Skip to content

Basic and Generic Governors

This page documents the four governors built into every RAMSES distribution, CONSTANT, 1ST_ORDER, HYDRO_GENERIC1 and THERMAL_GENERIC1, which take no prefix and are written in uppercase, together with HYDRO_DG, registered since 3.76.

The TOR sub-record these models are entered with is described on the Governor Models page.

The data-file model name is CONSTANT.

The tor_constant model provides a constant mechanical torque source. The mechanical torque is fixed at its initialisation value and does not respond to speed deviations or load changes:

Tm(t)=Tm0∀tT_m(t) = T_{m0} \quad \forall t

The single algebraic state satisfies:

f(1)=x(1)−Tm0=0f(1) = x(1) - T_{m0} = 0

The observable output is mechanical power:

Pm=Tm⋅ωP_m = T_m \cdot \omega

Implementation details:

  • nbxvar = 1, nbzvar = 0, nbdata = 0
  • One additional parameter: prm(1) = Tm0 (set at initialisation from load-flow)
  • Observable: Pm = x(1) * omega
ParameterDescription
(none)No user-supplied parameters. Tm0T_{m0} is taken from the load-flow initialisation.
Show the data record
SYNC_MACH g6 g6
# FP FQ P Q: share of the power-flow injection (FP, FQ) or initial power (P in MW, Q in Mvar)
1. 1. 0. 0.
# SNOM: Nominal apparent power, used as base power in the machine model (MVA)
400.
# Pnom: Nominal active power of the turbine, used as base power for the turbine model (MW)
360.
# H: Inertia constant (s)
6.
# D: Damping coefficient (usually set to zero when the damper windings are modelled) (pu)
0.
# IBRATIO: Field current base ratio I_fB^mac/I_fB^exc (see above) (pu)
2.05
# TYPE_MOD: Parameter format keyword: RL or XT (case-insensitive)
XT
# Xl: Leakage reactance L_l (pu)
0.15
# Xd: d-axis synchronous reactance (pu)
2.2
# X'd: d-axis transient reactance (must be smaller than Xd) (pu)
0.3
# X"d: d-axis subtransient reactance (pu)
0.2
# Xq: q-axis synchronous reactance (pu)
2.
# X'q: q-axis transient reactance (pu)
0.4
# X"q: q-axis subtransient reactance (pu)
0.2
# m: Saturation coefficient (set to 0 to neglect saturation)
0.1
# n: Saturation exponent (ignored when m = 0)
6.0257
# Ra: Armature resistance (pu)
0.
# T'do: d-axis open-circuit transient time constant (s)
7.00
# T"do: d-axis open-circuit subtransient time constant (s)
0.05
# T'qo: q-axis open-circuit transient time constant (s)
1.5
# T"qo: q-axis open-circuit subtransient time constant (s)
0.05
EXC GENERIC1
# IFLIM: OEL field current threshold (pu)
3.0618
# d: OEL timer band, usually negative: timer runs down at rate 1 below IFLIM+d, holds between IFLIM+d and IFLIM (pu)
-0.1
# f: OEL timer charging rate offset above IFLIM
1.
# S: OEL timer charging rate per unit of field overcurrent (1/pu)
0.
# K1: Not used by the code
100.
# K2: OEL gain on (if - IFLIM), negative so that the OEL lowers excitation (pu/pu)
-1.
# L1: OEL timer lower limit and initial value, negative; the OEL acts when the timer exceeds 0 (s)
-20.0
# L2: OEL timer upper limit (s)
10.
# G: AVR gain (pu/pu)
120.
# TA: AVR lead time constant (s)
5.
# TB: AVR lag time constant (s)
12.5
# TE: Exciter time constant (s)
0.1
# L3: Minimum field voltage (pu)
0.
# L4: Maximum field voltage (pu)
5.
# SPEEDIN: PSS input weight: 1 = speed, 0 = active power
1
# KPSS: PSS gain, derivative washout KPSS s/(1+sTw); 0 disables PSS (pu s)
75.
# Tw: PSS washout time constant (s)
15.
# T1: PSS first lead-lag numerator time constant (s)
0.22
# T2: PSS first lead-lag denominator time constant (s)
0.012
# T3: PSS second lead-lag numerator time constant (s)
0.22
# T4: PSS second lead-lag denominator time constant (s)
0.012
# DVMIN: PSS output lower limit (pu)
-0.1
# DVMAX: PSS output upper limit (pu)
0.1
TOR CONSTANT
;

The data-file model name is 1ST_ORDER (uppercase, with the underscore).

