Harvesting moving fluids
Water wheels and windmills were humanity's first engines. Vertical-axis windmills ground grain in Sistan, on today's border between Iran and Afghanistan, more than a thousand years ago; water wheels powered mills across the Roman world, China and medieval Europe; Dutch windmills drained a country. Their descendants are some of the most refined machines engineers build. In 2025, according to Ember, hydropower generated about 14% of the world's electricity and wind about 8.5%, with wind and solar together at 17.3%. A modern wind turbine's blades are longer than a football pitch, and the largest hydro turbines, at China's Baihetan dam, each deliver 1,000 MW.
Lumenary EE's renewables and storage derives the Betz limit and follows wind and hydro power into the grid. This lesson takes the mechanical engineer's view: the forces on a rotor and its tower, how a blade is shaped and why, how the shape determines the power curve and the energy a site will yield, and how hydro turbines are chosen for very different rivers and mountains.
A wind turbine and a hydro turbine do the same job, taking energy out of a moving fluid, but from opposite ends of a scale. Wind is a thin fluid at low pressure spread over a huge area, so a wind turbine must be enormous and light; water is dense and can be gathered behind a dam at high pressure, so a hydro turbine can be compact and immensely strong. Most of the differences between them follow from that.
Momentum theory: power, thrust and the Betz limit
Model the rotor as an actuator disc: a thin, permeable disc of area that extracts energy uniformly from a steady, incompressible stream of speed . The air slowing down ahead of the disc must spread out, so the flow through the disc is a stream tube that widens downstream. Define the axial induction factor by the speed at the disc, .
Theorem— Momentum theory of an ideal rotor
For an actuator disc in a steady, incompressible, frictionless stream, the speed at the disc is the average of the far upstream and far downstream speeds, so the wake speed is . The thrust and power are is largest, (the Betz limit), at , where .
Proof
Apply the momentum equation to the stream tube: with mass flow and wake speed , the thrust is . Apply Bernoulli's equation separately upstream and downstream of the disc (not across it, where energy is removed): the pressure drop across the disc is , and . Equating the two expressions for gives , so and . The power is the thrust times the speed at the disc, . Setting gives . Lumenary EE's renewables and storage gives the same derivation through the kinetic energy of the stream.
∎
The thrust is the mechanical engineer's first concern, because it acts at the top of a tower. The program tabulates the two coefficients and computes the thrust and overturning moment for a 120 m rotor on a 100 m tower.
import numpy as np
# One-dimensional momentum theory of an ideal rotor (an "actuator disc").
# a is the axial induction factor: the wind slows from V far upstream to V(1 - a) at the disc.
print(" a Cp = 4a(1-a)^2 CT = 4a(1-a) wake speed / V")
for a in (0.1, 0.2, 1 / 3, 0.4, 0.5):
print(f"{a:6.3f}{4 * a * (1 - a) ** 2:13.3f}{4 * a * (1 - a):16.3f}{1 - 2 * a:15.2f}")
# What the rotor does to the tower: thrust T = CT * 1/2 rho A V^2 acts at hub height.
rho, R, hub = 1.225, 60.0, 100.0 # kg/m^3, rotor radius (m), hub height (m)
A = np.pi * R**2
for V, CT in ((10.6, 0.8), (25.0, 0.1)): # near rated wind; at cut-out with the blades pitched
T = CT * 0.5 * rho * A * V**2
print(f"V = {V:4.1f} m/s, CT = {CT}: thrust {T / 1e3:5.0f} kN, "
f"overturning moment at the base {T * hub / 1e6:5.1f} MN m")Beyond the rotor extracts less power but pushes ever harder on the air; at the wake would stop completely, which is impossible, and above about real rotors enter a turbulent wake state that momentum theory cannot describe. Engineers then switch to empirical corrections first proposed by Hermann Glauert in the 1920s. Near rated wind, the 120 m rotor pushes on its tower with about 623 kN, a moment of 62.3 MN m at the base, which the tower, its foundation and every bolt between them must carry, millions of times over the turbine's life as the wind gusts.
