diesel smoke and its control
Page 52
Page 53
If you've noticed an error in this article please click here to report it so we can fix it.
INFORMATION given in a symposium on motor vehicle air pollution control at the Institution of Mechanical Engineers on Monday and Tuesday threw new light on a number of controversial smoke control and measurement problems. The symposium was jointly sponsored by the British Technical Council for the motor and petroleum industries, the Institute of Petroleum and the Automobile Division of the Institution.
In a paper on "The influence of fuel properties on diesel exhaust emissions", Mr. R. Burt and Mr. K. A. Troth of Shell Research Centre said that smoke emission depended to some extent on the molecular structure of the fuel, apart from any additive that it might contain. For example, aromatic hydrocarbons produced more smoke than the paraffins; the effect of the physical properties of a fuel on fuel /air mixing could vary according to the type of engine.
Assessing the merits of different types of additive, the authors stated that alkaline earth metal additives were the most effective smoke depressants. Since barium had the highest atomic weight of the alkaline earth metals, the use of barium compounds enabled the total amount of additive to be minimized for a given smoke reduction, which was important to reducing chamber deposits to a minimum. While alkaline smoke depressants 'did not significantly affect the combustion characteristics of the fuel, they could contribute to deposits in the fuel system, injectors and combustion chamber. And such deposits could cause black smoke by interfering with the fuel spray and air movement.
A graph published in the paper showed that a barium-treated fuel reduced light obscurity of the smoke from 28 per cent to about 17 per cent at a b.m.e.p. of 70 p.s.i. and from 96 per cent to 75 per cent at 105 p.s.i., which the authors claimed was typical.
A new form of "aerosol" pollution was produced by the use of barium additive which could be potentially harmful. Barium compounds that were soluble in water or in the stomach acids were potentially toxic. While barium additives reduced blacksmoke emissions they did not have a similar effect on carbon monoxide emissions.
After stating that fuel properties had little influence on the formation of nitrogen oxides, the authors said that production of these oxides was affected by engine design mainly because of variations in the fuel /air mixing process. Direct-injection engines gave the highest emissions up to 4,000 rpm and pre-chamber engines the lowest. Mr. B. E. Knight and Mr. C. H. T. Wang. of CAV Ltd., Acton, stated that a diesel fuel additive in the form of a bariumcontaining soap in a hydrocarbon carrier fluid reduced exhaust carbon by a "large fraction". They read a paper on "Some experiments on the mode of action of a diesel smoke suppressant additive".
Substantial smoke reductions were obtained in tests of an indirect-injection engine, with variations between 40 per cent and 70 per cent according to the nature of the test. There was a slight increase in exhaust temperature when the additive was used and also an increase in the carbon monoxide and carbon dioxide contents of the exhaust gas. The carbon monoxide content could increase up to 0.7 per cent. A marginal improvement in fuel consumption could be expected and air utilization was increased from 75 per cent to over 90 per cent, which represented a gain of up to 15 p.s.i. in the b.m.e.p. obtainable.
In tests of a direct-injection engine a reduction of about 50 per cent of soot content in the exhaust was achieved throughout the load range. Soot content was measured in milligrams /cu.ft.
Experiments with fumigating the engine, running without an additive, With 18 per cent pure petrol reduced the soot content from around 30 mg /cu.ft. to 20 mg /cu.ft. Additive sprayed in with the petrol was not as effective as additive in the fuel. Cinephotography had shown that the smoke depressant was active in the early part of the combustion.
In a paper on "The nature and cause of diesel emission", B. W. Millington of Ricardo and Co. Engineers (1927) Ltd., Shoreham, Sussex, said that the diesel produced lower levels of objectionable emissions than the petrol engine, other than diesel smoke. These included carbon monoxide, unburnt hydrocarbons and oxides of nitrogen.
Combustion in a diesel tended to be a confused flame condition in contrast to the premixed flame condition of petrol-engine combustion. Consequently the smoke level of the diesel was higher than the level of a petrol engine, and to keep emission within reasonable limits the diesel was always operated at overall fuel /air ratios that were weaker than stoichiometric. This explained why the level of other objectionable emissions was lower.
Carbon monoxide concentration was usually below 0.1 per cent, rising to 1 per cent at full load. If, however, the rating was increased "well into the black smoke condition" it could increase to 2 per cent. Commercial pressures on the oil companies had resulted in the production of diesel fuels with low sulphur contents.
