26.1.3.1 Liquid fossil fuels

Fuel oil, owing to its relative ease of handling and storage, the absence of large quantities of residual ash and its high calorific values, had by 1970 become a most attractive fossil fuel. Since then, however, its relative cost with regard to other fuels and the uncertainty of its long-term availability have both tended to reduce its competitiveness. Nonetheless, particularly in the case of smaller boiler installations, oil remains a viable fuel.

The type of oil burned for steam generation can vary from light distillates having low viscosity and low specific gravity to the heaviest residual black fuel oils with very high viscosities. Occasionally, crude (or unrefined) oil may be considered for use directly in the boiler furnace; however, the presence of highly volatile fractions (normally removed in a refining process) may prove troublesome. Typical characteristic of fuel oils are shown in Table 26.2. Despite the low percentages of ash indicated in Table 26.2, trace elements such as vanadium, sodium and sulphur can be responsible for a number of operating problems.

A further source of liquid fuel is oil derived from oil-shale, which is a fine-grained, compact, sedimentary rock containing an organic material known as kerogen. Shale oil is obtained by heating the rock to 470 °C. (The yield is normally about 115.1 per tonne of rock.) Substantial reserves are known to exist in North America, in some parts of the Middle East, in China and in Australia. The use of shale oil in steam generators may be restricted by cost, but consideration is currently being given to the direct combustion of oil-shale in boilers.

26.1.3.2 Gaseous fuels

In many ways gas can be considered the most easily used of all the chemical fuels. It is capable of easy transportation from the producing or gathering plant to the consumer, and the use of pipelines can eliminate storage problems, particularly for very small installations. Care is required, however, in large boilers to avoid the occurrence of explosions during unit ignition or at other times when ignition may be lost. As may be expected, for clean gases complete combustion with a low excess air requirement is possible and the combustion does not produce smoke; additionally, it is substantially free of ash. Typical analyses of gases burned in steam generators are shown in Table 26.3. Blast-furnace gas is included in Table 26.3 for completeness, although it is usually available only as a by-product of a steelworks. It is normally heavily contaminated with dust, and great care must be taken to avoid plugging of fuel pipes or furnace fouling. Clean-up is usually achieved by a washing process.

26.1.3.3 Solid fossil fuels

Coal is a major fossil fuel throughout the world and has been in use for many decades by the industrialised nations. It is found in many forms, ranging from anthracitic through bituminous and sub-bituminous to lignitic. With so many sources and with such variations of type and quality, it has been found impossible to produce any generalised analysis for this fuel which will predict its behaviour in the boiler. As opposed to oil and gas, coal contains relatively high ash levels, which in turn will contain trace compounds (such as SiO₂, AlO₂, TiO₂, Fe₃O₄, CaO, MgO, Na₂O, K₂O, Mn₃O₄ and P₂O₅) whose presence and effects need to be recognised if efficient and reliable coal combustion is to be achieved.

With these reservations in mind, however, it is possible to divide coals into certain classifications. In Europe the International System of Classification is generally used, while in the USA the ASTM System is used. Both systems have attempted to group coal together by their heating values, moisture contents and volatile matter contents.

Anthracitic coals have a high heating value, are non-agglomerating, have low volatility and very high fixed carbon content (in excess of 85%). Bituminous coals have a heating value not less than 5700 kcal/kg, and are agglomerating, while sub-bituminous coals can have heating values as low as 4600 kcal/kg. Lignitic coals are characterised by high moisture contents (up to 60%) and heating values lower than 4600 kcal/kg.

To provide an indication of the variations of coals worldwide, some of the more significant producers are listed in Table 26.4, together with corresponding typical hard coal properties.

26.1.3.4 Other solid fuels

Other fuels, mainly vegetable or vegetable waste, have been found to be useful. Peat is, in fact, burned by certain public utilities. Wood and wood wastes as well as bagasse waste are also well-known fuels, while, more recently, trials have been conducted into the use of industrial and domestic refuse as a boiler fuel. Some indication of the properties of these fuels is given in Table 26.5.

26.2.1.1 Reheat cycle

Most fossil-fuel-fired generating stations use the reheat cycle. Currently, outputs span the range 200–1300 MW. Common steam conditions are 16 MPa, 540°C (2350 lbf/in.², 1000 °F). Occasionally supercritical conditions, e.g. 22MPa and above, are met, and sometimes there are two stages of reheat; but the one-stage reheat subcritical cycle here described is more common. The British advanced gas-cooled nuclear-reactor plant also uses this cycle, with outputs currently of 660 MW.

In Figure 26.3, ABCD shows the conditions of steam throughout the cycle, and Figure 26.4(a) gives the plant arrangement. Steam is expanded in the high-pressure (h.p.) stage down to near-saturation and is then returned to the boiler for reheat to the original temperature but at lower pressure. It then expands in the low-pressure (l.p.) stage until it becomes as wet (usually not more than 13% wet) as the final blades can tolerate.

The cycle is efficient and eases the design of last blading, which is usually the critical problem in turbine development. By using reheat, less steam per megawatt-hour is needed and this means shorter blades at exhaust. The cycle usually gives an exhaust somewhat drier than others, reducing the blade erosion hazard. However, all this is at the expense of some complication and extra boiler size and the smaller units—say under 100 or 200 MW—use the simpler non-reheat cycle.

26.2.1.2 Non-reheat cycle

This is shown by the line EF, in Figure 26.3, and the plant layout in Figure 26.4(b). Steam usually at 5–8.5 MPa (900–1250 lbf/in²) is expanded through the turbine to exhaust from the condenser without return to the boiler. This simple and compact machine is usually constructed in a single cylinder. The plant has a lower first cost than that for reheat, but is 8% or more higher in fuel consumption.

26.2.1.3 Pressurised-water and boiling water reactor cycle

The steam conditions are given by line GHJF and the layout in Figure 26.4(c). Water-cooled reactors require a special cycle because steam is supplied at an unusually low temperature, about 280°C (540°F). The steam is saturated and when expanded, quickly becomes wet. Water in the h.p. stages can cause damage, so the steam is taken out of the turbine after the h.p. stage and put through a drier—usually a device that uses the motion of the steam to separate water from it. It is then reheated by steam bled from the main supply to a surface heat exchanger alongside the turbine.

The cycle is not efficient and turbine design must take care of the potentially damaging effects of water in the steam. However, it suits the pressurised-water boiling-water reactors and their economics. Machines of up to 1300 MW are in service using such cycles.

26.2.1.4 Extraction cycle

This cycle, line EKF in Figure 26.3 with layout in Figure 26.4(d), is particularly economical where steam is used to supply a process as well as to generate: a typical example is the desalination plant fed with steam from a turbogenerator, an arrangement in demand in countries where industry and population develop in desert areas. Here, power units of 20–120 MW are in operation, and pass-out steam feeds multi-flash sea-water desalination plants.

Steam is expanded in the h.p. stage of a turbine and is then withdrawn. A control system acts to take off as much steam as the process requires. The steam that the process does not need is then returned to the turbine to expand and do work in the l.p. stages and ultimately to exhaust to the condenser under vacuum.

26.2.1.5 Back-pressure (‘total energy’) cycle

Figure 26.3 shows the cycle by line EK, and Figure 26.4(e) shows the plant diagram. While the turbine is efficient in itself, with stage efficiencies of around 80%, the basic cycle has a much lower efficiency (usually 30–38% overall for power stations), mainly because over half the energy in the steam is lost in the exhaust. With the exhaust pressure usually as low as 5 kPa (0.75 lbf/in²) the steam is at only 33°C (92°F). It is difficult to find ways to use energy at such a relatively low temperature and so the very considerable heat in the exhaust is discarded to a condenser.

Some factories need steam for process work, e.g. at upwards of 150°C and at a pressure useful enough to drive the steam to and through the process plant—say 350–1000 kPa. District heating schemes usually require water at 65°C and obtain it from steam at around these conditions.

The turbine can be arranged to exhaust at these higher pressures and to pass all its exhaust steam to the process or to the district-heating scheme. Although the higher back pressure of, say, 700 kPa makes the machine much less efficient than with a more normal condensing turbine, the cycle as a whole is efficient because the turbine exhaust heat is not lost. Thus, a back-pressure plant generating electricity and exhausting to a factory process can have a thermal efficiency of 70% compared with 35% for a normal power station.

The basic problem with back-pressure units is that the heat required near to the turbine-generator rarely keeps pace with the electrical power required locally. Basically, it is easier to distribute electricity to more distant parts than to send heat: back-pressure units are therefore usually small. Confining their power and heat output to nearby use, they are rarely of more than 20 MW rating and often much less. There may well be a future for such cycles in the conservation of increasingly costly fuel, and the potential of such ‘total energy’ schemes must inevitably attract more attention.

26.2.1.6 Combined cycle

The ideal engine has a high inlet thermodynamic (absolute) temperature T₁ and a low exhaust temperature T₂. The highest attainable Carnot efficiency is (T₁ − T₂)/T₁. The steam turbine has only a moderate T₁ but a very low T₂, so that its efficiency is normally higher than that of the gas turbine, for which T₁ is high but so is T₂.

The combined cycle aims to optimise conditions with a cycle that uses the high T₁ of the gas turbine and the cool exhaust of the steam turbine, Figure 26.4(f). The cycle works as follows.

Fuel is burnt in the combustion chamber of a gas turbine, which generates electricity and exhausts to a heat exchanger. Here the hot gases boil water to raise steam which then drives a steam turbine which also generates electricity. The steam turbine exhaust is changed to boiler feed water in a condenser.

A number of such plants are in service and more are being built because of their high efficiency. They also have the quick-starting ability that the straight steam turbine lacks. Also they can offer power early in the building programme by installing the gas turbine first and the steam turbine later.

They have as yet a reliability somewhat lower than that of simple steam plant, and the gas turbine has a shorter life. The availability is more acceptable if natural gas is used rather than oil. A common arrangement is to take about two-thirds of the power output from the gas turbines and one-third from the steam turbine; say, two gas turbines and one steam turbine per unit.

While this cycle is as yet limited to gas and carefully prepared oil, there are prospects of its use with coal. The pollution problem is creating so strong an emphasis on clean burning that new combustion methods are being developed for coal. Pressurised, fluidised bed combustion is one; gasification is another. Such arrangements promise to produce a gas clean enough for gas turbines to use. So there is a prospect of using the combined cycle to give thermal efficiencies of around 45% using coal, which is the most profuse fossil fuel.

Combined cycles are not necessarily arranged as in Figure 26.4(f). For instance, fuel can also be burnt in the gas-turbine exhaust to raise more steam (see also Section 26.3.3).

26.2.1.7 Alternative fluids

From time to time research is done into fluids alternative to steam. The physical properties of freon and ammonia have certain advantages, but so far the disadvantages have over-ruled them. Energy costs, however, may yet impose a change from the steam cycle.