Those two thrust cases are only the start of a load analysis. The IEC 61400-1 standard defines dozens of design load cases: normal turbulence, extreme gusts and changes of wind direction, faults during operation, emergency stops, and a parked rotor in the strongest gust expected in 50 years, which for a parked turbine can produce larger tower loads than anything in operation. Fatigue, the accumulation of 10⁸ or more load cycles, often sizes the blades and tower more than any single extreme load.
Rotation and blade elements
A real rotor turns. Its tip-speed ratio compares the speed of the blade tips, , with the wind; large three-bladed turbines work best at to 9, so in a 10 m/s wind their tips move at 70 to 90 m/s. Turning the rotor costs something: by Newton's third law, the torque the air exerts on the blades is matched by a torque the blades exert on the air, so the wake rotates, and the kinetic energy of that rotation is lost. The faster the rotor turns for a given power, the smaller its torque and the smaller this loss, which is one reason modern rotors are fast and slender rather than slow and many-bladed like an old farm windpump.
To design a blade we cut it into short blade elements at radius and treat each as a two-dimensional aerofoil. The element meets a relative wind made of the axial wind at the rotor, , and the speed at which the blade moves through the air, , where is a small tangential induction factor from the wake's rotation. The relative wind arrives at the inflow angle to the rotor plane, with . The blade's chord is set at the twist (plus pitch) angle , so the aerofoil sees an angle of attack .
Lift acts perpendicular to and drag along it. Resolving them gives a force along the rotor plane, which drives the rotor, and a force along the axis, which is thrust: Near the tip is large and small, so the blade is set nearly flat to the rotor plane; near the root is small, is large, and the blade must be twisted steeply to keep the angle of attack near its best. That is why every wind-turbine blade is twisted, typically by 20° or more from root to tip, and why its chord shrinks towards the tip: the outer blade moves fastest and needs less area to produce the same force.
A blade element momentum solver
Blade element momentum (BEM) theory joins the two pictures. For each annulus of the rotor, momentum theory gives the thrust and torque in terms of and , and blade-element theory gives them in terms of the aerofoil's lift and drag at the angle of attack, which itself depends on and . Equating the two gives a pair of equations for each annulus, solved by iteration: where is the local solidity ( blades of chord ), and is Ludwig Prandtl's tip-loss factor, which accounts for air leaking round the tips of a rotor with a finite number of blades.
The program designs a 120 m, three-bladed rotor with the classic optimum chord and twist for a tip-speed ratio of 8 (capping the chord at 4.5 m, as road transport demands), gives it a simple aerofoil with a lift slope of per radian and a lift-to-drag ratio of about 100, and computes and across a range of tip-speed ratios.
import numpy as np
# Blade element momentum (BEM) theory for a three-bladed rotor, 120 m in diameter.
B, R, r_hub = 3, 60.0, 6.0 # blades, tip radius and root radius (m)
lam_design, cl_design, alpha_design = 8.0, 1.0, np.radians(6.0)
def polar(alpha):
"""A simple aerofoil: lift slope 2*pi, zero lift at -3.5 deg, stall above 12 deg."""
a_stall = np.radians(12.0)
cl_max = 2 * np.pi * (a_stall + np.radians(3.5))
cl = np.where(alpha < a_stall, 2 * np.pi * (alpha + np.radians(3.5)),
cl_max - 1.5 * (alpha - a_stall)) # lift falls after stall
cd = 0.010 + 0.01 * alpha**2 + 0.6 * np.maximum(alpha - a_stall, 0.0)
return cl, cd
# The blade: the classic optimum shape for the design tip-speed ratio, 40 elements.
dr = (R - r_hub) / 40
r = r_hub + dr * (np.arange(40) + 0.5) # mid-radius of each element
lam_r = lam_design * r / R
phi_d = (2 / 3) * np.arctan(1 / lam_r) # design inflow angle, with wake rotation
chord = np.minimum(8 * np.pi * r * (1 - np.cos(phi_d)) / (B * cl_design), 4.5) # capped for transport
twist = phi_d - alpha_design
def bem(lam, pitch=0.0, V=1.0, rho=1.225):
"""Power and thrust coefficients (Cp, CT) at tip-speed ratio lam and blade pitch (rad)."""