Commenting on the recently awakened interest in oxides of nitrogen, Mr.
Millington pointed out that the reaction that produced oxides was very sensitive to a rise in combustion temperature, which tended to Increase the concentration of oxides. Quoting another authority, Mr. Millington said that there was a progressive increase of nitrogen oxides in diesel exhausts with an increase in load. Above a certain load, the 1 production of oxides was reduced, which (it was supposed) was due to the decreasing availability of oxygen. Retarding the injec tion timing reduced the concentration as did recirculation of exhaust gas to the intake.
Reducing the oxygen concentration in the intake from 21 per cent to 18.5 per cent by recirculation produced an oxide reduction of 45 per cent. This, however, entailed a substantial worsening of the fuel consumption and an increase in smoke unless the engine were considerably derated„ A review of smoke meters was included in a paper "Smoke measurement—instruments and comparisons of methods" presetited by Mr. A. E. Dodd (Motor Indus try Research Association) and Mr. J. Spiers of the Perkins Engine Co. Ltd., Peterborough. In their conclusions, the authors observed that no available meter was ideal for the "actual measurement" of diesel smoke, while some forms were restricted in the type of test for which they could be used.
Defining the nature of smoke and its significance, the authors said there were two classes of matter that made up diesel ex haust smoke. Blue-white components were largely a mixture of fuel and lubricating-oil particles in an unburnt, partly burnt or cracked state, that were sometimes referred to as aerosols. Grey-black components con sisted mainly of solid particles of unburnt carbon residuum from the otherwise complete combustion of fuel. The aerosols mainly appeared during a cold start or when the engine was running light before it had warmed up; they could also be emitted
transiently during a rapid change in engine operation. If they persisted during normal operation under load they indicated that the engine or fuel-injection equipment was malfunctioning.
Grey-black smoke increased at a higher rate above a critical engine loading, which was associated with reduced thermal efficiency and set a limit to power output long before toxic constituents, such as carbon monoxide, formed a serious pro portion of the emission. The objective assessment of grey smoke density and its correlation with carbon concentration was regarded by engineers as an essential function in determining the power rating of an engine, and was the principal reason for requiring an accurate and reliable smoke meter. But it did not follow that such an instrument was a satisfactory enforcement device for the legislator who was mainly concerned with visible appearance. This could be largely influenced by subjective factors. A lack of precision could be expected in smoke measurement because the composition of the exhaust gas was variable and even under conditions of constant speed and load there were momentary and unrepeatable variations in the combustion process.
In observations on the choice of smoke meter the authors pointed out that this depended on the range of tests to be performed. The Hartridge smoke meter and the Bosch EFAW65 meter were "quite versatile" instruments; they could be suitably employed on test beds and on vehicles, but the latter was basically unsuited to tests of a transient nature. -For these a continuousreading instrument or the integrating filter type Bosch meter (the EFAW65B) was essential. Installation problems could arise in mounting the Hartridge meter on a certain type of vehicle, particularly if the tests were to be completed quickly. It was much more cumbersome than the Bosch.
The inherent single-shot feature of the Bosch could, however, prolong the time needed to carry out vehicle tests unless the filter paper magazine were incorporated, and provision were made to allow priming of the sample pump while the vehicle was running at the test speed. Various full-flow smoke meters had been designed to eliminate sampling problems but all were limited in application. While this type of meter was not subject to common types of error it was adversely affected by high temperature. A number of types were not suitable for fitting to a moving vehicle.
The authors stated that it had not been possible to show how the Hartridge meter reacted to aerosols. It was probable that the light scattered by particles of unburnt fuel or lubricating oil would produce an obscuration effect similar to that produced by solid carbon particles. The Hartridge meter had a greater response to aerosols than the Bosch filter types or the UTAC meter.
Commenting on test methods, the authors said that the free-acceleration test was highly inaccurate in most cases; there was no consistent relationship between the maximum steady-speed smoke level and the freeacceleration results. Such tests were disliked by engine manufacturers principally because they were unrealistic. A test involved operation of the engine under conditions that could never be obtained on the road.
For control purposes the steady-state test was unpractical; a controlled-acceleration test provided the most accurate indication of smoke emission at a steady speed. It was preferable to carry out tests on a chassis dynamometer. Mounting a meter on a vehicle could present insuperable difficulties. And other traffic had to be considered.