CHAPTER 3

EMBEDDED GENERATION AND EVALUATION TECHNIQUE

3.1. Steps for Reliability Evaluation

Firstly, choose the distribution network. And data is collected in the
distribution network. In this data, system operating and outage data and customer
outage data are included. Reliability indices can be calculated by using this data for
three cases. Thus, result can be seen. In this result, interruption is high. For reduce
interruption, distributed generator (DG) is installed in distribution network. Suitable
DG must be chosen and re-calculated, results can be seen. These results can be seen
differently.
Figure 3.1. Steps for Reliability Evaluation

System operating
and outage data

Compute Compute reliability

reliability Choose DG indices by adding
Data indices with DG
size and
collect three cases location
 

Customer
outage
Results of Results of
Calculations Calculations

Reliability improvement
 11

3.2. Basic Consideration

For the reliability analysis, need to know three typical criteria as described in
below.
 Planning generating capacity-installed capacity= the expected maximum
demand+ a fixed percentage of the expected maximum demand.
 Operating capacity – spinning capacity = expected load demand + a reverse
equal to one or more largest units.
 Planning network capacity – construct a minimum number of circuits to a load
group (generally known as an (n- 1) or (n-2) criterion depending on the
amount of redundancy), the minimum number being dependent on the
maximum demand of the group. [91Ron]

3.2.1. Basic steps for Reliability Evaluating

It is possible to calculate reliability indices for a particular set of system data
and conditions. These indices can be viewed as either absolute or as relative measures
of system reliability.
Absolute indices are the values that a system is expected to exhibit. It can be
monitored in terms of past performance because full knowledge of it is known.
However, they are extremely difficult, if not impossible, to predict for the future with
a very high degree of confidence. The reason for this is that future performance
contains considerable uncertainties particularly associated with numerical data and
predicted system requirements.
Relative reliability indices, on the other hand, are easier to interpret and
considerable confidence can generally be placed in them. In these cases, system
behavior is evaluated before and after the consideration of a design or operating
change. The benefit of the change is obtained by evaluating the relative improvement.
Indices are therefore compared with each other and not against specified targets. This
tends to ensure that uncertainties in data and system requirements are embedded in all
the indices and therefore reasonable confidence can be placed in the relative
differences.
The most important aspect to remember when evaluating these measures is
that it is necessary to have a complete understanding of the engineering implications
of the system. No amount of probability theory can circumvent this important
 12

engineering function. It is evident therefore that probability theory is only a tool that
enables an engineer to transform knowledge of the system into a prediction of its
likely future behavior. Only after this understanding has been achieved can a model
be derived and the most appropriate evaluation technique chosen. Both the model and
the technique must reflect and respond to the way the system operates and fails.
Therefore the basic steps involved are:
-understand the ways in which components and system operate;
-identify the ways in which failures can occur;
-deduce the consequences of the failures;
-derived models to represent these characteristics;
-only then select the evaluation technique. [96Bro]

3.3. Evaluation Technique

There are two set of indices, namely the basic load point indices and the
system performance indices for distribution systems. The basis indices are important
with respect to a particular load point.

λ
Unit up μ Unit down

Figure 3.2. Two-state model for a base load unit

The basic component model used in these applications is the two-state
representation as shown in figure. The rate of departure from the component up state
to its down state is the component failure rate. The restoration of the component to its
opening state is denoted the component repair rate. [96Bro]

3.4. Reliability Improvement by Protection System

Adding protective device is one of the most straightforward and effective
methods for improving distribution system reliability. Assuming proper coordination,
increasing the number of customers that experience interruptions after a fault occurs.
States differently, increasing the number of protective devices increases the selectivity
of the production system.
 13