omega = lam * V / R
sigma = B * chord / (2 * np.pi * r) # local solidity
a, ap = np.full_like(r, 0.3), np.zeros_like(r) # axial and tangential induction
for _ in range(500):
phi = np.arctan2((1 - a) * V, (1 + ap) * omega * r) # inflow angle
cl, cd = polar(phi - twist - pitch) # angle of attack = phi - twist - pitch
cn = cl * np.cos(phi) + cd * np.sin(phi) # force coefficients normal and
ct = cl * np.sin(phi) - cd * np.cos(phi) # tangential to the rotor plane
F = 2 / np.pi * np.arccos(np.exp(-B * (R - r) / (2 * r * np.sin(phi)))) # Prandtl tip loss
K = 4 * F * np.sin(phi) ** 2 / (sigma * cn)
a_new = 1 / (K + 1)
ac = 0.4 # above a = 0.4 momentum theory fails;
heavy = 0.5 * (2 + K * (1 - 2 * ac) - np.sqrt(np.maximum((K * (1 - 2 * ac) + 2) ** 2 + 4 * (K * ac**2 - 1), 0.0)))
a_new = np.where(a_new > ac, heavy, a_new) # use a straight-line (Glauert-type) fit
ap_new = 1 / (4 * F * np.sin(phi) * np.cos(phi) / (sigma * ct) - 1)
done = np.max(np.abs(a_new - a)) + np.max(np.abs(ap_new - ap)) < 1e-9
a, ap = 0.5 * (a + a_new), 0.5 * (ap + ap_new) # relax for stability
if done:
break
else:
raise RuntimeError("BEM iteration did not converge")
W2 = ((1 - a) * V) ** 2 + ((1 + ap) * omega * r) ** 2 # relative wind speed squared
dT = B * 0.5 * rho * W2 * chord * cn * dr # thrust of each element
dQ = B * 0.5 * rho * W2 * chord * ct * r * dr # torque of each element
q = 0.5 * rho * np.pi * R**2 * V**2
return omega * dQ.sum() / (q * V), dT.sum() / q
print("r/R chord (m) twist (deg)")
for k in (0, 10, 20, 30, 39):
print(f"{r[k] / R:4.2f} {chord[k]:9.2f} {np.degrees(twist[k]):10.1f}")
print("\ntip-speed ratio Cp CT")
for lam in range(3, 13):
cp, ct = bem(lam)
print(f"{lam:10d} {cp:6.3f} {ct:6.3f}")The blade comes out as expected: twisted by 27° from root to tip, its chord falling from the 4.5 m cap to 1.75 m. The rotor's power coefficient peaks at 0.490 at a tip-speed ratio of 7, about 83% of the Betz limit, and falls away on either side: at low the inner blade stalls, and at high drag and over-induction take over. The best is below the design value of 8 because the simple design formula ignores tip losses and drag, which the solver includes. Real large rotors achieve power coefficients of about 0.45 to 0.50.
BEM is fast, which is why it sits inside the design tools of every manufacturer and inside simulation codes such as OpenFAST and QBlade, with corrections for yawed flow, dynamic stall and the time the wake takes to respond. When BEM's assumptions fail, for very large flexible blades, wakes interacting across a wind farm or a rotor in yaw, engineers use vortex methods or the computational fluid dynamics of a previous lesson.
Power curves and capacity factors
A modern turbine runs at variable speed. Below rated wind it adjusts its rotor speed to hold the best tip-speed ratio and so the best (region 2); if the tips reach their speed limit, set by noise and by erosion of the leading edges by rain, it holds that speed and accepts a lower (region 2.5); above rated wind it pitches its blades to hold rated power (region 3); and at the cut-out speed, typically about 25 m/s, it feathers its blades and stops. The program builds the power curve of a 4.0 MW turbine with the rotor computed above, then averages it over a Weibull distribution of wind speeds, whose shape parameter describes how steady the wind is (higher , less spread) and whose scale sets the mean speed, . Lumenary EE uses the Rayleigh distribution, the special case .