The tor_1storder model is a first-order speed governor with droop and a two-mass (HP/LP) turbine representation. It captures the primary frequency response through a proportional-droop speed controller with a first-order lag.

Speed controller with droop RR:

The mechanical torque setpoint is:

Tset=Tm0−ω−1RT_{set} = T_{m0} - \frac{\omega - 1}{R}

The HP torque tracks TsetT_{set} through a first-order lag with time constant T2T_2:

dx1dt=1T2(Tset−x1)\frac{d x_1}{dt} = \frac{1}{T_2}\left(T_{set} - x_1\right)

The LP (shaft) torque is a weighted combination:

x2=(1−FHP)⋅x1+FHP⋅Tsetx_2 = (1 - F_{HP}) \cdot x_1 + F_{HP} \cdot T_{set}

where FHPF_{HP} is the fraction of power generated in the HP stage.

From the Fortran equations:

f(1) = (-x(1) + (prm(4) - (omega-1.d0)/prm(3))) / prm(2)
f(2) = (1.d0-prm(1))*x(1) + prm(1)*(prm(4)-(omega-1.d0)/prm(3)) - x(2)

i.e.:

  • prm(1) = FHP (HP fraction), prm(2) = T2 (time constant), prm(3) = R (droop)
  • prm(4) = Tm0 (setpoint, set at init)

Implementation details:

  • nbxvar = 2, nbzvar = 0, nbdata = 3
  • State 1 = HP torque x(1), State 2 = LP torque x(2)
  • Observables: HP torque, LP torque
1ST_ORDER governor block diagram. The speed deviation omega minus one is subtracted from the initial torque Tm0 through the droop one over R to give the torque setpoint Tset. Tset passes a first-order lag of time constant T2 to give the high-pressure torque x1, and the shaft torque is one minus FHP times x1 plus FHP times Tset. 1ST_ORDER governor block diagram. The speed deviation omega minus one is subtracted from the initial torque Tm0 through the droop one over R to give the torque setpoint Tset. Tset passes a first-order lag of time constant T2 to give the high-pressure torque x1, and the shaft torque is one minus FHP times x1 plus FHP times Tset.
ParameterDescription
FHPFraction of torque produced by the HP turbine stage (0–1)
T2Governor/HP turbine time constant (s)
RSpeed droop (pu/pu)
Show the data record
SYNC_MACH g1 g1
# FP FQ P Q: share of the power-flow injection (FP, FQ) or initial power (P in MW, Q in Mvar)
1. 1. 0. 0.
# SNOM: Nominal apparent power, used as base power in the machine model (MVA)
800.
# Pnom: Nominal active power of the turbine, used as base power for the turbine model (MW)
760.
# H: Inertia constant (s)
3.
# D: Damping coefficient (usually set to zero when the damper windings are modelled) (pu)
0.
# IBRATIO: Field current base ratio I_fB^mac/I_fB^exc (see above) (pu)
0.95
# TYPE_MOD: Parameter format keyword: RL or XT (case-insensitive)
XT
# Xl: Leakage reactance L_l (pu)
0.15
# Xd: d-axis synchronous reactance (pu)
1.1
# X'd: d-axis transient reactance (must be smaller than Xd) (pu)
0.25
# X"d: d-axis subtransient reactance (pu)
0.2
# Xq: q-axis synchronous reactance (pu)
0.7
# X'q: q-axis transient reactance (pu)
*
# X"q: q-axis subtransient reactance (pu)
0.2
# m: Saturation coefficient (set to 0 to neglect saturation)
0.1
# n: Saturation exponent (ignored when m = 0)
6.0257
# Ra: Armature resistance (pu)
0.
# T'do: d-axis open-circuit transient time constant (s)
5.00
# T"do: d-axis open-circuit subtransient time constant (s)
0.05
# T'qo: q-axis open-circuit transient time constant (s)
*
# T"qo: q-axis open-circuit subtransient time constant (s)
0.1
EXC GENERIC1
# IFLIM: OEL field current threshold (pu)
1.8991
# d: OEL timer band, usually negative: timer runs down at rate 1 below IFLIM+d, holds between IFLIM+d and IFLIM (pu)
-0.1
# f: OEL timer charging rate offset above IFLIM
0.
# S: OEL timer charging rate per unit of field overcurrent (1/pu)
1.
# K1: Not used by the code
100.
# K2: OEL gain on (if - IFLIM), negative so that the OEL lowers excitation (pu/pu)
-1.
# L1: OEL timer lower limit and initial value, negative; the OEL acts when the timer exceeds 0 (s)
-11
# L2: OEL timer upper limit (s)
10.
# G: AVR gain (pu/pu)
70.
# TA: AVR lead time constant (s)
10.
# TB: AVR lag time constant (s)
20.
# TE: Exciter time constant (s)
0.1
# L3: Minimum field voltage (pu)
0.
# L4: Maximum field voltage (pu)
4.
# SPEEDIN: PSS input weight: 1 = speed, 0 = active power
1
# KPSS: PSS gain, derivative washout KPSS s/(1+sTw); 0 disables PSS (pu s)
75.
# Tw: PSS washout time constant (s)
15.
# T1: PSS first lead-lag numerator time constant (s)
0.2
# T2: PSS first lead-lag denominator time constant (s)
0.01
# T3: PSS second lead-lag numerator time constant (s)
0.2
# T4: PSS second lead-lag denominator time constant (s)
0.01
# DVMIN: PSS output lower limit (pu)
-0.1
# DVMAX: PSS output upper limit (pu)
0.1
TOR 1ST_ORDER
# FHP: fraction of torque following the setpoint without lag (HP part)
0.30
# T2: time constant of the lagged (LP) torque part (s)
0.50
# R: speed droop (pu)
0.05
;

The data-file model name is THERMAL_GENERIC1.

The tor_thermal_generic1 model is a generic multi-stage steam turbine governor, suitable for both single-shaft and cross-compound units. It includes a speed controller with rate-limiting, a three-section steam chest/reheater, and separate HP, MP, and LP turbine stages.

THERMAL_GENERIC1 block diagram. The speed deviation from unity is divided by the droop sigma and filtered with time constant Tmes. That term and the gate position are subtracted from the power setpoint P0, divided by the servo time constant Tsm, rate-limited between zdotmin and zdotmax, and integrated between zmin and zmax to give the gate opening z. The gate drives three lags in series with time constants Thp, Tr and Tlp; the three stage powers are formed with the fractions Fhp, Fmp times the intercept valve factor ivo and one minus Fhp minus Fmp, summed into the mechanical power, and divided by the rotor speed to give the torque. THERMAL_GENERIC1 block diagram. The speed deviation from unity is divided by the droop sigma and filtered with time constant Tmes. That term and the gate position are subtracted from the power setpoint P0, divided by the servo time constant Tsm, rate-limited between zdotmin and zdotmax, and integrated between zmin and zmax to give the gate opening z. The gate drives three lags in series with time constants Thp, Tr and Tlp; the three stage powers are formed with the fractions Fhp, Fmp times the intercept valve factor ivo and one minus Fhp minus Fmp, summed into the mechanical power, and divided by the rotor speed to give the torque.