26.2.2.1 Blade types

There are two basic types of blade—impulse and reaction. In the impulse type (Figure 26.5(a)) all the expansion of steam is done in fixed nozzles and the high-velocity jets so created drive the rotor blade. In most machines the expansion is in successive stages, of which there are usually as many as 20 from inlet to exhaust, each comprising a row of nozzles and a row of moving blades.

Some turbines use a ‘Curtiss’ stage (Figure 26.5(b)) to drop the temperature quickly at inlet or to shorten the machine. This stage takes a more than usual pressure drop. An exceptionally high jet velocity results and this is converted to work by driving two or more rows of blades from the one jet. The sketch shows how this is done using guide blades to re-direct the steam leaving the first row.

In reaction blading (Figure 26.5(c)) the moving blades are driven not only by the steam jets from preceding nozzles: they act as nozzles themselves and the jet they create helps to propel them. Therefore, pressure is lost both in the fixed and moving blades and power is produced both by impulse and jet reaction.

In practice, few machines are purely impulse and none are purely reaction. In the modern so-called impulse machine all blades have a degree of reaction and generally this increases towards the exhaust. In some designs the amount of reaction is more marked, depending on the background experience of the manufacturer. Properly designed, there is little difference in efficiency between the impulse and reaction types.

In all but a few small machines, the steam flows along the shaft, parallel to the axis. There are two unusual designs in which this is not so. One is the Ljungström turbine, in which steam flows radially outwards through concentric rings of counter-rotating blades. The other is the Terry turbine where the steam jets travel at an angle inwards to drive against hollows in a rotating disc. The Ljungström turbine is confined to medium outputs, while the Terry turbine is used only for very low powers, e.g. for driving small auxiliaries.

Both these types are now rare. Most modern machines have axial flow with either reaction blading or a combination using h.p. stages of mainly impulse and l.p. stages of mainly reaction design.

26.2.2.2 Size effects

The steam turbine is notable in that higher outputs can be achieved without a pro-rata increase in mass and cost. A power increase of 20% is achieved by using blades 20% longer and this increases the mass of the machine by much less than 20%. So the mass/output ratio is a curve such as that shown in Figure 26.6. Also, parasitic losses such as gland loss, etc., do not increase at the same rate as power does when greater outputs are used. Thus, increased size produces some gain in efficiency (Figure 26.7).

Taking these two factors together there is a financial gain to be got from increased size (Figure 26.8). Two important points to be noted, however, are:

26.2.2.3 Part load

The steam turbine performs best at full load, where the whole system of nozzles and blades operates with proper steam velocities, with minimal throttle losses in valves. It will be less efficient on part load: energy is wasted by throttling down the pressure in valves. The steam jets are slow and do not strike the blades at the right angle, and the exhaust is too hot. If a turbine has to be used frequently at part load, it can pay to adopt nozzle rather than throttle control. The arrangement shown in Figure 26.9 could show advantage over that shown in Figure 26.10.

In nozzle control, part load is achieved by using fewer nozzles and unthrottled steam. The technique is more common overseas than in the UK, though mechanical difficulty tends to make it less suitable for large machines.

26.2.3.1 Form and arrangement

It would be ideal to have all turbines as simple as the single-cylinder machine, this is shown diagrammatically in Figure 26.11(a). In fact, these are built for outputs up to 80 or 90 MW. Above these, the longer blades make two or more exhausts necessary and these need a separate shaft. The next step is to Figure 26.11(b), which in the non-reheat form provides outputs of 250 MW or more.

When reheat is adopted there are two hot inlet flows, one at high pressure and a second at an intermediate pressure. Some makers mount each on a separate shaft, giving the h.p., intermediate-pressure (i.p.) and l.p. arrangement of Figure 26.11(c). Others take both inlets to a common h.p.-i.p. rotor, Figure 26.11(e). The latter is common American practice.

The arrangements so far described are for rated speeds of 3000 or 3600 rev/min. An important design factor concerns the last blade size: increase in blade height permits the number of exhausts to be reduced. Current UK designs for 3000 rev/min turbines include blading heights up to 0.95 m on a mean diameter of 2.65 m. For 300 MW these use two exhausts as in Figure 26.11(c) and four for 600 MW, as in (d), or alternatively six where the higher efficiency obtained is economically justified.

Developments may lead to even longer blades, capable of 600 MW with two exhausts as in Figure 26.11(c) provided that low coolant temperatures are feasible. In some areas of the world these cannot be obtained, and only low area exhausts are justified.

Where the power rating is such that the exhaust area is beyond the capacity of the largest blades, the necessary area can be provided by a two-shaft or cross-compound arrangement, e.g. Figure 26.11(f), or by halving the speed as in the 1800 rev/min 1300 MW machines such as that in Figure 26.11(g). These arrangements accept, as most do, that eight exhausts on a single high-speed shaft line would not be a preferred solution.

There is a compromise between the two methods shown by using a cross-compound with a high-speed, h.p.-i.p., line and a low-speed, low-pressure line, either 3000/1500 or 3600/1800 rev/min depending on grid frequency. Inherent in these schemes is the fact that the lower speeds make it possible to use much longer blades and so enlarge the exhaust area and machine output. However, they do so at the expense of more metal for the low-speed rotors and casings, which are massive in comparison with those for a high-speed set.

Engineering solutions to the problem of increased output are always evolving and much larger turbine sizes than the current 1300 MW units could, no doubt, be devised. The limit tends to be imposed not by technical aspects of turbines, but by economics and by factors other than the turbine itself. An influence that the turbine contributes to this limiting of growth, however, is that the return tends to level off as size increases and can easily be nullified or even reversed if availabilities of the large machines fall even slightly beneath those of the smaller ones.

26.2.3.2 Foundations

The large turbine generator is supported some 10–12 m above the basement, a space that accommodates pipes, condensers and drainage. Supports were formerly mostly of reinforced concrete, but steel supports are increasingly preferred. They can be accommodated more readily into a building programme and are more predictable in their dynamic characteristics.

The block usually takes the form of an entablature supported on columns. It must be designed so that the shaft runs smoothly even though it may be slightly out of balance. The foundations must sometimes accommodate earth tremors. Both these requirements are better served if the structure is in steel than in concrete, because with steel the stresses, loadings and masses commercially used give a lower and more predictable and natural frequency. Vertical frequencies of about 10 Hz and sway frequencies somewhat lower are typical of such designs. These frequencies are substantially less than the shaft frequency of 50 Hz or half-shaft frequency of 25 Hz, both of which must be avoided if resonance is not to occur.

Concrete foundations tend to give calculated frequencies nearer 50 Hz, either above or below it. If the softness of the earth beneath the foundations is allowed for, these fundamental frequencies fall usually to well below 50 Hz, but not so low as with steel. While concrete foundations have given good service, their response is more difficult to predict and they are less likely to accommodate shaft and earth dynamics so acceptably as do steel foundations. Also, steel is usually more amenable to a quick station-building programme. However, in many countries concrete may be more economic and the choice has to be made on the circumstances.

26.2.3.3 Lubrication

The security of the turbine depends on a continuity of oil supply to its bearings and on the safe functioning of its hydraulic control gear. Further, the fire record in power stations shows that the oil system may play a part in starting or aggravating a fire. It pays to invest thought and money to make the oil system completely reliable and safe. Oil systems involve quite large flows with large turbines, partly because the bearings (especially the thrust block) have turbulent conditions in them unlike the less power-absorbent laminar conditions in former 30 MW sets. Special measures aim to reduce these power losses but they can still reach 3.5 MW for a 660 MW set.

Most machines have shaft-driven main oil pumps, usually centrifugal, primed by jet pumps or pumps driven by a hydraulic motor. The main pumps are backed by a.c. full-duty pumps and by a battery-operated d.c. emergency pump to provide oil until the shaft stops should there be an a.c. failure. Sometimes a small a.c. pump is fitted for use on barring. Shafts are generally jacked by h.p. oil to help starting.

In many arrangements the main oil pump also supplies the control gear with fluid; in others, separate pumps are used and feed control gear at much higher pressures than are used in the lubrication system. A fire-resistant fluid is sometimes used for control but is rarely considered for lubrication. The lubricant fluid is still oil, which can burn: it is therefore essential to design the oil system and fire-fighting plant to reduce fire risk. Attention is needed especially to pipe flange and coupling design. Any item that may fracture or leak oil should be eliminated or designed to reduce risk.

26.2.3.4 Governing

The governing and protection system operates control valves both to keep the speed steady within narrow limits and to prevent overspeed on sudden loss. The large turbine has stop and throttle valves at its inlet, and stop and interceptor valves at the i.p. inlet from the reheater. Usually, speed control is by adjusting the h.p. throttles. Overspeed on sudden load loss is prevented, first by closing the throttle valves and interceptors and then, should speed still rise, closing the h.p. and i.p. stop valves. Overspeed prevention requires fast valve action, usually well under 0.5 s closure and in some cases down to 0.15 s. The modern large machine has a high power/mass ratio and a sudden loss of full load can cause accelerations of 10–15% of full speed per second. This allows only a short time to cut off the steam supply, a situation aggravated by the stored energy of steam already in the pipes and spaces downstream of the valves. It is an important aspect of machine arrangement to minimise such spaces, e.g. using short steam loop pipes.

Other protective circuits besides overspeed operate the valves. The list usually includes low bearing-oil pressure, high vibration and poor vacuum.

Mechanical governors of large machines operate the steam control valves via hydraulic relays and power amplifiers and, eventually, by a servo-motor whose position is determined by speed level and held by a pilot valve and feedback. There are variations on this theme, some having an acceleration element.

Electric governing is becoming more general. Speed accelerations are sensed electrically and translated to an error signal in some adjustable way within the electric governor. The final movement and positioning of the steam valve is still hydraulic.

Some small machines use mechanical relays between speed sensor and valve, but most use hydraulics or electrical signalling and hydraulic power for valve movement. Lubricating oil is normally used, with pressures taken from the main oil pump at 0.35–2 MPa (59–300 lbf/in²) but occasionally separate pumps are used for control and with pressures as high as 17 MPa. Fire-resistant fluids are used in the control systems of some large machines. In US practice, double piping is often called for if lubricating oil is used. The reason is that oil leaks are retained within the pipes rather than spill to atmosphere and cause a fire hazard.

26.2.4.1 Condensing plant

The condenser itself is large and costly consisting essentially of tubes through which cooling water is pumped and about which steam and air move. The steam is condensed and the air extracted. The condensate is drawn out by extraction pumps and in an efficient condenser it is nearly at the same temperature as the steam. It is returned to the boiler to use again. For every 1 °C by which the exhaust temperature is lowered in a typical 100 MW set, up to 0.5% extra fuel is burnt. Thus, a larger surface area of tubing allows the exhaust to be brought down more nearly to the coolant temperature; but, of course, this means a higher plant cost.