import math
import numpy as np
# Power curve of a variable-speed, pitch-controlled turbine with the rotor computed above:
# 120 m diameter, rated 4.0 MW (electrical), tip speed limited to 80 m/s.
lam_table = np.arange(3, 13)
cp_table = np.array([0.135, 0.261, 0.398, 0.467, 0.490, 0.484, 0.451, 0.399, 0.335, 0.257])
R, P_rated, eta_elec, v_tip_max = 60.0, 4.0e6, 0.94, 80.0
v_in, v_out = 3.0, 25.0
A = np.pi * R**2
lam_best = lam_table[np.argmax(cp_table)]
def power(v, rho=1.225):
"""Electrical power (W) at hub-height wind speed v (m/s)."""
if v < v_in or v >= v_out:
return 0.0
omega = min(lam_best * v / R, v_tip_max / R) # track the best tip-speed ratio, up to the speed limit
cp = np.interp(omega * R / v, lam_table, cp_table)
return min(eta_elec * 0.5 * rho * A * cp * v**3, P_rated) # above rated, pitch the blades
print("wind (m/s) power (kW)")
for v in (3, 5, 7, 9, 10, 11, 12, 15, 25):
print(f"{v:8} {power(v) / 1e3:8.0f}")
# Annual energy from a Weibull distribution of hub-height wind speeds.
v = np.linspace(0, 30, 3001)
dv = v[1] - v[0]
P = np.array([power(x) for x in v])
def capacity_factor(v_mean, k):
c = v_mean / math.gamma(1 + 1 / k) # Weibull scale from the mean speed
pdf = (k / c) * (v / c) ** (k - 1) * np.exp(-(v / c) ** k)
return np.sum(P * pdf) * dv / P_rated
print("\nmean wind capacity factor (gross) for Weibull shape k")
print(" (m/s) k = 1.6 k = 2.0 k = 2.6")
for vm in (6.0, 7.5, 9.0):
print(f"{vm:6.1f} " + "".join(f"{capacity_factor(vm, k):10.1%} " for k in (1.6, 2.0, 2.6)))
cf = capacity_factor(7.5, 2.0)
net = cf * 0.97 * 0.90 * 0.98 # availability, wake losses, electrical losses
print(f"\nat 7.5 m/s, k = 2: gross {cf:.1%}, net {net:.1%}, {net * P_rated * 8760 / 1e9:.1f} GWh a year")
# Two site effects: wind shear (the power law) and thin air at altitude.
for alpha in (0.14, 0.20):
print(f"mean 5.5 m/s at 10 m becomes {5.5 * (100 / 10) ** alpha:.2f} m/s at 100 m if alpha = {alpha}")
print(f"at 2000 m altitude (air 1.01 kg/m^3), output at 9 m/s falls from "
f"{power(9.0) / 1e3:.0f} to {power(9.0, rho=1.01) / 1e3:.0f} kW")The table makes several engineering points. The turbine reaches rated power at about 10.8 m/s, before its tips reach 80 m/s, so for this machine region 2.5 never occurs. Its gross capacity factor rises from about 27% at a mean of 6 m/s to 53% at 9 m/s (for ). The shape of the distribution matters too: at a low-wind site a gustier wind () is slightly better, because more of the energy arrives in strong winds that the turbine can use, while at a windy site a steadier wind () is better, because less time is wasted above rated power. After realistic losses for availability, the wakes of neighbouring turbines and electrical losses, the 7.5 m/s site's 41.3% becomes 35.4%, or 12.4 GWh a year.
Real power curves are measured under IEC 61400-12-1, with a mast or lidar upwind of the turbine. Manufacturers have been adding rotor area faster than generator rating, lowering the specific power (rated watts per square metre of rotor) so that turbines reach rated power in lighter winds and run at higher capacity factors; our turbine has 354 W/m². In 2026 onshore turbines of 5 to 7 MW with rotors of 150 to 175 m are common, and offshore turbines of 15 MW with rotors of about 240 m are being installed, with prototypes of 20 MW and more tested in China. Blades are built from glass- and carbon-fibre composites (Lumenary Materials' composites), and recycling them at the end of their 25 to 30-year life is an active engineering problem.
