Wind energy conversion systems

In short

A wind energy conversion system (WECS) converts the kinetic energy of moving air into mechanical rotation at a turbine shaft and then into electrical energy through a generator and, in modern machines, a power-electronic converter. The power electronics decide almost everything that matters: whether the turbine can run at variable speed to capture maximum power, how reactive power and voltage are controlled, and whether the machine can ride through a grid fault. India has over 48 GW of installed wind capacity (largely Tamil Nadu, Gujarat, Karnataka, Maharashtra and Rajasthan) and the segment is governed by MNRE policy, NIWE type certification and the CEA grid-connectivity regulations. The topic appears in GATE (EE) under power-electronic applications and renewable generation, and in interviews at Suzlon, Siemens Gamesa, Vestas India, Inox Wind, ReGen, Senvion, Hitachi Energy, Adani Green, Tata Power Renewables and the NTPC/PGCIL grid-side roles.

Explained

Wind resource and the power in the wind

• Air of density ρ moving at velocity V through a swept area A carries kinetic power P = ½ρAV³. The cube law dominates everything: a 10% rise in wind speed gives about 33% more power, so site selection and hub height matter more than machine refinement.
• ρ ≈ 1.225 kg/m³ at 15 °C and sea level; it falls with altitude and temperature, so the same turbine yields less in Rajasthan summer than in coastal Tamil Nadu winter.
• Wind shear: V(h) = V_ref (h/h_ref)^α, with the power-law exponent α ≈ 0.14 over open terrain and 0.2–0.3 over rough/forested ground. This is why hub heights in India have grown from 50 m to 120–140 m.
• Statistics: wind speed over a year is modelled by the Weibull distribution (shape k, scale c); Rayleigh is the special case k = 2 used in textbook problems. Annual energy = Σ over wind-speed bins of (probability × power × 8,760 h).
• Characteristic speeds: *cut-in* (3–4 m/s, generator starts), *rated* (11–14 m/s, rated electrical output reached), *cut-out* (20–25 m/s, machine shut down for safety), *survival* (~60–70 m/s, structural design).
• Capacity factor = actual annual energy / (rated power × 8,760). Indian onshore sites typically give 20–35%; modern high-hub-height machines reach above 35%. Offshore is higher.
• Classification: HAWT (horizontal axis, almost all grid machines, upwind three-blade) and VAWT (Darrieus — lift-type, not self-starting; Savonius — drag-type, self-starting; both low efficiency, niche use); also constant-speed vs variable-speed, and stall- vs pitch-regulated.

Betz limit, Cp and tip speed ratio

• The turbine extracts only part of the wind power because the air must keep moving to leave the rotor. The extracted fraction is the power coefficient Cp = P_mech / (½ρAV³).
• Betz limit: using one-dimensional momentum theory with axial induction factor a (V_rotor = V(1−a), V_wake = V(1−2a)), Cp = 4a(1−a)². Maximising gives a = 1/3 and Cp,max = 16/27 ≈ 0.593. No turbine of any design can exceed this; real large HAWTs reach 0.45–0.50 after blade, tip, drivetrain and generator losses.
• Tip speed ratio λ = ωR/V (blade tip speed ÷ wind speed), dimensionless. Cp is a single-peaked function of λ (and blade pitch β): Cp(λ, β) peaks at λ_opt, typically 7–9 for a three-blade HAWT, *higher* (≈10) for two-blade because fewer blades need faster tips, and below 1 (≈0.8–1) for a Savonius drag machine.
• Key consequence: for every wind speed there is one rotor speed that gives Cp,max. A fixed-speed turbine therefore sits at Cp,max for only one wind speed; a variable-speed turbine tracks ω ∝ V and stays near Cp,max over the whole below-rated range. This is the entire reason power electronics is in a wind turbine.
• Thrust and torque: thrust coefficient Ct = 4a(1−a) (= 8/9 at Betz point); shaft torque T = P_mech/ω. Torque at the optimum follows T = K_opt ω², which is the basis of the simplest MPPT law.