Having decided the surface area, however, it is important to use it properly. A particular offender is air, which can increase the exhaust temperature if it enters the condenser in large quantities, blanketing the tubes from steam. Excluding air is important: the condenser must extract such air as does enter without its forming pockets and blankets.

In the past, tubes have usually been made of brass. Where there is a severe corrosion risk, cupro–nickel is sometimes used. Dosing is usually employed, e.g. by ferrous sulphate injections, especially where there is a risk of corrosion or of erosion. Sometimes a cleaning gear is used in which shoals of floating balls pass through the tubes, are recollected and put back into the tubes.

Thin tubes of titanium are now becoming more usual. The evidence so far is that they offer much greater resistance to corrosion and erosion and often promise economic advantage.

With the high-steam conditions now used, it is vital to keep salts from the boiler and tube fixtures usually involve expansion or welding. Packings are rare. Sometimes anti-leak measures extend to the use of double tubeplates, with pressurised distilled water filling the space between the two plates. This ensures that, if there is a leak into the steam space, it will be of clean water and can be monitored. Large condensers usually have arrangements for on-load cleaning.

Ejectors are used to remove air from the condenser. Currently, steam-jet ejectors appear to be in favour but there are various alternative electrically driven ejectors.

The cooling water system can be quite elaborate and usually involves a considerable amount of both mechanical and civil engineering. In a typical British coastal 660 MW unit, for every 1 kg of steam that enters the condenser, about 30 kg of cooling water must be pumped through it. This means a flow of some 1000 m³/min. A 1300 MW nuclear unit on the Californian coast uses as much as 3000 m³/min of cooling water.

There is no standard cooling system using sea-water: the arrangement and equipment are made to suit the circumstances. Most have one or all of the elements shown in Figure 26.12. The system shown is elaborate because it is designed for the extreme conditions met, for instance, in the Mediterranean, where pollution causes growth of sea-grass. Periodic disturbance by storm can bring large quantities of this into the cooling system of a power station. The choking which results can take the turbines off load and is always a worry as regards chemical attack of the tubes. The weed has to be trapped and extracted before the condenser is reached if trouble is to be avoided.

Sand can also be a hazard, especially if it is hard and sharp. Settling tanks are often used to take sand out of the system.

In such systems the water inlet is taken well out to sea. This usually avoids the areas of higher weed density. The inlet is kept well above the seabed to avoid intake of sand, yet well beneath the surface to avoid oil slick. Water inlet positions must be planned most carefully with all these factors in mind and with others, such as avoiding warm water from the discharge re-entering the system. The work involves a careful on-the-spot survey with seasonal observations. It usually entails a hydraulic model.

Power stations are often sited on rivers. In some estuaries the water may be seriously short of oxygen or have aggressive pollutants, which promote tube damage. In others, the water is warm through use by other power stations or may be in seasonal short supply. Cooling towers then become necessary. Their essential principle is shown in Figure 26.13. Warm cooling water is sprayed into the atmosphere within the tower and passes its heat to the air, which in turn becomes warm and rises. The tower is thus filled with air which is warmer than that outside and the thermal syphon effect creates a through-draught. Wind across the top of the tower can stimulate the process.

In a sea-water system the turbine exhaust temperature is set by the temperature of the sea plus the temperature difference required by the condenser to transfer heat. In cooling-tower plants a further temperature rise occurs because the air is normally warmer than the sea; thus the station efficiency is lower and the cooling plant more costly.

Some water is lost by a wet cooling tower. About 1.4% of the total throughput is carried away by the air—about 27 000 t/day for a 660 MW plant. At some sites a reasonable source of water is too remote for economic use and a dry cooling system must be adopted.

Figure 26.14 shows the principle of the dry cooling tower. There is no direct contact between the water coolant and air, so no water is lost. Cooling takes place in large heat exchangers at the base of the tower across which air is drawn either by natural draught or by fans. The cooled water, which must be pure enough to use in the boiler, is sprayed into the steam entering a jet condenser. Such a system was installed at Rugeley in England in 1961 as an experiment at 120 MW and others have been built overseas since. An alternative arrangement is to use a surface condenser instead of a jet condenser. Raw water is then used as a coolant instead of feed water.

The dry-tower system is less efficient than the wet because the terminal temperature differences are higher so poorer vacua are obtained. Also the tower itself is much larger and so the system costs more. It can, however, enable large amounts of power to be produced in a dry terrain.

Where lower-power units are required, an air-cooled condenser (Figure 26.15) may be more economical than a dry tower. In the air-cooled condenser, the turbine exhausts to a heat exchanger mounted some distance above it. Steam goes through tubes and is condensed by air drawn or blown across them. The choice between air condensers and dry towers depends on many factors, but currently the former is used below about 100 MW and dry towers above about 200 MW.

26.2.4.2 Feed plant

The feed water produced in the condenser is taken back to the boiler through a train of devices. First the extraction pump raises it to well above atmospheric pressure and drives the water through a series of l.p. heaters, which warm it with steam drawn from the turbine. The heaters are usually of tubular type, the direct contact type is out of favour after accidents in which water spilled back into the turbine.

After leaving the l.p. heaters, the feed water is lifted to the deaerator where it is mixed with more steam bled from the turbine, and releasing dissolved air that would otherwise appear in the boiler. Sometimes the deaerator is incorporated in the condenser itself.

Water is then passed through a network of further tubular heaters and the feed pump, before passing to the boiler via flow-regulating valves. A typical arrangement is shown in Figure 26.16.

The feed-heating system is complicated, expensive and often takes up almost as much station space as the turbine itself. In spite of the fuel saved by it (and even for a small station and medium steam conditions this can amount to over 10%) there is a trend to simplify by reducing the number of feed heating stages, more particularly on stations for developing countries where simple robustness is desirable and where fuel is not expensive.

The feed pumps of a large power station absorb considerable power—up to about 15 MW for a 600 MW generator. There is much to be gained by driving them in an economical way.

26.3.1.1 Simple power relations

Figure 26.20 shows the ideal pressure–volume diagram representing the ideal cycle work derived from the turbine and the ideal work necessary to drive the compressor. In practice the work done in the turbine is less than that for the ‘ideal’, i.e. the isentropic efficiency η_t of the turbine is less than unity. Similarly the work done in the compressor is greater than for the ideal case, i.e. the isentropic efficiency η_c is also less than unity. The useful work per unit mass of air is the difference between the work done by the turbine and the work required by the compressor, namely

where k = r^{(γ − 1)/γ}, r is the pressure ratio of compression and expansion, γ is the ratio of specific heats, c_p is the specific heat at constant pressure, T₁ is the absolute temperature at compressor inlet, and T₂ is the absolute temperature of the air after combustion and before entry to the turbine.

It is evident that if η_t, η_c and T₂ are too low this expression could have a negative value. The turbine output would not be sufficient to drive the compressor and there would be no useful work. Hence the success of the gas turbine depends on the attainment of high isentropic efficiencies in the compressor and turbine and the ability to operate at a high turbine inlet temperature.

The gas turbine must operate at very high temperatures to develop efficiencies comparable with those attained in steam turbines and diesel engines. The turbine blades must withstand both high stresses due to rotation and the high temperature of the combustion gases passing over them. The success of the gas turbine has been made possible by the development of steels which will permit stressing at high temperatures without excessive creep. To allow the use of higher gas temperatures the turbine blades are often cooled by fluid circulating through passages in them. The designed maximum temperature of operation is determined by the required ‘life’ of the turbine. A lower temperature of operation will reduce the creep rate of the stressed parts and increase the number of hours of running life of the turbine. Temperature gradients in rotors and blades also induce stresses and alternating stresses due to blade vibration could cause failure by fatigue. With the materials and design experience now available, efficient and reliable gas turbine plants are being built and many are in successful operation.

The gas turbine is not a self-starting machine; it must be brought up to speed by a starting motor until the compressor is running fast enough to attain an adequate pressure ratio and component efficiencies for a self-sustaining or viable cycle. After ignition of the fuel the machine will increase in speed until the turbine drives the compressor and provides the useful output. If the gas turbine drives a generator with separate exciter, the exciter may be used as a starting motor.

As in the operation of a steam turbine, starting up and shutting down must take place with due precaution to avoid excessive temperature gradients, particularly in the rotors.

Contemporary gas turbines have compressor pressure ratios (delivery pressure/inlet pressure) of as much as 30:1 resulting in combustion chamber pressures up to 3000 kN/m² (435 lbf/in² gauge).

26.3.1.2 More complex plant

The efficiency of the plant described above is somewhat low (25–30%). It has advantages of simplicity, low cost and independence of water. Higher efficiencies (30–40%) may be attained at the expense of greater complication and cost.

The addition of an efficient heat exchanger has a most pronounced beneficial effect on efficiency. Figure 26.21 shows a gas turbine plant in which a heat exchanger is included. Heat in the turbine exhaust is transferred to the air delivered by the compressor so that less heat is required from the combustion of fuel. The output of the plant is not increased but the reduction in the amount of fuel burned improves the efficiency. To justify its space, weight and cost the heat exchanger must have a high thermal ratio. It usually consists of a shell containing nests of small-bore tubes over which the turbine exhaust passes giving up heat to the compressor delivery air passing through the tubes on its way to the combustion chamber; however, heat exchangers have not been adopted widely for reasons of cost, bulk and reliability.

For electric power generation the gas turbine is essentially a constant-speed machine. The simple, single-shaft open-cycle gas turbine is not generally flexible as it needs to operate at full load at constant speed to give its highest thermal efficiency. Some flexibility is possible by adopting a ‘split shaft’ (sometimes known as a ‘two-shaft’) machine, but much better flexibility and performance can be obtained at the cost and inconvenience of complexity. There are many possible combinations of compressors, intercoolers, turbines, combustion chambers and heat exchangers. Figure 26.21 shows a plant in which the compressor and turbine are each in two stages. Interstage cooling of the compressed air increases the efficiency of compression. Another combustion chamber may be interposed between the h.p. turbine and l.p. turbine to increase the output of the plant. The separation of the turbine producing useful output from that driving the compressors makes the plant more adaptable to load variations with less reduction of efficiency at part load. The cost of more complex plant is, of course, justified only when the load factor and fuel costs are high.

26.3.1.3 Aircraft-type gas turbines

Since the 1960s there has been widespread application of aerojet gas turbines in which the propulsion jet is coupled to a heavy duty power turbine. The development of this type of turbine makes use of the extensive research carried out by the aero-engine industry and employs one or more aircraft jet engines to discharge into a gas turbine. Such units are reasonably inexpensive, compact and have exceptionally quick starting properties, e.g. full load from a cold start in 2 min.

The jet engine on aircraft has frequent skilled maintenance and normally operates at a high altitude with low atmospheric pressure. By de-rating the engine and incorporating certain small modifications it can be operated successfully at sea level at a power between its normal flight and take-off ratings and can give 20 000–25 000 h operation between overhauls. Its compactness also makes it suitable for mobile power plants.

