Energy and Power Engineering, 2013, 5, 32-38

http://dx.doi.org/10.4236/epe.2013.52A005 Published Online April 2013 (http://www.scirp.org/journal/epe)

Virtual Synchronous Generator: A Control Strategy to

Improve Dynamic Frequency Control in Autonomous
Power Systems
Miguel Torres, Luiz A. C. Lopes
Department of Electrical and Computer Engineering, Concordia University, Montreal, Canada
Email: mi_torre@ ece.concordia.ca, [email protected]

Received February 28, 2013; revised March 28, 2013; accepted April 15, 2013

Copyright © 2013 Miguel Torres, Luiz A. C. Lopes. This is an open access article distributed under the Creative Commons Attribu-
tion License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

ABSTRACT
Ideally, diesel hybrid autonomous power systems would operate with high penetration of renewable energy sources
such as wind and photovoltaic to minimize fuel consumption. However, since these are inherently intermittent and
fluctuating, the grid-forming diesel engine generator sets are usually required to operate with larger amounts of spinning
reserve, often at low loading conditions what tends to increases operating and maintenance costs. Frequency stability is
of great concern in “small” systems, such as mini-grids, where any individual generator in-feed represents a substantial
portion of the total demand. There, the initial rate of change of frequency is typically larger and a lower value of fre-
quency can be reached in a shorter time than in conventional systems with all generation supplied by rotating machines,
possibly resulting in under-frequency load shedding and tripping of renewable energy generators. The first part of this
paper, discusses some general concepts regarding frequency stability in a diesel hybrid mini-grid and how energy stor-
age systems can be used to enhance system performance. Then, a particular technique based on a virtual synchronous
generator is presented and its effectiveness is demonstrated with simulation results.

Keywords: Frequency Control; Mini-Grid; Energy Storage; Inverter

1. Introduction (secondary control) optimize the operation of the power

sources, energy storage units and controllable loads, if
Historically, renewable energy sources (RESs) are inte-
applicable.
grated into autonomous power systems (mini-grids)
Figure 1 shows a schematic diagram of a supervisory
already supplied by a diesel power plant with multiple
controller of a diesel power plant. Each gen-set presents
diesel engine generator sets (gen-sets). With the addition
a primary controller, implemented in the governor,
of RES, it is more likely that the gen-sets will operate
whose main function is to adjust the speed of the gen-set.
under low load conditions resulting in lower efficiency
Since most gen-sets employ a synchronous generator, the
and potential of carbon build up and premature ageing of
governor indirectly controls the frequency of the gene-
the diesel engines. Re-sizing of the gen-sets and imple-
rated voltage. If the frequency is to be kept constant, an
mentation of appropriate cycling strategies, according to
isochronous governor can be used. However, if multiple
daily and seasonal variations of the mini-grid load and
gen-sets are to operate in parallel, it is more convenient
power output of the RESs to maintain optimum loading
to use a droop based governor where the speed and
under various conditions, can help improve the RES con-
frequency of a gen-set decrease as its output power in-
tribution and displacement of diesel fuel [1]. Besides,
creases. This behavior can be represented in steady-state
one can also include short- or long-term energy storage
by
systems (ESSs) and controllable (secondary) loads that
1
can be used for improving stability and economic per- P  Psch   f 0  f  (1)
formance of the mini-grid. This can be done by a super- R
visory controller which should, at one level, (primary where P is the actual power supplied by the gen-set,
control) maintain grid stability and, at the other level, f is the grid frequency, Psch and f 0 are input values

Copyright © 2013 SciRes. EPE

 M. TORRES, L. A. C. LOPES 33

Figure 1. Diesel power plant.