Hydro turbines: Pelton, Francis and Kaplan
Hydropower converts the potential energy of water falling through a head : for a flow (m³/s) and an overall efficiency , which for large modern turbine-generator sets exceeds 90%. All hydro turbines rest on Euler's turbine equation: the torque on a runner equals the rate at which it reduces the angular momentum of the water, , where is the swirl velocity at the inlet (1) and outlet (2). They differ in how they apply it.
- A Pelton wheel is an impulse turbine for high heads, from a few hundred to almost 1,900 m. Nozzles turn the whole head into high-speed jets, which strike split buckets on the rim of a wheel turning in air at atmospheric pressure. The buckets reverse the jet, and the wheel works best when they move at a little under half the jet speed. Its relatives, the Turgo and the crossflow (Banki–Michell) turbines, serve small hydro schemes.
- A Francis turbine is a reaction turbine for medium heads, from about 20 to 700 m. Water fills a spiral casing, passes adjustable guide vanes that give it swirl, and flows inwards through a closed runner, turning from radial to axial and leaving through a draft tube that recovers kinetic energy. The world's largest hydro turbines, of 700 MW at Itaipu (Brazil and Paraguay) and Three Gorges (China) and 1,000 MW at Baihetan, are Francis turbines.
- A Kaplan turbine is an axial-flow reaction turbine, a propeller in a pipe, for low heads and large flows. Both its guide vanes and its runner blades are adjustable, which keeps its efficiency high as the flow varies. Bulb turbines, with the generator in a capsule in the flow, serve the lowest heads and tidal barrages.
The specific speed tells which kind suits a site. It combines the rotational speed (rpm), flow and head in the group , which is the same for geometrically similar turbines whatever their size: low values mean high head and little flow (Pelton), high values low head and much flow (Kaplan). The rotational speed itself is not free, because a generator connected directly to a 50 Hz grid must turn at rpm, where is its number of pole pairs.
import math
rho, g = 1000.0, 9.81
def specific_speed(n_rpm, Q, H):
"""n_q = n sqrt(Q) / H^(3/4), with n in rpm, Q in m^3/s and H in m."""
return n_rpm * math.sqrt(Q) / H**0.75
def turbine_type(nq):
"""A rough guide; near the boundaries, cost, cavitation and part-load behaviour decide."""
if nq < 25:
return "Pelton (impulse)"
if nq < 110:
return "Francis"
return "Kaplan or bulb"
# Illustrative sites: head H (m), flow Q (m^3/s), a synchronous speed for a 50 Hz grid (rpm),
# and the overall efficiency of turbine and generator.
sites = [
("Alpine high-head plant", 1000.0, 8.0, 500.0, 0.90),
("mountain storage plant", 300.0, 50.0, 375.0, 0.92),
("large river dam", 110.0, 650.0, 90.9, 0.93),
("lowland river barrage", 12.0, 400.0, 75.0, 0.90),
("village micro-hydro", 60.0, 0.12, 1000.0, 0.70),
]
print(f"{'site':24}{'power':>11}{'n_q':>7} type")
for name, H, Q, n, eta in sites:
P = rho * g * Q * H * eta
power = f"{P / 1e6:.0f} MW" if P >= 1e6 else f"{P / 1e3:.0f} kW"
nq = specific_speed(n, Q, H)
print(f"{name:24}{power:>11}{nq:7.0f} {turbine_type(nq)}")