26.3.1.4 Free-piston gas generator

Instead of the compressor and combustion chamber of the conventional gas turbine, one or more free-piston gas generators may be used to produce the hot gas for discharge into the turbine. This comprises a cylinder containing two free pistons; fuel is admitted between the pistons and, by compression ignition and appropriate operation of valves, hot gas is emitted from the exhaust ports. The chief advantage is a higher overall efficiency (30–35%) than the conventional gas turbine, but cost and maintenance requirements are generally greater so that applications are limited. A few successful plants of between 1 and 10 MW are, however, in operation.

(1) Gates and/or valves used for isolating the turbine, together with the associated pumping sets and pressure receivers. Note that closure must be effected either under the action of gravity or by utilising stored energy.

(2) Governors, actuators and servomotors and a power oil system containing sufficient stored energy for operation of a closing–opening–closing cycle.

(3) Duplicated de-watering system with a proportion of the de-watering pumps supplied from a secure battery based supply.

(4) A compressed-air system capable of charging and topping up the pressure receivers under (1) and (2). Where synchronous condenser operation is involved or where pump turbines are employed, the compressed air system should have sufficient stored capacity to permit the blow down of one unit at a time and have the additional capacity to make up air losses following the blow down.

(5) Cooling systems for generators, transformers, pumping sets and for thrust and journal bearings of the turbogenerator.

(6) A heating and ventilating plant capable of maintaining the required degree of comfort.

(7) Voltage regulators and controllers.

(8) Instrumentation and controls for monitoring and operating the units and their auxiliaries, including automatic synchronising equipment.

26.4.4.1 General principles

Consideration of hydraulic conditions at the turbine runner leads to the division of water turbines into two main groups: the impulse turbines represented in modern turbine practice mainly by Pelton wheels; and reaction turbines, a group covering both mixed flow and axial flow machines.

Impulse turbines are driven by jets of water issuing from one or more nozzles distributed tangentially around the periphery of the wheel. The power output is controlled by adjusting the opening of the nozzles. As such adjustment does not affect the direction of the jet, the part-load performance of such machines is on a par with that at the design point.

With a few exceptions, reaction turbines are normally equipped with movable guide vanes. These are disposed symmetrically around the runner and control both the velocity of flow and its direction at the entry to the runner. The majority of reaction turbines have runners whose geometry is fixed and invariable. Modern, mixed flow machines, equipped with such runners are known as ‘Francis’ turbines. The flow at inlet is invariably inward and the flow at exit is usually axial. The axial-flow machine with a fixed geometry runner is known as a ‘propeller’. Turbines in which the guides are movable but the runner blades are fixed are said to have ‘single regulation’. Part-load performance of such machines (see Figure 26.28) tends to be poor. In order to improve performance, machines with both movable guide and movable runner blades have been introduced. Such machines are said to have ‘double regulation’.

Axial-flow turbines with double regulation have been employed since the 1920s. They are known as ‘Kaplan turbines’, after their inventor. Double regulated mixed-flow turbines have been used since the 1950s. This type of turbine is now designated as either a ‘diagonal’ or a ‘Deriaz’ turbine, after its inventor. It may be of interest to note that the Deriaz turbine was first developed in Rugby, England. Perusal of Figure 26.28 will show the degree of improvement, of part-load performance, resulting from the use of double regulation, on both axial- and mixed-flow machines. It will be appreciated that double regulation increases both the size of a machine and its costs. However, the resulting improvement in performance under a range of operating conditions can make their use economically justified, either where single or very few machines are installed, or where substantial head variations are encountered.

The range of heads for which the various reaction machines operating as turbines and as pump turbines are suitable is shown in Figure 26.29 This figure also indicates the outlines and the proportions of the runners of such machines. The shape of runners corresponding to progressively higher heads are shown from left to right.

With the advent of physically large units operating at very low and varying heads, as encountered on some run of river schemes and on tidal applications, yet another variation on the design of an axial flow turbine has been introduced. This type of turbine, known as a ‘Kapeller’, has fixed guide vanes but the runner is equipped with movable runner blades. Such a combination produces a plateau of relatively high turbine efficiencies, along the line of maximum output for a substantial range of heads.

In an impulse turbine all the available energy is converted into velocity, before the water enters the runner, while in a reaction turbine the process of conversion takes place partly before and partly after the water has entered the runner. The division of water turbines into these two groups is based on general usage and does not imply any difference in the method of energy transfer between the water and the runner. At its simplest the principle of a water turbine is that of a rotating duct, through which flows a stream of water. The stream and the duct interact; the stream is deflected and, as a result, a force is exerted on the duct. The moment of this force, about the axis of rotation of the duct is equal and opposite to the change in the moment of momentum of the stream. If the mass of water, flowing in a unit of time is ζQ, the driving torque M is given by

(26.4)

where C_u1 and C_u2 are the whirl velocities at the inlet and exit, and r₁ and r₂ are the corresponding radii. Note that only the whirl components of the absolute velocities of flow contribute to the driving torque.

The driving torque can also be determined from considerations of power output from the turbine and of angular velocity ω. Hence it can be shown that

(26.5)

where u₁ and u₂ are the peripheral velocities of the runners at the inlet and exit, and η_t is the turbine efficiency.

If the effect of the small mechanical and volumetric losses on turbine efficiency is neglected and the concept of effective head H_e = H_nη_t is introduced, we get the Euler turbine equation in its canonical form

(26.6)

This equation demonstrates the interdependence between head, velocities and the geometry of the turbine, at entry and exit from the runner.

It should be noted that, although equations (26.4) to (26.6) are constantly used in the literature to describe the performance of water turbines, they are only strictly applicable to machines in which the flow can be fully described by a single representative stream line.

26.4.4.2 Fundamental similarity

Turbine characteristics are normally established on the basis of tests carried out in hydraulic laboratories on homologous models of the machine and of certain associated water passages, be it distributor pipework and nozzles in the case of a Pelton wheel, or a portion of the inlet pipe, the spiral casing and the draft tube in the case of a reaction machine. For the model to be homologous it must be:

In order to meet these conditions, in machines operating at speeds which are inversely proportional to size it is necessary to ensure that Reynolds numbers for both the model and the prototype are sufficiently high to assure similarity of viscous forces. In addition, the consistency of head coefficients (ψ) at corresponding points must be assured. This last requirement can be expressed mathematically as

(26.9)

Using the relationships established above, and if small differences in the respective prototype and model efficiencies are disregarded, it is possible to derive a number of parameters in terms of which the performance of a whole family of homologous machines, of various sizes and operating at differing heads can be reduced to just a single diagram, applicable to the whole range of similar turbines. This is achieved by reducing all the data to those appertaining to a turbine with a runner of unit diameter, operating under unit head. These parameters are:

(26.10)

(26.11)

(26.12)

(26.13)

In the case of turbines, the performance is nowadays most commonly expressed in terms of unit speed against unit quantity, with efficiency curves (mussel curves) superimposed on the diagram. Occasionally, cavitation data are also superimposed. In the case of pump turbines the above diagrams would normally be extended to cover both turbine and pump operation. In addition, a four-quadrant plot of unit speed against unit torque is often given.

Typical turbine characteristics for a high-head Francis and a propeller turbine are shown in Figures 26.30 and 26.31. Figure 26.32 shows a four-quadrant characteristic for a reversible pump turbine.

26.4.4.3 Specific speed

In order to extend codification of hydraulic performance from that covering just one family of homologous turbines at a time to all turbines at large, it was found necessary to introduce yet another parameter. Theoretically, this parameter, known as ‘specific speed’ (N_s), should be obtained by combining unit speed and unit power at the point of best efficiency (b.e.p.) in such a manner as to eliminate the size of the unit from the resulting expression

(26.14)

However, the published data are more often than not based on the value of this parameter for the rated output rather than the b.e.p.

It should be noted that specific speed is a dimensional quantity and, therefore, the system of units must be specified. Thus in SI units, speed (n) is in revolutions per minute, power (P) is in kilowatts and head (H_n) is in metres. Unfortunately most of the available data are classified in terms of specific speed expressed either in metric or British units. In both systems, the horse power is taken on the reference unit and heads are in metres or feet, as appropriate. Thus, when referring to such data, it should be borne in mind, that

Whichever system of units is used, the numerical value of the specific speed represents the rotational speed at which a homologous machine would operate if it were designed to give an output of unit power, under unit head, when operating in the same fluid and at the same efficiency.

At this point, it should be mentioned that, while machines having different specific speeds cannot be geometrically similar, it is possible for machines of the same specific speed to be geometrically different, because the same hydraulic conditions can be met by different designs. Thus at certain ranges of specific speed there are overlaps between machine types. The two such ranges of interest concern the overlaps between Francis and multi-jet Pelton wheels at the lowest specific speeds and between Francis and propeller turbines at the highest specific speeds corresponding to the Francis range.

In selecting an appropriate specific speed for large reaction turbines of a particular installation, use may be made of a number of empirical expressions which summarise the current achievements world-wide. These expressions are normally given in the form

(26.15)

where A is a constant whose magnitude has been steadily increasing with the passage of time, following the adoption of higher specific speeds.

In the case of major reaction machines this constant is now taken as 2400 or so. Some 25 years ago, on Francis units US practice was based on half of this value. This improvement was made possible by better understanding reached through research and testing of the conditions at the turbine runner, where the danger of damage by cavitation represented the major risk. The steady increase in the magnitude of this constant represents a substantial economic benefit because the higher specific speed leads to smaller runner diameters and higher running speeds.

In the case of impulse machines the specific speed is based on a per-jet principle. Thus if a wheel is equipped with four jets its specific speed would be twice that of a wheel with just one jet. The range of specific speeds covered by Pelton wheels equipped with a single jet extends approximately from 10 to 80 rev/min, the lower figure corresponding to the highest available heads.

Because of the overlap between Pelton wheels and Francis turbines, the choice of machine type for a given combination of head and flow becomes too complicated to be expressed by a simple formula. Instead, the reader is referred to Figure 26.33 which demonstrates the design trends.

For many years specific speed has been used as a general guide to the shape, relative size and speed of machines. It also provides access to information regarding typical efficiencies, flow and head coefficients as well as to values of runaway speeds appropriate to a given class of machinery. The relevant information is summarised in Figure 26.34.

26.4.7.1 Historical development

In many countries the available water-power resources are becoming fully utilised and, as environmental considerations lead to resistance to further encroachment by developments leading to the inundation of large tracts of land, modern hydroelectric practice is tending towards construction of pumped storage schemes. Currently some 20% of the world’s largest turbines are used in this mode.

The location of such installations is not critically dependent on the quality of the catchment area, but there must be sufficient water to fill and make good any losses due to seepage and evaporation. However, any natural inflow to the upper reservoir produces a bonus in output.