The first step towards improving reliability is to place to protective devices,

typically a fuse, on all radial branches. Both field experience and reliability studies
show conclusively that laterals should be fused. The only compelling reasons not to
fuse a lateral are nuisance fuse blowing (which can generally be avoided by
specifying larger fuses) and the inability to coordinate. Three-phase laterals may
require devices with three phase lockout capability if the serve large motors which
may be damages by unbalanced voltages, or transformers with primary delta-
connected windings, which may create safety problems due to the possibility of back
feeding.
The effectiveness of lateral fusing increases as total lateral exposure increases
and as average lateral length decreases. Assuming perfect fuse operation, a fault on an
un-fused lateral will interrupted the entire feeder while a fault on a fused lateral will
only interrupted customers on the lateral.
The analysis of reliability improvement will be conducted for the following
cases.
 Effect of lateral protection
 Effect of disconnection switches
Using protective device on radial system has a great effect on reliability of the
system. In this thesis the protection failure is not considered and only permanent fault
is taken into account. [10Roy]

3.5. Reliability Improvement by Distributed Generation

There has been a considerable revival in interest in connecting generation to
the distribution network and this has come to be known as embedded or dispersed
generation. The term “ embedded generation ” is used to distinguish it from central
generation. The two terms can be considered to be synonymous and interchangeable.
No universally agreed definition of what constitutes embedded or dispersed
generation and how it differs from conventional or central generation. Some common
attributes of embedded or dispersed generation may be listed as :
 Not centrally planned (by the utility)
 Not centrally dispatched
 Normally smaller than 50-100 MW
 Usually connected to the distribution system.
 14

The distribution system is taken to be those networks to which customers are

connected directly and which are typically of voltage from 230/400 V up to 145 kV.
The CIRED survey asked representatives from 17 countries what were the policy
drivers encouraging embedded generation. The answers included:
 Reduction in gaseous emission (mainly CO2)
 Energy efficiency or rational use of energy
 Deregulation or competition policy
 Diversification of energy sources
 National power requirement.
Environment impact is a major factor in the consideration of any electrical
power scheme, and there is a generally accepted concern over gaseous emissions from
fossil fueled plant. As part of the Kyoto Protocol, both the EU and the UK have to
reduce substantially emissions of CO2 to help counter climate change most
governments have programs to support the exploitation of the so – called new
renewable energy resources, which include wind power, micro-hydro, solar
photovoltaics, landfill gas, energy from municipal waste and biomass generation.
Renewable energy sources have a much lower energy density than fossil fuels and so
the generation plants are smaller and geographically widely spread. For example,
wind farms must be located windy areas, while biomass plants are usually of modest
capacity due to the cost of transporting fuel with relatively low energy density. These
small plants, typically of less than 50MW in capacity, are then connected into the
distribution system. In many countries the new renewable generation plants are not
planned by the utility but are developed by entrepreneurs and are not centrally
dispatched but generate whenever the energy sources is available. There are two main
consideration concerning DG in distribution network.
 Embedded generation cannot be run independently of the BSP:
In this case, the situation when the embedded generation cannot be allowed to
continue to operate following a disruption of supply from the BSP: this may occur due
to a failure of the BSP, of the transmission system leading to the BSP, or of the
distribution feeder between the BSP and the load center being assessed.
 Embedded generation can be run independently of the BSP:
 15

If the embedded generation can be used independently of the main supply

from the BSP, then this source can continue to supply all or some of the load when a
failure occurs between the BSP and the load center being assessed. This will mean an
increase in the system reliability. Three case studies are included to illustrate this
aspect:
Case 1: Embedded generation is always available and has a capability greater
than the maximum demand.
Case 2: Embedded generation is always available but had a partial capability
which is less than the maximum demand.
Case 3: Embedded generation is available only for 25% of the same and has a
partial capability. [10Kan]

3.5.1. Assumption for DG Placement

The following assumptions are made for DG placement.
 DG is considered as negative load.
 DG injects only active power.
 The maximum DG size is assumed to be total load demand plus total active
 Power losses of the system.
 For DG placement, the source node is not to be taken into account [15Sah]

