BATTERY SIZING

example
The following example illustrates the use of this formula to size the
battery. Suppose we want to design a battery for a stand-alone power system,
which
charges and discharges the battery from a 110-V DC solar array.
For the DCDC buck converter that charges the battery, the maximum
available battery-side voltage is 70 V for it to work efficiently in the pulsewidth-modulated (PWM) mode.
Forthe DCDC boost converter discharging the battery, the minimum required
battery
voltage is 45 V.
Assuming that we are using NiMH electrochemistry, the cell voltage
can vary from 1.55 V when fully charged to 1.1 V when fully drained to the
maximum
allowable DoD.
Then, the number of cells needed in the battery is less than 45
(70/1.55) and more than 41 (45/1.1). Thus, the number of cells in the battery,
estimated from voltage considerations, must be between 41 and 45. It is
generally
more economical to use fewer cells of a higher capacity than to use more
lowercapacity

PV array sizing
The basic tenet in sizing a stand-alone power system is to
remember that it is
really a stand-alone energy system. It must, therefore, maintain
the energy balance
over the specified period.
A simple case of a constant load on a PV system using solar arrays
perfectly pointed toward the sun for 10 h of the day is shown in
Figure.

The solar array is sized such that the two shaded areas
on the two sides of the load line must be equal. That is, the area oagd
must be equal
to the area gefb.
The system losses in the round-trip energy transfers, e.g., from and
to the battery, adjust the available load to a lower value.
In general, the stand-alone system must be sized so as to satisfy the
following
energy balance equation over one period of repetition:

WIND FARM SIZING

In a stand-alone wind farm, selecting the number of towers and
the battery size depends on the load power availability
requirement.
A probabilistic model can determine the number of towers and
the size of the battery storage required for meeting the load
with a required certainty.
Such a model can also be used to determine the energy to be
purchased from, or injected into, the grid if the wind power plant
were connected to the grid.
In the probabilistic model, wind speed is taken as a random
variable. The load is treated as an independent variable.
The number of wind turbines and the number of batteries are
also variables in the analysis.

Each turbine in a wind farm may or may not have the same
rated capacity and the same outage rate.

In any case, the hardware failure rates in individual turbines

are independent of each other.

The resulting model has a joint distribution of the available

wind power (wind speed variations) and the operating mode
(each turbine working or not working).
The events of these two distributions are independent.

For a given load duration curve over a period of repetition, the

expected energy not supplied to the load by the hybrid system
clearly depends on the size of the battery, as shown in Figure .

The larger the battery, the higher the horizontal line, thus
decreasing the duration of the load not supplied by the system.

With such a probabilistic model, the time for which the load is not
supplied by
the system is termed as the Expected Energy Not Supplied (EENS).
This is given by the shaded area on the left-hand side. The Energy
Index of Reliability (EIR) is then given by the following

GRID CONNECTED SYSTEM

Wind and photovoltaic (PV) power systems have made a

successful transition from small stand-alone sites to large gridconnected systems.

This improves the overall economy and load availability of the
renewable plant site the two important factors of any power
system.
The grid supplies power to the site loads when needed or
absorbs
the excess power from the site when available.
A kilo-watt hour meter is used to measure the power delivered
to the grid, and another is used to measure the power drawn
from the grid.

INTERFACE REQUIREMENTS
Both the wind and PV systems interface the grid at the output
terminals of a
synchronizing breaker after the inverter. The power flows in either
direction depending on the site voltage at the breaker terminals. The
fundamental requirements on the site voltage for interfacing with the
grid
arevoltage
as follows:
The
magnitude and phase must equal that required for

the desired
magnitude and direction of the power flow.
The voltage is controlled by the transformer turn ratio or the
power electronic converter firing angle in a closed-loop control
system.
The frequency must be exactly equal to that of the grid or else
the system
will not work.
In the wind system, the base-load synchronous generators in
the grid

The interface and control issues are similar in many

ways in the PV and wind systems.
The wind system, however, is more involved. The
electrical generator and turbine, due to their large
inertia, introduce certain dynamic issues not
applicable in the static PV system.
Moreover, wind plants generally have much greater
power capacity than the PV plants.
For example, many wind plants that have been
already installed around the world have capacities
of hundreds of megawatts each.
More wind plants exceeding 300-MW capacity are
being installed, and even larger plants are being
planned.

SYNCHRONIZING WITH THE GRID

The synchronizing breakers have internal voltage and phase-angle sensors to
monitor the site and grid voltages and signal the correct instant for closing the
breaker.
As a part of the automatic protection circuit, any attempt to close the breaker at an
incorrect instant is rejected by the breaker.
Four conditions that must be satisfied before the synchronizing switch permits the
closure are as follows:

1. The frequency must be as close as possible to the grid

frequency, preferably
about one third of a hertz higher.
2. The terminal-voltage magnitude must match with that of the
grid, preferably
a few percent higher.
3. The phase sequences of both three-phase voltages must be
the same
4. The phase angle between the two voltages must be within 5.

INRUSH CURRENT
A small unavoidable difference between the site and grid voltages
results in an inrush
current flowing between the site and the grid. The inrush current
eventually decays
The
initial
of this
current
at the instant
the circuit
to zero
at magnitude
an exponential
rate
that depends
on theofinternal
resistance
breaker
closing
and inductance.
depends on the degree of mismatch between the two voltages.