that can be set by the secondary control and R is the only require the frequency in autonomous power systems
droop factor usually selected so that f decreases by about to be within ±2% of the rated value (50 Hz) during 95%
3% - 5% as P varies from no-load to full-load. In the of the week and within 15% during 100% of the week [2].
isochronous mode, R is set to 0, meaning that the grid Besides, operation with variable frequency can be ad-
frequency will not change regardless of the output power vantageous because it naturally conveys the information
level. of shortage (low frequency) or excess (high frequency)
Frequency stability issues usually appear up to a few of active power in the mini-grid. This would allow con-
seconds following variations in the power demanded trollable loads, ESSs and power sources dispersed in the
from the diesel plant. They are addressed primarily by mini-grid to adjust their active power levels accordingly,
the primary control blocks which should make sure that without the need for a dedicated communication channel
the rate of change of frequency (ROCOF) and frequency [3-5].
excursions remain within acceptable values. One impor- The supervisory control system of a diesel-hybrid
tant element in the variation of the grid frequency after mini-grid with short and long-term ESSs and controllable
variations in demand is the amount of kinetic energy loads can be made very sophisticated, with multiple hier-
stored in the rotating masses of the diesel power plant, archical control levels, employing modern information
which tend to reduce frequency variations. When a diesel and communication technology (ICT) [6]. This paper
hybrid mini-grid operates with high penetration of iner- however focuses on frequency transient stability issues
tia-less power sources, such as photovoltaic (PV), fre- which can be appropriately studied considering only the
quency stability can deteriorate. Wind turbines contribute primary control only of the gen-sets and ESSs.
to a certain extent with inertial response, but not as a
gen-set of equivalent power since they operate at low 2. Energy Storage Systems
rotational speeds and many of the small ones are of the
fixed pitch type, directly connected to the mini-grid via a As mentioned before, the stochastic nature of wind and
squirrel-cage induction generator. PV can put a lot of stress on the grid-forming diesel
The secondary control, adjusts the values of and/or to power plant to balance demand and supply. As a result,
bring the frequency of the mini-grid back to a desired the system might present frequency stability (power
value and/or to adjust the amount of power provided by quality) issues due to a lower amount of rotating masses
each parallel gen-set so as to keep them loaded at safe (inertia), and require the diesel power plant to operate
and more economical levels as computed by an economic with higher spinning reserve (economical operation). The
dispatch algorithm. However, due to a slow speed of use of energy storage units can solve not only these
response (bandwidth), the secondary control cannot make problems but also minimize fuel consumption in the long
a meaningful contribution regarding frequency stability. term. However, since conventional battery storage is
One approach based on the use of a short-term ESS and usually one of the most expensive components of a mini-
that can reduce this problem is discussed in Section 3. grid over its life-time, there are cost advantages to eli-
Without the secondary control, the frequency of the minating energy storage entirely, or incorporating only
mini-grid will vary according to the loading of the diesel smaller units to assist with transient dynamics. Another
power plant and the power vs. frequency droop charac- way to avoid higher diesel based spinning reserve is load
teristics of the gen-sets in operation. Although systems shedding, but its “cost” to system users, in terms of per-
operating with tightly regulated frequencies are custo- ceived loss of power quality and reliability, has to be
mary, some standards such as the European EN50160 balanced with the other options.

Copyright © 2013 SciRes. EPE

34 M. TORRES, L. A. C. LOPES

2.1. ESS Power Configurations power reference of distributed ESS (pESSref) is frequency
droop [9]. A major advantage of this technique is that
The choice of the ESS configuration depends on the
allows the autonomous control of ESSs, i.e. no need for
requirements of the application. Since in this particular
communication link, since the control relies on local
case the objective is to improve the frequency stability of
measurements. However, in its simplest form (static fre-
the diesel hybrid mini-grid and provide storage based
spinning reserve, the discussion will be constrained to quency droop) it only guarantees frequency regulation
short-term storage media. support for a reduced amount of time, and it does not
Short-term media like super-capacitors present fast cope with the problem of dynamic frequency control. To
dynamics, being able to supply and absorb large and fre- address the latter, a more sophisticated frequency droop
quent bursts of power, but present lower energy density control is needed.
than batteries [7]. Conversely, batteries present higher A strategy that indirectly deals with dynamic fre-
capacity and efficiency for storing energy which would quency control is the output power smoothing of inter-
be better employed in a more paced way. High speed mittent sources [10]. This technique entails generating a
flywheels could be a good choice for both requirements suitable power reference for the ESS that cancels the
[8]. Another option would be to use hybrid ESS [7] like high frequency components of the intermittent power. By
the ones shown in Figure 2, with two different types of doing this, the fast frequency variations induced by
storage technologies, batteries and super-capacitors (SC). intermittent sources are attenuated. One disadvantage of
The cascaded connection implies that the dc bus voltage, this technique is that it requires the measuring of the
vdc , must be controlled by the converter C1, while C2 is output power of the intermittent source, which restricts
controlled to supply the low frequency components of the location of the ESS or establishes the need for a
the current demanded by the dc-ac converter, lowering communication link to transmit the measurements.
the stress on the batteries. A disadvantage is that con- Although different approaches to support dynamic fre-
verter C1 has to support all the current flow and, if it fails, quency control can be found in the literature, one of the
then C2 cannot be used. On the other hand, in the parallel most discussed is the emulation of inertia [11]. In general,
connection both converters can work independently and inertia emulation techniques support diesel generator
their state-of-charge (SoC) can be used to adjust their response right after a load change occurs, keeping fre-
compensation efforts accordingly. quency under more restricted limits, and therefore con-
The dc-ac converter is controlled in the PQ mode, tributing to maintain or enhance power quality. In the
injecting the required amounts of active (P) and reactive following section we discuss this technique as part of a
(Q) power independently. The references values for P control strategy called virtual synchronous generator.
and Q can be provided locally or sent by the supervisory
controller of the mini-grid through a slow communi- 3. Virtual Synchronous Generator
cation channel.
Virtual synchronous generator (VSG) is a control strat-
egy that has been proposed to mitigate stability issues in
2.2. ESS Control Strategies
power systems with large fraction of distributed genera-
One of the most common strategies to generate the active tion [12]. The concept behind VSG is to control the grid-

