Automatic power generation control structure for smart electrical power grids

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
Volume: 03 Special Issue: 03 | May-2014 | NCRIET-2014, Available @ http://www.ijret.org 75
AUTOMATIC POWER GENERATION CONTROL STRUCTURE FOR
SMART ELECTRICAL POWER GRIDS
G.B.Hangargi1
1
Associate Professor, Electronics and communication Enggdept, BheemannaKhandreInstitute of Technology Bhalki,
Karnataka
Abstract
Smart electrical power grids with high saturation of intermittent, non dispatchable distributed energy resources (ING), and virtual
networks with real-time rating for a new tools for automatic electrical power generation with control (A EPGC) structure. The main
drawback of the intermittency are higher use of expensive spinning reserves, usually gas-fired units, nonrenewable sources units and
thermal units because of variation of turbine valves of steam units and hydro units for load regulation system. This paper presents a
new AEPGC structure to overcome the drawback of intermittency in electrical smart power grids. The constant pressure steam (CPS),
thermal units and nuclear system of variable pressure steam (VPS) are controlled to low and high frequencies load dynamic based
responses on a one-step head in advance load target. The distribution community storage (DCS) together with load dynamic response
(DR) and aggregate DR as virtual generator (DR/VG) assets are used to control high-frequency load fluctuations. The INGs and
nuclear units are used as base loads system.
Keywords: Automatic electrical power generation control, Cyber virtual network, Intermittent, renewable, non renewable,
smart power grids system.
----------------------------------------------------------------------***--------------------------------------------------------------------
1. INTRODUCTION
There port on annual bases of energy system [1] It observed
that up to by 2040, the combined intermittent, non dispatchable
ING sources are expected to deliver more than5% of total
generation in electric power(energy). Impediments to
increasing this percentage of electric power from (i) to install
costly spinning-reserves needs (ii) for load control purposed
the wear and tear of thermal and gas fired units are used and
(iii) carbon track of nonrenewable units are used. The author (I)
submit that by the use of a grid-pervasive installation of
distributed community storage (DCS) and demand
response/virtual generator (DR/VG) units, both in utility
substations and micro grids and automatic monitored by
independent system operator (ISO), automatic electrical power
generation control (AEPGC) will have the needed assets to
control the intermittency of INGs. In the proposed AEPGC
structure, AEPGC controls variable pressure steam (VPS),
constant pressure steam (CPS) units used for low frequency
load fluctuations proposed, DCS and DR/VG assets effectively
control the high frequency load variations this paper (I)
addresses the following:
1. Current AEPGC software and with its drawbacks.
2 .A new AEPGC and virtual cyber architecture to used in the
place of the intermittency of INGs
3 One-step-ahead on load forecast (Future load) for valve point
can be controlled by a CPS assets.
4 DCS and (VGs) are to control intermittency of intermittent,
non dispatch able distributed energy resources (INGs).
5 Optimal dynamic load controls of VPS units to follow a one-
step-ahead load goal.
2. AUTOMATICPOWER GENERATION
CONTROL SYSTEM
The objectives of automatic power generation control
(AEPGC) are as follows:
i) Matching of electrical power generation system to the power
consumed area load. It means that matching control of the
system frequency to the tie-line system frequency interchanged
with the time period.
ii)for security consideration [2]–[6] the changing the
distributing loads among generators so as to reduced the
operating costs subject
The first objective is met additional purposed controller in
which the concept of tie-line bias is permanently used for a
small change in the system load frequency produces directly
changes in the system frequency. The area control error is
given by
𝐴𝐶𝐸 = ∆𝑃𝑇𝐿 − 𝛽∆𝑓
In this case each area with fairly accurate knowledge of the
load changes and directs the additional controller for the area to
small changes of the turbine values of steam units as well as to