As the cost of machinery per unit output depends on the available head, there is an incentive to construct such schemes at sites providing potential heads in the range 300–600 m. Should such sites not be available, lower heads could still be exploited, but at a greater cost. The pump turbines employed on schemes with heads of up to 600 m would nowadays invariably be of the single stage, reversible type. Where higher heads are encountered, the pump turbines would employ several stages of runners/impellers. Alternatively, the scheme itself can be based on the use of two or more pump turbines operating in series. However, such a solution would involve substantial additional costs.

Wherever possible, the existing storage capacity of conventional hydroelectric plant is used as the base for the pumped-storage plant. In the absence of such facilities, large existing underground excavations may be used as the lower reservoir with the pump turbines installed in a deep shaft, well below the free level of the underground reservoir.

Historically, on pumped-storage schemes separate pumps and turbines were employed with the generator doubling up as a motor in the pumping mode. The four 90 MW units operating at a maximum head of 320 m installated at Ffestiniog, North Wales, are typical of this type of installation. All the machinery is arranged in line and the length of the main shaft is around 30 m. When pumping, the turbine is run in the dry and the thrust load is supported on two separate thrust bearings, one of which is located within the generator and the other under the pump. A coupling is provided between the turbine and the pump. Separate water conduits lead to the turbine and the pump and these are guarded by separate sets of valves (see Figure 26.35(a)). In 1960, when the scheme was commissioned, Ffestiniog was the largest pumped-storage project in the world. Even at the time, the design concept was already obsolescent. The next generation of pumped-storage plants depended on the use of reversible pump turbines which were developed at that time. Figure 26.35(b) demonstrates the simplification of layout arising out of the use of reversible units.

In Great Britain, because of constraints imposed by the terrain, conventional hydroelectric installations provide a very small fraction of the electrical output. Such schemes as are in existence are invariably small and are mainly located in the north of Scotland. However, there is a substantial investment in pumped storage. In addition to the 360 MW installation at Ffestiniog, there is a four 100 MW turbine installation at Loch Awe, operating at a maximum head of 365 m; a two 150 MW turbine installation at Foyers, Loch Ness, operating at a head of 180 m and the six 300 MW turbine station at Dinorwig, North Wales, operating at a maximum head of 545 m.

At the time of writing, the reversible pump turbines giving the largest output are installed at Bath County, USA, where six 457 MW units operate at a maximum head of 329 m. Next in terms of output come the units at Helms, USA, operating at heads of up to 540 m with an output of 400 MW per unit. Comparable outputs are obtained at Racoon Mountain, USA, where the operating heads are 310 m.

At the heads beyond the range of Francis turbines, where Pelton wheels are employed, application of pumped storage involves the use of multistage centrifugal pumps. Because of the need to ensure that the Pelton wheels are set above tailwater level, while high-head pumps require substantial submergence, the generator can no longer be used to drive the pump.

The Roncovalgrande plant in Hago Delio, Italy, operating at a maximum head of 747 m is equipped with eight Pelton wheels and a single pump. Similar arrangements have also been used at San Fiorano, Italy, where four Pelton wheels and one pump are installed. These operate at heads in excess of 1400 m.

Yet another solution to the problem presented by pumped storage at high head is presented by the use of multistage reversible pump turbines. An example of such an application is provided by the Edolo plant in Italy where eight such machines each rated at 132 MW operate at a head of 1287 m. In connection with the use of such machines it should be mentioned that, while their performance at or about the design point can be excellent, their output cannot be varied other than by natural changes in head. However, with eight units installed, the output of the scheme can of course be varied in stepwise fashion between 12.5 and 100%.

26.4.7.2 Mode of operation

Pumped storage plants provide the means for system regulation, peak lopping capacity and spinning reserve. Where a correct mix of thermal stations and pumped-storage plant has been installed, the system operator can maintain continuous output from base-load plant even at times of low-load demand. At times of high-load demand, the pump turbines provide additional output, and when operated in the spinning reserve mode they are capable of very quick response to system demands. Their use, unlike that of conventional hydroelectric units, does not in the main introduce the utilisation of renewable energy. Their economic and operational benefits arise mainly from a reduction of the high costs incurred in both fuel and in the consequences of physical damage incurred by thermal stations when operated in a cycling mode, such as ‘two shifting’, so common in the UK and elsewhere. They also ensure that the uneconomic use of gas turbines in their peak-lopping mode of operation is no longer necessary, thus removing the need for the provision of such facilities.

The rapid response capacity of pumped storage plant can be seen from consideration of the operating times, under automatic control, achieved at the Dinorwig power station:

Modern reversible pump turbines when starting in the pump mode are nowadays run up to speed and synchronised while the generator that acts as a motor is under the control of static, variable-frequency convertors. These convert the fixed voltage, mains frequency supply into a variable voltage, variable frequency output. Fixed-blade machines are normally started with the runner operating in the dry, in compressed air. Movable-blade runners are normally started with the unit watered, but with the blades fully feathered, to minimise resistance.

Starting a set for generation simply involves, after starting the essential auxiliaries, admitting water to the turbine. Where separate pump and turbine units are installed, like at Ffestiniog, the start of pumping involves: running the set up to speed by the turbine and then synchronising; closing the turbine guide and inlet valves so that the set is motoring; dewatering the turbine casing using compressed air; and opening the pump discharge valve. At Dinorwig changeover times from pumping to turbining under normal procedures take just under 10 min. In an emergency this time can be reduced substantially. It is claimed that at Dinorwig under an emergency procedure the changeover can be achieved in 90s. The changeover from turbining to pumping is much slower, mainly because the rates of acceleration of the unit in the pump start mode are controlled by the capability of the static, variable-frequency converter. However, as already pointed out, the initial reduction in load, from full to no load, is only a matter of 10 s or so.

26.4.7.3 Economics of pumped storage

The overall energy efficiency of a modern pumped-storage scheme is usually about 75%; the losses arise chiefly from the need for double energy conversion and from loss of water from the upper reservoir due to seepage and evaporation.

The running costs of the station comprise the cost of the energy used in pumping (including the above losses), wages and maintenance, and the costs of transmission losses arising from the remoteness of the station. Pumping energy can be purchased at off-peak times at between one-half to two-thirds of the price at which it can be sold at times of high demand.

Capital costs depend critically on the topography of the site and on the availability of efficient turbo-machinery of sufficiently high specific speed. The latter determines the size of the power station and the cost of generators.

Availability of pumped storage increases the efficiency of operation of the whole interconnected system and, because of its rapid response capability, removes the need for investment in gas-turbine based peak-lopping units and of operation of thermal units in the spinning reserve mode. Given the correct mix of pumped storage and thermal stations, the need for two shifting is largely eliminated.

In Europe, the USA and Japan, many pumped-storage plants have been built. Many more are projected or are in the course of construction. Pumped plants comprise the single most important area of growth of hydroelectric power generation.

26.5.1.1 Working cycles

The compression ignition engine operating on liquid fuels, or in the dual-fuel mode, works on the principle of fuel being injected into a charge of compressed air and spontaneously ignited by the high temperature of the induced air by the heat of compression. The process converts the heat energy of the fuel into mechanical work.

The two basic working cycles are four-stroke and two-stroke. These are represented diagrammatically in Figures 26.36 and 26.37 together with the appropriate indicator diagrams, which portray the events within the engine cylinder during each cycle.

In the two-stroke engine the working stroke occurs in each revolution of the crankshaft, whereas in the four-stroke engine it occurs once in every two revolutions. It does not follow that, because the two-stroke engine has twice as many power strokes as the four-cycle engine it will produce twice the power. The down stroke of the two-cycle engine (Figure 26.37) combines both power and exhaust strokes. As the intake and exhaust ports are cleared by the piston some mixing of fresh air charge and burned gases takes place (scavenging). Not all the burned gases are exhausted, which prevents a larger fresh charge of air being induced into the cylinder. The resulting power stroke has, therefore, less thrust.

In the four-stroke engine, however, nearly all the burned gases are forced out of the combustion chamber by the up-stroking piston. This allows almost a full air/fuel mixture to enter the cylinder since a complete piston stroke is devoted to induction of the mixture. The power stroke therefore produces relatively more power than its two-cycle counterpart.

26.5.1.2 Combustion

The major advantage of the reciprocating internal combustion engine is that its design is not limited by the properties of the materials of its construction, since none of its parts is required to work continuously at maximum-cycle temperature. This allows high maximum-cycle temperatures to be used, which results in a high thermal efficiency—this is a measure of the efficiency with which the fuel is burned during the combustion process to produce engine power.

Whilst compression-ignition engines are generally about 5% more efficient than their prime-mover competitors, there is appreciable variation amongst them in thermal efficiency. Much depends on the size of engine and the type of combustion chamber.

Combustion chambers are basically of two types: those designed for indirect injection and those for direct injection. The former employ pre-combustion chambers in the cylinder head into which a relatively coarse fuel spray is injected at low pressure. They are popular with European and American engine manufacturers and have the advantage of being able successfully to handle a wide range of fuels. When required to operate over a wide band of environmental conditions, however, they compare unfavourably with direct-injection chambers on fuel-consumption performance. Also, since heat loss from the pre-combustion chamber is high, cold starting can be difficult without prolonged cranking or recourse to external heating (such as glow plugs).

In direct-injection systems the underside of the cylinder head is usually flat and clearance volume on compression is mainly contained within the piston crown. Crown depressions are so shaped as to effectively induce swirled air turbulence, as the piston rises on its compression stroke. Fuel is then injected in the same direction as this flow of swirling air. The direct-injection principle is almost universally employed in modern medium-speed and on many high-speed engines.

Small-bore engines tend to have lower thermal efficiencies because their high surface-area/cylinder volume ratios give larger heat losses. Again, the greater heat losses from the larger exposed surfaces of indirect-injection systems means that they give lower thermal efficiencies than direct-injection ones. For these reasons, small indirect-injection engines may have thermal efficiencies as low as 28% whilst larger engines, particularly those using direct-injection techniques, may have efficiencies as high as 40%.

26.5.1.3 Pressure charging and inter-cooling

In the naturally aspirated four-stroke engine the working cylinder is almost (but not fully) charged with fresh air at atmospheric temperature and pressure at the end of the suction stroke. The density of this aspirated air regulates the weight of fuel which can be burned during the working stroke and this in turn determines the maximum power that can be developed. If a compressor were to be employed to supply the engine with intake air at a pressure higher than atmospheric, the mean effective pressure (and, therefore, the power output) of the engine would be increased without altering crankshaft speed or cylinder volume. This is effectively what pressure charging does. It can increase output by as much as 50% over an equivalent naturally aspirated engine of similar speed and dimensions. Furthermore, appreciable reductions are achieved at all loads in specific fuel consumption rates and the less arduous working conditions at the cylinders give increased engine reliability and reduced maintenance. On the debit side, however, it is impractical to expect a pressure charged engine to accept more than about 85% of its full load capability in one step in less than 10 s from crank initiation.