# A Pelton wheel: how efficiency depends on bucket speed u relative to jet speed V.
H = 1000.0
V = 0.98 * math.sqrt(2 * g * H) # jet speed after a nozzle losing 4% of the head
print(f"\nPelton jet at {H:.0f} m of head: {V:.1f} m/s")
k, beta = 0.90, math.radians(165) # friction in the bucket, deflection angle
for x in (0.3, 0.4, 0.47, 0.5, 0.6, 0.7):
eta = 2 * x * (1 - x) * (1 - k * math.cos(beta))
print(f" u/V = {x:4.2f}: bucket efficiency {eta:.3f}")
D = 60 * 0.47 * V / (math.pi * 500)
print(f"at 500 rpm, u/V = 0.47 needs a runner {D:.2f} m across (pitch diameter)")
# Tidal stream, wind and waves: power per square metre of flow, and per metre of wave crest.
print("\ntidal stream at 2.5 m/s: %.1f kW/m^2" % (0.5 * 1025 * 2.5**3 / 1e3))
print("wind at 12 m/s: %.1f kW/m^2" % (0.5 * 1.225 * 12**3 / 1e3))
for Hs, Te in ((2.0, 8.0), (3.0, 10.0)):
print(f"waves Hs = {Hs} m, Te = {Te} s: {1025 * g**2 * Hs**2 * Te / (64 * math.pi) / 1e3:.0f} kW per metre of crest")The program sorts the five sites as an engineer would: a Pelton wheel for 1,000 m of head (), Francis turbines for the 300 m storage plant and the 110 m river dam ( and 68), a Kaplan or bulb unit for the 12 m barrage (), and for a village scheme on a Himalayan or Andean stream, 49 kW from 120 litres a second falling 60 m, a small Pelton or crossflow turbine. The Pelton jet at 1,000 m of head travels at 137.3 m/s. In this simple model the buckets do best at exactly half the jet speed, with 93.5% of the jet's energy; real wheels run at about 0.46 to 0.48 of jet speed because of windage and the changing angle as each bucket sweeps through the jet.
Tidal and wave energy in outline
The oceans offer two more resources. Tides are as predictable as the Moon's orbit. A tidal range scheme traps water behind a barrage at high tide and releases it through low-head bulb turbines: France's La Rance barrage (240 MW) has run since 1966, and South Korea's Sihwa Lake plant (254 MW) since 2011. They are reliable but rare, because suitable estuaries are few and barrages change the ecology of the estuaries behind them. Tidal stream turbines work like underwater wind turbines in fast currents. Seawater is about 840 times denser than air, so a 2.5 m/s current carries 8.0 kW per square metre of flow, more than seven times as much as a 12 m/s wind (1.1 kW/m²); rotors can be small, but they must survive salt water, enormous thrust and the difficulty of maintenance underwater. Arrays of a few megawatts operate in Scotland's Pentland Firth (MeyGen), in Orkney and Shetland, and in China's Zhoushan archipelago, with larger projects planned in France and elsewhere.
Waves carry wind energy gathered over thousands of kilometres of ocean. In deep water the power per metre of wave crest is about , where is the significant wave height and the energy period: 16 kW/m for 2 m waves of 8 s, and 44 kW/m for 3 m waves of 10 s, more on exposed Atlantic and Southern Ocean coasts. Devices include oscillating water columns, where waves pump air through a turbine (Spain's breakwater plant at Mutriku has run since 2011), floating point absorbers and hinged attenuators. As of 2026 wave energy remains at the demonstration stage: the engineering problem is a machine that captures energy efficiently from ordinary waves yet survives the largest storms, which carry a hundred times more power, at an affordable cost. Test centres such as EMEC in Orkney exist to let developers learn this at sea.
Common misconceptions
Misconception
"A turbine with more blades captures more energy." More blades add drag, weight and cost, and slow the optimum rotor speed, which increases wake-rotation losses. Three blades are a compromise that gives nearly the maximum power coefficient with smooth dynamics; two-bladed rotors are cheaper but noisier and harder on their hubs, and many-bladed windpumps are designed for high starting torque, not efficiency.
Misconception
"A turbine rated at 4 MW produces 4 MW." It produces 4 MW only above its rated wind speed, about 10.8 m/s for the turbine in this lesson; at 7 m/s it produced 1,094 kW. Over a year its output averages a third to a half of its rating at good sites, which is what the capacity factor measures.
Misconception
"The Betz limit applies to hydro turbines too." Betz bounds a rotor in an open stream, which the flow can go around. A hydro turbine in a closed penstock takes the water's pressure energy directly, and every kilogram of water must pass through it, so efficiencies above 90% are possible. Tidal stream turbines in open currents, by contrast, are bounded by Betz.