__________________________________________________________________________________________
reduce ACE to zero valves. Through ACE control, the load
demands of the grid are observing every few seconds.
The second objective is met by observing the load every few
minutes (i.e.5-10 min) and as long as the changing load system
among different units for to lowest cost of the operation based
on economic dispatch calculation (EDC). During in this
period, the load demand remains constant of EDC. Fig. 2
shows that ACE of an ISO with a sampling of ACE to an ISO
with a sampling rate of 5 s.
Fig. 2 ACE data with mean and standard deviation
When ACE occurs changes in daily load pick up/drop or loss
of INGs. Through the rotor kinetic energy stored and the
controller of governor detect the change in the ACE by change
in constant speed. The control of governor action take place
than turbine valves are opens to the power grid at a constant
speed and at a new steady-state. However, since the set of
inputs are reference point such as the incoming steam demand
with corresponds to input power demand, so that does not
change the constant speed drops, which results in a drop in the
grid-operating frequency. At the certain instant, when the grid
load is dropped, the speed and operating frequency is
increased. The too much drop in constant speed can issue the
grid to parallel failure because cooling fan system of steam
units should be slow down, due in high temperature in motor
windings of cooling fans, and this happening steam units to
trip. Hence, for stable operation, the grid frequency must be
accurately controlled. From APGC controls frequency in
response to load changes to use the opening or closing the
turbine valves However, the primary energy source as boiler
overrides the opening of turbine vales as shown in Fig. 1.
However, the primary energy source (boiler) overrides the
opening of turbine vales if the steam pressure drops below a
secure limit.
To implements the above objectives, all APGC software is
used on single base unity control system. For expression (1),
the desired generation at time instant k which is normally
observed every4-10s is given by
Where PiE(k), PiR(k),PiEA(k) are the economic, regulating,
and emergency source assist components of desired generation
for unit i respectively. k is the time step in seconds. However,
the time step of EDC is kk which is 5–15 min. The desired
generation of unit i is calculated using equation from 2-4
respectively
kk is the time step of (EDC) in minutes at which the economic
participation factors of EPFi and economic base points PiEB
are computed. The values of EPFi(kk) and PiEB used in the
calculation of PiE are those calculated by the most
advanced execution of the EDC. AFi, RFi, are the emergency
factor and regulating factor for unit i, respectively. In
expression (3), G is the regulating system gain factor, AF, RF,
and G are tuned in the field for proper unit control. In
expression(2), is the change in total unit economic desired
generation at time step (k) (in seconds)based on the
most advances execution of the EDC function at time instant
kk (in minutes). The term (k) is calculated using

__________________________________________________________________________________________
Fig.1 Current automatic generation control (A PGC). AVR: Automatic voltage regulator; Vref: AVR reference voltage; VM: measured
voltage; Exc: Excitation; Factual: generated power unit i; P desire: desired power output unit i.
Where PiAis the actual generation of unit i at time k, ACEis
the area control error at time instant k, is the economic base
point for unit i calculated by the most advances execution of
the EDC at time kk, and N is the number of units operating in
the APGC control system Nearly all of the APGC software
systems have some kind of clean for avoiding the noise part of
ACE which cannot be controlled [2]–[8]. Fig 3 Control time scale of power grid assets systems
It can be the EDC and ACE both control the task ofmatching
the area for power generation to power consumed area load,
such that the resultant of tie-line exchange and frequency of the
area are their programmed values. Both of the control functions
are handled by variation governor motors to demanding of MW
load due to the adjusting of MW generated system ( in Fig. 1).
The high selection rate of the area load provides a quite
accurate direction of the changing load demand which allows
the tie-line bias method to follow the load.