Figure 2. Hybrid ESS with two possible configurations and the PQ control scheme for the dc-ac converter.

Copyright © 2013 SciRes. EPE

 M. TORRES, L. A. C. LOPES 35

interface converter of a distributed generator or ESS in

order to emulate a synchronous generator or a desired
characteristic of it. Different realizations of the VSG can
be developed depending on the application. In this par-
ticular case, it is desirable to emulate characteristics that
can be used to support frequency control, such as inertial
response, damping and droop control. Therefore, the
output power of the VSG can be synthesized according
to:
f  f df
pVSG 
R 

  

k D f  f  k J v
dt
(2)
 damping

  Figure 3. Block diagram of a VSG control performing only
droop inertia the virtual inertia function.
where f is the measured grid frequency. The droop term
is calculated from a frequency droop curve as indicated measured frequency, the virtual inertia algorithm cal-
in [9]. The other two terms in (2), damping and inertia, culates the output power of the VSG, which turns out to
are present only during the transient, and therefore they be the reference for the output power of the ESS.
are more related to the power capacity of the ESS rather
than the energy capacity. 3.1.1. Linear Variation
The damping term, with k D being the damping This algorithm starts to increase the value of the virtual
coefficient and f  the stabilization frequency, is the inertia if the frequency varies faster than a predefined
power that helps to attenuate oscillations. Note that in maximum rate (MAX). On the other hand, it starts to
order to provide damping power in a proper way, the linearly decrease the virtual inertia when the frequency
stabilization frequency should be known a priori. In the varies slower than a minimum rate (MIN). The algorithm
case of a system operating with isochronous primary fre- is executed in every sampling time of the controller and
quency control, f  should be the nominal frequency of it can be summarized in the following expressions:
the system. However, in the case of droop control, the df  MAX i  i 1
stabilization frequency will change with the load level  (3)
and, therefore, an estimation technique is needed to dt  MIN i  i 1
determine its value. kJv  k i 
Finally, the inertial term, with k Jv representing the
virtual inertia, emulates the inertial response of a con- df
ventional generator, whose main effects over the system pVSG  k J v π
dt
frequency are: decreasing the rate of change and the
where k is a predefined monotonically increasing vector
maximum deviation following a power disturbance. In
of values of virtual inertia, i is an integer counter used in
order to limit the extension of the paper, only the effects
the algorithm to access the elements of the vector k, and
of the emulation of inertia on the dynamic frequency
π is the conversion factor between frequency and
control are presented. To the best knowledge of the
generator speed.
authors, studies published so far have focused on tech-
niques that consider a constant virtual inertia. The strategy
3.1.2. Optimal Inertia
proposed in this paper entails the online modification of
This algorithm finds a value of virtual inertia, from a
this parameter.
predefined finite set of values, which minimizes the cost
function g . The optimization problem is solved in every
3.1. Self-Tuning Virtual Inertia
sampling time and it can be formulated as follows:
The idea was motivated by the fact that when frequency df
starts to deviate from steady-state, a larger inertia would min g   π
k J v k dt
 
Ts , k J v  k J v (4)
present a stronger opposition to the speed change of  0
generators, limiting the initial rate of change of fre-
df
quency and its peak deviation. On the other hand, when pVSG  k J v π
frequency starts to return to steady-state after the peak dt
deviation, a larger inertia is no longer required. Figure 3 where df dt  Ts , k Jv  is the one sampling time ahead
depicts a block diagram of the VSG control for emulating prediction of the rate of change of frequency. This
inertia. The grid frequency is locally measured at the prediction is based on Equation (2.1) in [13] that des-
point where the ESS is connected and then, based on the cribes the dominant dynamics of the system frequency.

