In many ways hydropower is the ideal alternative energy source. Its delivery is much more reliable, consistent and predictable than wind or solar energy. Modern turbines can extract up to 90% of the kinetic and potential energy in the water that passes through them -although an overall efficiency of 70% is more typical for small installations. More power is usually available during the wet winter months than in the summer, a supply profile that more closely matches normal demand. Unlike some other alternative energy technologies, hydropower installations generate many times more energy than is needed to build and run them; as much as 200 times more, it is claimed. Small hydropower installations are usually unobtrusive, quiet, emission free, and have little environmental impact.
Currently, around 20% of the electricity used across the world comes from hydropower. Some installations are very large - the largest, in South America, has an installed capacity of 12,600MW. Most are much smaller, producing little more than 5 kW. Definitions vary, but small hydropower installations are usually considered to have a generation capacity below 10MW. In some countries this dividing line is as low as 3MW. Mini hydropower definitions also vary. Usually the upper limit is 1MW; below 300 kW lies the category dubbed micro hydro . One other important distinction is also made: large hydropower installations use dams to create a high head of water above the turbines. The vast majority of hydropower installations, however, are run of river, i.e. no dams or significant water storage are involved (although weirs often are); water is taken straight from the river above the turbine and discharged back into it below.
Waterwheels have been used to extract energy from water for at least two millennia. For most of that time the energy abstracted was converted into mechanical energy - the wheels drove millstones, trip hammers and bellows in forges, looms, cranes, fulling hammers. By the end of the nineteenth century, the latest designs were achieving better than 60% efficiency. However, as generators of electricity, waterwheels leave a lot to be desired. In most potential locations maximum capacity will rarely exceed 5kW, but even if this is acceptable, the low rotational speed of the wheel means that a gearbox is needed to increase this to the levels at which electrical generators are effective. Nevertheless, on the smaller scale a waterwheel might well be an appropriate solution, and can also make an
aesthetic contribution to many projects. In use, waterwheels are simple to control, long-lived, and relatively cheap to make. A number of companies in the UK and US now offer modern steel waterwheels using off the shelf gearboxes and generators, which are said to be effective solutions on many projects, especially in the Third World (see Chapter 19).
Many of the earliest watermills used vertical axis wheels. A stream of fast-flowing water drove paddles mounted on a vertical shaft, which typically was connected directly to the upper of a pair of millstones. These mis-named Norse wheels appeared all over the world and continued in use for many generations, but were always limited in power and low in efficiency. Horizontal axis wheels could be made much larger, but needed a mechanism for converting their horizontal rotation into vertical rotation before they could be used to mill grain or crush olives - the Romans were the first to use cast bronze gearwheels to achieve this goal. Roman engineers also developed large installations of up to 16 wheels stacked up a hillside so that the water emerging from the tailrace of a higher wheel flowed straight onto the wheel below.
Choice of wheel type is largely a function of the head of water available at the site. The oldest, simplest and least efficient type is the undershot wheel, which utilises the kinetic energy of the water flowing beneath it, and can work with almost no head at all. Flat paddles are used, output is low and efficiency is 30% at most. If water levels drop in the summer the blades of an undershot wheel can be left dangling in mid-air. This disadvantage could be overcome by mounting the wheel on floating pontoons or between boats - these were often moored immediately downstream of multi-arch bridges, to take advantage of the increase in water speed as the river rushed between the bridge piers. A much later development is the Poncelet wheel, where the paddles are curved and the water is diverted into a pipe system and forced out as a jet that strikes the wheel at its base. This variant on the basic undershot concept can more than double the efficiency, but it requires a higher head of water than the simple undershot design.
Zuppinger undershot wheels manufactured by Hydrowatt of Germany have seen something of a renaissance. To be pedantic, these are not classic undershot wheels, being designed to work with heads of 0.5 m to 1.5 m. Installed wheels have ranged up to 7.5 m in diameter, producing 45kW at an estimated efficiency of 65%. These levels of efficiency are achieved by utilising as much as possible of what potential energy there is in the water flow: the flume closely follows the curve of the wheel downwards to the tailrace and fits closely around the wheel profile. Blades are curved and inclined 'backwards', minimising energy losses.