__________________________________________________________________________________________
3. STRUCTURE OF APGC FOR SMART
ECTRICAL POWER GRIDS
The steam units as operated presently, have very limited
changeable operations. The ACE and EDC control functions
are handled by controlling methods under APGC by to be
adjustment of the MW demand of the electrical grid load to
MW generated system.. If the system load is changes rapidly
i.e increasing or decreasing, or if intermittency occurs in INGs
due to the of changing wind-patterns, or cloud formation, the
situation is appreciably different. In a smart grid, ING assets
and electric vehicles, loads introduce a major obstruction in
determining satisfactory of APGC control [7]–[10]. Under this
situation, a new important set of control system must be used
and also a new set of economic factors is needed. Since INGs
are nonfarm resources, gas-fired units are used to moderate
sudden outage of INGs. Moreover, the cost of gas-fired units
is costly spinning reserves. Therefore, DCS and demand
response (DR) are necessary to control the intermittency of
ING assets. The average of DR can be used as agent of VGs of
various sizes to control the grid, when ACE is very high (see
Fig. 3). In order to assurance of the minimal carbon footprints
and low cost, all providing ING assets can be used as base-load
system. All nuclear units and hydal units must be used for
base-load system because they cannot contribute in regulation
due to protection purposed. In the proposed control structure
(see Fig. 4), VPS, , and CPS units can be used for low-
frequency load fluctuations based on a one-step-ahead load
target. A load aim with time horizon of 5–15 min can facilitate
control of these systems to follow an assigned targeted load.
Note that hydro units, depending on dynamic response, may be
used in both low- and high-frequency load controls. The
combined use of high- frequency load control. The combined
use of hydro, DCS, and DR/VG assets can hold a high level of
diffusion of ING assets. The proposed APGC control (see Fig.
4) can follow the low frequency dynamic load trends using a
one-step-ahead load for CPS units and optimally for VPS units.
This in turn can be lead to a reduction of ACE at the sampling
rate of the economic dispatch. The ACE reduction can reduce
the control actions by the APGC controller, which in turn
eliminates the unnecessary pulsing of the steam regulating
units. The high value of ACE is due to high frequency load
fluctuations or sudden decrease in wind speed (generation).The
control of the high value of ACE is achieve by using DCS and
DR/VG grid assets (see Fig. 3).
4. CYBER-ARCHITECTURE SMART
ELECTRICAL GRIDS
Fig:.5 shows that, the proposed cyber-architecture to facilitate
the goals of a smart grid with large numbers of ING system in
the form of PV, wind, tidal, and micro grids in distribution
systems. To confutes VGs of different sizes for an ISO, the
proposed cyber-architecture should be developed for a smart
electrical grid that consists of large number of nodes, with
sensors for measuring the grid states (bus bar voltages) and
phase or angle measurement units (PMU)[10]
Optimal resource distribution decisions are to be made based
on output signals and inputs signals from the sensor nodes.
Hence. the smart electrical power grid control structure, fully
depends on the availability of a reliable moving forgathering
sensor information and disseminating control information. Fig.
5 shows, a server fusion node (SFN) that can be used to
observe and control at various locations in the grid for
collection data and aggregation, pre filtering, and post filtering.
From the smart meters are the points of communication with
end energy user/provider (E/P), where residential, industrial,
and commercial consumers are located. At sub transmission
level, larger energy providers, such as independent power
producers and large serious power sites, are located. At the
transmission level, tie-line power (TPL) to other related power
grids is metered and data are sent to the ISO control center. The
VG units can be construing from a big number DRs in each
zone through smart meters. DRs most be aggregate to a sum
load in a few demands under the control of command for
APGC software. Then, VGs can be controlling through the
APGC when ACE exceeds its limited value. The design of a
virtual-architecture faces a large number of challenges, as well
as the possibility of incomplete information flow due to
message losses and inhibited computation resources. If wireless
networks, a highly popular choice of message system are used,
the messages are transmitted in a multi hop fashion from the
source to the intention via a number of fusion nodes, where
resource allocate decisions are made. While there have been a
great deal of work, taking place the design of control
algorithms used for wireless networks, the smart grid
applications is require several special cases, such as the need to
collected information from a large number of sensor nodes
moreover in real-time, stringent delay and reliability
requirements data. for APGC. Moreover, the delay in
aggregating a number of DRs may be controlled in distribution
systems for low loss of data. The virtual structure must have a
tree-like structure of a power grid, with high density in some
places such as the distribution systems; this is the backbone of
a network for APGC is likely to be comprised of a fiber-optic
transmission-line access toward sensors nodes at detached
locations. Moreover, a wireless network can be used in
residential distributions.

__________________________________________________________________________________________
Fig 4 The proposed APGC control structure for electrical mart power grid.
5. DECISION AND CONTROL
5.1 One-Step ahead Load Model System
The discussion taken for above case, it can be obvious, from
the outset the control plan should identify the load changes due
to high-frequency fluctuations, which are within low-frequency
fluctuations and dynamic response of APGC resources.
APGC must adopted decision and control plans to managed
with the low and high frequency load fluctuations, through help
of modeling the load dynamic responses using the one-step-
ahead load model system, both high- and low-frequency load
fluctuations can be controlled by carefully selection of APGC
resources The idea behind this methodology is based on system
identification [11]using a sample of observations Y(k) and
identifying the basic process of model system and its
parameters. The load monitoring(observing)sequence
corresponds to stochastic processes [12] and is produced by an
autoregressive moving average (ARMA) model given below

__________________________________________________________________________________________
This model can be expressed in compact form as (7)
The residuals error W (k+1) can be used to estimated and
constructed the ARMA model, if the past residual error has
steady dynamics. The one-step-ahead predictor model and error
are given as
Where W (k) is an immeasurable disturbance and is white with
zero mean, when a suitable ARMA model is identified for the
process. The one-step-ahead predictor of the load sequence is
given below in Fig. 6.
The parameter vector is identified from observation as
[13]. I submit the proposed method can be used to model
the load sequence base on the last large samples of
observations rates. Anther, one-step-ahead load model can be
the after a few observations, imprison the load dynamic
responses dazzling and identifying both high and low-
frequency fluctuation loads.
The ACE fluctuation is a feedback signal that identifies the of
ING generators. Note that during any sudden loss occur of
wind, load controlled by the energy consumers remains the
same. Whenever, the unbalance between generation and load is
measured trough high value of ACE.
The ACE value can be used as a feedback signal to select
suitable system to control the system frequency as described
below.
Fig 6 The one-step-ahead load model predictor [13]