Operating regions and control

1. Region 1 (below cut-in): no generation; rotor idles or is parked.
2. Region 2 (cut-in to rated, partial load): maximise energy — MPPT. Pitch held at fine/optimum (β ≈ 0°), generator torque used to hold λ = λ_opt.
3. Region 2½/3 (at rated power): wind exceeds rated; power must be limited. Speed is held at rated by pitching the blades and torque is held at rated.
4. Region 4 (above cut-out): shutdown — blades feathered to ~90°, rotor braked and parked; yawed out of wind in some designs.
5. Power limiting methods: *passive stall* (blade aerofoil designed to stall as wind rises — simple, no actuator, but power overshoot and higher loads), *active stall* (pitch towards stall), *pitch-to-feather* (pitch towards lower angle of attack — standard in modern MW machines; hydraulic or electric pitch drives with battery/ultracapacitor back-up for emergency feathering).
6. Yaw control: motorised yaw drive keeps the nacelle facing the wind using a nacelle anemometer/wind vane; small turbines use a tail vane.
7. MPPT methods for wind:
   • Tip-speed-ratio (TSR) control: measure V, command ω* = λ_opt V/R. Accurate but needs a reliable anemometer.
   • Optimum torque (OT) control: command T* = K_opt ω², K_opt = ½ρπR⁵Cp,max/λ_opt³. No wind measurement; most used industrially; slow to respond to gusts.
   • Power signal feedback (PSF): look-up table of P_max vs ω stored from the turbine's known curve.
   • Perturb and observe / hill climbing (P&O): step the duty ratio or speed and keep the direction that increases power. Sensorless and generic, but oscillates around the peak and can be confused by fast wind changes in large-inertia machines.
8. Mechanical/drivetrain: gearbox (ratio 1:70 to 1:100) between a 10–20 rpm rotor and a 1,500 rpm generator; direct-drive machines delete the gearbox and use a large low-speed multi-pole synchronous generator — PMSG (Siemens Gamesa direct drive) or electrically excited (Enercon-style); medium-speed hybrid drives use a single-stage gearbox.

Generator and converter topologies (Types 1–4)

• Type 1 — fixed-speed SCIG: squirrel-cage induction generator directly across the grid through a soft-starter; needs a capacitor bank for magnetising reactive power; speed varies only by slip (1–2%). Simple and rugged, but all gusts pass into the drivetrain as torque spikes, no MPPT, no reactive-power control. Older Suzlon/Vestas Indian fleet.
• Type 2 — variable-slip WRIG: wound-rotor induction generator with external rotor resistance varied by a chopper (OptiSlip); gives ~10% speed range and smoother power, but the slip energy is burnt as heat in the resistors.
• Type 3 — DFIG (doubly fed induction generator): wound rotor fed through a back-to-back AC–DC–AC converter of rotor-side (RSC) and grid-side (GSC) VSCs with a DC link; stator connected directly to the grid. Converter handles only the slip power, s·P, so it is rated at about 25–30% of machine rating for ±30% speed range — the cost advantage that made Type 3 the dominant Indian topology.
  • Slip power recovery: sub-synchronous operation (s > 0) → power flows into the rotor; super-synchronous (s < 0) → rotor power is fed back to the grid; total output P_stator + P_rotor.
  • Vector control: RSC in stator-flux (or stator-voltage) oriented d-q frame — q-axis rotor current controls electromagnetic torque (and hence speed/MPPT), d-axis rotor current controls reactive power/stator voltage. GSC regulates the DC-link voltage and its own reactive power, usually at unity power factor.
  • Weaknesses: slip rings and brushes need maintenance; stator is grid-connected so faults hit the machine directly — needs a crowbar (thyristor-switched rotor short) or DC-chopper to protect the converter during voltage dips; prone to sub-synchronous resonance/torsional interaction with series-compensated lines.
• Type 4 — full converter (PMSG or SCIG/WRSG): generator fully decoupled from the grid by a full-rated back-to-back converter. Machine-side converter (active rectifier, or a diode bridge plus boost in small machines) performs MPPT; grid-side inverter controls DC-link voltage and injects current at the required power factor, with an LCL filter.
  • PMSG: no excitation losses or field winding, high efficiency and torque density, allows direct drive; but costly rare-earth (NdFeB) magnets, risk of demagnetisation at high temperature, and full-rated converter.
  • Full converter gives the widest speed range (0 to ~1.2 pu), best fault ride-through and full four-quadrant reactive capability — the standard for offshore and new large onshore machines.
• Converter details: two-level IGBT VSC with SPWM/SVPWM at 2–5 kHz for MW units; three-level NPC or modular/parallel converters at higher ratings and for lower losses; press-pack IGBTs/IGCTs in multi-MW designs.