Several types of compressor (driven by chain or gearing from the crankshaft) are available, but since they deprive the engine of a portion of its shaft output, they are not as economical as the turbocharger which utilises the otherwise wasted energy of the engine’s exhaust gases. The turbocharger very simply consists of a gas turbine, driven by the exhaust gas flow, mounted on a common spindle with a blower or compressor placed in the air intake path. Figure 26.38 illustrates in schematic form the application of the turbocharger to a four-stroke engine. Further engine upratings are now being achieved in some medium-speed, four-stroke engines by employing two-stage turbocharging to give higher air intake densities.

The full potential of this increase in air inlet density by pressure charging is, however, marginally offset by an increase of air temperature due to adiabatic compression in the turboblower. This loss is recoverable by the use of charge air coolers (inter-coolers) placed downstream of the turboblower, which have the effect of increasing the fuel/air ratio, allowing more fuel to be injected into the cylinder and so raising the engine’s power output. The lower air intake temperature has the further effect of reducing not only the maximum cylinder pressure but also the exhaust temperature, and with it the engine’s thermal loading. Increase in engine power over a straight turbocharged model is usually of the order of 20–25% and thermal efficiencies of over 40% are obtainable.

Air-to-air charge cooling is usually carried out in a section of the engine’s radiator. An alternative water-cooled arrangement uses either a separate radiator circuit for charge-air cooling water or, in marine installations, sea-water-cooled heat exchangers with finned tubes carrying the sea-water and over which the charge air passes.

26.5.2.1 Basic classification

In generator drives, the compression-ignition engine is usually a multi-cylinder unit classified by the synchronous speed required, the type of fuel to be used, and the mechanical arrangement (i.e. the geometric arrangement of the cylinders).

26.5.2.2 Synchronous speed

Consistent with the output power required, the operating speed N (revolutions per minute) is determined by the frequency f (hertz) and the number p of generator pole-pairs in accordance with the relation N = 60f/p. Thus a two-pole generator must be driven at 3000 rev/min for 50 Hz or 3600 rev/min for 60 Hz; and for an eight-pole generator 750 or 900 rev/min.

26.5.2.3 Fuels and operating modes

26.5.2.4 Mechanical arrangements

Perhaps the most widely applied engine format is the vertical design in which the cylinder axes are perpendicular and in-line. In the medium-speed range vee-form engines offering high power in a small bulk are attractive, particularly for trailer-mounted and transportable generator plants. The included angle of the vee may range from about 35° to 90°. Other arrangements occasionally used are horizontal designs, opposed piston engines of various forms and vertical twin crankshaft or double-bank engines.

26.5.3.1 Fuel injection

Most modern compression ignition engines have a mechanical or airless fuel injection system embodying jerk-pumps. Medium-speed engines tend to use individual camshaft-actuated pumps for each cylinder. Higher-speed engines employ block pumps within which all the jerk-pump elements are incorporated and driven from a self-contained camshaft which, in turn, is coupled to an auxiliary drive from the engine. Certain two-stroke engines use the common rail system, wherein fuel is maintained at a constant pressure by a pump and a hydraulic accumulator. A fuel valve at each cylinder, driven and timed from the main camshaft, delivers the fuel to the engine.

The injector, the device that introduces the fuel by spray into the combustion space, is in essence a spring-loaded needle valve, whose tip covers the injector nozzle hole(s). The number of holes, their angles and the angles of spray are largely dependent upon the shape of the combustion chamber. Since fuel systems must have small passages and nozzle holes, it is of paramount importance that the fuel within these systems is well filtered.

26.5.3.2 Lubrication

The lubricant in an engine performs many tasks. In addition to reducing friction (and the potentially considerable power loss due to it) and minimising wear, its purpose is to provide:

It also affords a seal between piston rings and cylinder walls to reduce the seepage of gas that passes between the piston and cylinder walls from the combustion chamber into the crankcase.

Most engines use a pressurised or force-fed system to circulate the lubricant from an external drain tank or from a sump in the base of the crankcase. The main components of any system are: the circulating pump, which may be either of the gear or the multi-lobed rotor type; a pressure relief valve; oil filter(s) and an oil-to-engine coolant heat exchanger fitted between the feed pump and the filters. Delivery pressure is normally in the range of 50–200 kPa (0.5–2 kg/cm²) but it may even be as much as 400 kPa in high-speed engines.

In the so-called dry-sump system two pumps are employed. Oil that has circulated through the engine oilways returns by gravity to the crankcase pan. The task of the first pump is to transfer this oil from the pan to a reservoir tank external to the engine. The second pump draws oil from this tank and delivers it via the heat exchanger and filters to the bearings, etc.

The arrangement, mainly used on high-speed engines, in which the crankcase oil pan is in itself the reservoir is known as a wet-sump system.

Probably the most severe cylinder wear conditions occur just after an engine has been started, when piston lubrication is at its poorest. For this reason pre-priming systems are incorporated into both the manual start procedures and automatic start controls of low-speed and higher-rated medium-speed engines. It is also advisable to use periodic priming of lubricant on the larger high-speed engines, when these are operating in the automatic stand-by mode. Oil fed from a separate electric-motor-driven pump is circulated at periodic intervals through the oilways to flush the liners and generally wet the engine moving parts in readiness for an automatic start. Periodic priming also has the advantage of reducing the severe wear that can occur on cylinders due to condensed combustion products when an engine is at standstill.

26.5.3.3 Cooling

Engines may be either air cooled, using a mixture of air and oil as cooling medium, or water cooled with water and oil as the cooling media. Air cooling is simpler and is satisfactorily applied to high-speed engines up to approximately 400 kW output. Air is drawn into an impeller (usually secured to the engine flywheel or vee-belt—driven off the crankshaft) and discharged through shrouding across the finned external surfaces of the cylinder and cylinder heads. Whilst very low output engines may not require separate oil-to-air lubricating oil heat exchangers, they are standard features on larger engines.

Good installation of air-cooled engines is critical especially in high ambient conditions and in confined spaces. Care must be taken with the design and application of air intake and hot air outlet trunking to avoid, in particular, the possibility of hot discharged air being recirculated within the generator housing.

Water-cooling by circulation of water through cylinder jackets is the cooling method most frequently applied to engines in generating plants. Detailed arrangements vary considerably but most installations employ some form of closed-circuit jacket cooling system to transfer the engine heat to a heat exchanger. This may be a fan-cooled radiator or a shell-and-tube type heat exchanger. Thermostatic elements are incorporated within the systems to bypass the heat exchanger when starting from cold so as to allow the engine to attain its operating temperature more quickly.

Radiators may either be mounted on the same baseframe as their engine-generator assemblies, or remote mounted. Set-mounted radiators usually have engine-driven cooling fans, whilst remote units incorporate single or multiple fans driven by electric motors. At ratings below 1 MW, single or double sectioned radiators are employed. On larger engines multi-sectioned units are used to provide separate circuits for jacket water, lubricating oil and charge air cooling.

Since most modern medium- and high-speed engines are designed for high-temperature coolant conditions they employ pressurised closed circuit cooling systems. System pressures vary between 30 and 70 kPa (0.3–0.7 kg/cm²) in practice. They equate to system boiling points at sea level of 107°C and 115°C, respectively, so that engine makers may design for corresponding maximum operating temperatures at engine outlet of 80°C and 90°C.

Where a plentiful and cheap raw water supply of good quality is available at less than 30°C, it may be used as a secondary fluid to circulate through an engine’s tubular heat exchangers before it is run to waste. Pre-knowledge of the character and quality of the water is necessary to ensure that correct selection of materials for the tubes and shells of the heat exchangers is made.

Where a raw water supply is of dubious quality, clean (or treated) water may be used in a secondary closed circuit, which is in its turn cooled by the raw water in, for example, a cooling tower. Figure 26.39 illustrates diagrammatically an arrangement for a system of this kind. Whatever the system, it is advisable to use a high rate of jacket circulation with a small temperature difference between engine inlet and outlet, rather than a slow circulation and a large temperature rise.

26.5.3.4 Induction

Every engine should have air-intake filtration/silencing equipment. The induction system must be designed to supply clean dry air to the engine at as near ambient temperature as is possible and with the minimum of restriction. Engine makers stipulate the maximum permissible restriction at induction manifold or turbocharge inlet. The quality and quantity of the air supply has a direct bearing upon the engine output, fuel consumption and life.

Choice of filter depends upon the plant environment and the service life required. The following suggest the duty category in which the various filter types may be applied. The engine maker’s recommendations should be sought and complied with.

It is good practice to fit air restriction indicators to dry type filters to warn when the elements have reached a pre-set limit of fouling.

26.5.4.1 Starling equipment

The two main energy sources for engine starting are batteries for electric-start systems and air receivers for air-start systems. On small high-speed engines hydraulic energy or spring type inertia systems are occasionally employed. The simplest method of starting is of course manual cranking, but it is practicable only with the smallest engines. On the very largest low-speed engines small pony i.c. engines are sometimes employed to give crankshaft rotation at starting through a clutched pinion engaging with the engine flywheel gear ring.

Most electric-starting systems use motors fitted with Bendix-type pinions on their armature shafts to mesh with a toothed rim on the engine flywheel. Lower output engines in the high-speed range may require only one such motor whereas larger engines in that range, and medium-speed engines, need two motors with a common synchronising control.

Starter motors are either of the axial or co-axial type. In the former the complete armature assembly and pinion move forward axially to engage with the flywheel teeth. On the latter type only the pinion dome moves forward to engage under reduced power. This minimises engagement shock and reduces wear on the gear teeth. Starter windings may be of the ‘hold-on’ or ‘non-hold-on’ type. The non-hold variety is usually applied to motors for remotely or automatically started engines. The most frequently used voltages are 12 and 24 V, but 6 V for small engines and 32, 48 or 64 V for the top end of the electrically started range of engines are not unusual.

A cost-effective alternative to starter motors may be applied on generating sets below 15 kW rating, making use of a special d.c. starting winding within the generator to motor the engine. The same winding is used for charging the starter battery when the engine is running.

Starter batteries are either of the heavy-duty lead–acid type or alkaline type. The latter, in its nickel–cadmium form, has been favoured for stand-by generators because it retains its charge better over longer periods without use. Disadvantages are: that it is bulkier than the lead–acid battery of similar capacity; it is more expensive; and it tends to have a larger terminal voltage drop with heavy starting current drain. Maintenance-free, sealed storage batteries, using gas recombination technology, have been increasingly applied on stand-by plants in the last 5 years.

As a typical illustration the capacity of a lead–acid battery required to give six consecutive 20 s cranking periods with 5 s rest periods between each at ambient temperatures down to −7°C on an eight-cylinder vee-form engine rated at 600 kW is 236 A-h at the 5 h rate for a 25% discharged condition. At this temperature the steady cranking current demand at 80 rev/min is of the order of 1 kA corresponding to an engine cranking torque of 1 kN-m. Breakaway current and torque figures are as much as 150% above these values.

