Hydro · Choosing a turbine
Head and flow between them narrow the choice to one or two turbines before preference enters into it. What is left after that is a judgment about how much the stream changes through the year, and whether the machine can follow it.
Two families
Every turbine in micro-hydro belongs to one of two groups, and the difference is physical rather than cosmetic.
A nozzle converts pressure into a fast jet, and that jet strikes the runner in open air. The Department of Energy describes impulse turbines as the least complex design and the type most commonly used for high-head micro-hydroMicrohydropower Systems">[1].
The runner is not submerged, the casing carries no pressure, and the whole assembly is simpler to build and inspect. Pelton, Turgo and cross-flow machines are all impulse designs.
The runner sits full of water and every blade stays in contact with it. Power comes from the pressure drop across the blades rather than from a jet. Reaction turbines are highly efficient and suit lower head with higher flow[1].
Propeller, Kaplan and Francis machines belong here, along with pumps run backward. They are more complex, they need a sealed casing, and on low-head sites they carry a draft tube that recovers energy below the runner.
How selection actually works
Turbine selection is conventionally based on specific speed, a non-dimensional parameter combining head, output power and shaft speed. From it, engineers derive application domains: regions on a chart of head against flow where each machine type belongs[2].
The result sorts into three bands. Pelton and Turgo turbines at high heads, cross-flow and radial Francis turbines at mid heads, propeller turbines and waterwheels at low heads[2]. Commercially available machines follow the same pattern, which is the practical consequence for anyone shopping.
Pelton and Turgo. A small jet moving fast. The machinery is compact, the pipe is narrow, and the cost per watt is the lowest available.
Cross-flow and Francis. More water at moderate pressure. Cross-flow in particular is common in small hydro because it tolerates a wide operating range.
Propeller and Kaplan, plus the waterwheels from the previous guide. Large volumes at low pressure, and machinery sized to match.
These domains are conventions rather than physical walls. The Bristol study that produced the chart above went on to find that a single-jet Turgo, usually reserved for medium and high head, scored best for a specification at 1.5 to 3.5 meters of head, because at pico scale the qualitative factors matter as much as the efficiency curve[2].
For a household the practical reading is that if your numbers land in an overlap, more than one machine will work, and the decision moves to availability, serviceability and how the stream behaves in August.
The four machines
Specially shaped buckets around the rim of a disc, struck by one or more jets. Each bucket is split into two halves so the center does not become a dead spot unable to deflect the oncoming jet, and the lower lip is cut away so the following bucket enters the jet cleanly[3].
The classic high-head machine and the simplest to understand by looking at it. Its efficiency holds up well as flow varies, which matters on a seasonal stream. Adding jets extends it toward lower head, though at pico scale the extra pipework to feed multiple jets costs volume and gets penalized on compactness[2].
An impulse machine like the Pelton but designed for higher specific speed. The jet strikes the plane of the runner on one side and exits the other, so water passes through rather than reversing[3]. That lets a given runner size pass more water than a comparable Pelton.
The Department of Energy notes it is less bulky than a Pelton, needs few or no gears, and has a good reputation for trouble-free operation, working at low flow but needing medium or high head[1].
The tradeoff: a Turgo runner is harder to manufacture than a Pelton, and its vanes are more fragile than Pelton buckets[3].
A drum-shaped runner made of two parallel discs joined near their rims by curved blades. A rectangular nozzle directs water along the full length of the drum. The water strikes the blades, passes through the interior, and strikes them again on the way out, giving up most of its energy on the first pass and a smaller amount on the second[3].
The shaft is always horizontal, unlike Pelton and Turgo machines which can be mounted either way[3]. Strictly it is an impulse turbine, though pressure forces are involved and a mixed-flow description is more accurate.
Its reputation in small hydro rests on tolerance rather than peak numbers. The drum can be divided so that part of it takes low flows and the whole of it takes high ones, which is why cross-flow machines suit streams that change through the year.
Reaction machines resembling a boat propeller running in reverse, with three to six blades set at different angles on the runner. Bulb, tubular and Kaplan tubular are variations, and the Kaplan, with adjustable blades, is described by the Department of Energy as a highly adaptable propeller system usable at micro-hydro scale[1].
These are the low-head answer. The Bristol study found a propeller turbine with a draft tube to be the best solution for heads between 0.5 and 1.5 meters, and reaction turbines generally to have superior power density, meaning more watts from a smaller machine[2].
A fixed-blade propeller loses efficiency quickly away from its design flow. A Kaplan's adjustable blades address exactly that, at the cost of mechanism.
Worked example
The property followed through this section has 54 feet of net head and diverts 150 gallons per minute. In the units the charts use, that is 16.5 meters of head and 0.0095 cubic meters per second, producing about 810 watts.