Speed governor:

The speed deviation is divided by the droop σ\sigma and measured with time constant TmesT_{mes}:

Tmes x˙1=−x1+ω−1σT_{mes}\,\dot{x}_1 = -x_1 + \frac{\omega - 1}{\sigma}

The gate demand is what remains of the power setpoint after that term and the current gate position are taken off:

egov=P0−x1−ze_{gov} = P_0 - x_1 - z

It is divided by the servo time constant TsmT_{sm}, rate-limited to [zdotmin, zdotmax][zdot_{min},\, zdot_{max}], and integrated into the gate opening zz within [zmin, zmax][z_{min},\, z_{max}].

Turbine power stages:

The gate opening drives three lags in series, the reheat stage carrying the intercept valve factor ivoivo:

Thp x˙hp=−xhp+z,Tr x˙r=−xr+xhp,Tlp x˙lp=−xlp+ivo xrT_{hp}\,\dot{x}_{hp} = -x_{hp} + z, \qquad T_r\,\dot{x}_r = -x_r + x_{hp}, \qquad T_{lp}\,\dot{x}_{lp} = -x_{lp} + ivo\,x_r

Each stage contributes its own fraction of the power, and the three fractions sum to one:

PmHP=Fhp xhp,PmMP=Fmp ivo xr,PmLP=(1−Fhp−Fmp) xlpP_{mHP} = F_{hp}\,x_{hp}, \qquad P_{mMP} = F_{mp}\,ivo\,x_r, \qquad P_{mLP} = (1 - F_{hp} - F_{mp})\,x_{lp}

Total mechanical power, and the torque delivered to the shaft:

Pm=PmHP+PmMP+PmLP,Tm=PmωP_m = P_{mHP} + P_{mMP} + P_{mLP}, \qquad T_m = \frac{P_m}{\omega}

Parameters:

ParameterIndexDescription
sigma1Speed droop (pu/pu)
Tmes2Speed measurement time constant (s)
Tsm3Servo/gate time constant (s)
zdotmin4Minimum gate rate (pu/s)
zdotmax5Maximum gate rate (pu/s)
zmin6Minimum gate opening (pu)
zmax7Maximum gate opening (pu)
Thp8HP turbine time constant (s)
Fhp9HP power fraction
Tr10Reheater / MP time constant (s)
Fmp11MP power fraction
Tlp12LP turbine time constant (s)

Implementation details:

  • nbxvar = 10, nbzvar = 2, nbdata = 12, nbaddpar = 2
  • Additional: prm(13) = P0 (initial power), prm(14) = ivo (intercept valve factor)
  • Observables: z, PmHP, PmMP, PmLP, Pm
Show the data record
SYNC_MACH g1 g1
# FP FQ P Q: share of the power-flow injection (FP, FQ) or initial power (P in MW, Q in Mvar)
1. 1. 0. 0.
# SNOM: Nominal apparent power, used as base power in the machine model (MVA)
800.
# Pnom: Nominal active power of the turbine, used as base power for the turbine model (MW)
760.
# H: Inertia constant (s)
3.
# D: Damping coefficient (usually set to zero when the damper windings are modelled) (pu)
0.
# IBRATIO: Field current base ratio I_fB^mac/I_fB^exc (see above) (pu)
0.95
# TYPE_MOD: Parameter format keyword: RL or XT (case-insensitive)
XT
# Xl: Leakage reactance L_l (pu)
0.15
# Xd: d-axis synchronous reactance (pu)
1.1
# X'd: d-axis transient reactance (must be smaller than Xd) (pu)
0.25
# X"d: d-axis subtransient reactance (pu)
0.2
# Xq: q-axis synchronous reactance (pu)
0.7
# X'q: q-axis transient reactance (pu)
*
# X"q: q-axis subtransient reactance (pu)
0.2
# m: Saturation coefficient (set to 0 to neglect saturation)
0.1
# n: Saturation exponent (ignored when m = 0)
6.0257
# Ra: Armature resistance (pu)
0.
# T'do: d-axis open-circuit transient time constant (s)
5.00
# T"do: d-axis open-circuit subtransient time constant (s)
0.05
# T'qo: q-axis open-circuit transient time constant (s)
*
# T"qo: q-axis open-circuit subtransient time constant (s)
0.1
EXC GENERIC1
# IFLIM: OEL field current threshold (pu)
1.8991
# d: OEL timer band, usually negative: timer runs down at rate 1 below IFLIM+d, holds between IFLIM+d and IFLIM (pu)
-0.1
# f: OEL timer charging rate offset above IFLIM
0.
# S: OEL timer charging rate per unit of field overcurrent (1/pu)
1.
# K1: Not used by the code
100.
# K2: OEL gain on (if - IFLIM), negative so that the OEL lowers excitation (pu/pu)
-1.
# L1: OEL timer lower limit and initial value, negative; the OEL acts when the timer exceeds 0 (s)
-11
# L2: OEL timer upper limit (s)
10.
# G: AVR gain (pu/pu)
70.
# TA: AVR lead time constant (s)
10.
# TB: AVR lag time constant (s)
20.
# TE: Exciter time constant (s)
0.1
# L3: Minimum field voltage (pu)
0.
# L4: Maximum field voltage (pu)
4.
# SPEEDIN: PSS input weight: 1 = speed, 0 = active power
1
# KPSS: PSS gain, derivative washout KPSS s/(1+sTw); 0 disables PSS (pu s)
75.
# Tw: PSS washout time constant (s)
15.
# T1: PSS first lead-lag numerator time constant (s)
0.2
# T2: PSS first lead-lag denominator time constant (s)
0.01
# T3: PSS second lead-lag numerator time constant (s)
0.2
# T4: PSS second lead-lag denominator time constant (s)
0.01
# DVMIN: PSS output lower limit (pu)
-0.1
# DVMAX: PSS output upper limit (pu)
0.1
TOR THERMAL_GENERIC1
# sigma: speed droop (pu)
0.05
# Tmes: speed measurement time constant (0 = no lag) (s)
0.02
# Tsm: valve servomotor time constant (s)
0.20
# zdotmin: minimum valve rate, applied as given so normally negative (pu/s)
-0.10
# zdotmax: maximum valve rate (pu/s)
0.10
# zmin: minimum valve position (pu)
0.0
# zmax: maximum valve position (pu)
1.05
# Thp: HP steam chest time constant (0 = no lag) (s)
0.30
# Fhp: HP power fraction
0.30
# Tr: reheater time constant (0 = no lag) (s)
7.0
# Fmp: MP power fraction
0.40
# Tlp: LP stage time constant (0 = no lag) (s)
0.50
;

The data-file model name is HYDRO_GENERIC1.

The tor_hydro_generic1 model is a generic hydraulic turbine governor with a penstock (water column) and PI controller. It implements the standard hydraulic governor structure following IEEE Std 1110 conventions.

PI speed governor:

The speed error with droop σ\sigma:

e=(P0−Pmeas)⋅σ+(1−ω)e = (P_0 - P_{meas}) \cdot \sigma + (1 - \omega)

A PI controller with gains KPK_P and KIK_I drives the gate demand:

x˙I=KI⋅e\dot{x}_{I} = K_I \cdot e zdemand=KP⋅e+xIz_{demand} = K_P \cdot e + x_I

The gate is rate-limited by LIMz˙LIM_{\dot{z}} and position-limited by [0,1][0, 1], with a servo time constant TsmT_{sm}.

Water column (penstock):

The flow QQ and head HH are related by the water-starting time TWT_W:

TW⋅dQdt=1−HT_W \cdot \frac{dQ}{dt} = 1 - H

The head is:

H=(Qv+Tm(1−Qv)z)2H = \left(\frac{Q_{v} + T_m (1 - Q_v)}{z}\right)^2

where QvQ_v is the no-load flow fraction.