Starting difficulties, particularly in low ambient temperatures, are more usual on the smaller high-speed engines because their large surface area-to-cylinder volume ratios tend to dissipate the heat of compression. Moreover, restrictions on the size and weight of starting equipment limit the amount of starting torque available. Various proprietary starting aids are available: devices such as electric glow plugs or those using ether–air mixtures pumped into the air inlet manifolds to give access to the engine combustion chambers to promote combustion. Decompression mechanisms may also be used to hold off either the inlet or the exhaust valve on each cylinder, during the initial starting period. Once the engine is up to steady cranking speed, full compression is restored. The engine should then fire and run up to self-sustaining speed.

Compressed air for engine starting may be expanded either in an air motor engaging with the flywheel or directly within the engine cylinders, to move the pistons downwards on their working strokes until firing occurs. As in the electric starter the drive of the air motor is through a sliding pinion. Air motors may be applied to the range of engines that otherwise uses electric starter motors.

Direct air starting applies to the larger medium-speed engines and to low-speed units. Air is directed to each of the engine cylinders, in their proper firing sequence, through non-return valves either from a camshaft driven distributor or through mechanically operated valves.

Compressed air for either form of starting is stored in one or more receivers at pressures between 100 and 300 kPa (1–3 kg/cm²). Air charge is maintained by a small single-stage auxiliary compressor driven by an electric motor, an i.c. engine or by the main engine itself. On large or critical installations it is usual to provide back-up auxiliaries; the primary compressor being perhaps electric-motor-driven with an i.c. engine-driven stand-by unit.

26.5.4.2 Governors

Diesel generators use variable speed governors set to operate at the predetermined synchronous speed. Choice of governor is dictated by:

Governors vary from the simple all-speed mechanical type through various forms of mechanical-hydraulic and electro-hydraulic types to all-electric or electronic types.

The mechanical types use rotating flyweights to measure engine speed. The weights move radially and assume a position related to the speed of the engine. In the straight mechanical governor this position is directly translated to the fuel pump rack. Speed setting is usually fixed (or adjustable through only a very narrow speed range) and speed droop is non-adjustable. Pre-selection of speeder springs provides a choice of droop settings—between 4 and 12%. Most high-speed diesel generators up to 1 MW capacity use block type fuel pumps with an all-speed mechanical governor fitted to one end of the injector pump housing. Mechanical–hydraulic governors amplify flyweight movement through a hydraulic servo system to the fuel rack(s). Droop is readily adjustable from 0 to about 8%.

In the electronic governor, engine speed is measured through a magnetic pick-up usually mounted in the flywheel housing to detect the gear teeth on the flywheel rim. The signal so derived is compared with a potentiometer-set speed reference. Any detected difference is amplified within a control unit to adjust the signal to an hydraulic or electrical actuator fitted to the engine fuel rack to correct the fuel and return the engine to its preset speed. Response to speed changes is much faster than with the mechanical—hydraulic types and fully isochronous load sharing is possible on paralleled generators with these types of governor. Moreover, they may be used with a wide selection of control modules to provide fully automated and integrated multi-generator installations.

Where a governor is independently mounted from fuel injector pump(s) it is critical that the engine be fitted with some form of overspeed shutdown device, either acting directly on the fuel pump rack(s) or completely cutting off the intake air flow to the engine. This is necessary to prevent over-fuelling of the engine should any casual jamming of racks and fuel control levers take place or should any of the elements within the governing system fail.

26.5.4.3 Engine monitoring

It is essential that the strategic temperature, pressure, speed and flow parameters of an engine system are regularly monitored to interpret its behaviour and performance and relay this information to planned maintenance operations.

Maker’s instruction books give a good indication of the degree of instrumentation required. Much also depends upon the build specification for the particular engine but parameters such as those listed below may be monitored by instruments, wherever applicable.

Abnormal operating conditions may be detected by sensors whose output signals are fed into alarm/shutdown logic controls. Various combinations of indicative and protective action are possible:

The conditions to be covered may include:

This list is by no means exhaustive and the engine maker’s advice should be sought on the extent of protective insurance to be taken.

26.5.5.1 Construction

Types of construction and mounting arrangements are designated in BS 4999: Part 107 (IEC 34-7).

Almost all diesel-generator sets employ direct coupling of the prime-mover and generator. The latter may be treated either as a separate machine flexibly or solidly coupled to the engine, or flange-mounted to the flywheel housing and close coupled.

Most two-, four- and six-pole machines incorporate end-shield-mounted, grease-lubricated ball and roller bearings. Above about 3 MW, plain bearings are sometimes employed, self- or oil-scavenge lubricated. The large slow running machines tend to use single or dual pedestal outer bearings, mounted on a common baseframe with the generator stator casing.

26.5.5.2 Protection

Where generators are housed in buildings or canopies and enclosures, protection to IP23 of BS 4999: Part 105 is sufficient. (The technically equivalent standard is IEC 34-5.) Ventilation is provided by an internal shaft-mounted fan drawing air from the non-driving end and discharging it at the driving end. Dust filters may be fitted to this type of enclosure when the inlet air contains fine dust, sand, moisture or oil vapour. An output power reduction factor of about 0.95 should be applied to compensate for the resulting restricted cooling air flow. It is advisable to use thermistor temperature-sensing probes, either embedded in the stator windings or placed in the cooling air flow, to guard against damage by overheating, if the filters are not cleaned at regular intervals. For operation in extremely dirty conditions and on outdoor sites such as quarries, a totally enclosed construction, using closed-air-circuit ventilation, should be specified. The closed-circuit air is directly cooled by air-to-air (TECACA) or air-to-water (TECACW) heat exchangers, which may be attached to the generator casing. These enclosures are defined as IP45.

26.5.5.3 Voltage

In the UK the preferred output voltages for three-phase 50 Hz supply, and the appropriate ratings of engine-driven generators, are:

Voltages of 2.4, 4.2, 6.9, 13.2, 13.8 and 15kV are encountered in American and European practice.

BS 4999 requires the output voltage to be held to within ±5%, but most a.v.r.s will control to ±2 1/2% over the load range. Regulation down to ±1% is possible with solid-state devices in closed-loop control systems.

Standard classification for insulating materials is given in Chapter 7. Most machines up to 3 MVA employ class E or F insulation, or combinations thereof, for stator and rotor windings.

26.5.5.4 Parallel operation

Active power load-sharing is a function of the engine and its governing system. The sharing of reactive power is determined by excitation and synchronous impedance: proper sharing is obtained by applying quadrature current compensation (q.c.c.) to give automatic controlled droop of the output-voltage/reactive-current relation. The amount of droop should be the same for each generator paralleled into a power system: it is typically about 5%.

26.5.5.5 Short-circuit performance

Fast-response brushless and compounded machines can have subtransient short-circuit current levels of 6 to 10 times full-load current. But where excitation is derived from the output voltage, a short-circuit removes the excitation supply and the generator voltage collapses.

In distribution networks it is essential that an adequate ‘permanent’ short-circuit current is maintained by the energy source in order to allow discriminative operation between protective devices such as circuit-breakers, fuses and overcurrent releases. This is necessary if faults on final circuits are to be cleared as quickly as possible. Maintenance of short-circuit current at a level 2 to 3 times full-load current may be achieved by excitation power derived from current transformers in series with the generator output leads. This power is fed to the exciter field through a relay which closes when the output voltage collapses.

26.5.5.6 Generator selection

In selecting the right size of generator for a specific application the following factors, all influencing the rating, should be considered.

26.5.6.1 Planning

Factors affecting the choice and design of switchgear will be:

26.5.6.2 Fault considerations

Overcurrent protective devices must operate to isolate short-circuit faults safely, minimise damage to circuit elements and avoid, if possible, shutdown of plant. An accurate knowledge of prospective fault currents throughout the system is essential for the correct application of protective devices and the design of bus-bars and terminal arrangements to withstand consequential mechanical and thermal stresses.

Generator short-circuit fault current decreases from a high initial value determined by the subtransient reactance X″_d of the machine, through a lower value determined by the transient reactance X′_d, settling after 0.6–2.0 s to a steady-state level determined by the synchronous reactance X_d. Circuit-breakers and fuses should operate before the steady-short-circuit condition is reached.

System faults are also fed by synchronous and induction motors, which may generate by release of kinetic energy by their rotating masses. The fault contribution from an induction motor ceases after a few periods. Typical values of subtransient and transient reactance, in per-unit on a machine-rating base, are:

In the calculation of system fault levels it is sufficient to employ reactances only, except that in low-voltage (l.v.) systems the resistance of cables cannot be ignored. Arcing impedances should be included in l.v. system calculations. Typical values of per-unit arcing-fault current on 0.415 kV three-phase systems are: 0.70 for a three-phase arc, 0.57 for line-line and 0.14 for line-neutral.

26.5.6.3 Instrumentation and metering

Single running sets require an ammeter in each line (or an ammeter and a selector switch), a frequency indicator and a wattmeter. Optionally, a watt-hour meter and a power recorder may be fitted.

The instrumentation on each of any parallel running sets should include: an ammeter in each line (or a single ammeter with selector switch) a wattmeter and a reactive volt-ampere (var) meter. Power-factor meters may sometimes be substituted for var meters but as they develop low torque at low load they are prone to reading errors below 25% of rated current. The power factor calculated from kilowatt and kilovar values provides a more reliable and accurate indicator. Additionally, a set of synchronising instruments, perhaps mounted on a swivelling frame, is required, comprising a synchroscope, a double movement voltmeter and a double movement frequency indicator. Any incoming generator before being paralleled to the busbar should be connected by plug or switch to the ‘incomer’ movements of this voltmeter and frequency indicator, the second movements of both instruments being permanently connected to the bus-bars.

Recording instruments provide useful information for reconstruction of events, for registering operational patterns and indicating trends. In their multi-channelled form they give a time-sequenced record of events during disturbances and faults, particularly when they simultaneously monitor closing and tripping of key circuit breakers in the system.

26.5.6.4 Protection

The parameters to be monitored on the prime-mover have been dealt with in the preceding text. Protection related to the generator for which provision should be made, might in addition to system-fitted protective devices safeguarding against external short circuits and overcurrent, include the following:

Voltage surge protection should be considered where vacuum breakers or contactors are used, particularly with larger and higher voltage rated generators, if surges induced by lightning discharges on lines external to the diesel generator plant or initiated by switching are likely.

26.5.6.5 Control gear

Control gear should be designed to give a comprehensive indication of the state of a generator plant at all times and provide the means for modifying that state. Equipment for this purpose would include: measuring instruments, condition indicators, alarm annunciators, prime-mover and generator regulating controls and command signalling devices to engines and switchgear.

On the simplest single-generator installations, instrumentation and logic controls are usually incorporated within the switchgear cubicle(s) to give convenient operation of the plant from one location. On the more complex multi-generator plants, since adjustments on one machine affect all others in parallel with it, simultaneous observation of all the effects of any one action is essential. Where only two or three generators are involved, individual switchboards arranged as for the single generator may suffice, provided that they are placed as close together as possible and preferably in a quiet location away from the noise of running machinery.