Grid integration, codes and protection

• Requirements a grid-connected WECS must meet: active-power control and ramp-rate limits, reactive power/power-factor range, voltage and frequency operating windows, LVRT/FRT (low-voltage/fault ride-through), flicker and harmonic limits, SCADA telemetry and remote dispatchability.
• Indian framework (verify the current amendment — these regulations are revised often):
  • CEA (Technical Standards for Connectivity to the Grid) Regulations, 2007 with later amendments — the 2013 amendment introduced fault-ride-through and reactive-power requirements for wind and solar generators.
  • Indian Electricity Grid Code (IEGC) issued by CERC, with forecasting, scheduling and deviation-settlement (DSM) obligations for wind; state codes for intra-state projects.
  • Electricity Act 2003 (open access, RPO), and MNRE/NIWE for wind resource assessment, the Revised List of Models and Manufacturers (RLMM) and type certification; IEC 61400 series (61400-1 loads/classes, -12 power performance, -21 power-quality measurement, -3 offshore) is the international reference, adopted in Indian practice.
• Harmonics and power quality: converter-side switching harmonics limited by LCL filters; IEEE 519 / IEC 61000 limits commonly specified; fixed-speed machines cause flicker and inrush instead.
• Reactive power: Type 1 absorbs reactive power (capacitor banks, SVC/STATCOM at the feeder); Types 3 and 4 can supply or absorb within converter current rating, acting as plant-level voltage support.
• Fault ride-through hardware: crowbar and DC chopper (DFIG), series dynamic braking resistor, converter current limiting, and a pitch-based power reduction so the turbine does not overspeed when its electrical load vanishes.
• Protection and safety: overspeed (two independent systems — pitch/feather and mechanical disc brake), vibration, over-temperature, lightning protection down the blade to the earthing system, anti-islanding, and an interface circuit breaker. Collector system is typically 690 V generator → 33 kV padmount/nacelle transformer → 33 kV collector → 132/220 kV pooling substation.
• Variability management: aggregation over a wind zone, forecasting (mandatory in India for scheduling), curtailment, and increasingly BESS hybridisation; synthetic inertia and fast frequency response from the converter are now specified in several codes.

Standalone, hybrid and emerging systems

• Standalone/isolated WECS: SEIG (self-excited induction generator with capacitor bank) or small PMSG with diode rectifier → DC bus → battery and inverter; needs dump load/electronic load controller to keep frequency and voltage when load drops.
• Wind–solar hybrid: shared inverter/DC bus and shared evacuation; MNRE's National Wind-Solar Hybrid Policy (2018) encourages this, raising the use of the transmission capacity.
• Wind pumping and wind–diesel systems in remote and island grids (Lakshadweep, Andaman).
• Offshore wind: higher capacity factor and bigger rotors; India's National Offshore Wind Energy Policy (2015) with identified zones off Gujarat and Tamil Nadu — tendering is ongoing; HVDC or 66 kV AC export, monopile/jacket foundations.
• Repowering: replacing clusters of old sub-500 kW fixed-speed machines in Tamil Nadu and Gujarat with few MW-class variable-speed units on the same land; MNRE's repowering and life-extension policy supports it.