Copyright © 2013 SciRes. EPE

36 M. TORRES, L. A. C. LOPES

On the other hand, the second term of the cost function is Table 2. Parameters of wind generator.
included to represent the output power and the weight Description Value Unit
factor  determines the relative importance between
Power at 9.5 m/s 50 kW
the two terms.
Cut-in wind speed 3.5 m/s
4. Performance Verification Nominal frequency 60 Hz
Power regulation Stall control n.a.
To verify the performance enhancement offered by the
proposed control strategy, a simulation is conducted for Rotor speed 42 rpm
the system of Figure 4, which represents a diesel-hybrid Rotor diameter 19 m
autonomous power system. It consists in two identical Equivalent Inertia 5.83 kgm2
diesel gen-sets as primary sources of energy (DG1, DG2),
a type 1 wind generator (WG) used to complement power The proposed algorithms are compared with the con-
production, an energy storage system, and several re- stant virtual inertia in two cases of transient condition: a
sistive loads. The main parameters of the diesel gensets load increase and wind variations. In all cases, the con-
and wind generator are presented in Tables 1 and 2 re- stant virtual inertia has a fixed value of 5 kgm2, the
spectively. Since the secondary control is slower than the vector k is defined with 20 elements linearly spaced from
primary one, it will have no impact over the system
0 to 5 kgm2, the sampling time is Ts = 10 ms, MIN = 0.1
frequency during the time frame of the simulations and it
Hz/s, MAX = 1 Hz/s, and   4 .
will not be considered in this study. The system is si-
mulated for the critical case when only one gen-set is in
4.1. Load Increase
operation (DG1).
The system starts operating with DG1 supplying a load
of 25 kW and with pESS  0 kW . The switch SL is
closed at t  0 s , producing a load increase of 5 kW.
Figure 5 shows the simulation results.
In comparison to the gen-set only response (waveform
labeled DG1), all the inertia emulation methods reduce
the maximum frequency deviation. The rate of change of
frequency is also reduced. However the constant and
variable inertia result in noticeable oscillations during the
first 0.2 s. This might be a consequence of the numerical
calculation of the derivative with insufficient pre-
Figure 4. Schematic diagram of the diesel-hybrid autono- filtering of the measured frequency. On the other hand,
mous power system. the optimal algorithm did not present much oscillation,
which might be because it used a first order model to
Table 1. Parameters of diesel generators. calculate the derivative of the frequency and its pre-
Description Symbol Value Unit diction.
In the bottom plot it is observed that the optimal
Rated power n.a. 37.7 kW
inertia algorithm started to decrease the value of inertia (t
Nominal speed n.a. 1800 rpm ≈ 0.4 s) before than the variable inertia algorithm (t ≈ 2.1
Nominal frequency n.a. 60 Hz s). The ESS output power started to decrease accordingly,
Inertia J 1.6 kgm2
and therefore the frequency deviation was slightly greater
for the optimal inertia.
Friction losses D 0.18 kgm2/s

Scheduled power Psch 0 kW 4.2. Wind Variations

No-load frequency f0 63 Hz The system starts operating with DG1 and the WG
Droop coefficient R 0.5 (kWs) –1 supplying a load of 30 kW. The ESS is enabled at
t  0.1 s . The wind speed varies according to a real wind
Conversion factor kc π n.a.
profile obtained from [14]. Figure 6 shows the simul-
Proportional gain kp 1.9 kWs ation results. It is observed that all the algorithms of
Integral gain ki 5.7 kW inertia emulation achieved a similar performance in
terms reduction of frequency variations. However, from
Time constant v 43 ms
the middle plot, it can be noted that the optimal inertia

Copyright © 2013 SciRes. EPE

 M. TORRES, L. A. C. LOPES 37

storage systems (ESSs) that can enhance system stability

and reduce fuel consumption. While controlling a ESS so
as to smooth the fluctuating power of a co-located RES is
beneficial, it would not mitigate problems caused by load
variations, that are relatively large and common in re-
mote community type systems. The control strategy
called virtual synchronous generator (VSG), which can
tackle both problems was described. Then, a case study
of a wind-diesel system with a short-term ESS controlled
as a virtual inertia, a sub-set of a VSG, was presented.
Simulation results showed that an ESS controlled as a
virtual inertia can enhance the dynamic frequency control
of a mini-grid subject to fluctuating renewable energy
and load variations. The concept of variable/optimal
inertia was introduced in this paper. In general, it pro-
vided an improvement in the grid frequency response
while requiring less power as compared to the con-
ventional fixed virtual inertia scheme.