Misconception
"Doubling the wind speed doubles the power." Below rated speed, power grows as the cube of the wind speed, so doubling the speed gives about eight times the power. This is why a site's mean wind speed, and the shear profile up to hub height, matter so much.
Exercises
Use kg/m³ for air, 1000 kg/m³ for fresh water and m/s² unless an exercise says otherwise.
1. Wind and water turbines
Warm-upHumanA wind of 8 m/s blows through a rotor 100 m in diameter (air density 1.225 kg/m³). What is the most power any rotor of that size could extract, according to the Betz limit? Give your answer in kW, to the nearest whole number.
2. Wind and water turbines
CoreHumanAn ideal rotor 150 m in diameter operates at the Betz optimum () in a 9 m/s wind (air density 1.225 kg/m³). Using momentum theory, what thrust does it exert on its tower, in kN, to the nearest whole number?
3. Wind and water turbines
CoreHumanWhy is a wind-turbine blade set at a much steeper angle to the rotor plane near its root than near its tip?
4. Wind and water turbines
CoreComputationalWrite capacity_factor(power, rated, v_mean, k). power is a function giving a turbine's output (in any unit) at a hub-height wind speed v (m/s), rated is its rated output in the same unit, and the wind follows a Weibull distribution with shape k and mean speed v_mean. Return the capacity factor as a fraction, by integrating numerically from 0 to 40 m/s.
Recall with (math.gamma).
5. Wind and water turbines
CoreHumanA hydro site has a net head of 200 m and a design flow of 80 m³/s per unit. The generator will run at 250 rpm. What is the specific speed , to the nearest whole number? (Think also about which type of turbine it points to.)
6. Wind and water turbines
CoreHumanA Pelton wheel works under a net head of 600 m. Its nozzle has a velocity coefficient of 0.98 (jet speed ), the buckets move at 0.47 of the jet speed, and the wheel turns at 600 rpm. What is the wheel's pitch diameter (the diameter of the circle where the jet strikes the buckets), in centimetres, to the nearest whole number?
7. Wind and water turbines
StretchComputationalWrite the heart of a BEM solver for a single blade element, without tip loss (F = 1) and without the heavy-loading correction. Take the wind speed as 1, so the element moves at its local speed ratio lam_r .
element_induction(lam_r, sigma, theta, cl_slope=2*math.pi, cd=0.0) should iterate:
- , ,
cl_slope,cd; - , ;
- , ;
starting from , and updating with relaxation, (likewise ), until the change is below . Return (a, a_prime).
Then check the lesson's design rule: for an element designed with , and , the converged should be close to 1/3.
8. Wind and water turbines
CoreHumanWhich of these statements are true? Select all that apply.
Further reading and resources
- Aerodynamics of Wind Turbines · Martin O. L. HansenA standard, compact text on momentum theory, BEM and the unsteady aerodynamics of rotors.
- Das Maximum der theoretisch möglichen Ausnützung des Windes durch Windmotoren · Albert Betz · 1920The paper that derived the limit on the power a rotor in an open stream can extract.
- Global Wind AtlasFree maps of mean wind speed, Weibull parameters and power density at several heights, for anywhere on Earth; from the Technical University of Denmark and the World Bank.
- OpenFASTOpen-source software for the coupled aerodynamics, structural dynamics and control of wind turbines, developed at the US National Laboratory of the Rockies (formerly NREL).
- IEA 15 MW offshore reference turbineA complete public design of a 15 MW, 240 m rotor turbine from IEA Wind, with blade geometry and simulation models, used by researchers worldwide.
- QBladeWind turbine design and simulation software from TU Berlin, with a free community edition for non-commercial use.
- International Electrotechnical Commission (IEC)Publishes the IEC 61400 series on wind turbine design, load cases and power-curve testing, and IEC standards for hydraulic turbine acceptance tests.
- Global Wind Energy CouncilAnnual global wind reports on installations and markets by country.
- International Hydropower AssociationData and reports on hydropower and pumped storage worldwide.
- European Marine Energy Centre (EMEC)The test centre in Orkney, Scotland, where many tidal and wave energy devices have been tested at sea.