For heads of up to 3 m, the breast shot wheel was the popular choice. Water entered the wheel close to, or slightly above, its axis and was captured by bucket-like paddles. Potential as well as kinetic energy was available as a result. The wheel rotated in the same direction as the water in the tailrace, eliminating any energy-sapping counter rotation against the flow by the empty paddles.
Water flow - Flume
Overshot waterwheel schematic
This was a problem with the classic overshot wheel design, which otherwise was the most effective choice for higher head situations. Normally water enters the bucket type paddles just past the highest point of the wheel, rotating it in the direction of the incoming flow. After rotating through 100 ° or so the paddles empty and begin to turn against the flow in the tailrace. This problem can be eliminated and the efficiency of the wheel significantly enhanced by converting it into a backshot design - also known as a pitchback wheel. Water enters the paddles before the highest point, rotating the wheel in the opposite direction to the incoming flow. Water leaving the paddles flows in the same direction that the paddles are rotating, adding some kinetic energy to the potential energy already extracted by the wheel. It is claimed that a modern steel backshot wheel is more efficient than most available microturbines, and needs a much simpler supply and tailrace configuration (see below).
At their peak in the eighteenth and nineteenth centuries there were hundreds of thousands of waterwheels in operation across the world. Few are left and even fewer are still being used for their original purpose - but their ancillary works often remain. Weirs, mill leats and sluices tend to be long-lived and relatively simple to refurbish. Taking advantage of such existing infrastructure can be economically very effective. It can also be much easier to obtain permission from the relevant authorities to re-open a defunct site than to construct new infrastructure. It is currently estimated that there are around 20,000 former watermills in the UK alone that could be upgraded to modern standards without great difficulty.
All run of river hydropower installations, ancient or modern, have certain components in common. There is usually a weir across the river, to ensure that the water level is always higher than the intake for the water heading off to the wheel or turbine. This intake is usually protected by a screen or trashrack, to stop fish and debris entering the system. (One of the advantages of the overshot waterwheel is that it is very tolerant of foreign objects in the water flow, and therefore needs a less elaborate trashrack.) If the water then enters a leat, or small canal (also known as a headrace or lade), this screen could be located just before the forebay, a tank where sediment is allowed to settle out. From the forebay the water flows into the penstock, a pipe that takes it down to the powerhouse, and then into a tailrace which discharges it back into the river.
Sometimes there is no leat, and the water passes straight from the river into the forebay and the penstock. Sometimes the leat stretches right to the powerhouse - this is usually the case on most former waterwheel sites. Leats need spillways and sluices to control water levels. An alternative where the weir creates a reasonable difference in water level
above and below is the barrage option. Here the turbines are installed almost immediately downstream of the weir, usually without forebay and only a minimal penstock. As no water is abstracted from the river proper there are fewer regulatory hurdles to be surmounted. This type of installation is now being tried in the outflows from both water and sewage treatment plants, with considerable success.
past the turbine
A simple layout without a forebay can be used where silt is not a problem past the turbine
A simple layout without a forebay can be used where silt is not a problem
The most important factor in any installation, however, is the head, the vertical distance between the intake and the wheel or turbine. Gross head is the actual distance; net head is the effective head at the turbine intake, which will normally be less than the gross head due to losses to friction in the penstock.
High head hydropower installations are usually classified as those with heads above 50 m, medium head would be between 10 and 50 m, and low head would be below 10 m. Most former waterwheel sites will have heads of 5 m or less, although overshot or backshot wheels over 20 m in diameter were built. There is a limited choice of turbine types which can function with such low heads, but there are commercially available units that can extract useful power from heads as low as 2 m, provided there is enough water passing through them.
Almost as important as head, therefore, is flow. The key value is the available flow; that proportion which can be abstracted from the river to drive the turbine. This value will normally vary throughout the year, so the most critical information is contained in the flow duration curve, from which the potential output and capacity factor of the installation can be calculated. Typically, the capacity factor for a mini hydropower installation is greater than 50%, significantly higher than wind or solar power.