Turbine mechanical torque:

Tm=H⋅(Q−Qv⋅H)T_m = \sqrt{H} \cdot (Q - Q_v \cdot \sqrt{H}) Pm=Tm⋅ωP_m = T_m \cdot \omega

Parameters:

ParameterIndexDescription
SIGMA1Speed droop (pu/pu)
Tmes2Speed measurement time constant (s)
Qv3No-load water flow fraction (pu)
KP4PI proportional gain
KI5PI integral gain
TSM6Gate servo time constant (s)
LIMZDOT7Gate velocity limit (pu/s)
TW8Water starting time (s)

Implementation details:

  • nbxvar = 6, nbzvar = 2, nbdata = 8, nbaddpar = 1
  • Additional: prm(9) = P0 (initial electrical power)
  • Observables: z, Q, H, Pm
HYDRO_GENERIC1 block diagram. One minus omega plus the droop term, the power error times sigma, drives a PI controller with gains Kp and Ki. The gate demand is rate-limited, position-limited between zero and one, and follows a servo of time constant Tsm to give the gate z. The water column relates flow and head through the water-starting time TW, and the turbine power follows from the head and flow. HYDRO_GENERIC1 block diagram. One minus omega plus the droop term, the power error times sigma, drives a PI controller with gains Kp and Ki. The gate demand is rate-limited, position-limited between zero and one, and follows a servo of time constant Tsm to give the gate z. The water column relates flow and head through the water-starting time TW, and the turbine power follows from the head and flow.
Show the data record
SYNC_MACH g1 g1
# FP FQ P Q: share of the power-flow injection (FP, FQ) or initial power (P in MW, Q in Mvar)
1. 1. 0. 0.
# SNOM: Nominal apparent power, used as base power in the machine model (MVA)
800.
# Pnom: Nominal active power of the turbine, used as base power for the turbine model (MW)
760.
# H: Inertia constant (s)
3.
# D: Damping coefficient (usually set to zero when the damper windings are modelled) (pu)
0.
# IBRATIO: Field current base ratio I_fB^mac/I_fB^exc (see above) (pu)
0.95
# TYPE_MOD: Parameter format keyword: RL or XT (case-insensitive)
XT
# Xl: Leakage reactance L_l (pu)
0.15
# Xd: d-axis synchronous reactance (pu)
1.1
# X'd: d-axis transient reactance (must be smaller than Xd) (pu)
0.25
# X"d: d-axis subtransient reactance (pu)
0.2
# Xq: q-axis synchronous reactance (pu)
0.7
# X'q: q-axis transient reactance (pu)
*
# X"q: q-axis subtransient reactance (pu)
0.2
# m: Saturation coefficient (set to 0 to neglect saturation)
0.1
# n: Saturation exponent (ignored when m = 0)
6.0257
# Ra: Armature resistance (pu)
0.
# T'do: d-axis open-circuit transient time constant (s)
5.00
# T"do: d-axis open-circuit subtransient time constant (s)
0.05
# T'qo: q-axis open-circuit transient time constant (s)
*
# T"qo: q-axis open-circuit subtransient time constant (s)
0.1
EXC GENERIC1
# IFLIM: OEL field current threshold (pu)
1.8991
# d: OEL timer band, usually negative: timer runs down at rate 1 below IFLIM+d, holds between IFLIM+d and IFLIM (pu)
-0.1
# f: OEL timer charging rate offset above IFLIM
0.
# S: OEL timer charging rate per unit of field overcurrent (1/pu)
1.
# K1: Not used by the code
100.
# K2: OEL gain on (if - IFLIM), negative so that the OEL lowers excitation (pu/pu)
-1.
# L1: OEL timer lower limit and initial value, negative; the OEL acts when the timer exceeds 0 (s)
-11
# L2: OEL timer upper limit (s)
10.
# G: AVR gain (pu/pu)
70.
# TA: AVR lead time constant (s)
10.
# TB: AVR lag time constant (s)
20.
# TE: Exciter time constant (s)
0.1
# L3: Minimum field voltage (pu)
0.
# L4: Maximum field voltage (pu)
4.
# SPEEDIN: PSS input weight: 1 = speed, 0 = active power
1
# KPSS: PSS gain, derivative washout KPSS s/(1+sTw); 0 disables PSS (pu s)
75.
# Tw: PSS washout time constant (s)
15.
# T1: PSS first lead-lag numerator time constant (s)
0.2
# T2: PSS first lead-lag denominator time constant (s)
0.01
# T3: PSS second lead-lag numerator time constant (s)
0.2
# T4: PSS second lead-lag denominator time constant (s)
0.01
# DVMIN: PSS output lower limit (pu)
-0.1
# DVMAX: PSS output upper limit (pu)
0.1
TOR HYDRO_GENERIC1
# SIGMA: permanent droop, gain on the power error P0 minus measured P (pu)
0.04
# Tmes: electrical power measurement time constant (s)
2.0
# Qv: no-load water flow (pu)
0.
# KP: PI proportional gain
2.00
# KI: PI integral gain (1/s)
0.40
# TSM: gate servomotor time constant (s)
0.2
# LIMZDOT: gate velocity limit, applied as plus or minus (pu/s)
0.1
# TW: water starting time (s)
1.0
;