REFERENCES
Figure 5. Comparison of virtual inertia strategies for a load
increase. [1] F. Katiraei and C. Abbey, “Diesel Plant Sizing and Per-
formance Analysis of a Remote Wind-Diesel Microgrid,”
IEEE Power Engineering Society General Meeting, Tam-
pa, 24-27 June 2007, pp. 1-8.
[2] European Committee for Electrotechnical Standardization,
“EN50160 European Standard on Voltage Characteristics
of Electricity Supplied by Public Distribution Systems,”
European Committee for Electrotechnical Standardization,
Brussels, 2010.
[3] R. Tonkoski and L. A. C. Lopes, “Enhanced Part Load
Operation of Hybrid Mini-Grids with High Penetration of
Photovoltaics,” 3rd Brazilian Conference on Solar En-
ergy, Belém, 21-24 September 2010.
[4] K. Elamari and L. A. C. Lopes, “Frequency Based Con-
trol of Electric Water Heaters in Small PV-Diesel Hybrid
Mini-Grids,” 25th Canadian Conference on Electrical
and Computer Engineering, Montréal, 29 April-2 May
2012.
[5] L. A. C. Lopes and M. Dalal-Bachi, “Economic Dispatch
and Demand Side Management via Frequency Control in
PV-Diesel Hybrid Mini-Grids,” 6th European Conference
on PV-Hybrid and Mini-Grids, Chambéry, 24-26 April
2012.
[6] F. Katiraei, R. Iravani, N. Hatziargyriou and A. Dimeas,
Figure 6. Comparison of virtual inertia strategies for wind “Microgrids Management,” IEEE Power and Energy Ma-
variations. gazine, Vol. 6, No. 3, 2008, pp. 54-65.
doi:10.1109/MPE.2008.918702
algorithm used less power. [7] S. Lukic, S. Wirasingha, F. Rodriguez, J. Cao and A.
Emadi, “Power Management of an Ultracapacitor/Battery
5. Conclusion Hybrid Energy Storage System in an HEV,” IEEE Vehi-
cle Power and Propulsion Conference (VPPC’06), Win-
This paper addressed the issue of dynamic frequency dsor, 6-8 September 2006.
control in diesel-hybrid autonomous power systems [8] N. Hamsic, A. Schmelter, A. Mohd, E. Ortjohann, E.
(mini-grids) with intermittent renewable energy sources Schultze, A. Tuckey and J. Zimmermann, “Increasing
(RES). It discussed basic concepts related to the con- Renewable Energy Penetration in Isolated Grids Using a
figuration and control of a diesel power plant and energy Flywheel Energy Storage System,” International Confer-

Copyright © 2013 SciRes. EPE

38 M. TORRES, L. A. C. LOPES

ence on Power Engineering, Energy and Electrical Dri- Penetration,” 2010 International Conference on Power
ves, 12-14 April 2007, pp. 195-200. System Technology (POWERCON), Hangzhou, 24-28 Oc-
[9] P. C. Loh and F. Blaabjerg, “Autonomous Control of Dis- tober 2010.
tributed Storages in Microgrids,” 2011 IEEE 8th Interna- [12] Q. C. Zhong and G. Weiss, “Synchronverters: Inverters
tional Conference on Power Electronics and ECCE Asia That Mimic Synchronous Generators,” IEEE Transac-
(ICPE ECCE), Jeju, 30 May- 3 June 2011, pp. 536-542. tions on Industrial Electronics, Vol. 58, No. 4, 2011, pp.
[10] W. Li, G. Joós and C. Abbey, “Wind Power Impact on 1259-1267. doi:10.1109/TIE.2010.2048839
System Frequency Deviation and an ESS Based Power [13] H. Brevani, “Robust Power System Frequency Control,”
Filtering Algorithm Solution,” 2006 IEEE Power Systems Springer Science+Business Media, New York, 2009.
Conference and Exposition (PSCE’06), Atlanta, 29 Oc- [14] “Renewable Energy Cheaper than Coal,” Wind Data Files.
tober-1 November 2006, pp. 2077–2084. http://code.google.com/p/google-rec-csp/downloads/list
[11] Y. Z. Sun, Z. S. Zhang, G. J. Li and J. Lin, “Review on
Frequency Control of Power Systems with Wind Power

Copyright © 2013 SciRes. EPE