All turbines fall into one of two categories. Impulse turbines depend on the potential energy of the water being converted into the kinetic energy of a high velocity jet of water. The turbine runs in air and is rotated by the jet striking its blades or buckets. In the classic Pelton design the jet is aligned with the plane of the turbine wheel - or runner, in the Turgo
variant the jet is angled, reducing the effect of water splashing back from the runner and affecting the incoming jet. Crossflow turbines - also known as Michell-Banki or Ossberger turbines - have horizontal axis drum shaped runners, and act rather like a sophisticated overshot waterwheel. A rectangular nozzle directs the flow across the curved blades mounted around the rim. Water enters at the top of the wheel and passes through the blades again as it leaves. Crossflow turbines are relatively simple and cheap to construct and easy to maintain.
Reaction turbines are normally more expensive than impulse turbines because their runners are usually enclosed in pressure casings. Runner blades are carefully profiled to extract kinetic and potential energy from the water. Water pressure falls back to atmospheric as it flows through the runner and out through the obligatory draft tube, which reduces water velocity and lowers pressure across the runner. The first true reaction turbine was the Fourneyron, invented in 1827, a vertical axis outward flow design using blades curved in only one direction. It was something of a sensation at the time, as it could operate at more than 2,000 rpm and achieve 80% efficiency. In 1895, Fourneyron turbines were used to tap the energy of Niagara Falls, its high rotational speed making electricity generation much simpler.
Inward flow reaction turbines are inherently more efficient than outward flow designs, and from the late nineteenth century onwards the Fourneyron design was superseded by the Francis turbine, first invented in 1849, and still the most common turbine used worldwide. Water flows tangentially into the runner through a series of adjustable guide vanes and out along the runner axis. Early installations were open flume - the runner with its guide vanes and draft tube were simply immersed in the supply channel, leat or headrace, with the draft tube turning through 90° to discharge at a lower level. Later a spiral casing was added, which helped accelerate the flow into the runner. Later still came various forms of propeller turbines, which, as the name suggests, work like ships' propellers in reverse. Smaller versions generally have fixed pitch blades: on larger installations the Kaplan variant is usually a better choice. This has variable pitch blades, enabling it to work efficiently over a wider range of heads and flows.
Recently there has been considerable interest in and research into the concept of using readily available water pumps in reverse to generate power rather than consume it. The attraction is the significant potential cost savings. Turbines are virtually hand built in small numbers, pumps are mass-produced, and are generally rugged and reliable. Calculating the potential performance of a pump as turbine is far from straightforward, however, and some types are basically unsuitable. Pump as turbine installations have demonstrated efficiencies as high as 90% when used with high heads.
Measuring the gross head at a particular site is relatively simple, and this helps in a preliminary assessment of the likely optimum type of turbine. Potential suppliers would need more information than this, of which the key item is the flow duration curve (FDC). Obtaining a reliable flow duration curve for a particular intake location takes time. Ideally, measurements of flow are taken over a period of at least three years. A reasonably reliable curve can be derived from national gauging stations or from hydrological records, although these days allowance has to be made for the likely effects of global warming on rainfall patterns. A flat FDC is the optimum, indicating a largely spring-fed river with low risk of flooding. The authority responsible for the catchment will determine the volume of water that can be diverted into the hydropower installation. It will demand that a certain minimum compensation flow is maintained through the river at all times, to minimise the environmental impact. On low head run of river schemes the head will tend to vary with river level as the changes in level at head and tail of the scheme are never exactly in phase.
Typically the design flow for the site will be the mean river flow averaged over several years. Peak power in kilowatts can be estimated as 7 x design flow in cubic metres per sec x head in metres. Energy output will be a function of the capacity factor, which in turn will depend largely on the size and type of the turbine selected. In essence, the choice is between a larger installation which can handle high flows but will be working at less than full capacity for much of the year; and a smaller, less expensive unit which will be working harder for longer but will generate less electricity overall. The first will have a lower capacity factor and a lower rate of return on the original capital investment, but will be less stressed, last longer and probably be cheaper to maintain.