The data-file model name is HYDRO_DG. Registered since 3.76.

A hydraulic turbine governor for a dispersed generation unit that holds an active-power setpoint and does not provide primary frequency response.

The penstock and turbine are exactly those of HYDRO_GENERIC1. What differs is the governor control law, and the difference is the point of the model. HYDRO_GENERIC1 is a speed governor: its error signal carries an unconditional frequency term, so the unit responds to system frequency whether or not that is wanted. HYDRO_DG regulates power alone.

PI power regulator:

The error is the deviation from the setpoint captured at initialisation, with no droop term and no speed term:

e=P∗−Pe = P^{*} - P x˙I=Ki⋅e,zdemand=Kp⋅e+xI\dot{x}_{I} = K_i \cdot e, \qquad z_{demand} = K_p \cdot e + x_{I}

There is no measurement lag on PP, so the regulator acts on the instantaneous electrical power. The gate follows through a servo of time constant TsmT_{sm}, rate-limited to ±LIMz˙\pm LIM_{\dot{z}} and position-limited to [0,1][0, 1].

Water column and turbine, identical to HYDRO_GENERIC1:

TW⋅dQdt=1−H,H=(Qz)2T_W \cdot \frac{dQ}{dt} = 1 - H, \qquad H = \left(\frac{Q}{z}\right)^{2} Tm⋅ω=(Q−Qv) H1−QvT_m \cdot \omega = \frac{(Q - Q_v) \, H}{1 - Q_v}

Parameters:

ParameterIndexDescription
Qv1No-load water flow fraction (pu)
Kp2PI proportional gain
Ki3PI integral gain
TSM4Gate servo time constant (s)
LIMZDOT5Gate velocity limit (pu/s)
TW6Water starting time (s)

These are exactly parameters 3 to 8 of HYDRO_GENERIC1, in the same order. HYDRO_GENERIC1 precedes them with SIGMA (droop) and Tmes (speed measurement lag), and those two are what HYDRO_DG does not have. Setting SIGMA to zero in HYDRO_GENERIC1 does not reproduce HYDRO_DG, because the frequency term survives.

Implementation details:

  • nbxvar = 5, nbzvar = 2, nbdata = 6, nbaddpar = 1
  • Additional: prm(7) = Pset, the initial electrical power, captured at initialisation
  • Observables: z, Q, H, Pm, Pset
HYDRO_DG block diagram. The error is the power setpoint minus the instantaneous electrical power, with no droop and no speed term, and drives a PI regulator with gains Kp and Ki. The gate demand is rate-limited, position-limited between zero and one, and follows a servo of time constant Tsm. The penstock and turbine are exactly those of HYDRO_GENERIC1. HYDRO_DG block diagram. The error is the power setpoint minus the instantaneous electrical power, with no droop and no speed term, and drives a PI regulator with gains Kp and Ki. The gate demand is rate-limited, position-limited between zero and one, and follows a servo of time constant Tsm. The penstock and turbine are exactly those of HYDRO_GENERIC1.
Show the data record
SYNC_MACH G4 N4
# FP FQ P Q: share of the power-flow injection (FP, FQ) or initial power (P in MW, Q in Mvar)
1. 1. 0. 0.
# SNOM: Nominal apparent power, used as base power in the machine model (MVA)
5
# Pnom: Nominal active power of the turbine, used as base power for the turbine model (MW)
4.5
# H: Inertia constant (s)
3.
# D: Damping coefficient (usually set to zero when the damper windings are modelled) (pu)
0.
# IBRATIO: Field current base ratio I_fB^mac/I_fB^exc (see above) (pu)
1.9333
# TYPE_MOD: Parameter format keyword: RL or XT (case-insensitive)
XT
# Xl: Leakage reactance L_l (pu)
0.0937
# Xd: d-axis synchronous reactance (pu)
2.027
# X'd: d-axis transient reactance (must be smaller than Xd) (pu)
0.210
# X"d: d-axis subtransient reactance (pu)
0.173
# Xq: q-axis synchronous reactance (pu)
2.027
# X'q: q-axis transient reactance (pu)
*
# X"q: q-axis subtransient reactance (pu)
0.173
# m: Saturation coefficient (set to 0 to neglect saturation)
0.1
# n: Saturation exponent (ignored when m = 0)
6
# Ra: Armature resistance (pu)
0.00
# T'do: d-axis open-circuit transient time constant (s)
3.1
# T"do: d-axis open-circuit subtransient time constant (s)
0.0387
# T'qo: q-axis open-circuit transient time constant (s)
*
# T"qo: q-axis open-circuit subtransient time constant (s)
0.0387
EXC AVR_DG
# IFLIM: OEL field current threshold (pu)
2.865
# d: OEL timer band, usually negative: timer runs down at rate 1 below IFLIM+d, holds between IFLIM+d and IFLIM (pu)
-0.1
# f: OEL timer charging rate offset above IFLIM
0.
# S: OEL timer charging rate per unit of field overcurrent (1/pu)
1.
# K1: Not used by the code
100.
# K2: OEL gain on field current error if minus IFLIM (pu/pu)
-1.
# L1: OEL timer lower limit (initial value) (s)
-20
# L2: OEL timer upper limit (s)
10
# G: AVR gain (pu/pu)
50
# TA: AVR lead time constant (s)
4
# TB: AVR lag time constant (s)
20
# TE: Exciter time constant (s)
0.1
# L3: Minimum field voltage (pu)
0
# L4: Maximum field voltage (pu)
5
# SPEEDIN: PSS input weight: 1 = speed, 0 = active power
1
# KPSS: PSS gain, derivative washout KPSS s/(1+sTw); 0 disables PSS (pu s)
0.
# Tw: PSS washout time constant (s)
5.
# T1: PSS first lead-lag numerator time constant (s)
1.
# T2: PSS first lead-lag denominator time constant (s)
1.
# T3: PSS second lead-lag numerator time constant (s)
1.
# T4: PSS second lead-lag denominator time constant (s)
1.
# DVMIN: PSS output lower limit (pu)
0.
# DVMAX: PSS output upper limit (pu)
0.
# KQP: Reactive power regulator proportional gain (pu/pu)
0.04
# KQI: Reactive power regulator integral gain (1/s)
0.06
# L5: Reactive power regulator output lower limit (pu)
-0.2
# L6: Reactive power regulator output upper limit (pu)
0.33
TOR HYDRO_DG
# Qv: no-load water flow (pu)
0.0
# Kp: PI proportional gain on power error
0.1
# Ki: PI integral gain on power error (1/s)
0.4
# TSM: gate servomotor time constant (s)
0.2
# LIMZDOT: gate velocity limit, applied as plus or minus (pu/s)
0.1
# TW: water starting time (s)
1.0
;