Table 3.1 Design Rules for Installing DG to network [15Sah]

Network location Maximum capacity of

generator

out on 400 V network 50 kVA

at 400 V busbars 200-250 kVA
out on 11 kV or 11.5 kV network 2-3 MVA
at 11 kV or 11.5 kV busbars 8 MVA
on 15 kV or 20 kV network and at busbars 6.5-10 MVA
on 63 kV to 90 kV network 10-40 MVA

3.5.2. Selection of Location

 16

In order to reduce the efforts to select appropriate nodes for the placement of
DG devices a priori, the sensitiveness of nodes toward the change in active power loss
with respect to change in active power injection at various nodes are identified. The
loss sensitivity factors at different buses have been evaluated using a load-flow
program to select appropriate nodes for DG planning. These sensitivity factors reflect
how the feeder power losses change if more real power is injected at a particular node
and also allow the obtained candidate nodes to locate DG. Loss sensitivity factors are
initially evaluated for the base case (initial condition). In order to decide the next
location(s), the computation of loss sensitivity has been repeated, taking into account
the previously allocated DG, which is termed successive sensitivity analysis. This
process is repeated until the optimum number of locations is ascertained. The optimal
number of locations is that number of buses that give maximum benefits (savings) for
optimal size DGs placed at selected locations. [97Sch]

3.5.3. Increase In Network Fault Levels

Most embedded generation plant uses rotating machines and these will
contribute to the network fault levels. Both induction and synchronous generators will
increase the fault level of the distribution system although their behaviour under
sustained fault conditions differs. In urban areas where the existing fault level
approaches the rating of the switchgear, the increase in fault level can be a serious
impediment to the development of embedded generation. Uprating of distribution
network switchgear can be extremely expensive and, under the charging policies
currently used in the UK, this cost will be borne by the embedded generator. The fault
level contribution of an embedded generator may be reduced by introducing an
impedance between the generator of the network by a transformer or a reactor but at
the expense of increased losses and wider voltage variations at the generator. In some
countries explosive fuse type fault current limiters are used to limit the fault level
contribution of embedded generation plant. [97Sch]

3.5.4. Power Quality

Two aspects of power quality are usually considered to be important: (i)
transient voltage variations and (ii) harmonic distortion of the network voltage.
Depending on the particular circumstance, embedded generation plant can either
 17

decrease or increase the quality of the voltage received by other users of the
distribution network. Embedded generation plant can cause transient voltage
variations on the network if relatively large current changes during connection and
disconnection of the generator are allowed. The magnitude of the current transients
can, to a large extent, be limited by careful design of the embedded generation plant,
although for single generators connected to weak systems the transient voltage
variations caused may be the limitation on their use rather than steady-state voltage
rise. Synchronous generators can be connected to the network with negligible
disturbance if synchronized correctly, and antiparallel soft-start units can be used to
limit the magnetizing inrush of induction generators to less than rated current.
However, disconnection of the generators when operating at full load may lead to
significant, if infrequent, voltage drops. Also, some forms of prime-mover (e.g. fixed
speed wind turbines) may cause cyclic variations in the generator output current
which can lead to so-called ‘flicker’ nuisance if not adequately controlled [9,10].
Conversely, however, the addition of embedded generation plant acts to raise the
distribution network fault level. Once the generation is connected any disturbances
caused by other customers’ loads, or even remote faults, will result in smaller voltage
variations and hence improved power quality. Similarly, incorrectly designed or
specified embedded generation plants, with power electronic interfaces to the
network, may inject harmonic currents which can lead to unacceptable network
voltage distortion. However, directly connected generators can also lower the
harmonic impedance of the distribution network and so reduce the network harmonic
voltage at the expense of increased harmonic currents in the generation plant and
possible problems due to harmonic resonances. This is of particular importance if
power factor correction capacitors are used to compensate induction generators. A
rather similar effect is shown in the balancing of the voltages of rural MV systems by
induction generators. The voltages of rural MV networks are frequently unbalanced
due to the connection of single phase loads. An induction generator has a very low
impedance to unbalanced voltages and will tend to draw large unbalanced currents
and hence balance the network voltages at the expense of increased currents in the
generator and consequent heating. [96Dug]

