The Engineer, March 1st 1907. The Generation of Power at Neasden

Through the kindness of Mr. Thomas Parker, consulting engineer for the electrification of the Metropolitan Railway, under whose guidance all the arrangements at Neasden have been effected, we are enabled to state the technical and economic results of the working of this remarkable power station and to describe the methods whereby these results are obtained.

The plan of the power house is a rectangle about 325ft. long by 102ft. wide, with an excrescence 70ft. by 114ft. on its south side containing economisers and chimney, and with the condensers and cooling towers standing also outside these limits on the north side.

There is also a lake of 2½ acres area and 7 million gallons capacity, and two artesian wells yielding respectively 32 and 14 thousand gallons an hour of clean water. The cooling powers of the lake are not drawn upon except under stress of exceptional load, because the scour of the water flow brings with it a little troublesome mud, which is undesirable in the circulating pumps. The boiler feed is wholly from the wells.

Neasden Power Station. It opened in December 1904, and closed in 1968. It was later demolished.

The coal is brought by railway sidings to the south west corner of the building where the wagons are tipped direct into a large hopper 16ft. square. From this it drops through crushing rolls when of too large size for mechanical stoking, which rolls are not, however, often in use. From the rolls a shoot takes it straight on to the buckets of a great chain-and-bucket conveyor, the buckets each carrying 28 lb. and being spaced 2ft. apart. This conveyor travels at speeds up to 50ft. per minute and its full carrying capacity is about 20 tons per hour.

This conveyor ascends vertically between 50ft. and 60ft. to the top of the ‘hopper tower,’ then travels horizontally the full length of the building, distributing the coal into the bunkers over the boilers, which bunkers have a combined capacity of 2,000 tons; descends vertically at the east end of the building; and returns horizontally in front of the ash doors. The conveyor is used alternately for feeding coal to the bunkers and for the removal of ash. The ash is raked by hand direct into the buckets of the conveyor which in turn empty themselves into an ash bin situated in the ‘hopper tower’ over the coal truck siding. This conveyor chain, which has a total length of about 730ft. has worked satisfactorily with only minor repairs since the start.

From the bunkers the coal descends by automatic weighing shoots to the mechanical stokers, which are driven by eccentrics on a longitudinal shaft with an electro-motor on each end of it. The shoots drop the fuel evenly on travelling chain grates driven by the same shaft. Roney  grates were fitted in the first place, but were replaced by chain-grates on account of their being unsuitable for the kind of coal found to be most economical, which is North Staffordshire double-screened nuts. Preparations are now in hand for a large extension of this building and of the plant it contains.

The four turbo-generator sets at present installed are served by fourteen water-tube boilers designed for 200 lb. pressure and worked at 180 lb. to 185 lb. with 180 deg. Fah. superheat. They were hydraulically tested to 300 lb. per square inch for one hour. Each has 260 tubes, 18ft. long by 4in. diameter, and two drums 23½ft. long by 4½ft. diameter. This gives each boiler 5730 square feet of heating surface, and its grate area is 100 square feet. Their guaranteed evaporative power is 20,000 lb. per hour each, with 25 per cent. overload capacity; this corresponding to 3½ to 4⅜ lb. per square foot heating surface. The superheater of each boiler contains 128 tubes, l½in. in diameter, giving 894 square feet of external heating surface. Four banks of economisers serve the fourteen boilers with a total of 1760 tubes. These deliver the feed at an average temperature of about 275 deg. Fah. The feed is brought to them from the hot well through feed heaters, which raise the temperature about 100 deg. Fah. by two compound vertical double acting steam pumps, each of which is equal to supplying three turbines at full load.

Each turbine is served by a separate barometric condenser, the rising exhaust main being 54in. in diameter and 34ft. in height. Each is designed to condense normally 66,500 lb. of steam per hour, and will deal satisfactorily with about 60 per cent. in excess of this. The remainder steam, that is to say, the steam that is not carried away as condensed water with the circulating water, and the air are drawn off by a steam-driven vacuum pump, its two pump barrels being 24in. in diameter by 24in. stroke, while the one steam cylinder is 10in. diameter by 24in. stroke, all three being arranged tandem.

This pump runs at 100 double strokes per minute. and consumes 55 horsepower. No auxiliary jet ejector condensers are employed. The circulating pumps are centrifugal, one 18in. pump to each turbine, delivering 250,000 gallons per hour, or not much short of forty times the weight of steam to be condensed. There is also a 16in. hot water pump to each turbine.

It is to be noted that the range of boilers is not ring connected. One header connects the whole set of fourteen. The main steam pipes to each turbine from this header are 12in. in diameter. The steam pipes are of steel, and exhaust pipes of cast iron.