__________________________________________________________________________________________
Fig 5 The proposed cyber architecture
5.2 VGAND.DCS Units Control System
The final decision taken place, DCS and VG system is made
on the basis of ACE. If ACE is extremely negative, indicating
the deficiency in energy balance (a drop in frequency), the
APGC controller identifies, this state as an emergency state
(see Fig. 4.).The VG system can be used for an instant drop of
load and maintained the energy balance, thus, restored the
frequency to a normal state from the ACE method, presented in
Fig. 2 , that the need for VG system can be used for a very
short time, in the order of few seconds to a few minutes, unless
the grid remains in the emergency state. The alert state occurs
when ACE has a to determined dynamic trend toward the
emergency state. For the alert state, DCS system should be
used to maintain the energy balance in a few seconds. The
response of a DCS-inverter is shown in Fig. 7. I have studied
the dynamic response of the DCS-inverter[8].
5.3 Control of CPS Units
The on-line system one-step load model can used to assigned a
maximum load to the units under CPS control. The CPS units
comes under the boiler control system and have been multi face
ted thermal processes. The models of CPS units for APGC
system do not have been developed because CPS units has a
limited regulating capabilities, the APGC software controls ,the
CPS units from the turbine valves. Valve point of CPS units
reduces the throttle loss and wear of the valves. The one-step-
ahead ,load can be used for valve point for loading of the CPS
units. Control of this units for APGC documented in [4].

__________________________________________________________________________________________
Fig 7P.Q , regulation of a laboratory microgrid [8].
5.4 VPS Control Units
The VPS units can be accurately controlled by APGC by using
dynamic response of VPS units. In [14], the gross drum model
type boiler was developed. In this paper, this model is used to
for accurately control of VPS units, to maintained
demonstrated of model. The control problem is calculated by
using the optimal control theory by Bryson and Ho [15] and
implemented in Matlab. The discrete state-space model of the
VPS model is given in (10).
Where the variable X is the drum pressure and U is the input
control of fuel flow, the control valve setting, and the feed
water, respectively.
The vector difference (10) indicates the dynamic performance
of the VPS unit, where X(k+1) belongs ton dimensional space
and is the state variable at time k; U(k) belongs to m
dimensional space and is the input at time k; A is n xn ; and B
is n x m matrix. The MW output of the VPS unit can be
expressed as a function of state variables, and the input
variables at any time instant k. That is,
Where P(t), is the MW output at time instant k.
Where T is the period of control and is the subinterval of
control represented by k. The energy delivered to the grid is at
the same time as presented by [15].
The control problem is formulated as a linear separate control
problem for small load instability and the constraint are as
follows:
a) Final time of control is fixed
b) Initial and terminal boundary values are given
c) The control variables and the state variables must satisfy a
magnitude constraint
d) The rate of change of the state variables must satisfy a
magnitude constraint
e) The rate of change of the state variables must be zero atk=N:
at the end of control cycle, i.e., the system must reach their a
steady state at the final time
X (N-1)--X (N) =0.
The problem is to find the optimizing control sequence
U=(U0,U1….U N-1)and the corresponding route
X=(X0,X1……X(N))and subject to conditions a) through e),
While minimizing the performance index given by (14):
Where d is a weighing vector on control actions. The
accurately dynamic generation response control is proved,
when the VPS units are controlled by the above constraints.
Equation (13) indicates, power consumed by the load and
power supplied to the system must be equal. Therefore, the
power balance enables APGC to control the system frequency
and the planned exchange of power with the nearest grids. The
control constraints require that the VPS unit to reached a
constant state at the final control stage, this will be happened in
about 2-3times of the constant of the VPS unit. This is a very
important if VPS unit is to be steered again to another target
MW. The rest of the situation and constraints are used by
practical considerations. For example, the rate of change of the

