What each part does, the numbers that decide whether a quotation is credible, and where the machine stops being the right answer.
A Pelton runner looks simple: a wheel with a ring of spoons. At 600 m of head each spoon is struck twice per revolution by a jet at roughly 100 m/s, and the whole machine turns on how that jet is made, split and deflected.
Pressure becomes velocity in the nozzle, outside the runner, so the runner sees only a free jet at atmospheric pressure. No draft tube, no cavitation limit in the Francis sense, good part-load behaviour, runner above the tailwater.
| Pelton | Turgo | Francis | Kaplan | |
| Head (m) | 80 – 1600 | 20 – 300 | 10 – 600 | 2 – 70 |
| ns (m·kW) | 10 – 70 | 30 – 90 | 60 – 300 | 200 – 900 |
| Load at 30 % | 85 – 90 % η | 80 – 85 % | 70 – 80 % | 60 – 75 % |
The boundary against Francis is specific speed, not head. Below about ns = 60 at the required output a Francis runner gets narrow, disc friction and leakage climb, and the Pelton wins. Above it, the diameter needed to keep D/d sane becomes uneconomical.
Theory says u = 0.5 v, on the assumption of zero exit whirl; friction and imperfect axial exit push the real optimum lower. Chase the theoretical value and the runner runs slightly too fast, a point under the model test.
Figure 1 — Jet in, jet out. The nozzle converts pressure to velocity; the bucket only redirects it.
Force is F = ρQ(v − u)(1 − cos β2). At β2 = 180° the force doubles, and no production bucket does it: the leaving jet runs into the back of the next bucket. Practical buckets use 165°, recovering 96.6 % of ideal force with clean two-sided discharge. That 3.4 % is the price of a machine that survives.
Figure 2 — Bucket detail: working face, water inlet edge, water outlet edge.
A bucket is two half-ellipsoidal cups joined by a splitter that knives the jet in two. Three details carry the load: a splitter edge radius of 1–3 mm, sharp enough to cut but not to erode; the root cutout that lets both jet halves clear; the outer notch that stops the discharge clipping the next bucket.
Count follows Z ≈ 0.5 (D1/d0) + 15. Too few and water slips between buckets, too many and they crowd the jet. D1/d0 is the master parameter: 8–10 multi-jet, 10–14 typical single or twin-jet, 14–20 conservative high head.
Material is 13Cr-4Ni martensitic stainless throughout — ASTM A473 S41500 forged, 0Cr13Ni4Mo in Chinese designation, ASTM A743 CA6NM cast — at 250–300 HB. A forging has no shrink porosity at the splitter root, where stress and erosion peak, so forged plus CNC is defensible above about 400 m or on abrasive water.
|
Q = φ (π d02 / 4) √(2gH) |
With head fixed the only variable is jet diameter, set by needle withdrawal from the seat. That is the whole power control system.
Figure 3 — Injection mechanism. The needle slides axially inside the nozzle to change jet area.
The needle closes slowly, 20–60 s, because stopping flow in a kilometre of penstock can split the pipe. The deflector covers that gap: on a trip it swings into the jet within one to two seconds and throws the water to the tailrace without reducing penstock flow. The runner decelerates, the needle then closes slowly, the deflector lifts. A slow deflector overspeeds the unit; a fast needle overpressures the penstock. Both timings are set against the reflection period.
Figure 4 — Normal running (left) and deflector action (right).
A throttled needle also makes a poor jet, so multi-jet machines open jets one at a time. If load sits near 50 %, specify the sequence.
Figure 5 — Section through the casing. Runner, nozzle and deflector visible; tailwater leaves through the foundation trough.
A Pelton casing is a splash cover, not a pressure vessel. It is vented and sees atmospheric pressure plus spray, so the tailrace pit must take a full jet if a seat fails.
Horizontal with one or two nozzles suits small and medium units. Vertical with four to six takes over from about 5 MW, nozzles spaced to balance radial thrust. Jets raise specific speed as √z, so four jets run at about twice the ns of one at the same output: smaller, lighter, faster runner. That, not flow capacity, is why multi-jet dominates above a few megawatts.
Horizontal units: 80–800 m, 100 kW–20 MW, one or two nozzles. Vertical: 80–1000 m, 500 kW–20 MW, four to six nozzles. Runners are forged plus CNC or cast plus ground, matched to generators at 0.4, 3.3, 6.3 or 10.5 kV, 50 or 60 Hz.
The head range moves between pages — 80–800, 100–800, 80–1000, and 80–1600 m in one description — which are brackets for different constructions rather than contradictions. None of them says anything about your site. Send net head, rated discharge, unit count, frequency and voltage, and ask for a quoted point.
|
Metric |
1–20 MW class |
Ask for |
|
Peak efficiency (model) |
90 – 92 % |
IEC 60193 model report |
|
Guaranteed efficiency |
88 – 90 % |
guaranteed point + tolerance |
|
At 50 % / 25 % load |
88–91 % / 80–86 % of peak |
part-load curve per jet sequence |
|
Runaway speed |
1.8 – 1.9 × rated |
guaranteed, at slowest deflector travel |
Runaway reaches 1.8–1.9 × only if the deflector works, so size the generator and shaft against its slowest permitted travel. There is no NPSH limit: the runner runs in air, so Thoma sigma does not set elevation. Erosion is the wear mechanism that matters — needle tip, nozzle seat, splitter, bucket outlet edge, in that order. Above roughly 0.5 kg/m³ of abrasive solids, hard facing on needle and seat is not optional.
IEC 60193, IEC 60041, IEC 61362, IEC 61116, GB/T 15468, GB/T 10969, ASTM A473 and A743. What matters is not the list but whether the supplier will witness-test to IEC 60193 at a certified laboratory and accept IEC 60041 field testing as the basis of the efficiency guarantee.
Given H = 300 m, Q = 1.2 m³/s, 50 Hz:
1. v = 0.975 √(2 × 9.81 × 300) = 74.8 m/s
2. u = 0.47 × 74.8 = 35.2 m/s
3. A = 1.2 / 74.8 = 0.0160 m², so two jets of d0 = 101 mm
4. D1 = 2u/ω = 2 × 35.2 / 78.5 = 0.895 m at 750 rpm
D1/d0 = 8.9, Z ≈ 19, ns = 750√3178 / 3001.25 ≈ 34. That sits inside the 20–40 band where Pelton efficiency peaks, so the selection holds together. Had it returned 65, the layout would be wrong and Turgo or Francis the better question. Run this before comparing offers: a quoted runner diameter that will not reconcile with your head and speed is the clearest sign the offer was templated.
Below 60 m of head, where the runner grows large for the power. At large discharge and moderate head, where unit and civil cost overtake the machine saving. On abrasive rivers without a wear budget, since replacement buckets are not a stock item. And below 30 % load, where a double-regulated Kaplan wins.
Drawings from Hydrotu's impulse turbine technical material. Performance figures are indicative ranges for the 1–20 MW class; guaranteed values must come from a model test for the specific project.
コンタクトパーソン: Mr. Tu mingqi
電話番号: +86 13003602610
ファックス: +86-571-85194290
Japanese