Key points

• Wind power - P = ½ρAV³; cube law means wind speed, not machine size, dominates yield.
• Betz limit - maximum extractable fraction Cp = 16/27 ≈ 0.593 at axial induction a = 1/3; practical Cp 0.45–0.50.
• Power coefficient Cp - mechanical power ÷ available wind power; function of tip speed ratio λ and pitch β.
• Tip speed ratio λ - ωR/V; λ_opt ≈ 7–9 for three-blade HAWT; holding λ_opt is the aim of MPPT.
• Variable speed - keeps Cp at maximum across the below-rated range, giving typically 5–15% more annual energy plus lower drivetrain loads.
• Operating regions - cut-in 3–4 m/s, rated 11–14 m/s, cut-out 20–25 m/s; MPPT below rated, pitch regulation above rated.
• Pitch vs stall - pitch-to-feather actively limits power and loads; passive stall is simpler but overshoots power.
• Type 1 SCIG - fixed speed, soft starter, capacitor bank, no MPPT or reactive control.
• Type 2 WRIG - rotor resistance chopper, ~10% slip range, slip energy wasted as heat.
• Type 3 DFIG - partial-rated (≈30%) back-to-back converter in the rotor circuit; vector control of torque and reactive power; needs crowbar for FRT.
• Type 4 full converter - PMSG or SCIG with 100%-rated converter; full decoupling, widest speed range, best ride-through, often direct drive.
• Back-to-back converter - machine-side converter (MPPT/torque) + DC link + grid-side converter (DC-link voltage and reactive power), with LCL filter.
• Optimum torque MPPT - T* = K_opt ω² with K_opt = ½ρπR⁵Cp,max/λ_opt³; needs no anemometer.
• Capacity factor - annual energy ÷ (rated power × 8,760 h); 20–35% typical onshore in India.
• Weibull/Rayleigh distribution - describes the annual wind-speed spread used for energy yield estimates.
• LVRT/FRT - mandated by CEA connectivity regulations; turbine must stay connected and support voltage through a dip.
• Grid code and standards - CEA Connectivity Regulations, IEGC, Electricity Act 2003, NIWE RLMM, IEC 61400 series.
• Gearbox vs direct drive - 1:70–1:100 gearbox with high-speed generator, or multi-pole low-speed generator with none.

Formulas and examples

Example 1 — Available power, Betz limit and actual output. A turbine has rotor diameter 90 m, Cp = 0.45, drivetrain + generator + converter efficiency 0.93, wind 12 m/s, ρ = 1.225 kg/m³.

• A = π × 45² = 6,361.7 m²
• V³ = 12³ = 1,728
• P_wind = 0.5 × 1.225 × 6,361.7 × 1,728 = 6.73 MW
• Betz ceiling = 0.593 × 6.73 = 3.99 MW
• P_mech = 0.45 × 6.73 = 3.03 MW
• P_elec = 0.93 × 3.03 = 2.82 MW → a 2.8 MW-rated machine is about at rated power at 12 m/s.

Example 2 — Optimum rotor speed, gearbox and optimum-torque constant. Same rotor (R = 45 m), λ_opt = 8, at V = 9 m/s. Cp,max = 0.45, ρ = 1.225.

• ω = λ_opt V/R = 8 × 9/45 = 1.6 rad/s → N = 60 × 1.6/(2π) = 15.3 rpm
• To drive a 4-pole, 50 Hz generator (ns = 1,500 rpm) the gear ratio G = 1,500/15.3 ≈ 98
• P_mech = 0.5 × 1.225 × 6,361.7 × 729 × 0.45 = 1.278 MW; T = P/ω = 1.278×10⁶/1.6 = 799 kN·m (low-speed shaft)
• K_opt = 0.5 × 1.225 × π × 45⁵ × 0.45/8³ = 0.5 × 1.225 × 3.1416 × 1.8453×10⁸ × 0.45/512 ≈ 3.12×10⁵ N·m·s²/rad²
• Check: T = K_opt ω² = 3.12×10⁵ × 1.6² = 799 kN·m ✔ (matches the direct calculation)

Example 3 — DFIG slip, rotor power and converter rating. A 2 MW, 4-pole, 50 Hz DFIG runs at 1,750 rpm with the stator delivering 2 MW (losses ignored).