__________________________________________________________________________________________
state variables is constrained, which in turn limits the rate at
which the load can be picked up or dropped. The boiler-turbine
model developed by Astrom and Eklund [14] with a maximum
output of 160MWis used to demonstrate the control
method.Fig.8
Fig 8 Optimal dynamic generation control of a VPS
Fig:8shows eight periods of control cycles (i.e one Hour) the
variable pressure operations with turbine valves,U2 set at
95%. Also shown are the corresponding fuel flow U1 and feed
water flow U2, according condition (police). The proposed
variable pressure operation is based on a one-step-ahead, load
to reached maximum valves for a VPS unit. Therefore, for this
type operation, an online load modeling method with observing
intervals based on dynamic VPS units must be developed.
Since the important boiler dynamics is used in the control
problem formulated, the temperature changes within the
turbine when the throttling losses must be deleted. Than
automatically reduces the thermal weakness and increases the
efficiency of operation.
6. CONCLUSIONS
The proposed work, APGC control structure can contained
ING penetrations by identified the load fluctuations is easily
measured by ACE method, high and low frequency system.
The levels of INGs in the smart grid must corresponding to
DCS and VG systems under APGC control system in the order
to as per as possible to minimum of the costly installation of
gas-fired units for spinning reserves. The CPS units can also be
controlled to follow of steam or objective load with a longer
control of time cycle to keep the steam pressure within a not
dangerous limit and realize valve-point loadings. The given
proposed method of control minimum both the steady state and
dynamic cost of units. During the normal operation conditions
of the electrical power grids, the VPS and CPS units can be
used for controlled as changeable units for low frequency load
fluctuations with hydro, wind, DCS, and DR/VG as
complimentary agents for corresponding for high-frequency
load fluctuations. During the alert state, the EDC is suspended
and DCS system are dedicated to regulation. In the case of
emergency state, DR/VG systems are dedicated to stabilizing
the grid and controlling the system frequency. As penetration
of INGs increase, the DCS and VG units must be installed. The
proposed control structure mitigates the INGs’ control
problems in the smart electrical power grids. However, the
development of APGC simulation based, on this paper needs
further research. The research problems are as follows: i) A
cyber-structure power grid system that can observe the state of
power grid; ii) Online unpleasant dynamic models of boilers.
iii) A smart power grid APGC simulator for operator training
system.
REFERENCES
[1]. J. J. Conti et al., “Annual energy outlook 2011 with
projections to 2035,” U.S. Energy Information Administration,
2011.
[2]. N. Jaleeli, L. S. VanSlyck, M. M. Yao, R. R. Shoults, and
R. Kelm,“Discussion of AGC logic based on NERC’s new
control performance standard and disturbance control
standard,” IEEE Trans. Power Syst., vol. 15, no. 4, pp. 1455–
1456, Nov. 2000.
[3]. A. J. Wood and B. F. Wollenberg, Power Generation
Operation and Control, 2nd ed. New York: Wiley, 1996.
[4]. P. Kundur, Power System Stability and Control. Palo Alto,
CA: Mc-Graw-Hill, 1994.
[5]. V. Donde, M. A. Pai, and I. A. Hiskens, “Simulation and
optimization in an AGC system after deregulation,” IEEE
Trans. Power Syst., vol. 16, no. 3, pp. 481–489, Aug. 2001.
[6]. J. Kumar, K. Ng, and G. Shelbe, “AGC simulator for price-
based operation part 1: a model,” IEEE Trans. Power Syst.,
vol. 12, no. 2, pp. 527–532, May 1997.
[7]. A. Keyhani, M. N. Marwali, and M. Dai, Integration of
Green and Renewable Energy in Electric Power Systems.
Hoboken, NJ: Wiley, 2010.
[8]. A. Keyhani, Design of Smart Grid Renewable Energy
Systems. Hoboken, NJ: Wiley, 2011.
[9]. A. Nourai and C. Schafer, “Changing the electricity game,”
IEEE Power Energy Mag., vol. 2, no. 4, pp. 42–47, 2009.
[10]. K. Moslehi and R. Kumar, “Vision of a self-healing
power grid,” ABB Rev., pp. 21–25, Apr. 2006.

__________________________________________________________________________________________
[11]. L. Ljung, “Perspectives on system identification,” in
Annual Reviews in Control. : Elsevier, 2010.
[12]. A. Keyhani, “Stochastic load modeling and optimal
dynamic generation control,” Ph.D. dissertation, School of
Electrical Engineering, Purdue Univ., West Lafayette, IN,
1975.
[13]. A. Keyhani and A. H. El-Abiad, “One step-ahead load
forecasting for on-line applications,” presented at the
IEEEWinter Power Meet., New York, 1975, Paper CP 221-
PWR.
[14]. K. J. Astrom and K. Eklund, “A simplified non-linear
model of a drum boiler-turbine unit,” Int. J. Control, vol. 16,
no. 1, 1972.
[15]. A. E. Bryson and Y. C. Ho, Applied Optimal Control:
Optimization, Estimation, and Control. Waltham, MA:
Blaisdell, 1969, p. 481

Automatic power generation control structure for smart electrical power grids

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Automatic power generation control structure for smart electrical power grids (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Automatic power generation control structure for smart electrical power grids