Another important factor is how well the turbine copes with low flow conditions. Pelton and Kaplan turbines retain high efficiencies when running well below design flow, cross-flow and Francis turbines lose efficiency sharply below 50% of design flow, fixed pitch propeller turbines need to operate above 80% of design flow. There are several possible ways of mitigating this problem. One of the most effective is to use multiple turbines, so that the flow can be directed away from some to keep the remainder running at high efficiency. (This approach also simplifies maintenance, as individual turbines can be taken off line without interrupting power generation.) The drop off of efficiency with flow on cross-flow turbines can be minimised by concentrating what flow is available onto a smaller section of the perimeter. On Kaplan turbines the runner blade pitch can be adjusted to suit. The flow duration curve should indicate how much of a problem low flow is likely to be, so that the appropriate installation can be selected.
Turbine speed was traditionally controlled by complex mechanical governors that opened and closed sluices and gate valves to vary the flow of water into the turbine. However, these are not very practical for smaller installations. Luckily, a more convenient modern alternative now exists. Electronic load controllers (ELC) effectively add an artificial load as needed to maintain total load on the turbine at its design level. This artificial load produces surplus electricity, which can be used for any number of purposes.
Generators also have optimum revolutions per minute (rpm), typically around 1,500. Some turbine types are relatively slow running, particularly in low head installations, so some form of ' gearing-up' is needed between turbine and generator. This is one reason why faster running propeller-type turbines have taken over from Francis turbines at the smaller end of the market. Generally, single-phase alternating current is produced by installations below 20 kW. Above this, three phase generation is the norm. Until recently, turbine speed had to be closely controlled to maintain the frequency of the supply at the desired level. However, turbines fitted with direct drive permanent magnet synchronous generators and electronic power conditioning, which allows them to operate over a much wider speed/flow range, are now available.
Using pumps as turbines (see above) leads logically to the use of motors as generators. Induction motors can be run above their rated speed to generate both single-phase and three phase current. Such off the shelf units can be significantly cheaper and still produce highly effective results. Integral gearing means they can be successfully coupled with modern waterwheel designs, such as the Pedley Wheel, which is claimed to be capable of producing up to 20kW from a 6m diameter wheel (see Chapter 19).
Hydropower is a mature technology and there is a wide choice of equipment suppliers. Development continues, however, with an increasing use of polymers to produce both runners and pressure casings. Modular construction is becoming the norm, and there is an increasing emphasis on rapid installation and minimal maintenance. Minimal visual impact is also a priority. Almost any desired output is readily available, from picohydro units with outputs of less than 1 kW upwards; and there is a wide choice of turbines with ratings of 10 kW to 50 kW. Many of these are specifically designed for low head situations where the risk of flooding should not be ignored, and are proofed and protected against prolonged immersion.
Assuming a capacity factor of 50%, a typical small office building consuming 200,000 kWh per annum would need a 45-50 kW turbine to supply 100% of its power. Reaction turbine units are available that can produce this sort of power from a 5 m head and a flow rate of around 1500 l/s (litres per second) - alternatively, five similar units could provide the same output from a head as low as 2 m and flow of 5000 l/s. Even with only 200 l/s and a head of 2 m a single reaction turbine could produce more than 2kW, which might be best employed to power a water, ground or air source heat pump (see Chapter 12).
Grid connection, if available, is still worth considering even when there is plenty of hydropower potential. Intakes can become blocked, turbines can break down, the powerhouse could be flooded. There are other benefits, as detailed in Chapter 1. In either case, hydropower will always be a serious option for those projects where the opportunity exists.
Was this article helpful?
Do we really want the one thing that gives us its resources unconditionally to suffer even more than it is suffering now? Nature, is a part of our being from the earliest human days. We respect Nature and it gives us its bounty, but in the recent past greedy money hungry corporations have made us all so destructive, so wasteful.