• ns = 120 × 50/4 = 1,500 rpm → s = (1,500 − 1,750)/1,500 = −0.1667 (super-synchronous)
• P_rotor ≈ −s × P_stator = 0.1667 × 2 = 0.333 MW fed back to the grid through the converter
• Total output ≈ 2 + 0.333 = 2.33 MW; for a speed range of 1,000–1,800 rpm, |s|max = 0.333, so converter rating ≈ 0.333 × 2 = 0.67 MW ≈ 33% of machine rating — versus 100% for a Type 4 machine.

Example 4 — Energy yield and capacity factor with the cube law. A 2 MW turbine at a Tamil Nadu site: 2,400 h at 12 m/s (rated, 2 MW), 3,000 h at 8 m/s, 1,500 h at 5 m/s, remaining 1,860 h below cut-in or above cut-out (0 output). Assume Cp·η = 0.42 combined and A = 6,361.7 m².

• At 8 m/s: P = 0.5 × 1.225 × 6,361.7 × 512 × 0.42 = 0.838 MW
• At 5 m/s: P = 0.5 × 1.225 × 6,361.7 × 125 × 0.42 = 0.205 MW
• Energy = 2,400 × 2 + 3,000 × 0.838 + 1,500 × 0.205 = 4,800 + 2,514 + 307.5 = 7,621 MWh
• CF = 7,621/(2 × 8,760) = 7,621/17,520 = 43.5% (high — this illustrative duration curve is more favourable than a typical Indian onshore site at 25–30%)
• At ₹3.00/kWh, revenue ≈ 7,621,500 × 3 = ₹2.29 crore per year

Example 5 — Wind shear and the gain from a taller tower. V = 7.5 m/s measured at 80 m; α = 0.20. Find the speed and power gain at 120 m hub height.

• V(120) = 7.5 × (120/80)^0.20 = 7.5 × 1.5^0.2 = 7.5 × 1.0845 = 8.13 m/s
• Power ratio = (8.13/7.5)³ = 1.0845³ = 1.276 → about 28% more power for the same rotor, which is why Indian OEMs moved to 120–140 m towers.

Where it is used

• Turbine OEMs and their power-electronics teams: Suzlon, Inox Wind, Senvion India, ReGen Powertech, Siemens Gamesa, Vestas India, Envision — roles in converter design, control software, pitch systems, drivetrain and type testing to IEC 61400.
• Converter and component suppliers: Hitachi Energy, ABB, Danfoss, Amara Raja/Exicom (back-up systems), gearbox and generator makers — IGBT stack design, LCL filter design, DC-link and crowbar sizing.
• IPPs and O&M: Adani Green, Tata Power Renewables, ReNew, Sembcorp/Vector Green, NTPC Green — SCADA, performance analysis, availability and generation-loss accounting, repowering studies.
• Grid and regulatory side: POSOCO/Grid-India, CEA, state load despatch centres, NIWE — forecasting and scheduling, DSM accounting, grid-code compliance and FRT testing, wind resource assessment (NIWE 100 m/120 m wind atlas).
• Consulting and EPC: feasibility and energy-yield studies (Weibull-based), collector-system and 33 kV/220 kV substation design, wind–solar hybrid and BESS integration.
• Small and standalone systems: micro-wind and wind–solar street lighting, telecom towers, island grids (Lakshadweep, Andaman) and wind pumping.