The synchronizing power produced by the

inrush current brings the wind system and the
grid in synchronism after the oscillations decay
out.
Once synchronized, the wind generator has a
natural tendency to remain in synchronism with
the grid, although it can fall out of synchronous
operation if excessively loaded or a large load
step is applied or during a system fault.
Small load steps induce swings in the load
angle decay out over time, restoring the
synchronous condition.
The magnitude of the restoring power, also
known as the synchronizing power, is high if the

Synchronous operation

Once synchronized, the voltage and frequency of the wind

system need to be controlled.
The grid serves as the frequency reference for the
generator output frequency when the induction generator is
directly connected to the grid.
The grid also acts as the excitation source supplying the
reactive power. Because the torque vs. speed characteristic
of the induction generator has a steep slope near zero slip,
the speed of the wind turbine remains approximately
constant within a few percentages.
A higher load torque is met by a proportionately increased
slip, up to a certain point beyond which the generator
becomes unstable.
If the load torque is immediately reduced, the generator
returns to the stable operation.
From the operating point of view, the induction generator is
softer, as opposed to a relatively stiff operation of the
synchronous generator, which works at an exact constant
speed or falls out of stability.

 Once synchronized and connected with the lines, the

synchronous generator has an inherent tendency to remain in
synchronous lock with the grid.
The synchronism can be lost only during severe transients and
system faults. The generator must be resynchronized after such
an event.
In the variable-speed induction generator system using the
inverter at the interface, the inverter gate signal is derived from
the grid voltage to assure synchronism.
The inverter stability depends a great deal on the design. For
example, there is no stability limit with a line-commutated
inverter.
The power limit in this case is the steady-state load limit of the
inverter with any short-term overload limit.

Load transient

The grid will pick up the area load during steady-state

operation if the renewable power system output is fully or
partially lost. The effect of this is felt in two ways:

The grid generators slow down slightly to increase their

power angle under the increased load.
This results in a momentary drop in frequency until the
governors allow more fuel.
The grid conductors now carry more load, resulting in a
small voltage
drop throughout the system.
The same effects are felt if a large load is suddenly
switched on. Because starting the wind turbine as the
induction motor draws a large current, it also results in the
aforementioned effect.
Such load transients are minimized by soft-starting large
machines.
In wind farms consisting of many generators, individual

Safety
Safety is a concern when renewable power is
connected to the utility grid lines.
The interconnection may endanger the utility repair
crew working on the lines by continuing to feed
power into the grid even when the grid itself is down.
This issue has been addressed by including an
internal circuit that takes the inverter off-line
immediately if the system detects a grid outage.

Because this circuit is critical for human safety, it

has a built-in redundancy.

 The grid interface breaker can get suddenly

disconnected either accidentally or to meet an
emergency situation.
The high-wind-speed cutout is a usual condition
when the power is cut off to protect the generator
from overloading. In a system with large
capacitors connected at the wind site for power
factor improvement,
the site generator would still be in the selfexcitation mode,
drawing excitation power from the capacitors and
generating terminal voltage. In the absence of
such capacitors,
one would assume that the voltage at the
generator terminals would come down to zero.
The line capacitance, however, can keep the

When the grid is disconnected for any reason, the

generator will experience a loss of frequency regulation as
the frequency-synchronizing signal derived from the grid
lines is now lost.
When a change in frequency is detected beyond a certain
limit, the automatic control can shut down the system,
cutting off all possible sources.
of excitation.

Operating limit

The link connecting a renewable power

site with the area grid introduces an
operating limit in two ways, the voltage
regulation and the stability limit.
In most cases, the link can be considered
as an electrically short transmission line.
The ground capacitance and the ground
leakage resistance are generally
negligible.
There are two major effects of the
transmission line impedance, one on the
voltage regulation and the other on the
maximum power transfer capability of the

Voltage regulation

STABILITY LIMIT

The direction of power flow depends on the

sending- and receiving-end voltages and the
electrical phase angle between the two.
However, the maximum power the line can
transfer while maintaining a stable operation
has a certain limit.

We derive in the following text the stability limit,

assuming that the power flows from the renewable site
to the grid, although the same limit applies in the
reverse direction as well. The series resistance in most
lines is negligible and hence is ignored here

ENERGY STORAGE AND LOAD SCHEDULING

 Even with a grid connection, large wind

and PV plants may find it economical to
store some energy locally in a battery.
The short-term peak demand is met by
the battery without drawing from the
grid and paying the demand charge.
For formulating an operating strategy
for scheduling and optimizing the use of
grid power, major system constraints
are first identified.

 The usual constraints are the battery size,

the minimum on/off times and ramp rates for
thermal power plants,
the battery charge and discharge rates,
and the renewable power capacity limits
For arriving at the best short-term and longterm scheduling,
the optimization problem is formulated to
minimize the cost of all thermal and
renewable plants combined subject to the
constraints.

 Such an optimization process determines the hours for

which the base load thermal units of the utility company
should be taken either off-line or online.
The traditional thermal scheduling algorithms, augmented
by Lagrangian relaxation,
branch and bound,
successive dynamic programming,
heuristic method (genetic algorithms and neural
networks), can be used for minimizing the cost of operating

In a case study of a 300-MW thermalPVbattery power

plant, the authors have arrived at the total production
costs shown in Table 13.1 that a battery hybrid system
saves $54,000 per day compared to the thermal power
plants alone.

UTILITY RESOURCE PLANNING TOOLS

Wind and PV power,

in spite of their environmental,
financial, and fuel diversity benefits,
have been slow to enter utility resource planning
because of the planners lack of familiarity with them,
and analytical tools for non dispatchable sources of
such power are not available on demand.