There are three duplex cooling towers, each essentially composed of an immense number of timber slats, which baffle and spray the water as it falls through a height of about 20ft. Each can cool 400,000 gallons per hour down to at least 85 deg. Fah. The hot water, after being pumped to the heads of these towers, flows in a number of streams along wooden troughs perforated with a large number of holes through which it falls upon the topmost tiers of slats. These towers are substantial well-braced timber structures, which have withstood without any wracking or other injury the winds of a somewhat exposed and elevated locality. The hot well, from which the feed is taken direct to the economisers, and from which also the bulk of its contents, namely, the circulating water, is pumped to the cooling towers, in a wide concreted trench running along, and outside the north wall of the station.

The four turbines, of Westinghouse-Parsons pattern, are duplex, each of 3,500 kilowatt normal power, and capable of 50 per cent. overload for one hour. There is a daily momentary peak in the use of each of them of about 100 percent overload. They run at 1,000 revolutions per minute, with a permitted variation of 2½ percent; but 1½ percent variation is very seldom exceeded. The steam consumption is I7 lb. per kilowatt-hour at full load and 20¼ lb. at half load. The vacuum maintained is about 1¼ lb. per square inch absolute. The variations in load are naturally heavy during daytime running from about 35 percent to 175 percent of normal, and after 7 p.m. from little over 20 percent up to 100 percent of normal. The governor, which is of the well known oscillatory type, is electrically controlled, and the above figures prove that it performs its function or maintaining approximate constancy of speed very satisfactorily.

The gust admission of steam, which is the peculiarity of this method of governing, is practically the only cause of noise in the Neasden station. The bearings are continuously water cooled and are practically without appreciable tremor, while the pedestals are always quite cool. The level of the lubricating oil tank gives an available oil pressure of 15 lb. per square inch but not more than 10 lb. is found to be necessary. The oil is cooled and filtered and returned to the overhead tank by a steam-driven pump. The consumption of lubricating oil in the whole station for the four turbo-generators, the exciters, alternator for donkey work, pumps etc. is under 100 gallons per month. From each end of the casing of the turbine, which is 11ft. in diameter by 13ft. long, the exhaust is led off by a 40 inch pipe, the two branches joining in one 54 inch uptake to the condenser.

The alternators are three phase, with 11,000 volts in each phase on a non inductive load. The normal speed of 1,000 revolutions per minute gives a frequency of 33⅓ cycles per second, the machines being four-poled. They were tested with 20,000 volts across the armature for half an hour; with a 30,000 volts flash test; and with a dead short circuit under full excitation maintained for one minute. There are three 100 kilowatt exciters installed. A small 100 kilowatt alternator is run to supply power to various small motors and for lighting the station and yards.

The switchboards occupy three galleries in the west end of the generator house. The bus-bars are on the ground and first floors; the three-pole high tension oil switches on the first floor; and the master control board upon the top floor. There are two sets of bus-bars. Each generator and its main switch can be isolated from the bus-bars; and each feeder can be connected to either bus-bar, or to both, or be isolated.

The feeders are three core, paper insulated, with a lead sheath, and steel armoured. They are tested under water to 30,000 volts, and again to 22,000 volts after being laid in place. Four sizes are used; of copper sections 0.1, 0.15, 0.2, and 0.25 square inch. Three of the largest size lead to Baker Street sub station, which is the distributing centre for the Circle sub stations, and two each of smaller size lead from Neasden to Harrow and Ruislip.

There are in all nine sub stations. Most of these have in each three 800 kilowatt rotary converters; but Charlton Street and Moorgate Street stations have each three 1,200 kilowatt, and Baker Street station four 1,200 kilowatt rotaries. In all there are twenty eight sub station rotaries, of a total normal capacity 25,200 kilowatts. These run at 375 revolutions per minute, and with 550 to 600 volts between the outgoing terminals. Each can carry a 25 percent overload maintained for twenty four hours with 50 deg. Cent. rise of temperature, or a 50 percent overload for one hour with 60 deg. Cent rise. Each rotary converter is served by three static transformers reducing from 11,000 to about 440 volts.

The conductor rails are of soft steel of a conductivity equal to one seventh that of copper. They are 100 lb. per yard in weight, or 10 square inches in section. They are copper-bonded, with 1½ square inch copper section. The insulators, placed 9ft. apart, are of highly vitrified porcelain, and have plain cylindrical bodies, which are 5½ inches in diameter by 4½ inches deep for the positive rail, and 3 inches depth with the same diameter for the negative or return rail.

The present rolling stock comprises twenty eight trains, each of a total weight of 180 tons plus the passenger load, the passenger seated capacity being 350. The motor coaches weight 40 tons and the trailers 25 tons each. Each full train consists of two motor coaches and four trailers.