3.5.5. Protection
 18

A number of different aspects of embedded generator protection can be

identified:
• Protection of the generation equipment from internal faults
• Protection of the faulted distribution network from fault currents supplied by the
embedded generator
• Anti-islanding or loss-of-mains protection
• Impact of embedded generation on existing distribution system protection
Protecting the embedded generator from internal faults is usually fairly
straightforward. Fault current flowing from the distribution network is used to detect
the fault, and techniques used to protect any large motor are generally adequate. In
rural areas, a common problem is ensuring that there will be adequate fault current
from the network to ensure rapid operation of the relays or fuses. Protection of the
faulted distribution network from fault current from the embedded generators is often
more difficult. Induction generators cannot supply sustained fault current to a three-
phase close-up fault and their sustained contribution to asymmetrical faults is limited.
Small synchronous generators require sophisticated exciters and field forcing circuits
if they are to provide sustained fault current significantly above their full load current.
Thus, for some installations it is necessary to rely on the distribution protection to
clear any distribution circuit fault and hence isolate the embedded generation plant
which is then tripped on over/under-voltage, over/under frequency protection or loss-
of-mains protection. This technique of sequential tripping is unusual but necessary
given the inability of some embedded generators to provide adequate fault current for
more conventional protection schemes. Loss-of-mains protection is a particular issue
in a number of countries, particularly where auto-reclose is used on the distribution
circuits. For a variety of reasons, both technical and administrative, the prolonged
operation of a power island fed from the embedded generator but not connected to the
main distribution network is generally considered to be unacceptable. Thus a relay is
required which will detect when the embedded generator, and perhaps a surrounding
part of the network, has become islanded and will then trip the generator. This relay
must work within the dead-time of any auto-reclose scheme if out-of phase
reconnection is to be avoided. Although a number of techniques are used, including
rate-of-change-of-frequency (ROCOF) and voltage vector shift, these are prone to
nuisance tripping if set sensitively to detect islanding rapidly. The neutral grounding
 19

of the generator is a related issue because in a number of countries it is considered

unacceptable to operate an ungrounded system and so care is required as to where a
neutral connection is obtained and grounded. [97Sch]
3.5.6. Stability
For embedded generation schemes, whose object is to generate kWh from new
renewable energy sources, considerations of generator transient stability tend not to be
of great significance. If a fault occurs somewhere on the distribution network to
depress the network voltage and the embedded generator trips, then all that is lost is a
short period of generation. The embedded generator will tend to over speed and trip
on its internal protection. The control scheme of the embedded generator will then
wait for the network conditions to be restored and restart automatically. Of course if
the generation scheme is intended mainly as a provider of steam for a critical process
then more care is required to try to ensure that the generator does not trip for remote
network faults. However, as the inertia of embedded generation plant is often low and
the tripping time of distribution protection long, it may not be possible to ensure
stability for all faults on the distribution network. In contrast, if an embedded
generator is viewed as providing support for the power system then its transient
stability becomes of considerable importance. Both voltage and/or angle stability may
be significant depending on the circumstances. A particular problem in some countries
is nuisance tripping of ROCOF relays. These are set sensitively to detect islanding
but, in the event of a major system disturbance (e.g. loss of a large central generator),
may mal-operate and trip large amounts of embedded generation. The effect of this is,
of course, to depress the system frequency further. The restoration, after an outage, of
a section of the distribution network with significant embedded generation may also
require care. If the circuit was relying on the embedded generation to support its load
then, once the circuit is restored, the load will demand power before the generation
can be reconnected. This is, of course, a common problem faced by operators of
central generation/transmission networks but is unusual in distribution systems.
Synchronous generators will pole-slip during transient instability but when induction
generators over speed they draw very large reactive currents which then depress the
network voltage further and lead to voltage instability. The steady-state stability limit
of induction generators can also limit their application on very weak distribution
 20