Very little water, a useful amount of drop, and an output well inside the pico range, which the literature defines as generation under 5 kilowatts[2]. That combination lands squarely in impulse territory, and specifically in the region where Pelton and Turgo overlap.
A propeller or Kaplan turbine wants high flow at low head. At 9.5 liters per second there is nowhere near enough water to turn one usefully. Waterwheels are out for the same reason the previous guide gave: 16.5 meters is far past what a wheel can use.
At 16.5 meters of head a jet leaves the nozzle near 18 meters per second, and a Pelton runner works best at roughly half the jet speed. To turn at 1,000 to 1,800 revolutions per minute, the runner would be somewhere between 95 and 172 millimeters across. Commercially available generators run between about 200 and 3,000 rpm[2], so this site can drive one directly.
Both work here. A Pelton is simpler to obtain and repair and holds efficiency well across varying flow. A Turgo passes more water for a given runner size and is more compact, at the cost of a more fragile runner that is harder to replace. On a site with this little flow, the Pelton's tolerance and repairability are the stronger argument.
As head rises, impulse turbines spin faster, which removes the need for a gearbox and raises overall system efficiency[2]. The waterwheel guide showed the opposite case, where 10 revolutions per minute demanded 180 to 1 gearing and gave away a quarter of the output. A site with real head buys its way out of that problem, and this one does.
The criterion nobody checks
Manufacturers quote a peak. A peak is achieved at one combination of head and flow, and a stream visits that combination for part of the year at best. What determines annual energy is the shape of the efficiency curve either side of that point.
A machine holding 70 percent from a quarter flow to full flow will out-produce one that touches 85 percent at design flow and collapses to 40 percent in August. The first number sells the machine. The second one runs the house.
The Bristol analysis treated part-flow and part-head efficiency as its own weighted criterion alongside rated efficiency for exactly this reason[2]. It deserves the same weight in a household decision.
The third question matters because penstock friction is an estimate until the pipe is in the ground. A machine that falls apart when head comes in slightly low is a machine that has left you no margin for the one number you could not measure precisely.
The fifth option
Reverse the action of a centrifugal pump and it behaves as a turbine. Because pumps are mass produced for every conceivable flow and pressure, they are far easier to obtain than purpose-built micro-hydro turbines and considerably cheaper[1].
That availability argument is stronger than it sounds. Only a few companies build micro-hydro turbines, most of what they build is high-head, and low-head low-flow machines can be hard to find or need to be made to order[1]. A pump can be bought locally and replaced the same way.
The Department of Energy states the conditions plainly. The site must have fairly constant head and flow for a pump to perform adequately, and pumps are less efficient and more prone to damage than a matched turbine[1].
Spring-fed sources and piped supplies with a steady year-round flow, which is precisely the constant-conditions case the caution describes. Parts availability in a remote place is a genuine engineering consideration, not a compromise.
Streams with a wide seasonal swing. A pump has no adjustment for varying flow, so it runs off its best point for most of the year. On the creek in this section, with flow varying several times over between seasons, it would be the wrong choice.
Common mistakes
The headline percentage describes one operating point. Annual output depends on the whole curve, and the stream decides which part of it you live on.
A turbine matched to snowmelt spends summer far below its design point. The machine should be matched to the flow you measured in the driest month, with any surplus spilled.
The turbine sees net head. Ordering against the hose-and-gauge figure specifies a machine for a site that does not exist once the penstock is carrying water.
Sediment is abrasive and jets are fast. On a silty stream, runner material and the ease of replacing the runner matter more than the last two points of efficiency.
If head and runner size can put the shaft inside a generator's operating band, direct drive removes a loss, a maintenance item and a failure point. It is worth asking for explicitly.
Pipe diameter sets net head, and net head sets the machine. Deciding the turbine first and the pipe second is the wrong order, and it is the usual one.
Next
Every figure on this page assumed a net head, and net head is decided by the pipe. The intake and penstock guide covers screens that clean themselves, sizing the pipe so friction does not eat the site, and the freeze protection that keeps all of it running through January.
After that comes the electrical side, where a turbine that cannot simply be switched off has to be given somewhere for its power to go.
Where the turbine sits in the chain, and why the penstock decides the budget.
Read the guideThe low-head alternative, and the speed problem that turbines avoid.
Read the guideThe two numbers this entire page reads off. Measure them before shopping.
Read the guideSources
Head, flow and runner figures in the worked example are illustrative and carried forward from earlier guides in this section. Application domains are engineering conventions with real overlap at their edges, not fixed limits. Confirm performance at your own measured head and flow with the manufacturer before ordering, and ask for the part-flow curve rather than the peak figure.