The motors are rated at 150 horsepower, and there are four of these on each motor coach, giving 1,200 horsepower to the full train. This yields an acceleration of 1½ft. per second per second, or 30 miles per hour in about half a minute. There are in addition ten trains of a total weight 185 tons each, the motor equipment being some. what heavier in these. The turret controller which was at first used has been substituted by a rectangular form, which weighs considerably less and is also less bulky. Besides these there are ten locomotives of 1,000 rated horsepower each in four motors. These weigh from 45 to 50 tons, and are capable of drawing 250 tons up a gradient of 1 in 40 at 10 miles per hour.

In November last; complete systematic tests of the whole generating plant were made. It is unnecessary to give here all the details of these tests. Only the main results need be mentioned. The boiler test lasted six hours, the gauge pressure being 188 lb. per square inch, and the superheat 150 deg. Fah. The feed reached the economisers at 160 deg. Fah. and left them at 253 deg. Fah. The water actually evaporated was 3¼ Ib. per hour per square foot of heating surface, the coal consumed being 23½ lb. per hour per square foot grate area.

This corresponds to a steaming power of 7.94 lb. actual superheated steam per lb. of coal burnt. Apart from the action of the feed heater and the economiser, which together raise the feed to 253 deg. Fah. the boiler proper and its superheater have, therefore, an evaporative efficiency of 9.0 lb. water from and at 212 deg. Fah. per lb. of coa1 burnt. (7 .94 x 1095.6/965.7 = 9.0). Taking the boiler along with the feed heater and economiser, the whole plant has a heating efficiency of 1263.6/965.7  x 7.94 = 10.4 lb. from and at 212 deg. Fah. per lb. of coal. This means 84½ percent efficiency, since the coal used has 11.980 as its calorific value. The percentage efficiency of the boiler and superheater separately is 73, deducting 12 percent of moisture contained in the fuel, the consumption of dry coal per kilowatt-hour was 2.4 lb. Considering that there was 8½ percent of ash as well as 12 percent of moisture in the fuel, these results are very good.

The turbo alternator test was carried out with 184½ lb. steam gauge pressure and 150 deg. Fah. superheat, the average vacuum pressure being 1¼ lb. per square inch absolute. During a six hour full-load" run the actual variation of load was from 1,200 to 6,000 kilowatts and averaged 3,850 kilowatts. During a subsequent two hour half load run it varied from 800 to 3,400 kilowatts, and averaged 2,175 kilowatts. The actual steam consumption per kilowatt-hour was 18.3 at full, and 22.2 at half power; but with the allowance made on account of the above large variation and for other reasons, the full load consumption is within the guaranteed 17 lb. per kilowatt-hour. If 85 percent be taken as the mechanical and electrical efficiency of the combined turbine and alternator, this figure of 17 means 14½ lb. per kilowatt-hour, or 10.8 lb. per horsepower hour of mechanical work done by the steam on the turbine blades.

During the test of No.4 turbine the condensing water passed over the measuring weir was 3¼ million pounds per hour, and it was raised in temperature from 60¾ deg. to 84¼ deg. Fah. in passing through the condenser. This condensing water was 46½ times the weight of steam it condensed. Since 84¼ - 60¾ = 23½ and 23½ x 46½ = 1060, this last is the units of heat extracted from each pound of exhaust steam. Adding 84 - 32 = 52 to this, we find 1,112 British thermal units as the total heat from 32 deg. Fah. of the condensed steam, which corresponds to slightly under l lb, per square inch absolute pressure in dry saturated steam. Some of the steam, however, is withdrawn by the vacuum pumps without mixing with the above water, and it is also not certain that the exhaust is quite dry steam. The exhaust also loses heat, of course, from the surface of the large 54in. uptake pipe, and these items amply account for the difference between this 1 lb. pressure and the actually measured l¼ lb. per square inch, the latter corresponding to a total heat of 1,115, or an excess of three units only above that resulting from direct measurement. This small difference points to great accuracy in the measurements made in these tests.

Table 1. gives a complete return of the coal consumption and power generated and work done for each week during the last four months of 1906. The chief contents of this table for three months are also given in diagram form in Figs. 1, 2, 3, and 4.

For the purpose of obtaining a definite figure for train-miles all the trains have been reduced to equivalent trains of 170 tons weight, although many of the trains are three-car or ‘half’ trains, and also the real weight of a full train approximates more nearly to l80 tons than to 170 tons. In calculating the ton-mileage the passenger load has been neglected. The seating capacity gives a full passenger load of about 20 tons, and seeing that during certain hours there is a great deal of strap-hanging, 10 tons of passengers per full train is probably an under estimate of the average. This is 6 percent of 170 tons, and this ought to be remembered in considering the other figures of the table. For instance the average of the column giving energy from sub stations per ton-mile is 74 watt-hours, but reducing this by 6 percent, the real figure, taking into account passenger load, is only 70 watt-hours.

