fibres forming the filtering media. Although on a clean woven filter the pore size between adjacent fibres is generally much larger than the finer particles, as the bigger particles are collected, these form a layer through which the gas must pass. Consequently, with time, the finer particles are collected on the deposited layer of particles rather than on individual fibres. Even with felted and some surface treated media there is always a risk of fine particle penetration when the filter is clean and some time must elapse before the optimum filtering is re-established.
In operation, to minimize the increasing pressure drop across the media, it is essential that the deposited layer is adequately removed to ensure that the gas flow through the system is not compromised. For most large installations the preferred method of removal is by reverse gas flow, or by injecting short duration pulses of air, which is blown through the media in the opposite direction to normal.
Regardless of which system of cleaning is used immediately following removal of the deposited layer, there is a risk of fine particle penetration, which may compromise the fine particle emission requirement. To meet the more stringent regulations concerning heavy metals, the manufacturers of filter media have developed improved materials and finishes to minimize fine particle penetration. In addition to special surface finishes, such as polyurethane and terpolymer coatings, there has been a trend towards the use of microporous expanded PTFE membranes such as Gore-Tex.
The latest legislation in the incineration field, demanding more stringent emission levels for both particulates and acid gases, means that many installations are now designed with bag filters, rather than electrostatic precipitators, as the most cost effective particulate collection system. The reasons for this choice rests with the overall improved absorbent stoichiometry of the dry or semi-dry acid gas scrubbing system, coupled with the need to collect activated carbon used to absorb specific particulate gaseous species, particularly mercury, to comply with legislation.
Mercury, dioxins and furans
These materials are of increasing concern to the WHO and other agencies because of their perceived effect on health. Mercury can reportedly lead to blindness, muscle deterioration and possible birth defects; and because of its low melting point can exist in waste gas in either solid, liquid or gaseous form depending on the gas temperature.
Dioxins and furans are understood to be possibly carcinogenic and can lead to skin disease and liver damage and are regulated in Europe to a discharge limit of 0.1 mg/Nm3.
These are mainly associated with waste incineration processes, where the gas conditions, e.g. temperature and presence of chlorine and carbon act as precursors to the formation of these
Typical figures for uncontrolled mercury emission from utility plants firing eastern US coals is 5-30 mg/dNm3, which is very much less than the regulated emission figure of 80 mg/dNm3 for MSW incinerators. From measurements carried out on several US utility plants, the ratio of oxidized mercury to elemental mercury is dependent on the fuel type and the presence of chlorine in the coal. For bituminous coals, up to 94 per cent of the mercury has been found in an oxidized form, whereas for sub-bituminous coals the ratio falls to around 30 per cent.
The determination of mercury is difficult and for accuracy the recommended approach is based on the “Ontario Hydro” method rather than EPA 29, which has shown a tendency to oxidize elemental mercury in the absorption train.
While both the electrostatic precipitator and bag filter have been shown to be relatively effective in removing oxidized mercury in particulate form, neither can capture gaseous phase material. Measurements for mercury across a wet scrubber, used for desulphurization on a utility plant, have shown that mercuric chloride (oxidized mercury) can be collected at relatively high efficiencies, whereas it had little effect on the collection of gaseous mercury.
Over the past decade, to collect gaseous phase mercury and other gaseous phase materials, activated carbon has been injected into the gas stream ahead of the collection device. Activated carbon, dependent on its particle size and reactivity, has been shown to be very effective and is now one of the preferred methods adopted by the incineration industry to meet the prescribed emission limits for mercury, dioxins and furans.
No commercial utility plants, as far as it is known, have adopted an activated carbon injection approach for mercury control, mainly because of the large gas flows and hence quantities of activated carbon required.
For power plants employing bag filters, because of the retention of the activated carbon on the media and therefore longer gas exposure times, this approach could be adopted to meet any proposed regulations, but further investigative work would be required to optimize injection rates and absorption coefficients.
As an alternative to carbon injection, finely ground limestone has been injected into the furnace area of a unit firing eastern US coals producing an untreated mercury emission of 25 mg/Nm3. Although limestone was initially injected to reduce sulphur dioxide emissions, tests showed that it was quite successful in reducing specific mercury emissions.
At a Ca/S molar ratio of 2.0, 74 per cent of the mercury was captured by a downstream electrostatic precipitator. At lower Ca/S injection molar ratios of 0.35 and 0.04, the mercury capture fell to 56 per cent and 46 per cent, respectively. These figures relate to the capture of oxidized mercury and compare to an untreated mercury capture by the precipitator of around 18 per cent.
In Europe, in addition to the injection of activated carbon into the gas stream, carbon beds have been used for mercury control. In these, the carbon is contained either in a fixed bed or slowly descending bed arrangement through which the pre-cleaned waste gases are passed.
While this approach can effectively remove the gaseous mercury from the gas stream, one is left with a highly contaminated carbon residue, which requires special disposal techniques to avoid the release of the captured mercury.
Measurements on utility plant precipitators have shown that particulate heavy metal components, generally classified as HAPs, can be reduced by at least 95 per cent on a correctly sized and operated precipitator. If the regulated emission level should demand the collection of vapour phase material, then additional gas cooling or wet precipitation may be required.
Investigations using carbon injection in front of utility electrostatic precipitators has shown promise, but additional work is required to establish injection rates, since operating conditions and the presence of acid gases etc., can impact on the absorption efficiency of the activated carbon.
Investigations using limestone furnace injection as part of a desulphurizing process on a utility plant using electrostatic precipitators, has indicated that mercury oxidation can occur which makes capture easier. This approach may be more cost effective than using activated carbon injection, but depends on the final degree of mercury removal required, since only oxidized mercury is captured.
Additional cooling and calcium carbonate injection would also apply to plants employing bag filters/fabric filtration, since satisfactory removal can only be obtained with particulate material. Even then there is a risk of penetration of the finer heavy metals following cleaning through some filtration media.
In the incineration industry, many operational units employ activated carbon injection ahead of the final particulate removal stage to capture gaseous phase material, which is absorbed and collected as part of the injected material. In addition to mercury, dioxins and furans show significant removal efficiencies being obtained using this approach.
The pollution control industry has reacted to meet ever-decreasing emission levels and while some of the existing control technologies can be used, it is likely that operational changes and plant modifications will be necessary to minimize both capital and operating costs for either the precipitator or bag filter.
Perhaps the greatest problem facing industry will be the accurate measurement of the untreated emission of the various materials in order to determine the optimum removal efficiency required and hence plant costs to satisfy any possible legislation.