Interview questions

• Q: Why is the Betz limit 16/27 and can a better blade beat it? A: Momentum theory gives Cp = 4a(1−a)², maximised at a = 1/3 to give 16/27 ≈ 59.3%; no rotor design can exceed it because the air must retain enough velocity to leave the rotor disc.
• Q: Why do modern wind turbines run at variable speed? A: Cp peaks at one tip speed ratio, so holding λ = λ_opt needs ω ∝ V; variable speed captures 5–15% more annual energy and absorbs gusts as rotor inertia rather than drivetrain torque.
• Q: Why is the DFIG converter rated at only about 30% of the machine? A: Only the slip power (s·P) flows through the rotor circuit, and the speed range is limited to roughly ±30% around synchronous speed, so the converter handles about one-third of rated power.
• Q: What does the rotor-side converter control in a DFIG, and the grid-side converter? A: The rotor-side converter controls electromagnetic torque (q-axis rotor current, hence MPPT) and stator reactive power/voltage (d-axis); the grid-side converter regulates the DC-link voltage and its own reactive power.
• Q: What is a crowbar and why does a DFIG need one? A: A thyristor-switched resistor network that shorts the rotor during a grid dip, diverting the large induced rotor current away from the converter so it survives the fault; a DC-link chopper is often used with it.
• Q: Compare DFIG and PMSG with full converter. A: DFIG is cheaper (partial converter) but has slip rings, a grid-connected stator and weaker ride-through; PMSG with a full converter is fully decoupled, gives the widest speed range and best FRT and reactive capability, at higher converter and magnet cost.
• Q: What is pitch control and how does it differ from stall control? A: Pitch control rotates the blades to change the angle of attack and actively limits power above rated speed; stall control relies on the aerofoil stalling passively — simpler and cheaper but with power overshoot and higher loads.
• Q: How does optimum-torque MPPT work without measuring wind speed? A: At λ_opt the shaft torque follows T = K_opt ω², so commanding that torque from the measured generator speed drives the rotor to the optimum tip speed ratio as the wind changes.
• Q: Why do turbines cut out at 20–25 m/s rather than generating more? A: Thrust and fatigue loads rise steeply with wind speed; shutting down and feathering protects the blades, tower and drivetrain, and the hours spent above cut-out contribute little annual energy.
• Q: What is LVRT and is it mandatory in India? A: Fault ride-through requires a generator to stay connected and support voltage during a dip rather than trip; it is required for grid-connected wind plants under the CEA Technical Standards for Connectivity Regulations (introduced by the 2013 amendment, with later revisions).
• Q: Why does a fixed-speed SCIG wind farm need capacitor banks or a STATCOM? A: An induction generator draws magnetising reactive power from the grid; capacitors supply the steady component and a STATCOM/SVC gives the dynamic voltage support the machine itself cannot provide.
• Q: What does a 30% capacity factor mean, and why is it not efficiency? A: It means the plant produced 30% of what it would at rated output all year; it reflects the wind resource and availability, not conversion efficiency, which is measured by Cp and drivetrain losses.
• Q: Why is a 10% drop in wind speed so serious? A: Power varies as V³, so 0.9³ = 0.729 — about 27% less power — which also explains the push for taller towers and better siting.

Common mistakes

• Using P = ½ρAV³ as the turbine output; it is the wind's power — multiply by Cp and the drivetrain/generator/converter efficiency.
• Writing V² instead of V³, or forgetting that A = πR² with R the blade length, not the diameter.
• Treating Betz's 59.3% as an achievable figure, or adding generator efficiency *to* Cp instead of multiplying.
• Confusing capacity factor with efficiency or with availability.
• Assuming the DFIG converter must be rated for full power, or that the DFIG stator current flows through it.
• Mixing rpm and rad/s in λ = ωR/V and in T = P/ω, and forgetting the gearbox ratio between rotor and generator speed.
• Claiming pitch control is used for MPPT below rated — below rated, pitch is held at its optimum and torque does the tracking.
• Averaging wind speed and then cubing it; because of the cube law this underestimates energy, so the Weibull/bin distribution must be used.
• Assuming a two-bladed rotor runs slower than a three-bladed one — fewer blades need a *higher* tip speed ratio.

Related topics

Solar PV systems and MPPT · Inverters · Rectifiers and converters · Electric drives · Multilevel inverters · Power semiconductor devices