During these four months the ton-mileage has steadily increased from under 5 to over 6½ millions. This is also shown by a diagram line on Fig. 1. The only deviation from regularity is the great drop in the week ending 29th December. This is due to the withdrawal of a considerable number of ‘full’ trains, and the substitution of ‘half’ trains for them. Experiments in this direction were made at the end of October, and the disturbances thus caused at these two particular periods account for the only marked irregularities in the curves of the diagrams in Figs. 1 and 2. Fig. 2 gives the costs per ton-mile, and apart from the above-mentioned disturbances no cost curves could be more satisfactorily steady. Fig. 4 gives the cost per unit generated in the power house. Here the coal cost keeps perfectly uniform, and wages and consequently the total costs exhibit a slow steady decrease.

During the 24 hours there is on the average one unit at work during 1.85 hours, two units during 13.75 hours, and three during 5.15 hours, The seventh column in Table 1. gives the turbine unit-hours used per week. Near the end of the year this was about 280, or 40 per day. Averaged over the twenty-four hours, this means 1⅔ turbines at work continuously. Averaged in the same way, the output of each is somewhat over 2,000 kilowatts, or about 60 percent of their rated power. This rate of generation of 280 turbine-hours per week corresponds to about 590,000 B.T. units generated, and to 1,100 tons of coal consumed, costing £600, or between 0.24d. and 0.25d. per B.T. unit. This coal consumption includes all banking-up time and losses on uneconomically small loads. The output from the sub stations is between 82 and 83 percent of that generated. Of this latter output between 9 and 10 percent is spent on lighting, machine driving etc. etc. Thus about 75 percent of the power generated is used for actual traction work upon the track. It is this last consumption that is dealt with in columns 11 and 13, headed ‘Consumption per train-mile and per ton-mile from sub stations.’ Per train-mile of 170 ton trains the average is 12,800, and per ton-mile 74, watt-hours. Per ton-mile ⅜ lb. of coal are used, costing 0.23d. The cost in coal consumed varies very uniformly in proportion to the output, while the other costs in this station are phenomenally low, being only about 0.008d. for labour, and 0.0005d. for lubricant and stores. The total cost thus comes to only 0.0315d. on the average per ton-mile. Per kilowatt-hour generated the average figures are: 4¼ lb. coal, costing ¼d.; wages, 0.1d.; stores, 0.005d.; total, .0.355d. This figure includes no depreciation or repairs, nor does it cover interest on capital outlay. Nevertheless, it is hardly needful to point out that it is remarkably low, probably the lowest figure that has yet been well authenticated.

In the diagrams Figs. 1, 2, 3, and 4 the scales have been chosen very open in order to show the variations from week to week very markedly; but those who are accustomed to read and consider such diagrams will recognise that the uniformity of running, allowing, of course, for the gradual increase of work done, has been unusual. The installation is fortunate in the fact that the morning and evening demand peaks are not nearly so pronounced as they are on very many electric traction systems. This is shown in Fig. 3 which gives the hourly output throughout three typical days. The Metropolitan Railway is used as a means of business communication right through all the mid-day hours. Also for at least at night there are no passenger trains running, and for three out of every twenty-four hours none of the big turbo-generators are at work. The load factor is thus high, being about 42 percent, reckoned on the normal capacity of the four units installed.

The chief element in the economy of this station is however, undoubtedly the completeness and steady reliable working of the arrangements for almost entirely automatic working. The staff works in three shifts, and over its eight hours, each shift has a low load factor of employment. Considering that over half a million B.T. units are sent out on it per week, the place has at all times a weirdly lonesome aspect. The three shifts all told of high and low degree, number 42 persons, or 0.0030 of an individual per rated kilowatt. The total number of stokers is 14 to take charge of boilers whose normal duty is to evaporate and superheat 280,000 lb. or 125 tons of water per hour.

There is a legend that on one occasion 41 of the staff of 42 took a twenty-fours’ holiday, and that, in consequence of their taking the precaution to chain with a sufficiently free tether the master control switchboard man to his gallery, neither the board of directors nor anyone else outside the staff ever heard of the incident. The chief engineer however, denies that there is any basis founded in truth for this legend. He says that if it had been fact he would certainly have heard them-through the telephone - coming home again.

Seriously, this generating station is an important example of the important dictum that the best commercially economic results are obtained when there is no excessive straining after the last possible fraction in the percentage of physical efficiency. The most skilful engineering is not that yielding the exact maximum of physical efficiency, but that that gives maximum commercial economy - reliability and freedom from breakdown being a very influential factor in this economy - which is always obtained at the sacrifice of some small degree of physical efficiency.