networks because a very high source impedance, or low network short-circuit level,
can reduce their peak torque to such an extent that they cannot operate at the rated
output. [96Dug]
3.5.7. Network Operation
Embedded generation also has important consequences for the operation of
the distribution network in that circuits can now be energized from a number of
points. This has implications for policies of isolation and earthing for safety before
work is undertaken. There may also be more difficulty in obtaining outages for
planned maintenance and so reduced flexibility for work on a network with embedded
generation connected to it. [96Dug]

3.6. Economic Impact of Embedded Generation on the Distribution System

The issues of the technical impact of embedded generation on distribution
systems are generally well known and calculation and design techniques well
established. The only new issues emerging are those concerned with the overall
penetration magnitude of the embedded generation and the particular difficulties of
very large schemes being connected to very weak networks. In contrast, the economic
impact of embedded generation on distribution systems is only now being seriously
addressed. Embedded generation alters the power flows in the network and so will
alter network losses. If a small embedded generator is located close to a large load
then the network losses will be reduced as both real and reactive power can be
supplied to the load from the adjacent generator. Conversely, if a large embedded
generator is located far away from network loads then it is likely to increase losses on
the distribution system. A further complication arises due to differing values of
electrical energy as the network load increases. In general there is a correlation
between high load on the distribution network and the use of expensive generation
plant. Thus, any embedded generator which can operate in this period and reduce
distribution network losses will make a significant impact on the costs of operating
the network. At present, embedded generation generally takes no part in the voltage
control of distribution networks. Thus, in the UK, embedded generators will generally
choose to operate at unity power factor to minimize their electrical losses and avoid
any charges for reactive power consumption, irrespective of the needs of the
 21

distribution network. In Denmark some progress has been made in this regard with
embedded CHP schemes operating at three different power factors according to the
time of day. During periods of peak loads reactive power is exported to the network
while during low network load the generators operate at unity power factor.
Embedded generation can also, in principal, be used as a substitute for distribution
network capacity. However, most high-voltage distribution circuits are duplicated or
meshed, and embedded generation can reduce the requirement for these assets. At
present the concept, that embedded generation can substitute for distribution circuit
capacity, is not generally accepted by distribution utilities. [97Sch]

3.7. Impact of Embedded Generation on the Transmission system

In a similar manner to the distribution system, embedded generation will alter
the flows in the transmission system. Hence transmission losses will be altered,
generally reduced, while in a highly meshed transmission network it is easier to
demonstrate that reduced flows lead to a lower requirement for assets. In the UK, the
charges for use of the transmission network are currently evaluated based on a
measurement of peak demand at the Grid Supply Point. When embedded generation
plant can be shown to be operating during the periods of peak demand then it is
clearly reducing the charges for use of the transmission network. [97Sch]

3.8. Impact of Embedded Generation on Central Generation

The main impact of embedded generation on central generation is to reduce
the mean of the power output of the central generators but, often, to increase the
variance. In a large electrical power system, consumer demand can be estimated quite
accurately by the generator. Embedded generation will introduce additional
uncertainty in these estimates and so may require additional reserve plant. In
Denmark, considerable effort has been made to predict the output of wind farms by
forecasting wind speeds, and embedded CHP plants by forecasting heat demand. Both
forecasts are based on meteorological techniques. As embedded generation supplies
an increasing proportion of the customer load, particularly during times of low
demand, the provision of generation reserve and frequency control becomes an
important issue. Conventional central generating plant (i.e. steam or hydro-sets) is
able to provide these important ancillary services which are necessary for the power
 22

system to function. If embedded generation displaces such plant then these services
must be provided by others and the associated additional costs will then reduce the
value of the embedded generation output. [97Sch]