But where adequate observation of a greater number of switchboards is difficult from one vantage point, it is advisable to use a control desk placed some distance in front of the composite switchboard. This may be used to house the generator regulating and control equipments whilst instrumentation and switchgear indicators are retained on the individual switchboards. Where size and cost of an installation warrants, a mimic diagram, representative of the complete generator plant and its feeder networks, may be incorporated into this control desk. Alternatively, it may be fitted into a separate diagram panel surmounting the composite switchboards. The diagram should incorporate pilot lights and semaphores to represent circuit-breakers, bus couplers, isolators and selector switches with miniature indicating instruments to cover feeders and generator output conditions.

26.5.7.1 Engine ratings

Since the i.c. engine is an air-aspirating machine its output is affected by changes in the temperature and the pressure of the air it breathes. Power ratings should be quoted to the Standard Reference Conditions (S.R.C.) contained in the applicable National Standards of the engine maker concerned. For British built engines this is BS 5514: Part 1, which is equivalent to ISO 3046/1. Although corresponding German (DIN 6270), American (SAE J-243) and Japanese National Standards cover the same technical subject matter, their treatment of it differs, particularly with regard to the standard reference conditions adopted and definitions of kinds of power.

BS 5514, which covers reciprocating i.c. engines using both liquid and gaseous fuels, applies the following standard reference conditions:

The ISO standard power rating of the engine has to be adjusted for ambient conditions falling outside the standard reference condition, to arrive at a predicted on-site or service power rating. Section 10 of BS 5514: Part 1:1987 sets out the method for re-rating RIC engines. It lists the formulae to be applied for this purpose; and Annexes B to G offer tables and examples to help simplify and explain the methods of calculation. Since the calculations are fairly complex (see Annex G of BS 5514: Part 1), most engine manufacturers reduce the formulae to simplified forms stating: the percentage reduction in power for a certain increase in temperature and altitude, above the BS 5514 SRC levels. As it is rare, in any part of the world, for high humidity to be combined with very high temperature, a de-rating exceeding 6% for humidity is seldom, if ever, warranted.

The power categories of importance to the diesel generator user are ‘continuous power’; and ‘overload power’, as defined in BS 5514: a footnote states that it is customary to permit an overload of 110% power ‘for periods and speed corresponding to the engine application’. Engine makers almost universally permit 10% overload for 1 h in any 12 of continuous operation on generator application. In so doing they rightly perpetuate a requirement of BS 649, which was withdrawn on publication of BS 5514 in 1977.

Prospective engine users should be wary of rating classifications other than those above, for generator applications, e.g. stand-by rating, continuous duty with time-limitation rating, reserve rating, intermittent rating and maximum rating, to name but a few of those cited. American, European and Japanese manufacturers will usually declare ISO standard power ratings, when asked.

26.5.7.2 Fuel oils

Liquid fuels used in compression-ignition engines are categorised as either distillate fuels or residual (blended) fuels. These are further subdivided into classes based upon viscosity, measured in seconds, using either the Redwood or the Saybolt systems. A fuel with a viscosity greater than 150 s Redwood No. 1 (expressed as 150 SRI) at 37.8°C is normally considered to be a residual oil.

BS 2869 lists nine classes of fuel oil, two of which, classes A and B, are specifically produced as engine fuels. Of the remaining seven, classes E, F, G and H are classified as industrial and marine fuels: E and F are light residuals or boiler fuels, G and H are heavy residuals or bunker fuels. The distillates, class A (gas oil) and class B (diesel oil), are the most widely used fuels on medium- and high-speed engines. Their American ASTM diesel fuel classification equivalents are D 975-66T Nos 1-D and 2-D. Grade 4-D of ASTM.D 975-66T, although considered a distillate oil, is intolerable to some high-speed engines and is best suited to low- and medium-speed units.

Almost all medium-speed engines will operate on the class E and F light residual fuels, whilst crosshead two-stroke low-speed engines and some of the slower running medium-speed four-stroke engines can successfully operate on class G and, to a lesser extent, class H heavy residual fuels. Residual fuels must be pre-treated, i.e. settled, heated, separated by centrifuge and filtered or waterwashed before being transferred to an engine fuel system. Heavy fuels should not be mixed with gas, distillate or light residual fuels.

The effects that fuel constituents have on engine performance and their influence on maintenance may be briefly summarised as follows:

26.5.7.3 Lubricating oil

Selection of lubricant depends on whether the engine has separately lubricated bearings and cylinders (low-speed and some medium-speed engines) or whether it has combined lubrication of bearings and cylinders (all high-speed and most medium-speed engines).

Where separate lubrication is used, specially formulated and refined mineral oils containing anti-oxidant additives are specified for bearings. Some engine operators favour detergent additives despite their tendency to thicken relatively rapidly. Cylinder lubrication calls for a heavier oil than that suitable for bearings, but on many engines the same viscosity grade is satisfactorily applied to both.

For combined lubrication, oils need: a viscosity applicable to both cylinders and bearings; high oxidation stability; sufficient detergency and corrosion inhibition properties. A big proportion of medium- and high-speed engines use a straight mineral oil of the type applied to separately lubricated bearings; but others, especially the highly rated high-speed engines, require heavy duty oils with detergent and contaminant dispersive properties, e.g. MIL-L-46152 specification for naturally aspirated engines and MIL-L-2104C for turbocharged units. If there is any possibility of an engine being subjected to prolonged light-load running, such as on many telecommunication transmitter applications, it is advisable to employ a heavy duty oil even though the engine might not normally require it.

Engine makers will always specify the oil to be used consistent with the duty and rating of an application and the fuel which it is intended to use. Mono-grade oils are normally recommended but makers will countenance the use of multi-grade types, provided that the operator obtains the oil supplier’s assurance that the type proposed meets the certified performance level of the equivalent and acceptable mono-grade oil.

26.5.7.4 Maintenance

The diesel engine and the gas turbine are both i.c. heat engines but since the former employs reciprocating motion it is mechanically the more complex machine and demands a greater skill in its maintenance. Also its wear rate is higher. However, repair services for diesel engines are more readily available world-wide than they are for gas turbines.

Plant operating personnel should be fully familiar with all its component parts and be capable of maintaining it in its optimum condition. Operation must be continuously monitored and accurate records kept so that incipient faults are detected and repair effected before major, unscheduled and costly stoppages occur.

The starting point of any preventive maintenance programme must be the manufacturer’s operator and service manuals. They indicate what needs to be checked and how frequently. Planned schedules should at first rigidly adhere to the maker’s recommended frequencies for inspection checks and maintenance tasks. Only after sufficient operational experience has been accumulated should one contemplate any modification to fit the particular installation and its operating conditions, and certainly not before the first general overhaul (stripping to crankshaft bearing level) has been undertaken. This gives the opportunity to assess achievement against the maker’s wear and renewal limit schedules.

Whilst the service intervals recommended by manufacturers are conservative and based on average experience and temperate conditions, periodicity of inspection and service may have to be increased where, for instance, the quality of fuel is in question or where corrosive and very dusty environments pertain, or where less than 50% loading conditions are initially expected. Over the last decade, engine and component manufacturers have achieved impressive improvements in times between overhauls. Typically, major overhauls on well-maintained high-speed engines may now occur at 15 000 h intervals, whilst those on medium-speed engines could be at between 20000 and 30 000 h—the longer interval applying to the slower running engines.

Careful consideration must be given to plant spares stocks. A more comprehensive holding of both non-consumable and consumable items would be necessary at remote sites or in certain developing countries, where there are no local accredited spares stockists. At locations where there is ready access to good stockists, a much-reduced inventory on consumables is justified.

Even with the most vigilant monitoring, problems can arise at any time. Running plants are usually most vulnerable immediately after commissioning and after general overhauls. Table 26.7 based on 4 years of analyses of the excellent Annual Working Cost and Operational Reports published by the Institution of Diesel and Gas Turbine Engineers, shows the major classes of failure (expressed as rounded-off percentages of total unscheduled stoppages) reported by over 100 diesel and gas engined plants, world-wide.

(1) Capital costs, embracing land; site preparation and access; foundations; buildings; workshops and tools; fuel storage; the power plant and its ancillary equipments; cranes; stores and non-consumable spares and heat recovery equipment (if applicable).

(2) Installation costs.

(3) Operating costs; fuel; lubricating oil; service spares; wages of operating and maintenance staff; insurance; depreciation, interest on capital and rates.

(4) Costs of any consequent supply outages.

(5) The size of the installation and its mode of operation:

26.5.9.1 Fuel costs

The cost of fuel to generate electricity only can be estimated as one-eighth of the cost of purchased electricity expressed in unit currency per kilocalorie (e.g. p/kal, where 1 kW-h = 0.863 kcal). Assume a consumer purchased 150000 kW-h of electricity over a given period at a total price of 1.5 p/kW-h (= 1.74 p/kcal). This would equate to a cost of purchased power of £2610 over the period. The cost of fuel to privately generate 150 000 kW-h would then be £326.

26.5.9.2 Lubricant costs

Lubricating oil consumption may be taken to be 1.5% of the fuel-oil consumption at full load. To this must be added the quantities representative of any oil changes at routine service intervals. This may vary from between 250 h to every 5000 h of running, depending upon the size and speed of the engine. The sump capacity of a 1.2 MW engine is of the order of 8001. For first estimates of lubricating oil costs, one may work on the basis of 5% of fuel costs for the same period.

26.5.9.3 Maintenance costs

Estimate on the basis of 5% of the combined costs of fuel and lubricating oil over the given period. This covers spares and labour.

26.5.9.4 Depreciation, interest, insurance and rates

Engineers do not always appreciate how substantial these costs can be. A recent economic appraisal for a private generating scheme, employing two 1.2 MW sets (with a third in reserve) operating at a 92% load factor and written down over a 7-year period, revealed that the annual sum of these costs was 70% of the combined total of the other running costs, i.e. those attributable to fuel, lubricant, maintenance, spares and labour.

26.5.9.5 Capital costs

Table 26.8 offers a very rough guide for first estimates on capital costs (including installation), using the cost of the engine generator and its auxiliaries as base 100.

(1) hotels;

(2) hospitals;

(3) leisure centres;

(4) small-scale district heating;

(5) sewage treatment farms; and

(6) industrial processes.

(1) The Electricity Council Engineering Recommendations G59;

(2) The Electricity Council’s publication ET 113—Notes on guidance for the protection of private generating sets up to 5 MW for operation in parallel with Electricity Boards’ distribution networks;

(3) British Gas publication IM 17—Code of Practice for natural gas fuelled spark ignition and dual-fuel engines; and

(4) H.M. Department of Energy, Energy Efficiency Office’s Good Practice Guide No. 1—Guidance notes for the implementation of small-scale packaged combined heat and power.